4:2:2 vs 4:2:0 (with examples)

EDV

EDV

Well-known member
I know this has been debated before, but it seems to me few people realize how superior 4:2:2 color can be.

There are many other examples but lets have a look at this one:

http://www.youtube.com/watch?v=0z0xFk6Ca7E#

Those colors are insane! (especially after 3:00)

shot with the HPX170/Letus Ultimate and Nikon primes.

now have a look at a 7D clip:

http://www.youtube.com/watch?v=aiWy8dga4kg

never mind the cool effects, take a look at the 4:2:0 color.

No comparison! And this is a very well lit video!

It kind of reminds me of the HDV look.

Granted, they both look good enough for HD broadcast but...







Keep in mind these are compressed JPEGs...
 
Last edited:
There is no doubt that there is better color on the first video vs. the Satin Dolz but I'm not sure what I'm seeing I would blame on the 7d. I shoot 7d footage and HMC 150 footage, both in less than 4:2:2 but depending on my lighting either produced better footage than the Lola footage. I can put my raw footage up on a 42" HDMI screen and it just blows the socks off anything remotely like it. But then again, I have some color work that was marginal, and it takes my footage down to semi-pro. So I don't know what to make of the Lola footage. The production values are very good. Actually I think they are higher than the first one. I'm sort of tired of seeing poorly lit video pretending it's film. I can always tell the difference right away, as they look a bit like 50s film without really looking like 50s film. I think that the first video would actually have been helped by dropping the color entirely. would have seemed more real to me. Probably Lola would have also. Sometimes color just doesn't cut it. Give it a try. Take out your 4:2:2 color and see what you get. Might be surprised! But I thought both were clever. The first one seemed a bit 'overacted". The Lola video obviously was supposed to be a bit 'over the top' 40s piece. But nice work anyway! We are picking nits here. You did a good job on both as filmmaker!
 
People often seem to confused color space and color rendition. Not the same thing at all and one doesn't have anything to do with another. Color rendition is programmed into the camera and is usually characteristic of the company that produces the camera. How the color comes out has nothing to do with what color space it was shot. You can take a DVX and under identical lighting condition it would probably produce very similar color because Panasonic tends to have their camera produce more saturated color tone versus say, Canon. It's undeniable that it's better to shoot at 4:2:2 versus 4:2:0 for a multitude of reason (better ability to CC, chromakeying) but it's misleading to claim that the footage looks better because it was shot at 4:2:2.
 
Last edited:
Huy Vu said:
You can take a DVX and under identical lighting condition it would probably produce very similar color because Panasonic tends to have their camera produce more saturated color tone versus say, Canon.
Click to expand...
Similar, but still inferior, since one is SD and the other one HD. But yes, there are other factors involved, such as the Nikon Prime lenses they used. The Canon example would have used pretty good lenses too, so no unfair advantage.
 
The point is that it's an apple and orange comparison to start with since color space really has nothing to do with how the color comes out.
 
8string said:
I'm sort of tired of seeing poorly lit video pretending it's film.
Click to expand...

I hope you weren't referring to Gordon Evolution. I was the DP on that film and I think my crew did one hell of a job lighting that, especially since it was a film competition piece. Ryan Wood and I had 50 hours to come up with a script and another 50 hours for production and post production.

Also in case anyone is wondering, 99% of the color was achieved in camera. We only spent 10 minutes color correcting tops.
 
Funny thing. Color point, well taken. That said....this is QUITE a video on a 1700usd body. I don't own a 7d, but if I did....this would certainly VALIDATE my meager investment. NO...owning one doesn't mean you can shoot this. No....the 1700 body is NOT the main cost here. YES...they could have used anyone of 50 cameras to shoot this...it is only web afterall. That said...your point WAS NOT how great the 7d held up. You point was the superiority of the 4:2:2 color space compared to the 4:2:0 color space. Point taken. Thanks for sharing!
 
Whatever the source was, it's 4:2:0 after YouTube re-encodes it (rather brutally). For all practical purposes, you can get footage out of a cheap consumer HDV camcorder (in the right hands) that will look just as good as footage originating from way, way better cameras - after it's been thoroughly stomped on by YouTube's crude re-encoding.
 
Don't get me wrong. YouTube is great for being able to share video, but it's absolutely, utterly worthless for any sort of production quality comparison/analysis.
 
I know what you're saying about YouTubes' rather crude encoding, but both of those videos looked pretty clean to me, inspite of that. Even the screen captures, which are again brutally compressed JPEGS, look pretty damn good to my eyes.
 
Not sure if it was the way it was lit or what, but I like the colors in the Panasonic example I posted better...
 
These are two different types of productions that have different grading requirements. This is a really poor comparison on a myriad of levels.
 
Huy Vu said:
People often seem to confused color space and color rendition. Not the same thing at all and one doesn't have anything to do with another. Color rendition is programmed into the camera and is usually characteristic of the company that produces the camera. How the color comes out has nothing to do with what color space it was shot. You can take a DVX and under identical lighting condition it would probably produce very similar color because Panasonic tends to have their camera produce more saturated color tone versus say, Canon. It's undeniable that it's better to shoot at 4:2:2 versus 4:2:0 for a multitude of reason (better ability to CC, chromakeying) but it's misleading to claim that the footage looks better because it was shot at 4:2:2.
Click to expand...

Well, be careful not to confuse color space with color sampling, either. You're describing color sampling here.
 
If you properly expose the DSLR's I think the color is okay and cuts unnoticeably between HPX and 7D/t2i/5D... I just finished a shoot with an HPX and a t2i, the difference in the colors you get depends on who is holding the camera. When I (DP of the film) was holding the t2i, the colors looked identical between the two cameras, and you couldn't tell a difference when projected. When I was on the HPX and my AC was on the t2i, none of it was near usable next to the awesome color of the 170, the slightest overexposure/underexposure really wreaks havoc on your colors. The biggest difference between 4:2:2 and 4:2:0 when you work with it is how many of your shots you throw away due to operator errors with exposure (which will happen to any DP without the proper monitoring tools).

With the HPX and 4:2:2 you get great colors, and the tools to properly monitor your exposure. But when the time comes that you overexposed a run and gun shot, its no big deal just adjust the colors in post.
 
Great point David. However, that falls more into the discussion of using the right tools for the job I think lol. I was thinking more just general narrative shooting to compare the colors. If you don't know the difference amongst your tools when pulling a key, you shouldn't be doing anything in green screen.
 
Both are well done clips visually (besides the acting/directing in the first one)... but the first one still looks like video. I don't know what it is about the DSLR's that make the so film like, but it's awesome. They have worse compression, worse colorspace, but damn it looks filimc. The lighting was good in the first clip with the HPX, but I just don't think it has the shadow-sensitivity/range that the 7D has... and the highlight roll-off. You can argue all you want that an HPX with 4:2:2 color has more dynamic range (which is true if you're talking about the range of in-between luma values) but there is just no doubt that you can physically expose for a bigger range of light with the 5D/7D.
 
You must log in or register to reply here.
Back
Top