Scanning 4 x 5 color negative

[Wilbur Wong]

A decade ago I was encouraged by Richard Misrach to change from color transparency to color negative, for the extended range which negative format offers. I have never looked back on following up the suggestion.

Starting 4 years ago or so, I started having problems purchasing color chemistry in small kits, a problem which I attribute to the overall move to digital by the vast majority of photo "enthusiasts". I am an amateur photographer shooting only for my personal satisfaction, and while I continue to embrace view camera technique and the shooting experience, I have found my wet 4 x 5 color printing darkroom increasingly idle.

As I am getting deeper into digital processing of the printing end of my endeavors, I am finding a dearth of information regarding work flow from color negatives. (Most digital threads seem to relate to black and white or color transparency - even manufacturers of equipment and hardware seem to gear primarily to these choices, like not offering IT8 targets in negative form for example.)

I do not have a large budget for major changes, after having just upgraded my computer this year and my camera a couple of years ago. I scan with an epson 4870, print to my several year old epson photo 2000p. My current computer has 2 g of ram and 256 mb of video ram - so I think it is fairly decent.

I just ordered an upgrade to SilverFast AI from SE

How do other photographers handle work flow from color negatives? I would like to know if there are better directions for my current situation.

Thanks

Wilbur

[Kirk Gittings]

Negs scan beautifully with a couple of small problems. I use SF AI 6 "Studio" which has some additional features that are worthwhile like frame alignment in Multipass which greatly helps reduce noise. I shoot NPS which scans with a slight magenta cast to the blues, but it is easily correctable in Curves or in SF. I run the scanner calibration anyway just to make sure everything is running properly. Since there are not targets available for color negs it becomes imperitive to use the right film profile and make sure that the scan frame is within the image area so that it is not calculating in any part of the frame mask including the little hole in Fuji readyloads.

[ronald moravec]

May seem simplistic, but I scan Portra 160 on 4870 using the Epson software.I take a stab at getting the end points at the end of the scale and ocassionally make the curve s or reverse s shapped. I will do a colorbalance correction, but 99 times of 100, it is not required. The idea is to get as close to perfect a scan to Photoshop.

Perspectice correction or local color correction or any other PS work is done that can`t be done with the scanner software.

Send print to Epson printer. All Epson printers have a color correction in the advanced printing tab to make the print match the monitor. So far I used it once.

I have never profiled or done anything to color calibrate. I was very leary at first, but friends and people who know nothing about color printing were getting decent prints, so I figured I would too. Maybe I`m lucky, I don`t know. I picked the Epson because of the controls. I thought I could do corrections just like the color balance on my enlarger. The few times I used it, it worked like a charm.

I do make small test prints on the exact paper I will use for larger prints. Darkroom technology seems to transfer right along.

[Leonard Evens]

I've always preferred color negative film over transparencies because of the advantages when printing. When I took up digital, I continued with color negative film, primarily Portra VC 160. I use Veuscan, and I process with the Gimp under Linux, but the principles are the same.

Color negative film typically has a lower Dmax than transparency film, so almost any modern scanner is adequate. I use an Epson 3200, but I gather the Epson 4990, which is what is available today, is a bit better with respect to resolution and Dmax. Because of the narrower range of values in the negative film, you want to avoid too much manipulation in a photoeditor which may spread values apart, but I haven't found this to be a problem despite the fact that the Gimp is limited to 8 bits per channel.

When you scan, the most important thing is to have the RGB values within range. I usually aim for maximum white at about 235 or slightly higher, although in some cases it may be necessary to allow some highlights to go to 255. The deepest shadows I try to keep above about 35 but I'm not fanatical about it. Vuescan allows you to choose an element of the scene which you want to be a neutral color and then determine the color balance from that by right clicking. When that doesn't work, I use a collection of typical values for different colors which I can set insead. Also, I use a small digital camera to take a picture of a gray card, and then i can use that to get some idea of what the RGB values should be for different parts of the scene. Of course, you don't have the advantage of looking at the film itself on a light table, but I think that is much overated. There is no special reason to believe that what a transparency shows you is in fact an accurate rendition of the colors in the scene. In any case, with a little practice, I seldom find that color balance is an issue.

Probably the most useful tool when photoediting is the curves tool. It is well worth the effort trying to master its use.

[Ron Marshall]

Kirk, When you do multipass scanning to reduce noise do you find there is a loss of image definition , or does the Studio version of Silverfast preserve image alignment?

I have Silverfast Ai, but have hesitated to multiscan because I worried that successive scans would not register. If the Studio version is worthwhile I will upgrade.

[Mike Lewis]

Prof. Evens-- I tried negative film, in my case Fuji NPS, for the reasons you described. However, after scanning (Epson 4990) and printing, the results do not seem as sharp as with Provia. Admittedly I've only tried this a couple of times so I don't really know what I'm doing yet. I'd like to print 16x20 with my own scans. If I can get sharper prints I'll continue, otherwise I'll give it up and go back to Provia. What size do you print? Any other hints? Thanks.

[Kirk Gittings]

Ron,

The studio version has an alignment program built into the multipass. It works well except with excessive movement (stretching from heat etc.). I solve this by doing a batch scan-first one is with one pass to heat up the film and the next one is the real scan at 8X.

Mike I print 4x5 NPS negatives at 16x20 all the time with appropriate sharpening. I use Photkit Pro Sharpener.

[Ed Richards]

All versions of Silverfast has the alignment feature now, so you can try it with base line program. Vuescan has multisampling for many consumer scanners, so it all done in one pass. This prevents alignment issues, but slows the scan so sometimes the negatives "pop" from the heat.

As one of our other members has pointed out, the Negpos system is by far the best thought out approach to handling color negative scans:

http://www.c-f-systems.com/

I do not do much with color, so I have not gone beyond reading their papers, but their theory seems well thought out and is based on dealing with both gamma and color together.

[Leonard Evens]

Mike,

The largest print I've made digitally is 13 x 17 because that is all my Epson 1280 can do. But examining images on the screen and comparing with prints, I think I could go to 16 x 20 and produce prints sharp enough for me. But they might not be sharp enough for you. I haven't noticed any difference with the few transparencies I've scanned, but I haven't made any systematic tests. It may well be that some transparency films surpass Portra VC 160 in resolution, but I would be surprised if one could see much difference with any of the Epson scanners.

[Kirk Gittings]

"All versions of Silverfast has the alignment feature now, so you can try it with base line program."

"Get rid of Noise! SilverFast Multi-Sampling with Auto-Alignment is here!
Ships with SilverFast SE Plus and SilverFast Ai Studio." from the SF site yesterday.