J
JGeddes
Guest
David: I have had the same horrible results if I use this setting in USM AFTER adjusting levels. It seems that doing the unsharp mask first is the ticket.
jgeddes
jgeddes
-
Welcome to the new forums! Please read this first. For known issues we are working to resolve, click here.
Sharpening Images in Photoshop
- Thread starter LittleJim
- Start date
Mike Kelley
Senior Member
- Messages
- 1,445
- Reaction score
- 0
Dan,
All of my reference books on PS say there is no loss -- the conversion to LAB (and back) is harmless.
"Professional Photoshop" (the classic guide to color correction, written by someone who is about as conservative as they come) advocates LAB as being, without qualification, the best colorspace in which to retouch (as well as sharpen). That book is also the one that primary advocates a CMYK approach to color correction (although that does entail some loss going back and forth).
Now, obviously there is always the chance for computer error (dropping data) although checksums and other techniques usually obviate such problems in these routines. But to my knowledge no rounding or other such conversion is taking place, and I see no where in any of my references that suggest otherwise.
Mike Kelley
Senior Member
- Messages
- 1,445
- Reaction score
- 0
Actions are just macros -- which basically just collect your keystrokes into one keystroke.
Go to Action, select New and after having the chance to name it you're off. Every thing you do will be recorded, until you press the stop button in the Action rollout. If you make a mistake you can either rerecord, delete and start again, or edit (it takes a little, but only a little, skill to edit a macro). See my response below for other things you can do with an Action.
Mike Kelley
Senior Member
- Messages
- 1,445
- Reaction score
- 0
There are at least three approachs to this:
1) You can record the Action and place a pause at the USM step. This is done by just putting a check mark at the pause or step to the left of the particular step. PS will execute the macro and wait for you to enter values at that stage. I have such an action for my critical USM adjustments.
2) You can record the action without adjusting the USM parameters during your action recording. Now the USM parameters that are CURRENT will be applied when your action is executed.
This way if you bring up a group of photos where you can anticipate the sharpening needed, you go in and adjust the USM parameters, apply and then undo the results. Now run your action and it will use those last parameters you used in USM.
3) You build two or three different actions for various purposes. For the web I have a certain amount of sharpening I always use -- this is my Web sharpening Action. I find it doesn't change much from image to image when I'm only doing web stuff.
Noel is the defender of the GF sharpening process. I don't see a difference -- and to me that's the bottom line. Don't worry about whether, if the image is magnified three thousand times, you can see artificats or not. That's ridiculous. The only test you need apply is whether the best tools in the world for discerning such a difference tell you if there is one -- your own eyes.
If you're particularly insecure, you can ask your wife or significant other or (in the case of true neurotics) you can gather a group of your friends together and have them vote on it. I think you'll find that they not only won't see any difference, they'll wonder what the heck you're trying to prove (but if you have beer and pizza it won't be a total waste of time).
From PhotoShop's help file:
When you convert an image from one mode to another, the transition between modes creates a permanent change to the color values in the image. For example, when you convert an RGB image to Indexed color mode, the RGB colors that do not appear in the color table are converted to the nearest matching colors in the table. Consequently, it’s best to convert to another mode only after doing as much editing as possible in the image’s original mode. If you think you may need an image for more than one purpose, save a backup copy of that image before converting it.
When you convert an image from one mode to another, the transition between modes creates a permanent change to the color values in the image. For example, when you convert an RGB image to Indexed color mode, the RGB colors that do not appear in the color table are converted to the nearest matching colors in the table. Consequently, it’s best to convert to another mode only after doing as much editing as possible in the image’s original mode. If you think you may need an image for more than one purpose, save a backup copy of that image before converting it.
koumou29739
Well-known member
Instead of spending your time discassing the merits of the sharpening in lab
mode,why don't you guys download the ltra-sharpen at http://www.ickybits.com that Mohit suggested,and try it out.I think it's very good and applys sharpening selectively without creating artifacts.
Make sure you read the instactions so that you understand how it works.
This way you gain more control over it.
Many thanks to Mohit for the link!
Koumou
mode,why don't you guys download the ltra-sharpen at http://www.ickybits.com that Mohit suggested,and try it out.I think it's very good and applys sharpening selectively without creating artifacts.
Make sure you read the instactions so that you understand how it works.
This way you gain more control over it.
Many thanks to Mohit for the link!
Koumou
D
Daniel Lauring
Guest
I recall someone running tests where they resaved into jpeg a 1000 times with next to no degradation. I wonder if this is a similar case to that.
Noel Carboni
Senior Member
Mike, that's a sensible comment, but I've come to realize that, though you might not be able to see the difference directly, the "feel" of an image can be greatly affected by the quality of algorithms one applies to it. Plus, in the case of images displayed on computers, not everyone is looking at the image on the same monitor. I have some monitors where digitally sharpened images tend to look overprocessed, and others where everything looks okay.
I feel an image is better if a transition from one color to another is accomplished by the first color changing directly to the the second in the fewest possible pixels, versus the first color going to a lighter color, then to a darker color than the second, and finally to the second (in other words, "halos" around sharpened edges). GF upsampling, sharpening, then downsampling to original size, can accomplish this. More typical sharpening will leave halos.
Note that you don't need to use GF at all if you're going to be downsampling an original to a much smaller size. Simply sharpen it (heavily) first, then downsample. The smoothing tendency of the downsampling will result in a nice final result that can be very sharp with few artifacts.
-Noel
Mike Kelley
Senior Member
- Messages
- 1,445
- Reaction score
- 0
I'm glad you didn't take my comments the wrong way, Noel. I actually appreciate all the thought (and effort) you've put into establishing some methods of handling these problems. And someday I might even use the technique.
But I was actually referring to the way printouts look -- I rarely judge anything I do by the monitor (since that's not my target in most cases). And the prints I make that I (and others) view don't show any difference between the two approachs (admittedly I've only made two sets of prints trying to see if there was a difference -- they were both landscapes where I felt a difference might be appreciated, but I'm more than willing to admit this ain't a scientific test nor is it close to being conclusive. It's just enough to convince me).
I can't even get my wife to pick which image is "better" -- and my friends pick the Lab approach 60/40 (with ten friends, so not a very large sample either). I've tried it myself, with a blind test, and I get around 50/50 so I obviously can't see the difference, even with a loupe applied to the prints. I think about all I can establish for sure is the limits of my printer come out far before the limits of any sharpening approach.
You're probably right, particularly when it comes to onscreen images, but I don't see these halos or any such artificating in my prints, so I think it makes it moot.
Now I see that earlier in this thread we have advocates for some sharpening program (plugin?) so I guess the best thing about all this is to note that nearly any image from the D30 will benefit from sharpening, and that sharpening, like the seasoning in a stew, is best when applied to taste by the cook using whatever cooking tools they feel the most comfortable with.
B
Bill(y)
Guest
Mike,
excellent. I hear you, and agree whole heartedly.
I am pretty confident in the USM parameters and even spot-sharpen when I think it appropriate. I LIKE applying the filter at the anticipated size of print rather than up and downsampling. As such I DO use my eyes to measure, rather than guess, the appropriate sharpening amounts and have not had jaggy or halo issues in prints....
But , I wanted to add that I read a review a while ago of a plugin (and for the life of me I don't know now its name) that really did sound interesting. The plugin includes some of the considerations that you and I bare in mind while applying USM: Output media (and I think there were plently of choices), viewing distance, relative size of detail, as well as size of output....
I can't remember it all, but I like this approach. It would be nice if in conjunction with these choices an advanced mode allowed a user to fine tune the resultant USM parameters after the initial selections are made!
Also, I do seem to remember the cost of the plugin being somewhat prohibitive (to my budget at least).
Bill
excellent. I hear you, and agree whole heartedly.
I am pretty confident in the USM parameters and even spot-sharpen when I think it appropriate. I LIKE applying the filter at the anticipated size of print rather than up and downsampling. As such I DO use my eyes to measure, rather than guess, the appropriate sharpening amounts and have not had jaggy or halo issues in prints....
But , I wanted to add that I read a review a while ago of a plugin (and for the life of me I don't know now its name) that really did sound interesting. The plugin includes some of the considerations that you and I bare in mind while applying USM: Output media (and I think there were plently of choices), viewing distance, relative size of detail, as well as size of output....
I can't remember it all, but I like this approach. It would be nice if in conjunction with these choices an advanced mode allowed a user to fine tune the resultant USM parameters after the initial selections are made!
Also, I do seem to remember the cost of the plugin being somewhat prohibitive (to my budget at least).
Bill
D
Dale Case
Guest
Mohit,
The author states that Photoshop 5.5 is a prerequisite to using Ultra-Sharp Pro. I was able to make it work with Photoshop 5.0 quite effectively. Is there a chance that it could work with Paint Shop Pro V 7.0? If so, how would you do it?
Regards,
Dale Case
The author states that Photoshop 5.5 is a prerequisite to using Ultra-Sharp Pro. I was able to make it work with Photoshop 5.0 quite effectively. Is there a chance that it could work with Paint Shop Pro V 7.0? If so, how would you do it?
Regards,
Dale Case
Dale,
I really don't know ... like I said, I just found that link on Adobe's forums and was passing it on. Perhaps you can email the author to ask him?
- Mohit
Keyboard shortcuts
- f
- Forum