Scanning Debate.

[false_Aesthetic]

Hey,

There's a debate going on at school about how to scan.

A few of us think that you should get as close as possible to the final image while scanning and then use photoshop to do minimal corrections.

Others say that you should just scan with as much data as possible and do all your corrections in photoshop.

Both groups agree that it should be done at 16 bit.

Does anyone have an opinion on this? (I'm sure ya'll do)

Thanks

T

[David A. Goldfarb]

Either way it's post-processing for the most part. Use the tool that works better.

Here's another suggestion--get as close to the final image as possible in the camera so you can get the most out of the post-processing options that are available.

[Brian Ellis]

Back when Photoshop was limited as to what editing could be done in 16 bit I used to do as much as possible in the scan. But now that Photoshop can do most things in 16 bit I do most editing in Photoshop and try mainly to just get as much information as possible in the scan. I agree with David that ideally you get the image as close to final as possible in the camera and minimize the editing regardless of where it's done. But of course inherent limitations in the media and equipment can sometimes make that difficult to do.

[Gordon Moat]

Partially it could depend upon which scanner you are using. When you have a high end scanner, like a Creo or Screen, or a drum scanner, the tendency is to adjust everything in the scanner software; sometimes there is no need to even go into PhotoShop. On low to mid range scanners (Epson, MicroTek, Canon, et al) you could do many things in software, though there is often a desire to do better than what the scanner can output; some might consider that a need for manipulation in PhotoShop.

You should approach scanning as an output and results procedure. When you know the final parameters needed, and you scan directly to that, your workflow is more streamlined. This becomes important as a professional, in that billing your time is an issue; amateurs might enjoy spending more time in PhotoShop and might be less concerned about time spent at a computer. However, this is also somewhat software dependent, in that some scanning software either does not give a very good preview, or simply lacks a good range of adjustments.

If you plan on manipulating an image, it can be better to do so using 16 bit images. Unfortunately some scanning software leaves many adjustments in 8 bit only. The majority of commercial printing uses 8 bit image files. If you can adjust everything to suit in 8 bit, are outputting for CMYK, and not limited by the scanner or scanner software, then there is nothing wrong with an 8 bit scan.

So the simple answer to your question is it depends upon your scanner and software. There is rarely a situation when one procedure should be blindly followed or in which only one approach should always be used. The other answer to this is to try it out, compare variations to see the affects upon output results, then find the workflow that is best for what you want to achieve with the gear you use.

Ciao!

Gordon Moat
A G Studio

[false_Aesthetic]

Hey,

We're using an Imacon 848 and a 646. Most of us are outputing on an Epson 9800.


Those of us that shoot 8x10 are limited to a Microtek or Epson 1680.

Thanks again for the input.


T.

[Bruce Watson]

So the simple answer to your question is it depends upon your scanner and software. There is rarely a situation when one procedure should be blindly followed or in which only one approach should always be used. The other answer to this is to try it out, compare variations to see the affects upon output results, then find the workflow that is best for what you want to achieve with the gear you use.

What Gordon said!

[robc]

it's a myth that all things can be fixed in photoshop. Many photoshop actions are destructive and so the old computer adage from well before the days of pc's still holds true. i.e. GIGO which stands for Garbage In Garbage Out. Or to put it simply, at every stage of the process you should try to maximise quality. Stick to that and you won't go far wrong. At any stage that you don't bother with that principle, then you are just passing poor quality data to the next stage in the process.

[Saulius]

My understanding is that when you work images in Photoshop you are losing something because it's a destructive process. Someone corretct me if I'm wrong. So I've always thought it best to do most of your image corrctions during the scanning process. Your scanner will have more data to work with than Photoshop will so when you make corrections during the scanning phase you will have a higher quality image. Now to what degree it is higher quality I can't exactly say but this is what I've read and heard about before. As others have said it also depends on your software, and definately scan in 16 bit mode if possible. Once you have the image in Photoshop you will still have manipulations to do but they will be less drastic so that should help with the image quality. And as David says

get as close to the final image as possible in the camera so you can get the most out of the post-processing options that are available

[Marko]

My understanding is that when you work images in Photoshop you are losing something because it's a destructive process. Someone corretct me if I'm wrong. So I've always thought it best to do most of your image corrctions during the scanning process. Your scanner will have more data to work with than Photoshop will so when you make corrections during the scanning phase you will have a higher quality image. Now to what degree it is higher quality I can't exactly say but this is what I've read and heard about before. As others have said it also depends on your software, and definately scan in 16 bit mode if possible. Once you have the image in Photoshop you will still have manipulations to do but they will be less drastic so that should help with the image quality. And as David says

You loose in processing, not in Photoshop per se. With scanning, like with any other kind of digital capture, you have to have some processing in order to be able to see the image as we are used to. And that's even before any corrections are applied. The question, therefore, is not if but where do you do it.

So the real question is where do you loose less data and where do you achieve better quality output and, last but not least, which process is most efficient.

Processors in digital capture devices are highly specialized but have nowhere near the power and capacity of even the wimpiest PC. It sounds logical that only the most specialized processing, such as demosaicing should be done on the spot and everything in Photoshop or whatever other program one might be using. That, among other things, is the real advantage of RAW format, either in scanners or digital cameras.

[buze]

I have more trust in Adobe's Engineers ability to write image processing than, say, Epsons. Therefore I do try to scan to get as much of the Dmax of the image as I can (just set the white and black point), get it into a file, then process it in photoshop.

Using correction "layers" you do not touch the image until the time you are really happy with the rendition; you then just have to "commit" the changes by flattening the layers in 16 bits (even if photoshop is actualy 15 bits only) and export to your favourite format.