PDA

View Full Version : Flash XDR on HPX300



Big Brother
03-10-2009, 10:27 AM
1. I hope its compatible for AVC100/50 1080/50i recording
2. How does it gets mounted on AB Mount batts or recorders

Views invited !

David Saraceno
03-10-2009, 10:52 AM
Doesn't that just capture to Sony XDCAM Long GOP HD, at 100 Mbps or 50 Mbps 4:2:2 data-rates?

Barry_Green
03-10-2009, 10:57 AM
The Flash XDR doesn't record AVC-Intra, it records MPEG-2. It would certainly work with an HPX300 recording the SDI output, but the recording won't be AVC-Intra, it'll be MPEG-2.

Big Brother
03-10-2009, 11:12 AM
thx barry for your reply. But i still am amused how does it get mounted between camera's AB mount & my AB Battery ?

DavidNJ
03-10-2009, 11:14 AM
Their FAQ says at the current time it only supports 50 and 100Mb/s CBR MPEG-2 with 8-bit depth. So AVC Intra 100 would be significantly better.

.

BerkeleyBob
03-10-2009, 03:50 PM
Barry, I guess Panasonics answer to the XDR will be the HMR10 which also records to SDHC (due out later this year). Do you happen to know if it's 10 bit?

David Saraceno
03-10-2009, 04:46 PM
My understand is that AVCHD is 8 bit and the HMR 10 records AVCHD.

Barry_Green
03-10-2009, 05:02 PM
As David said, AVCHD is 8-bit 4:2:0. That HMR10 is an HD-SDI recorder that records to an SDHC card in the AVCHD format. So, depending on price, it would be in competition with the XDR or NanoFlash in that it's a commodity memory HD-SDI recorder, but... if the price is the same or near, the XDR/Nano will have a significant advantage as far as maximum potential quality goes. AVCHD can match XDCAM-EX recording, but the XDR and Nano can go to 100mbps 4:2:2, which should be a lot cleaner and better than AVCHD.

Of course, if the HMR10 is substantially less expensive, then it's a whole different ballgame.

Big Brother
03-11-2009, 03:23 AM
BUt i wonder if i record on MPEG2 Long GOP mode 100mbps of XDR then will it be better then DVCPRO HD or AVC50/100 of HPX300 ? Apart from saving a huge cost of P2 cards. Also will the XDR unit record with pause/record command from the camcorder body. Also is it necessary to have a P2 slot loaded with P2 card to shoot & record on XDR if so then may be one can use just One P2 card in continue mode (if available) to keep reusing the single P2 card & keep shooting on XDR !

Barry_Green
03-11-2009, 06:23 AM
No possible way will 100mbps long-GOP MPEG-2 be better than AVC-I @ 100mbps. But it would almost certainly be better than DVCPRO-HD, and should be better than AVC-I 50mbps.

You shouldn't need any P2 cards in order to record on the XDR. But when you talk about the "huge cost" of P2 cards, don't forget the "huge cost" of a $5,000 XDR; you can buy an awful lot of gigs of P2 for the cost of one XDR. At today's prices you could get about 128GB of P2... that's almost six hours of 720/24pN or twelve hours of 720/24pN @ AVCI-50, so you'd have to decide if that's enough storage for what you need, versus having a secondary recorder and battery and tracking four CF cards and whatnot.

Lonnie
03-11-2009, 06:45 AM
No possible way will 100mbps long-GOP MPEG-2 be better than AVC-I @ 100mbps. But it would almost certainly be better than DVCPRO-HD, and should be better than AVC-I 50mbps..

Barry, in theory I agree that 100Mbps Long Gop Mpeg 2 will not be as good as 100Mbps Avc Intra on codec alone. But, the final image will be a culmination of not just the codec, but also the lens and the front end electronics (chips, AD processors, etc.) of the camera as you know...

The Jist:
An EX1 with XDR is $11K-ish and will record from 1/2"chips to 422 Long Gop Mpeg2.
An HPX300 with ample P2 is $11k-ish and will record from 1/3"chips to 422 Avc-Intra.

So here's where I'm at: if spending $11K-ish, which will deliver a superior image?

... or will the imagery be just so close, it will just come down to workflow, personal tastes, and where our past investments will play into our next purchase workflows as well (ie. if you already have a P2 investment - this may be a no-brainer.)

Your thoughts...

Thanks,
Lonnie

Barry_Green
03-11-2009, 07:07 AM
Well, again, it depends on a million other factors too, such as the lens -- the EX1's lens has chromatic aberration, the HPX300 has CAC to eliminate it. And you can change the HPX300's lens.

The EX1 is going to have a small advantage in sensitivity and noise and dynamic range and DOF. The HPX300 is going to have an advantage in rejecting infrared; if you've got an IR-sensitive shot you're going to have to deal with filtration and perhaps green vignetting etc. on the EX1.

The HPX300 is going to have the "panasonic" gamma and color. The EX1 is going to have more tuneability, but not necessarily the "same" look. Do you want that look? That's not something that comes down to chips and recording formats, that's internal processing, etc.

So - the thread here wasn't about EX1 vs. HPX300, it was about using the Flash XDR on an HPX300. All the discussion here was about using the HPX300 as the camera head, and then the question was about recording on P2 or on the XDR, and in that case I'd take 10-bit intraframe AVC over long-GOP 8-bit MPEG-2 any day of the week.

The question you're asking is entirely different, and there's merit to the question of course. I think the better question would be EX3 vs. HPX300 as the camera head though, because then you can eliminate the EX1's fixed lens from the equation.

Dan Keaton
03-11-2009, 08:21 AM
The AVC-Intra has an advantage in that it is 10-bit while all of the MPEG-2 options on the Flash XDR are 8-Bit.

The Flash XDR does give you flexibility in that you can choose 50 or 100 Mbps Long-GOP (8-bit), or 100 Mbps or 160 Mbps I-Frame Only (MPEG 8-Bit). We also support 4:2:0 35 Mbps files in case someone needs a very long recording time.

In addition, a future (extra cost) option for the Flash XDR will provide full uncompresed 4:2:2 10-Bit recordings.

However, we do recommend the 100 Mbps Long-GOP or 160 Mbps I-Frame Only as the workflow is so much easier than uncompressed.

Another factor to consider is editing in general. I believe that it is safe to say that at this time, our MPEG files are easier to edit than AVC-Intra. Our files can just be dropped onto a HD timeline and edited without any transcoding.

Many equate all MPEG-2 Long- GOP with HDV. Our files, since they are full raster, are much easier to edit than the 1440 x 1080 HDV files.

With the 1440 x 1080 files, with a pixel aspect ratio of 1.3333, there is a lot of math being performed in the NLE in order to convert the image to full 1920 x 1080. With our full raster files, this computer intensive processing is not needed, nor used thus making the editing process smoother.

One other difference is that AVC-Intra 100 is only 100 Mbps in some modes, in others it is less than 100 Mbps.

Our 100 Mbps I-Frame Only mode and our 160 Mbps I-Frame Only mode use the full specified Constant Bit Rate modes in all formats and frame rates.

With the highest of high-end cameras, the Thomson Viper, with the Flash XDR, both our I-Frame Only and Long-GOP modes have been thoroughly evaluated and approved for quality by independent experts.

The advantage still goes to the AVC-Intra in that it is 10-bit.

The Flash XDR does have an advantage in that 2 Hours and 22 minutes of 100 Mbps CompactFlash card media costs only $360. At 50 Mbps, one gets 4 Hours and 44 minutes using the same 128 GB of CompactFlash card media.

I hope this helps and I hope this is perceived as being fair and factual.


Disclaimer: I am the Director of Sales and Marketing for Convergent Design, makers of the Flash XDR.

Lonnie
03-11-2009, 08:46 AM
Dan,
I understand that the XDR, off the shelf right now, is an 8 bit, Long Gop, Sony codec. But with a future firmware upgrade, the XDR is poised to do uncompressed 10 bit - that does indicate the XDR is hardware capable of 10 bit.

Thus, would the XDR be able to hardware handle a future 10 bit compressed codec if you were able to license one (like if you and Cineform shook hands for example)?

I realize this is all speculative, but it is a cool & unique feature that your little box has the potential of "future proofing" via a firmware codec upgrade...

Thanks,
Lonnie

Christian m
03-11-2009, 08:56 AM
Another factor to consider is editing in general. I believe that it is safe to say that at this time, our MPEG files are easier to edit than AVC-Intra. Our files can just be dropped onto a HD timeline and edited without any transcoding.

Avid are able to do that with Avc intra.

One issue with the EX1+recorder, the ergonomics for hand held use are...well...to be modest not good at all.

Barry_Green
03-11-2009, 10:10 AM
Lots of editors support AVC-Intra files on the timeline, including Avid and EDIUS and higher-up. It's not yet the case to be able to do that with FCP or Adobe.

Dan Keaton
03-11-2009, 10:27 AM
Dan,
I understand that the XDR, off the shelf right now, is an 8 bit, Long Gop, Sony codec. But with a future firmware upgrade, the XDR is poised to do uncompressed 10 bit - that does indicate the XDR is hardware capable of 10 bit.

Thus, would the XDR be able to hardware handle a future 10 bit compressed codec if you were able to license one (like if you and Cineform shook hands for example)?

I realize this is all speculative, but it is a cool & unique feature that your little box has the potential of "future proofing" via a firmware codec upgrade...

Thanks,
Lonnie

Dear Lonnie,

We have two options, a future 10-bit Full uncompressed option, or the current MPEG-2 codec.

We use a module, from Sony, which does the MPEG-2 compression. The Sony PDW-700 has one of these chips, we have two.

We are not able, without hardware changes, support other codecs.

But, full uncompressed does not require a codec, so we can do that as a future firmware upgrade (an extra cost option).

Dan Keaton
03-11-2009, 10:34 AM
Avid are able to do that with Avc intra.

One issue with the EX1+recorder, the ergonomics for hand held use are...well...to be modest not good at all.


Dear Christian,

Agreed, it is very difficult to mount anything, including the Flash XDR, on a Sony EX1.

The Flash XDR can be used with the EX1 in other configurations, such as in a backpack, or belt mounted. (This may work for some but not others, so this is not a universal solution). We have a wired remote control to make this feasible.

For some, the Flash XDR does not need to be mounted on the camera at all.

Dan Keaton
03-11-2009, 10:35 AM
Lots of editors support AVC-Intra files on the timeline, including Avid and EDIUS and higher-up. It's not yet the case to be able to do that with FCP or Adobe.


Dear Barry and Christian,

I stand corrected.

Lonnie
03-11-2009, 10:47 AM
Dear Christian,

Agreed, it is very difficult to mount anything, including the Flash XDR, on a Sony EX1.

The Flash XDR can be used with the EX1 in other configurations, such as in a backpack, or belt mounted. (This may work for some but not others, so this is not a universal solution). We have a wired remote control to make this feasible.

For some, the Flash XDR does not need to be mounted on the camera at all.




I believe it would be safe to say that the XDR is for more sophisticated users... and as such a lot of these users will be operating their cameras on rails. Mounting the XDR with a rigging that can attach/slide/undersling on the rails would be an approach I'd recommend...

Lonnie

Dan Keaton
03-12-2009, 07:47 AM
Dear Lonnie,

We do have a rail mount for the Flash XDR.

This mount attaches to any 15mm (60 mm spacing) rails. But typically, this is used with Zacuto rails. The rail mount also allows for a Anton Bauer Gold Mount battery.

We are currently designing a rail support system that allows 15mm rails out front and 15 mm rails in the back.

Jan_Crittenden
03-12-2009, 07:52 AM
Perhaps we could move the EX1/XDR conversation over to the EX1 forum. Frankly I don't see the need to record to an 8 Bit 4:2:2 Long Gop codec for the same price as a couple of P2 cards which would get me 10bit 4:2:2 I-Frame only.

Thanks,

Jan

Big Brother
03-12-2009, 07:56 AM
do post pix of setups where one can sandwich XDR between Camera & Anton Batt system ! would be a sight to see.

Cost wise (APPROX)

HPX300 = 8500
XDR = 5000
4x32GB cards CF = 250

All for Two hours & 22minutes of footage I have spent = 13750$


HPX300 = 8500
2x64GB cards P2 = 5200

All for Two hours of footage I have spent = 13700$
( i still lose 22minutes recording capab here)

But in long run say when i have to shoot about 6-8 hours long event then may be XDR way is cost effective (no comparison between XDR's Lgop & P2 AVCintra100 capabilities or 8vs10bit info) plain business maths to asses the investment, I can use my personal laptiop to copy onto to external HDDs from CF readers in XDR way or I can burn a BIG WIDE hole in pocket to buy the New P2 store (i dont get the model # right here i suppose) to read P2 & copy to external HDD. I am not too much into techincal aspects or quality diff between the XDR or P2 way but i am trying to make a way out to have a HD shoot at the best possible way. i am desperate to debut my HD Shoot FULL raster's !

Big Brother
03-12-2009, 08:00 AM
When going the XDR WAY then i can use http://www.bhphotovideo.com/c/product/573864-REG/Epson_B31B192002_P_7000_Multimedia_Photo_Viewer.ht ml or similar products & can swap internal HDD with some 500+ Gigs and have my CF cards copied as ordinary DATA (though not viewable on the data assistant) but stil i am mobile with field recording of data with a fraction of cost of what panny offer us as their P2store types. pls correct me if my dream sounds vague !

Bassman2003
03-12-2009, 09:46 AM
Cost wise (APPROX)

HPX300 = 8500
XDR = 5000
4x32GB cards CF = 250

All for Two hours & 22minutes of footage I have spent = 13750$


HPX300 = 8500
2x64GB cards P2 = 5200

All for Two hours of footage I have spent = 13700$
( i still lose 22minutes recording capab here)


To make this comparison equal, you need to mention that AVC-Intra 50 is 1440x1080 or 960x720 with 4:2:0 color but still 10 bit.

So the real comparison is 4:2:2 8bit vs 4:2:0 10bit

Call me a curmudgeon, and I know that 1/3" is the new 2/3", and it's all about the codec, but are we really talking about $13,500 for four hours of recording with a 1/3" chip camera?

Seems expensive to have the hole world in focus...

Jan_Crittenden
03-12-2009, 10:29 AM
do post pix of setups where one can sandwich XDR between Camera & Anton Batt system ! would be a sight to see.
Cost wise (APPROX)
HPX300 = 8500
XDR = 5000
4x32GB cards CF = 250

All for Two hours & 22minutes of footage I have spent = 13750$

HPX300 $8495
2 64GB cards, $2399(B & H) X 2 = $4798
Total $13,293
And this gets you 4:2:2 10 Bit recording that is I frame only instead of 8 bit 4:2:2 Long Gop.

So for a sacrifice of 22 minutes you can improve the overall quality of your image by going 10 bit and I-Frame only. Seems like a better deal when you are considering the overall objective is to get the best quality you can get in acquisition.



XDR way or I can burn a BIG WIDE hole in pocket to buy the New P2 store (i dont get the model # right here i suppose) to read P2 & copy to external HDD.

You don't need to buy the P2 Portable, you can do this off of the camera or using the P2 Card slot or the Duel Adapter.


I am not too much into techincal aspects or quality diff between the XDR or P2 way but i am trying to make a way out to have a HD shoot at the best possible way. i am desperate to debut my HD Shoot FULL raster's !

And if you don't fully understand the difference between the impact of 8 bit vs 10 bit, think which gives me more choices, 256 or 1024? Math would point you to 1024. Long GOP or Individual Frames. Individual Frames will give you more control over the compression scheme.

Add onto that other factors that are relevant in a file based workflow. Like Clip repair in the case of a loss of power. Metadata, when trying to find a singular clip in a 300 hour pool of clips, MetaData is your friend. Parity checking during record. Robustness of the media. You need to see my demo on this one, because my phone is a hot line, I hear what people do and what accidently happens.

Again this is tunring into one of those debates that basically assumes that the cards are the same, they are not. There is a lot more muscle in the P2 than what most will ever use, I hope. Built to take a beating and keep on working. Designed to close the recording every 2 seconds so that in the case of power failure, the most you will ever lose is 2 seconds. Yes they both record footage but there are a good number of failsafes that are built into P2.

Best,

Jan

Big Brother
03-12-2009, 11:14 AM
Jan, Speaking theoretically,

when i have to shoot & yes we do shoot events for about 6-8 hours a day typically. Then i do need at least 4 64GB cards which will let Two shoot & other two cards consume time siting on a P2store or some laptop adopter. so the calc has to be as under

- HPX300 = 8500
- 4 64GB cards = 9600
- P2 store = approx 5500

I dont want to have a laptop thereby freeing myself from keeping one more operator (hidden cost which panny doesnt think) so i am FORCED into a P2 store ! so a cool drool of $23,600 WOAH ! pls correct me ...

Where as if i go the xdr way then i save good $$$ on card cost & can use ordinary Stand alone HDD store to copy the CF data without being connected to a combersome laptop etc.

The whole idea is to get the best juices from HPX300 & also by keeping the cost contained, cox as of now whas looking to ordinary users like me that Panny is keeping the cost of Cameras low there by recovering the balance amount thru the cards !

I think for me an ideal way out for event shoot is

- HPX300
-XDR with say aboput 8 CF 32gb cards
-one Standalone HDD store

I do end up with 422 Lgop footage though 8bit,,,, whats the big deal for me 8 instead of 10 bit in my projects, we shoot live events, weds, conferences etc. May be i am right or may be you proove me otherwise. The debate goes on.

best way to keep the discussion contained is to have TWO HPX300 shooting same event with two setups
- One on Exorbitantly priced P2 cards (i feel the cost should be reduced not as an annual fair but more often) to get 422 10 bit thing
- Second one with XDR loaded with CF cards to get the lGop 422 8 bit color space

may be then we see the videos with lay mans eyes & decide.

period.

Big Brother
03-12-2009, 11:54 AM
MPEG2 offers very good quality video at relatively low data-rates. Flash XDR offers a
very wide range of MPEG2 data-rates from 18 Mbps up to 160 Mbps. Additionally, the
MPEG2 CODEC can be programmed to either Long-GOP or I-Frame only recording
modes, allowing the user to make trade-offs between recording efficiency and ease of
editing.

Compared to JPEG2000, MPEG2 is substantially faster to decode (improved playback
performance) and has been widely adopted for broadcast industry (ASI requires MPEG
compression). The MPEG2 hardware CODEC can provide much greater compression
efficiency (storage time) compared to strictly I-Frame based CODECs such as ProRes 422
or DNxHD CODEC. (However, the MPEG2 CODEC can be configured for Long-GOP or
I-Frame encoding). MPEG2 is also supported by a much wider range of NLE programs.

H.264 / AVC CODECS are not widely supported in NLE programs or by hardware
manufacturers at this point in time.

-- all thats laid out on XDR website !

Does that mean anything over AVC-I frame something like 100mbps vs MPEG2 - I frame only 160Mbps

Christian m
03-12-2009, 05:04 PM
H.264 / AVC CODECS are not widely supported in NLE programs or by hardware
manufacturers at this point in time.
Avid, full support with no transcoding
Edius, states support on their website?
FCP, via transcoding to prores

Adobe are the only I remember not supporting avc in any form yet.

Basicly the suites most used professionally support avc-i in some form, FCP need native support asap.

The P2 cost have been discussed before, yes there is a investment to take in P2 cards as with all proven broadcast gear, with the support to back it up. This is an important factor for many costumers. Sony and Panasonic, maybe Ikegami are the only suppliers that have the network of dealers to support their gear around the world. The college where I study recently bought 10 hpx171, 4 hpx2100 and 1 hpx2700 to replace the older dvx and dvcpro50 cams. As far as I can see one important factor are P2 as a workflow, no need for expensive decks and its rock solid and flexible. It's expensive for the smaller firms or freelancers as me, but it's worth it.

Mike Schell
03-12-2009, 08:00 PM
MPEG2 offers very good quality video at relatively low data-rates. Flash XDR offers a
very wide range of MPEG2 data-rates from 18 Mbps up to 160 Mbps. Additionally, the
MPEG2 CODEC can be programmed to either Long-GOP or I-Frame only recording
modes, allowing the user to make trade-offs between recording efficiency and ease of
editing.

Compared to JPEG2000, MPEG2 is substantially faster to decode (improved playback
performance) and has been widely adopted for broadcast industry (ASI requires MPEG
compression). The MPEG2 hardware CODEC can provide much greater compression
efficiency (storage time) compared to strictly I-Frame based CODECs such as ProRes 422
or DNxHD CODEC. (However, the MPEG2 CODEC can be configured for Long-GOP or
I-Frame encoding). MPEG2 is also supported by a much wider range of NLE programs.

H.264 / AVC CODECS are not widely supported in NLE programs or by hardware
manufacturers at this point in time.

Does that mean anything over AVC-I frame something like 100mbps vs MPEG2 - I frame only 160Mbps

We now have several Thomson Viper users shooting documentaries using the Flash XDR, with plans to print to film. I personally reviewed some of the footage shot at 1080psf23.98 at 160 Mbps; it was awesome. The detail was really amazing.

If you compared the two CODECs side by side when recording at 1080psf23.98/24 (which is the most popular format for film), you have the AVC-I 10-bit at 80 Mbps (effective) vs the MPEG2-I 8-bit at 160 Mbps. 10-bit is better than 8-bit, but 160 Mbps is better than 80 Mbps. However, we are talking about 10-bit compressed vs 8-bit compressed video. The 10-bit / 8-bit quality differences are certainly diminished after going through a compression cycle, as some of the resolution is lost.

So, it's really hard to compare just by looking at the numbers. The real test comes when you sit down and look at the images side by side. I do think there will be differences, but they will likely be very subtle. 10-bit should look better for banding issues, 160 Mbps will provide better detail. (Yes, I know AVC-I is more efficient than MPEG2, but the difference is very small at the higher bit-rates. And it all depends on the silicon implementation).

Just for completeness, we will very soon introduce the lower-cost ultra-portable nanoFlash, which fits in the palm of your hand, weights under 1 lb, draws about 7 watts of power and has both HD/SD-SDI and HDMI I/O. (Yes, it still uses the same affordable Compact Flash cards). It should be easy to mount to any camera.

Mike Schell
Convergent Design

Big Brother
03-12-2009, 08:51 PM
waoh I m impressed but hope Jan has more to reason 8 bit vs 10 bit vs MPEG2 vs AVC-I ?

Big Brother
03-12-2009, 08:55 PM
BTW is this the kind of plate to be used for mounting BATT & XDR back to back on Camcorder ---

http://antonbauer.com/hotswap.htm

Or something else...

DavidNJ
03-12-2009, 09:30 PM
Two questions:

1) The website says only 50 & 100Mb/s Long GOP are shipping. Is 160Mb/s also shipping?

2) Is the nanoFlash XDR available? Does it also do 160Mb/s as it said in the preliminary specs? The website has no availability info.

Big Brother
03-12-2009, 11:48 PM
i feel AVCintra 10bit at 100mbps wil have no major/drastic visual impact over MPEG2 Iframe 160mbps unless otherwise proved by Panasonic.

i am very keen to travel the HD path with HPX300 mated with XDR to its back ! i hope HPX doesnt need P2 card in it belly to allow external recording. Also XDR can be triggered into rec/pause mode from the camera switch on LENS. it wil be very exciting to tread the path without shelling out the $$$ expensive P2 cards :)

I was hearing from (gossipMONGERS) that coming NAB the p2 cards will be slashed by half their cost or near by ?

Big Brother
03-13-2009, 12:30 AM
Image Comparison Uncompressed Single Frame (19-June-08, 8MB)

http://www.convergent-design.com/CD_Products_nanoFlash.htm

Couldnt get whats being shown on the XDR site thru this image...

Also cannanoFLASH draw power from AB mounts power port or from camera power output port ...

Paul Bryer
03-13-2009, 03:02 AM
Two questions:

1) The website says only 50 & 100Mb/s Long GOP are shipping. Is 160Mb/s also shipping?

2) Is the nanoFlash XDR available? Does it also do 160Mb/s as it said in the preliminary specs? The website has no availability info.

Hi David,

In answer to your questions:

1) The Flash XDR can record at 50 or 100Mb/s long-GOP and at 100 or 160 Mb/s I-Frame only, and it can do that today. If you plan on using 160Mb/s I-Frame you will need to use the faster 300x speed CompactFlash cards. The standard 133x cards are not fast enough to deal with this bitrate.

2) The nanoFlash is not currently shipping. And yes, it will record at 160Mbps I-Frame only. Again, you will need to use the faster 300x cards.
A comprehensive FAQ on the nanoFlash is available here: http://www.dvinfo.net/conf/attachments/convergent-design-flash-xdr/11154d1236373753-nanoflash-faqs-nanoflash-faqs.pdf

Hope that helps :beer:

Paul Bryer
03-13-2009, 03:32 AM
i feel AVCintra 10bit at 100mbps wil have no major/drastic visual impact over MPEG2 Iframe 160mbps unless otherwise proved by Panasonic.

i am very keen to travel the HD path with HPX300 mated with XDR to its back ! i hope HPX doesnt need P2 card in it belly to allow external recording. Also XDR can be triggered into rec/pause mode from the camera switch on LENS. it wil be very exciting to tread the path without shelling out the $$$ expensive P2 cards :)

I was hearing from (gossipMONGERS) that coming NAB the p2 cards will be slashed by half their cost or near by ?

You can start/stop recording on the XDR using the record button on your camera, using incrementing timecode embedded in the HD-SDI stream.

How this works:
-You connect your camcorder's HD-SDI output to the input on the XDR.
-You set the XDR's record trigger option to Timecode, and set the timecode source option to Embedded.
-You set your camcorder to REC-RUN timecode.
-When the camera is switched on and in idle mode, the timecode is stationary.
-When you hit the record button on your camcorder, the timecode starts to increment.
-The XDR detects the incrementing timecode in the HD-SDI stream from the camera and starts recording (there is about a 2-3 second delay from hitting the button to the recording starting).
-When you hit the record button on the camcorder to stop recording, the timecode stops incrementing.
-The XDR detects that the timecode has stopped incrementing, and stops recording.

For the camcorder to use REC-RUN timecode in this fashion typically requires there to be media present in the camcorder for it to record to, so in your case you would need a P2 card present.
This can present issues in itself as with most camcorders when you run out of in-camcorder media, the REC-RUN timecode stops running as you'd need it to.
I believe some Panasonic camcorders have a loop-record feature, which is ideal if you're not bothered about making a back-up copy to P2, as it essentially gives you infinite in-camcorder media and the REC-RUN timecode will keep going as you need it to.

In addition to this there is another, simpler option: a remote trigger.
-Attach a remote trigger to the 4-pin hirose connector on the XDR, tie this to the HD-SDI cable and run it back up to your camcorder.
-Tape the trigger next to your camcorder's record button.
-The trigger now acts as your REC button for starting and stopping the XDR recording.

The beauty of this is the simplicity of it; you remove all the hassle of checking settings and media housekeeping inherent with timecode triggering.
It also gives you the freedom to use timecode modes other than REC-RUN.
Triggers are available with an inbuilt tally light, so you know when the XDR is recording and when it's not without having to look at the unit itself.

As for the AVC-I versus MPEG-2 debate, here's a CODEC comparison chart based on research done by Sony's Chief Technology Officer:

http://www.dvinfo.net/conf/attachments/convergent-design-flash-xdr/7941d1213657171-codec-quality-chart-codec-quality-chart.jpg

Food for thought.

Big Brother
03-13-2009, 04:52 AM
it will be really nice f you could post few pix of such trigger sin use on large Camcorders etc.

I was also wanting to know if this is of any use

http://antonbauer.com/hotswap.htm

Is this what one can use for Batt & XDR mounting on Camcorder.

Also are there any parale video comparision of AVCintra100 vs Iframe @ 160

Jan_Crittenden
03-13-2009, 04:58 AM
MPEG2 offers very good quality video at relatively low data-rates. Flash XDR offers a very wide range of MPEG2 data-rates from 18 Mbps up to 160 Mbps. Additionally, the MPEG2 CODEC can be programmed to either Long-GOP or I-Frame only recording modes, allowing the user to make trade-offs between recording efficiency and ease of editing.

The XDR does not do I-Frame at this time.



Compared to JPEG2000, MPEG2 is substantially faster to decode (improved playback performance) and has been widely adopted for broadcast industry (ASI requires MPEG compression). The MPEG2 hardware CODEC can provide much greater compression efficiency (storage time) compared to strictly I-Frame based CODECs such as ProRes 422 or DNxHD CODEC. (However, the MPEG2 CODEC can be configured for Long-GOP or I-Frame encoding). MPEG2 is also supported by a much wider range of NLE programs.


How fast it decodes is not as nearly as important as to how it encodes. Efficiency is not the primary goal in the acquisition of footage, or at least it shouldn't be. Getting the image encoded so that it can be used later is the primary goal. It is one of the reasons that in our upper end of the product line we use only I-Fram codecs. And actually AVC-Intra is pretty well supported at this time, including Apple, Avid, GV on the Edius platform, and a number of others. We ar only 3 weeks or so away form NAB where the landscape cna change again.


H.264 / AVC CODECS are not widely supported in NLE programs or by hardware manufacturers at this point in time.

Wrong.


-- all thats laid out on XDR website !

Perhaps they might want to make corrections on their website as they are outstnandingly incorrect. If you want more information about how advanced the AVC codecs are, don't ask they guy that is making the MPEG2 codecs. Read the two articles that are in the latest SMPTE Journal for an unbiased Scientific Review.



Does that mean anything over AVC-I frame something like 100mbps vs MPEG2 - I frame only 160Mbps


256 vs 1024, see you in color correction.

Best,

Jan

Jan_Crittenden
03-13-2009, 05:01 AM
I was also wanting to know if this is of any use

http://antonbauer.com/hotswap.htm

Is this what one can use for Batt & XDR mounting on Camcorder.

Also are there any parale video comparision of AVCintra100 vs Iframe @ 160

The camera draws 18W. The AB Device is more than a mounting bracket. It will be looking to swap power to the second slot.

There is not a parallel of the 8Bit 160 to a 10 Bit codec. 256 levels vs 1024.

Jan_Crittenden
03-13-2009, 05:05 AM
As for the AVC-I versus MPEG-2 debate, here's a CODEC comparison chart based on research done by Sony's Chief Technology Officer:


Do you think that maybe Juan had a motive to show AVC-Intra like this? Please look in the SMPTE Journal January/February 2009 issue for a factual look at how the AVC-Intra really holds up. This chart from Sony is silly wrong. They have said more silly things about AVC-Intra in the last two years it is pathetic, and I am still surprised at how many people believe it.

Best,

Jan

Jan_Crittenden
03-13-2009, 05:07 AM
i feel AVCintra 10bit at 100mbps wil have no major/drastic visual impact over MPEG2 Iframe 160mbps unless otherwise proved by Panasonic.


Bring it on!

Do you have any concept of what 256 vs 1024 looks like? Let me see if I can get Barry to host something for me.

I'll be back

AJ101
03-13-2009, 05:39 AM
Remember Sony also said their XDCAM EX was MPEG4??>>...Why??

Big Brother
03-13-2009, 05:55 AM
The camera draws 18W. The AB Device is more than a mounting bracket. It will be looking to swap power to the second slot.

There is not a parallel of the 8Bit 160 to a 10 Bit codec. 256 levels vs 1024.

with the AB dual slot i am thinking that one one side i slot in th XDR with other occupied with a AB Batt ... he AB Hot power tab will give power to the 4pin xlr of XDR or perhaps the camcorder's power outlet socket wil do the job ... Hope m righ t if i dream so ... :)

Big Brother
03-13-2009, 05:57 AM
The XDR does not do I-Frame at this time.



256 vs 1024, see you in color correction.

Best,

Jan

Is the 8bit vs 10bit debate solely restorcted to color corecction or chrom eying ... if so then why should i go the MPEG2 way on XDR with 160Mbps as I dont really see any such chroma or keying thing happening my way. f at any point it does then i still have p2 slots to be filled with the "PRINCELY EXPENSIVE" p2 cards to get the 10bit thing rolling.

Big Brother
03-13-2009, 05:59 AM
Bring it on!

Do you have any concept of what 256 vs 1024 looks like? Let me see if I can get Barry to host something for me.

I'll be back

Yes, actually i dont get the diff here. I hope you do help me out here.

Dan Keaton
03-13-2009, 06:23 AM
The Flash XDR does indeed give you the option of choosing 4:2:2 I-Frame Only at 100 Mbps or 160 Mbps (with faster CompactFlash Cards), or 4:2:2 Long-GOP at 50 or 100 Mbps.

We are about to release 4:2:0 35 Mbps support for those who need to record even longer.

From this forum yesterday, we learned that the AVC-Intra codec is now supported in some Non-Linear Editors. We will update our website as soon as practical.

Big Brother
03-13-2009, 06:25 AM
the debate for alternate to EXORBITANTLY/Princely/STEEPLY/price P2 cards will never end ...

Look at todays available CF/SDHC media available upto 100GB with enough bandwidth to support high data rates. I wish if panasonic makes apart from Great camera ... some nicely priced media. In the long run they will surely end up selling more number of cams like Sony XDCAMs... :( I dont love sony but i hate panasonic for P2 cards cost :(

why cant panny make some adopters for us to mate in such cards

Pretec unveils 666x Compact Flash Cards: Digital Photography Review

The ranting will go on...sadly till this years NAB hopefull panny wakes up to some proper price correction rather then merely doubling up the internal SD cards capacity. HUH

Christian m
03-13-2009, 07:29 AM
Look at todays available CF/SDHC media available upto 100GB with enough bandwidth to support high data rates.
And people should look into what's actually available, not what technically possible. 32 and 16gb are the largest flash memory when we are talking SD or CF, and SxS. We'll see 128gb later this year I think and as long as SD evolves the P2 will follow.


And the Mpeg2 vs avc-i, neither panasonic nor sony is the right to ask. For the technical aspect you need a more independent source. Asking Sony if mpeg2 is the best is about as smart as asking Bill gates if he got an ipod....or iphone.

Big Brother
03-13-2009, 07:41 AM
its been talked about what available or in nar future. can panny do more thn just doubling the GB space of their COSTLY/EXPENSIVE P2 cards. i look forward to soe good price reduction, seems like panny should devote some more time to evolve tech to make the p2 cards more affordable rather then making 128 so on so forth & making them only for Prim TV operators...

A poor mans soul just cried ! (ok not so poor to not to dream in HD)

Jan_Crittenden
03-13-2009, 08:36 AM
with the AB dual slot i am thinking that one one side i slot in th XDR with other occupied with a AB Batt ... he AB Hot power tab will give power to the 4pin xlr of XDR or perhaps the camcorder's power outlet socket wil do the job ... Hope m righ t if i dream so ... :)

That isn't how that device works. The secd mount does not draw power it supplies power, since the XDR needs to draw power, it is not going to waork.

Best,

Jan

Jan_Crittenden
03-13-2009, 08:40 AM
why cant panny make some adopters for us to mate in such cards



But you assume that the cheap memory does everything that the P2 does. It doesn't.

Now as promised here is the difference, in a black and white graph the difference between 8 bit and 10 bit.

http://www.dvxuser.com/barry/8-vs-10-bit.jpg


Best,

Jan

Big Brother
03-13-2009, 08:41 AM
That isn't how that device works. The secd mount does not draw power it supplies power, since the XDR needs to draw power, it is not going to waork.

Best,

Jan

well you didnt get what i meant.

I meant, on the AB dual mount sledge !

One side I put AB bat with Power Tap facility
One other side mounts my xdr which gets power thru power tap cable from AB batts power tap port. Hope i am easy this time.

---

Another reson for XDR (at least as long as FCP dont supports native P2 footage) it records straight into MOV which gets edited in FCP in no time.

Jan_Crittenden
03-13-2009, 08:42 AM
Yes, actually i dont get the diff here. I hope you do help me out here.

As soon as Barry can fix my link, it should be relatively apparent.

Best,

Jan

Big Brother
03-13-2009, 08:48 AM
thx for the link ,,, it indeed demos the diff in B & W image. But my mind just dont gos only for color details

I see 160MBps of I frame 422 8 bit compression giving me over all more image info at lesser compression unlike AVCintra 100mbps yes at 10 bit 422 (correct me if wrong).

i was discussing with a Local panny Engg & he was of the view that visually it wont be of too much diff between both the codecs !

i hope to see some Same image comparison by pany or -D of their products in general interest to proove either POV !

Jan_Crittenden
03-13-2009, 08:49 AM
well you didnt get what i meant.

I meant, on the AB dual mount sledge !

One side I put AB bat with Power Tap facility
One other side mounts my xdr which gets power thru power tap cable from AB batts power tap port. Hope i am easy this time.

---

Another reson for XDR (at least as long as FCP dont supports native P2 footage) it records straight into MOV which gets edited in FCP in no time.

I understood exactly what you meant. We do sell these devices. That two gold mount device is meant to supply power to the camera. It will be looking at the device that is there and will probably fail to function properly becaus ethe cicuit is interupted by one side of the device not standing by and being eady to deliver power. A momentary loos of one battery is one thing, but not sure it will work over the long haul.

FCP inputs AVCIntra in no more time than it takes to input DVCPRO HD and have it re-wrap.

Best,

Jan

Big Brother
03-13-2009, 08:54 AM
But you assume that the cheap memory does everything that the P2 does. It doesn't.


Best,

Jan

What else or where else is it different.

Instant Edit
Safe Data storage random access

in any language or words it(read EXPENSIVE P2 cards) seems to do the same job as a relatively cheap memory card (300x UDMA type) I see it hard to make my self understand what panasonic it self feel as a tough terrain to explain to competition. i did a small maths for cost evaluation a few pages back on this thread.

I see the only dif which panny is beating around on HPX300 is a AVC-I @ 100with 422 codec. I am not very satisfied with Panny explanation on --- either P2 cards exorbitant cost esp when we see the media cost going down every passing day & where we see so many CF or SDHC cards coming with large space & fast speeds enough to sustain a bit rate of 160mbps --- or with geting a bit too much detail with the extra 2but color space !

:dankk2:

Jan_Crittenden
03-13-2009, 08:55 AM
thx for the link ,,, it indeed demos the diff in B & W image. But my mind just dont gos only for color details

I see 160MBps of I frame 422 8 bit compression giving me over all more image info at lesser compression unlike AVCintra 100mbps yes at 10 bit 422 (correct me if wrong).

i was discussing with a Local panny Engg & he was of the view that visually it wont be of too much diff between both the codecs !

i hope to see some Same image comparison by pany or -D of their products in general interest to proove either POV !

Our DVCPRO HD is 8 bit. This is represented in the 8 bit line. The AVC-Intra is 10 Bit it is represented by the 10 bit line. What is it don't you get? This is the number of levels you have in the image, be it black and white or color, in a color image in 256 you start to get ti to the right color of red and it makes a big jump. In 1024, I have a much more graceful sliding scale. 4:2:2 in 10 bit means that I am drawing from the image that is 1024 levels as compared to 256. Read the SMPTE Journal articles, written by two different learned individuals, without a marketing position to make. The overall summation: It clearly is the best codec of the day.

Thanks,

Big Brother
03-13-2009, 08:58 AM
I understood exactly what you meant. We do sell these devices. That two gold mount device is meant to supply power to the camera. It will be looking at the device that is there and will probably fail to function properly becaus ethe cicuit is interupted by one side of the device not standing by and being eady to deliver power. A momentary loos of one battery is one thing, but not sure it will work over the long haul.

FCP inputs AVCIntra in no more time than it takes to input DVCPRO HD and have it re-wrap.

Best,

Jan

well both of us stand confusd on the point to use dual batt AB mount ... lets leave it to further experimentation. Yes i feel that if AB mount dont gets power from (which it wont) XDR side it will treat it as a dead batt & continue to draw power from AB batt side till end ... but still some one may try it for us.

But when we talk on FCP works natively with AVCintra ... i donnt agree cox i had to transcode it thru ProRes....also i had to waste enough time to transcode the P2 file to MOV...all this gets saved by use of an XDR or similar unit this the best way to go to edit straight from edit.

P2 shoot but one has to transcode the whole footage into MOV & that takes a considerable amount of time for edit; whereby defeating the very purpose of Direct To Edit.

I might be happy with 8bit DVCPRO HD geting rcorded n cheap media thru nanoFLASH or XDR in 422 8 bit rather then too much of expense on P2 cards for the extra 2bit color of space.

Agreed to the AVC-I codec the best codec of the day (that makes panny folks happy) but P2 cards are the most exorbitant priced media (sadly when compared to available options, ) panny wont like to hear this !

Barry_Green
03-13-2009, 09:11 AM
I see 160MBps of I frame 422 8 bit compression giving me over all more image info at lesser compression unlike AVCintra 100mbps yes at 10 bit 422 (correct me if wrong).
It's pretty simple, really. Both companies have claimed that their codecs are nigh unto uncompressed, visually. But one of them has 4x as many shades of gray or shades of color as the other. If you're looking at "virtually uncompressed" 8-bit vs. "virtually uncompressed" 10-bit, then 10 bit wins. Easy.

MPEG-2 at 100mbps is going to look great. But it can't ever look the same as AVC-I at 100mbps, because it's 8-bit instead of 10-bit.

I'm surprised there's even that much debate about it. Unquestionably, without a doubt, the XDR is capable of delivering much better quality than the onboard codec of an XDCAM-EX or XDCAM-HD or XDCAM-422 camcorder. But when you put it up against AVC-I, you're talking about the codec that Discovery certified Gold, that the BBC is using to archive all their content in their Digital Media Initiative, and that the Hollywood Post Alliance has just given a special award for technical achievement. None of them did that for MPEG-2.

Barry_Green
03-13-2009, 09:19 AM
But when we talk on FCP works natively with AVCintra ... i donnt agree cox i had to transcode it thru ProRes....
That's the way FCP is set up to work. It doesn't HAVE to work that way, and you can get around it.


also i had to waste enough time to transcode the P2 file to MOV...
Only if you refuse to use one of the many workarounds. You never *have* to transcode or convert the files to MOVs at all. This is the single most irritating thing about FCP, and it's easily overcomeable.


all this gets saved by use of an XDR or similar unit this the best way to go to edit straight from edit.
You can go straight to the edit with P2 as well. Just use Raylight, or Calibrated's MXF importer, or HD Log Pro, or MXF4QT, or ... heck, there are probably others. You can edit directly off the card, even with AVC-Intra. You can edit the MXF files directly, even with AVC-Intra.

You don't have to ever do the silly .MOV re-wrap or a ProRes transcode unless you simply just won't use one of the above tools.



P2 shoot but one has to transcode the whole footage into MOV & that takes a considerable amount of time for edit; whereby defeating the very purpose of Direct To Edit.
No, it's FCP that defeats the very purpose of direct to edit. You can do direct to edit on *any* Windows NLE. Only FCP requires the transcode or rewrap, and as stated before you can avoid the transcode and rewrap entirely, which brings FCP back into the real world of tapeless editing.

Big Brother
03-13-2009, 10:11 AM
thx Barry for the few insight which panny folks not able to. BUT can you also give me an idea will a AVI-1@100 be better or near XDR MPEG2 -I @ 160MBPS. i feel that @ 160MBPS i get more data so better Pix quality.

Barry_Green
03-13-2009, 11:03 AM
But it's still going to be 8-bit. How can 8-bit ever be better than 10? It can't, provided that there's enough bandwidth overall.

You can't just compare megabits to megabits, because you're talking about an entirely different compression system. AVC is far more efficient than MPEG-2. According to Sony themselves, 9 megabits of AVC long-GOP is equivalent to 25 megabits of MPEG-2 long-GOP.

It's true that efficiencies are greatest at the smaller data rates, and once you allocate enough bandwidth you're getting close to uncompressed quality; I would expect that 160mbps MPEG-2 would be on par with 100mbps AVC-I in that both should deliver close to uncompressed, and maybe even hit the magic level of "virtually" uncompressed. But even then, you're going to have 10 bits of data vs. 8. You're going to have 1,024 shades of each color, instead of 256. You're going to have smoother gradients instead of banding.

Big Brother
03-13-2009, 11:14 AM
Cool Barry, Thx again. But isint 160mbps on xdr an I-frame compression. one similar to panny Iframe compression @ 100 mbps. As thats what XDR site states.

Jan_Crittenden
03-13-2009, 12:18 PM
What else or where else is it different.

Instant Edit
Safe Data storage random access



How about loss of footage if you lose power?

How about a metadata wrangling?

How about Parity check on the recording of the footage?

What about an ever increasing speed for read or write? Just off loaded 2 64 GB cards in 34 minutes.


in any language or words it(read EXPENSIVE P2 cards)


Only expensive if you use them once and put on a shelf. If you used the 64GB card 1000 times, you are loking at a cost of $2.40, 10,000 times, .24 cents. It is a relative thing, certainly more expensive as the front end of the purchase but the cost goes down but the reliability stays the same.


Seems to do the same job as a relatively cheap memory card (300x UDMA type) I see it hard to make my self understand

Only cheap if you don't lose the power in the middle of recording. P2 closes the recording every 2 seconds so that the most you would ever lose is 2 seconds.

Only cheap if you aren't concerned with a data rich archive that allows for searching for clips.

Only cheap if your investment will handle what might come down the stream. Keep in mind the 4GB cards we started to deliver 5 years ago handle the DVCPRO and DVCPRO50 which were the first P2 cameras, but they can als handl DVCPRO HD and now AVC-Intra. And if we were to develop product warpped around the 4:4:4 version of the AVC-H.264 family that same 4GB card could handle it. So you make your invesment once and it continues to work.


what panasonic it self feel as a tough terrain to explain to competition. i did a small maths for cost evaluation a few pages back on this thread.

Been meaning to go back to that math and show you that your estimates of needing 4 cards is not necessary, you could do what you are doing very easiliy with 3 cards. Seen it done.



I see the only dif which panny is beating around on HPX300 is a AVC-I @ 100with 422 codec. I am not very satisfied with Panny explanation on --- either P2 cards exorbitant cost esp when we see the media cost going down every passing day & where we see so many CF or SDHC cards coming with large space & fast speeds enough to sustain a bit rate of 160mbps

The difference between the Memory is different but because it is not commercially available, you cannot buy it and thus you do not have a reference. The Memory in the CF Cards and SDHC cards are a step down from the memory that is in the P2 cards. But it is hard to prove without ruining a card. And frankly that is not worth it to me. I look at all of the failsafes built into the card and feel comfortable in recommending them and have my customer's careers ride on them.


--- or with geting a bit too much detail with the extra 2but color space !

This detail is only a portion of the entire story but I have spent the time trying to educate you in the difference between 8 bit and 10 bit and the difference it will make in color correction. It will also make a diffi=erence in the tonal gradations of a singular color, or flesh tones. It does make a difference in color, just like the gray scale, imagine it in color, it is the same number of steps.

Best,

Jan

DavidNJ
03-13-2009, 01:40 PM
Hi Jan, Mike, and Dan (did I miss someone?),

Let me first say that both Panasonic's AVC Intra and the Flash/nanoFlash XDR show both vendors are pushing the envelope. Let me add the GH-1 to Panasonic's feather collection, although that is a different division.

There have been a few statements here that I would love to get clarified.

1) What rate does AVC Intra 100 record at? Is it VBR Max 100? If so, what is that target rate? What is the actual rate?

2) With the introduction of the impressive HPX300 and HPG20 AVC Intra is available at a lower price point. That would allow capturing AVC Intra from any HD-SDI (or HDMI with a conversion box like the ones sold by Convergent). Has it been used with the EX3?

3) I thought the Thompson output uncompressed 4:4:4 RAW format. If so, how does it interface to the Flash XDR?

4) Has the XDR been used on a camera that outputs AVC Intra 100? If so, have the formats been compared?

5) Convergent records both Long GOP and I-frame outputs. Why did Panasonic chose not to offer 10-bit 4:2:2 AVC output? That would have fit with the 100Mb/s P2 speeds and offered more information picture detail.

6) Panasonics AVC is a VBR, effectively using the bandwidth while producing a consistant image quality. Why is the Flash XDR CBR? Does it cause an inconsistent image quality? Was it a codec limitation, as was 8-bit recording?

Thanks for offering such great products (that we are having a discussion of competing 100Mb/s protocols!!!)

David

Jan_Crittenden
03-13-2009, 02:03 PM
Hi Jan, Mike, and Dan (did I miss someone?),
1) What rate does AVC Intra 100 record at? Is it VBR Max 100? If so, what is that target rate? What is the actual rate?

If you are are 24P, just like the DVCPRO HD, there are fewer bits used but not quite the savings as DVCPRO. There is no VBR in the top end, only in the bottom end. AVC-Intra is CBR, as each frame stands alone.


2) With the introduction of the impressive HPX300 and HPG20 AVC Intra is available at a lower price point. That would allow capturing AVC Intra from any HD-SDI (or HDMI with a conversion box like the ones sold by Convergent). Has it been used with the EX3?

Yes, the EX3 can become a Mastering Quality camera as well.


4) Has the XDR been used on a camera that outputs AVC Intra 100? If so, have the formats been compared?

AVC-Intra is a Codec, what would ome out of the HD_SDI is uncompressed HD.


5) Convergent records both Long GOP and I-frame outputs. Why did Panasonic chose not to offer 10-bit 4:2:2 AVC output? That would have fit with the 100Mb/s P2 speeds and offered more information picture detail.


The Output of the HPX300 is 10 bit uncompressed which is by its nature 4:2:2



6) Panasonics AVC is a VBR, effectively using the bandwidth while producing a consistant image quality. Why is the Flash XDR CBR? Does it cause an inconsistent image quality? Was it a codec limitation, as was 8-bit recording?

Depends on which end of the AVC spectrum you are on. The AVCHD while in the H.264 family, it is not the same as AVC-Intra, it is able to do VBR. AVC-Intra is CBR. At the low end of the AVC familiy is 4:2:0, 8 Bit, long GOP. At the top end it becomes I-Frame 4:2:2 100 Mbs.



Thanks for offering such great products (that we are having a discussion of competing 100Mb/s protocols!!!)

It is more like competing Codecs. Not protocols, just to be specific. protocols would be like the protocol that is embedded in the HD-SDI stream to trigger the HPG20 into record from the camera's record trigger. Codecs are what will determine quality that ends up on the timeline for editing.


Hope this helps,

Jan

Barry_Green
03-13-2009, 02:16 PM
But isint 160mbps on xdr an I-frame compression. one similar to panny Iframe compression @ 100 mbps.
Yes, they're both I-frame. But the XDR is MPEG-2, and the Pansonic is AVC. AVC is a lot more efficient...

To break out the eternal comparison to cars, which one can drive farther, the one with a 10-gallon tank or the one with a 16-gallon tank? And the answer is, you don't know until you know how many miles per gallon the car gets! AVC gets way more miles per gallon than MPEG-2 does.

MPEG-2 at 160mbps is probably about equivalent to AVC at 100mbps, in terms of resolved detail and resistance to artifacting. But AVC-Intra is 10 bit instead of 8 bit. So it's still recording way more information, above and beyond the nominally-stated 1920x1080 and 4:2:2.

Mike Schell
03-13-2009, 05:39 PM
But it's still going to be 8-bit. How can 8-bit ever be better than 10? It can't, provided that there's enough bandwidth overall.

You can't just compare megabits to megabits, because you're talking about an entirely different compression system. AVC is far more efficient than MPEG-2. According to Sony themselves, 9 megabits of AVC long-GOP is equivalent to 25 megabits of MPEG-2 long-GOP.

It's true that efficiencies are greatest at the smaller data rates, and once you allocate enough bandwidth you're getting close to uncompressed quality; I would expect that 160mbps MPEG-2 would be on par with 100mbps AVC-I in that both should deliver close to uncompressed, and maybe even hit the magic level of "virtually" uncompressed. But even then, you're going to have 10 bits of data vs. 8. You're going to have 1,024 shades of each color, instead of 256. You're going to have smoother gradients instead of banding.

Hi Barry-
It's a little more complicated than just a straight 8-bit to 10-bit comparison. Remember these are compressed images, so you undoubtedly loose some of the resolution through the compression process. Most compression routines reduce the resolution of the values and introduce round-off errors. So, I bet the 8-bit / 10-bit difference becomes much less after compression through the respective CODECs.

You also have to consider the final compression ratio. As you noted, AVC-I is more efficient than MPEG2, but the efficiency difference is greatly reduced as the bit-rate increases. I will do some more research on this difference at the 100+ Mbps level, but I bet it's not very large.

Let's do some compression ratio comparisons at 1080psf23.98, the most popular format for output to film (where quality is most critical). AVC-I bit rate = 80 Mbps (effective) at 4:2:2 10-bit for an effective compression of 12.43 to 1. MPEG2-I = 160 Mbps at 4:2:2 8-bit for an effective compression of 4.97 to 1. In other words, the MPEG2-I has 2.5 times less compression. (Both CODECs are 4:2:2, full-raster, I-Frame only).

So, I bet the quality differences are very small and may be very difficult to discern. Naturally, we do need to test and confirm.

Having spent many years in video CODEC design, I know that the actual silicon implemetation of any compression algorithm can made a huge difference in the final quality, especially with regard to the encode quality (much less for the decode quality). So, it is a very complex comparison, and the numbers can be very misleading.

So, let's try to arrange a test in the near future.

Best-
Mike Schell
Convergent Design

Jan_Crittenden
03-13-2009, 06:02 PM
So, it is a very complex comparison, and the numbers can be very misleading.

So, let's try to arrange a test in the near future.



Hi Mike,

If you want Barry there, no problem, but frankly I think you just need to figure out the 8bit-10bit reality and if you really want that to be part of the "public record." Look back at the pic I posted earlier in the thread, trust me we will go to that level in the test, color and all. And we will do the full work flow with the power outage and see how much we lose and what the th metadata capabilities are. In addition we need to look at during post color correction to the ideal color can we say a standard like Coca-cola Red and Panasonic blue. Do you have any favorites?

Yes, 160 sounds better than 100, but the total, needs consideration. I can arrange a test after NAB. Of course that test is here in NJ with our $16,000 Digital Scope and our BT-LH2550 Full Raster LCD.

See you at NAB


All the best,

Jan

Mike Schell
03-13-2009, 07:31 PM
Hi Jan, Mike, and Dan (did I miss someone?),

Let me first say that both Panasonic's AVC Intra and the Flash/nanoFlash XDR show both vendors are pushing the envelope. Let me add the GH-1 to Panasonic's feather collection, although that is a different division.

There have been a few statements here that I would love to get clarified.

1) What rate does AVC Intra 100 record at? Is it VBR Max 100? If so, what is that target rate? What is the actual rate?

2) With the introduction of the impressive HPX300 and HPG20 AVC Intra is available at a lower price point. That would allow capturing AVC Intra from any HD-SDI (or HDMI with a conversion box like the ones sold by Convergent). Has it been used with the EX3?

3) I thought the Thompson output uncompressed 4:4:4 RAW format. If so, how does it interface to the Flash XDR?

4) Has the XDR been used on a camera that outputs AVC Intra 100? If so, have the formats been compared?

5) Convergent records both Long GOP and I-frame outputs. Why did Panasonic chose not to offer 10-bit 4:2:2 AVC output? That would have fit with the 100Mb/s P2 speeds and offered more information picture detail.

6) Panasonics AVC is a VBR, effectively using the bandwidth while producing a consistant image quality. Why is the Flash XDR CBR? Does it cause an inconsistent image quality? Was it a codec limitation, as was 8-bit recording?

Thanks for offering such great products (that we are having a discussion of competing 100Mb/s protocols!!!)

David

Hi David-
I'll answer the questions directly at our products:

3) The Thomson Viper can output either 4:4:4 dual-link or 4:2:2 single link HS-SDI, so we can connect directly to the Flash XDR / nanoFlash.

4) I don't think anyone has made a direct comparison. I will speak to our Viper users next week about doing a comparison shot. We can use an HD-SDI DA and simultaneously send the same signal to an AVC-I recorder and our Flash XDR. We can compare some complex sky shots, facial shots and some gradient charts. I bet the differences will be very small and we will find some areas of the image where AVC-I excels and some where MPEG2-I excels.

The Thomson Viper is actually an ideal camera for this compariosn since it has a wide dynamic range and and low-noise sensors. The image quality is really amazing.

6) The choice of CBR was made to match with software CODEC specifications. However, I am fairly certain we can switch to VBR with little effect. The CBR choice is not based on a CODEC limitation or because of 8-bit encoding. I'll ask our engineers to test VBR next week.

Mike Schell
Convergent Design

DavidNJ
03-13-2009, 07:51 PM
Hey everyone...everyone please calm down...

If the Flash XDR is better...so what? The HPX is the same price as its competitors and my all accounts a bit better. It can use the XDR the same way.

If AVC Intra is better...so what? The Flash is smaller, uses less expensive media, can interface to an HDMI device (in nanoFlash format).

And other than the XDR, its native recording is significantly better than the XDCAM EX used by both Sony and JVC. 8-bit editing isn't a problem. I know if I use a nanoFlash XDR from a GH-1, I will convert to Cineform 10-bit 4:2:2 for editing. It was only 5 years ago our choices were DV, HDV, or maybe DVCProHD. XDCAM EX seems like HDV to me.

There is no need for a battle...the playing field is different.

And maybe complementary. For $8500 B&H lists the camera with a very nice Canon lens. An Flash XDR is $4800. And a HPG-20 with monitor is $4600; that device is probably more comparable.

Flash questions: What speed is 720p24 recorded at? Can we get 160Mb/s with no pulldown?

P2 questions: With regards to power loss...that facility is in the camera or on the P2 card? How does the battery powered Flash XDR deal with power loss?

Thanks,

David

Mike Schell
03-13-2009, 07:52 PM
Hi Mike,

If you want Barry there, no problem, but frankly I think you just need to figure out the 8bit-10bit reality and if you really want that to be part of the "public record." Look back at the pic I posted earlier in the thread, trust me we will go to that level in the test, color and all. And we will do the full work flow with the power outage and see how much we lose and what the th metadata capabilities are. In addition we need to look at during post color correction to the ideal color can we say a standard like Coca-cola Red and Panasonic blue. Do you have any favorites?

Yes, 160 sounds better than 100, but the total, needs consideration. I can arrange a test after NAB. Of course that test is here in NJ with our $16,000 Digital Scope and our BT-LH2550 Full Raster LCD.

See you at NAB


All the best,

Jan

Hi Jan-
Thanks for the offer, I think we can do some tests in the meantime. I think the differences are going to be very small. Our Thomson Viper users have examined the quality of the images to the nth degree. I am quite comfortable with their findings.

As I have stated before, I think AVC-I is a great CODEC, but it's not the only possible or viable solution on the market (you do sell lots of DVCProHD based cameras, for example). And, as you mentioned, many other factors come into play including metadata support, ease of use, cost, portability, connectivity, etc. Like image comparisons, it's not a simple black and white (no pun intended) decision.

Best Regards-
Mike Schell

PS We will use our $24K Tektronix HD-SDI waveform monitor and look at the images on a Sony TriMaster monitor.

Big Brother
03-13-2009, 09:01 PM
Hi Mike,

If you want Barry there, no problem, but frankly I think you just need to figure out the 8bit-10bit reality and if you really want that to be part of the "public record." Look back at the pic I posted earlier in the thread, trust me we will go to that level in the test, color and all. And we will do the full work flow with the power outage and see how much we lose and what the th metadata capabilities are. In addition we need to look at during post color correction to the ideal color can we say a standard like Coca-cola Red and Panasonic blue. Do you have any favorites?

Yes, 160 sounds better than 100, but the total, needs consideration. I can arrange a test after NAB. Of course that test is here in NJ with our $16,000 Digital Scope and our BT-LH2550 Full Raster LCD.

See you at NAB


All the best,

Jan



Hi, Any ida what goes in the inside of P2 card. Can Panasonic shws us an open P2 card, it shall help (at least) me to beileve on why P2 cards are Expensive.

I spoke with C-D folks & was told that the power failiure result in last clip being stored geting lost but they are upto some development thru firmmware where we can recover the last lost clip if in case of any power failure. But a power failure can happen in only Two cases if we shoot on bats. Firstly if a foolish cameramen fails to notice Batt end or near end warning. secondly if bat goes dead in the middl of shot for some tech fault.

In any which case I am really impressed with HPX300 in anyy senses but still not happy with P2 prices,,, hopefully Jan has some images of an open P2 card which shouldnt be too hard for her to Get being in PANNY.Unless Panny doesnt want to reveal the simple layout plan inside of p2thing.

I am sure Jan will oblidge us all with the tests mentioned above to proove the discussion with facts. B it codec 100vs160mbps or P2 inners so much fan faired so called complex design which alow them to double the capacity each NAB but not drop the prices. :)


Hi Jan-

Best Regards-
Mike Schell

PS We will use our $24K Tektronix HD-SDI waveform monitor and look at the images on a Sony TriMaster monitor.

:) I liked the last line where you mentioned the $$$ a cost comparison to Pannys DigiScope.

Earlier when Panny had 1280x720 then also they were the BEST (in their own thoughts) now that they have been able to finlly offer full raster they still are the best Wonder what sony is feeling & saying when they were full raster much earlier & still leads the pack with the benchmark Product to which all products aim HDCAM SR (800+mbps to my knowledge) well no comparision between Panny & sony tech. I feel i am attracted to panny as of now cox of the low entry cost of their HD full rasters.

i am no big film maker am a simple events guy wored way up to HD ri8 thru the good Old VHS, Umatic Low/High Band, Betacam SP, DVCAM & now aspiring for HD. But the cost have been really the biggest hurdle to make a decision, I must thank panny for stealing the show with low cost HPX300. Now my aim here was solely to compare the much lower vost media & its recorder C-D to the MUCH expensive P2 card.we need four p2 cards per camera as we dont want to have a shoot held up cox the card simply is wasting its our time while it gets copied onto the P2 gear. We shoot average 5hours per camera effectively 10Gig P2 cards !

I use generally Three Cam setup any given event. My maths there has been like this for each HPX300 I will need aleast 4 P2 cards which jacks up the cost considerably & when i theritically speaking go the nanoFLASH way then this is what i calculate. (Not counting the cost of Hard Disks as it gonna be cooman to p2 or nanoFLASH way)


3 Panasonic HPX300 = 8500x3 = 25,500
P2 Gear = 5000
12 P2 cards = 30,000 (2 hour average on each camera)

Cost of ownership = 60,500 WOAH all to get the 10bit recorded & will not want to have a laptop instead of P2 ear cox then i will have to have atrained person baby sit the transfer process. p2 gear can be opeated with ease by any of my assistants without any extra cost.

Now the nanoFLASH way.

3 Panasonic HPX300 = 8500x3 = 25,500
3 indivudual data copier with 500GB hDD = 1000x3 = 3000
12 cheap 300x CF cards = 1200
3 nano Flash = 3400x3 = 10,200

so the cost of owner ship = 40,000 On the higher side.

And if i decide to add 2 P2 cards for record operation thru camera on the wonderfull nanoFLASH then i can burn some more hard earned money like this

6 p2 32 GB cards = 8400

now these cards can be re used after formatting inside the corder or on laptop with a trained guy or if programmable then tp be reused after they reach end there by they sit USELESSLY in my HPX300 just to enable me to trigger record on the wonderful nanoFLASH from the lens trigger button. WOAH still I am way under 61,000$ thats a hell lot load of moolah saved for my next pruchase. This calc wil got up & up ...

BTW why cant C-D design their prods around AVC-I format 10 bit codec...Ok may be some Licenses issue & more over i feel panny wont let their P2 Profit go away like this. If panny says they do have a real tech inside then am sure Jan will help us all with the images of open P2 cards to take her facts further.

So just to get the 10bit recording i will have to burn extra 20,000 now thats no brainer for a simple event shooter company. Jan should feel that here on this forum she is interacting with the END most user, & they comprise of high end fellas to lower end ones like me. Here she will get what end users feel & want from panny.

Believe me if panny decides to reduce the HUGE margin they make on P2 sales this sort discussion will die off suddeny. but no its not gonna happen & so will the companies making cost effective solution will win hands down.

My thoughts flushed out on this page.... No personal attacks so pls dont get back with vengeance I am a end user who is expressing his simple thoughts. i hope I am helped by Jan to start thinking Panasonics way...

ROne
03-14-2009, 01:06 AM
Why are we still doing the cost thing on P2? It's Sesame street maths.

My company makes a reasonable 85,000 profit on 120,000 turnover. That includes the cost of buying new equipment and running the business. P2 costs and camera are maybe 10% of that figure spread over several years! Say 3.33% a year to allow upgrade every 3 years. (Not to mention tax write off.)

If it's too expensive buy the alternative then or get more business.

Expensive is relative, expensive is subjective - but expensive is not a fact.

What I like is quality and reliability. P2 achieves that mecca.

Jason Stewart
03-14-2009, 02:34 AM
OK Big Brother, we get it. You don't value p2 cards because of the premium price, etc, but understand that others do. You don't know what pannasonic's profit margin is on the cards or how complex or simple the internals are(does it really matter??) so some of your remarks on these subjects aren't that creditable and kinda immature.

You definitly put some thought into your price comparisons that we saw many times and it seems for what you think you need purchasing the external recorder will serve you better. Since this topic interests you so much, once you get this setup, you should post an extensive review on it for us. I'm sure we'll all be intersted to hear the outcome.

Years ago i did something similar with my dvx. I needed long uninterupted recording times so I purchased a firestore (same recording format though)and it got me tapeless and served me well but wasn't a solution for all my needs. When I got a P2HD camera, I was already familiar with the tapless workflow but loved the 100% reliability and performance I got from P2 even with that so called HUGE profit margin you talk about. Even though my comments or others won't change your opinion on this, just know that I, like many others, value P2 cards/products and the INVESTMENT was well worth it.

What I find interesting how some people changed their tune about P2 once avc-i became available at this price point. I also understand that C-D needs to sell their products and shouldn't be overlooked. The cameras internal avc-i codec and C-D's device will BOTH provide much better quality compared to other codecs in this price range. There is enough room for both.



I think everything productive on this thread's topic has been said, again and again. So relax, that's enough.

Jan_Crittenden
03-14-2009, 03:39 AM
If the Flash XDR is better...so what? The HPX is the same price as its competitors and my all accounts a bit better. It can use the XDR the same way.

Point is that why would you? The FLASH XDR is not better, you can get a better recording using the on camera recorder. Why would you spend the $5000 on the XDR to get a lesser recording than what the camera can offer? It makes no sense to me.



If AVC Intra is better...so what? The Flash is smaller, uses less expensive media, can interface to an HDMI device (in nanoFlash format).

The Camera is able to record a better picture with having anything attached to it. The flash is smaller but it doesn't have a camera. So if that is the system that you want to use, find a camera that doesn't have a 10 bit recorder to hang it on.



And other than the XDR, its native recording is significantly better than the XDCAM

But AVC-Intra is better than the XDR. 10 Bit will always be better than 8 bit any way you slice it.


There is no need for a battle...the playing field is different.
An Flash XDR is $4800. And a HPG-20 with monitor is $4600; that device is probably more comparable.

Exactly but the difference will be that the HPG20 will record 10 bit and the XDR will only record 8 bit.


P2 questions: With regards to power loss...that facility is in the camera or on the P2 card? How does the battery powered Flash XDR deal with power loss?

That facility is part of what is in the card. There is an LSI in the card. It is part of why we say the card is format agnostic. If we came out with a 200Mbs AVC-Intra codec the cads we produced in 2004 could still handle it. It handles the parity checking, metadata once generated, and file structure. It also controls the Raid on the memory sites, which is how we get the read/write speed, currently at 800mbs on the 64GB cards, and it closes the recording every two seconds so that the most you would ever lose is 2 seconds. You would have to repair the clip that was recording but that would only be because it would need a finishing close.

I will let the XDR guys answer the question on how the XDR handles power loss.

Best,

Jan

Big Brother
03-14-2009, 04:21 AM
I will let the XDR guys answer the question on how the XDR handles power loss.

Best,

Jan

i hope XDR folks do respond to it. I by all times have been looking like a fool to market my pov on XDR ... I will let C-D take over from here to do better explaining :)

Jan_Crittenden
03-14-2009, 04:37 AM
[QUOTE=Big Brother;1577396]Hi, Any ida what goes in the inside of P2 card. Can Panasonic shws us an open P2 card, it shall help (at least) me to beileve on why P2 cards are Expensive.

Well if you have failed to pick up any of what has been said ovre the last four years and what has been said in this thread I can't help you. In direct response to you I have said what happens inside the P2 card. In response to others I have said what is in the card.


I spoke with C-D folks & was told that the power failiure result in last clip being stored geting lost but they are upto some development thru firmmware where we can recover the last lost clip if in case of any power failure. But a power failure can happen in only Two cases if we shoot on bats. Firstly if a foolish cameramen fails to notice Batt end or near end warning. secondly if bat goes dead in the middl of shot for some tech fault.

So in other words currently you lose the last clip. I know a pretty good DP and he by no means is foolish, but he lost his shot due to things just going wrong, and got to pay for another helicopter ride.


In any which case I am really impressed with HPX300 in anyy senses but still not happy with P2 prices,,, hopefully Jan has some images of an open P2 card which shouldnt be too hard for her to Get being in PANNY.Unless Panny doesnt want to reveal the simple layout plan inside of p2thing.

The cost of 2 P2 Cards is equal to the cost of the XDR, add a third card and you could record all day without stopping, as one card could be offloading.

I am not going to cut open a $2000 card so that you can see the inside. We do not make them in the US. They are made in a Clean Room in Osaka. I have a picture of an 8GB card cust open but they no longer look like that on the inside.



I am sure Jan will oblidge us all with the tests mentioned above to proove the discussion with facts.

The P2 card has performed flawlessly for the last 4 years for our customers. That I think is the best test one can have. We have over 100,000 P2 devices in the world working for some of the most demanding customers in the world. I have laid out facts, I have stated how the card works. The proof is in customer satisfaction.



B it codec 100vs160mbps or P2 inners so much fan faired so called complex design which alow them to double the capacity each NAB but not drop the prices. :)


Well here you are wrong, the price has come down. The 8GB card when intro duced was $2400. The 16GB Card was $900 when introduced. The memory we use offers 100,000 rewrites.


Now my aim here was solely to compare the much lower vost media & its recorder C-D to the MUCH expensive P2 card.we need four p2 cards per camera as we dont want to have a shoot held up cox the card simply is wasting its our time while it gets copied onto the P2 gear. We shoot average 5hours per camera effectively 10Gig P2 cards.

Frankly you wouldn't need 4 cards per camera and yes the P2 Portable would be a very handy thing to have in the tool kit. The cards offload faster than they can be recorded. We have a white paper on our website www.panasonic.com/p2hd called Time Management, you should take a look at it for solid information about how fast these cards are.


I use generally Three Cam setup any given event. My maths there has been like this for each HPX300 I will need aleast 4 P2 cards which jacks up the cost considerably & when i theritically speaking go the nanoFLASH way then this is what i calculate. (Not counting the cost of Hard Disks as it gonna be cooman to p2 or nanoFLASH way)

I have redone your math below

3 Panasonic HPX300 = 8500x3 = 25,500
P2 Portable= 4600
9 P2 cards = 21,492 (2 in camera, one at the offload station) Given that each card can be offloaded faster that one is filled, there is ample time for offloading the singular card. If you also equip that offload station with a mirrored raid, you will have a back up copy ready for the archive.

Total cost$ $51,592

However this assumes that you fill the card slots with 64 GB cards. You wouldn't have to. Why not put 32GBs in the slots? I mean the guys shoot the Iditarod with 16s and 32s and they come away with about 120 hours of footage at the end. Then the table would look like this:

CamerasX3= $25,500
Cards X9 = $12,600
P2 Portable= $4,600
Total= $42,700

The benefit of this calculation is that the cards offload even faster(because they are smaller), and they are less expensive than the 64GB. Since you had in the earlier budget the P2 Portable this is still in this one, and there is no need for the computer on set. Frankly there is no need for a trained person to baby sit this process, it is so simple that you basically tell it to offload the card and walk away. The Serial number of the copied card is on the screen and if you want to verify you can match the number on the card or play back off the HDD a couple of seconds. Then reformat the card and put it back into the camera. It really is not rocket science.

Plus having the P2 portable there can allow for playback at the event. Another function you can offer your clients.


Believe me if panny decides to reduce the HUGE margin they make on P2 sales this sort discussion will die off suddeny. but no its not gonna happen & so will the companies making cost effective solution will win hands down.

You are so misinformed about what it costs to make these cards with the memory we are using we do not make huge profits with these cards. If we change to memory that is comparable to the CF we could reduce the price.

Best,

Jan

Jan_Crittenden
03-14-2009, 04:42 AM
PS We will use our $24K Tektronix HD-SDI waveform monitor and look at the images on a Sony TriMaster monitor.

Either we paid a lot less than you for the same scope or my tech gave me the wrong price, as we have the same scope.

While you are doing your tests, check out the color gradations.

Best,

Jan

Big Brother
03-14-2009, 06:45 AM
Jan, I thank you for your replies but must let you know that they very hostile in tone. Pls do let us know more on your wonderfull products but with more friendly approach. half of your marketing is done by that it self.

Chill & Peace.

Bassman2003
03-14-2009, 07:28 AM
To me. the P2 price reaction seems to be related to how long one typically records.

With the introduction of ACV-I, that time has gone up to two hours.

With the plan just laid out using a computer and P2 portable, one must juggle data during recording or on a break.

To me this is dangerous. Wiping a card on location has some serious human error potential, which I see as a great weakness to this setup.

Also, the sheer cost of the larger P2 cards prohibits one from buying and extra just in case one gets lost, dropped or has whatever happen to it.

Back in the tape days, you would never show up with the exact amount of record time from what the client told you. With P2, as record times increase, you are operating at 100% efficiency as your cards might actually be less than the length of the entire event.

This is where I see the benefit of the CF cards as they are cheap enough to have plenty on set/location and you don't have to wipe any at the shoot.

This lets the shooter focus more on shooting and less on being an IT manager.

Big Brother
03-14-2009, 08:38 AM
i agree with bassman perhaps he has given words to my though process a better way. THX Man.

Dan Keaton
03-14-2009, 09:27 AM
Panasonic's AVC-Intra has the advantage in that it is 10-bit.

The Flash XDR, at this time, is 8-bit, as is all MPEG-2.

For color correction, we recommend editing the 8-bit MPEG-2 data in a 10-bit timeline (or canvas).


Panasonic has an advantage if one loses power.

The Flash XDR, internally and seamlessly, breaks up lengthly recordings into sub-clips (files). At this time, if the power is lost, the last clip is not closed. It is our intention to improve this in the future by recovering the last clip.

We do expect one to properly manage their battery system.

With the Flash XDR this is relatively easy; as commonly available, industry standard batteries will power the unit continuously for well over 6 hours straight.


I previously made statements that AVC-Intra required a transcode before editing and was not supported in some NLE's. I stand corrected.


The Flash XDR records in I-Frame Only and Long-GOP modes, your choice. The recording bit-rate can also be selected via menu options to tailor the recording quality / recording time to your current requirements.


The small, portable, battery powered temporary storage for CompactFlash media, the Nexto DI, model ND-2700, in 500 GB version is available for $330. A version without the hard disk drive is available for $130 and a great 500 GB hard disk drive is available for $100.

The Transcend 133x write, 300x read 32 GB CompactFlash card is available for $84 each.
These are reliable, inexpensive, high-quality CompactFlash cards with a lifetime warranty.

These work with all of our recording options, except for 160 Mbps I-Frame Only mode, which requires a 300x UDMA card, which is more expensive. The SanDisk Extreme III 32 GB card is a good option at $205.00.

The Flash XDR was released to the public before all promised features were implemented. Our customers asked us to do this.

The Flash XDR is being enhanced to fulfill our promises as well as to respond to customer's requests on a regular basis. We typically add significant new features at least twice a month. We also respond as quickly as possible to any problems reported by our users.


I hope one notices that I have attempted to state facts and have not attacked or been disrespectful.

Big Brother
03-14-2009, 09:43 AM
I am very keen on trying out your product on HPX300 (as it gets launched).

i was wondeing Does HPX300 give a option to have the P2 card run in cycle meaning that it restarts recording over the previous recorded footage when the p2 card reaches its end. this way i would love to have the p2 card as a dummy recording device which will be reused again & again ... w/o formatting after it end its storage.

can you pls post few pix of XDR or NanoFlash in use with Full body camcorders ... Being used with Pro batt etc type connection.

Not even Nexto but any stand alone Data CF copier can suit the job for copying XDR/nano Cf recording off the field.it will really be very handy . The cost of High Speed CF cards is almost around 200$ that one can maintain enough number to shoot all thru the day w/o bothering to copy for reuse. One would like t copy to have a safe backup ... If i am using a P2 card then I am sure to backup & free up the EXPENSIVE card to be reused. In this P2 fashion I am mostly engaged in keeping track of when card gets filled , when it gets dumped thru a P2 store/gear & then again concentrate on shoot. in a way i will never feel lazy on field with p2 ... I am gonna be more alert. :)

Dan pls have more worked out Nano to let us remain a shooter then a IT manager on location. Esp on event days when we are working hard to keep track of shots rather then tracking EXPENSIVE media getting copied/filled/formatted etc.

Wish list for XDR/nano
- Composite out of recorded stuff from the device, in a way it should play back what been recorded on the CF card to store confidence on it recording or a small LCD monitor on the unit itself which play whats been recorded (in a way like pre-read what we had in Beta days where the head reads out what been just recorded, hope i remember it properly) this will instill faith on your product

- camcorder hooking device which can be used to power the unit & also keep it hooked to the camera

- Audio output thru Ear phone

- custom file naming or something were we can give identity of the shooter/camera

- instead of rubber cover of CFcards etc we should get plastic cover which can slide up/down this way we wont have to spend time fitting the rubber covers and/or risk it being lost in field.

can you pls post some pix of the connector side of XDR & Nano unit. Also whats the power consumption of Nano & XDR units in Watts.

DavidNJ
03-14-2009, 10:36 AM
Ok...while 10-bit has more color depth, recording that color depth takes away for other compression information without an algorithmic improvement. Given a fixed data rate and not necessarily having a critical color detail loss, it is most likely a trade off.

Second, the HPX300 (that is the thread, isn't it?) comes with the AVC Intra as part of the camera. It is the same price as an EX3 and only slightly more the XDCAM EX JVC HM700. Net, it is the best ENG/event camera in that price range.

P2s are nice...a little RAID 0 memory card. Their speed was an issue in 2004. Now, CF can match it so it just an ease of handling and reliability issue. Important for ENG. Less important for independent film makers and budget conscious event videographers.

Lacking any comparisons recording on both formats, it is hard to judge the difference between the two formats. Both have more compression than DVCPro50. Which is also 10-bit :)

I calculated 5 compression for the XDR (800Mb/s/160Mb/s)) and 10:1 for AVC Intra 100 (1Gb/s over 100Mb/s or 800Mb/s over 80Mb/s). However, assuming 13Mb/s AVCHD equals double that speed MPEG-2, and applying a log scale for reduction that drops AVC Intra to a 'virtual' MPEG-2 of 6.6...versus 5. Not much of a difference.

The variable block size of H.264 may allow AVC Intra to capture stuff that is otherwise missed in the compression. However, we would need to do tests to determine this.

To really do the tests well, it would be ideal if in addition to a HPX300 or EX3 the compression could be applied to an oversampled image or computer image. That is not from a camera with 1:1 sensor to image sampling. The low pass optical and digital filters would remove a lot of the higher frequency detail. You would want an actual camera image also to show what it is on an image that went through those filters. What is important detail on an image from a Red One may not be important if it comes form an HVX200.

SPZ
03-14-2009, 11:02 AM
I've been using P2 for 4 years, and it has never failed me. Not once. It has proven to me to be reliable in the most demanding situations. They are expensive. Never used professional CF cards and don't know if they fail, but P2 really did give me safety and was a very solid way to go Tapeless.

AVC Intra is a fantastic codec. Sincerely, If this was the codec that came with the camera I wanted next, I wouldn't need an external recorder. But I would also be happy with both 100mbs Mpeg 2 Long GOP 8bit and the 160Mbs Intraframe Mpeg2, if they came with any camera. So I do not see any problem with this. I also don't see what's the need of a Nanoflash for any HPX300 owner. Panasonic is using AVCIntra 100 for their top of the line cameras, cameras that are used in multi million dollar productions. So this is. obviously, a master quality codec.

For non Panasonic AVCintra cameras, however, I can see were the Nanoflash could be a better value. I'm currently on the fence not on the codec, but on the camera I will go with. If I go HPX300, I won't get a Nanoflash. If I get an EX3 or JVC700 , however, the nanoflash will be on my purchase list. Even if there are differences between the Nanoflash top of the line codec and Avcintra, it will be minimal. Avcintra or Nanoflash vs XDCAM EX or DVCPRO HD, however, is Apples and oranges...

Mike Schell
03-14-2009, 11:10 AM
Ok...while 10-bit has more color depth, recording that color depth takes away for other compression information without an algorithmic improvement. Given a fixed data rate and not necessarily having a critical color detail loss, it is most likely a trade off.

Second, the HPX300 (that is the thread, isn't it?) comes with the AVC Intra as part of the camera. It is the same price as an EX3 and only slightly more the XDCAM EX JVC HM700. Net, it is the best ENG/event camera in that price range.

P2s are nice...a little RAID 0 memory card. Their speed was an issue in 2004. Now, CF can match it so it just an ease of handling and reliability issue. Important for ENG. Less important for independent film makers and budget conscious event videographers.

Lacking any comparisons recording on both formats, it is hard to judge the difference between the two formats. Both have more compression than DVCPro50. Which is also 10-bit :)

I calculated 5 compression for the XDR (800Mb/s/160Mb/s)) and 10:1 for AVC Intra 100 (1Gb/s over 100Mb/s or 800Mb/s over 80Mb/s). However, assuming 13Mb/s AVCHD equals double that speed MPEG-2, and applying a log scale for reduction that drops AVC Intra to a 'virtual' MPEG-2 of 6.6...versus 5. Not much of a difference.

The variable block size of H.264 may allow AVC Intra to capture stuff that is otherwise missed in the compression. However, we would need to do tests to determine this.

To really do the tests well, it would be ideal if in addition to a HPX300 or EX3 the compression could be applied to an oversampled image or computer image. That is not from a camera with 1:1 sensor to image sampling. The low pass optical and digital filters would remove a lot of the higher frequency detail. You would want an actual camera image also to show what it is on an image that went through those filters. What is important detail on an image from a Red One may not be important if it comes form an HVX200.

Hi David-
A small correction on your math (and Jan please correct me, if needed). First, let's assume 1080psf23.98 format, which is popular in the film industry, where quality is critical. (BTW, this format is curiously missing in the new AG-HPG20 AVC-I recorder specs).

So, AVC-I is 100Mbps for 1080i60 or 1080p30, but scales with the frame rate, so at 23.98p, the bit rate drops to 80 Mbps. When recording 1080psf23.98, the Flash XDR / nanoFlash runs at the full 160 Mbps data-rate (no reduction).

The uncompressed bit-rate for AVC-I compression is 1920x1080x2(luma/chroma)x10(bits)x23.98(frame rate) = 994.49 Mbps. So the compression ratio = 994.49/80 = 12.43:1

The uncompressed bit-rate for MPEG2-I is 1920x1080x2x8x23.98 = 795.59 Mbps. So the compression ratio = 4.97:1

So at this format (1080psf23.98), the MPEG2-I has 2.5 times less compression. This comparison changes depending on the video format, so at 1080i59.94, MPEG2-I has a 2X advantage, while at 720p23.98 the advantage increases to 5.6X.

Hence my poiint that while AVC-I has an advantage of 10-bit (vs 8-bit for MPEG2-I), the lower compression ratio will greatly reduce the difference. Again, we are comparing compressed images and compression always reduces the resolution of the data and introduces round-off errors, especially in the least significant bits (the two extra bits found in 10-bit encoding).

Mike Schell
Convergent Design

Big Brother
03-14-2009, 11:33 AM
Mike pls addres these points -
1. Does XDR/nano records in 1080i50 (pal land) so on XDR/Nano
2. whats the power consumption of these units in WATTS.
3. can the wish list be thought for ( i stated above ) incorporation in future variants
4. if i go 422 160mbps in 1080i50 then how much recording time can i expect on a 32GB card at this data rate

Finally i saw on HPX300 user guide that loop recording is very much possible on HXP300. Now only concerning me is how to get confidence that the clips are getting stored on CD prods safely...

DavidNJ
03-14-2009, 01:13 PM
I've been using P2 for 4 years, and it has never failed me. Not once. It has proven to me to be reliable in the most demanding situations. They are expensive. Never used professional CF cards and don't know if they fail, but P2 really did give me safety and was a very solid way to go Tapeless.

AVC Intra is a fantastic codec. Sincerely, If this was the codec that came with the camera I wanted next, I wouldn't need an external recorder. But I would also be happy with both 100mbs Mpeg 2 Long GOP 8bit and the 160Mbs Intraframe Mpeg2, if they came with any camera. So I do not see any problem with this. I also don't see what's the need of a Nanoflash for any HPX300 owner. Panasonic is using AVCIntra 100 for their top of the line cameras, cameras that are used in multi million dollar productions. So this is. obviously, a master quality codec.

For non Panasonic AVCintra cameras, however, I can see were the Nanoflash could be a better value. I'm currently on the fence not on the codec, but on the camera I will go with. If I go HPX300, I won't get a Nanoflash. If I get an EX3 or JVC700 , however, the nanoflash will be on my purchase list. Even if there are differences between the Nanoflash top of the line codec and Avcintra, it will be minimal. Avcintra or Nanoflash vs XDCAM EX or DVCPRO HD, however, is Apples and oranges...

If you already have P2 cards there are few reasons not to chose HPX300 over the EX3 or HM700. The EX3 is just not new technology. And the HPX300 is loaded with it. And nice features...the view finder, the spot zebra (The Sony has a version also), the waveform monitor and vectorscope. The difference in sensor size is about the same as a 4/3rds vs APS-C... would it be really noticeable?

Now, if your customer requires XDCAM EX, it would be different...or maybe adding an XDR?

DavidNJ
03-14-2009, 01:24 PM
Hi David-
A small correction on your math (and Jan please correct me, if needed). First, let's assume 1080psf23.98 format, which is popular in the film industry, where quality is critical. (BTW, this format is curiously missing in the new AG-HPG20 AVC-I recorder specs).

So, AVC-I is 100Mbps for 1080i60 or 1080p30, but scales with the frame rate, so at 23.98p, the bit rate drops to 80 Mbps. When recording 1080psf23.98, the Flash XDR / nanoFlash runs at the full 160 Mbps data-rate (no reduction).

The uncompressed bit-rate for AVC-I compression is 1920x1080x2(luma/chroma)x10(bits)x23.98(frame rate) = 994.49 Mbps. So the compression ratio = 994.49/80 = 12.43:1

The uncompressed bit-rate for MPEG2-I is 1920x1080x2x8x23.98 = 795.59 Mbps. So the compression ratio = 4.97:1

So at this format (1080psf23.98), the MPEG2-I has 2.5 times less compression. This comparison changes depending on the video format, so at 1080i59.94, MPEG2-I has a 2X advantage, while at 720p23.98 the advantage increases to 5.6X.

Hence my poiint that while AVC-I has an advantage of 10-bit (vs 8-bit for MPEG2-I), the lower compression ratio will greatly reduce the difference. Again, we are comparing compressed images and compression always reduces the resolution of the data and introduces round-off errors, especially in the least significant bits (the two extra bits found in 10-bit encoding).

Mike Schell
Convergent Design

I made the correction...however, isn't 4:2:2 already a form of chroma compression? Ditto 8 vs 10-bit. There is a small bit of resolution lost. Doesn't going to 10-bits for 8-bits in the same output bit rate reduce the spatial resolution?

If the Flash XDR can do I-frame rates to 160Mb/s, what was the constraint preventing long GOP at 160Mb/s? Does 160Mb/s I-frames have more resolution than 100Mb/s long GOP? I would have thought they would be close. At the same bit rate doesn't it become a trade off of spatial resolution vs. temporal resolution?

For a large part, doesn't the I-frame vs. long GOP depend on whether you have the opportunity to transcode in your workflow?

Dan Keaton
03-14-2009, 01:37 PM
Mike pls addres these points -
1. Does XDR/nano records in 1080i50 (pal land) so on XDR/Nano
2. whats the power consumption of these units in WATTS.
3. can the wish list be thought for ( i stated above ) incorporation in future variants
4. if i go 422 160mbps in 1080i50 then how much recording time can i expect on a 32GB card at this data rate

Finally i saw on HPX300 user guide that loop recording is very much possible on HXP300. Now only concerning me is how to get confidence that the clips are getting stored on CD prods safely...


1. Yes.

2. Flash XDR 14 to 16 watts, nanoFlash around 7 watts.

3. I will answer those questions separately.

4. Approximately 24 minutes. This is a calculated figure for 1080i60. I let the Flash XDR calculate this value for me. I do not have a 1080i50 source with me today.

We recommend that you run your own quality tests to compare 160 Mbps I-Frame to 100 Mbps Long-GOP which gives you longer recording time and great quality.

Typically, one uses an HD-SDI monitor: Camera to Flash XDR to HD-SDI monitor.

Confidience monitoring, which typically involves tape media, involves a read after write function. This is not required with modern solid-state media.

I will answer your other questions shortly.

Barry_Green
03-14-2009, 01:50 PM
So at this format (1080psf23.98), the MPEG2-I has 2.5 times less compression.
But you guys are going in circles trying to prove with math, something that can't be proven that way. You can't just say "we're 5 to 1, they're 12 to 1" because a) you're not using the same compression technology, and b) you're not even compressing the same data, you're comparing 10-bit vs. 8-bit.

AVC has an advantage in compression efficiency. Arguing the math is only going to work to influence those who don't know what is going on, and that doesn't really serve anyone.

Just shoot some images. Strap both recorders onto the two SDI outputs of an HPX3000 and shoot some footage and let us see it without the numbers argument, because frankly that's a non-starter.

If someone was going to follow the numbers, they might come to the conclusion that MPEG-2 is better. But that's a just plain silly conclusion, and every standards organization that I know of has concluded that AVC, and especially the AVC-Intra 100 implementation of it, beats MPEG-2. I can't think of anyone mounting a credible argument against it, in fact. Even Sony Broadcast themselves have indicated that when they move to 1080/60p, they're going to abandon MPEG-2 and go with AVC. Sony Consumer, while currently actively selling MPEG-2 HDV, has said that AVC is the "format best suited for the coming High Definition era." So arguments about bits or compression ratio are avoiding the central core issue, which is: does one look substantially better than the other?

Because if they look *identical*, then surely you'd want half the file size (AVC-I 1080/24pN 80mbps vs. MPEG-2 160mbps) and 10-bit color, right?

Wouldn't you have to prove that the XDR is providing a noticeably superior image over AVC-Intra in order to make it a viable or even attractive alternative to the built-in AVC-Intra recording on the HPX300? (seeing as this thread was, and perhaps still is, about using the XDR with an HPX300). If you're asking someone to swallow using twice the storage space and twice the bandwidth, there needs to be a corresponding increase in image quality to justify that. And I don't expect that to happen, when comparing a 10-bit codec to an 8-bit codec. Especially when that particular 10-bit codec has been certified "Gold" by the pickiest nitpickers of all, Discovery HD and the BBC.

So, can we dispense with the numbers and get on with the visuals please? :)

Mike Schell
03-14-2009, 02:11 PM
Hi Barry-
Agreed, let's move onto the visuals, the whole matter is way too complex to analyze on a forum. Let me try to arrange a shoot-out, that's the most definitive answer.

Mike Schell
Convergent Design

Dan Keaton
03-14-2009, 02:16 PM
Ok...while 10-bit has more color depth, recording that color depth takes away for other compression information without an algorithmic improvement. Given a fixed data rate and not necessarily having a critical color detail loss, it is most likely a trade off.

Second, the HPX300 (that is the thread, isn't it?) comes with the AVC Intra as part of the camera. It is the same price as an EX3 and only slightly more the XDCAM EX JVC HM700. Net, it is the best ENG/event camera in that price range.

P2s are nice...a little RAID 0 memory card. Their speed was an issue in 2004. Now, CF can match it so it just an ease of handling and reliability issue. Important for ENG. Less important for independent film makers and budget conscious event videographers.

Lacking any comparisons recording on both formats, it is hard to judge the difference between the two formats. Both have more compression than DVCPro50. Which is also 10-bit :)

I calculated 5 compression for the XDR (800Mb/s/160Mb/s)) and 10:1 for AVC Intra 100 (1Gb/s over 100Mb/s or 800Mb/s over 80Mb/s). However, assuming 13Mb/s AVCHD equals double that speed MPEG-2, and applying a log scale for reduction that drops AVC Intra to a 'virtual' MPEG-2 of 6.6...versus 5. Not much of a difference.

The variable block size of H.264 may allow AVC Intra to capture stuff that is otherwise missed in the compression. However, we would need to do tests to determine this.

To really do the tests well, it would be ideal if in addition to a HPX300 or EX3 the compression could be applied to an oversampled image or computer image. That is not from a camera with 1:1 sensor to image sampling. The low pass optical and digital filters would remove a lot of the higher frequency detail. You would want an actual camera image also to show what it is on an image that went through those filters. What is important detail on an image from a Red One may not be important if it comes form an HVX200.

Dear Barry,

This post was written while you were creating your post. Please feel free to delete it if you wish.

We have images that show the quality, I will be happy to email them to you, if desired.


Dear DavidNJ,

In general, please remember we are trying to answer specific questions as asked on this forum. The Flash XDR is a general purpose, high-quality HD-SDI recorder. It just happens to be able to record from the HPX-300, as well as any other HD-SDI source.

Whether or not it provides value, or not, to an owner of a HPX-300 is up to the owner. We have never promoted the Flash XDR for use with the HXP-300. However, the two devices will work together and may provide value to some users.


DavidNJ, your observation that after compression, differences between 10-bit vs 8-bit is less than the differences before compression, is very astute. We agree.


The obvious: Not all CompactFlash cards are created equal in quality and performace.

However, high-quality CompactFlash cards that we have qualified for use with the Flash XDR are very reliable and are ideally suited for the task at hand.

In a future firmware release, we will support redundant recording. Thus, one will be able to record to two separate CompactFlash cards simultaneously, thus creating two separate "Masters".

This will reduce the risk of a media failure to near zero and allow one card to be held safely while the other card can be mailed directly from the set location.


We are in complete agreement, the proof of any camera or recording device is in the images.

The Thomson Viper is considered by many to provide the highest quality video images available today.

We have multiple Thomson Viper camera users using the Flash XDR for film-out.

One Emmy Award winner has been testing the Flash XDR for use with his Viper for approximately one month now.

A very skilled and experienced DP tested the Flash XDR yesterday.

All of these experts have found our images to meet their high quality standards after exhaustive testing.

I mention the above in response to your request to record a highly detailed image to our Flash XDR. I did not have to resources available today to generate and record a highly detailed computer image.

I attempted to post an image that shows the quality of a Thomson Viper image, as recorded by the Flash XDR, but the file was too large.

I hope this helps.

DavidNJ
03-14-2009, 04:31 PM
This thread is beginning to get funny. So sample clips, Flash XDR vs AVC Intra off a HPX 300 will answer the thread. However all this little minutia really is off the point and clouding two (three with the nanoFlash, maybe four with the HPX 200) great products.

The Flash XDR and nanoFlash XDR give all camcorders (and hopefully soon some hybrids) with HDMI or HD-SDI output the ability to record a high quality image that make the lens and sensor the image quality issues. I would love a lower price, but then...where is the competition? A great concept and I'm sure a great product.

The HPX300 redefines what can be expected in a around $10k professional camcorder. All the great shooting features (great viewer, waveform, vectorscope, spot zebra), excellent sensor with cute features like chromatic aberration detection, and so far and away the best recording on any camcorder within 5x it price that we have 10 pages of knockdown dragout comparisons with the highest bandwidth recorder otherwise available.

When more footage is out there I can imagine anyone with a prosumer camcorder with the sensor and lens to take advantage of it buying a Flash...or a least considering it.

Meanwhile, the HPX300 stands a good chance of pushing the EX3 and maybe the slightly less expensive EX1 aside simply because it is a much more complete package. Would you seriously compare either to a 35Mb/s long GOP 4:2:0 MPEG-2 configuration? It is like trying compare Tri-X pushed to 800 to Panatomic-X.

Will HPX300 buyers get a Flash or nanoFlash XDR? Some will, even if the tests show an advantage I imagine it will be so small as few will unless it meets their workflow needs. However, if you are considering a Sony or JVC, it could even be considered a requirement to gain an image competitive with the Panasonic.

My 2, probably only worth 1.5

SPZ
03-14-2009, 05:23 PM
If you already have P2 cards there are few reasons not to chose HPX300 over the EX3 or HM700. The EX3 is just not new technology. And the HPX300 is loaded with it. And nice features...the view finder, the spot zebra (The Sony has a version also), the waveform monitor and vectorscope. The difference in sensor size is about the same as a 4/3rds vs APS-C... would it be really noticeable?

Now, if your customer requires XDCAM EX, it would be different...or maybe adding an XDR?

Hmm, I'm simply looking for the best camera. I've already stated my case on another thread. If it was for on camera recording, the HPX300 would be the clear winner. But we all know this is only a part of the equation. We're talking about 10k here. Small money for some, but a big investment for others. It is a big one, for me. An investment that must last for years. So, even tough I invested in P2 and am happy with the workflow, ultimately, I need a big jump in performance from my HVX200.

DavidNJ
03-14-2009, 06:20 PM
Ok...let me try this...under 10k, it is all about compression. I see you don't believe me...so let me provide evidence...http://vimeo.com/1082016. That link is an HR11 captured via HDMI (uncompressed 8-bit 4:2:2) to Cineform. Now compare that to EX1, HPX170, HXV200, XL H1/A1/G1, etc. The big hammer on quality is the compression.

The HPX170/HVX200/HMC150 all have undersampled sensors, yet are still viewed favorably. They are recording DVCProHD when their competition was HDV. HDV was, to quote Sean Connery in The Untouchables 'taking a knife to a gunfight'. They had nearly 2x more compression and gave away 4:2:2 chroma subsampling.

MTF is the product of the MTF of each component, but good compression is the weak link, IMHO. Given equal capture, the differences between the EX1/3 and HPX300 lens/sensor are probably fairly small. The difference in recording is huge.

Everyone is always saying Scarlet. Sensors start at 2/3" and 3k resolution oversampling. The image is recorded at over 200Mb/s in RAW non-demosaiced format. The only way to address that otherwise is a hybrid digital camera, a technology in its infancy. There are examples of it being used for corporate videos though, and of course indepent film. It is more work, but those using it think it is worth it. But you have prime lenses, an large oversampled sensor (with some issues) and a recoring mecahnism with 20-40% more compression that AVC Intra 100, depending how you value long GOP.

If you have a camera with HD-SDI or HDMI output, trying a better encoding mechanism is a first step.

Mike Schell
03-14-2009, 07:18 PM
If you have a camera with HD-SDI or HDMI output, trying a better encoding mechanism is a first step.

Hi David-
We are in complete agreement. The encoder (compressor) is by far the most quality robbing part of most cameras. The quality of almost every camera would be improved with a better recorder, be in on-camera or captured via a laptop and I/O box.

Mike Schell
Convergent Design

Big Brother
03-15-2009, 12:17 AM
I hope if we can get some same scene shot on HXP300 on P2 10bit 422 thing and parallely on XDR/naon @ 160mbps quality in Iframe mode. I hope we get to see for ourselves visually. It shall be nice if small portion of same footage be uploaded in original format for evaluation.

For me now HPX300 with XDR is a winner , i will need two p2 cards in loop mode so no matter what they record on shall be useless forme. i wil like to go the CF way cost effective/qualityful/safe/ease of being free from dumping on field :Drogar-Smoke(DBG): i shall love to be a shooter than an IT manager !

DavidNJ
03-15-2009, 06:34 AM
Both companies CFOs will be happy but I fear Jan will frown. :)

Mike, does the XDR record 720p24 at 100Mb/s GOP and 160Mb/s I-frame? If so, the bandwidth difference would be 4x. That would only 2.2:1 compression.

Why do that? IF the XDR gives you the full bandwidth at 720, reasons are:

1) Less compression information losses

2) Not full spatial or chroma information loss. To avoid capturing noise, the camera has optical and digital filters around the Nyquist limit of 540lp/ph on this camera. So frequencies between 360 and 540lp/ph (roughly 15-27lp/mm on FF35) are at much lower contrast (MTF percentage). This is an advantage of oversampling cameras like the Red One.

One advantage of that combo is that you could always just shoot the HPX300 natively of for some reason your couldn't manage both devices, you wanted to be compatible with other cameras in the shoot (my WAG is for an HPX200 at NAB) or AVC Intra output was preferred by the customer. Does the XDR record regular XDCAM also, making it a univeral XDCAM source?

http://www.dvxuser6.com/uploaded/113/1237126304.jpg

P.S.
One Red One user noted that because of the weight, the support equipment would cost as much as the camera.

Dan Keaton
03-15-2009, 07:50 AM
Mike, does the XDR record 720p24 at 100Mb/s GOP and 160Mb/s I-frame? If so, the bandwidth difference would be 4x. That would only 2.2:1 compression.

Yes, in all of our modes, we use full bandwidth.

LightPrism
03-15-2009, 09:28 AM
The XDR does not do I-Frame at this time.


.....


Best,

Jan


1. I am looking at my XDR menu as we speak and it most certainly DOES do I-frame. The XDR does I-frame at 100 and 160mbps data rates.

2. I am wondering how many people have actually SEEN the various 50, 100, Long GOP, 100, 160 I-Frame recordings out of the XDR?
My only experience with the P2 card visual quality was with the limited DVX200 - it appeared too soft - but I am sure it was a limitation of the camera and not the codec. At the time, off loading was a pain but that has since been fixed as well.
My experience with the XDR has been wonderful. Huge quality and portability improvement with our XLH1 and XHG1.

I wished Adam Wilt would get his hands on an XDR and do some in depth testing and comparisons of these 2 options. And of course in should include color correction testing results.

And even the 8bit vs. 10bit debate will be moot when C-D adds their 10bit option. I do wonder how many folks will see the added value of 10 bit when considering the extra data wrangling and storage and actually use it.

DavidNJ
03-15-2009, 09:52 AM
You can't compare the HVX200...totally different compression. It is DVCProHD, an 8-bit MPEG2 I-frame protocol that was far and away the best for years. The new stuff is AVC Intra, an H.264 codec 10-bit 4:2:2 I-frame protocol at bit rates (depending on frame size and frame rate) not that far off the XDR.

Also, the HVX200 is a little prosumer camera with a fixed lens and undersampled 960x540 sensors. The HPX300 is has interchangable lenses and a latest technology full 1920x1080 3 sensor setup up.

DavidNJ
03-15-2009, 09:55 AM
Yes, in all of our modes, we use full bandwidth.

How about 10-bit 4:4:4 to 720p? Everyone is so used to 1:1 photo site to pixel processing they will scream...but my guess is that format, which loses no chroma information per 720p pixel to 1080p but still has significantly less compression could result in the best images with 1080 sensors for the reasons I posted earlier.

Is it possible?

LightPrism
03-15-2009, 10:08 AM
You can't compare the HVX200...totally different compression. It is DVCProHD, an 8-bit MPEG2 I-frame protocol that was far and away the best for years. The new stuff is AVC Intra, an H.264 codec 10-bit 4:2:2 I-frame protocol at bit rates (depending on frame size and frame rate) not that far off the XDR.

Also, the HVX200 is a little prosumer camera with a fixed lens and undersampled 960x540 sensors. The HPX300 is has interchangable lenses and a latest technology full 1920x1080 3 sensor setup up.

And that is why I said I am sure it was a limitation of the camera.

The point is there are a lot of folks here debating without having actually seen the footage out of the XDR or the new Panny. Its all about what actually ends up on screen.

That is what I am looking to hear about - or better yet see.

One other point - what is the bit depth of the AUDIO of AVC? The XDR records in 24bit; I honestly don't know about the audio bit depth out of the new Panny to AVC.

DavidNJ
03-15-2009, 10:39 AM
It is 16-bit. However, few on this forum discuss sound...I will pick up a Samson Zoom H4n soon as its out with 24-bit/96Khz 2-channel.

Big Brother
03-16-2009, 06:31 AM
looking for footage test from Convergence D with XDR/nano at various bit rates on Iframe v/s p2 panny.

JohnnyD
03-16-2009, 08:12 PM
Wow, I just caught this thread and had to grap another beer.

You guy's are way caught up in the specs. The HPX300 is a lower-end camera, at least by my standards. I loss track of how many $10,000 to $30,000 cameras I have bought in the last 25 years. If you're shooting weddings, corporate, industrial, , this could be the camera for you. But so could a Sony, JVD or Canon. The client isn't going to see the difference. He WILL see if you can tell a story.

It's all about money and where you're at. If you're just starting, buy a lower end camera that you can afford and rent a high-end cam for those jobs that need them. Later on in life, when you become Spielberg's DP, then shoot a $200,000 camera, etc.

Years ago, I shot a insert for an infomercial in a car. My BetaSP was too large. So,I pulled out my Hi-8mm video cam, and mixed the shot later in editing with my Beta video. No one called me and said: "Hey, what's with that lousy video in the car?" . They were more interested in hearing the story.

Anybody, remember Blue Blockers? All shot on Hi-8 and grossed around 12 million.

How many of you have HD scopes and HD-SDI monitors? And, I'm not talking about a computer screen.

I had a good weekend watching sports in HD. I also noticed all of the aritfacts of the video compression on all of the networks. So what are we seeing? 4:2:0? It looks worse than that. Yes, the cable and dish guy's are really compressing the heck out of the feed.


Forget about bits, and enjoy life with a bite of an In-N-Out Burger.

JohnnyD

Big Brother
03-16-2009, 11:36 PM
Let the discussion roll on for knowledge purpose.

LightPrism
03-17-2009, 08:59 AM
Wow, I just caught this thread and had to grap another beer.



I had a good weekend watching sports in HD. I also noticed all of the aritfacts of the video compression on all of the networks. So what are we seeing? 4:2:0? It looks worse than that. Yes, the cable and dish guy's are really compressing the heck out of the feed.


Forget about bits, and enjoy life with a bite of an In-N-Out Burger.

JohnnyD

Whenever I hear reference to the Discovery Channel and BBC spec requirements I get a kick out the logic of shutting out content by raising the bar and producer's cost by setting that spec bar so many levels above the quality that is broadcast to endviewer's screens. It doesn't make a lot of sense to me. As stated, what comes off the sattelite or down the cable is so highly compressed and bit rate limited that is not even anywhere close to their spec.

I can understand a level or two up for editing, archiving, manipulation headroom. But let's allow some common sense into the spec based upon the final broadcast output spec.

Big Brother
03-17-2009, 09:33 AM
Wish list for XDR/nano
- Composite out of recorded stuff from the device, in a way it should play back what been recorded on the CF card to store confidence on it recording or a small LCD monitor on the unit itself which play whats been recorded (in a way like pre-read what we had in Beta days where the head reads out what been just recorded, hope i remember it properly) this will instill faith on your product

- camcorder hooking device which can be used to power the unit & also keep it hooked to the camera

- Audio output thru Ear phone

- custom file naming or something were we can give identity of the shooter/camera

- instead of rubber cover of CFcards etc we should get plastic cover which can slide up/down this way we wont have to spend time fitting the rubber covers and/or risk it being lost in field.

Also

I have asked it earlier if we use this type of a adapter then on one side we mount any AB Mount batt & on other side we mount XDR. We provide XDR power from power tap plug on abtts or AB mount... will this combo work

http://antonbauer.com/hotswap.htm

http://switronix.com/product_details.php?cmd=info&id=200

or if not from the Power tap port can we use the power output from the camera back the 12v 5a dc out .. will that supply enough power to run xdr or nano flash !

Christian m
03-17-2009, 09:39 AM
Whenever I hear reference to the Discovery Channel and BBC spec requirements I get a kick out the logic of shutting out content by raising the bar and producer's cost by setting that spec bar so many levels above the quality that is broadcast to endviewer's screens. It doesn't make a lot of sense to me. As stated, what comes off the sattelite or down the cable is so highly compressed and bit rate limited that is not even anywhere close to their spec.

If the idea is good enough and it as for example Deadliest catch outrules large cams for most of the production they even accept dv/hdv. Discovery's shows are mostly relativly high end anyway so getting one financed won't be impossible, rates for renting Gold standard equipment aint that high.

Well take a look at BBC HD and planet earth which are broadcasted from time to time. At least here in Norway where we can get BBC HD(nordic version) its basicly a showcase of how good TV can be technically...with todays equipment. And if tomorrow allows for way more efficient compression or more bandwith, you will gett better results if the source for compression are a HDcam-sr master(even if the recorded codec was xdcam/dvcprohd, you are likely to color correc it and choose to render in much higher bitrates) and generally the images look as good as possible from the start. Recompressing an already compressed signal like hdv is not the best idea for good pictures...

Big Brother
03-17-2009, 09:45 AM
thread topic is " Flash XDR on HPX300 "

And i still wonder how does CD suggest to mount XDR or Nano on HPX300 type corders. still no clear pix on this.

DavidNJ
03-17-2009, 10:24 AM
Wow, I just caught this thread and had to grap another beer.

You guy's are way caught up in the specs. The HPX300 is a lower-end camera, at least by my standards. I loss track of how many $10,000 to $30,000 cameras I have bought in the last 25 years. If you're shooting weddings, corporate, industrial, , this could be the camera for you. But so could a Sony, JVD or Canon. The client isn't going to see the difference. He WILL see if you can tell a story.

It's all about money and where you're at. If you're just starting, buy a lower end camera that you can afford and rent a high-end cam for those jobs that need them. Later on in life, when you become Spielberg's DP, then shoot a $200,000 camera, etc.

Years ago, I shot a insert for an infomercial in a car. My BetaSP was too large. So,I pulled out my Hi-8mm video cam, and mixed the shot later in editing with my Beta video. No one called me and said: "Hey, what's with that lousy video in the car?" . They were more interested in hearing the story.

Anybody, remember Blue Blockers? All shot on Hi-8 and grossed around 12 million.

How many of you have HD scopes and HD-SDI monitors? And, I'm not talking about a computer screen.

I had a good weekend watching sports in HD. I also noticed all of the aritfacts of the video compression on all of the networks. So what are we seeing? 4:2:0? It looks worse than that. Yes, the cable and dish guy's are really compressing the heck out of the feed.


Forget about bits, and enjoy life with a bite of an In-N-Out Burger.

JohnnyD

You are absolutely right...TV is hammered. I usually ask people to compare ABC to NBC...as part of a 1080 vs 720 discussion.

Now, I totally absolutely 100% disagree with you on the importance of the recorder. A $1000 consumer camera with a high end recorder (I don't have and from the XDR, but is a link to HR11->Cineform and there will be HF S10->Cineform footage in a week or so: http://images.vimeo.com/71/97/03/71970372/71970372_160x120.jpg (http://vimeo.com/1082016) (Click picture for link to video)

Without a high quality recording, many lens and sensor differences appear small. With it, huge.

Production quality is an important part of the perceived quality of the product. A great script and editng inside poor quality leads to 'it's a shame it didn't look better'. A bad script in high production quality leads to yawn's a trips for coffee, etc. Good production quality sometimes keeps your audience over the weak points.

Going from SD to HD created a 6x increase in data. Temporal compression was added in some cases, and some had higher data rates. HDV and AVCHD is the same data rate as DV and relies on temporal compression or algorithms. XDCAM EX raises the core rate 40%.

We are advancing production quality in three areas this year. First, AVC Intra, the 45Mb/s Class 6 SDHC Canons, and the XDR are a major advance in recording quality. The digital camera hybrids and the CMOS consumer cameras are using oversampled sensors where dramaitcally increase the MTF in mid and higher spatial frequences were detail lives. And while it hasn't hit most of the professional camcorders, the digital hybrids, the RED, and some high end professional systems are using larger sensors that offer reduced noise in addition to increased DOF control.

A good 2008 camera, like the EX1/EX3 will seem very much like DV in a few years. People still use this forums namesake, the DVX100, but the expectation will be much higher.

Mike Schell
03-19-2009, 01:14 PM
Wish list for XDR/nano
- Composite out of recorded stuff from the device, in a way it should play back what been recorded on the CF card to store confidence on it recording or a small LCD monitor on the unit itself which play whats been recorded (in a way like pre-read what we had in Beta days where the head reads out what been just recorded, hope i remember it properly) this will instill faith on your product

- camcorder hooking device which can be used to power the unit & also keep it hooked to the camera

- Audio output thru Ear phone

- custom file naming or something were we can give identity of the shooter/camera

- instead of rubber cover of CFcards etc we should get plastic cover which can slide up/down this way we wont have to spend time fitting the rubber covers and/or risk it being lost in field.

Also

I have asked it earlier if we use this type of a adapter then on one side we mount any AB Mount batt & on other side we mount XDR. We provide XDR power from power tap plug on abtts or AB mount... will this combo work

http://antonbauer.com/hotswap.htm

http://switronix.com/product_details.php?cmd=info&id=200

or if not from the Power tap port can we use the power output from the camera back the 12v 5a dc out .. will that supply enough power to run xdr or nano flash !


Hi Big Brother-
Attached is a comprehensive FAQ on the nanoFlash, this should answer most of your questions. Here's a couple of comments on the immediate concerns.

1) We do have HDMI output on the nanoFlash, so you will be able to view footage on low-cost portable LCD monitors at much higher resolution than composite video.

2) We have mounting points on 3 sides of the nanoFlash. The bottom of the box has a 1/4-20 tripod mount. We are also developing AB and IDX mounting plates which bolt onto the back of the nanoFlash. We will also have a battery plate on the top of the nanoFlash for Sony camcorder battery.

3)Audio output is not possible on the nanoFlash, we simply ran out of room on the box. Can you monitor off your camera?

4) We are adding metadata support for cameraman, camera number, scene and take numbers.

5) Yes, we agree that a sliding cover would be better than a removable cover, I will speak to our mechanical designer about this concept.

Best-
Mike Schell

Noel Evans
03-19-2009, 02:38 PM
No possible way will 100mbps long-GOP MPEG-2 be better than AVC-I @ 100mbps. But it would almost certainly be better than DVCPRO-HD, and should be better than AVC-I 50mbps.



Respectfully disagree with you Barry. Because there is more than one way to look at an image when you use the word "better".

For a small image quality increase, OK. But, then what are your workflow options? You could take all that footage and transcode to an I frame codec, which essentially means your re-compressing your footage and then whats the difference in image quality? Not to mention the amount of time I have just wasted to do so. This is certainly something I cannot ask a client to do. Oh by the way, you will need to transcode all that footage OR have horrible render times due to mpeg2. Oh you have a deadline? Bah, just pay your editor overtime!

I know this is just one case scenario and if you shot and are editing yourself and are prepared to eat that - fair enough. But my clients (I know them) would NOT be happy. And what of my clients working in p2? ME: Ok heres your mpeg2 files. THEM: Are you stupid?

Bassman2003
03-19-2009, 02:50 PM
Noel.

I don't know which NLE you or your clients use, but I have worked with the XDR 100mbps footage and it edits easy in Edius.

In fact, I can edit from the CF card in the USB reader.

And I agree, these higher level/bitrate codecs are going to be closer than farther apart and any image difference will be splitting hairs from a final delivery point of view imho.