Yahoo Groups archive

Digital BW, The Print

Index last updated: 2026-04-28 22:56 UTC

Message

Re: [Digital BW] Understanding channel mixer

2005-02-17 by Steve Kale

Thanks Paul.  I'm slowly getting it.  It's understanding the greater than
100% and -% that needed explaining. Cheers, Steve


> From: Paul Williamson <kb5mu@...>
> Reply-To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
> Date: Thu, 17 Feb 2005 12:19:20 -0800
> To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
> Subject: Re: [Digital BW] Understanding channel mixer
> 
> 
>>  I am still not conceptually getting
>> what is happening when, with monochrome checked, a channel is set to greater
>> than 100% or at a negative value...
> 
> I think you're making it more complicated than it really is. The
> Channel Mixer just does addition.
> 
> Remember that each of R, G, B for each pixel is just a number, 0 to
> 255 (in 8-bit mode). Channel Mixer in monochrome mode just takes each
> number, multiplies it by the percentage set, and adds them all up to
> get the luminance number for the result. If the sum is more than 255,
> it gets clipped to 255; if it's less than 0, it gets clipped to 0.
> 
> Each channel contributes to the sum according to the percentages, in
> proportion to the brightness of that channel. Where the R channel is
> bright, it contributes a lot to the sum. Where the R channel is dark,
> it contributes little or nothing to the sum, so the end result comes
> mainly or entirely from the G and B channels. And so on.
> 
> Where all the channels are dark, the result will be dark regardless
> of the percentages.
> 
> Increasing the percentage of R means that R values will contribute
> more to the sum. So areas that are bright in Red will be brighter in
> the result.
> 
> If you set R to a negative value, that reverses the contribution of
> the R channel. So areas that are bright in Red will become *darker*
> in the result. This is exactly as if you inverted the R channel and
> then added it in at the same positive percentage.
> 
> There's nothing magic about 100% per channel. The magic combination
> is 30% red, 59% green, and 11% blue. That mixture is what Photoshop
> uses when you just use Desaturate, and it corresponds to some notion
> of "normal" luminance perception. If you want red tones to be lighter
> than "normal", use more than 30% for the R channel (and don't worry
> if it goes over 100). If you want them to be darker than "normal",
> use less than 30% (and don't worry if it goes below 0).
> 
> This is like using colored filters. They reduce the effect of certain
> colors of light on the result. With colored filters, you have to
> compensate for the lost light by increasing the exposure. That allows
> more of the colors that aren't filtered out to get to the film. Same
> thing here. If you reduce the percentage of one channel, you will
> probably want to increase the percentage of the other channels to
> compensate for the overall loss of brightness. Making them add up to
> 100 is like using the filter factor out of the book: it gets you into
> the right ballpark. Your creative vision may require a different sum.
> Keep the histogram palette open and check it for clipping while you
> adjust the percentages.
> 
> There ought to be a button that says "scale my percentages to add up
> to 100", but there isn't. So you have to do it manually.
> 
>    -Paul (another one)
>

Attachments

Move to quarantaine

This moves the raw source file on disk only. The archive index is not changed automatically, so you still need to run a manual refresh afterward.