midan_smith

<p>camera raw and capture nx for detail, capture one for colour, especially skin tones is my preference.. it used to be capture nx i used, but now for most conversions the new camera raw is good enough to avoid switching all the time.</p>

<p>"sharpening using multiple passes of the unsharp-mask filter in photoshop (with different settings each time) on multiple layers you then merge (using lighten for some and darken for others)"</p>

<p>i havent come across this before (and this is suprising). I presume you talking about high radius, lo amount HIRALOAM contrast enhancement, which i use all the time, but masked, then over several layers with various blends.. interesting.</p>

<p> </p>

<p>ah thats a shame, as there are so many images that deserve a 5/4,6/4 or 7/4, the "well done but so what? images" and those few which deserve a 4/6, for really good ideas that arent quite there yet. This is going to result in even more well executed but completely uncreative shots getting great ratings. Is it time for me to shoot old men with great beards, convert to B&W, dodge and sharpen that beard and those eyes and wrinkles, and burn everything else down to a minimum: instant classic 7 out of 7! Or the alternative, an ordinary landscape slightly cross-processed, some cyan and maybe magenta added to the shadows, again dodging the highlights like crazy "ooh it looks like a painting (and not like a photo) 7 out of 7!"<br>

lots of my work was only worth a 5/4 or 6/4, and those parts that merited a geniune 6/6 were the good ones...<br>

but you are completely right, its the small minority thing :)</p>

<p>pixel level dodging and burning is a good place to start. use a soft light layer at 50% grey, with a 1 pixel size brush at 5% or so opacity, zoom in to 400% and start painting white on the darker pixels and black/very dark red on the lightest pixels. Do that for a few hours and see how the skin is looking.</p>

<p>one other thing, CS5 makes vastly improved use of a computers graphics card: it runs much better on my meagre-2gbram-machine-but-with-good-graphics-card than CS4 did, even at 16bit.</p>

<p>Hi patrick! thanks for the great tutorial, and earlier recommending the gry garness ebooks! having studied (and practised, practised, practised) all this i was wondering: do you ever use frequency seperations to seperately heal/clone (or even curves etc) on the HF and LF layers? i know some people use up to 5 frequency layers, im quite happy with 3 at the moment for some skin work, is this something you are into? Also, do you exclusively use liquify, or throw in some warping too? Thanks again for your generosity in sharing your expertise!</p>

<p>just like Les says: healing brush, clone tool, possibly a bit of patch tool. perhaps a selective colour adjustment, lowering magenta slightly in the red and magenta sections, adjusting yellow and possibly black to get it right. i dont think NIK dynamic skin softener can do this, as you said the level of de-grunging required blows the rest of the skin to pieces, so you would have to be doing loads of masking, and if the image is going to require that much attention, you might as well do it with the proper tools and get much better results.</p>

<p>whilst i edit colours in 16bit adobe 1998, then switch to 8bit to speed things up when it gets down to pixel retouching, i have heard from several experienced pro-retouchers (who completely understand the theory) that they do everything in sRGB because in practice you cannot tell the difference.. just food for thought!</p>

<p>ah, got it, apologies!!</p>

<p>To me, the problem with your images is that there are sharp parts, but they are not where you want them! on one, bits of the bike are sharp, on another there are some sharp stray hairs. The eye on your third shot looks sharp enough, though the depth of field is very narrow. Shoot with the aperture a little bit more closed, and manually focus if necessary. The sharp parts look sharp enough to me, and with a little discreet PP work look fine. On your larger images, apart from the focal point not being in the right place, you cant expect to see too much fine detail on the face/hair for the bike shot at a 100% crop, unless you shoot with a mega high MP, and then the 100% crop will be a completely different view anyway.</p>

<p>are you shooting raw, are you converting with capture nx2 or at least camera raw 6.0 (i know there are others..), to get the most details out of the shot?</p>

<p>i applied 30 secs of sharpening work in photoshop to the third shot, and even with the compression of re-saving a jpeg, parts are plenty sharp enough.</p><div></div>

<p>well it all depends on what you are up to.. but essentially you are asking about workflow as much as layers, as that will determine the order. Heres what i do in order in Ps, starting in 16bit colour:</p>

<p>1. maybe some slight sharpening to begin with, prefer to do at end but depends on shot<br>

2. colour layers: curves, hue/saturation etc. this could be as few as 2-3 layers or as many as 20 for a complex piece, of which most would be local (eg. eyes, lips, arm, clothes etc), a few global<br>

3. contrast/light layers, if thats what it needs, could have already been done in curves etc, but maybe something else here in soft light/overlay/multiply<br>

4. reduce image down to 8bit now the global colours are done with, onto retouching, as in clone, patch and healing, each on a new blank layer if possible<br>

5. anything special the shot needs, could be anything at all<br>

6. some output sharpening, must look natural!</p>

<p>the only exception to this is that if i have to do some masking for background alteration work, i usually get it done with at the start. i cant be more precise, as each picture is different, and requires different treatment.</p>

<p>In full agreement with Carl! What a superb E-book, read about half way through it, and already getting to sections where i want to spend a full day learning each new technique! Also got the CS4 book - i think i'll skip the CS5 one, content aware and puppet warp are obvious enough tools.. but having trawled the web for the last year picking up every retouching tip i could, i am staggered by the quality of this book! Thanks Patrick!</p>

<p>Thanks for your advice Patrick! I practise for hours every day, and am downloading Gry Garness' ebook right this second!</p>

<p>the overlay is b&w - fine, but steps 4-5 are actually 4-5-6: as in a desaturation to taste, using opacity. So we have a hyper-real (due to overlay - equally just use black and white which will also allow more control over luminance values), desaturated, brownish image. I dont have time to look into it today, but tomorrow i will figure out the last steps, as practise in Ps is never wasted! I am thinking some more blend modes, maybe some replace colour on highlights etc etc</p>

<p>I upgraded from my D80 to a D700, and i had to get new lenses, but boy it was worth it. I would still be perfectly happy to stick a 70-300mm on my D80 to use that crop factor, but the D700 is simply ten times the camera for me. Having said that, I would not buy a D700 now, as its cycle is almost over, and it therefore does not represent great value for money. When the replacement is available, you will be able to pick a d700 up for much less.</p>

<p>dont know about the bronze as it hasnt come up for me yet, but i would say that patrick is responsible for about 70% of my fashion post processing technique. As you are probably aware now, the premise is no-blur, just cleaning up details to minute levels, but it works sooo well! learn to love the healing and clone brushes, smart sharpen, and liquify, thats about 60-80% of the job done depending on the shot.</p>

<p>the D90 shouldnt be the problem, my d80 was fine for stock.</p>

<p>all to be done on seperate new layers (i usually end up with 15-20 varied-type layers on a fashion shot that would take me 2-3 hours to edit)</p>

<p>for lines under eyes: healing brush tool, if you have cs5 the content aware can help<br /> for bags under eyes: careful use of the clone tool, getting just the right opacity and sampling from appropriate texture<br /> for eyelashes: a 50% grey overlay layer, paint with black, opacity 20-40%, brush size 1-3 depending on size of your image, paint over the existing eyelashes to give them clarity. at the end perhaps a "dupilcate layer, filter:other:high pass" layer set to overlay mode, with radius around 3, masked just for the eyelashes (and i would include eyes and lips, and any fashion item that needs emphasising)<br /> i never blur skin, im a big fan of Patrick Lavoie's technique of simply using healing brush and clone to make the skin perfect (though it takes a long time). these techniques should leave nothing blurry, everything should be fine and natural viewed at 100%, if it isnt then tone down the opacity until it looks right zoomed in. Maybe its just me, but i would not make a single selection in doing what i have just described, other than painting a layer mask for the highpass layer - i would use Ctr-Alt-E quite a lot. I understand some people like to have every selection copied-pasted then worked on as a seperate layer, i use blank layers and duplicate layers for this. The only times i use selections is for changing/fixing the background. Anything else gets the appropriate painted layer mask, eg curves adjustments, colour/light/contrast adjustments, black and white balancing.</p>

<p>check your brush settings just in case, which brush view you have it set on, possibly if you are using a very soft brush it is collecting data from a larger area than you think?</p>

<p>Nick,</p>

<p>Hmm i havent run extensive tests yet, but my brief ones saw NX2 get the same detail as camera raw 6 (NX2 was definately significantly better than the previous camera raw). One possibility to consider is that capture nx applies the in-camera settings, and on mine they are set to neutral with no sharpening, so i am comparing unsharpened images in both cases. If you have sharpening set in-camera, then this is being applied and you are comparing unsharpened with sharpened. Dont hold me to this! I will run some proper tests at the weekend when i have some spare time!</p>

<p>For nikon, capture nx2 used to get better details out of the raws than the competition, but the camera raw 6.0 on cs5 seems to be able to match it. camera raw is a faster, more efficient program than capture nx2, but despite it being slow capture nx2 has some excellent features. phase one capture one is good, excellent workflow, same conversion quality as the old camera raw (so not as good as the new one or nx2). what do i use? capture nx2 if i need to use its functions, cs5 if i dont as it is quicker. if capture one got the same quality i would probably move back to it, they are all good programs. i just wish nikon would make more user friendly software that was less of a system hog. my detail observations are all with sharpening turned off, i only sharpen in photoshop.</p>

<p>nothing special, spot healing brush, shortcut 'J', the options at the top are then proximity match/creature texture/content aware, hit the content aware (presumably on a new layer, with sample all layers switched on)</p>

<p>hmm... have you tried just painting with the spot healing brush on content aware setting? works well for me, i got rid of an entire wire fence from a picture just testing the tool. One time i made a large selection and hit content aware fill i got an error because i didnt have enough ram. This alteration took me 2 minutes, its not completely perfect but doesnt need too much more cleaning up.</p><div></div>

<p>I have had the same pen and touch for a few months now. For some things i think it is brilliant. I like to use it with opacity sensitivity in photoshop, mostly for overlay layers and skin retouching, though on that setting masking is tricker as the opacity is only at 100% if i hold it down hard. I use it in conjunction with a mouse for that reason; the other thing is that i noticed that my intensive use was showing on the tablet, so i put a plastic cover over it, and write onto the cover to save some wear on the tablet. great device, probably saves me 30 mins on a 2 hour edit, with increased accuracy too, so it paid for itself almost immediately. another point is i use mouse wheel zoom and with that the pad zoom function operates differently. i really like mouse wheel zoom though so im happy enough to deal with that.</p>

<p>do other people use them in photoshop with size or opacity sensitivity?</p>

<p>I found content aware to be pretty useful, on one image it made short work of a wire fence, on another it fixed a lot of chips and scuffs on textured wood on an old church organ. Sure its easy enough to fool, but does most of what it is asked, then a little bit of 'pre-content aware' healing brush and stamping finishes the job. on one image i tried to do a very large content fill but it told me i didnt have enough ram (12mp image in 16bit, only 2 gig ram so thats fair enough)</p>