rolling shutter hxp300?

ksteiger said:
All I know is I am eagerly awaiting the $500 camera using Kodak's amazing chip that Combatentropy is apparently about to start manufacturing. And if it doesn't have amazing depth of field capability, 8 channels of PCM audio and a free 64 GB stick of digital recording media, I will be P.O.'d.:beer:
Click to expand...

People continue to misread my post. I said I would rather have CCDs. puredrifting said, expect to pay more --- more than a $10,000 camcorder!

I do not think Panasonic should sell a 3-CCD 2/3" camcorder for $500 or even $1,500.

I do think someone could sell a camcorder I have in mind for about $5,000. And Ikonoskop and RED have announced cameras for about that price that are about what I am looking for. They are not quite right (RED's Scarlet still will be too big and heavy when kitted out, Ikonoskop has impractically priced media cards for me. Something like an Ikonoskop that records to SDXC would fit the bill nicely)
 
Last edited:
Do you have any idea what the cost is of a 1920 X1080 sensor that outputs 60 frames?

You need to work for a manufacturer for a while. If we even had a chip set, then there is the rest. Everyone thinks that R & D is inexpensive. For perspective, the DVX100B chipset cost $5,000,000 to develop. This is not an inexpensive business to develop product for, we as do all manufacturers work hard. We do try to listen and try to deliver where we can.

Best,

Jan
 
3 sensors from the GH-1 would each capture 60 frames a second (actually at 4k) and would use a prism to avoid Bayer artifacts. Down scaling to 720p or 1080p wouldn't be any harder then down scaling 1080p to 720p...possibly easier. The prism would increase low light sensitivity.

Three GH-1s are less than on HVX200, even at EOL.
 
Okay, guys, what am I missing?

Here's are two camcorder recipes:

- 2/3" 1920x1080 Bayer CCD
- AVCHD to SDHC
- HMC150 body
- $2,500

This is the least that I want. Even better would be:

- 4/3" Bayer CCD
- AVC Intra to SDXC
- interchangeable lenses
- with 10x lens, an HMC150 body
- $5,000

None of these ingredients are new. Yes, I believe R & D is expensive. How much is a good engineer? $100,000 a year, at least? I would guess it takes a dozen for a year or two to put together a camcorder from scratch. That's 12 * 100,000 * 2 = 2.5 million. Plus marketing, logistics . . . I can totally believe it cost 5 mil. I don't doubt it for a second!

But either one you could sell 10,000 of them, wouldn't you? At $5,000 apiece, you gross $50 million.

Not only that, but there is not a single new technology in either of my above camcorders. You are already recouping R & D for these technologies in other camcorders and even still cameras.

So what am I missing?


Panasonic, I will give you this. You are the best of the bunch. You have introduced features in camcorders more aggressively than Sony and Canon. Your build quality is solid. The color and tones of the DVX100 and others are beautiful. The size, shape, and weight of the HMC150 sounds like it is perfect.
 
Jan_Crittenden said:
Today, that is not possible.

Best,

Jan
Click to expand...

Well on lighter note, I am in a way happy for these cameras to be expensive as i have an edge over competition since most of them still work with PD150/170 type cameras as they are not able to jack up their affordability. kudos to the companies to make products for the privilege few.

Ok it was a joke so no leg pulling on it :)
 
Last edited:
combatentropy said:
None of these ingredients are new. Yes, I believe R & D is expensive. How much is a good engineer? $100,000 a year, at least? I would guess it takes a dozen for a year or two to put together a camcorder from scratch. That's 12 * 100,000 * 2 = 2.5 million. Plus marketing, logistics . . . I can totally believe it cost 5 mil. I don't doubt it for a second!
Click to expand...
If Japan has about the same living standards as Norway you probably are going to pay a lot for good engineers, if most RD are undertaken in Japan then. I might be wrong, but about 100,000$ are about average for the engineers in private sector up here in the cold north, the best heads often end up in the offshore business where they top out way beyond that...even in the times the world economics are in today good engineers are expensive.

A "hidden" cost for many are the support network the large manufactures have, RED and others that make the impression that you can have 4k at a budget are lacking that. It will be quite interesting to see if RED got the logistics to handle the price level they are aiming for with scarlet...it should sell loads. If a xdcam or p2 cam breaks down it will be possible to repair it relatively fast and get it back into making money relatively fast, with RED you ship the dam thing back to the US....
 
Creating a new camera is very expensive. Creating a new chip set alone costs about $5 mill. And it takes about 2 1/2 years to get a camera from concept to reality. At least that is what it has taken with the HPX300.

The cost of a 2/3" full raster CCD with the prism assembly, that is appropriate for video, starts around $10,000 and I might be low.

Best,

Jan
 
any handheld camcorder (based on the hpx300) being considered? Thatd be a good option.

I really dont see a cheap ccd based camera like combat entropy is saying.
 
I'll preface my post by saying that we regularly use VariCam, 2000, 500, 200 series Panasonic cameras in our productions - and that I (and my clients) love the Panasonic color and cadence.

I really didn't intend to get into the fray on this, but the 300 skew / warp issue is far more pronounced than anyone is talking about. I've been in the room with a Panasonic rep (who shant be named unless Panasonic does so) and we went through many framerates to try and find ones that performed better or worse. All of this was conducted on a BTLH1760 monitor - where smear won't manifest.

pN rates (1080 / 720) were unusable except in static shots, and everything else exhibited the skew issue to a great extent. Not just a little, but to a degree where we opted out of purchasing the unit.

You can test this for yourselves on demonstration cameras or the one you've already purchased. Point the camera, at any focal length (longer is more pronounced, of course), at any framerate other than 1080/60i (AVC or DVCProHD) at a doorway and pan from it to another point in the room, as if following a character in or out. You will see the skewing is a prominent effect, with doorways, switchplates and anything else perpendicular in your frame effectively becoming a parallelogram.

The rolling shutter on the 300 manifests during pans at all frame rates except 1080/60i, where it is diminished by over 50%.

It is far more pronounced in 24fps than the Z7U 3 CMOS cameras we're using for TLC, which are strange little cameras with strange connections, but they seem to have defeated the skewing issues somehow.

If you have trouble with my statements, conduct the test with a 300 pointed at a doorway and post your results here. Just follow some action - not a "whip-pan", but normal follow for talent or panning.

That's what you should be able to do with a camera. And this ongoing suggestion of "just shoot around" the skew effect is a laughable one at best for real world production.

e
 
BTW how much can the cost jack up if we get a 1/3" CCD in-place of 1/3" CMOS on HPX300 !

Hope Jan give us an idea ... may its workable for future ... 1/3" full raster AVC Camcorder WOAH ! All rolling issues tumble down ... :)
 
I think you have that already in the spatially offset LCD HPX500. As close as they've come anyway. Save for nifty features like the dual HDSDI outputs and HD viewfinder.

I still think the 2-slot P2 location on the 300 is also odd. No assistant can get there during shooting, and it's more important than ever with only two slots. The dumb side of the camera or the rear makes more sense to me. But so does four slots.

The rep asked me what I thought about this thing, and I told him it was nothing less than a "silver-bullet camera" - except for the rolling shutter issues. I really hope Panasonic can sort it (at least to the level of a Z7U), because we'd own a couple if they did.

e
 
Erik Olson said:
I really didn't intend to get into the fray on this, but the 300 skew / warp issue is far more pronounced than anyone is talking about. I've been in the room with a Panasonic rep (who shant be named unless Panasonic does so) and we went through many framerates to try and find ones that performed better or worse. All of this was conducted on a BTLH1760 monitor - where smear won't manifest.

pN rates (1080 / 720) were unusable except in static shots, and everything else exhibited the skew issue to a great extent. Not just a little, but to a degree where we opted out of purchasing the unit.
Click to expand...


puredrifting said:
6. Rolling shutter artifacts. Ah, yes, the $64,000 question. Well, what I observe seems to concur with Barry's skew test. The skewing is really not too noticeable in 720 24pN AVC Intra 100 or 50. It is pretty noticeable on 1080 24pN AVC Intra 100. I have not had a chance to shoot any 1080 60i or 1080 30p yet. I would normally shoot in 720 anyway most of the time but I do understand that most of you have 1080 24pN on the brain. I would agree, less skewing in 720 than the Sonys and more in 1080 24pN. Hopefully Panasonic's engineering wizards can improve the 1080. The clip I posted is a mixture of mostly 720 24pN AVC 100 mixed with some AVC 50 shots and a couple of 1080 24pN AVC 100 shots. If there is enough interest, I will post another version next week that identifies which clip was which, but for now, you have to guess.

Dan
Click to expand...


Okay, how do we reconcile these two posts?

Is there a straight answer in here somewhere?

best
 
My answer is based upon my observations and the footage I have shot so far. You will notice that my footage is real world subjects in the real world, not test charts and doorways. I don't shoot doorways for a living, I guess it all depends on if you do.

Yes, the skew is there, it is just a matter of how noticeable it is to you and your audience. The skew is always there on the EX1 and EX3 as well and I have seen plenty of successful programming shot with those cameras. My notes are in my post, yes, there is more skewing at 1080 24pN, less skewing at 720 24pN. As far as if that is acceptable to you, here is my straight answer. Use your eyes. Rent or borrow one and go out and shoot like I am doing. I can post all of the footage in the world but that doesn't really matter because everyone is going to have their own take and opinion on this.

For me, right now, I could be very happy with this camera because I mostly shoot 720 24pN in the HPX170 world and I find the amount of skewing in 720 24pN totally acceptable for my clients, I seriously doubt if any of them would ever even notice it. As a matter of fact, this would be an improvement for my clients who own the EX1, this camera has less skew at 720, of that I am sure. Everyone seems obsessed with 1080 24pN, which is fine, even though my personal opinion is that we could do a double blind test with the same monitor and 90% of people wouldn't be able to tell the difference between 720 24pN and 1080 24pN from a resolution standpoint, but that is just my personal view. 1080 is really most important for those of us lucky enough to have our work projected on the big screen but for most consumer televisions, 720 origination looks about the same or even slightly better than 1080. In other words, 1080 has always been overrated for many usages and for many viewers. Many of us just can't accept that something with a smaller number could be as good or even better than something with a larger number.

Dan
 
puredrifting said:
In other words, 1080 has always been overrated for many usages and for many viewers. Many of us just can't accept that something with a smaller number could be as good or even better than something with a larger number.

Dan
Click to expand...

On what finding is the logic based. Pl enlighten my soul. :Drogar-Love(DBG):
 
We just don't see much if any difference between 720/24p and 1080/24p with our HPX170.

No charts, or anything like that -- just viewing it.
 
Jan_Crittenden said:
Do you have any idea what the cost is of a 1920 X1080 sensor that outputs 60 frames?

You need to work for a manufacturer for a while. If we even had a chip set, then there is the rest. Everyone thinks that R & D is inexpensive. For perspective, the DVX100B chipset cost $5,000,000 to develop. This is not an inexpensive business to develop product for, we as do all manufacturers work hard. We do try to listen and try to deliver where we can.

Best,

Jan
Click to expand...

Jan... Jan... Jan! I was only keeeeeding! You are one SERIOUS person! I was actually agreeing wid yew, but in a goofy way. Of course you can make a 3 2/3" CCD camera and sell it for $5,000, you could sell them for $1 if you wanted, but neither would be even remotely profitable. Man, when I have to explain my own silliness, ha!
 
Big Brother said:
On what finding is the logic based. Pl enlighten my soul. :Drogar-Love(DBG):
Click to expand...

Adam Wilt for one. He knows more about video, resolution and HD than you and I put together probably http://provideocoalition.com/index.php/awilt/story/how_important_is_resolution/

I have also had this discussion with several video engineers whose opinions I respect.

And then there is my own usage. I deliver programming to all kinds of entities, from small shows for corporate clients that will be seen as web clips all of the way to broadcast clients and all of the studios. I was the DP on a 90 minute broadcast doc for TCM called, "The Dawn of Sound: How Movies Learned to Talk", I have delivered several A&E Biographies and have delivered hundreds of projects to all of the major studios.

1080 is overrated for most users, not all, but most. Sure, if you are shooting Planet Earth, 1080 is nice, although, oh yeah, most of Planet Earth was photographed in 720 on the Varicam. Hmm... Well, most of the networks require 1080 delivery, don't they? Well, um except for these little outfits called ABC and FOX, sure. ABC and Fox, yes, they like 720.

Video and film people (and I am one in case you were wondering) tend to think that bigger is always better. A 103" Plasma must be better than that measly 42" plasma you have, my native 1080 camera must be better than that 720 camera you are shooting with, etc. I simply disagree. The only place I have seen a noticeable difference where 1080 is a definite practical (not theoretical) advantage is large screen projection, either digital or film. Then it matters. But for broadcast, delivery to the studios and the work that I do, 1080 is fine but so is 720. Even when I have to deliver 1080 masters, usually most of the source material sitting on my timeline is 720. The Aja Kona 3 does a beautiful real time uprez that all of the QC and studios have been happy with for going on three years.

Dan
 
You must log in or register to reply here.
Back
Top