scanning preference

S

shaaktiman

Guest
Do you do minor corrections in pre-scan at all? Stuff like curve adjustment, color, sharpening, dust removal, etc. I've been doing it all later in PS. Which is better?

adam
 
I have wonderd about this myself. I haven't made adjustments in prescan because I can do so afterwards in Photoshop and I believe that I can do so with more tools and flexibility than what is allowed with the scanning software.

Also, I don't have much to judge by other than a preview scan which is too small to make any judgements from.

I would like to see what others do. I suppose with practice and knowledge of how well your scanner does with different films, you could save different settings for later use.
 
I use the Minolta software with my KM SDIV. The only feature I use during scanning is the "grain dissolver" which reduces noise (if it is necessary). The rest I do in Photoshop. I think you may find that those who buy 3rd party software like Vuescan tend to use more pre-scan functions.

 
I believe that there is a good reason to make prescan adjustments. I think because your resultant adjusted scan has more resolution than a post adjusted scan. In other words, if the scanner spews out a 50mb file and you adjust the levels before the scan, let's say by changing from 0-255 to 25-240, then you still wind up with a 50 mb file with those levels. However if you adjust the levels afterward, you reduce the file size.
 
Like, Huh?

Adjustments like levels don't affect the file size of an image. Only the dimensions, mode and resolution affect the file size, not any image adjustments. Likewise, levels don't have any bearing on image resolution. Resolution is independant of all of these factors, once you dial it in, that's what it is unless you specifically change it.

I was just wondering which would be better in terms of image quality. I tend to think that PS can do better (or at least more controlled) adjustments than the scanning software. (esp. sharpening) But I don't REALLY know.
 
shaaktiman said:
Do you do minor corrections in pre-scan at all? Stuff like curve adjustment, color, sharpening, dust removal, etc. I've been doing it all later in PS. Which is better?

adam
Click to expand...



The scanner should be set to get as much information as possible from the film WITHOUT clipping either the whites or blacks of a given image.

You will find that most scanners, when using the presets, usually clip the highlights and do fairly well at extracting shadow detail... at least they are set to TRY and get as much shadow as possible. The actual amount captured depends upon the dMax capabilities of the scanner.

If you adjust anything pre-scan, it should only be 'Levels' or the equivalent adjustment on your scanner. Try to get a histogram that encompasses both the highlight and shadow range without cutting off or over-compressing either.

I would absolutely NOT attempt any contrast adjustments, especially with B&W film. PhotoShop is a much more powerful tool in that regard.

The goal for scanning is to get the most information you can from the film. Follow on manipulations are best done with software that specializes in that area.

Tom
 
shaaktiman said:
...

I was just wondering which would be better in terms of image quality. I tend to think that PS can do better (or at least more controlled) adjustments than the scanning software. (esp. sharpening) But I don't REALLY know.
Click to expand...

I don't really know either, but like you have always assumed that PS would do a better job than the Epson scanning software I use. If I'm not mistaken, high-end pro shop scanning operators, who use drum scanners (liquid mounting, and all that high-end quality stuff), apply some corrections during scanning. So there probably is some precedent, or reason to do so.

As a rank flatbed amateur, I tend to want as raw a scan as possible, thinking of it as my "digital negative." I try to work on the development end to control contrast, viewing my scanner as I would an elarger. Whether that is an accurate analogy (scanner to enlarger), I don't know. :)
 
T_om said:
The scanner should be set to get as much information as possible from the film WITHOUT clipping either the whites or blacks of a given image.

You will find that most scanners, when using the presets, usually clip the highlights and do fairly well at extracting shadow detail... at least they are set to TRY and get as much shadow as possible. The actual amount captured depends upon the dMax capabilities of the scanner.
Click to expand...
Good advice Tom. I hadn't considered clipping. Now, I need to figure out how to determine that with my particular scaner.
 
I'm lookng for an image of a histogram that shows clipping.
Basically goes like this, say you have a scan of an image that has deep darks and brilliant highlights- the histogram that represents the image will have peaks and valleys along a curve the low end being the shadows and the high end being the highlights.).

A clipped histogram will have a flat area beginning the curve, and a flat area at the top of the curve. This is where data was lost in the shadows and highlights.

best description I could come up with on short notice.
 
OK- I made one up using an image I had on hand. I have tried to make one extreme example to help depict what happens during clipping.

In the first two images I have an image almost as it came from the scanner. It has been reduced from the original scan. There is deep shadow in this first example but the details are not completely lost. The second image is the histogram for image 1. It shows a a complex curve representing the dark to light areas in the image. Along the top of this curve you can see that there is only one peak that shows a small "clip" where the peak kisses the top of the graph.

In the second two images I have manipulated it to clip the shadow detail by duplicating the image layer, then setting that duplicate to "multiply." The resultant histogram (image 4) shows considerable clipping in the shadow regions.
 
Last edited:
Just want to add one bit of levity- in Rock and Roll clipping is good, In photography clipping is not good (unless you are attempting to make it look like a graphic)
 
shaaktiman said:
What's clipping?
Click to expand...


Here are some histogram examples.

1. Full range, no clipping and no compression to speak of.
2. Shadows compressed and highlights clipped. This is what my scanner tends to give me if I do not adjust it. This is not bad though.
3. Shadows clipped
4. Shadows and highlights compressed.

Note that in 2 & 3 above information is lost forever. Usually, 3 is worse as we like to keep shadow detail even if highlights get a little blocked up.

In #4, the information is there in both shadow and highlight areas, but it is compressed. You are not getting the maximum tonal range available, but you are getting all the highlight and shadow detail available.

Tom
 
Tom, I'm a touch confused. I don't see the shadows clipped in #3. Is it that tiny bit at the very edge that you mean. Also in #4 it looks like the shadows are clipped pretty severely. Am I not understanding?
 
But Tom, it seems that in example #3 you're saving a lot of highlight info that you'll want to clip later. Wouldn't it be better to clip that area in prescan and let the scanner use that info (those bits) to record the mid-tones, or shadows data?
 
Nick R. said:
But Tom, it seems that in example #3 you're saving a lot of highlight info that you'll want to clip later. Wouldn't it be better to clip that area in prescan and let the scanner use that info (those bits) to record the mid-tones, or shadows data?
Click to expand...


Ah, but what I am showing is what the SCANNER wanted to deliver, not the final setting.

These are mock-ups to illustrate various 'irregularities' you want to avoid.

Tom
 
with color images, it's important to glance at the preview a bit too. If I simply move the white point to the lightest pixels (the bottom of the right slope) and the black point to the darkest pixels (the bottom of the left slope), I usually exceed my Epson 1260's dynamic range and get some major noise (I think this is what's sometimes called banding) in the highlights. I can usually get a little more than Epson's auto levels, but not much.

At least I think that's what's going on. But I totally agree that you should do a levels adjustment before scanning, so that the scanner uses a range that matches the image. This is especially true where you want the most contrast to work with.
 
You must log in or register to reply here.
Back
Top Bottom