color problems

I was making a project and I had some problems with the final result in flv format. I used ramp effect and the color wasnt linear, but it was making “color levels”. I used too an item in perfect red color and it was destroyed, when i changed the color, it was ok. Is it a “compression” problem?

I believe you are referring to ‘banding’ in the ramp.

Try 16-bit and/or add a bit of noise. I’ve also heard adding some extra colors (perhaps a bit of 4-color gradient) can help.

Don’t know what you mean by red being “destroyed.”

Compression problems. flv format in most cases “destroy” the red color, and the ramp isn’t smooth.

Generator said
I believe you are referring to ‘banding’ in the ramp.

Try 16-bit and/or add a bit of noise. I’ve also heard adding some extra colors (perhaps a bit of 4-color gradient) can help.

Don’t know what you mean by red being “destroyed.”

yes I mean banding, I couldnt find the right word.

videologio said

Generator said
I believe you are referring to ‘banding’ in the ramp.

Try 16-bit and/or add a bit of noise. I’ve also heard adding some extra colors (perhaps a bit of 4-color gradient) can help.

Don’t know what you mean by red being “destroyed.”

yes I mean banding, I couldnt find the right word.

http://greyscalegorilla.com/blog/2009/10/how-to-remove-banding-artifacts-in-after-effects/ watch this video. It will help you.

Compression can work in many ways, but one of the chief ways that it makes data saving is to split the image into its luminance and chrominance components, and then break the chrominance component into large block at low bit depth. This is because the eye is much more sensitive to luminance than chrominance. Similar principle to tinting a black and white photograph.

When the image is put back together, then lack of resolution in the chrominance channels can cause banding.

Because luminance is largely created from the green component of an RGB image, the greens in an image rarely look compressed. The red and the blue channel by comparison are heavily compressed. Our eyes are more sensitive to red than blue, so if you’re going to have a problem with any color in compression, then it’s going to be red.

It’s quite possible that some compression systems turn the red saturation down a little to overcome these artefacts.

The other big color issue to look out for with compression is the gamma level. Devices that measure light (i.e. camera / human eye) and those that emit light (i.e. computer / tv screen) rarely have a linear response to light. Double the light hitting the chip in a camera won’t create double the electrical signal, double the light to the eye won’t stimulate the nerve cells doubly and create something perceptually twice as bright. etc.

That means that each device in a production chain needs to be gamma managed (i.e. the response curve compensated for) in order that it looks right. If the image looks too dark and shadowy or too washed out, then it’s almost certainly the gamma level that’s at fault.

-f.

markoO said

videologio said

Generator said
I believe you are referring to ‘banding’ in the ramp.

Try 16-bit and/or add a bit of noise. I’ve also heard adding some extra colors (perhaps a bit of 4-color gradient) can help.

Don’t know what you mean by red being “destroyed.”

yes I mean banding, I couldnt find the right word.

http://greyscalegorilla.com/blog/2009/10/how-to-remove-banding-artifacts-in-after-effects/ watch this video. It will help you.

thanks, great video!

felt_tips said
Compression can work in many ways, but one of the chief ways that it makes data saving is to split the image into its luminance and chrominance components, and then break the chrominance component into large block at low bit depth. This is because the eye is much more sensitive to luminance than chrominance. Similar principle to tinting a black and white photograph.

When the image is put back together, then lack of resolution in the chrominance channels can cause banding.

Because luminance is largely created from the green component of an RGB image, the greens in an image rarely look compressed. The red and the blue channel by comparison are heavily compressed. Our eyes are more sensitive to red than blue, so if you’re going to have a problem with any color in compression, then it’s going to be red.

It’s quite possible that some compression systems turn the red saturation down a little to overcome these artefacts.

The other big color issue to look out for with compression is the gamma level. Devices that measure light (i.e. camera / human eye) and those that emit light (i.e. computer / tv screen) rarely have a linear response to light. Double the light hitting the chip in a camera won’t create double the electrical signal, double the light to the eye won’t stimulate the nerve cells doubly and create something perceptually twice as bright. etc.

That means that each device in a production chain needs to be gamma managed (i.e. the response curve compensated for) in order that it looks right. If the image looks too dark and shadowy or too washed out, then it’s almost certainly the gamma level that’s at fault.

-f.

thanks felt_tips!