Videoshader and colour blind people

[K.I.L.E.R]

AFAIK VS works by changing certain colour values of certain pixels. That is how it deblocks videos.

What happens to people if they are colour blind?

Wouldn't that mean that colourblind people would not see the effects of deblocking or it wouldn't look like it's deblocking effectively?

When I view DivX videos they look extremely good with VS, but others say it doesn't make much of a difference.

I have taken a colourblind/vision test and I have perfect vision and I am not even slightly colourblind.

This makes me believe that VS is less effective if it's even effective on colourblind people.

 

[Captain Chickenpants]

The colorblindness would apply equally to the actual video, so their should be no difference for people who are color blind.

 

[Dio]

It is an interesting point, though.

Many a time I've coded RGB -> intensity calculations and automatically used the 0.6 / 0.3 / 0.1 formula (or some variant) without wondering what difference there is if someone's got a duff set of receptors...

 

[Ailuros]

The colorblindness would apply equally to the actual video, so their should be no difference for people who are color blind.

ROFL (no I'm not laughing with certain people's misfortune; rather the obviousness of the above reply).

 

[Hanners]

You mean ATi's crappy drivers haven't fixed colour blindness yet? Will they ever fix the important issues?

 

[Dio]

Actually I'm not sure if he has a point or not. As I inferred above, this kind of correction is (probably) performed based on some type of intensity calculation. That's going to be wrong for colourblind people.

Whether or not the video is 'equivalently wrong' or would be better with some kind of differential correction, I can't work out in my head right now...

 

[PrzemKo]

Since there are different "versions" of colour blindness (these people don't see R, G, B, RG, RB or GB colours), best solution for them could be to use only grayscale modes. This way they'll see at least correct brightness levels. Other thing is, that such images will look somewhat different to seen by them in actual world or TV, so IMO the best thing here is just to do nothing...

 

[Dio]

But is the perception of greyscales by colourblind people the same as that by non-colourblind people?

 

[Captain Chickenpants]

The filtering will be done on the Y, Cr and Cb planes individually and then they are merged before display (automatically merged when using overlays).
As colorblindness presumably affects the R,G,B color space perception then the actual relationship between changes in the YCrCb domain is probably not easy to predict, but as the colors would be blended to look natural to non-color blind people, then people affected with color-blindness whose perceptions are used to this should see a comparable affect.

 

[K.I.L.E.R]

The filtering will be done on the Y, Cr and Cb planes individually and then they are merged before display (automatically merged when using overlays).
As colorblindness presumably affects the R,G,B color space perception then the actual relationship between changes in the YCrCb domain is probably not easy to predict, but as the colors would be blended to look natural to non-color blind people, then people affected with color-blindness whose perceptions are used to this should see a comparable affect.

However minor quality is lost.
Thanks

 

[Dio]

Makes sense.