Relative to comments in the thread, is the Local Color Enhancement step an additional step in addition to the 3 USM passes Colin recommends, and, is it in fact best placed between 1st & 2nd pass? If the Local Color Enhancement step is still needed in this sharpening workflow, why does it not seem to produce any effect rather than the very dramatic effect when used separately without the other sharpening steps? Another question: If one has no further edits to be done when you exit ACR and enter PS, does one simply process the 3 sharpening steps immediately following one another in succession? In any case, I presume you put each on its own layer and merge it down in between steps? The 3rd USM step, I presume, would only be completed when/if you are on your way to print or web display, etc., correct?
My experiments with this sharpening flow, so far, have been exciting. Thanks for the additional comments.
Gerry
Hi Gerry,
I'm guessing that you mean "local contrast enhancement"?
Bit busy right now, but I'll add something to this when I get home later tonight.
*Contrast. Yes.
It's probably the best place for it.is it in fact best placed between 1st & 2nd pass?
Sorry Gerry - not sure what you mean.If the Local Color Enhancement step is still needed in this sharpening workflow, why does it not seem to produce any effect rather than the very dramatic effect when used separately without the other sharpening steps?
In theory I guess so (it's never happened to me - there's always SOMETHING else I can do to improve an image).If one has no further edits to be done when you exit ACR and enter PS, does one simply process the 3 sharpening steps immediately following one another in succession?
No. Sharpening is applied directly to a pixel layer (as opposed to an adjustment layer), although you can duplicate the layer prior to applying any kind of sharpening to give you an escape plan and additional options of you want to vary the opacity.In any case, I presume you put each on its own layer and merge it down in between steps?
Yes. It's been suggested that the image be saved following the 2nd pass, and the 3rd pass applied to another copy since it'll depend on the image size, and output device and medium, althou personally I don't bother since I pretty much only ever print on canvas, and at large sizes.The 3rd USM step, I presume, would only be completed when/if you are on your way to print or web display, etc., correct?
Great to hearMy experiments with this sharpening flow, so far, have been exciting. Thanks for the additional comments.
It seems that when I have previously applied LCE (Local Contrast Enhancement) without applying any sharpening, it produced an very dramatic effect and change that does not seem to be evident when applying it following the 1st pass sharpening. Not a big deal...I'll just follow the drill. Thx.
Sorry, yes I meant duplicate layer. Both the CIC tutorials on Sharpening and LCE recommend placing this process on duplicate layers and blending as a luminosity layer in order to avoid any of the negative effects of both.
Thanks, again, Colin, for the additional clarity. I'll now retire to a dark hole and bone up on the basics with Fraser's book.
You can do this without leaving RGB (or CMYK, if you prefer)...the trick is to use the "fade" command, in the "Edit" menu. Change the mode to "Luminosity", and adjust opacity to suit. This enables some aggressive sharpening.
HTH
proseak
In Windows, the command is Shift + Ctl +F
Last edited by proseak; 10th March 2010 at 07:22 PM. Reason: missed a bit...
Hi Peter,
Yes - I'm aware of that. LAB colour isn't suited to all image types - but - for the ones that it IS most suited (which just happens to be many that I shoot) I find it makes the workflow so much more streamlined; so one quickly gets to the point where although it COULD be done in RGB, it's usually more work & often doesn't produce the same result. As a result I tend to think of these RGB work-arounds as being mainly for those who don't feel confident using LAB.
I'm probably a bit "unusual" in that I find LAB very intuitive (moreso than RGB anyway)
Colin -
Yes, I take your point on this - working in LAB does give more control over the sharpening process, but my trick means not having to leave RGB; unlike you I find RGB more intuitive for a lot of the image process.
Still, here's another sharpening technique, which I got from Dan Margulis. It's not suited to wide-gamut images, but very useful for others. It involves working in CMYK, and rests on exploiting the "unwanted colour"
To use it, Image > Mode > CMYK Color. If you then "tab" through the four channels (Ctrl +1, 2, 3, 4) you'll see that one of them is almost always much weaker than the others; for greenery, it will be the Magenta(CTRL +2) and for skin tones the Cyan (CTRL +1).
Go to this channel, and first of all open up the "Curves" dialogue (CTRL +M), and experiment with steepening it. You can be quite aggressive here - it's the "unwanted colour" after all, and the weakest, but it is very important; it can really add "pop" or even "zing"!
Now apply USM to this channel. Once again, you can be quite aggressive here. then, go to the Black channel (Ctrl +4). This needs rather more care; Steepening the curve isn't something I do at this point - overall tonal corrections happened earlier. Zoom in to 100%, and experiment; The figures used vary so much that I can't really give any advice here.
Then Image > Mode > RGB colour and the pic is ready for export.
That's it from me until I can persude a portrait subject to let me upload a couple of pic's - this technique is particularly powerful for skin tones.
HTH
Peter
Bump for good stuff in this thread
I never thought to sharpen a jpg after reducing size. It makes sense though if you are after the most control.
I upload to flickr and facebook from time to time, using the full size jpg (anywhere from 3 to 10 megs) and I have noticed that when they downsize it, I don't always like the results!
this is my first time posting here; congratulations for this wonderful site I've been reading since long.
my question is:
isn't the first sharpen step done in-camera? even with raw?? (unless we change the default settings of our cameras)
commonly said raw is an untouched file, as a matter a fact it is not exactly like that, once we can choose quite different settings before shooting (in nikon case: vivid, neutral, etc)
Hi jovelino, welcome to CiC.
I only know what my 400D and 60D do.
The RAW file is unaffected by your shooting style. However, when you view in DPP, the style settings are already applied, though they can be changed in any way you like - the 'raw' data is still there and unchanged by your style setting. Once saved as a jpeg, the file settings are applied and become a permanent part of the image.
Hi Jovelino,Originally Posted by speedneeder
It is also as Brian says for Nikon - the setting you apply in camera isn't affecting the data at all, it is only 'making it known' (in the image EXIF data), that that is what you had selected. So if you import and display via Nikon Transfer > ViewNX (or NX2), it will 'honour' that selection, but you are able to change it on the Quick Adjustment tab.
The ultimate demonstration of this is to set MC (monochrome) in camera, download the RAW - in ViewNX it is monochrome, but in ACR (Adobe Camera RAW) as used in Elements, Photoshop CS and LightRoom, you'll see it is colour.
Welcome to the CiC forums from ...
Hi Colin,
I just bought the second edition (new, but second hand ). Thank you so much, this is just what I needed.
Cheers, Pierre
I am always concerned about the level of noise that sharpening introduces,so mine is always a light Unsharpen-mask as the last process(not that i process in any great depth,as i dont have the patience for it ).
How can this trade-off be minimised?..if at all.
Hi Martin (and welcome to the CiC forums)
I find that in doing the downsize, typically from 4000px wide and/or 3000 px tall (after cropping) to the 1200 - 1600px and/or 900-1000px tall, for web display "averages out" a lot of the noise, then you can sharpen with USM with settings of 70 - 110% at 0.3px with a threshold of between 0 and 5, dependent upon how much noise is left (and what your camera is, etc.) ~ 0 for low noise, 5 if you can still see it easily.
Cheers,
The one exception I would add to all of the above comments about capture sharpening is that I never, ever include the plain blue part of the sky in any capture sharpening. If you later reduce the size of the file to the native size of a digital projector or a television, sharpened blue skies can be not so smooth as they should be when projected as a large image.
I do generally apply sharpening to the entire image when downsizing files because I do so in a batch process. Even so, every once in a while the only solution is to custom sharpen the downsized file to exclude the sky.
Hi Dave,and thanks .
I am a bit of a focus/noise nazi..so i doubt i will ever be happy(full frame is but a distant dream).those settings are pretty much what i use(i am using cs3),though playing with the threshhold is something i should do more..thanks for that tip..
BTW i use Nikons,D60 and D90..
Low ISO is always preferable,but not always possible of course..i have found Gimp to be a bit harsher that Photoshop for sharpening though..