Color Grading with the A1.

Brandon Rice

Brandon Rice

Distributed Producer
Ok, so I've read the in depth article on colorspace that was recently posted here... I've come to the conclusion that 4:2:2 picks up better color, and is also able to be pushed around a lot more in post. Am I wrong in thinking that? I know that the A1 is 4:2:0, which by my conclusion from reading the article isn't as good at capturing true color, in fact, by the charts, it's nearly the same as regular DV 4:1:1. Please enlighten me.
 
There is as much info is 4:2:0 as there is in 4:1:1 (not much)

Yes, 4:2:2 is much better in post.

I have heard, however, that when you down convert HD, you get the equivalent of 4:4:4 SD color space but I am not sure how this works.
 
http://www.dvxuser.com/V6/showthread.php?t=77926

Perfectly good example of just how well the 4:2:0 holds up. And it's much better than 4:1:1.

Might capture more color info. But it's no good if the camera is noisey.

Hence why people shoot 1080p or DVCpro50 for Green Screen work. Because 720p, regardless of all that "color", is too noisey to work with. But, I already said that before. =P This is for those who want to ask the same question.
 
Yeah, it's really more along the lines of actually seeing narrative footage from this camera, and seeing the specific looks that can be achieved in post production.
 
You shot with the DVX100 right? Ever find the color a problem? The DVX100 is no 4:2:2 either, its 4:1:1 Most like the DVX color and the possibilities in the editing.

The PAL DVX100 is a 4:2:0 system too. Never heard of problems but for multi generation work you better use another intermediate. The HVX200 gives 4:2:2 color that is the best of the bunch (its a more professional camcorder). Great for bluescreen work etc.

But in practice, you can go a long way with the HDV. BTW: the Canon A1 uses 4:2:2: color inside before the codec. So everything you change in color in the camera is in the 4:2:2: demain!
 
Last edited:
Bradon, try downloading some of the various XH A1 native M2T files available. If you search around at dvinfo there should be quite a few from Kaku Ito. Load em up into your favorite app and tweak them around for yourself.

I did this with some of the 24F clips that were posted and had no problems at all. Even the night time footage had a lot of shadow detail I could bring out without any weird compression blocks or issues. Running tests on that footage actually helped me to decide that the XH A1 would work for me.
 
Then Ill throw this in the mix. Effectively 4:2:0 and 4:1:1 is the same number of colour samples. But they work in a very different way.

Think of it in these terms

4:1:1

Cy Cb | Y | Y | Y |
Cy Cb | Y | Y | Y |

4:2:0
Cy | Y | Cy | Y |
Cb | Y | Cb | Y |

As you can see 4:1:1 picks up twice as much chroma info in the first block, but then ignores it totally for 3.

4:2:0

As you can see picks up half the chroma then luminence, then chroma again then luminence.

The end result is the same compression ratio. But I would suggest the continual reference to chroma by 4:2:0 is more accurate. And certainly HD broadcasters are feeling the same making more effective 4:2:0 specifiic decoders for just about all HD broadcasts. Which are of course in the mpeg2 format.
 
ncje said:
Then Ill throw this in the mix. Effectively 4:2:0 and 4:1:1 is the same number of colour samples. But they work in a very different way.

Think of it in these terms

4:1:1

Cy Cb | Y | Y | Y |
Cy Cb | Y | Y | Y |

4:2:0
Cy | Y | Cy | Y |
Cb | Y | Cb | Y |

As you can see 4:1:1 picks up twice as much chroma info in the first block, but then ignores it totally for 3.

4:2:0

As you can see picks up half the chroma then luminence, then chroma again then luminence.

The end result is the same compression ratio. But I would suggest the continual reference to chroma by 4:2:0 is more accurate. And certainly HD broadcasters are feeling the same making more effective 4:2:0 specifiic decoders for just about all HD broadcasts. Which are of course in the mpeg2 format.
Click to expand...

So in essence 4:2:0 is more balanced. Correct?
 
philnerd said:
Bradon, try downloading some of the various XH A1 native M2T files available. If you search around at dvinfo there should be quite a few from Kaku Ito. Load em up into your favorite app and tweak them around for yourself.

I did this with some of the 24F clips that were posted and had no problems at all. Even the night time footage had a lot of shadow detail I could bring out without any weird compression blocks or issues. Running tests on that footage actually helped me to decide that the XH A1 would work for me.
Click to expand...

That's exactly my point with the A1. Compression shows up when you go from dark to light in extremes, but nowhere near as bad as when you go from dark to light with the HVX footage. 4:2:2 or not.
 
You must log in or register to reply here.
Back
Top