Sunday, January 27, 2008

Learning - Scott KelbyTraining

One thing leads to another on the net and I somehow ended up at and found that for $200 a year, one could basically view as many 4 - 8 minute lessons as one cared to from around 20 instructors, includi Scott himself, JP Caponigro, Don Margulis and a variety of others.

George Purvis had written me a while ago about the possible advantages of L.A.B. colour and recommended one of Don's books. Well, here I was on a site which offered video lectures on a variety of subjects including Don talking about L.A.B. Better yet, without even signing up you could try the first three lessons of any topic by any instructor for completely free - impressive.

I watched Don's first three lessons on L.A.B., enough to whet my appetite. I brought an image into L.A.B. via Image/Mode/LAB and tried an adjustment curve - wow - what power. Where in RGB one adjusts all the colours in creating a curve in the default setting, here you start out with L or lightness and can basically increase contrast without increasing saturation - an important tool. Better yet, you can flip in to A and B and use the same kind of contrast increasing S curve to separate colours further, without changing the overall balance of colours - somewhat akin to increasing saturation, but not quite.

In a matter of a few minutes I was able to dramatically improve a few different images, in what would have taken a lot more time back in RGB - hmmn, maybe this guy is onto something. No doubt you can overdo it but by using the opacity slider in the curves adjustment layer, you can tune it back and forth between zero and 100% opacity and choose what looks best - almost always less than 100% as it's hard not to go too far with such powerful controls.

Will this change my image editing, am I abandoning working with RGB for a life in LAB?
Certainly not any time soon - I might well buy his book, I may even break down and spend the $200 for a years subscription to ScottKelbyTraining, it's hard to believe I won't geet $200 worth of information out of these famous instructors.

A little more careful perusal of the site indicates they are just getting going and don't have a huge selection of lessons at this time - but I just purchased my first studio flash for some still life work and I see they have a studio lighting series - handy. Some of their listed authors don't even have lessons listed yet - eg. J.P. Caponigro. I guess this would make it a leap of faith that more will be coming and that by year end there will be plenty - but I think the concept is sound and the price fair and I don't see a reason for it not to fly.

Anyone have any experience with it yet - care to comment? Anyone who hasn't signed up, what do you think of the concept, the value, the practicality?

Getting back to the L.A.B. thing, the top image is after using LAB, the bottom before. I was able to warm up the sun, without yellowing the whole image and without needing masks, add a little punch without taking the colours over the top. I'm sure I could have done the same in RGB, but probably not nearly as quickly - there may be something to this LAB - I'll let you know as I work on real images whether it gains or fades.


yz said...

How about applying an adjustment curve to an RGB image and setting the layer blending mode to luminance?

I think it should give the same result as working in LAB mode, as far as increasing contrast without increasing saturation is concerned.

pitchertaker said...

And how about merely applying an USM with setting of radius around 60 and percentage about 15-20. I copied your example image into CS2, and applied the USM as above, and achieved what looked to be the same results. I realize that monitor views are not really a way to judge any process, but the USM method works well creating internal contrast within an image.


George Barr said...

Both of the above suggestions are ones that I regularly use. My impression was that I was getting results different from these effects but only time will tell.

Matt Kloskowski said...

I think the concept, the value, and the practicality are wonderful. Of course I'm biased because I work there :)
On a serious note, I'm amazed at Dan's stuff. I have learned so much from him and it's refreshing to see his style. He's actually just finished another course for us and it goes into even more depth. The one that's there now is the "intro" course if you can believe that. Personally, watching Dan's material has made me better. I can't say I always use the techniques but when I do the results are great. And when I don't I typically get to the result I want faster because I at least know more of what I'm looking for after having seen the way that Dan suggests.

Btw... congrats on your book. I'll be picking up a copy during my next trip to Borders.

Take care,
Matt Kloskowski

Nick Jungels said...

I've read the LAB book by Dan Margulis, and liked it a lot (as well as his Professional Photoshop book). He is very thorough in his discussion and talks about the pros and cons of using LAB as a working space.

Mark Higgins said...

Hi George,

I have been using the online courses with the subscription fee. I'm a NAPP member so it's 17.99 a month. So far I have taken the Lab course with Dan, setting up your studio with Andy Greenwell, studio lighting with Joe McNally and a B&W course with Katrin Eismann. This all in one month. All this information is well worth 200.00. Actually quite more. They also have some courses with Moose Peterson that I am just checking out. I consider myself an expert landscape photographer, but Moose has some unique techniques for shooting ultra wide panos

George Purvis said...

Hi George,

I'm delighted that you've found Dan Margulis' work. What you've done with LAB is just the tip of the iceberg.

A year ago, I picked up his book Professional Photoshop 5th Edition (PP5E) and it changed my thinking about images and how to get the most out of them more than I would have ever thought possible.

Until I fully understood what it meant that every image has ten channels available (RGBLABCMYK) I had no idea what I could really do with Photoshop. While I might be able to get close to the same effects working totally in RGB, "close" is not good enough for me and probably not for those interested in fine art photography either.

Over the past year I have read Dan's 500 page book cover to cover three times. And I "never" read a book more than once. The concepts Dan introduces were so foreign to my thinking that it took two readings to "get it" and a third to enjoy the content.

As photographers who print our own work, we inevitably begin in RGB with a raw converter and end in RGB when we output to our inkjet printers -- unless we are printing a book when we need to end in CMYK. In between, we work mostly in RGB until we encounter situations where another color space allows us to make changes that would be very difficult in RGB.

Here are a few things I learned from PP5E:

0. Because of visual chromatic adaptation, it is impossible to color adjust an image accurately by viewing the image on a screen, even on a calibrated monitor. However, if you pay attention to the color numbers, you can get great images even if you are color blind. [One of the great problems with Lightroom for color adjustment is that it does not allow you to measure color values at fixed points on an image. By contrast, Camera Raw does allow it. Strange.]

1. LAB is wonderful for differentiating colors in monochromatic landscapes and still lifes. Separate colors by steeping the curves in the a and b channels 10-20%.

2. CMYK is great for dialing in skin tones (Caucasian skin is always Cyan much less than Magenta less than Yellow).

3. The Apply Image command can really pop a sky when the red channel is applied in luminosity mode.

4. Faces often benefit from applying the green channel in luminosity mode.

5. Wildlife survives by blending into the background. In our photographs, we usually want wildlife to pop out visibly just as we saw it. Applying a channel where the wildlife is light and the background is dark in luminosity mode will create pop by separating the wildlife from the background.

6. The foundation of a great color photo is a black and white image. You can use a black and white version of your image to make the color image better.

7. Always soft proof your images and turn-on the out-of-gamut-warning. When the image is out of gamut, details will be lost and the image flattened. Apply your channel knowledge and bring the image into gamut. [Lack of out-of-gamut-warning is another problem with Lightroom which is made more serious by Lightroom's use of ProPhoto RGB colorspace. ProPhoto RGB makes it very easy to create out-of-gamut images for printing]

The list could go on and on.

I have a shelves full of photography books. The concepts Margulis covers in PP5E are not widely known by, understood by, or taught by photographers.

I found that the time and effort I have taken to understand and apply Margulis' way of thinking about images has changed my image processing fundamentally more than I would have believed possible.

I hope that Margulis' techniques will be come better known among photographers, especially those who understand that three or four 10% improvements make a wow.

George Purvis

Other than owning Dan's book and applying his methods when appropriate I have no involvement with his businesses or interests.

I use Lightroom every day and think it is a great program... but a couple of improvements would make it better. The capture sharpening improvements they put into 1.3 are excellent and have mostly replaced my Bruce Fraiser derived CS3 capture sharpening actions. On the other hand, color correction is Lightroom's weakness that drives me to CS3.

After five years, I gave up serious photography in 1965 when I lost my darkroom. College got in the way. I started again in the '90s when Photoshop gave me back a darkroom. Half of an image happens after the capture. I have to be part of that second half of "the seeing".

Nick Jungels said...

George (and George) -

I don't mean to diverge too much, but isn't ProPhoto RGB intended to be used only so that you capture the color space of the camera vs. monitor vs. printer?

I've just recently watched the "From Camera to Print" series of videos found on Luminous Landscape, and both Jeff and Michael seemed to imply that as the reason that they save their images in the ProPhoto color space.

This makes sense to me technically, though I understand the argument that I need to worry about out of gamut colors. However, I need to worry about it no matter what (don't I) as when I print the dynamic range of the paper being utilized will affect my final output.

George Barr said...

Probably the best person to address this question to is Jeff Schewe who is often on the Luminous Landscape forum, but my understanding is that the latest pigment printers do exceed the limits of Adobe RGB colour gamut in some places and thus you are artificially limiting yourself if you use the more restricted colour space. Perhaps more importantly, it can be important to decide how you are going to deal with out of gamut colours. For example, if you are photographing a bright yellow flower, the intensity of the yellow can easily exceed adobe rgb and what your printer can do. Your options include simply taking every yellow that is beyond the limits of the colour space of the printer and have it set AT the limit, but doing this there will be no variation in the saturations of the yellows and you will end up with a bright yellow blob. Most of us feel it is better to reduce the intensity across the board so that even in gamut yellows are reduced proportionately - now you get the more natural looking variation in saturation and thus can see detail in the flower - a compromise to be sure but a better one than the other of lumping the yellows together at the limit of the colour space.

What I have described above is the difference between perceptual and relative colorometric - though I can never remember which is which and don't care since I want to deal with out of gamut colours rather than have Photoshop make assumptions about what I want done.

There are times when you might not care about the difference between a yellow that is at the limit and one that is beyond, you just need both to be as yellow as possible - in which case proportionately reducing saturation so you are within the gamut is not the way to go - setting both at the limit would be best.

George Purvis said...

Hi Nick,

You're exactly right. ProPhoto RGB is what should be used to include the colors of cameras, monitors and printers (although I've seen minor cases where even ProPhoto RGB is not large enough). Lightroom's use of ProPhoto RGB is a good thing.

The problem as I understand it is that if you Develop the image in Lightroom, you're editing colors in ProPhoto RGB space and can create out-of-gamut colors that degrade the print without any out-of-gamut warnings.

Printing from Lightroom is convenient especially if you don't have to take a trip into CS3 to check for out-of-gamut colors.

Anonymous said...

So if you bring all your colors back into gamut, should you be printing with relative colorimetric since that doesn't change any colors and clips the out of gamut colors, OR should you still print with perceptual, which tries to maintain the relationships among colors? Great comments in this thread that are clarifying issues I've been thinking about the past few nights.

George Barr said...

My understanding of it is that if all the colours are in gamut (ie. within the gamut of the printer), then it doesn't matter which rendering intent you use as neither affects the output if there are no out of gamut colours.

Which is of course the whole point to soft proofing and out of gamut warnings.