Scanning Negatives

[joeyk49]

Not sure where this belongs...but since it has to do with a process I want to place in between film development and printing...

I've been trying to scan negatives in order to get an idea of how they will look, or how I want them to look printed. My attempts thus far have been miserable.

Can anyone brief me on, or refer me to an article on scanning negatives with a flatbed and coverting them to positives. I use PS, but thus far, only to scan and store prints and work in that other format.

Thanks.

Joe

 

[magic823]

What size of negatives and what model of scanner?

 

[argus]

There's an article on scanning for web display on this very own website!
(there was a url link here which no longer exists)

Maybe that can help you a bit further.

G

 

[Helen B]

I understand from your post that you want to use the scanner and PS to previsualise how the neg will look when printed conventionally, not as an intermediate digital step. That sounds like an interesting project that isn't contrary to the spirit of APUG. A similar thing is done in the movies - digital simulation of how a neg will print. It's like BTZS with pictures.

Therefore what you want is not the best scan you can get, but a scan that is representative of how a piece of paper will respond, or preferably, will allow the paper response to be mimicked in PS.

I'd do it by producing a standardised file from the scanner - ie one in which a certain transmission density will always correspond to a certain greyscale value (I'm using B&W for simplicity of description, the same system would work for colour). I'd achieve that by using a step wedge with scanning software like Vuescan ($50) that allows you to fix the scanner 'exposure'. Then, knowing the paper's response curve to that step-wedge for each contrast grade, I'd write a little PS filter that applied the paper curve to the film scan file and allowed the choice of paper grade and exposure - ie the ability to select the portion of the film's response curve that would be translated to the paper. Easier to explain with a few diagrams than to put into words, especially as this should be explained at greater length than I've done quickly here. If I'm on the right lines let me know and I'll explain further.

Best,
Helen

 

[rfshootist]

Not sure where this belongs...but since it has to do with a process I want to place in between film development and printing...

I've been trying to scan negatives in order to get an idea of how they will look, or how I want them to look printed. My attempts thus far have been miserable.

Joe

Hi Joe,

we talked about it already, in general a flatbed is not suited well for 35mm negs, tho it works well for MF negsor prints. And for B&W it's even worse, if you want decent results you need a dedicated 35 mm scanner. A small Minolta DualscanIV is better than each flatbed, excepted maybe the expensive Epsons. But for those I haven't seen reliable results, for the Dualscan IV i have.
Regards,
Bertram

 

[Lee L]

Go with Helen's suggestions. I'm only chiming in to stress that you need good software like Vuescan to make your scanning consistent enough to be useful. A lot of mfgr supplied scanning software is automated in a way that won't let you be consistent about scanning, which would only be an exercise in frustration.

Lee

(Should this go to the gray area?)

 

[Helen B]

"(Should this go to the gray area?)"

It does appear to be about using digital as on offline (as we'd say in the movies) part of an all-traditional online workflow, as opposed to producing an inkjet neg for example. The 'line' in this case is the direct line between the camera original and the final print.

The preview system should work with any scanner that is supported by Vuescan or similar, and which can capture the full density range of the neg. The quality of the scan is not all that important because it isn't being used for the final print. This is all very similar to the relationship between offline and online post production in the movies.

Best,
Helen

 

[Lee L]

"(Should this go to the gray area?)"

It does appear to be about using digital as on offline (as we'd say in the movies) part of an all-traditional online workflow, as opposed to producing an inkjet neg for example. The 'line' in this case is the direct line between the camera original and the final print.

The preview system should work with any scanner that is supported by Vuescan or similar, and which can capture the full density range of the neg. The quality of the scan is not all that important because it isn't being used for the final print. This is all very similar to the relationship between offline and online post production in the movies.

Best,
Helen

Sometimes the "gray" line is unclear to me. I guess if you look at this as "scanner as densitometer", it's not an issue.

Like you, I think the greatest problem with some standard software packages is that they don't allow the user to set the "black" and "white" points, and are often set to drop too large a percentage of the extremes of the range, losing highlight and shadow detail. That's the aspect in which I'd be most concerned about the quality of the scan for this application.

I've been using Vuescan for so long that I can't really recall the "official" software for my flatbed or neg scanners. I do recall how tremendous an improvement Vuescan was. Vuescan also supports my neg scanner even after the mfgr decided not to release drivers for WinXP.

Lee

 

[Bighead]

Hi Joe,
we talked about it already, in general a flatbed is not suited well for 35mm negs, tho it works well for MF negsor prints. And for B&W it's even worse, if you want decent results you need a dedicated 35 mm scanner. A small Minolta DualscanIV is better than each flatbed, excepted maybe the expensive Epsons. But for those I haven't seen reliable results, for the Dualscan IV i have.
Regards,
Bertram

In general, it is not AS good. I think for this application, it would be just fine. I have a new Epson Perfection 4990 (flatbed). I have made test prints of both B&W and color 35mm negs and pos, at 5x7, 400 dpi... I have made the same prints from scans made on a Nikon Coolscan (dedicated 35mm), same specs, and you can't tell the difference. My goal was to compare the two to see which one I wanted to buy . I also used the prints to calibrate my monitor.

Consequently, I own both now but I use the flatbed for anything for th web, or for proofing needs... I only use the Nikon for color stuff that is going to be larger than 4x6... But, for your needs, I think any newer flatbed would be just fine.

 

[srs5694]

Sometimes the "gray" line is unclear to me. I guess if you look at this as "scanner as densitometer", it's not an issue.

I've never used a densitometer before, but I've been getting curious about densitometer measurements as a means of calibrating film speed and development times. I do own a scanner (a Minolta DiMAGE Scan Elite 5400), though, and was wondering if it's possible to get densitometer-style values out of it for this purpose. Any suggestions or pointers?

 

[Lee L]

I've never used a densitometer before, but I've been getting curious about densitometer measurements as a means of calibrating film speed and development times. I do own a scanner (a Minolta DiMAGE Scan Elite 5400), though, and was wondering if it's possible to get densitometer-style values out of it for this purpose. Any suggestions or pointers?

See Helen B's first post to this thread for a basic outline of how it would be done. I haven't done this myself.

The step wedge she's referring to would be something like a Stouffer multiple step wedge (probably a 31 step version preferred?) in whatever format you're going to scan. I think they make a version with 11 steps at one stop intervals specifically for scanners, but finer steps allow finer tuning of the process.

If you get the same size as the film you're shooting, you could use it in both the scanner and the enlarger for a fixed transmission standard that ties the two processes closely together for calibration and previewing as accurately as you can from a scan.

Lee

http://www.stouffer.net/Stoufferhome1.htm

 

[hortense]

Helen - I'm sure your suggestions are valid. To me however, this seem too complicated (at least for my generation). How about recommending a simpler method?

 

[joeyk49]

I was hoping for something that would get me close to print quality without too much investment in time or money.

I'm using an Epson Perfection 1660, which is already a couple of years old. I can't justify the expense of high end, nor am I talented enough to write software. My most basic aim is to find a way, through my present set up, to help evaluate my negatives and see where I need to go in the printing process...

If I had the time, Helen's suggestion really looks interesting. But alas, I'm lucky if I can into the darkroom once per month; and that includes developing exposed film.

 

[MattKing]

Therefore what you want is not the best scan you can get, but a scan that is representative of how a piece of paper will respond, or preferably, will allow the paper response to be mimicked in PS.

I'd do it by producing a standardised file from the scanner - ie one in which a certain transmission density will always correspond to a certain greyscale value (I'm using B&W for simplicity of description, the same system would work for colour). I'd achieve that by using a step wedge with scanning software like Vuescan ($50) that allows you to fix the scanner 'exposure'. Then, knowing the paper's response curve to that step-wedge for each contrast grade, I'd write a little PS filter that applied the paper curve to the film scan file and allowed the choice of paper grade and exposure - ie the ability to select the portion of the film's response curve that would be translated to the paper. Easier to explain with a few diagrams than to put into words, especially as this should be explained at greater length than I've done quickly here. If I'm on the right lines let me know and I'll explain further.

Best,
Helen

Helen, this looks like a great idea for those of us whose access to darkroom time is limited. Any further explanation would be appreciated.

See Helen B's first post to this thread for a basic outline of how it would be done. I haven't done this myself.

The step wedge she's referring to would be something like a Stouffer multiple step wedge (probably a 31 step version preferred?) in whatever format you're going to scan. I think they make a version with 11 steps at one stop intervals specifically for scanners, but finer steps allow finer tuning of the process.

If you get the same size as the film you're shooting, you could use it in both the scanner and the enlarger for a fixed transmission standard that ties the two processes closely together for calibration and previewing as accurately as you can from a scan.

Lee

http://www.stouffer.net/Stoufferhome1.htm

Lee, thanks for the further explanation, and the Stouffer link. I could see that it would also be useful to combine this with the use of an analyzer or enlarging meter, so as to improve the results on that "1st test print".

Thanks again, to both of you,
Matt

 

[Lee L]

Joe,

I understand about time constraints, as I have many of my own. But if you want to calibrate, you'll still need to do the work. Vuescan is $50 for the standard version as Helen mentioned, and it's great software. It does support your 1660, so you don't need to do any coding.
If the step wedge and Vuescan are too costly and time consuming, you might try making up your own transmission step wedges from negatives from photos of paint chips, the sample cards of a neutral gray scale at a hardware store. If you have a spotmeter or can get in close with no shadows from a reflectance meter, you could get a decent reading of the original scale to use for calibration. Other people have painted a flat board and lit it from one end, marking it with indicators for each stop of light falloff along the length of the board. (That might be a little hard to read densities from on a 35mm frame, but in a print, it's useful.)
I've found that reading densities of full frame single tone shots on 35mm can be done reasonably well with a reflectance meter, both with a spotmeter and with a Gossen LunaPro and the enlarging meter attachment. The same could probably be done relatively well with any smaller aperture reflectance meter and a diffuse light source. If you only get 1/3 stop resolution, you're still way ahead of not knowing anything.
PM me if you have questions about how to do this without a scanner.

Lee

 

[joeyk49]

Thanks, Lee:

You've got me scratching my chin and thinking, "...could make this work, yet..."

I'm glad I posted the question.