Zoom in to 100%. The resampled preview is not identical to the final result.
Whether I view actual size or actual pixels, it doesn’t work. My preview in the levels window looks great then I click ok and what I end up getting is a very medium noisy image. I get a similar result when I decrease the brightness and increase the contrast.
My preview in the levels window …
What’s that? My Levels window (dialogue is a more accurate term) has no preview. The preview is simple a live update of the document window. If I set my view to actual pixels, then OK the Levels dialogue, I see no change from the preview to the final result.
Post a screen grab at http//:www.pixentral.com so we can see exactly what you are up against.
Sweet, good idea. This is what it looks like before I aply the change:
And this is what I get when I click "ok":
Make sure you zoom to 100% to judge the effect.
You are viewing your image at 24%. 100 pixels of image get squeezed into 24 pixels on screen — i.e., 76% are left out and the 24 pixels do not necessarily represent actual pixels, since they are de-interpolated. Also, the previews use a cached, less accurate version of the image for speed. Do the levels adjustment at 100% for the most accurate preview. (And I’d advise doing it with an adjustment layer, not an adjustment.)
And I’d advise doing it with an adjustment layer, not an adjustment.
And why is that? I mean, it’s a little presumptious, isn’t it, that he or she wants a non-destructive edit. I don’t mean to sound snotty, but I know that for a LOT of my work, my images are processed destructively with deliberation. Images that I need to preserve, of course, are not, but not all of my images fall into that category.
I’m just saying that the blanket recommendation of a non-destructive edit (say by using an adjustment layer) isn’t always sound.
Tony, what possible advantage could there be to using a destructive adjustment? An adjustment layer increases the file size more, but disk drives are cheap. I’ll always prefer to sacrifice disk space to gain the ability to edit an adjustment at a later time. Or to be able to drag an adjustment layer onto another file that needs the same adjustment.
And, regarding destructive processing in general, it may well be that if ALL my work were destroyed, no one would ever notice or care! 🙂
I think that, unless you are prepared to go into the pros and cons of destructive vs non-destructive editing, it’s better to advise on the side of caution ie non destructive.
It’s one of the disadvantages of trying to keep advice short enough here on the forum.
I implied one advantage of the destructive approach; did I miss others?
Sometimes short enough is so cryptic that it takes dozens of posts to convey what one thoughtful, long enough post could have conveyed. Life is hard!
No.
Life is easy.
Comedy is hard.
🙂
Tony, what possible advantage could there be to using a destructive adjustment?
Sigh. I take a picture of my dog. Then I take a picture of my cat. I put the head of the dog on the cat. I flatten it and put it up on myspace. I give less than ratz fuzzy bottom about preserving the work. I just wanted a JPG. That’s just an example.
For MY work, there are reasons to do destructive edits, because of how the work is distributed along with a host of other factors.
My only point was, considering the dog head transplant scenario, it just seems presumptuious for some one to adivse an adjustment layer. The intent is obviously honorable, not not always sound. You just don’t know what the application is, so an adjustment layer may be entirely moot.
it’s better to advise on the side of caution ie non destructive.
If it’s asked. I am only pointing out, again, that it presumes a lot. Since the OP didn’t ask about it, I think it nice to let them know that it’s an alternative method in case they don’t know, but to suggest it as a preferred method is, well, are we saying "always"? <grin>.
…to suggest it as a preferred method is, well, are we saying "always"?
Why not?
And you really did not respond to my "what possible advantage" question.
Sigh.
I suggested that as my preferred method, for his consideration. I didn’t say "You shouldn’t do that as an adjustment, do it as an adjustment layer, dagnabit!"; I said, "I’d advise doing it with an adjustment layer, not an adjustment." That’s my advice, based on my experience and practice. Others are free to offer their own advice, based on their experience and practice, which may be (and is, in Tony’s case) far more extensive than mine.
For my part, if I were doing the dog’s head on the cat (or vice versa), I would use an adjustment layer whether or not I was going to save the work file with layers; I may want to go back and tweak it after putting in the background of the Folies Bergère.
And you really did not respond to my "what possible advantage" question.
So is it your position that when using the Levels command, you always use an ajustment layer, unless there is a distinct advantage in not doing so? I work the other way ’round. I will use it if it’s an advantage.
In my specific case, I’m producing thousands and thousands of PDFs weekly. I don’t want to preserve layers, type, or anything else. I have the original, unedited raw files, and the finsihed versions (PDF). I do not need the intermediate steps that got me to PDF, not no way, not no how. There is NO, let me be very clear, NO advantage, for what I do to keeping intermediary files. In fact, if you want to talk about advantages, it is FAR AND AWAY more work for me to have a semi-final version of these files that contain these extraneous layers.
The advantages for me, my product, and my business are frankly, too numerous to mention.
Some images that I produce do, however need to have adjustment layers. They are images I know I may have to go back to at some point in time and either adjust, or will want to preserve them for the sake of "safety".
But the conculsion that I should use, for example, adjustment layers instead of destructive edits is simply a false conclusion. It’s a Frank Lloyd Wright thing – form follows function.
Michael,
I would use an adjustment layer whether or not I was going to save the work file with layers;
Personal preference. I might too. But then agian, with history states and snapshots, if it’s not required in the final file, I most likely wouldn’t.
I may want to go back and tweak it after putting in the background of the Folies Bergère.
I laughed. Thanks. I know – there are some images like that. My real point in bringing it up was, there are some "generally accepted practices" which are not always applicable. In the OP’s case, if they’re just creating a flattened JPG, who cares if it’s an sdjustment layer?
We don’t need to beat this to death, I am only continuing to reply so as to assure clarity. I feel that I’ve probably done that (and beyond) <grin>.
Take a look at the image the OP was trying to work with, and you’ll see why I suggested an adjustment layer. This is not a yearbook photo; it’s something that will require a lot of tweaking. And Kellen was getting something different in the preview from what ended up after hitting OK.
Michael,
I’m sorry to be agrumentative – I know that you’re exceptionally bright. I saw the image, and you obviously see something I don’t. I understood the problem and agreed with your solution (magnification), but from that snapshot I, for one, cannot discern intent.
But it’s cool.
So is it your position that when using the Levels command, you always use an ajustment layer, unless there is a distinct advantage in not doing so?
Yup, I have Ctrl-M mapped to add a Curves adjustment layer because I do a fair amount of retouching and restoration. I think using layers would be a good approach for most people. Not for you as it turns out, but I had no way to know that until post #18. [See post #13.] 🙂
DISCLAIMER #1: To make sure there’s no ambiguity, I did not write post #13 with you in mind. Rather, I was expressing my bemused frustration with the game of 20 Questions that plays out here all too frequently when the original poster writes "My Adobe’s broke" and expects someone to tell him how to fix it or where to find a wheelbarrow and mud.
DISCLAIMER #2: Nothing in DISCLAIMER #1 reflects on the OP of this thread (anybody still remember Kellen?).
Jeez, I didn’t mean to start any fights here. Let’s just all calm down. Here is my problem (which persists). In my example what you saw what "actualy size." My image res. is 300. If I were to print my final image I would get a messy grey brick. If I view pixels It does look decent, but this image isn’t for web, it’s for print. Therefore, I need a 300dpi image not 72 and it needs to look like a starfield at "actual size." I’m sure I am missing something obvious here… nevertheless help me out!
I’m sure I am missing something obvious here
what you’re missing is you need to view it at 100% to determine the real effect of any changes you make. any other percentage (up or down) and you’re viewing an "approximation" of the changes.
Kellen, no fights going on–we’re just being boys, holding each other’s feet to the fire! Ok, so which one of you guys stole my Prozac and Ritalin?
<
http://www.pixentral.com/show.php?picture=18aqbNGM9XojjlIkmx pemHZIAfhQsZ>
I notice in the Levels histogram of your screengrab that, in addition to the high spike at level 0 (that’s a good thing), there seems to be a lot of dark gray areas up to about 30 or so. Have you tried increasing the shadow contrast in Curves? And your gamma slider appears to be set to 0.10. What does it look like if you move that one toward the right to darken the image? Also, the histogram shows absolutely nothing in the highlights. Shouldn’t there be bright stars up in that area?
Post a JPG of the image (not just a screenshot) to pixentral so we can see what you see.
Kellen,
Your "Actual Size" on screen does NOT represent a print. It’s just to give you an impression of it’s size.
A printer’s dots are much closer to each other than the pixels of a monitor. They are even less related than that.
Follow the advice given: to really judge things, view at 100%!
Rob
Right. And if it isn’t already clear, let me try it this way.
Double click the magnifying glass icon in the tool box to view at 100% magnification. Image representations at any magnification other than 100% are considered an approximation of the image.
