3 STEP SHARPENING

<p>Patrick, what are your thoughts about doing initial "sharpening" (in CS4) on RAW files by tweaking the "clarity" setting? I seem to remember reading that clarity, by adjusting the mid-tones, is somehow superior at least to the sharpening option in the RAW conversion stage.</p>

<blockquote>

<p>Yeah, ya know if you had some better creds as a photographer that statement might have some relevance but alas, I don't think you are seeing any trends perhaps your own lack of understanding?</p>

</blockquote>

<p>Jeff, what do you know about my creds. I understand clearly. I understand oversharpening for print, I stated above that I do that myself. I know that the print loses sharpness unless over sharpened before hand. Theres a big difference in oversharpened that looks oversharpened and oversharpened that looks like cooked crap.<br>

Patricks example here shows an image thats close to ready for print, if not already, yet his doesnt look bad. Following the steps in the PK sharpener manual, you get cooked crap as shown above. Eye lashes become blobs of black dots instead of lines of hair. Maybe thats what some peoples idea of sharp is, but mine is a sharp image that looks as clear as it does in person and hair that still looks real. But, not everyone sees images in real life with clarity. It took glasses for me, which was my point.</p>

<p>And I'm not in business to print 11x14s of every image just to know how it looks in print. Its extremely wasteful and expensive, but then again, this isnt hobby for me, its my business so I have to keep tabs on waste to survive. I develop a feel for what looks right, then add a touch more sharpening for the print once I have determined I will be making a print of that image.</p>

<blockquote>

<p>If you wish to advance your knowledge of the subject, I suggest you read this "<a rel="nofollow" href="http://www.luminous-landscape.com/essays/rant23b.shtml" target="_blank">The Fallacy of Judging Image Quality Online</a>"</p>

</blockquote>

<p>Really, are you kidding. I read his stuff quite often. And he's refering to looking at down sized web images of 800pixels wide to compare resolution of a lens or camera vs another. If you take a Canon 1DsIII and use a Canon 300mm f2.8L IS and compare the Sigma version of a 300mm f2.8 in terms of resolution, I promise you a 100% view of those full resolution images will tell you all you need to know. Yes, downsized web images are no test, but dont kid yourself. It doesnt take a 20x30 print on $100 off of a $100K Inkjet to see the difference. I love a good print as much as the next guy, but its not sensible to do that in real world. And like it or not old timers, digital was brought about for this purpose. Viewing without needing to print every image you take. Speed in sharing images etc. Like old guys complaining about the onset of fuel injection in automobiles. Took my grandfather 10 years to get the fact that a fuel injected truck would start without issue in 20deg weather where his carburated truck needed you to pump the gas, screw with the choke and freeze to death all just to get it to idle.<br>

I'm looking at full resolution images here Jeff. And at 100% its a pretty damn close representation of what the image will or can look like in print. And its aweful funny how he and some others on the web can be one way one day, and another on day 2. I read his Leica S2 review and he gawked at 100% until he messed his pants. So dont let his 90 year old man rant fool you. Theres not a lot of people who make a living selling 20x30 prints of a farm field for thousands of dollars in an art gallery either. That guy will run off a 10ftx20ft print of his dog drooling just to do it. Most business owners cant and dont do this. They are too busy selling wedding images or school senior portraits or product adverts etc.<br>

Give me a break. Patricks method is about all that needed at even the extremes. The simpler method in most cases will suffice.</p>

<p>Rob, sorry, I know this is intended for Patrick, but yes, sometimes clarity can produce enough apparent sharpness without the need of cooking images in a sharpening machine. Its all about apparent sharpnes to the eye. What appears clear and sharp. Sometimes its a clarity bump, sometimes just a contrast bump, sometimes sharpening if those wont do. Sometimes its all of the above.</p>

<p>Id play with it. If you like it, its good. I have done this in many cases. Just a quick clarity bump does it without adding sharpening which can increase noise sometimes. I find it great for brides in wedding dresses. It increases detail in dress without sharpening the whole image.</p>

<p>Silly question, but why separate capture and output sharpening? Why not do them both as a last step? Creative sharpening is clear (pardon the pun), but I'm missing the point on the value of separating capture & output sharpening, except to have a 'good' image to work with, but technically I'd think the result is the same if it's done once. No?</p>

<p>David thanks.</p>

<p>It would have been easier to print this without the ridiculous kvetching about capitals that could just have easily been in a different, relevant post.Since this is a topic that concerns other posts by the same author, that would make more sense and wouldn't have wasted many readers' time.<br>

Anyway, Patrick, you mentioned in the document "Remember that you should never add digital sharpness with help from a Raw-conversion software or with your camera." I use Adobe Lightroom with RAW (oops, raw) photos, which has its sharpening feature built in. Do you know if this uses the same engine as Photoshop or a different one? Thanks for the post and doc. Ray</p>

<p><strong>Ziggy</strong>, I kinda wonder the same thing.<br>

So I did that. And it made no difference. It didnt matter what step I did color, cropping and sharpening. The exported files showed no difference whatsoever. I looked at even 200% and there wasnt 1 pixel that was different no mater if the total sharpening was done first, last or split into "capture first, then color and crop, then "output sharpening. It really is personal preference. And just becasue Bruce Fraser says do it in 3 steps and no RAW capture sharpening, doesnt mean thats correct. Just because he is Bruce Fraser doesnt mean everyone else is wrong and hes the only right way.<br>

You know, the ancient greek thought the world as flat, until <strong>Pythagoras said I think iits round. </strong>It took almost a century for someone else to go against the grain and say that <strong>Pythagoras was right and everyone else was wrong. </strong>And look. So if you know something works for you and you cant find any logical reason to believe your thinking is incorrect, then go with it. Look at Patrick. He followed Bruce. But he clearly states that he cant see a benefit to PK Sharpener vs Smart Sharpen based on his tests. And hes right! Lets all remember...while Bruce was indeed very knowledgeable, he wasnt the be all end all and also.......he was trying to sell his program plug in ok.<br>

Just like Ken Rockwell. That guy will talk digital is great till the cows come home, but when he needs to make a film statement, then its all digital is shit and film rules the world. BS! There isnt a 35mm film out out there these days that can show superior over 35mm digital. Example look <a href="http://www.stepheneastwood.com/tutorials/Tutorials_Canon_Film_Test.htm">here</a> There are even examples of the 8mp 1DII even beating film, and his film was bigger than 35mm. Its clear whats better. But ask a 75yr old film guy and he'll swear its not possible. And that who are you gonna believe, Ansel Adams or a 29yr old man that uses digital. How dare anyone argue with Ansel. But hey, I dont care, my eyes clearly tell me whats superior, not what someone else tells me is superior.</p>

<p>It makes no difference. Now, if you downsize an exported file after all its adjustments, you do need to resharpen the downsized image. I'm talking about taking a full size image and resizing for web at say 1024 pixels wide. That image needs resharpening. But for those who say sharpening always has to be done last, nonsense. It didnt matter in my work flow the least. Because its all cumulative. It doesnt matter when you add all your adjustments, if they are all the same numbers, the exported file will be exactly the same. LR for example will make the adjustments you applied to the expoted file in the order Adobe deemed ideal. Now some like to go ahead and add sharpening first, becasue they cant work on a dull file, but this is just personal preference, not right or wrong.</p>

<p><strong>Ray </strong>Yes, Lightroom uses the same engine as PS. Its all ACR so long as its the same version ACR the defaults it sets is the same..<br>

Now for Patricks workflow, its a bit different. What he is saying(Patrick, hope I'm not butchering this too much) is for "his" type workflow, you dont add capture sharpening i RAW. Reason is for one, his tutorial shows that his flow involves adding the 1st round of Smart Sharpen at 125%, then making a background copy, addig another 125% to that copy which would in deed way over sharpen the whole image, so he then adds a mask to the new background layer and only allows the 2nd 125% to show where he wants it to show(i.e eyes etc creative sharpening), then once he likes that, he flattens those layers. So in reality, the only parts of the image at this point that has a total of 250% SS, is the eyes or where ever he unmasked. The rest was covered by the mask and shielded the 2nd round of SS at 125%. So the flattened image now is sharpened as a whole by Smart sharpen for a 3rd round....but only at say 25%.<br>

So the whole image technically only recieved say 150%, but the eyes receieved 275% total including all three rounds.<br>

So by that process, had you added sharpening at RAW conversion, his method would be too much, because it was sharpened already going into his 3 step process.<br>

For me, I think you could do RAW sharpening as capture, but just tone down the Smart Sharpen process you do so its not too much. But I think Patrick likes to work the other way. I personally think that capture sharpening should be done in your RAW converter. Thats when its the best. Especially Canons DPP, as in my examples, it does the best job at applying the sharpening. So for me, I say let DPP do the capture to give me what the file was intended to look like. Then I'll just sharpen that to taste or media.</p>

<p>@<a href="../photodb/user?user_id=359944">Allard K</a><br>

Inexcusable omission from mine, of course I forgot the whole bunch of deconvolution algorithms. With EXIF you should have the necessary data to parameterise the deconv-kernel, and also the noise statistics from each camera should be available. So you might get decent sharpening results.</p>

<blockquote>

<p>Silly question, but why separate capture and output sharpening? Why not do them both as a last step? </p>

 

</blockquote>

<p>Because capture sharpening is agnostic in terms of output sharpening. Output sharpening is based on the device and size of the document. Do you always print to the same device using the same number of pixels? If so, in theory, you could run capture, then output sharpening at the same time, as one action in fact. But if you decide today you’ll capture sharpen the full rez original and output an 11x14 to an ink jet, the an 8x11 halftone, the output sharpening has to be different and of course based on the capture sharpening. Its like asking “<em>why keep data in an RGB working space, just convert from the get-go into U.S. Web Coated (SWOP) v2“</em>. Well that’s fine if every time you print, you’ll always be targeting that web press. In all likelihood, you’ll converting to many different devices so you keep the archive in an RGB working space (Capture Sharpen), then at the last step, convert to the output space (size and output sharpen). The master archive can always be used in the future for other needs.</p>

<p>http://www.creativepro.com/story/feature/20357.html </p>

<blockquote>

<p>I use Adobe Lightroom with RAW (oops, raw) photos, which has its sharpening feature built in. Do you know if this uses the same engine as Photoshop or a different one?</p>

</blockquote>

<p>Both capture and output sharpening in LR and ACR are based on the Pixel Genius products. No, the processing is not the same for a number of reasons (one being, Photoshop operates on gamma adjusted data, LR linear encoded data). The engines if you will are different.</p>

 

<blockquote>

<p><strong>Ray </strong>Yes, Lightroom uses the same engine as PS. </p>

 

</blockquote>

<p>No, it doesn’t at all, unless you’re being quite sloppy here and referring to ACR which DOES share the same processing engine as LR. Photoshop, nope, no way. </p>

<blockquote>

<p>I'm looking at full resolution images here Jeff. And at 100% its a pretty damn close representation of what the image will or can look like in print. </p>

</blockquote>

<p>Well, you may think that...my experience indicates that a properly output sharpened image will NOT look good at 100% zoom. The degree of increased sharpening in anticipation of printed output is not something you can judge on screen. Nor can you modify output sharpening for matte or glossy media on screen.</p>

<p>If you think output sharpening is useless, I wonder why you think Adobe worked with Pixel Genius to incorporate PK Output Sharpening into Lightroom and Camera Raw? Did they waste their time? I think not :~)</p>

<p>Andrew, yes thats what I mean. If you process a RAW in LR, it uses the same ACR as Photoshop.</p>

<p><strong>Jeff, </strong>I said,</p>

<blockquote>

<p>And at 100% its a pretty damn close representation of what the image will or can look like in print.</p>

</blockquote>

<p>Looking at an image at 100% will give you a good enough idea of what image will or possibly look like in print. I didnt say looking at an image thats been cooked for print at 100%. Obviously it will look bad. I also said, I process in my RAW converter(DPP) to make my image look as good as it can first, which is the right way to do it. Then I said, if I know I'll be printing that particular image, I'll add some heavy sharpening just before I print it to counter the print softness. Call that output sharpening if you want, but its just sharpening.<br>

My statement was in response to the nonsense statement that you can never know how good a print will look until its printed. Thats total BS. I can look at my RAW and know immediately what my image needs for print. Usually its always the same. Only time I need creative sharpening for eyes for instance is when the focus wasnt spot on. If the image is in focus, not much sharpening is needed for print or web.<br>

My images always look like my RAWS at 100%. I use a calibrated workflow and start with good images. When it looks right and sharp enough but doesnt look like its been sharpened, its right. At this point I stop. If its to be printed, I go back to it and crank the slider a little further. Thats all it takes.</p>

<blockquote>

<p>If you think output sharpening is useless, I wonder why you think Adobe worked with Pixel Genius to incorporate PK Output Sharpening into Lightroom and Camera Raw? Did they waste their time? I think not :~)</p>

</blockquote>

<p>I never said output sharpening was useless. I said the 3 step process bologne was useless and calling this sharpening process output sharpening was useless. Its just sharpening. You add extra sharpening to offset print softness or less shrpening for web. It doesnt take a special term or additional program to do this. This is exactly what confuses new users and spawns the questions about sharpening. People over complicate the process to the point where is u nbearable. Look, output is just sharpening for the use. More heavy for print, less for web.<br>

And why do I think Adobe worked with Pixel Genius to add PK Sharpener. Because of $$$$$$ man. Dont kid yourself. They saw the hold PG had on having everyone convinced that their method was the only way to do it right, and so they wanted a piece of the pie. Not because it was so good. And I personally wont use it. I think its not as good as Smart Sharpen. I proved that in the post above. Waste your money and time if you want to, I'm not buying it. I dont need it.</p>

<blockquote>

<p>I never said output sharpening was useless. I said the 3 step process bologne was useless and calling this sharpening process output sharpening was useless. Its just sharpening. </p>

</blockquote>

<p>You sure about that or do you want to actually learn something?</p>

<p>If you are sure that 3 phase sharpening of capture, creative and output sharpening is useless, goodie for you. You are wrong...but it seems you don't want to better yourself so, more power to you if you want to ignore the experts in the industry. Again, I would refer to your "creds"as perhaps not meeting the the expectations I have of being an "expert". Sorry, can't say you are adding much to the common knowledge...</p>

<blockquote>

<p>I never said output sharpening was useless. I said the 3 step process bologne was useless and calling this sharpening process output sharpening was useless. Its just sharpening. </p>

</blockquote>

<p>You sure about that or do you want to actually learn something?</p>

<p>If you are sure that 3 phase sharpening of capture, creative and output sharpening is useless, goodie for you. You are wrong...but it seems you don't want to better yourself so, more power to you if you want to ignore the experts in the industry. Again, I would refer to your "creds"as perhaps not meeting the the expectations I have of being an "expert". Sorry, can't say you are adding much to the common knowledge...</p>

<p>Good stuff in here. I've been using Lightroom sharpening mostly, but the nice thing about it is that it never modifies the original, so I can go back and revisit them later. As well as easily saving multiple edits of the same picture.</p>

<p>I'm always up for learning. But only when its learning something that helps, not adds more work for either the same results or worse all because someone tells you its supposed to be this way.</p>

<p>Never claimed to be an expert. Just simply stated that, I dont believe everything I hear. And that 2 experts can have differing ideas and opinions....but whos actually right. Thats my point. Its clear I got the same results without all this nonsense. I believe what my eyes show me, not what someone tell me is right. The world is full of people trying to make a point and a buck and will do anything they can to drive it home.</p>

<p>Honda is an expert in Automotive Manufacturing. So is Toyota. But Honda will tell you that their vehicle is the best and that their way is the only way that works for the best results. Bet many here will object to that having owned both Hondas and Toyotas.</p>

<p>I was also an Engineer for Mercedes Benz and you wont belivee the crap I was told by other engineers in the recent years about their product, only to see them be wrong and fired afterwords because it cost the company so much money. And they were experts as well. But wrong. Why, because another "expert" at the supplier told them that the way he was doing it was wrong(obviously, they have a product to sell) and so he went with it. The parts failed and cost the company millions becasue of warranty claims and yard sorting to contain the issue.</p>

<p>So you see, forgive me if I dont take the Bruce Fraser pill. I'm sure he was a wonderful photographer. But I have seen the results from his PK kit and I dont like it. I think its not needed and the results arent as good as others Ive tried.</p>

<p>And I do listen and learn. I looked at Patricks method of Smart Sharpen and agreed with him that its excellent. Thanked him and will be using it when needed from now on. So I moved up in ability now because I have another tool to use, thanks to him. But becasue I didnt take the Bruce Frasier pill, my creds are questioned...get outta here man.</p>

<p>And as a side note also, the last time an "expert" told me I should invest my 401K in HealthSouth Medical, I said "I'll pass".<br>

Because I saw that Richard Scrushy was in charge and knowing him from past having had family involved in business transactions with him and friends in business transactions with him and seeing how shady he was in business practice. Said no way, that guys a too shady for me to give my money to. Do a google search on the HealthSouth scandle and Richard Scrushy and see what it turns up, I'm sure you are not familiar with it if you are out of area. Dude went to prison for fudging books and scammed people out of millions who lost their retirements because of him. Why, because "experts" from all walks convinced people to go with it.</p>

<p>Moral of this story, Experts are not always right, so what makes Bruce Fraser any different or anyone else for that matter. If I see it for myself, I'll go with it. Just as I did with Smart Sharpen.</p>

<p>dam, im happy you like my method.. wouldtn want you to be on my back ; )</p>

<p>_______</p>

<p>I will answer all the questions later today or tomorow..im busy retouching lingerie shot today..so you will understand that my attention have to be put on something else for 24hre ; P</p>

<blockquote>

<p>So you see, forgive me if I dont take the Bruce Fraser pill. I'm sure he was a wonderful photographer. But I have seen the results from his PK kit and I dont like it. I think its not needed and the results arent as good as others Ive tried.</p>

</blockquote>

<p>Yeah, well you see that's where the problem lies...he wasn't a photographer...we was an imaging expert in color and image sharpness. He did considerable research which led to PhotoKit Sharpener and ultimately to the capture sharpening of Camera Raw & Lightroom as well as the output sharpening of Lightroom and Camera Raw. Bruce never claimed to be a good photographer but he sure did to a lot of research.</p>

<p>I guess you've never read any of his book huh? Real World Image Sharpness? Real World Camera Raw? Ring any bells? Since Bruce passed away, I've taken over both of those books. The new edition of Real World Image Sharpness is out...perhaps by reading it you might gain some insight to what Bruce and I know (that you don't seem to) that a 3 phase sharpening workflow can produce optimal results and is used efficiently without substantial increase in workflow. A lot of that comes from knowing how to use the tools at hand.</p>

<p>As for the rest of your, uh, "post" about Mercedes, Honda, Toyota and that last bit about 401K investments...yeah, right, whatever...</p>

<blockquote>

<p>As for the rest of your, uh, "post" about Mercedes, Honda, Toyota and that last bit about 401K investments...yeah, right, whatever...</p>

</blockquote>

<p>Yeah, thats exactley what people do when they dont have an answer after someone else showed a different side to things.</p>

<p>How did we get into this anyway. All I said was I didnt like it, showed examples of the results from several methods including his, and now I am in an arguement with someone I dont know...<em>about </em>someone I dont know.<br>

Look, he may very well have some insight that might turn out to be correct and worthwhile. I may read his book....I do know I wont spend 79.99 for it thats for sure.<br>

If he says anything in there that I havent heard of or tried, I'll do as I always do and give it a shot. If I like, I'll keep doing it.<br>

Just like I did with Patricks method. Look, I'm not beyond trying new things. I love to learn. I eat breath and live photography everyday. Thats what I do for a living. When I am not working on something, I'm reading new stuff, trying new things every spare moment I have. Almost to a fault....my wife hates it. <br>

I'm just beyond taking everyones word for it and not testing my own thoughts. I want to see results, not told that they work...and so far, I've seen nothing to suggest otherwise....yet</p>

<p>uh, is this too sharp?...</p><div></div>

<p>All I did is work on the contrast tonal edge roll off adjusting Clarity and Contrast in ACR 4.6.</p>

<p>look good, maybe just the top rock seem a bit too vibrant, like too sharp compare the whole image? it could be just a matter of DOF and the leaf seem out fo focus compare to the rock?</p>

<p>I'd say this is good. Looks just like you could reach out and touch it.<br>

From the small web pic, doesnt look cooked or over sharpened to me. This is what I strive for and have always gotten it without additional programs etc. As stated, if this pic was slid into PS for some retouching, then the included Smart Sharpen tool would be just as good or better. But, being you didnt opt for PS retouch, the USM in ACR did the trick.</p>

<p>That wasnt hard. And if someone new was to ask you, "How'd you get those so sharp". You could say, "I sharpened it in ACR" without confusing them with, "I applied capture sharpening in ACR, then output sharpening upon export." You did just that, sharpened the photo for its use.</p>

<p>Wow, Camera Raw 4.6...hum, ya know Photoshop CS4 and Camera Raw 5.x have been out for a year now...lot's of new stuff in Camera Raw...</p>

<p>Just saying...</p>

<p>From your comments I think I may have sharpened too much for the web. LOL! Patrick, that's bark from a pecan tree, not a rock next to the green plant, but that's what I was expecting. I would've mistook the bark for dried dog poop. But also this could be due to the abstractness of the image from it being a macro shot.</p>

<p>I didn't sharpen it in ACR. I left it at defaults. What I did was fix color temp,<br /> Exposure: -.45,<br /> Fill: 20,<br /> Brightness: 60<br /> Contrast: 40<br /> Clarity: 60<br /> With a "Strong Contrast" point curve instead of the default medium. Applied USM in Photoshop indicated above.<br /> Also selected a custom profile built from an X-rite Color Checker chart using the Wizard section in Adobe DNG Profiler Editor. This added additional definition to the roll off in the shadows.</p>

<p>The before image was quite flat and ugly looking because I shot AutoWB with an old legacy film Pentax lens with plenty of early Monday morning sunlight. See the 100% view before and after crops.</p>

<p>I did this to show that sometimes its more about massaging the contrast out of the image more than relying on sharpening. The problem with all sharpening techniques is it's hard to anticipate how the final results will look on a calibrated ultrasharp LCD downsized for the web. I would've had to keep going back to the original and readjusting sharpness and contrast which I didn't do here. However, it did look similarly to zooming down to 25% or whatever percentage that's an even multiple of 2. A relatively small rez 6MP size image helped as well. Larger rez images may require smaller zoom views to emulate the same crispyness in the final web size.</p>

<p>Jeff, ACR 4.6 works fine for me. I like its simplicity. It feels like a swiss army knife with all the tools I need right in front me. Don't need all the extra bells and whistles.</p><div></div>