Ad
Ad
Ad
Pages: « 1 2 3 [4]   Bottom of Page
Print
Author Topic: Lab Color for B/W - Why does it work so well?  (Read 11899 times)
john beardsworth
Sr. Member
****
Offline Offline

Posts: 2846



WWW
« Reply #60 on: February 23, 2010, 03:25:37 AM »
ReplyReply

Quote from: DeanSonneborn
...I'm beginning to think that difference techniques can lead to a similar outcome, so perhaps the specific technique is less important than I originally thought since I seem to be able to get the image to where I want it.
That's kinda what I was saying. Focussing too much on the technique takes away from your interpretation of the subject, or your ability to separate colours into distinct tones of grey. However, if you use Lightroom or Photoshop's targeted adjustment tools, you are keeping your eyes on the picture's appearance.

As well as taking your eyes off the picture, the 3 layer technique is pointless when you can achieve the same outcome by using adjustment layers (the B&W adjustment layer if you have CS3/4 and the channel mixer for earlier versions). And if you read that Gorman article quickly, it's more about toning than about the art of black and white and its Lab method is merely one whose conversion looks OK with skin tones. The results are always pretty neutral and the method gives little scope for interpretation - hence your need to do more work afterwards changing the midtones. With the more modern techniques, you can vary the conversion recipe to produce tonal separation (make greens look different from reds etc) or to interpret the image (eg make clouds stand out more or less).

John
Logged

Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6983


WWW
« Reply #61 on: February 23, 2010, 06:04:40 AM »
ReplyReply

If you make changes to Adjustment Layers regardless of where they are in the layer stack, you will see the changes in the image. If you make changes to pixel-based layers, the upper-most layer predominates.

Your observation that different techniques can produce very similar results is correct, and to the extent that's true it argues for using the least convoluted, most efficient approach available for achieving the B&W tonality you like.
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6983


WWW
« Reply #62 on: February 23, 2010, 06:10:15 AM »
ReplyReply

Quote from: johnbeardy
That's kinda what I was saying. Focussing too much on the technique takes away from your interpretation of the subject, or your ability to separate colours into distinct tones of grey. However, if you use Lightroom or Photoshop's targeted adjustment tools, you are keeping your eyes on the picture's appearance.

As well as taking your eyes off the picture, the 3 layer technique is pointless when you can achieve the same outcome by using adjustment layers (the B&W adjustment layer if you have CS3/4 and the channel mixer for earlier versions). And if you read that Gorman article quickly, it's more about toning than about the art of black and white and its Lab method is merely one whose conversion looks OK with skin tones. The results are always pretty neutral and the method gives little scope for interpretation - hence your need to do more work afterwards changing the midtones. With the more modern techniques, you can vary the conversion recipe to produce tonal separation (make greens look different from reds etc) or to interpret the image (eg make clouds stand out more or less).

John

I agree, and not only for the convenience of instant feedback re appearance,  but also of course those targeted adjustment tools really do a very good job.

As well, when you combine the B&W Adjustment Layer (or the Greyscale panel in LR/ACR) with Curves and Layer masking, one can get an infinity of effects. And if one wants even more easy options and localized control, and does a lot of B&W work, Nik SIlver Efex Pro (used within Photoshop as a plugin) is highly recommended. Conversion to L*a*b* is not necessary to produce good B&W skin tones.
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
DeanSonneborn
Newbie
*
Offline Offline

Posts: 44


« Reply #63 on: February 25, 2010, 08:59:22 PM »
ReplyReply

t seems that if I set the opacity of 1 one the layers to 100%, that then making any adjustment above that 100% layer, seems not to have any effect.

Quote from: Mark D Segal
If you make changes to Adjustment Layers regardless of where they are in the layer stack, you will see the changes in the image. If you make changes to pixel-based layers, the upper-most layer predominates.

Your observation that different techniques can produce very similar results is correct, and to the extent that's true it argues for using the least convoluted, most efficient approach available for achieving the B&W tonality you like.
Logged
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6983


WWW
« Reply #64 on: February 25, 2010, 09:07:06 PM »
ReplyReply

Quote from: DeanSonneborn
t seems that if I set the opacity of 1 one the layers to 100%, that then making any adjustment above that 100% layer, seems not to have any effect.

If you are trying to say that by setting an underlying layer to 100% opacity, any adjustments to the layer sitting above it have no effect, I don't see how this is possible. I can't replicate it. Or do you mean something else? And are you talking about layers containing pixels, or adjustment layers?
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
DeanSonneborn
Newbie
*
Offline Offline

Posts: 44


« Reply #65 on: February 25, 2010, 10:40:18 PM »
ReplyReply

I usually set my layers as red on the bottom, green in the middle, and blue on the top. If I set the opacity of the red layer to 100% then try to adjust the green layer opacity I don't see any changes to the over all image.

Quote from: Mark D Segal
If you are trying to say that by setting an underlying layer to 100% opacity, any adjustments to the layer sitting above it have no effect, I don't see how this is possible. I can't replicate it. Or do you mean something else? And are you talking about layers containing pixels, or adjustment layers?
Logged
john beardsworth
Sr. Member
****
Offline Offline

Posts: 2846



WWW
« Reply #66 on: February 26, 2010, 02:05:45 AM »
ReplyReply

Really, as shown by your confusion here, the layers method is a contorted way of working and produces no better results than more modern methods. You'd be better off using on or more adjustment layers.

John
« Last Edit: February 26, 2010, 02:07:46 AM by johnbeardy » Logged

Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6983


WWW
« Reply #67 on: February 26, 2010, 07:21:53 AM »
ReplyReply

Quote from: DeanSonneborn
I usually set my layers as red on the bottom, green in the middle, and blue on the top. If I set the opacity of the red layer to 100% then try to adjust the green layer opacity I don't see any changes to the over all image.

Oh - I didn't realize you were speaking of the contorted method of getting to B&W by converting channels to layers; I thought you were talking about the normal way of using Photoshop. If it's in the context of the former, good as its proponents say it is and I won't argue, I have no idea why it's not working because it isn't a procedure to which I would commit time.
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6983


WWW
« Reply #68 on: February 26, 2010, 07:22:27 AM »
ReplyReply

Quote from: johnbeardy
Really, as shown by your confusion here, the layers method is a contorted way of working and produces no better results than more modern methods. You'd be better off using on or more adjustment layers.

John

Indeed.
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
DeanSonneborn
Newbie
*
Offline Offline

Posts: 44


« Reply #69 on: February 26, 2010, 08:05:08 AM »
ReplyReply

I had heard that the layer technique was "the best" and so I thought I would give it a try but, as you all can see, it seems to be more trouble than it is worth. I too may return to LR B&W and local adjustment tools.

Quote from: Mark D Segal
Indeed.
Logged
Schewe
Sr. Member
****
Offline Offline

Posts: 5512


WWW
« Reply #70 on: February 26, 2010, 11:43:31 AM »
ReplyReply

Quote from: johnbeardy
Really, as shown by your confusion here, the layers method is a contorted way of working and produces no better results than more modern methods.


Actually, the channels>layers can produce results that are near impossible to reproduce using adjustment layers...

See, if you start with a color image and use the B&W adjustment, Desaturate, Channel Mixer, etc. you get a global conversion to B&W.

There is no easy way to have local control how the conversion will be handled...you can't use layer masks because, well if you mask part of the adjustment layer, the reveal will be back to the color image and trying to place yet another adjustment layer over it becomes problematic...

In terms of B&W conversion in the raw processor, same global deal.

Where the channel>layer is superior is that you can literally paint in a custom B&W conversion area by area and use masks to control what layer are visible. It's pretty easy to do an action that will do the base conversion for you and add hide all layer masks so all you need to do is choose the layer and paint it in via the mask.

I won't claim it's "the best" way...I often do B&W conversions directly in Camera Raw or Lightroom (usually Lightroom where I can make a virtual copy and keep the color & B&W separate). But if you need precise control over different areas in a B&W conversion then the channel>layer is a good solution. Of course, one does have to know how to do it...
Logged
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6983


WWW
« Reply #71 on: February 26, 2010, 11:52:44 AM »
ReplyReply

Quote from: Schewe
But if you need precise control over different areas in a B&W conversion then the channel>layer is a good solution. Of course, one does have to know how to do it...

Jeff, refined localized control is also very much achievable using Nik Silver Efex, and it's a helluva lot easier than replicating this kind of functionality directly in PS. Of course it's another 160 bucks, but for those who may be doing a lot of this stuff, most likely worth its cost in time saved and additional features/functionality. I've been playing with it on-and-off and like it.
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
kikashi
Sr. Member
****
Offline Offline

Posts: 4062



« Reply #72 on: February 26, 2010, 12:02:03 PM »
ReplyReply

Quote from: DeanSonneborn
I usually set my layers as red on the bottom, green in the middle, and blue on the top. If I set the opacity of the red layer to 100% then try to adjust the green layer opacity I don't see any changes to the over all image.
Well, I'm no PS wizard, but I believe that that's what "100% opacity" means. The red layer is completely (100%) opaque, so of course you don't see any difference if you fiddle with layers underneath. What did you expect?

Jeremy
Logged
DeanSonneborn
Newbie
*
Offline Offline

Posts: 44


« Reply #73 on: February 26, 2010, 05:12:08 PM »
ReplyReply

I was not quite sure what to expect so I thought I would "test" and see. Nothing wrong with that and your right 100% means 100%, end of adjustment. See, now I know.


Quote from: kikashi
Well, I'm no PS wizard, but I believe that that's what "100% opacity" means. The red layer is completely (100%) opaque, so of course you don't see any difference if you fiddle with layers underneath. What did you expect?

Jeremy
Logged
Schewe
Sr. Member
****
Offline Offline

Posts: 5512


WWW
« Reply #74 on: February 26, 2010, 08:43:09 PM »
ReplyReply

Quote from: Mark D Segal
Jeff, refined localized control is also very much achievable using Nik Silver Efex

So, you have the full capability of using Color Range and/or other Photoshop selection techniques such as paths or calculations to create your local selection of color>B&W conversions? You do if you use channels as layers (and have the color image open in order to make color based selections).

Really, haven't you ever seen me demo the channels to layers technique? If you want total, absolute control (without having to resort to actual tone changes) it's the most powerful method of controlling the result...
Logged
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6983


WWW
« Reply #75 on: February 26, 2010, 08:46:57 PM »
ReplyReply

No I haven't seen your demo. Can you point me to it?

But have you tried Silver Efex Pro's U-point selection technology along with their various effects filters?
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
Schewe
Sr. Member
****
Offline Offline

Posts: 5512


WWW
« Reply #76 on: February 26, 2010, 08:56:43 PM »
ReplyReply

Quote from: Mark D Segal
No I haven't seen your demo. Can you point me to it?

Uh well, it's old but try THIS LINK (800KB PDF)

Quote from: Mark D Segal
But have you tried Silver Efex Pro's U-point selection technology along with their various effects filters?

No, sorry, I don't do NIK–for obvious reasons if you think about it...

:~)

But I'm pretty darn sure the channels>layers is still more powerful and flexible (and easy if you record an action to create your layers and layer masks).
Logged
stamper
Sr. Member
****
Offline Offline

Posts: 2821


« Reply #77 on: February 27, 2010, 03:17:48 AM »
ReplyReply





No, sorry, I don't do NIK–for obvious reasons if you think about it...

:~)

I don't see why you can't use it as a plug in in Photoshop? It isn't unethical? If you find it to be good then it is to your advantage and if you don't then you are free to criticize it?
Logged

Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6983


WWW
« Reply #78 on: February 27, 2010, 06:51:04 AM »
ReplyReply

Quote from: Schewe
Uh well, it's old but try THIS LINK (800KB PDF)



No, sorry, I don't do NIK–for obvious reasons if you think about it...

:~)

But I'm pretty darn sure the channels>layers is still more powerful and flexible (and easy if you record an action to create your layers and layer masks).

Hi Jeff,

Many thanks for the paper. Now that I access it, I seem to recall that I may have seen it before. I have copied it to my hard drive and will do a tryout. BTW, your illustrations there are stunning.

As for NIK, yes of course I had thought about it, and you know that I know the context, but I was simply thinking of the product as a product, because that particular application is unique.  :-)
Anyhow, I'll select an appropriate image or two and do a side-by-side. Won't be this weekend, but I'll report back when I'm done.

(minor edit today 13:45 EST)
« Last Edit: February 27, 2010, 12:46:08 PM by Mark D Segal » Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
john beardsworth
Sr. Member
****
Offline Offline

Posts: 2846



WWW
« Reply #79 on: March 01, 2010, 09:31:43 AM »
ReplyReply

Quote from: Schewe
Actually, the channels>layers can produce results that are near impossible to reproduce using adjustment layers...

See, if you start with a color image and use the B&W adjustment, Desaturate, Channel Mixer, etc. you get a global conversion to B&W.

There is no easy way to have local control how the conversion will be handled...you can't use layer masks because, well if you mask part of the adjustment layer, the reveal will be back to the color image and trying to place yet another adjustment layer over it becomes problematic...

In terms of B&W conversion in the raw processor, same global deal.

Where the channel>layer is superior is that you can literally paint in a custom B&W conversion area by area and use masks to control what layer are visible. It's pretty easy to do an action that will do the base conversion for you and add hide all layer masks so all you need to do is choose the layer and paint it in via the mask.

I won't claim it's "the best" way...I often do B&W conversions directly in Camera Raw or Lightroom (usually Lightroom where I can make a virtual copy and keep the color & B&W separate). But if you need precise control over different areas in a B&W conversion then the channel>layer is a good solution. Of course, one does have to know how to do it...

My typo didn't help, but notice you quoted my saying "one or more adjustment layers". If you don't want a global conversion but do want to paint in custom conversions, then the second or later B&W adjustment layer would apply its conversion recipe only to whatever colour shows through after the initial, masked B&W adjustment layer - your "reveal". The topmost adjustment layer would need no mask (to be sure that no remaining colour is peeking through). So I'd have the image as a raw file smart object, plus local conversion with 2+ masked adjustment layers.

Where's the advantage? Well, to some extent we're different rather than better. Assuming we're both staying in high bit etc, we'll both end up with the same best quality data, and we don't care if your 3 pixel layers plus masks result in a bigger file than mine (the SO probably makes it even anyway). We can assume your action included a colour layer, so you can always output the same file as a colour image as easily as I can by switching off my adjustment layers. We'll even take the targeted adjustment tool's virtues out of the comparison. And let's assume in each case one does indeed know how to do it....  

I'd say the main difference is in flexibility, which is key because you don't always get things right first time. Let's say that after making the b&w conversion one wants to do something like tweak the lens vignetting correction or adjust capture sharpening or noise reduction - I'm going to find that a lot easier as my adjustment layers have entirely separated the b&w conversion from the image smart object which I can tweak in ACR. That advantage is even greater if one wants to do more substantial in-Photoshop retouching, whether that's cloning or something like perspective or lens correction. Again, the adjustment layers would provide flexibility that your pixels layers can only provide if one really, really knows how.

The end results can be the same, but the workflow's much sweeter.

John
« Last Edit: March 01, 2010, 09:36:11 AM by johnbeardy » Logged

Pages: « 1 2 3 [4]   Top of Page
Print
Jump to:  

Ad
Ad
Ad