Great Tutorial on LAB color (easy one!)

lynnma · September 4, 2009

Hi everyone.. like everyone else (except Rutt) I'm struggling with Dan Margulis's book on color corrections. I love the ideas but glaze over with the book. LAB color space is so much more flexible than rgb but it's hard to get your head around. Here is a really good link to a quick and easy action to really make your work POP.... it's not so good for portraits or skin but it's great for landscapes or streetscapes... animals too. I just tried it and am so excited about it I had to share :barb .... you can try it really quickly. Remember when you go to curve, use "image" adjustment - curves... as you'll need to retain the option to "not flatten" later.

Try it.. it's great! Let me know what you think. :clap

arodney · September 4, 2009

lynnma wrote:

Hi everyone.. like everyone else (except Rutt) I'm struggling with Dan Margulis's book on color corrections. I love the ideas but glaze over with the book. LAB color space is so much more flexible than rgb but it's hard to get your head around. Here is a really good link to a quick and easy action to really make your work POP

Or I could just move the Vibrance slider in ACR/LR on that difficult color space, RGB <g>.

Seriously, if we both agree that except for Rutt, many struggle with this book/techniques, and we can avoid having to correct color before we produce an RGB document, in a Raw converter, with all the data provided, non destructively and by simply building a text file that can always be undone or altered, never burning an edit onto pixels, why make life complicated after the fact and make so many people's heads explode? KISS.

pathfinder · September 4, 2009

I tend to use the Vibrance slider like Andrew mentions, but knowing tricks in LAB is still a valuable skill, and always kind of fun when you first learn its power.

Here is a link to the discussion we had about Margulis LAB book , Lynn. You may find these of some help as you work through the book.

lynnma · September 4, 2009

pathfinder wrote:

I tend to use the Vibrance slider like Andrew mentions, but knowing tricks in LAB is still a valuable skill, and always kind of fun when you first learn its power.

<A href="http://www.dgrin.com/showthread.php?t=18203"" target=_blank>Here is a link to the discussion we had about Margulis LAB book , Lynn. You may find these of some help as you work through the book.

Thanks Path... I must have been asleep when this was written<img src="https://us.v-cdn.net/6029383/emoji/mwink.gif" border="0" alt="" > I use Vibrance slider as well, but for bringing out detail and color I'm finding LAB a new fun toy!

lynnma · September 5, 2009

arodney wrote:

Or I could just move the Vibrance slider in ACR/LR on that difficult color space, RGB <G>.

Seriously, if we both agree that except for Rutt, many struggle with this book/techniques, and we can avoid having to correct color before we produce an RGB document, in a Raw converter, with all the data provided, non destructively and by simply building a text file that can always be undone or altered, never burning an edit onto pixels, why make life complicated after the fact and make so many people's heads explode? KISS.

So I'm assuming then that you do all your color adjustments and sharpening in ACR? and, if you do anything at all in PS do you stay with RGB? I've been thinking that sharpening in LAB using the lightness channel was a good way to sharpen without altering any color? Also, how much damage is done by converting an RGB document into LAB and back again.. my sense is that some, but how much. I'm interested to know your method.
Lynn

Richard · September 5, 2009

lynnma wrote:

So I'm assuming then that you do all your color adjustments and sharpening in ACR? and, if you do anything at all in PS do you stay with RGB? I've been thinking that sharpening in LAB using the lightness channel was a good way to sharpen without altering any color? Also, how much damage is done by converting an RGB document into LAB and back again.. my sense is that some, but how much. I'm interested to know your method.
Lynn

Sharpening the L channel works quite well, though you can get pretty much the same effect by putting your RGB sharpened layer into luminosity mode. Except for B&W conversions, I do almost all of my PS work in LAB. Keeping color and luminosity on different dimensions is just easier for me to understand and visualize. I find the Blend If sliders much more useful in LAB. Personally, I have never seen any image degradation as a result of conversion to and from LAB, though I suppose if you try hard enough and convert often enough, you can force some to occur. Why anyone would do that is another matter.

pathfinder · September 5, 2009

lynnma wrote:

So I'm assuming then that you do all your color adjustments and sharpening in ACR? and, if you do anything at all in PS do you stay with RGB? I've been thinking that sharpening in LAB using the lightness channel was a good way to sharpen without altering any color? Also, how much damage is done by converting an RGB document into LAB and back again.. my sense is that some, but how much. I'm interested to know your method.
Lynn

Lynn, Margulis addresses the results of conversion to and from the LAB channel in chapter 6.

Suffice to say that he is convinced it is harmless if the original image is a photograph.

One can create colors in LAB that aRGB or sRGB or CMYK cannot begin to display, so one does want to be careful and avoid computer created gradients in LAB and then convert back to another space.

There are those who do disagree this thesis, of course, but most agree that a trip to LAB and back in a 16 bit image is not discernible.

lynnma · September 5, 2009

pathfinder wrote:

Lynn, Margulis addresses the results of conversion to and from the LAB channel in chapter 6.

Suffice to say that he is convinced it is harmless if the original image is a photograph.

One can create colors in LAB that aRGB or sRGB or CMYK cannot begin to display, so one does want to be careful and avoid computer created gradients in LAB and then convert back to another space.

There are those who do disagree this thesis, of course, but most agree that a trip to LAB and back in a 16 bit image is not discernible.

Thanks Richard and Path... very helpful I shall keep playing

arodney · September 5, 2009

lynnma wrote:

So I'm assuming then that you do all your color adjustments and sharpening in ACR? and, if you do anything at all in PS do you stay with RGB?

Yup.

I've been thinking that sharpening in LAB using the lightness channel was a good way to sharpen without altering any color? Also, how much damage is done by converting an RGB document into LAB and back again.. my sense is that some, but how much. I'm interested to know your method.
Lynn

You don't have to convert to Lab to sharpen, just use the Fade, Luminosity option which will introduce no damage in conversions to Lab and back* and allows you to further control this using the opacity slider.

*Every time a conversion to LAB is produced, the rounding errors and severe gamut mismatch between the two spaces can account for data loss, known as quantization errors. The amount of data loss depends on the original gamut size and gamma of the working space. For example, if the working space is Adobe RGB, which has 256 values available, converting to 8-bit LAB reduces the data down to 234 values. The net result is a loss of 22 levels. Doing the same conversions from ProPhoto RGB reduces the data to only 225 values, producing a loss of 31 levels.
Bruce Lindbloom, a well-respected color geek and scientist, has a very useful Levels Calculator, which allows you to enter values to determine the actual number of levels lost to quantization (see the “Calc page” at http://www.brucelindbloom.com. If you do decide to convert into and out of LAB, do so on a high-bit (16-bit per channel) document.

lynnma · September 5, 2009

arodney wrote:

Yup.

You don't have to convert to Lab to sharpen, just use the Fade, Luminosity option which will introduce no damage in conversions to Lab and back* and allows you to further control this using the opacity slider.

*Every time a conversion to LAB is produced, the rounding errors and severe gamut mismatch between the two spaces can account for data loss, known as quantization errors. The amount of data loss depends on the original gamut size and gamma of the working space. For example, if the working space is Adobe RGB, which has 256 values available, converting to 8-bit LAB reduces the data down to 234 values. The net result is a loss of 22 levels. Doing the same conversions from ProPhoto RGB reduces the data to only 225 values, producing a loss of 31 levels.
Bruce Lindbloom, a well-respected color geek and scientist, has a very useful Levels Calculator, which allows you to enter values to determine the actual number of levels lost to quantization (see the “Calc page” at http://www.brucelindbloom.com. If you do decide to convert into and out of LAB, do so on a high-bit (16-bit per channel) document.

Right....

thanks for the info... I've gone to your link and will do some more reading. I'm finding this information very interesting.. I do like to at least try to understand as much as possible.
Thanks again.
Lynn

Nikolai · September 5, 2009

Sharpening wise I use variety of options:
1) initial rough sharpening can be easily done in ACR
2) if/when I need some PS-base level sharpening (which is typically a late/last stage of the process) I create a "global composite" (I believe the shortcut is ctrl+alt+shit+E

), convert it to a Smart Object and then go have fun with sharpening.
The advantage of this approach is that you don't have to rely on "one time deal" sharpening action and maybe another one luminosity fade fix later, but rather an endless number of attempts, so you can adjust your results as you go. And of course, you have the usual perks of the extra layer opacity...

lynnma · September 7, 2009

Nikolai wrote:

Sharpening wise I use variety of options:
1) initial rough sharpening can be easily done in ACR
2) if/when I need some PS-base level sharpening (which is typically a late/last stage of the process) I create a "global composite" (I believe the shortcut is ctrl+alt+shit+E ), convert it to a Smart Object and then go have fun with sharpening.
The advantage of this approach is that you don't have to rely on "one time deal" sharpening action and maybe another one luminosity fade fix later, but rather an endless number of attempts, so you can adjust your results as you go. And of course, you have the usual perks of the extra layer opacity...

Thanks Nikolai, I'm trying and trying to find a way (prob does'nt exist really) to sharpen certain channels so that the sky stays pretty much untouched. According to Dan Margulis this should be doable.. but I'm not finding it so easy. I've tried sharpening in all sorts of ways but it ALWAYS brings up a little noise in the sky. ... some worse than others of course. Any suggestions? Think of a blue sky with a leafy tree and complex skyline... lots of holes between the leaves so selection is not really appropriate.

arodney · September 7, 2009

lynnma wrote:

Thanks Nikolai, I'm trying and trying to find a way (prob does'nt exist really) to sharpen certain channels so that the sky stays pretty much untouched.

Far easier and more effective to sharpen though a mask built using the image itself (as described in Bruce's book). Basically you use the image itself to make a mask. Smooth areas, like sky would produce no sharpening as they are flat, smooth areas (and there's nothing to sharpen). The mask is based on using the Find Edges filter which as the name suggests, does just that. The net result is the sky and all smooth areas are black (no effect of sharpening), edges are white or shades of gray. Pure white gets 100% of the sharpening applied, levels of gray lesser amounts.

Here is just one flavor of the technique:

http://www.digitaldog.net/files/USM_tutorial.pdf

But again, the correct values to insert for blurring the mask, applying curves to it and of course the USM are all based on testing for the type of sharpening (Capture or output) and in the case of output, the device type and resolution of the file, something Bruce addresses in the Creative Pro article.

All of this masking work is possible within Lightroom! Its vastly more advanced at this kind of work than Photoshop unless you really know the exact values to insert. In LR, you build and view the mask on the fly as you zoom in at 100% and use the alt/option key on most of the sliders you find in Develop module. For output, its all automatic!

Richard · September 7, 2009

Masking is one solution. Here's another that often works for me with noise in a blue sky. In LAB, merge up to a new layer (ctl/cmd-shift-alt/option-E) and do your normal sharpening on the L channel. Then go to the blending options of the new layer, select the B channel and set the blend-if slider to exclude the blues on the underlying layer. Since the sky isn't yellow--and vegetation is--it will sharpen the vegetation without touching the sky. You will probably want to split the slider to give a slight gradient to the effect. I find this method much faster than masking, though it won't work if you have significant blue detail that's not in the sky.

lynnma · September 7, 2009

Richard wrote:

Masking is one solution. Here's another that often works for me with noise in a blue sky. In LAB, merge up to a new layer (ctl/cmd-shift-alt/option-E) and do your normal sharpening on the L channel. Then go to the blending options of the new layer, select the B channel and set the blend-if slider to exclude the blues on the underlying layer. Since the sky isn't yellow--and vegetation is--it will sharpen the vegetation without touching the sky. You will probably want to split the slider to give a slight gradient to the effect. I find this method much faster than masking, though it won't work if you have significant blue detail that's not in the sky.

Thanks Richard and Andrew... both great replies. I'll go study both these options. I did just get a fair job from doing everything in ACR... got some decent sharpening with a flawless blue sky using all the tools available... but thats much too easy

I'll go look at these other options now...thankyou both for taking the time to help me with this. I'ts so fascinating.

arodney · September 7, 2009

lynnma wrote:

I did just get a fair job from doing everything in ACR... got some decent sharpening with a flawless blue sky using all the tools available... but thats much too easy

Then you're a prefect candidate for reading Dan's books over and over again: wink.

Life's too short to fall in love with complex and unnecessary techniques when most of us would rather create pictures anyway. As my dear friend Bruce Fraser would say:

You can do all sorts of things that are fiendishly clever, then fall in love with them because they're fiendishly clever, while overlooking the fact that they take a great deal more work to obtain results that stupid people get in half the time. As someone who has created a lot of fiendishly clever but ultimately useless techniques in his day, I'd say this sounds like an example. Bruce

The beauty of ACR and LR is the team building them (at the top of the list Thomas Knoll) want to make them the anti-photoshop in terms of complexity.

lynnma · September 7, 2009

arodney wrote:

Then you're a prefect candidate for reading Dan's books over and over again: wink.

Life's too short to fall in love with complex and unnecessary techniques when most of us would rather create pictures anyway. As my dear friend Bruce Fraser would say:

The beauty of ACR and LR is the team building them (at the top of the list Thomas Knoll) want to make them the anti-photoshop in terms of complexity.

So True Andrew

I'm fast realizing that... and yes... I note Thomas Knoll's name every time I open PS and silently thank all those dedicated geniuses for their contribution to the digital arts.

Next week I'm starting on my 18 month cross America trip in a trailer with my camera... hopefully to get some great shots (would'nt that be nice) and I'M AS READY AS I'M EVER GONNA BE!!!

Lynn

BinaryFx · September 18, 2009

lynnma wrote:

Hi everyone.. like everyone else (except Rutt) I'm struggling with Dan Margulis's book on color corrections. I love the ideas but glaze over with the book.

Lynnma, perhaps you should take a look at:

http://www.kelbytraining.com/instructors/dan-margulis.html

There are two Lab mode videos:
- The LAB Frontier
- Introduction to Photoshop LAB Color

Subscription to the site entitles one to view all the instructors/tutorials, not just the two mentioned above!

Hope this helps,

Stephen Marsh

pathfinder · September 18, 2009

I vastly prefer the basic sharpening in Lightroom2, than Photoshop, like Andrew suggested. The ability to see the mask real time and on the fly is just super. I can make the mask as soft or as limited to edges as I usually desire.

That said, it is real easy to exclude the sky with Blend IF sliders in Photoshop. I find I can also do it either by Quick Mask, or Color Select very easily also. There really are many, many ways to avoid sharpening the sky. I frequently prefer to run my skies through NoiseWare, even if shot at lower ISOs so they are nice and smooth, then excluding them from any final local contrast enhancement.

Great Tutorial on LAB color (easy one!)

Comments