Has P2 been completely murdered by KxS?

[Noel Evans]

Joe, should have sold the HVX and got yourself an HPX500 if size isnt an issue for you?

To weigh in, SxS and the fact you can use SDHC cards is big factor IMO when new buyers are going to start to compare should I buy x or y. Forget if its the right way or not, they will and no Im not talking about users on this forum - there are millions of users out there that arent registered on here or any other forum. Id say an absolute majority of shooters dont use a forum like this to ask questions.

Anyway if one of the people do pass this forum, consider this point please. I use a mac pro, that whizzes through anything I throw at it. Its fast and efficient. Im hitting around 200mbps on my editing drives, have 6gb of ram and just dropped a new vid card in there.

I have used this machine to edit many different codecs from many different cameras. Ive shot and edited ex footage a number of times as well. And if you remove yourself entirely from every other consideration for a moment, and Im not sure of the exact numbers as I dont time these things --- wait a second -- OK I actually did just time it.

EX 1080/25p clip - 5 seconds duration - applied 3 filters (from Tiffen DFX) I know are notoriously horrible for render times on any footage.
Total render time for EX footage 1.55.

I then took this same 5 second EX clip and converted to dvcprohd 1080/25p clip. I then created a new 1080/25p dvcprohd sequence and imported. Then applied the same 3 render hungry filters exactly as they were on the ex timeline (I copy and pasted them to be sure nothing changed).
Total render time was 1.17.

So, in conclusion the DVCProHD codec rendered the same clip 38 seconds faster. Doesnt seem like a huge difference but lets assume the clip was 10 minutes long.

EX codec render time - 230 minutes
DVCProHD render time - 154 minutes.

36 minutes saved.

Can I reiterate the 3 filters I applied are disgusting on render times, and applied all over the top of each other, I can only imagine what is required to render it.

And there is another side to this coin. On DVCProHD, I can do many things and run RT previews whereas the EX footage will only output something like 2fps so to guage it I need to render.

Here I am not talking about what camera can do what (I mean cmon, if your that hung up on specs to think say the 170 or an ex arent going to do a good job, well you probably need to shoot more), I concerned myself trying to guage the REAL cost of one codec over another. And in real terms, if your paying an editor - I can only imagine how much more editing the EX codec over a full year will cost you - $20-30k over p2? Lots of factors effect that, are they on staff or freelance etc. Then factor in if you edit your own footage - how many jobs cant you do because your edit time is longer? Probably many.

Now, if your a person on the edge, maybe shoots one job a month, or someone that sees this as a hobby and are look at camera vs price - WHO CARES? My point is from a business standpoint well the numbers are there. From a personal standpoint where numbers dont matter etc

 

[RCFisher]

Not trying to add fuel to the fire, I purchased a EX1 because I was looking for better low light performance and resolution. I loved the look of the HVX200 even projected on a motion picture screen it looks awesome despite it's poor low light performance. So far the EX1 has done a great job. Yesterday I shot 4 hours of interviews on my new Delkin/Transcend setup, I guess you call it DxT. They performed flawlessly. The price for shooting long form projects can't be beat. I have 5 hours of media that cost me about $350 including adapters the same recording time would have cost me almost as much as the camera using SxS cards. Most of my shooting is at 24FPS so I really don't need the SxS cards speed most of the time. The XxX cards have been just awesome so I am happy, now if I can get a picture profile that emulates the look of HVX that's even better.

 

[PerroneFord]

Noel brings up an interesting point. And even though I work on my footage myself, it's a good one. When I first started working with HD footage (not my own) I was cutting and grading HDV. It was pretty slow on my old machine. I used two methods to deal with it.

1. Cut Proxies for everything. Though it was a bit cumbersome, eventually, I learned that I could essentially batch cut proxies overnight of all footage, and the next morning I could simply edit the proxies just like I could miniDV. Worked GREAT even if it was at a smaller size.

2. Cineform. Halfway through my first monthlong project I started using Cineform. This allowed me to work with full sized footage with no real visual loss. And it cut and rendered FAST.

Now that I have the EX1, I typically work in Cineform as my intermediate. Yes, it's an extra step. And perhaps one I wouldn't have to take were I using a more modern machine or cutting on DVCProHD.

But looking at how Hollywood seems to work, the idea of cutting on proxies seems well established, so I don't think my workflow is too far from the mainstream. If I was passing things off to an editor, I think I'd be pretty happy using either methodology.

 

[Buck Forester]

The longer record times are important to me because whether I'm shooting my kids at the S.F. Zoo or Monterey Bay Aquarium or on a multi-day trip in the wilderness, I don't want to have to dump my cards on my laptop every 15 minutes. I can go a long time before I have to do a harddrive dump with the EX1. Plus this is all in 1920x1080 full rez, which is mandatory for what I'm doing. I'm sure the intraframe would be nice in some situations, but it's not practical to have such short record times for what I'm doing, especially not having the resolution specs to boot. That said, like others have here too, I don't think it's a P2 killer because there are plenty of shooting styles and applications where P2 shines. I have spoken, you may all be seated.

 

[Green Hornet]

I think it COULD be a P2 killer if

1. Panasonic doesn't reduce the price for P2 cards.
2. Panasonic doesn't offer the full raster codec on a low end camera like the HPX170.
3. Pansonic doesn't offer it's own inexpensive P2 card reader.

The P2 cards are priced in a league of their own. It is a format where new laptops can not support it. Soon the lack of a method to offload will prevent a camera purchase at a higher rate than the price of the P2 card. Without a PCMCIA slot, you are screwed.

The pricing of P2 cards, to me, is like buying gas for $8 a gallon at one station, when everyone else is selling the same gallon for $2.00. And the $8 gas only comes in one gallon containers, that you have to stop and put in your tank one at a time.

I love the Panasonic cameras, and when I bought the HVX200 a few years ago, it addressed my needs,
Technology has moved on and offered similiar solutions with easier methods for offloading, capture at higher resolution.

 

[DanHD]

In my opinion, sony has taken a step back.

sure we're seeing longer recorder times and full raster, and it looks good and all

but long GOP? it's so highly compressed. Long gop is the way of the past. It may be more efficient but not worth the artifacting.

 

[Stevet]

In my opinion, sony has taken a step back.

sure we're seeing longer recorder times and full raster, and it looks good and all

but long GOP? it's so highly compressed. Long gop is the way of the past. It may be more efficient but not worth the artifacting.

It's this type of nonsense that's got some confused who do not know the extent of video knowledge that's many here understand.

Have you seen the fast motion 4:2:2 100mbps long-GOP examples that Jim Arthurs posted from an EX1 hooked to the new COnvergent Design XDR?
There were comparisons and layer subtractions with uncompressed that were outstanding.

Yes, it's true they were noticeably better than the stock EX XDCAM 35mbps, but only so under faster motion. Having said that, I've seen compression macroblocks in HVX200 DVCPRO-HD footage which was one of the reasons (inluding noise levels) I steered away from the HVX200.


I'd take this full 1920x1080 raster codec 4:2:2 100mbps long-GOP any day over DVCPRO-HD.

 

[brianluce]

I'm a new EX3 owner, before that I used HDV. The mantra for HDV world was "Just say NO to HDV". I used cineform, others I know used Canopus HQ. The work flow is something like this:
capture with the cineform module which flips the HDV to an IC>open in Vegas or whatever. Long GOP really isn't an issue for most HDV shooters in post. It's slice and dice once you're in an IC. So the whole pitfalls of GOP has always been a bit of a red herring as far as I'm concerned. And I expect the EX3 workflow to take a similar path.

 

[ChipG]

In my opinion, sony has taken a step back.

sure we're seeing longer recorder times and full raster, and it looks good and all

but long GOP? it's so highly compressed. Long gop is the way of the past. It may be more efficient but not worth the artifacting.

Dan's 19 he is here to learn...

 

[adamr316]

So, in conclusion the DVCProHD codec rendered the same clip 38 seconds faster. Doesnt seem like a huge difference but lets assume the clip was 10 minutes long.

EX codec render time - 230 minutes
DVCProHD render time - 154 minutes.

36 minutes saved.

Can I reiterate the 3 filters I applied are disgusting on render times, and applied all over the top of each other, I can only imagine what is required to render it.

And that math makes perfect sense considering this:

Divide 154 minutes/ 230 minutes = 0.669 (round up to 0.67) so 67%. The DVCPro HD footage rendered approximately 33% faster.

DVCPro HD at 1080/25p resolution is 1280x1080. XDCAM EX is 1920x1080. Guess what the differential in width resolution is? 1280/1920 = 0.66~ which is approximately 33% less resolution with the DVCPro HD format.

So...33% more resolution for 33% more rendering time. That's actually pretty impressive for a long GOP format vs. a (relatively) lightly compressed format that is smaller resolution.

A 33% increase in resolution is worth the extra time for me. Of course I'm the type who would do a 10 pass CBR encoding for DVDs to squeeze out every ounce of resolution or buy expensive/overpriced audio cable from Mogami with Neutrik connectors etc. ...but there's a lot of us out here. And 33% is substantial!

 

[Barry_Green]

Someone who looks at price alone will conclude that, sure. But you asked why anyone would continue with P2, and we're telling you why. Looking at price alone is a not a good way to make a choice.

But why would anyone, who looks at price, go with an EX1 instead of an HMC150? Long-gop 4:2:0 recording to solid state, but with one it costs $6500 and with the other it's $3400. With one you have to use a specific kensington adapter with certain pre-qualified forum-vetted memory cards, with the other you can slap any $10 SD card in there as long as it says "Class 4" on it somewhere.

So, at this point, someone will undoubtedly point out some irrelevant-to-the-recording media factor such as "EX1 is sharper" or "EX1 does variable frame rates" and at that point we're back to the crux of the same discussion, right? You can't separate the media from the camera and indict the media alone.

That, plus the fact that about 80% of the broadcasters going tapeless have chosen P2, and Fuji and Maxell and Hitachi are all making P2 equipment, would seem to imply that no "murdering" is going on.

I think the initial question is flawed based on the notion that one may sometimes project one's own opinion upon the rest of the world, seeing the whole world from their own perspective. It ain't so. For example, people clamor that 720p is "dead" and 1080 is the one and only true form of HD, right? As an experiment I went through my cable system to see what's being broadcast. I have the basic HD package, 28 channels broadcasting in HD. Of those 28 channels, 17 were 1080i, and 11 were 720P! Eleven! Who would have thought that in (almost) 2009, 40% of the channels broadcast by Time Warner Cable would be 720P?

Sometimes what people think, and what actually *is*, ain't the same thing.

There's a perception among some folks that Sony's EX1 is "murdering" the market. Today, Sony's electronics business laid off 16,000 people (8,000 full-time, 8,000 temporary workers) and announced that they were closing 10% of their factories, blaming lack of demand, and Sony chief Howard Stringer even said (paraphrased) that "it may hurt today, but if we don't take this step there won't be a tomorrow". These layoffs are of course spread across all of Sony, not just the (relatively tiny) broadcast division.

Perceptions can be altered by regional territories too; certain areas of the world are predominately sony or predominately panasonic, so residents there might say "all I see is XDCAM 330s everywhere I go, therefore Sony must have the market locked up." But just because Australia bought Betacam SX, that doesn't mean any other market in the world did.

Final thought on this is that our little corner of the market is not exactly a massive segment. If every DVXUser/DVInfo member with an EX1/EX3 used KxS, that might account for maybe 5% of the market. The rest of the world doesn't even know it exists. And I've got to believe the big section of the market, the broadcast buyers, are never going to be tempted to use something like KxS. If you ever see ABC or NBC or CBS or Fox or the BBC or anyone like that using KxS, let us know because frankly I'd be shocked.

Is KxS a lovely thing for the indie/prosumer guy? Sure. Is it going to make the slightest perceptible dent in either P2 or SxS sales? I extremely exceptionally doubt it. I mean, if HALF the people we get were swayed to make a decision based on KxS vs. SxS or P2, then you're talking about 1/2 of 5%, or less than 3%. Not enough to even pinch or slap the market, much less murder it.

 

[csundbom]

I'm a new EX3 owner, before that I used HDV. The mantra for HDV world was "Just say NO to HDV". I used cineform, others I know used Canopus HQ. The work flow is something like this:
capture with the cineform module which flips the HDV to an IC>open in Vegas or whatever. Long GOP really isn't an issue for most HDV shooters in post. It's slice and dice once you're in an IC. So the whole pitfalls of GOP has always been a bit of a red herring as far as I'm concerned. And I expect the EX3 workflow to take a similar path.

I fully agree. Long-GOP is an excellent method of compressing video data. I really don't understand the fuss about it. Today's solid state recording media are very reliable, and the danger of losing a frame that will make you lose a lot of footage is pretty remote.

Archiving is easier, and you save a lot of space (i.e. money)

Long-GOP is just a different and more efficient way of storing the same info as an I-frame-heavy codec. The main drawback is less detail with fast motion when you run out of bandwidth, but I never found that to be an issue. I guess you could pause a frame with fast action and detect some artifacts, but we are not showing stills; we are showing motion.

 

[ChipG]

Someone who looks at price alone will conclude that, sure. But you asked why anyone would continue with P2, and we're telling you why. Looking at price alone is a not a good way to make a choice.

Dave,

I made a decision to use P2 (1 16 gig card per camera) on a show I produced this spring, it was VERY tough, but I did it wanting 4:2:2 dvcpro and am glad I did.

Here is a previous post I made about the situation. And no I don't like the cost of P2 and will always feel that way

http://www.dvxuser.com/V6/showpost.php?p=1489567&postcount=12

 

[ChipG]

All the effort was worth it.

 

[brianluce]

There's a perception among some folks that Sony's EX1 is "murdering" the market. Today, Sony's electronics business laid off 16,000 people (8,000 full-time, 8,000 temporary workers) and announced that they were closing 10% of their factories, blaming lack of demand, .

Whoa, hang on a sec, if Sony were in fact "murdering" P2, and I don't know if they are or aren't, but let's say they were by a 10:1 margin, Sony would still have laid off 16,000 people because as you say, we're talking about a niche in a huge market and Sony seems to produce everything from video cameras to electric toothpicks -- so those layoffs cannot be seen as remotely relevant to EX market share. Btw, I'm dealing with the "Try and find the best third party media" for your EX circus and it's completely neurotic -- panicky emails to suppliers in Tasmania to see if they have any Kensington readers left and Uh oh! how about that new Indonesian made reader on Ebay! Anyone tried it? What? Only reads up to 34FPS and you need a peice of chewing gum to hold it in? No network on the planet would sign on that insanity.

 

[ChipG]

Sony needs to trim a lot of fat. Other companies will be announcing similar layoff's, I know from being a share holder. Not a big one but I can read financials.

EDIT: can't spell worth a crap but I can read spreadsheets.

 

[Joe Lawry]

Joe, should have sold the HVX and got yourself an HPX500 if size isnt an issue for you?

Size isnt an issue, i love broadcast cameras. However at 21 years old my current client base are those who want to shoot on small cameras. Not for budget reasons, but for the style of shows we shoot.

And when your clients start specifically asking for the EX1 by name.. then i have to make a decision about whether i want to keep getting the work, it was as simple as that.

And unforunately.. NZ is a very XDCam based country.. so if i was going to buy a full sized broadcast camera, it would have to be a PDW700 as that is becoming the standard here. The 500 is a great camera, but i've yet to find a client in Auckland asking for it. HVX on the other hand.. all the time, well it use to be. Hence me not selling my Panny.

 

[mrjazz]

EX 1080/25p clip - 5 seconds duration - applied 3 filters (from Tiffen DFX) I know are notoriously horrible for render times on any footage.
Total render time for EX footage 1.55.

I then took this same 5 second EX clip and converted to dvcprohd 1080/25p clip. I then created a new 1080/25p dvcprohd sequence and imported. Then applied the same 3 render hungry filters exactly as they were on the ex timeline (I copy and pasted them to be sure nothing changed).
Total render time was 1.17.

XDCAM-EX has 1920x1080 pixels. DVCPRO-HD has 1280x1080.
So if you have to render only 66% of the pixels it isn't remarkable that it takes just 66% of the time, is it?

 

[lawriejaffa]

Mr Jazz ain't that simple... and totally innaccurate way of comparing the two.

For a start, the long GoP on the SxS is utterly compressed, so it fits a higher visual resolution in terms of 'resolution' of one sort - with less colour information.

The HD Dvc Pro records also at 100mb per bit sec while the SxS is 35 (correct me if im wrong anyone as im no tech head lol.)

What your focussing on your comparison there is telling - this visual resolution, its all about the pixels haha, but folk forget how they get there, and how the quality of what pixels you have can vary. Important for determining what you need your camera to do too. Both cams produce 1080 and end up with as many pixels in their final image, but sensors derive those images differently.

 

[mrjazz]

Mr Jazz ain't that simple... and totally innaccurate way of comparing the two.

I didn't compare anything, I just explained where the difference in rendering-time comes from. (Your NLE doesn't care of 4:2:0 vs. 4:2:2)

For a start, the long GoP on the SxS is utterly compressed, so it fits a higher visual resolution in terms of 'resolution' of one sort - with less colour information.
The HD Dvc Pro records also at 100mb per bit sec while the SxS is 35 (correct me if im wrong anyone as im no tech head lol.)

I would prefer to call ex "highly efficient" compressed.
Yes, DVCPRO-HD has a little more color information (640x1080=691200 chroma-pixels of DVCPRO-HD vs. 960x540=518400 of EX), but those extra-chroma-pixels are useless, if you can't fill them, because your cameras sensors only have 960x540 pixels.
Btw, here's a comparison of chroma-information recorded by the ex1 and hvx:
http://www.barrhauss.com/projets/repaire/ex-vs-200/
This comparison shows clearly that ex1-35mbs-footage has more effective chroma-resolution than hvx-100mbs-footage, so the hvx can not benefit from the higher chroma-pixel-count of DVCPRO-HD.

What your focussing on your comparison there is telling - this visual resolution, its all about the pixels haha, but folk forget how they get there, and how the quality of what pixels you have can vary.

hvx-folks should better forget how the pixels get there, because it's an ugly dirty story.
Yes, ex-pixels can vary because it's an interframe-codec, but in the end experience shows just very high-quality in all situations. That's what's important.

Important for determining what you need your camera to do too. Both cams produce 1080 and end up with as many pixels in their final image, but sensors derive those images differently.

Much more important than pixel-counts is measured resolution in terms of "tv lines". And needless to say, the ex1 scores the hvx out in that domain.

But you're right, it's all about what we're doing with these cams. Let's forget these technical details!