Prints from Scanned B&W Negs?

[photophorous]

Can anyone comment on the quality of prints made from traditional B&W film that you scanned at home? How do they look compared to darkroom prints? I'm talking about 35mm only.

I used to make darkroom prints and I miss the look. Now I'm considering buying a negative scanner so I can scan at home and send the files off to have them printed, just like I do with my digital camera files. I've done a little bit of this already, with a cheap flat-bed scanner. The quality of those scans is fairly good, considering the cheap scanner, but they don't compare to darkroom prints. I want to know if a dedicated negative scanner will give results that are close to real darkroom prints.

I would probably get something like the low end Nikon Coolscan, but I'm not ready to pick out a specific model. I'm just toying with the idea right now.

Thanks,
Paul

[Medley]

My advice is to stick with darkrrom technology for negative printing.

When you scan a negative in, convert it to a digital print, and send it off to the printer, several things happen.

The scanner scans the print in, converting it from CYMK mode to RGB mode. It creates a digital image by converting the image into a grid pattern, and converting that grid pattern into a bitmap -a set of pixels. Some "averaging" takes place in converting it into bitmap form, how much depends on the scanning size. Some pixels are also goin to be changed in the conversion from CYMK to RGB, though at this end, the difference should be negligable.

Now you do whatever editing you're going to, and prepare the image for print. Before it prints, the image is going to be converted again, for RGB back into CYMK. In a black and white image, you have 255 possible brightnesses in RGB mode, but a range of only 100 (0% black to 100% black) in CYMK. So there's going to be a fair amount of averaging there. Still, if you haven't done a lot of editing, you may not notict without looking carefully.

Also in converting back to printable form, it's going to change the image from square pixels into a round dot halftone. The image is going to be softened some. the only way to counteract this is to sharpen the image before printing, to take the softening into account. But sharpening, by definition, increases the contrast between light and dark pixels. Now you've changed the brightness of pixels and when it converts, the averaging between RGB and CYMK mode is going to become more pronounced.

It becomes a vicious circle. It can be done, and done well. but it generally takes high-end equipment and image editors to do it.

So, as I said, it's just easier to keep analog images analog. Just my opinion though.

-Joe U.

[photophorous]

My advice is to stick with darkrrom technology for negative printing.

When you scan a negative in, convert it to a digital print, and send it off to the printer, several things happen.

The scanner scans the print in, converting it from CYMK mode to RGB mode. It creates a digital image by converting the image into a grid pattern, and converting that grid pattern into a bitmap -a set of pixels. Some "averaging" takes place in converting it into bitmap form, how much depends on the scanning size. Some pixels are also goin to be changed in the conversion from CYMK to RGB, though at this end, the difference should be negligable.

Now you do whatever editing you're going to, and prepare the image for print. Before it prints, the image is going to be converted again, for RGB back into CYMK. In a black and white image, you have 255 possible brightnesses in RGB mode, but a range of only 100 (0% black to 100% black) in CYMK. So there's going to be a fair amount of averaging there. Still, if you haven't done a lot of editing, you may not notict without looking carefully.

Also in converting back to printable form, it's going to change the image from square pixels into a round dot halftone. The image is going to be softened some. the only way to counteract this is to sharpen the image before printing, to take the softening into account. But sharpening, by definition, increases the contrast between light and dark pixels. Now you've changed the brightness of pixels and when it converts, the averaging between RGB and CYMK mode is going to become more pronounced.

It becomes a vicious circle. It can be done, and done well. but it generally takes high-end equipment and image editors to do it.

So, as I said, it's just easier to keep analog images analog. Just my opinion though.

-Joe U.

Hi Joe,

Thanks for commenting on this. I guess there aren't many folks here with this kind of experience to share.

I think I understand the theory behind what happens when you scan a negative to make a print, but I'm having trouble understanding how it translates to percievable print quality. Color film, for example, is routinely scanned and printed digitally by photo labs everywhere. That seems to be the standard these days. I've seen some very impressive prints made from home scans of 35mm slides. I'm just wondering if that process works as well with B&W. Have you ever compaired a digital B&W print to a wet B&W print to see the difference?

Unfortunately, I no longer have access to a darkroom to do the comparison myself. I'm just trying to decide if it would be worth my time and money to buy a better scanner and shoot more B&W than I currently do. Darkroom is no longer an option.

Thanks,
Paul

[Medley]

Sorry Paul. Comming from the image editing end of things, I tend to get wrapped up in the technical. It's the part that I know inside and out.

The advantage of scanning your prints is that they end up being much higher in resolution than most camera images. The disadvantage is that they tend to have a lot less information to begin with.

Every time you convert from one mode to another or, with few exceptions, apply an editing command, you discard information from the image. Discard enough information, and the eye will see a difference.

The average eye can detect about 200 shades in a given color. A trained eye can detect a bit more, maybe 250. A scanned image will start with 255, and have right around 250 when the initial conversions are done. Converting for print will take it down to the 240-245 range for tones (shades of color).

That doesn't leave a lot of room for editing, and yet scanned images will always have some noise associated with the scanning process- similar to taking the photo with a higher ISO. Use an editor to reduce the noise, and it will probably improve the photo, but will definately sacrifice information.

The bonus, from a scanning standpoint, is the absolutely huge size of the image. A size of 8100x5700 pixels is pretty easy to accomplish with scanning. Compare that to the 3456x2304 pixels of my 8mp DRebel XT. In many cases, flaws can be reduced or eliminated simply by downsampling (making the image smaller).

So, if you do get a scanner, pay attenetion to scan resolution. That is going to dictate the size of the file, and the amount of "fudging" that can be done.

So it's always a balancing act, and there's always a tradeoff. I just didn't want you to go into this thinking that you could buy a low-end scanner and get photo lab results. Generally speaking, it doesn't work that way. I would also look into some form of noise reduction software, as noise is nearly always a given with scanned images. I've heard that Noise Ninja is good, but I personally have no practical experience with anything but Photoshop.

-Joe U.

[photophorous]

Hi Joe,

Thanks again for the detailed reply. I don't think I'll be buying a dedicated film scanner anytime soon, but this is all good stuff to know. I have a Canon 8400F flatbed that I've been using, mostly for proofs, but it does surprisingly well as long as the negatives aren't too thick. I think I'll download the trial version of Hamrick's Vuescan software and push this set up to it's limits before I consider upgrading.

Paul