Author | Thread |
|
08/15/2004 01:28:26 PM · #1 |
Greetings...
I have been studying some alternative methods of converting a color image to black and white. I purchased this book:
"The Photoshop Book for Digital Photograhpers" by Scott Kelby
There are several methods outlined in this book, but the channel mixer method seems to be the most popular mode. In the past, my conversions to black and white have always been rather simple. My workflow usually starts with a simple conversion to grayscale. I follow that with any necessary level and contrast adjustments.
In an effort to learn the benefits of the channel mixer method, I played around with one of my recent photos. I started out by making the conversion my normal way. I got the photo to look exactly like I wanted it. Then I played around with the channel mixer method. With the channel mixer, I could not come up with anything I liked particularly better than what I came up with on my normal conversion. The book does a nice job of explaining the channel mixer and what it is actually doing, but I am struggling to find the benefit of using it. Adjusting the red/green channels seem to simply change the contrast in the image. The blue channel doesn't seem to have much of an effect on the photo I was working with, but there wasn't much blue in it to begin with.
Can someone pass me some tips on why I should use the channel mixer? How will it benefit me?
|
|
|
08/15/2004 01:49:28 PM · #2 |
The only benefit I've found so far is the ability to use it in an adjustment layer so that the change is non-destructive; the underlying color data is still present in the Background layer. Combined with a mask, it's also an easy way to do a selective desaturation-type effect.
Mostly though, I also find I get about the same or better result making a simple conversion of RGB > Grayscale followed by the judicious application of one or more Curves. I think you could also get some of the benefit of the Channel Mixer method by applying specific color Curves to the original image before making the conversion to grayscale.
Maybe it's what you're used to doing ... I've been making the Mode change to Grayscale since PS 2.0, and only started tryng out the Channel Mixer last year : )
Message edited by author 2004-08-15 13:49:46. |
|
|
08/15/2004 01:51:00 PM · #3 |
when i use it (which is rare, but occasionally i do it because i want to use an adjustment layer rather than duplicating the whole layer), the first thing i do is take the "constant" setting down to about -30. After that I bring up the blue and the green to get the look i want.
I find the red channel is a little overdone at 100 as well, so i drop it a bit. if i leave the red and the constant up, i find any adjustments to the other channels just blow out my highlights.
I get decent results, but not as consistent as my usual duo/quadtone method.
P-ness |
|
|
08/15/2004 01:53:13 PM · #4 |
Hi Jim, I occasionally use the channel mixer to get more severe contrast for b&w stuff, I quite often use the gradient map as well. What I find quite good for adding tonal range sometimes is to use the gradient map, then the highpass filter at 255 and then the gradient map again on top of it. |
|
|
08/15/2004 02:42:19 PM · #5 |
I found the same results John,If I want a black and white image, I do all the tonal and contrast things in colour then convert and tweak it. Might not be the correct way but it works for me.
If I am still not happy I resort to duotone.
|
|
|
08/15/2004 02:52:37 PM · #6 |
Cool -- I'm not the only one that is not finding it all that useful.
Don't know if this will be of benefit, but I was under the impression it served a similar purpose to using colored filters in B&W photography -- with a lot more control over which colors are accented. I had set it aside as something much further along the learning curve than I am, and I needed to concentrate on other things.
Although I generally desaturate an image in an adjustment layer, so I can apply adjustments and filters to the channels individually if I feel the need. So maybe what I am doing isn't all that different after all.
David
|
|
|
08/15/2004 03:10:12 PM · #7 |
I think the CM method simply provides more in-edit control (as an adjustment layer) and a greater ability to control constrast/tone relationships without the necessity of other tools (contrast/levels/etc). On the other hand, using CM does not mean that those other tools won't still come in handy to further refine the b/w conversion.
You can get a grayscale equivalent by using a (R,G,B) = (30,59,11) CM recipe. That would be a starting point for someone who wants to explore the benefits of using CM. I'm currently playing with the dual HS layer approach, but I still use CM most of the time. I never use grayscale because it throws a bit of a wrench into a color managed workflow.
Message edited by author 2004-08-15 15:10:30. |
|
|
08/15/2004 03:10:56 PM · #8 |
Jim, I find it a whole lot more useful on landscapes and still life images than i do on people. it's particularly good when one channel dominates the image (like a really red/orange fall scene, or green trees behind green grass etc). with people - especially brown eyed people - it does weird stuff to their eyes.
P
@david - looks like you've been playing in my backyard ;)
Message edited by author 2004-08-15 15:13:01. |
|
|
08/15/2004 03:20:07 PM · #9 |
John, Count me in...I really have never had any luck for the main use of the "channel mixer".....maybe for shifting colors on other photos than B/W's...even with useing the presets for b/w in channel mixer...I have to agree about converting to Grayscale and then add tones (if will) and then adjust the levels ect....From what I have found that this is the best way for now....Sorry, to have not any ideas how to make channel mixer better for b/w's....I think that if you keep asking around, you'd find out more....Btw, I'm looking forward to your new Website, "duotones.org." Waiting to add to the collections....it's a great idea you have there!
|
|
|
08/15/2004 03:20:14 PM · #10 |
Hi, I think using the channel mixer does give more control over the image if you want to create a particular effect. Obviously it depends very much on the subject and the main colours in the start image. For instance the blue channel can give really high contrast effects and will enhance things like freckles and wrinkles in portraits.
With portraits of older people it can really emphasise their age and skin condition. |
|
|
08/15/2004 06:41:39 PM · #11 |
Firstly it is very image dependant. What it allows you to do is change the amount of each channel in the final image. Image-Greyscale uses a set % of each channel. Channel mixer lets you do it yourself and vary the amount. So we are to the point of 'more control' again.
Usually, I start off by duplicating the image, going in to the menu on the channel palette and separating the various channels out - you can then see the usually stark difference between the red, green and blue channels. This is particularly true for white/light skin tones - the red channel is almost porcelin smooth, the blue channel has all the freckles, spots and skin marks.
From there, I get a good idea of how much of which channel I want to use - for an 'even' conversion, I use ratios so that they add up to 100%. For a high contrast effect, I'll mix over 100%, to about 150% or so.
It can also be interesting to subtract channels, using -% values.
But right at the heart of this - it really depends on the type of scene you've shot, and if you want to vary the tonal relationships between the colours or not - e.g., a sky shot were you could change it from an almost light sky to an almost completely dark sky, as you vary the blue component.
As a final aside, I never, never, ever use the constant adjuster. It is roughly equivalently useless as the brightness/ contrast adjustments.
Message edited by author 2004-08-15 18:42:25. |
|
|
08/16/2004 09:10:25 AM · #12 |
I still love the non destructiveness and control of DigiDan's Conversion the best.
|
|
|
08/16/2004 09:25:23 AM · #13 |
I'm sure someone will correct me if I'm wrong, but I am under the impression that using the channel mixer for b&w conversion keeps more tonal detail in the image. A black and white image contains 256 discrete tonal levels from black through to white. A colour one has 256*256*256 different levels, and using the channel mixer allows you to keep that information when converting. You can see that this is important when carrying out curves modifications, which can result in a posterised effect if the 256 levels are stretched out too far. |
|
|
08/16/2004 09:46:39 AM · #14 |
Have been experimenting a bit with this method to make high contrast B&W´s:
1. Enhance - Adjust Color - Remove Color
2. Enhance - Adjust Brightness/Contrast - Levels
3. Layers - Duplicate layer
4. Image - Adjustments - Threshold (For high key images lower threshold very much but less for normal and very litle for low key)
5. Blend the layers with "overlay" and find out what "opacity" suits best (I often but it to 50-70%).
This method doesnt work for all images but the results are sometimes sharp and high contrasted images.
Just recently started to experiment with this. Maybe post something here soon. |
|
|
08/16/2004 09:47:16 AM · #15 |
that is what i use almost solely. i like the results, and its a one slider deal. i edit the lower adjustment layers hue.
Originally posted by dacrazyrn: I still love the non destructiveness and control of DigiDan's Conversion the best. |
and i think its what dwoolridge is doing as well with his exploration of hue/saturation adjustment layers - just a key press to create those adjustment layers with digidaans action.
Originally posted by dwoolridge: I'm currently playing with the dual HS layer approach, but I still use CM most of the time. I never use grayscale because it throws a bit of a wrench into a color managed workflow.
|
Message edited by author 2004-08-16 09:47:48.
|
|
|
08/16/2004 09:53:40 AM · #16 |
Originally posted by dacrazyrn: I still love the non destructiveness and control of DigiDan's Conversion the best. |
I also like this dual-adjustment-layer approach. |
|
|
08/16/2004 09:58:01 AM · #17 |
I always just use image >> desaturate.

 
 
  

Think I will test some of those advanced methods out!!
Message edited by author 2004-08-16 10:00:43.
|
|
|
08/16/2004 09:58:40 AM · #18 |
the channel mixer conversion has worked very few times for me. Most of the time it just turns normal pictures into incredibly noisy ones. Perhaps I'm doing something wrong, but it just doesn't work for me. I'd like to understand why it does this.
June
|
|
|
08/16/2004 10:16:33 AM · #19 |
Originally posted by chiqui74: the channel mixer conversion has worked very few times for me. Most of the time it just turns normal pictures into incredibly noisy ones. Perhaps I'm doing something wrong, but it just doesn't work for me. I'd like to understand why it does this.
June |
Use less of the blue channel, more of the red & green channels, and make sure the %'s add up to around 100%. Don't use the constant adjuster, have monochrome checked.
The blue channel is usually far and away the noisiest part of the image, particularly on portraits.
|
|
|
08/16/2004 10:19:03 AM · #20 |
Originally posted by jmsetzler:
Can someone pass me some tips on why I should use the channel mixer? How will it benefit me? |
I knew I had this somewhere. This is something I wrote a couple of years ago, but the pictures at least illustrate the reasons why I find the channel mixer potentially useful.
B&W workflow
|
|
|
08/16/2004 10:20:26 AM · #21 |
The channel mixer simulates a B&W darkroom. By using red, green, or red filters in the darkroom you can increase the contrast of the print differently. Use the opposite color to bring out the tone you aiming for. |
|
|
08/16/2004 10:30:40 AM · #22 |
I don't always use the channel mixer, but even when I do, I look at the photo in each RGB channel first, to see what might work best in the channel mixer. For example, I'll look at the blue channel, green channel and red channel to see how the photo looks.
Sometimes, I'll just switch to Lab Color and simply convert to grayscale using the Lightness channel. Sometimes, in RGB, I'll simply convert to grayscale using one of those channels.
In this image, I used the channel mixer after seeing the effects in each of the different channels. I finally selected the red channel, which was like putting a red or orange filter on the sky, making it darker.:
In this photo, however, I used the green channel, which kept everything lighter, because there are no contrasting colors with green in the photo:
The red channel is great for caucassian skin in darker situations, since skin has lots of red tones, so I used the red channel for this one:
Here's one where I used the Lightness channel and simply converted to grayscale (and deleted the other channels):
So the channel mixer, as MeThos indicates, can be kind of like using filters. I check all methods out until I find the one that suits the subject best. But more often than not, now, I'll PLAN on using a certain channel, so I can do more special work in-camera, like put a blue or yellow cast over the photo so when I convert to b/w I'll have a certain effect.
(edited for grammar; added another sentence)
Message edited by author 2004-08-16 10:33:43.
|
|
|
08/16/2004 11:20:35 AM · #23 |
If you are a color photographer, and had used color film or get in direct in digital, you really don´t saw any great advantage using Channel Mixer method.
It method has the particular benefit simulating the color filters and filters stacking used in the old B&W Film pictures. I really agree that is more dificult to understand this method if you don´t do it phisically before. I finally understand it by viewing a lot of photos in B&W at art shows. The diference between this and the simple grayscale conversion is the power of control over each channel. That is only usefull if you really knows what you whant to get. |
|
|
08/16/2004 11:22:31 AM · #24 |
Originally posted by GoodEnd:
It method has the particular benefit simulating the color filters and filters stacking used in the old B&W Film pictures. I really agree that is more dificult to understand this method if you don´t do it phisically before. I finally understand it by viewing a lot of photos in B&W at art shows. The diference between this and the simple grayscale conversion is the power of control over each channel. That is only usefull if you really knows what you whant to get. |
Worth noting, this is the reason I suggested using the 'split channels' option to see each channel as a separate image. Really helps me work out which elements from which channel that I want to combine. Further along the path, is the option to use masks to select features from particular channels when you combine them too.
|
|
|
08/16/2004 11:50:59 AM · #25 |
I used to do Channel mixer method, results were quite flat, I was reading book on BW photography, one that changed my life was how some Photographer Print Fine BW using DUTONE and from that point my WORK flow is as Follow's
Image>Gradient Map>Convert Image to gray scale ( using Image Mode) > than convert into Dutone> there adjust the two curve.
Results are very crisp |
|
Home -
Challenges -
Community -
League -
Photos -
Cameras -
Lenses -
Learn -
Help -
Terms of Use -
Privacy -
Top ^
DPChallenge, and website content and design, Copyright © 2001-2025 Challenging Technologies, LLC.
All digital photo copyrights belong to the photographers and may not be used without permission.
Current Server Time: 08/09/2025 04:10:38 PM EDT.