Thanks to this site, I bought Dan's 'LAB' Book
I've followed the LAB Canyon Conundrum book discussions and tutorials here for a while. I've been enlightened by a few of the techniques mentioned so I figured it a good investment to buy the book.
I can say this site and the tutorials that Rutt, et.al. provided have prompted me to go and buy the book. It'll be here tomorrow.
Dan really needs to give you guys some kudos for what you've done here.
finally got my copy!
thanks rutt for all your encouragement to get this book. i've stopped buying application specific books for some years. a lot of them are so much rubbish to be thrown out with the next release of photoshop. this is a lovely book though. nice stock, well produced, and more to the point, something that'll be relevant for many years, and margulis writes very well.
i'd encourage others to buy the book. i've only read the first 2 chapters but there's a lot of subtilties to the "recipes" he gives. in fact, his recipes are a bit of a pedagogical carrot. i'm really enjoying reading it. read chapter 1 and said, ahhhhh, so *that's* the canyon conundrum...
margulis obviously knows a lot about colour. and even if you don't want to use some of his techniques or become a total convert to LAB forever and ever, it's worth reading this book to get some insight into colour in general. the more technical sections are definitely worthwhile persisting with.
Where do we stand with the synopses of the remaining chapters?
Pathfinder is singed up for Ch 11 and I have a draft from him. Edgework is signed up for Ch 14. That's all I know about that's outstanding. After the holidays, I'll try to reawaken the Ch 16 practice bakeoff, but this just doesn't seem to be the right time of year for it.
Is one of the ones that isn't spoken for interesting to someone?
Got the book for Christmas!
I just got the book for Christmas. So far I'm loving it, especially the second half of each chapter. I recently finished the Total Training for Photoshop CS (about 25 hours of video) and then started trying to extract some backgrounds and couldn't do it. That's one of the things that is really interesting about lab for me. I'm only in chapter 2 so I couldn't wait, I had to read the excellent summary by Rutt. When I first looked at the a & b channels they were so grey I thought there wasn't enough there to make masks. Rutt, you've explained it very well with your examples.
I got turned on to lab because I posted a message on dpreview about my workflow and someone said something like "that sounds complicated for each image, it would be must faster with lab". Since then I've been playing around and really having fun. Now that I have the book and this site I'm really looking forward to doing new things.
I'm curious, how many of you use lab as your primary colorspace? I mean you save your 'master' PSD file in lab and only convert to RGB to save as JPEG? I use a lot of adjustment layers and switching back and forth between lab & jpeg isn't possible, so I'm curious how people handle this. It sounds better to just stay in lab. Is it feasible?
I just got the book for Christmas. So far I'm loving it, especially the second half of each chapter. I recently finished the Total Training for Photoshop CS (about 25 hours of video)l
Goodness you sat through 25 hours of video. Please enlighten us on the fun things you've learned.
Goodness you sat through 25 hours of video. Please enlighten us on the fun things you've learned.
I've used photoshop a bit in the past but never learned advanced things, I always used basic stuff. So there were some things that I knew in the training, but what was interesting was I was constantly learning new tricks and tips. Sometimes I fast forwarded through boring parts but overall I think the Total Training program is excellent. Overall I learned a lot of theory and practical uses, like the channels which I had never used before and the blend modes. I find now that learning new techniques is much easier now that I have a good foundation! For example I'm picking up l*a*b really quickly, and am getting good at extractions using channels (in lab mode).
I've used photoshop a bit in the past but never learned advanced things, I always used basic stuff. So there were some things that I knew in the training, but what was interesting was I was constantly learning new tricks and tips. Sometimes I fast forwarded through boring parts but overall I think the Total Training program is excellent. Overall I learned a lot of theory and practical uses, like the channels which I had never used before and the blend modes. I find now that learning new techniques is much easier now that I have a good foundation! For example I'm picking up l*a*b really quickly, and am getting good at extractions using channels (in lab mode).
-Michael
Yeah, it's like once you learn spanish, italian is easy!
Darn, I just got my copy the 3rd of jan. Luckily I read a lot of the book at the bookstore. I actually haven't looked at it since I got it. I have been using LAB color since mid november. I love how it makes plain photos pop so much. I made this plain "portrait" of my little brothers toy amazing, well interesing at least. I know I went a lttle overboard but I like it.
I'm curious, how many of you use lab as your primary colorspace? I mean you save your 'master' PSD file in lab and only convert to RGB to save as JPEG? I use a lot of adjustment layers and switching back and forth between lab & jpeg isn't possible, so I'm curious how people handle this. It sounds better to just stay in lab.
-Michael
i work at an ad agency where everything ends up in CMYK. However, most of our images come in RGB. I have sometimes wound up saving three different PSD files, LAB, RGB and CMYK as a reflection of the actual sequence that I go through, LAB being the initial step with all the heavy lifting and either RGB or CMYK (or both) reserved for the types of subtle changes that aren't always convenient in LAB (which I sort of see as the atomic bomb of emage enhancement.) Which is not to say that LAB is incapable of subtlety. But If a face has overall good tone, but the highlights are pulling a bit yellow, there would be no point in trying to make that move in LAB since CMYK is perfect for such a shift.
There are two ways to slide through life: to believe everything or to doubt everything; both save us from thinking.
—Korzybski
I finally have my copy of Dan Margulis' new book, Photoshop LAB Color : The Canyon Conundrum and Other Adventures in the Most Powerful Colorspace. This book is a runaway bestseller and I know that at least some dgrinners have it.
What about organizing a reading group to help us all get through it? We can rea a chapter a week (or two weeks?) and share questions applications of the ideas to our own shots, &etc. I think we are more likely as a group to get through it with good understanding than as indivituals.
Who is up for this? Anyone?
At Rutt's request, I'll write up Chapter 14 in the next few weeks.
At Rutt's request, I'll write up Chapter 14 in the next few weeks.
Thanks! How cool is that? We're almost finished. One more chapter -- 15, which I think Edgework has agreed to do. Two chapters I'm inclined to ignore -- the one about the client wanting it one color and then one about the universal exchange standard.
I have some really stupid questions, where can I ask them anonymously. Things I didn't understand in the first place and don't understand now.
For instance: the numbers...............I have no idea how anyone sees this zero number thing.
At the bottom left??? I have learned to use those numbers in my own way, but I am fudging. I have a lot of stupid things like that that I don't know.
I noticed on Rutt's explanation on my challenge thread, I thought, "uh, I don't want Rutt ever to know this, and I am sure I did something close to what he is talking about, but I have no idea "what he is talking about".
That kind of thing. Where can one be stupid and not be told to do homework and stuff. I remember David, way back at the beginning asking about why his numbers didn't match Rutt's, and I did not understand the answer, so "search" would not be an option now, I still would not understand.
I, uh, did my dog photos, which I am sure you all rushed to see here
This one is not really that stupid, but it does show that you haven't cracked those two books I know you have.
The numbers can be viewed with the Color Sampler Tool. When you an image is in LAB, these numbers will naturally be represented in LAB. If A=0, and B=0, the sampled color is color neutral. If your image is not in LAB mode, you can still find out the LAB values of a particular point by reconfiguring the Info palette to show LAB as one of the two readouts. Your sample points won't show LAB values, but if you mouse over a particular point, you'll be able to see them.
I have some really stupid questions, where can I ask them anonymously. Things I didn't understand in the first place and don't understand now.
For instance: the numbers...............I have no idea how anyone sees this zero number thing.
At the bottom left??? I have learned to use those numbers in my own way, but I am fudging. I have a lot of stupid things like that that I don't know.
I noticed on Rutt's explanation on my challenge thread, I thought, "uh, I don't want Rutt ever to know this, and I am sure I did something close to what he is talking about, but I have no idea "what he is talking about".
That kind of thing. Where can one be stupid and not be told to do homework and stuff. I remember David, way back at the beginning asking about why his numbers didn't match Rutt's, and I did not understand the answer, so "search" would not be an option now, I still would not understand.
I, uh, did my dog photos, which I am sure you all rushed to see here
I found this forum, yesterday, and I am fascinated by this thread. I am at somewhat of a disadvantage because I am working with the GIMP and everything doesn't translate directly (or easily, for me).
Here is a shot I have tried to apply some of the LAB concepts with. I often go way too heavy-handed with the LAB adjustments, so I tried to hold back some, this time. How am I doing?
The photo was taken by my son with a little Canon 2 MP P&S camera. It was at the Oregon coast in the spring.
I would like to see the sky with more constrast so the clouds come out more. You did a good job on the rocks. The white in the waves is blown out, and the water in the foreground just doesn't look right. I don't know if there's too much contrast or too much sharpening.
What did you do to it? Can you do a screen grab of your curves?
Also, if you can post smaller images that would help to compare them side by side (or top to bottom).
Thanks, DavidTO and cunparis. I only realized they were too large after I posted them.
I am working in the GIMP and it seems to lose the settings after you accept them. But, here is another stab I took at the same shot. I steepened the curves of the A and B layers and applied sharpening to the L layer. The colors look off, but the overall effect is nice.
I still don't understand how to apply a "good" curve to the L layer. I know how to do the curve, just not how to select the points to raise and lower.
You didn't say if you're reading the book or not, if you don't have the book I highly recommend it. Steepening the curves is good, but in later chapters he explains how to bring out certain colors.
For the L curve, ask yourself if there is an area that you'd like to have more constrast than the rest of the photo. It could be that the entire photo is interesting and there is not one area that should be singled out. When you play with the L curve, you make one area more interesting but you sacrifice other areas in the process. So if you bring out the rocks, you might darken the water and sky and blow the water. If you bring out the water maybe everything else is darker. Sometimes you don't care, like if you have a bee on a flower, it's could be ok if the background gets darker. All depends on the image. For your image, I'm not sure, I'll make an attempt tonight from home.
I'm on chapter 16. The book has been fascinating. I've learned more about color correction than from anything else. I never knew how to use curves. Well I still don't but at least I know how to experiment to learn. I now need to go back and do the exercises and practice on the example images on the CD..
Hmm... I have not had any problem understanding the LAB book. But know am I at chapter #14 and its a bit hard to understand. I hope that the dgrin Lab-topic will add the 14th topic soon.
Hmm... If I have a picture, let say of a landscape with a blue sky, and the sky is a bit dull and need to be bluer. WHY use the Red channel? Maybe Im a bit lost in RGB color space. But the compliment color for Red is not Blue - right? Well, im sure I understand if I read the chapter again and again. Bu I want to get to chapter 16! So i can begin at my Professional Photoshop book!
One more question. In this picture. The Blue channel have most detail, right? And should be used for channel blending to get detail. This picture have I already blended the Bluechannel in Lighten blending-mode. Becous the cute cow where abit dark. I can post a high res if wanted, or the RAW file.
I hope everyone understand my english. If not, tell me and I will explain what I meant.
Hmm... I have not had any problem understanding the LAB book. But know am I at chapter #14 and its a bit hard to understand. I hope that the dgrin Lab-topic will add the 14th topic soon.
Hmm... If I have a picture, let say of a landscape with a blue sky, and the sky is a bit dull and need to be bluer. WHY use the Red channel? Maybe Im a bit lost in RGB color space. But the compliment color for Red is not Blue - right? Well, im sure I understand if I read the chapter again and again.
Timely question. I'm working on the chapter 14 writeup for posting on dgrin now. I will hopefully be done in a few days. I suspect lots of folks have had trouble with chap. 14 (I did). It's got some pretty new concepts in it and I don't think it's one of his better written chapters. I will explain all your questions (hopefully) in the full writeup of the chapter. The reason for using the red channel to enhance the contrast in skies has absolutely nothing to do with color. It's because the red channel has the most contrast of the three RGB channels and the right kind of blend can take advantage of the contrast in the red channel to add more contrast to the sky. I'll have more explanation and examples in my writeup for chap. 14 soon.
Timely question. I'm working on the chapter 14 writeup for posting on dgrin now.
Great, and thank you. I must say this is the best way to read a book on! If someone is not understanding something - just post the question in a forum and you will get an answer within an hour, mostly.
I hope it will be more book topics like this one. Maybe when Dan's Professional book arrives. ;-) Must say, I realy love LAB. Only thing I miss in it is the neautral pipet in the curves/levels. But its in ACR anyway.
Great, and thank you. I must say this is the best way to read a book on! If someone is not understanding something - just post the question in a forum and you will get an answer within an hour, mostly.
I hope it will be more book topics like this one. Maybe when Dan's Professional book arrives. ;-) Must say, I realy love LAB. Only thing I miss in it is the neautral pipet in the curves/levels. But its in ACR anyway.
Hmm... If I have a picture, let say of a landscape with a blue sky, and the sky is a bit dull and need to be bluer. WHY use the Red channel? Maybe Im a bit lost in RGB color space. But the compliment color for Red is not Blue - right?
The red channel is good for contrast for blue skys with white clouds because the blue sky has very little red and the white clouds need a lot of it in order to be bright white. The blue sky often is a bit cyan and so may have a little green. But it cyan is the opponent of red, so you won't find any red there. Only at sunset, sunrise, in rainbows, &etc.
Green is good for contrast in faces because faces are basically red. Sometimes the blue channel has stuff you want, but usually you can just take the green for a good B&W of a face. Why green and not blue? Maybe someone else knows.
The red channel is good for contrast for blue skys with white clouds because the blue sky has very little red and the white clouds need a lot of it in order to be bright white. The blue sky often is a bit cyan and so may have a little green. But it cyan is the opponent of red, so you won't find any red there. Only at sunset, sunrise, in rainbows, &etc.
Green is good for contrast in faces because faces are basically red. Sometimes the blue channel has stuff you want, but usually you can just take the green for a good B&W of a face. Why green and not blue? Maybe someone else knows.
The blue channel can be a problem because of artifacting concerns…. when it comes to jpegs. Depending how compressed they are. If the blue channel is clean, you can blend with that, or do a 50/50 blend with the green and blue into the composite channel. The blue has more contrast than the green when it comes to faces. The green is most of the time smoother (looks better).
Please don't use invisible (white) ink. Some of us use white backgrounds and white on white is invisible. Please just use default text colors unless you have a reason to do otherwise. Thank you.
For anyone who has missed it, we got a great summary of Chapter 14 from JFriend. This isn't an easy chapter and John has done a really fine job with it. His results are very dramatic and show just how valuable it is to separate the treatment of color and contrast in your mind. When I took his class, Dan has very adamant about making sure we understood this.
TODO:
This leaves 4 chapters unsummarized.
Pathfinder has spoken for chapter 11 and I even saw a rough draft before Xmas.
I think Edgework volunteered for 15 some time ago. Maybe I'll try to collaborate with him to get it done.
Chapters 10 and 13 are a little dry, even by Dan's standards. I was thinking we'd skip unless Edgework or someone else can make a case.
Finish the portrait assignment. A handful of people did some of these before Xmas but we never actually figured out how to pull it together and comapre them in (semi) real time. I'd love to get this to happen. Thoughts?
Comments
I've followed the LAB Canyon Conundrum book discussions and tutorials here for a while. I've been enlightened by a few of the techniques mentioned so I figured it a good investment to buy the book.
I can say this site and the tutorials that Rutt, et.al. provided have prompted me to go and buy the book. It'll be here tomorrow.
Dan really needs to give you guys some kudos for what you've done here.
that's all I wanted to say.
I have a confession to make.
I like LAB - used it tonight to correct an order for one of our pros. Will make a post in my blog about it soon.
Somewhere, I'm sure there's a teeny-tiny serving of Crow waiting for me. :eat
:food
Portfolio • Workshops • Facebook • Twitter
And where is this blog?
thanks rutt for all your encouragement to get this book. i've stopped buying application specific books for some years. a lot of them are so much rubbish to be thrown out with the next release of photoshop. this is a lovely book though. nice stock, well produced, and more to the point, something that'll be relevant for many years, and margulis writes very well.
i'd encourage others to buy the book. i've only read the first 2 chapters but there's a lot of subtilties to the "recipes" he gives. in fact, his recipes are a bit of a pedagogical carrot. i'm really enjoying reading it. read chapter 1 and said, ahhhhh, so *that's* the canyon conundrum...
margulis obviously knows a lot about colour. and even if you don't want to use some of his techniques or become a total convert to LAB forever and ever, it's worth reading this book to get some insight into colour in general. the more technical sections are definitely worthwhile persisting with.
i think dan should give rutt a commission. ;-)
Uhm look at the bottom of nearly every SmugMug page for a linky
Portfolio • Workshops • Facebook • Twitter
Dgrin FAQ | Me | Workshops
Pathfinder is singed up for Ch 11 and I have a draft from him. Edgework is signed up for Ch 14. That's all I know about that's outstanding. After the holidays, I'll try to reawaken the Ch 16 practice bakeoff, but this just doesn't seem to be the right time of year for it.
Is one of the ones that isn't spoken for interesting to someone?
I just got the book for Christmas. So far I'm loving it, especially the second half of each chapter. I recently finished the Total Training for Photoshop CS (about 25 hours of video) and then started trying to extract some backgrounds and couldn't do it. That's one of the things that is really interesting about lab for me. I'm only in chapter 2 so I couldn't wait, I had to read the excellent summary by Rutt. When I first looked at the a & b channels they were so grey I thought there wasn't enough there to make masks. Rutt, you've explained it very well with your examples.
I got turned on to lab because I posted a message on dpreview about my workflow and someone said something like "that sounds complicated for each image, it would be must faster with lab". Since then I've been playing around and really having fun. Now that I have the book and this site I'm really looking forward to doing new things.
I'm curious, how many of you use lab as your primary colorspace? I mean you save your 'master' PSD file in lab and only convert to RGB to save as JPEG? I use a lot of adjustment layers and switching back and forth between lab & jpeg isn't possible, so I'm curious how people handle this. It sounds better to just stay in lab. Is it feasible?
Thanks everyone for a great resource!
-Michael
Chris
Detroit Wedding Photography Blog
Canon 10D | 20D | 5D
I've used photoshop a bit in the past but never learned advanced things, I always used basic stuff. So there were some things that I knew in the training, but what was interesting was I was constantly learning new tricks and tips. Sometimes I fast forwarded through boring parts but overall I think the Total Training program is excellent. Overall I learned a lot of theory and practical uses, like the channels which I had never used before and the blend modes. I find now that learning new techniques is much easier now that I have a good foundation! For example I'm picking up l*a*b really quickly, and am getting good at extractions using channels (in lab mode).
-Michael
Yeah, it's like once you learn spanish, italian is easy!
Dgrin FAQ | Me | Workshops
i work at an ad agency where everything ends up in CMYK. However, most of our images come in RGB. I have sometimes wound up saving three different PSD files, LAB, RGB and CMYK as a reflection of the actual sequence that I go through, LAB being the initial step with all the heavy lifting and either RGB or CMYK (or both) reserved for the types of subtle changes that aren't always convenient in LAB (which I sort of see as the atomic bomb of emage enhancement.) Which is not to say that LAB is incapable of subtlety. But If a face has overall good tone, but the highlights are pulling a bit yellow, there would be no point in trying to make that move in LAB since CMYK is perfect for such a shift.
—Korzybski
At Rutt's request, I'll write up Chapter 14 in the next few weeks.
Homepage • Popular
JFriend's javascript customizations • Secrets for getting fast answers on Dgrin
Always include a link to your site when posting a question
Thanks! How cool is that? We're almost finished. One more chapter -- 15, which I think Edgework has agreed to do. Two chapters I'm inclined to ignore -- the one about the client wanting it one color and then one about the universal exchange standard.
For instance: the numbers...............I have no idea how anyone sees this zero number thing.
At the bottom left??? I have learned to use those numbers in my own way, but I am fudging. I have a lot of stupid things like that that I don't know.
I noticed on Rutt's explanation on my challenge thread, I thought, "uh, I don't want Rutt ever to know this, and I am sure I did something close to what he is talking about, but I have no idea "what he is talking about".
That kind of thing. Where can one be stupid and not be told to do homework and stuff. I remember David, way back at the beginning asking about why his numbers didn't match Rutt's, and I did not understand the answer, so "search" would not be an option now, I still would not understand.
I, uh, did my dog photos, which I am sure you all rushed to see here
http://dgrin.com/showthread.php?t=26019
were done almost entirely in LAB (blk and white started in LAB), without this stuff that I don't understand at all.
So, I can get by, fast and dirty, knowing little, but I feel a bit stupid.
ginger
The numbers can be viewed with the Color Sampler Tool. When you an image is in LAB, these numbers will naturally be represented in LAB. If A=0, and B=0, the sampled color is color neutral. If your image is not in LAB mode, you can still find out the LAB values of a particular point by reconfiguring the Info palette to show LAB as one of the two readouts. Your sample points won't show LAB values, but if you mouse over a particular point, you'll be able to see them.
Here is a shot I have tried to apply some of the LAB concepts with. I often go way too heavy-handed with the LAB adjustments, so I tried to hold back some, this time. How am I doing?
The photo was taken by my son with a little Canon 2 MP P&S camera. It was at the Oregon coast in the spring.
Thanks, Tim
Can you post them a tad smaller so they fit on most screens?
Dgrin FAQ | Me | Workshops
What did you do to it? Can you do a screen grab of your curves?
Also, if you can post smaller images that would help to compare them side by side (or top to bottom).
I am working in the GIMP and it seems to lose the settings after you accept them. But, here is another stab I took at the same shot. I steepened the curves of the A and B layers and applied sharpening to the L layer. The colors look off, but the overall effect is nice.
I still don't understand how to apply a "good" curve to the L layer. I know how to do the curve, just not how to select the points to raise and lower.
Anyway, here is my next attempt:
For the L curve, ask yourself if there is an area that you'd like to have more constrast than the rest of the photo. It could be that the entire photo is interesting and there is not one area that should be singled out. When you play with the L curve, you make one area more interesting but you sacrifice other areas in the process. So if you bring out the rocks, you might darken the water and sky and blow the water. If you bring out the water maybe everything else is darker. Sometimes you don't care, like if you have a bee on a flower, it's could be ok if the background gets darker. All depends on the image. For your image, I'm not sure, I'll make an attempt tonight from home.
I'm on chapter 16. The book has been fascinating. I've learned more about color correction than from anything else. I never knew how to use curves. Well I still don't but at least I know how to experiment to learn.
Hmm... If I have a picture, let say of a landscape with a blue sky, and the sky is a bit dull and need to be bluer. WHY use the Red channel? Maybe Im a bit lost in RGB color space. But the compliment color for Red is not Blue - right? Well, im sure I understand if I read the chapter again and again. Bu I want to get to chapter 16! So i can begin at my Professional Photoshop book!
One more question. In this picture. The Blue channel have most detail, right? And should be used for channel blending to get detail. This picture have I already blended the Bluechannel in Lighten blending-mode. Becous the cute cow where abit dark. I can post a high res if wanted, or the RAW file.
I hope everyone understand my english. If not, tell me and I will explain what I meant.
Timely question. I'm working on the chapter 14 writeup for posting on dgrin now. I will hopefully be done in a few days. I suspect lots of folks have had trouble with chap. 14 (I did). It's got some pretty new concepts in it and I don't think it's one of his better written chapters. I will explain all your questions (hopefully) in the full writeup of the chapter. The reason for using the red channel to enhance the contrast in skies has absolutely nothing to do with color. It's because the red channel has the most contrast of the three RGB channels and the right kind of blend can take advantage of the contrast in the red channel to add more contrast to the sky. I'll have more explanation and examples in my writeup for chap. 14 soon.
Homepage • Popular
JFriend's javascript customizations • Secrets for getting fast answers on Dgrin
Always include a link to your site when posting a question
Great, and thank you. I must say this is the best way to read a book on! If someone is not understanding something - just post the question in a forum and you will get an answer within an hour, mostly.
I hope it will be more book topics like this one. Maybe when Dan's Professional book arrives. ;-) Must say, I realy love LAB. Only thing I miss in it is the neautral pipet in the curves/levels. But its in ACR anyway.
I the mean time, don't be afraid to post questions. Some of us do know the current edition of Professional Photoshop pretty well.
The red channel is good for contrast for blue skys with white clouds because the blue sky has very little red and the white clouds need a lot of it in order to be bright white. The blue sky often is a bit cyan and so may have a little green. But it cyan is the opponent of red, so you won't find any red there. Only at sunset, sunrise, in rainbows, &etc.
Green is good for contrast in faces because faces are basically red. Sometimes the blue channel has stuff you want, but usually you can just take the green for a good B&W of a face. Why green and not blue? Maybe someone else knows.
The blue channel can be a problem because of artifacting concerns…. when it comes to jpegs. Depending how compressed they are.
If the blue channel is clean, you can blend with that, or do a 50/50 blend with the green and blue into the composite channel. The blue has more contrast than the green when it comes to faces. The green is most of the time smoother (looks better).
TODO: