Is HVX2OO really 4:2:2 or actually not?

bbalser

bbalser

Veteran
To quote someone talking about DVCPROHD not being true 4:2:2 colro space, "...but with it's small frame format, it's not at all 4:2:2. Taking a VHS signal and recording it to HDCAM can arguably be called 4:2:2/HDCAM format, but the resolution and signal aren't up to par with what the industry generally considers a true color sample. If it were, it would for instance, be acquiring 22 luma samples to every 11 blue and 11 red (minus luma) samples. But it's not. Do the math. It's truly more like 3:1.2:1.2 This is one of the reasons why it's a softer image, it's not recording full vertical or horizontal information to the small frame format of DVCProHD."

Is this true?
 
There's a saying in legal circles that says "when the law's on their side, argue the facts; when the facts are on their side, argue the law."

Why not just look at the pictures?

The guys from Peter Jackson's WETA Digital put 4:2:0 XDCAM HD from a $35,000 XDCAM HD setup through some chroma keying tests. And they also put the 4:2:2 HVX200 through those same tests. The HVX footage produced a better key than the $35,000 XDCAM footage did. And this is from some of the best chroma key experts in the world. I'd say that answers it definitively.

But heck, you could also just look at Jim Arthurs' work, or Anders Holck's work, and see the results.

DVCPRO-HD is absolutely a 4:2:2 format.

The question is whether the HVX is delivering enough resolution to fill the recording format's capabilities. There's not much of a question that the HVX isn't delivering a full 1920x1080 image with discrete resolved pixel detail for every one of the 1920x1080 pixels in a full HD frame, but does it deliver enough to fill out the recording format, enough to qualify as 4:2:2? We can debate statistics and theoretics all day long, or we can look at the actual results.

If someone wants to play word games and claim 3:1.5:1.5 or 3:1.2:1.2 or whatever, what does that say about XDCAM HD or other 4:2:0 systems, since the HVX outperforms them? Especially with the XDCAM's sharper resolution and bigger chips? The only rational conclusion is that the superior color sampling of DVCPRO-HD is making the difference, and that the HVX is obviously delivering that color information to the codec (or we wouldn't see the real-world results).

Oh, and about that stupidity about 3:1.5:1.5 or whatnot: are they prepared to debate Charles Poynton or Graeme Nattress?
http://poynton.com/PDFs/Chroma_subsampling_notation.pdf

4 is the leading digit. The reason "3" ever entered the equation was because of Sony's HDCAM, which employs 3:1:1 color sampling. It has nothing to do with the prefiltering of the recording format, it refers to the ratio of color samples to luma samples. Unfortunately some people insist that the leading "3" of HDCAM has to do with the prefiltering from 1920x1080 down to 1440x1080 for recording, so they think that the "3" in 3:1:1 comes from the fact that 1440 is 75% of 1920, and 3 is 75% of 4, so therefore the 3 references the prefiltering. It doesn't. It references that there are three luma samples for every one chroma sample.

I've never heard of 3:1.2:1.2. That's a new one, someone's obviously playing some sort of word games in an attempt to discredit the format. They obviously don't know what they're talking about.

Look at the footage. Look at the results that people (who know what they're doing) are getting.
http://www.dvxuser.com/V6/showthread.php?t=61098
http://www.dvxuser.com/V6/showthread.php?t=68217
http://hvx200.moctel.com/Users/Anders_Holck/2_carlas_kaptajn_amerika_hi.mov
http://www.imageshoppe.com/HVX_bluescreen2.html

Especially read that last one, as Jim takes us step-by-step through the process, and flat-out declares that "Bluntly put, the results were spectacular, and ground-breaking for a camera in this price range."
 
Barry_Green said:
http://poynton.com/PDFs/Chroma_subsampling_notation.pdf

4 is the leading digit. The reason "3" ever entered the equation was because of Sony's HDCAM, which employs 3:1:1 color sampling. It has nothing to do with the prefiltering of the recording format, it refers to the ratio of color samples to luma samples. Unfortunately some people insist that the leading "3" of HDCAM has to do with the prefiltering from 1920x1080 down to 1440x1080 for recording, so they think that the "3" in 3:1:1 comes from the fact that 1440 is 75% of 1920, and 3 is 75% of 4, so therefore the 3 references the prefiltering. It doesn't. It references that there are three luma samples for every one chroma sample.

I've never heard of 3:1.2:1.2. That's a new one, someone's obviously playing some sort of word games in an attempt to discredit the format. They obviously don't know what they're talking about.
Click to expand...

I agree 100%... its amazing how many stupid arguments one can justify with math. If four people get killed by buzzards one year, and three of those people happened to be over 6 ft tall, you can be sure somebody out there will claim that people over 6 ft tall have a 75% higher chance of being killed by buzzards than shorter people...

You can state 4:2:2 in many ways, and as long as you keep the ratio the same, it is still technically accurate. But that doesn't make it correct. Just because somebody wants to be silly and call the HVX's color space 2:1:1 or or 0.0625:0.03125:0.03125 only proves that the person arguing doesn't know what they're talking about. And changing what the ratios actually mean in your argument makes things even worse...
 
Hey, I'm 6'1"! Does that mean I should stay the hell away from snow?

Lord, some people's children. Yes, the poster I saw did strike me as playing a very slick word game. But you guys are way over my head on all the technical stuff. So I figured y'all would set me straight.

DVXuser.com Rocks, y'all!
 
You must log in or register to reply here.
Back
Top