Untitled Document

For those who have been interested in discussing my future projects, here's an opportunity, because I have to make certain decisions within the next week.

On Wednesday, Scott Kelby (head of NAPP, which runs Photoshop World; the bestselling author by far in the entire computer field, let alone Photoshop) announced at

http://www.kelbytraining.com/online an ambitious on-line training/DVD program through his Kelby Training subsidiary. The idea is an unlimited-use subscription service. He has contracted with many of the leading teachers in the field to produce 2-hour classes.

Once they're up (and I'm not sure of how many are up as of today, I know they're planning at least 30, with many more to come) you pay a subscription fee of $199/year (some kind of monthly plan available too, I think) and then you can watch as many of these classes as you like, as many times as you like. Eventually, as I understand it, these lectures will be released in DVD format.

I'm participating in this program; it was a secret until Wednesday, hence the lack of notice to the list. My first two videos are done, but they haven't been posted yet AFAIK, and I haven't seen them so I can't comment on how good or bad they are. Here's the issue, though.

I contracted with Scott to deliver five of these presentations. My calendar is full already for the entirety of 2008. These three videos (in addition to the two I've done already) are going to be done in early January, period. That's the only time I have available. So I will be working in earnest on these programs starting next weekend. I have changed my mind about what topics to do and if anybody thinks I should do something different, now would be a real good time to speak up.

The idea of the series is to simulate a live two-hour lecture. As, unlike other instructors, I do lectures at Photoshop World in two-hour increments, it's a convenient format because it enables me to expand on what I've previously done. In a truly live lecture, a lot of time gets wasted when I make mistakes and have to start over, when required announcements are made, when somebody's cell phone rings, and the like. All these things get cut out in a video, plus I don't have to stick strictly to two hours if I don't wish to.

The initial plan was, therefore, to do two videos on LAB, two on curves, and one, a popular yet less-attended session called "CMYK for Cowards" which attempts to explain, to RGB-oriented folk, what benefits CMYK correction offers, as well as a discussion about how to get the best results from commercial printers. The first two were to be the curves lecture, based largely on the four-hour preconference sessions, somewhat beginner-friendly, that I did until this year at each Photoshop World. Those were hot-selling lectures, usually filling up a month or more in advance, because curves are so fundamental to quality.

This year in Boston, we substituted a preconference session on "LAB in Depth" because of considerable demand for something at that level. Feeling that I should also be offering a generalized color correction with curves session, we decided to do a two-hour session in the main conference called "Color by the Numbers," explaining that philosophy and using an RGB orientation.

This turned out to be a mistake. My other two-hour session (the picture-postcard workflow) and the LAB in Depth session were well received; I knew this at the show because I saw the evaluations. I got a lot of oral compliments about the by-the-numbers session, but did not have an opportunity to review the written evaluations, which showed up a couple of weeks ago. The session didn't crash and burn; a lot of people really liked it. But it was too difficult for many people, and others complained that they didn't come to Photoshop World to hear me talk about things this elementary. Also, it was apparent that there was defects in delivery that had nothing to do with content. On the whole, this session got the worst set of evaluations since I started at Photoshop World several years ago.

Deciding to offer that session was foolish. At least three other Photoshop World instructors are qualified to present that material. They may not be up to my speed with curves but they're all better speakers than I am. I do not intend to do this show forever, and would like to maximize the benefit of my appearances to those who are interested in what I have to say. Since "by the numbers" was a new session in Vegas, it was scheduled to be presented again in Orlando at the next Photoshop World in April. I have cancelled that session and intend henceforth to present only material that intermediate and advanced users will find challenging. The show management fully supports that decision.

For the spring Photoshop World in Orlando, I have substituted a two-hour session called "Every File Has Ten Channels".This one isn't going to be found too easy by *anybody*. It will discuss the situations where we are working in one colorspace but where the best chance for success is to make a copy, convert to a different colorspace, and steal a channel therein for use in a blend or as a mask.

That brings us back to the question of what videos to do. It's clear in retrospect that I should have consulted the list before announcing a new agenda for the September Photoshop World in Vegas. (No problem with either the LAB in Depth precon or the Five Minutes to a Picture Postcard lecture; both got terrific responses and I'll repeat them in Orlando.) But the third session, I thought it was a good idea, and it wasn't.

Similarly, I think I have a good idea now for the remaining three videos, but I'm open to arguments against.

The above events made me question my original decision to do the first two videos on curves, so instead I did two LAB sessions. One covers the basics of how the colorspace works, and curving and selecting. The second is called "The LAB Frontier". It discusses parts of LAB technique that are either commonly misunderstood in my LAB book, or else covered incorrectly or not at all. These include retouching using imaginary colors, sharpening the AB channels, an improvement to the Man from Mars method, and the power of multiplying in LAB through a blurred layer mask. If you don't like this choice of topics, too bad, it's done, and I was contractually obligated not to make it public before doing it, so I had to live without the list's advice.

Anyhow, unless persuaded otherwise, I now intend to devote the other three videos not to channel blending or to CMYK phenomena, but entirely to the picture-postcard workflow. While it's gotten a great response at Photoshop World and in my classes, a two-hour presentation isn't long enough for anybody but an expert to understand much more than that there's a series of confounding steps that work together to get an outstanding result. With six hours, there's enough time for prolonged repetition.

Also, the topic breaks nicely into three parts, and affords the significant advantage that the viewer can jump from one lecture to another without finishing the first. This is helpful because the first part, which is a by-the- numbers RGB correction for color only, takes a fair amount of effort to learn for what seems to be only a marginal gain. Its benefits become apparent later, but meanwhile it can be boring to learn something that doesn't seem to have much effect.

I have continued to test this method extensively and the results are positive. Not every image is suitable for the approach but when it works the quality gain is large. Best of all, it's a quick method--three to five minutes per image. I've tested it in competition in the advanced classes, and also against some of my own previous corrections, including some that appeared in the two latest books. It seems that I can now get significantly better results in a quarter, maybe even a third, of the images in those books. While certainly my skills have improved from edition to edition and there are always some images that I discover could have been handled better, there's never been any time in my career where I have been able to make so substantial an improvement on so large a fraction of what I've done previously.

At the moment, I'm doing the final test: I've arranged with three different photographers (two world-famous, one serious amateur) to provide me large collections of images to correct using this method. I've almost finished with the second of these. Granted that when you work this quickly some duds are inevitable, many of the results are strikingly better than those obtainable by any other known method.

I have not reconsidered my decision to cease writing books. OTOH, I was not anticipating having anything more than marginal improvements to add. As you might expect, I am somewhat tired after several months of nonstop activity. In late January I am going to rectify that with ten days on a secluded Puerto Rican beach and NO INTERNET ACCESS. So, I'll give the matter more thought then. Meanwhile, I'm going to prepare these three videos, unless anybody on this list talks me into doing different topics.

Dan Margulis

____________________________________________________________________________

"Dan Margulis" wrote:

Dan - I'm glad you did this - it is like a Margulis class with a "pause" button - very helpful. I like your current plan to do your last lectures based on the "Picture Postcard" workflow.

For other list members interested, the first LAB video is now posted. There is a minor problem with the sequence of the segments at this time. I sent feedback to the Kelby folks about. I am sure they will fix it, but if you decide to watch the video before they do, the segment "A Closer Look at the LAB Channels Themselves" should be watched right after "When Not to Use LAB."

Good stuff, by the way. I have taken both of Dan's classes and managed to pick up some new tricks already from this first video. I also am excited to see some of the work by the other instructors.

Geoff Shearer

____________________________________________________________________________

I attended all of your pre- and conference sessions in Boston, having studied seriously the LAB book and PP5 for many months beforehand. My hope was to get some views of the forest for the trees. The Boston sessions helped some, but by the nature of such conferences, I don't think they are the best environment for consolidating learning.

I have also looked at the teaser parts of your Kelby videos, thinking the ability to replay would be a great help.

In this context, I offer two suggestions.

First, it would be great if your remaining videos were at the intermediate/advanced level. As far as I can see, this is where you shine.

Second, it would also be great if you could help us understand the framework that provides the foundation and context for developing new ideas. From my own experience as a teacher (albeit of quantum mechanics, not color theory) the gap from recipes and to deep fluency is wide, and very hard to help students bridge. But you sure have a willing audience.

Dan Dill

____________________________________________________________________________

Wonderful news Dan!

I feel you are going about this the right way. Your courses have always been geared to advanced users of PS and that is the way you should keep it. What you have to offer is way better than all the "Photoshop tips" out there rolled into one. It's hard stuff to understand and harder still to harness and learn when to use but the road to excellence in any discipline is not an easy one. You are an excellent teacher and technical magician. Learning from you makes the rewards intensely gratifying.

Keep your focus on the hard stuff. Not to belittle them but, let the giggly frat boys tackle the other stuff. It what they do best and their style is fun and entertaining but not in your league. Don't feel you have to change because you are in a "new" medium.

The new workflow is something amazing which can be followed like a recipe so a lot of folks, with a little hard work, will get a lot out of it. Hopefully, jut like in my case, it will make them eager to learn more and so be more willing to dig deeper through the books and your other videos.

Just one last thought. I hoped and got my very egotistical wish by your decision not to "retire." I remember writing to you back when you first mentioned your intentions, and one of the things I suggested were video tutorials over the internet. Talk about a great Christmas present!

You are special! A great technician, a great teacher and to top it all off, a great guy. I personally can attest to this and in so doing consider myself very fortunate in being able to do so.

� ?¡Todav� ?ía te quedan muchos gigantes por derrotar Don Quijote! � ?¡A la carga!

Paco Marquez

661 McKinley

San Juan, PR 00907

787-721-8554 Studio

787-587-7384 Cel.

http://www.pacomarquez.com

____________________________________________________________________________

Dan writes :-

I would be most grateful if somebody would explain to me whats the meaning of "picture-postcard workflow"?

Thanks!

Wai-hong Chung from Hong Kong

____________________________________________________________________________

Given your positive results and the extra speed, I don't see how you could do better for us than three sessions on "the picture-postcard workflow."

Jerry Fusselman

____________________________________________________________________________

Dan--I agree completely with Geoff Shearer. It's very exciting to have these videos to look forward to and the material for the lectures sounds perfect. I'm going to look at the first LAB video right away.

(I always question the use of the phrase "picture Postcard" because-- even though that's the point of the phrase--it doesn't reflect the complicated process that gets us there!)

Sally Cohen

____________________________________________________________________________

Sally writes,

A good point, and one that I should likely have brought up earlier. "Picture postcard" is surely not the most elegant of names for a workflow. Any suggestions for a better name would be welcome--as long as they come in before my next taping.

"Picture postcard", for the benefit of Wai-hong and anyone else in doubt, refers to the type of photograph found on such postcards, or used to promote tourism. The defining feature would be that the colors are rich and pleasing, but not necessarily accurate to the actual scene. Also, one would seek extreme depth and detail, much like what is now being described as High Dynamic range images.

Applying simple LAB curves moves in the direction of a picture-postcard look, but what I was showing at Photoshop World and intend to show again in these videos goes farther, and is quick. The problem of trying for artificially pleasing color is that it's easy to make the picture garish. The techniques I'm showing try to avoid that, and are based on some new understandings of what people actually find attractive.

Dan Margulis

____________________________________________________________________________

Hi Dan,

If I understand your response, and I might not, a rhetorical question would be, if a person is doing pleasing color workflow why would they want to mess around with going to LAB when and RGB workflow would allow you to produce the same results with less steps?

My sense (and some experiences I have) indicate that any type of color workflow takes skill and that the less steps are a good thing.

Jim Rich

____________________________________________________________________________

.

On Dec 25, 2007, at 12:29 PM, Dan Margulis wrote:

It looks like you found the new title yourself:

Rich, Attractive Color [and Depth]

Laurentiu Todie

____________________________________________________________________________

.

Dan writes :-

Thanks for the explanation.

Wai-hong Chung

____________________________________________________________________________

.

Yeah, I'm digging _that_ acronym...

"RATCOD"

hehheh

/andrew webb

On Dec 25, 2007, at 11:01 AM, Laurentiu Todie wrote:

____________________________________________________________________________

Hi Everyone,

Dan wrote:

I would buy a subscription to view those videos. As a photographer I really only print my own photos and I don't often come across photographs that require a great deal of difficult colour (I'm Canadian) work. I don't feel, however, that I've developed an adequate "5minute" workflow that I can apply to each and every photo I take to get the best I can from each one. So yes, for me personally I agree with your decision to concentrate on the 'picture-postcard' workflow.

Also Dan, ever since I read that you planned on retiring from writing I've been in mourning! It always seemed to me that, with nobody to take your place, it would be a tremendous loss to the color community. It's nice to hear that you may be reconsidering.

One final question: Has peace finally come to this list? Has the 'Dynamic Trio' taken their fight somewhere else? Maybe there is a Santa Claus!

Murray DeJager

____________________________________________________________________________

Jim Rich writes,

The rhetorical reply would be why would anyone want to go to Venice, when you can see the same thing in Las Vegas and not have to pay in euros?

There may be a question of terminology here--"pleasing color" used to be shorthand for "This client is not interested in paying for quality, he's not going to insist on a remake no matter what, so don't waste a lot of time on his work." In the old days, it meant that the scanner operator made a good-faith effort to get a nice scan but nothing would be done afterward. Today, I suppose a "pleasing color" workflow would consist of Image: Adjustments>Auto Color and Save.

Some of us encounter situations where we have so many images or so little regard for quality that just applying an automated correction is appropriate. Most of us, I think, are willing to spend a little more time than that with the typical image, but probably not as much as if it were a magazine cover. At that point the question becomes how to get the best overall results given the time constraints. And that boils down to deciding which steps to omit and which shortcuts to take.

If I had to process 400 images and wanted to do it in eight hours of work time (that is, I want to do way better than Auto Color), then I would do everything in LAB 100% of the time. Granted that these originals are probably not prizewinners, LAB is much faster at bringing out "pleasing colors" than RGB is, at the cost of some accuracy. Some of these one-minute specials might actually be fairly close to what a good retoucher could get in more time, but there would be a lot of duds because LAB is not appropriate for many images. And surely, nobody would claim that such a workflow would give *better* results than spending more time.

The 3- to 5-minute workflow I've been showing is a very different animal. It is appropriate no matter how strict the quality requirements are. It isn't for every image, but when it works well it gives *better* results than anything seen in the past AFAIK.

This workflow starts with RGB manipulation and does not always move into LAB. When it does, the reason is that the correction being contemplated plays to an LAB strength and would presumably not work as well in RGB.

That's right. If the job is of infinite importance, we take all the steps that are needed. Otherwise, we condense, and the faster we want to do the work the more we cheat.

In my workflow, for example, I use a lot of Multiply blends when a curve would offer more flexibility. I say to use a single one-size-fits-all setting of the Shadow/Highlight command when needed, instead of adjusting its half- dozen internal settings image-by-image. I admit up front that these are shortcuts and that marginally better results can be obtained by doing things the "right" way. But I feel that the gains, such as they are, do not warrant adding thirty seconds to the workflow.

Moving into and out of LAB when necessary takes a lot less than thirty seconds, and adds a lot to quality.

Dan Margulis

____________________________________________________________________________

.

Way to go Dan...

I can remember in the old wet days when I had a color lab, people would come in and you knew right from get-go that they were going to ask for a redo...no matter what....Boy you hit the nail on the head...it is refreshing to hear that someone has the same outlook as I do...

Thanks, Frank

William L. (Frank) Collins

336-887-2703

IPMS 29961

____________________________________________________________________________

.

Dan wrote:

Dan, I strongly agree with your considerations. The picture-postcard workflow is so flexible, complete and result-oriented that is a sort of recipe for a-better-looking-images never seen before, and I think that six hours of video will be enough.

It covers so many different kind of images that you should consider to get a patent pending (is this the right english for this?) as Scott Kelby has done with his 7-points-system for Photoshop.

I haven't seen Scott's system but I know very well your workflow and it was really challenging and amazing in the same time to use it in your class. And the results talked by themselves.

I know that somebody may say that some steps of the color correction could be done with different techniques (more common and easy but not so efficient) but I think that the philosophy beyond is the most important part.

It is also very useful for experiencing with the single channels of an image in any colorspace. So it's the right solution for anyone everyday work.

The only things I can suggest are:

- try to make the workflow non-destructive as much as possible using the Smart Objects for integrating the Lab or any other suitable part of the workflow into the same RGB file. It's always confortable to have all the things together and in addition to this you can also add a mask to the Smart Object and choose any layer blend method you like for the final result

- the use of Smart Objects is also ideal for the Shadows & Highlights step, but unfortunately works only in PS CS3

- a good idea could be to integrate at a certain point the trick for exhalt the washed out skies before the Overlay Blend through specific moves with the Selective Color command, layer blending and so on.

I think that in this way the workflow will be more complete, powerful and editable at any time. A really unbeatable tool for producing the best image you can, not only in five minutes.

If I think that the workflow will be realeased on a video reviewable many times you want, this will be simply one of the best ideas in the last years. It should be used like a medicine to take twice a day in a lot of prepress companies and photographic studios.

Thank you Dan for sharing with us your thoughts and your genius in color correction but aboveall in making our work better through simple good ideas.

Alessandro Bernardi

____________________________________________________________________________

Well said Murray!!!!!!! After reading the non-stop pissing match for years I hope the trio you are referring have just argued themselves into a ball that will take a machine generated "thingy" to release them.Of course the "thingy" will not be valid when you try to access it. Thanx for posting the classes Dan!! Looking forward to using them! The bang for the buck is unmatched! Jim Donovan

____________________________________________________________________________

Dan,

Lots has been said about this great idea of having your techniques available on videos, so I'll just add couple of thoughts to what it has already been written.

If at all possible, make the videos easy...but not too easy! What is great of your classes is that people learn to think. You may want to keep that in mind, because IMHO produces better results in the long run, and allows to reach incredible results also for people who are not professional.

I know you said it is your intention to leave USM out of your future videos, and I know this is something that other authors have spoken and written about. But none have ever treated USM the way you teached us in the advanced class and this new methods are not to be found in any book or video on the market as this time; and trust me I have seen many of both.

You are simply the best.

David Limentani

____________________________________________________________________________

Thanks Mike!

____________________________________________________________________________

Thanks Mike...you sometimes have to just noodle around....I just subscribed to Dan's lessons...look forward to them....

William L. (Frank) Collins

336-887-2703

IPMS 29961

____________________________________________________________________________

Alessandro writes,

Questionable. I'm already leaving stuff out that should really go in (noise reduction and sharpening, for example) but for sure six hours is a lot better than the two hours with interruptions at Photoshop World.

Yes, unfortunately, as I recall, your results were often better than mine.

That's right. Almost every individual step *can* be replaced by something else. But one has to know what the desired result is. And this is what was missing previously: neither I nor anyone else realized that smashing color variation as the image gets darker was a good idea.

I will certainly recommend this option to those people whose work is important enough that they are willing to spend the extra time, and also to those who may have to pick up and edit other people's work.

I will not, however, *demonstrate* it, because I think it's an optional addon that relates more to how to operate Photoshop safely than to ultimate image quality. IOW, in the presentations I've made so far, the assumption is that time is a serious constraint. If I truly have to process 15 to 20 images an hour I must accept that a few of them will not be all they should have been. If that is truly the time constraint, I am *not* going to go back and re-edit the initial steps unless the final result is *really* bad. And in the unusual case where that happens, I think I am better off just throwing it out and beginning again than I am saving Smart Objects that I am extremely unlikely ever to need into every single file.

Your case, high-end advertising, doesn't operate under such constraints. It certainly would be nice if every image could be processed in three minutes, but you would be willing to spend much longer than that to avoid a second- rate result. So, for you, Smart Objects make a lot of sense in this workflow.

Here is where the six-hour format shines. My current plan is to have a series of at least four and possibly six images with sky problems and to discuss the various options in depth.

I think it is probably a poor substitute for a glass or two of wine at lunchtime.

Dan Margulis

____________________________________________________________________________

if it's not too late for a suggestion... I would pay full price for the video program just to see 10-15 consecutive examples of utilizing the blend-if function.. they are scattered throughout dan's work, they are excellent/unique techniques... helping someone to recognize the better options when using this method would easily be worth the price of admittance... jim bean

____________________________________________________________________________

.

Jim,

Blends IF are not that difficult if you allow yourself to think in terms of color and/or darkness. You just need to practise a little bit.

As the place where to operate the "blend if" function depends on the image, it is hard to tell you exactly what to do. Personally I usually check if what I am trying to isolate is the only thing of that color in the image, or if it is the lightest or darkest object. For example skies in a panorama are usually the only blue thing in the image, and you can use either the Blue channel or the B of LAB. The more what we are trying to isolate is separated from the rest of the image in terms of color or darkness, the easier it is to use the blend if. Once you set the range on the blend if slider, don't forget to alt+click the slider to create a transition zone.

Should you need any more advice, or if I can be o any further help, do not hesitate to contact me offline.

David Limentani

____________________________________________________________________________

.

David Limentani wrote,

Well put.

One technique from an earlier edition of PP (4E, I think) really helped me see the effects and get the settings right (if they work at all, and you could also see this case). Maybe this would be useful to the originator of the thread, also.

Before starting, you have to know where you want the blending to happen. The object is to set the sliders so that only that part of the image is affected (with the transition zone mentioned in an earlier post).

Turn off the upper layer visibility, and add a new layer filled with 50% grey. Then adjust the sliders until there's grey in the part of the image where you want the blending to happen, and no grey anywhere else. When (if) you get it, you Copy/Past Layer Style to copy the slider settings to the upper layer. You'll have to change the blending mode to the proper value. This assumes blending using Underlying Layer. A similar technique works for blending using This Layer, except that you turn off visibility of the lower layer and put a 50% grey layer in its place in the chain.

If you get grey where you want it, but grey in other (disconnected) parts of the image where you don't want it, you can eliminate those with very easy layer masks.

John Bongiovanni

____________________________________________________________________________

Dan, just finished your Kelby Training exercises…FANTASTIC…I will be going over them again, probably many times, but I had heard about the LAB Color Space but never thought it would help this much…Thank you for doing this series and look forward to the LAB Frontier sessions…The Blend If and masking are great for my purposes…I have both of your books and re-read them constantly….That’s it and thank you again for the great sessions here in the Kelby Training Series…..Frank

William L. (Frank) Collins

336-887-2703

IPMS 29961

____________________________________________________________________________

Could someone "throw me a bone" about the using Smart Objects. I am working in PSCS3, but haven't touched Smart Objects yet.

Andy Adams

____________________________________________________________________________

For the most relevant application of Smart Objects to subjects in the this discussion group, here is a good pragmatic 'bone' to start with:

Before you apply a filter to a layer, choose Filter > Convert for Smart Filters (2nd down).

Then, when you apply your desired filter the layer will open up in a manner that offers you an adjustment layer like behavior. The main advantages, AISI, are that you can close the filter, then return to it like it was, and refine later, and, that you can apply several filters to that layer independently, and return to them independently. You do get a mask as well, but at least at this time, that is only one common mask for all filters you applied to that layer.

Robert Kenedi

____________________________________________________________________________

Smart Objects use a similar logic of adjustments layer. You can retrieve them at a later time - even if you have saved and closed the file - and change their function within your image. You may for example change curve shaping in a curve adjustment layer at any time that pleases you.

So here is what you could do (this is just an example): at a certain point in your workflow, you probably have some adjustment layer. You want to flatten the image, but are afraid to loose your ability to go back and make changes, should you please to do so. Or simply you need to move to a different color space and you know that this means you need to flatten the image.

To create a Smart object (copied from PS help):

Choose File > Open As Smart Object, select a file, and click Open.

Choose File > Place to import files as Smart Objects into an open Photoshop document.

Choose Layer > Smart Object > Convert to Smart Object to convert a selected layer into a Smart Object.

I suggest you place the smart object into a new file. This means that it points to your original layered document. Work as you please.

If later you decide that it would have been better to make same changes in steps you took early in the process, like for example change some adjustments you did in your original file, go ahead and do it! The smart layer, points at the psd file you saved, and will reflect any changes you do to its source.

It get even better: you could apply also smart filters to smart objects. This means that you can retrieve a filter you applied earlier in the process and make changes to it, the very same way you would do to an adjustment layer.

I hope my explanantion is clear enough. For further documentation you can access the PS help file and look for smart objects.

David Limentani

____________________________________________________________________________

Dan:

Just looked at your video series about Lab and I concur with others that it is very good.

Despite having read "Canyon Conundrum, I've already learned some, and I'm not finished the series for the first time. Putting curves on layer masks: that must have been in the book somewhere but I didn't know it! Sometimes, when things get heavy I just try to hang on to the basics. This just shows how watching someone do something is a lot better than hearing them give a description of the process.

Well, mark me down for encouraging you to put your entire teachings in this form.

The particular benefit this form of training has is that it is very efficient, as you mentioned, plus you can pause and return to the concepts that didn't take the first time through. Even classroom instruction doesn't quite offer this aspect despite the fact that the instructor is right there. Not to mention, as an instructor using the web, the world is now your oyster. There must be millions out there who couldn't travel to your classroom sessions but wanted to.

If it should happen that you are encouraged and able to put your all your lessons through Kelbytraining.com that's fine, I'm happy to go that route. The price is very reasonable, and I'm not counting anything that might be of value from the other instructors.

If you are not so encouraged (space limitations, conflicts with other instructors) I would encourage you to consider your own, dedicated video training website.

Boy, wouldn't it be fast and easy to update to a new workflow! Corrections and improvements could be made in a fraction of time compared to publishing a book.

I could say more, but I've got to get back to Lab http: //www.kelbytraining.com/instructors/dan-margulis.html

Ron Kelly

____________________________________________________________________________

Another good use for smart objects is embedding a Camera Raw file as a smart object. You can go back and readjust the ACR parameters at any time. There is a good example of this starting on page 192 of Real World Camera Raw with Photoshop CS3 by Bruce Fraser & Jeff Schewe.

Rick Gordon

____________________________________________________________________________

Ron Kelly writes,

Thanks much. I still haven't seen it myself, which is a disadvantage for preparing the next ones, but, as with the images we receive, we have to play the cards we're dealt.

I didn't know it particularly well at the time either. I'm using that technique a lot more these days. Like much of the rest of the workflow I'm introducing in the next set, PP5E's chapters on blurring and sharpening strongly hint at it, but I wasn't able to put all the pieces together until last year.

True enough, although classroom instruction does have other plusses. Early returns from those who saw the first video are encouraging, however it may be more telling when the rest of the videos, which are denser and harder to comprehend, go up and people have a chance to evaluate them.

This last is not going to happen. After PP5E I said that I wanted to stop writing and to cut back on other color-related activities in order to enjoy life more, and I meant it. Unfortunately, it has not worked out that way. 2007 was an extraordinarily busy year and 2008 is already completely booked. Barring a cancellation or two, always a possibility given this economy, I could not make further videos this year even if I wanted to. As indicated earlier, I hope to spend the later part of this month seriously contemplating how to implement a cutback in 2009, presumably with input from this list. But I can positively rule out setting up my own business.

We'll have more after these taping sessions, I'm sure, but for now I would like to thank you and the other list members who have commented for the kind words of encouragement and for the input about what to put in these videos.

Dan Margulis

___________________________________________________________________________

"Dan Margulis" wrote:

It was only because I added some retouching and painting to the photos....

You're too modest to compare this to a wine. It depends also by the quality of the wine. If you're talking about a Brunello di Montalcino or a Sassicaia or something similar we can discuss about it and in any case I suggest to use a combination of the picture-postcard workflow and a glass of this kind of wine to obtain an unbeatable result for color correction.

Alessandro Bernardi

___________________________________________________________________________

May I suggest RADICAL:

Rich

And

Deep

I

Color

Adjustment in

Lab

--Herbert Ripka

Greendale, WI

___________________________________________________________________________

I got back from my Kelbytraining shooting session last Monday night and since then have been trying to catch up with a mountain of work that had developed while I was preparing for those sessions. Having cleared up that pile, I'd like to thank those list members who helped me with planning for them, before I head off for nine days of Internet-free time on the beach tomorrow night.

Having made two trips to Tampa in the last month for these tapings (five sessions in all, of which two have been released), I was most impressed with the Kelby empire. When a company is growing rapidly, as his is, there's always a lot of enthusiasm and electricity in the air. They've got a lot of new hires and are bursting the seams of two different buildings. They now have a tremendous market presence in a lot of areas and seem to be doing well in most of them, the danger always being that at some point they may spread themselves too thin.

I was impressed in an entirely different way but what the real estate crisis has done to that area. Here in my part of New Jersey, home prices have dropped around 10% and realtors are very spooked by the fact that there don't seem to be any buyers around. It is not, however, disastrous. We don't see many foreclosures, and although new building has essentially ceased, that isn't a big part of our economy.

This part of Florida, unfortunately, is a very different story. There, the whole economy was being driven by speculation that the real estate boom would continue forever. New complexes were going up everywhere, builders were big employers; they would borrow money to build one development and the banks would give them more for a second development on the assumption that the first one would sell out instantly at 20% higher than it would have last year. Plus, there was a fair amount of panic buying right at the peak of prices.

Now that the whole house of cards has fallen down, half-finished, never-to-be-finished developments are everywhere. Lots of for-sale signs on houses, some of which have the helpful legend "pre-foreclosure". Unemployment is way up. The banks have nothing to do with all the property they're repossessing. Scary stuff.

But back to Photoshop. I have never done videos before and this isn't the way I would have liked to start. I haven't yet even seen the first two that have been released which means that I am sure these new ones repeat some obvious defect in the way that I present things. But there just wasn't time--as list members will recall, I was well prepared to do three different sessions, but after discussion here, I decided we had an opportunity that would not likely be repeated in the near future to explain the new workflow that I was showing last year.

Since this is a *fast* workflow aimed at processing large numbers of images in a relatively short (3-5 minutes per) period, I wanted to test it not on cherry-picked images but on a *series* of images shot by people who are better photographers than I am. So, I asked three people to supply me libraries of images for the specific purpose of testing the "postcard workflow." SInce I think that my own photographs represent the redneck point-and-shoot user pretty well, I wanted to get one serious pro-amateur and two well-known professionals. I am very grateful to list members David Cardinal and Fred Drury for filling these roles, along with the brilliantly talented Vincent Versace.

The format of the three videos (which I guess will get posted in a couple of weeks; they were working on them this past week because they knew I'd be incommunicado starting tomorrow) is convoluted. The workflow as I showed it last year consists of six steps, but they break very logically into three thirds. So, each video covers a third of the workflow. But the same images are worked each time: in each case I pick up a saved copy of where the last session left off, and in the third session I compare to the original.

The total number of images is a big one: 36, start to finish. I chose that number to demonstrate speed. Assuming five minutes per image, that's three hours of work. The videos together are somewhat more than six hours. I think it is obvious that, what with introductions, jokes, explanations of steps, demonstrations of why alternate methods don't work as well, and having to open and work on each image three different times rather than in a single pass, it's undeniable that in real life it would take me much less than half as long. So, having that many reconfirms the suggestion that it's a 3-5 minute workflow.

In retrospect I think this idea worked well. Once the videos are posted, it will be interesting to see if list members have the same reaction. For each session, I reordered the images to make a logical presentation for each session. That is, five different images may be handled in almost exactly the same way in step two. In that case, I start with a lengthy explanation of how the technique works with a single image, and then blow through the four that require similar treatment. Of course, the same five images might not handle identically in the *third* session, so I reordered them again to stress similarities with different images.

Also, doing that many images and looking at them from three different POVs added a lot of realism IMHO. This way, we can really see that a certain step is sometimes difficult, sometimes easy, and sometimes not worth doing at all. The problem in adopting a new tool is always the temptation to use it too much. So, I think it's valuable to be forced to call up each image for each step. That way, from time to time I am obliged to say, "right here we would ordinarily be doing something, but this image doesn't seem to have the problem that we're looking for, or our normal method of fixing it is going to have some bad consequence if we do it now, so we should just leave this image alone for the moment and pick it up again in the next session."

And with that many images to do, it's in effect a live performance, impossible to script, so we get a better sense both of pace and the actual quality level. If I take 10 minutes to explain how one step of one image works, that means I have to do several at high speed. And once a mistake has been made in an early session, it's hard to recover subsequently. At the end of the third session, having compared all 36 final results to the originals, I said that I was not satisfied with the results on four or five that I named. They were better than the originals, granted, but they wouldn't stand up in competition. As against that, there were about a dozen that I thought were extraordinary, in the sense that I don't believe anybody could have matched the results more than a year ago. The remainder were somewhere in the middle: good work, likely achievable in other ways but not as quickly.

There's not a lot of point in discussing specifics of the workflow before the videos become available, but I again want to thank those who encouraged the project. I'm off starting tomorrow nite and will be doing some serious thinking about the future while I'm gone.

Dan Margulis

___________________________________________________________________________

Davide writes,

I don't know what the schedule is. I had been hoping that they would release all three at the same time, but I suspect that they are under a good deal of pressure. They were all recorded in the same week.

I will be interested, later in the year after all three parts have been posted for a while, to find out whether people agreed with my recommendation in this first video that they only skim it, and then go on to the next two. Right now, of course, there's no such choice, so we will be able to compare the experiences of those who watched the first in its entirety with those who had the option of doing otherwise.

The first one is not nearly as sexy as the other two and I'm afraid people will get bored. But it's important information to have, how to correct for small imperfections in RGB, and I wonder whether people will consider it in this light.

Dan Margulis

___________________________________________________________________________ .

If they released all three at the same time, many of us would sign up for one month watch all three, and then cancel the subscription. That would be a bad business decision on their part.

I had been hoping that they would make a bad business decision, but no such luck.

The first video is intriguing and covers the whole workflow, as far as I can tell. I took notes and then turned around and tried it out on several shots. As you might expect, I had excellent results. I made the mistake of telling my wife about the whole idea, and now it looks like I'll be printing a lot of vacation snaps...

Um, thanks, Dan. A lot. Epson should sponsor you.

/asw

Andrew Webb

Creative Director

Serious Retouching & Color

303.682.9119/303.819.0480

--------------------------------------

___________________________________________________________________________

Dan, I watched the first of the three videos and it was outstanding in my mind. So many times I have seen or read photoshop learning material and my biggest complaint was not enough examples. Repetition is the key to consistency.The one or two examples that the rest of the training world uses just are not enough to get a real grasp of any concept.I believe most of the time one or two examples are used because the image fits the correction.Anyone who works on 100 images will have a much better grasp on correction than someone who has worked on only five,even if 50 out of the 100 are very similar. The info as you said is vital,proper correction cannot be made unless these first "baby" steps are taken.Thanx for providing many examples,that is the way to do it. I hope the other two have many examples as well.Looking forward to learning from the next two. Jim Donovan

___________________________________________________________________________

Part One of The Picture Postcard was great. I went through the whole thing non-stop. I guess I'm not ADD. The many examples were great and it really forced me to think about LAB #'s. I am really looking forward to the next two installments.

Have a Great Day!

Jim Lawson

___________________________________________________________________________

Jim Donovan writes,

The second two videos pick up the same 36 files used in the first one and continue the correction. Different order, though, because I group similar with similar so as to get the repetition you refer to.

While repetition is a good thing to drive home the point, I'd say that the more important payback from doing so many is that it gives us to see the whole range of difficulty. With that many images, some of them are going to be hard and some are going to be routine and others totally trivial and still others get no correction at all. When we just show one or two examples to illustrate a particular technique there's the tendency to think that it always applies to every image and to force it to happen whether the image needs it or not. If only two images are being shown as an example of a technique you can be sure that both of them are going to be hard--and in that case the viewer might not realize that 60 or 70 percent of images either don't need the technique at all or else are much easier.

Thanks for the very kind words about the video.

Dan Margulis

___________________________________________________________________________

[Private note to Moderator: If Dan doesn't want

the entire workflow summarized until all three

parts are available on Kelby, don't hesitate to

reject this post, and I'll revise it with only

piecemeal comments.]

[MODERATOR INSERT:

The workflow is not a secret. I showed it at Photoshop World, and ran last year's advanced courses around it. All ACT classes in the second half of the year saw pieces of it. As of the first class this year (three weeks ago in Washington) the curriculum was almost completely revised to stress this new way of doing things; the most thorough overhaul since the class started 15 years ago.

That said, the description below, although correct in its basics, is necessarily incomplete. The Kelby videos are 2+ hours apiece for a reason. None readily compresses into a two- or three- paragraph summary.

The first video, the one that's up now, is the exception: it's not showing anything new or exciting, just proposing a disciplined way of doing things that people already know how to do, and suggesting that using LAB numbers to analyze RGB pictures is considerably faster than any alternative.

The second one is more difficult but again is a repackaging of known methods, so it's not hard to discuss it.

The third one, as the description below hints, is full of nasty counterintuitive new stuff that seems to make very little sense but redeems itself with shockingly better results than anything previously known. The false profile being referred to is a particularly egregious example, but we DON'T use it on every image.

So my suggestion, but not my demand, is that we should tread lightly on discussions of the third part before the video gets posted.

Dan Margulis

/moderator insert]

Inasmuch as Kelby is stretching-out rollout of the second two parts of Dan's new rapid color corrections workflow (perchance to squeeze more monthly subscriptions out of we who only are in it for this), I'm anxious to begin discussion now. Accordingly, the following summary of the entire workflow is what I've been able to glean from Dan's lickety-split presentation in his overview.

Kelby video: Margulis "Picture Postcard Workflow," part 1, chapt. 3 "teaser":

1) "Once for color" - correct color casts: Conservative color prep in RGB to prevent amplification of small defects by later high-stress corrections.

- Analyze image in ten seconds.

- Remove color casts, often in Color blending mode if looks better, with Curves and/or channel

blending.*

- If lightest channel > 230, Shadow/Highlight: 18,25,30 / 8,26,30

- Save as preliminary conservative version, for possible use later.

2) "Once for contrast"

- Check to see if more highlight detail needs to be added.

- Increase contrast by steepening curves in individual RGB channels, with layer in Luminosity blending mode.

- Perhaps completely replace poor-contrast channel with another existing better-contrast channel.

3) "Twice for color" (vivid colors in Lab)

- Still in RGB, Assign 1.4 gamma false profile to above Luminosity-mode layer to lighten appearance. (Presumably to exploit the consequent increased range devoted to highlights-to-midtones during subsequent darkening moves below.)

- Convert to Lab without Flattening (to take advantage of superior color layer blending when Flattening in Lab).

- Flatten, add two consecutive duplicate layers in Multiply blending mode. (Presumably to restore original overall darkness while adding otherwise unattainable contrast and "luster" back to the pre-lightened false profile move above.)

- Merge-down third layer into second layer, add Layer Mask of L channel, lighten Layer Mask with Curves if still too dark.

- Gaussian Blur the Layer Mask (using Radius guidelines for contrast/detail preservation as for S/H or blurred inverted overlay blend). Image will appear sharper.

- Flatten. Duplicate layer, Apply Image: composite channel, Overlay mode.

- Apply Image: background layer's L channel, 80% Opacity, to above dupe layer's L channel.

- Now too intensely colorful, so add Layer Mask then Apply Image: L channel, so as to limit to lighter areas.

- Lower Opacity to taste, e.g., 40%-67%. Flatten.

- Final L channel contrast curve.

- Back to RGB.

- If these image corrections are too strong, Apply Image: the preliminary conservative version saved in step 1 above, at low Opacity, e.g., 25%.

--

* Channel Mixer: Make opposite change to other channel(s) by same total amount as changed in destination channel, to preserve neutrality. Also may need Fade to Lighten or Darken in order to avoid introducing casts in other non-neutral colors.

Comments:

I'm most intrigued with the pre-lightening by the low-gamma false profile assignment, followed by darkening with layers in Multiply mode. I believe these steps produce detail and a realistic attractiveness not easily achievable with other contrast-enhancing moves, drawing upon the same advantages from reallocation of greater range to the targeted area as the Shadow/Highlight command and the blurred inverted single-channel overlay blend have in shadows, or blurring-out shadow noise. It's like a "Dolby System" for color correction!

- George Machen