Dan Margulis Sharpening Action

[digitaldog]

I almost think I'm doing something wrong, because I'm seeing really substantial differences on the three images I've tried this with. Especially in the noisy sky of one image. I could post these, but people should do this themselves, so they can zoom in to really take a look.

I do agree, though, that I couldn't easily see much difference unless I used the Apply Image 'trick'.

I'm confused because you say you see substantial differences on the three images but then say you couldn't easily see the differences without the Apply Image 'trick'.

Btw, I also did this without applying any sharpening, just the conversions, and did notice some difference, sort of what I would call a smearing of some color areas.

So you compared a non edited version with one that simply made a move from RGB to Lab and back again. I'd expect to see that. This is a reason why, in 8-bit, you don't want to make such moves. Depending on the original color space, both gamut and gamma, you will toss between 20 and 30 levels of the original 256 levels of the document just converting into and out of Lab.

In fact you may be onto something. You could compare the differences in just the color space conversions using Apply trick compared to just the sharpening edit and get some idea which is providing more alterations to the data. But in the end, just look at the two images side by side at 100% view. When I did my tests (which you can see contains gradients which can show a lot of data loss damage), I could hardly see the differences between the two, hence the need for the Apply trick.

In my mind (and I really DO need to test this on more images in different color spaces), the Fade Luminosity technique works as well as going through the hoops of converting to Lab to sharpen. And that's why we have such a blend mode. To attempt to accomplish the kinds of moves we'd see just working on the L channel of a Lab file, without having to make conversions.

[laughfta]

I'm confused because you say you see substantial differences on the three images but then say you couldn't easily see the differences without the Apply Image 'trick'.

Sorry, what I meant is that comparing the images side by side at 100% (before the apply image trick) didn't show much difference. But the results of the Apply Image trick showed a lot of changes between the two files. In one of the three,  I could clearly see the entire image.

The thing I don't know yet is if the changes are significant/obvious in the final print. (if one is printing). I guess I should print them.  

Thank you, Andrew. This 'apply image trick' has many possibilities for comparing techniques.

Gloria

[Mark D Segal]

Mark,

Your suggestion seems like a very good one, but do you carry the process further by making a surface mask to protect the edges and allow stronger NR in areas with less detail. In his sharpening book, Bruce suggested that one could merely invert the sharpening mask for this purpose, but that better results would require some tweaking.

The case of the blue sky is illustrative. It makes no sense to sharpen blue sky, not so much because of its color, but because there is no detail to sharpen and the sharpening will only accentuate noise. The edge mask will protect the sky from the sharpening. With NR, the sky has little detail and strong NR may be employed and a surface mask will reveal the sky but protect the edges from the NR process.

The NR programs such as Noise Ninja may have edge protection but so far as I know this is not documented.

One could make his/her own surface mask, but perhaps the edge mask in PK Sharpener could be used as a starting point. Any suggestions from you (or Jeff and others too)?

Bill
[a href=\"index.php?act=findpost&pid=143142\"][{POST_SNAPBACK}][/a]

Bill, firstly, I have to say that noise is not a big-ticket issue for me. I shoot most of my images at ISO 100 or 200 with a Canon 1Ds, so the amount of noise reduction I need varies from none to little. When I need it, say higher ISO shots or where "cleanliness" is very critical, I do a number of things - depending. Before any sharpening (i.e. ZERO in Camera Raw, then just after rendering into Photoshop) I'm mainly using NoiseWare for noise reduction. It has a remarkably detailed UI and feature set for distingushing between detail and noise for both colour (even various hue categories) and luminance noise. But sometimes I want to be more extreme in how I handle noise vs sharpening in a single image.

Good example, I have images where I want zero noise reduction in some areas and heavy noise reduction in others. Large amounts of sky above a city-scape with much fine detail and texture which I want to preserve is good example. In a case like this I put the sky on a layer, apply heavy noise reduction to that layer, then capture sharpen the whole image, examine the result on the display at 50% and 100%; if OK, just flatten the whole thing and carry on. If not, adjust opacities of the noise and sharpen layers, re-examine and flatten. Take another example: I want sharpening on everything except skin. I’ll sharpen, then punch-through the sharpening layer masks with a paint brish (be it capture or output sharpen) to delete the effect of the sharpening in those particular areas. There are other ways of targeting of course, but these two are the more usual situations and these approaches are quick and effective.

[mistybreeze]

Well, with "fans" like you, I sure don't need enemies...

True, loyal fans often care deeply about the character of the men and women they admire. Enemies hate and typically seek to destroy. Jeff (and Andrew), I repeat, I am a fan. And I may continue to pay hard-earned money for your knowledge. That's up to you.

My purpose on this thread was to address "professional behavior," particularly the public kind. If the subject of behaving like a professional does not belong here, Michael Reichmann is free to edit the thread.

Much like bashing a fellow professional, calling a customer, any customer, a "yahoo," or any derogatory name in public, demonstrates disrespect and possibly contempt. It is my firm belief that any businessman who disrespects or maintains any obvious contempt for his customers or fellow businessmen, is a businessman out to destroy. He is his own worst enemy.

One final note: I came to this thread wondering who Dan Margulis was. I had never heard of him. I called friends at two ad agencies and they had never heard of him. Finally, I presented his name to my two favorite beauty retouchers at Box Studios. One said, "Never heard of him." The other said, "She's hot!" (It was noisy but I'm sure Cindy Margolis appreciated the compliment.)

Charlatans exist in every business. I'm sure Adobe software has spawned many. There is always a discreet and polite way to let your fans and customers know what you think of an opposing theory. Hopefully, if you regard yourself as a true professional and care to protect your business interests, you'll change the subject. The use of discretion can be as powerful and rewarding as Photoshop.

[digitaldog]

My purpose on this thread was to address "professional behavior," particularly the public kind.

I may have stepped over the line in professional behavior but I'm not convinced (yet) but will keep an open mind.

What is key here is that industry experts (guru's, pundits, whatever) should be able to discuss the ideals and techniques of others, its known as peer review. Its done in all kinds of industries. It should be applied to myself, Jeff and in this context, Dan. I've tried to explain a history of denial on Dan's part, censorship and worse, an inability to allow peer review. If in doing so I got personal, and in that respect unprofessional, I will gladly review these points when shown to me and if necessarily, apologize not to Dan but to the community. I don't recall saying he's a fat impotent slob that can't tie his shoes. I did say he's got a big BS factor, which I believe can be demonstrated by simply quoting him.

[Schewe]

One final note: I came to this thread wondering who Dan Margulis was.
[a href=\"index.php?act=findpost&pid=143383\"][{POST_SNAPBACK}][/a]

That explains a lot...your lack of knowledge about a published author and Photoshop expert of many years kinda puts the rest of your perspective in context.

As for your attempts at behavior modification bud, that's a lost cause...I'm way to old to care what others may think of me. It's just not relevant to who I am and what I do.

[Mark D Segal]

Mistybreeze, whoever you are, there is no disagreement between you and me about the desirability of maintaining professional standards on web forums. That would make it perhaps less than satisfactory to label customers as "yahoos". I think you would agree that any one who buys a book is a customer of the author. I quote for you from the left column of page 58 in Professional Photoshop 5th Edition by Dan Margulis as follows:

"The following practices identify you as a yahoo and should be avoided:
> Using the Master Curve.................."

Well, you know, when I read that (and I am a customer of Dan's books too - lots of useful stuff in them) I just passed it off as Dan's emphatic way of trying to drive home a point, but truly it's a bit condescending and demeaning, especially since the overwhelming majority of his peers teach their students how to use the Master Curve (known by the rest of the world as the RGB curve) appropriately.

So the message here is simply this: within the range of behaviours that are "not un-professional", there can be some wiggle room for a bit of feistiness in the give-and-take before it gets downright nasty (which must be avoided), and those people who are overly dogmatic about what to do and what not to do in a program like Photoshop can expect a reasonable dose of it.

[digitaldog]

Well, you know, when I read that (and I am a customer of Dan's books too - lots of useful stuff in them) I just passed it off as Dan's emphatic way of trying to drive home a point, but truly it's a bit condescending and demeaning, especially since the overwhelming majority of his peers teach their students how to use the Master Curve (known by the rest of the world as the RGB curve) appropriately.

Dan is a master of being condesending and demeaning on his list. Case in point was the post you Mark made about new usefulness and functionality in Camera Raw 4 (of which I know Dan doens't have, I saw him at Photoshop World last month using CS2 for his classes). He goes on a rant about how shills (who's he pointing his finger at?) and paid consultants:

Mark Segal writes,

>>This is out of date. The Raw Converter in CS3 (ACR-4) is far more
powerful and offers many new and improved options for correcting and modifying
colour balance, hue, brightness and saturation of individual colours, and
overall image brightness and contrast whether in colour or greyscale.>>

Dan replies:
This morning I started a new class with eight strangers. In light of your post, I asked them: What is the version of Photoshop that you currently use? The answer was what I would have expected at this time: one uses CS3, seven CS2. That's entirely typical of past releases of Photoshop--it's at least six months after the initial ship date before the new release has caught on with the majority of the professional user base. (AR: Dan using his idea of science to propose fact, he asked 8 people who gravitated to his class about their experiences, that sums up the entire Photoshop community).

What I asked was that "as a courtesy" people who were using tools from the new release to explain to the non-users (who are certainly the current majority) what is going on. This is no different from what has taken place in past releases; for example when we discussed the Surface Blur filter here in the early stages of the CS2 release most people were kind enough to indicate not just *what* it did but talk about specific images in which it was superior to previous methods.

There is an even more compelling reason to ask this here. It is difficult to think of anything in the last five years that was overhyped the way the CS2 Raw was. Among the phrases that the shills used to describe it was "an absolute revolution"; "incredibly powerful"; "renders Photoshop a mere plug-in for Camera Raw"; "gives infinite control", "the greatest advance in Photoshop ever," etc., etc. Worse, those users who hadn't hopped on board were belittled and informed that it proved that they were disinterested in quality, and that if they were really anything more than rubes they would be using the state-of-the-art stuff that all professionals know is the best ever. (AR: Utter nonsense. And where did these quotes come from? Can he back this up?).

Now, of course, it turns out that the CS2 release was not so hot--otherwise, how could CS3 be, as you say, "far more powerful"? And how could all these vital tools possibly have been left out in the past? (AR: With that attitude, why not stick with Photoshop 2.5 Dan?. And CS2 was not so hot? Why? Because Dan says so)

Whenever any software package is upgraded, you can be sure that certain
parts of it will be fluff and others actually useful. (AR: Its fluff based on what I say is Fluff, not you. Oh, and doesn't Adobe provide a demo so you can decide if this is full or not?) It is up to the users to sort it out and decide whether it is worth getting. Obviously the shills will be telling us, as they always do, that it is the greatest release in the history of the galaxy. Sensible people (AR: That means Dan) ignore this bought-and-paid-for commentary. (AR: I, Dan of course don't count here) Someone like yourself, OTOH, who does not depend on handouts from vendors, can promote the product considerably more than they can, by citing specific images that benefit from it and that could not have been worked as effectively in the version that previously was the greatest release in the history of the galaxy.

If you get a handout, your obviously biased. Hog wash. Read ANY review in a magazine or on line, the authors get the products to review. 9 times out of 10, they keep it. A good reviewer who's respected and has been doing reviews for any length of time is going to be unbiased. What Dan is pissed about is hardly anyone would send him product to review, he was kicked off the Photoshop beta years ago for being not only useless but worse, for writing scathing reviews of Photoshop that were either technically soft or to once again build attention to himself. Most well balanced reviews stress the good and the not so good.

On his list, he didn't imply, he said Adobe was Jeff's boss. He insulted Chris Murphy, a well respected author and expert. Chris replied and of course got nothing back from Dan:

Chris replies:

Who's the bought and paid for commentary? Who, specifically, depends on handouts from vendors? What's the nature of the handout? How is the advice adversely tainted? You brought this up and have made a charge of serious impropriety on the part of unnamed digital imaging professionals, and you've done so with a VERY broad brush. Your statement categorizes anyone who says this or that is the best release ever is a shill. Watch out where you point that shotgun.

So there's more to all this than Dan's questionable techniques, lack of methodology and peer review. He's a hypocrite and he's just as nasty as anyone here or elsewhere with respect to professional courtesy.

[mistybreeze]

That explains a lot...your lack of knowledge

It sure does bother me, Jeff, that you seem to take pleasure in pointing out how everyone is NOT as smart as you. As brilliant as I think you are, I find this to be a "small" quality in a man.

I'm way to old to care what others may think of me.

They say you can't teach an old dog new tricks but I rarely listen to what "they" say. Plus, it's never too late to love yourself and others more. My thoughts on professionalism were meant for any photographer trying to grapple with the business side of doing business. My perspective was not meant for you alone.

It's just not relevant to who I am and what I do.

I see a couple of I's there which suggests your focus on only yourself. It seems you're professionally aligned with others, though. Is their professionalism worthless? I guess if they choose to tolerate the selfishness of the old dog than so be it. I still believe you can do better.

your lack of knowledge about a published author and Photoshop expert of many years

I can't read and/or follow every so-called "Photoshop expert." I have a business to run, mouths to feed and people who need my love and attention. If you want to criticize me and "shoot me down" because I was smart enough to latch my brain and focus onto Bruce Fraser, Jeff Schewe, and Andrew Rodney, while not wasting my time and energy on most others, be my guest.

there is no disagreement between you and me about the desirability of maintaining professional standards on web forums

"Professional behavior" in public includes more than just web forums.

I think you would agree that any one who buys a book is a customer of the author.

Buys a book, attends a seminar, and purchases software, yes, including those of the experts who are professionally aligned with each other.

I quote for you from the left column of page 58 in Professional Photoshop 5th Edition by Dan Margulis as follows:  "The following practices identify you as a yahoo and should be avoided:
> Using the Master Curve.................."

Maybe there was a good reason why the experts in my life never mentioned or steered me in the direction of Margulis. Like I said, I don't have time to waste on bullshit and charlatans. Every expert in my circle knows I spend my money and time wisely. A thoughtless recommendation that proves worthless will get you beheaded in my studio.

I just passed it off as Dan's emphatic way of trying to drive home a point

Nobody is perfect and fans tend to overlook a flaw or two. Plus, drama attracts attention. Schewe's seminar style shocked me at first but I kept giving him another chance. The brains matter more to me than the insecure bully factor.

In the "serious" business world, especially when you are aligned with other professionals, public behavior matters because respect and character matters. There isn't much "wiggle room" for tolerating flaws.

I fully agree, I'm grateful when someone steers me away from a charlatan or bad advice. But there is a professional way to bring differences to light without doing harm to your professional reputation or attaching your "shoot-em-down" reputation to others in alignment.

I've been told some photographers are worse than $2000-a-day hairdressers. Their ego and need to proclaim their greatness leaps beyond any form of professional standard. I don't enjoy the comparison and I'm proud of my professional woman behavior, bud.

[Mark D Segal]

Andrew, you know, I was puzzled by the result of his poll in the Advanced Class - one of eight people having up-graded to CS3. I'm not saying he wasn't telling the truth, but the puzzling thing is that (i) it's the advanced class, meaning they all took his basic class which is a pre-requisite, and (ii) these are people willing to commit 1600 in the US for that class, so there is something serious about their interest in Photoshop. Could it be that the class occurred so soon after the new release that few had up-graded yet? I'm not sure about the exact timing of these events. Anyhow, he should better have asked how many of them WILL up-grade and I'll bet he would had a much different answer.

The other aspect of puzzlement is that when we go to PSW, more often than not people like Scott Kelby will start a session with "how many of you have up-graded to CS3; how many on CS2; how many on CS; how many still on PS7......(security! laughter.....). And the audience can be anywhere from 500~2500 people, so the sample is much bigger than 8. I look around the room at that question time and I see (each time that I've been there, and I chose the PSWs closest after a Photoshop upgrade) that AT LEAST half the audience if not more is already up-graded. Personally, I was "upgraded" from the day the Beta was announced.

I read most of the reviews and commentary on new releases and I just haven't seen the extent of hype and exaggeration about them Dan talks about. Most of them are enthusiastic about the new features, but that's fine - par for the course. Most of them also use the opportunity to communicate their wish-lists - and that helps Adobe in charting the future work program. And some of them do indicate that they don't see the obvious improvement of this or that llittle bit. As well, it is technically undeniable that ON THE WHOLE each release is better than the previous one. So what's his problem?

You know, after reading all that stuff you just quoted about what Dan responded to me, the only things I thought it worthwhile to say at the time are that I've processed about 1400 images of all kinds in the new software since it became available so I have some experience to rely upon, and just because successive versions of Photoshop were advances over previous ones, as they should be, it doesn't mean the previous ones were no good or that commentators had no right to tell their readers what improved since the previous version. "Rome wasn't built in a day" but people were living in it during the construction phases.

Oh - and I may have mentioned that professional reviewers are entitled to remuneration for the time they spend (and the expertise they've invested in)working the software and writing their stuff. It's real work and they deserve to be paid for it.

I found it very unfortunate how technical discussions about a piece of new software got diverted into all this kind of stuff. It would indeed have been preferable to see the colour space discussion, the bit-depth discussion, the Camera Raw discussion, the colour management discussion (trusting "machines"), the B&W Adjustment Layer discussion and now the sharpening discussion stick to their technical and aesthetic aspects.  But it was not to be.

(Edited: unintended emoticon eliminated)

[laughfta]

In all fairness, (sigh) and in spite of the fact that I find some of the shenanigans really frustrating, I have to consider this: isn't it amazing that we are able to actually access people like Jeff Schewe and Andrew Rodney at all? I wonder how many times they have seen the same questions and responded to the same arguments in the last couple of years?

Andrew just provided me a very courteous response to a question that has been plaguing me for some time. Jeff was one of the portfolio reviewers at Photoshop World; I benefited greatly from a very patient and encouraging exchange.

Cost and benefit? Benefits far outweigh the cost, in my opinion. Simple as that  

Gloria

[Jonathan Wienke]

It sure does bother me, Jeff, that you seem to take pleasure in pointing out how everyone is NOT as smart as you. As brilliant as I think you are, I find this to be a "small" quality in a man.

And others could say it bothers them that the majority of your posts on this forum have been self-righteous pontifications proclaiming your moral superiority over others here, while in several instances the invective and epithets you've directed toward others has been far harsher and more disrespectful than the statements that got you on your high horse in the first place. If you want to be the guardian of civility and decorous public discourse here, the first place you should apply your standard of civility and respectfulness is to your own postings.

[Peter Stacey]

I have been a long time lurker on LL, recent joiner and more recent poster, and having watched this thread and others, I'm now motivated to post on this subject (ie. the DM subject).

From the start I'll admit to believing in the benefits of a calibrated workflow and I think that DM writes at a level that involves some abstract thought beyond the basic user of PS, but that DM also acknowledges this in his books. His books are not meant for the basic or even intermediate user and DM fails to acknowledge this to the extent necessary.

Why I have been motivated to post is that I see this type of thread on a few forums (eg. DGRIN (user: peterst6906) and dpreview) and, in my view, to some/many it can come across a bit like school-yard bullying.

DM doesn't (for whatever reason) move outside his own tightly controlled list (ACTL) and so, to people who don't know much about the list, it seems like there are plenty of personal attacks from one side, without any defence from the other (not that there isn't the chance to defend himself, just that DM doesn't move outside ACTL), and it is natural for this to attract sympathy.

This makes it seem like a snap-shot photograph. The people in the know are able to draw on personal experience from ACTL in order to write their posts, but some of the message is in their experience, it isn't in the personal exerience of what general people see. So it's like a snap-shot because while the photo triggers memories in the photographer, others just don't see the connection because the composition doesn't stand on it's own. All they see is part of a message without the back-ground.

I think this is where the claims of "unprofessionalism" come from. However, as PhD scientist myself, I totally agree with the idea of peer review, but think that peer review should be confined to assessing the technique/method/hypothesis and not the man. It's not the man that is important, only the ideas.

This is where some disagreement might occur, because while it may seem that the ideas are being attacked, others may see that the man is being attacked instead (without any defence, despite his ability to defend himself here).

Perhaps a better strategy to show that the Emperor has no clothes would be to only make comments about the technique and to leave the names, egos and histories out of it.

If this is a thread about a sharpening technique, then perhaps the other field experts might discuss a response privately and then one post (with supporting comments from others) on some of the technical limits of the technique and ways in which general readers can test this (ie. 16 bit v. 8 bit, Lab, CMYK, ACR, etc. aren't relevant to this discussion for most people, only the technical merits of the sharpening method are).

That would allow anyone to make their own emprirical conclusion about the method, without being clouded by the character assassination (whether you agree or not, this is a perception).

Then, if the method doesn't stand up, people will see that the Emperor has no clothes and his influence will be diminished.

Ultimately, attacking the ideas is a much better peer review then attacking the character, but unfortunately, for those without a personal knowledge of ACTL, it seems like the reverse is happening.

MarkDS in his replies attracts some praise, I think, specifically because he focusses more on the technical aspects of the topic at hand, rather than the other topics that cloud the issue.

If you just chose to completely ignore DM is association with this method, and only discussed the technique, then the value might be increased.

Hope that helps. But if it doesn't, I will keep my head down to avoid that hand-grenades (luckily being Australian, I eat them for breakfast, so can handle criticism if you want to dish it out. Also, I live in The Netherlands and the Dutch aren't known for any sort of tact, so I guess you can say I'm an Australian who has even tougher skin because of my Dutch cultural influence).

Regards,

Peter

[Schewe]

Hope that helps. But if it doesn't, I will keep my head down to avoid that hand-grenades (luckily being Australian, I eat them for breakfast, so can handle criticism if you want to dish it out. Also, I live in The Netherlands and the Dutch aren't known for any sort of tact, so I guess you can say I'm an Australian who has even tougher skin because of my Dutch cultural influence).
[a href=\"index.php?act=findpost&pid=143437\"][{POST_SNAPBACK}][/a]

Gotta love those Aussies...

Personally, I think the whole "professionalism" aspect of this thread has been blown WAY out of proportion–particularly by mistybreeze's attempts at behavior mod. Nothing he, (or I suppose she) has written has been on point nor on topic...and believe me when I tell ya getting a finger shaken in my face doesn't phase me one bit...

The simple fact is that ol' Dano has been TRYING for years to regain some level of relevance in an industry that has moved beyond his area of expertise–prepress and has tried to branch out in areas where his lack of knowledge and expertise tends to get pretty quickly exposed.

I seriously doubt he will leave the cocoon of the ACTL list because then he's exposed to a far wider ranging group of peers...

But I have no problem sticking to the technical merits (or lack thereof) of a a debate. However, I think the first salvo was lobbed by the OP himself, not those following up on the discussion...and I'll never back down from a good dustup!

:~)

[digitaldog]

Hope that helps. But if it doesn't, I will keep my head down to avoid that hand-grenades (luckily being Australian, I eat them for breakfast, so can handle criticism if you want to dish it out. Also, I live in The Netherlands and the Dutch aren't known for any sort of tact, so I guess you can say I'm an Australian who has even tougher skin because of my Dutch cultural influence).
[a href=\"index.php?act=findpost&pid=143437\"][{POST_SNAPBACK}][/a]

Can't disagree with anything you've said, certainly in spirit. It would be useful if we all calmed down a bit and got back to Dan's Sharpening techniques. Lets all be nice and friendly now.

Gotta love those Aussies...

That I do! Any excuse to go there, great people, great food, great wine.

[Peter Stacey]

Earlier in the thread there was some discussion of the need for some comparative testing (perhaps more extensive than what is already available; and including a comparison with PKS).

So, after spending the last 6 weeks in the centre of Russia, I'm finally heading back home this afternoon and I'm happy to volunteer to do some testing. I have all next week off work, so I'll do as much as I can then and post a paper for review.

As a result, expect something no later than 12th October.

After having spent the last year developing a sharpening workflow that I'm comfortable with, I think I understand sharpening enough now to be able to contribute something of value. So I'll host a paper on my domain, but link directly to the paper from this thread.

Regards,

Peter

[sniper]

Perhaps you'd be kind enough to share the workflow you came up with for yourself with us Peter?  I for one would be glad to see any that might help me.  Wayne

[PeterLange]

In the Nikon forum I came across an interesting post regarding a method devised by Dan Margulis for sharpening of RGB images. It uses an artificial black channel derived from a CMYK conversion to make a sharpening mask which restricts the sharpening to the darker and less colorful areas of the image. This mask It is implemented as a Photoshop action.

Dan Margulis Action
...

Quote: >> Doesn't sound very impressive. If you're going to do tonal masking, you want to mask off the highlights and shadows to avoid/reduce halos, especially shadows, since dark tones have the most noise. You also want to mask off highlights to avoid clipping. Focusing sharpening on darker areas makes no sense at all. <<

Quote: >> First off, he doesn't mention the obvious (that I can tell) that the sharpening layer wants to be set to Luminosity only in the blend mode. This is so obvious that it makes me wonder why he left it off.  Second, the sharpening is directed to the wrong area/areas of the image (particularly for digital capture). Part of PhotoKit Capture Sharpener's emphasis is to REDUCE sharpening in shadow areas where there is already more noise and in general less edges (which generally occur between light/dark contours). Dan's recipe concentrates sharpening in the WRONG areas and in the wrong ways.<<


I’m typically keeping myself out of sharpening discussions, nonetheless, I downloaded and tried the action:

/> Interestingly enough, the first step of this Action is to change to 16 bit/ch

/> In agreement with above comment, I’d change the sharpening layer to Luminosity blend mode. Also, I would go for a broader split of the Blend-IF sliders e.g. 5/35 – 220/250.

/> The mask obtained from the K-channel seems to be somewhat close to a Luminosity mask which can be obtained without leaving the RGB sphere by duplicating the image, converting to CMYK, etc. I have yet to understand the merits from this effort.

/> Anyway it’s not an edge mask which differentiates by frequency. In a sample chapter of his book, Professional Photoshop (fifth edition) Dan Marglis demonstrates K-channel sharpening with a person’s portrait. It happens that this person has black hair and (light) Caucasian skin. In this case, the K-channel (tonality) correlates with frequency, but that is of course not a general rule.

/> Sharpening is directed to the shadows where there is more noise, thus, making USM prone to amplify the noise. Perhaps – maybe – this Action is meant to be applied on an image where a stronger noise reduction was applied on the shadows first. But even then, a differentiation by frequency would make sense.

I assume that at the end this points to a matrix where the needs for noise reduction and (re-)sharpening are defined depending on tonality and frequency. At the moment, I find this too complex to sort it out and to action it. Above Action for sure does not. So I’m staying with the simpler tools which work well enough for me and leave it with this.

My 2 ct. Peter

--

[Mark D Segal]

I assume that at the end this points to a matrix where the needs for noise reduction and (re-)sharpening are defined depending on tonality and frequency. At the moment, I find this too complex to sort it out and to action it. Above Action for sure does not. So I’m staying with the simpler tools which work well enough for me and leave it with this.

My 2 ct. Peter

--
[a href=\"index.php?act=findpost&pid=143547\"][{POST_SNAPBACK}][/a]

Peter,

Your findings correlate with others and the supporting reasons you mention appear to me well-based. Of course the matrrix for all of this exists already in the form of PK Sharpener. I reported on the extent of that matrix several days ago above.

Mark

[Mark D Segal]

So, after spending the last 6 weeks in the centre of Russia, I'm finally heading back home this afternoon and I'm happy to volunteer to do some testing. I have all next week off work, so I'll do as much as I can then and post a paper for review.

As a result, expect something no later than 12th October.

After having spent the last year developing a sharpening workflow that I'm comfortable with, I think I understand sharpening enough now to be able to contribute something of value. So I'll host a paper on my domain, but link directly to the paper from this thread.

Regards,

Peter
[a href=\"index.php?act=findpost&pid=143504\"][{POST_SNAPBACK}][/a]

Peter,

This is excellent, and very much looking forward to your findings. If you would like to have a copy of the spreadsheet where I charted all the possible combinations of PK Sharpener (for Capture and Output Sharpening) please send me a private email through this website. It all allows one to define and calculate the size of a sample set for testing.

While the most obvious tests of interest to you would be comparisons with your own sharpening approach (I too would be interested in seeing it), it would also be of considerable interest if you were able to include comparisons with PK Sharpener. If you do not own a licensse to it yet, you can download it on trial and give it a whirl with this testing.

Mark