Sort of Newbie with Some Stupid Questions

[mikemeister_admin]

Apologies up front if this becomes a long post or if I really am too dumb to understand.

I'm a semi pro photographer in the UK who started to use Curvemeister to help me do basic adjustments. So far so good. I then bought Dan Margulis' book Professional Photoshop - tried to read it and put it on one side because I didn't understand a word. Gradually as my knowledge of Photoshop has increased i've come back to it and I am just starting to grasp some of the power that I could possibly unleash by using curves in different channels and colour (or color for our American friends) spaces.

Again so far so good and I have even considered enrolling on one of Dan's courses in the USA to try and understand things more. However, reading around the internet it seems that a number of well respected colour professionals are critical of Dan's approach for not being scientific and some say that he is just downright wrong on many points (they often quote the 8/16 bit argument amongst other issues see http://brucelindbloom.com/index.html?DanMargulis.html). So first question is he wrong on many points and if so why should I consider paying a lot of money to go on one of his courses?

Now onto a  more specific question If I have an RGB image open in Photoshop then use Curvemeister to make say a LAB L curve correction I'm presuming that it doesn't actually change the colour space to LAB then back again? I'm told that changing colour space can introduce loses so does working totally within Curvemeister minimise these loses (if you see what I mean)? What I'm saying can I have an RGB image open, make CMYK or LAB curve corrections, but to all intents and purposes not introduce the losses that would normally occur by changing colour space? Secondly following on from that is there any way to have Curvemeister sample point display the CMYK numbers whichever curves I'm using?

As i said at the beginning sorry if this is a bit dumb but i'm kind of floundering. Setting highlight, shadows and neutrals I understand but getting under the hood and using specific color spaces for specific tasks and making specific correction curves in particular channels as well as plate blending is a totally new ballpark. If anyone can point me to good info on the net I'd be grateful.

[ggroess]

Greetings!

While I may not be able to answer all of your questions I will take a shot at a few.

Having read Dan's book and then re-read it..I can tell you that he can be a lightning rod for many people. He challenges people to be better and re-introduces concepts that many people thought were long gone with the shift from film to digital.  All that said you certainly will learn something in one of Dan's courses.

Personally, I find scientific color to be a bit of a problem.  I think that if you lose sight of why you are mucking about in the image and shoot for perfection you just might lose an opportunity to create something unique.  As for bit depth, I think that most people will agree that more is better.  Having just typed that I can think of a few cases where that is almost certainly dead wrong.  Who you choose to follow is less important than what you choose to use in your images.

Using Lab or any other color space in CM can make adjustments that are not possible in RGB and vise versa.  So from that standpoint yes losses are introduced.  Remember that Lab can create a color that cannot be reproduced in any other space.  When you make changes using CM you make the changes as though actually working in the color space but you do not change color spaces.  I'm sure Mike can explain the technical side much better than I. 

You can set the hue clock to display CMYK by clicking on the little arrow in the upper right corner of the display and changing the settings.  See shot 1 below.

While you sound like you understand the basics you might want to look at the Curvemeister Class.  In six weeks of CM 101 class we cover the basics and so much more..  We cover creating masks using any channel in CM and copying them to a adjustment layer in PS.  Using CM to boost color, contrast, and sharpening effects.  You also get a lot of experience correcting class and class members images.  We strongly encourage students to post images to the class boards so everyone can take a swing at the corrections. 

We recognize that CM is a tool for your use and our goal is to make it easy and fast for you to add it to your work flow.  We use other techniques and tools as a part of the class with the goal being superior image correction not necessarily scientific or technical perfection.

Please continue to ask questions and we will do our best to answer them.
Greg

[derekfountain]

Again so far so good and I have even considered enrolling on one of Dan's courses in the USA to try and understand things more. However, reading around the internet it seems that a number of well respected colour professionals are critical of Dan's approach for not being scientific and some say that he is just downright wrong on many points (they often quote the 8/16 bit argument amongst other issues see http://brucelindbloom.com/index.html?DanMargulis.html). So first question is he wrong on many points and if so why should I consider paying a lot of money to go on one of his courses?

Now onto a  more specific question If I have an RGB image open in Photoshop then use Curvemeister to make say a LAB L curve correction I'm presuming that it doesn't actually change the colour space to LAB then back again? I'm told that changing colour space can introduce loses so does working totally within Curvemeister minimise these loses (if you see what I mean)? What I'm saying can I have an RGB image open, make CMYK or LAB curve corrections, but to all intents and purposes not introduce the losses that would normally occur by changing colour space? Secondly following on from that is there any way to have Curvemeister sample point display the CMYK numbers whichever curves I'm using?

It's all a lot more subjective than you might think. When someone says "I want to improve this image" they're not telling you anything concrete. It all depends on their taste and objectives. I appreciate it's confusing; the key is not to take anyone's word as gospel. There are a number of high profile and well recognised experts in the field of image correction and they all have their views and methods of getting to what they feel is the "right" endpoint. What you ultimately want is your own method of getting to your own "right" endpoint. Once you're there, anyone else's opinion is moot.

It makes sense while you're learning to pick one expert and learn from them, up to the point where you start to disagree, or want to head towards your own endpoint as opposed to theirs. You appear to have made a connection with Dan Margulis's work, and Dan would therefore probably be a fine leader to follow. It never hurts to read up on counter opinions so you keep a balanced view, but if you really don't understand the arguments enough to form your own opinion, you have to set your stall somewhere. No one is right or wrong, so you could do a lot worse than follow Dan's advice and methodologies to the letter and go with his concept of "right" until you know better.

Your Curvemeister question is an easy one: Curvemeister uses its own internal lossless colourspace conversions. You can convert and curve all you like inside CM and you won't lose anything. As for the CM sample point, yes, you can change that. There's a drop down menu accessed from the little icon in the hue clock's top right corner.

I couldn't comment on Dan's course. I guess you'd learn an awful lot from it, provided it doesn't start off too fast and leave you behind at the first lesson. You do, however, have a no-brainer option: take the CurveMeister class. It costs pittance and starts at the very beginning. You sound like the perfect class participant: the amount you learn from it will absolutely blow you away.

[mikemeister_admin]

I would strongly recommend the CM course - it has taught me an enormous amount and has the great advantage of interacting with like minded people.  One learns from others as much as doing things oneself.  I too (attempted to) read Dan's Canyon book - it took about 3 goes to begin to understand what he was saying.  Reading the book alone gives one nothing to bounce off and one never quite knows if one has the right idea or not, you need others to help you and the input from Mike and Greg is really good, even when 'daft' questions are asked!
All the best
Chris

[mikemeister_admin]

Didn't your Mom tell you not to believe everything you read? Well, for the Internet - don't believe anything you read. Really! Try it out for yourself. Test everything. If you are a mathematician, really get into the pixels and the numbers. If you are a photographer, use what makes your photographs better and ignore the rest. The key criteria, if you can't see the difference or the improvement, it doesn't matter.

So, try all the color spaces. Convert back and forth. Try 8/16/32 bits. Check everything out on your calibrated monitor and print everything. If it makes your pictures better, use it. If it doesn't, ignore it. If you can't tell, it doesn't matter.

I've never taken a photography or Photoshop class in my life. I've taught a few, but never taken one. I have tested out a ton of techniques, ideas, and products. I use a lot of Dan's ideas and techniques because they make my pictures better. I don't use everything, as some of it doesn't fit my needs and/or I can't see a significant difference. CM helps me do many of those techniques easier and better. That's why I use it too.

For example, I do most of my editing in LAB mode. That means that the number of bits doesn't really matter. I save everything in 8 bit sRGB though. Yes, I've been following the 8/16 bit debate for many long years. I have also tested it out many times. I can't SEE any difference between the two. (No surprise, as most monitors don't show more than 8 bits and many of the new ones are only 6 bit.) So, I save in the smaller file size. I don't really care what anyone else says about something I can't SEE.

BTW, check out his book on "Photoshop LAB Color". It's great.

Clyde