Scanning color transparencies with a digital camera : LUSENET : Large format photography : One Thread

What kind of quality do you get by photographing a transparency on a lightbox with a P&S digital camera ? The goal here is obviously not to print, but rather to display on the web. At a low resolution such as 768x512 pixels, will the quality of the image be obviously lower than a scan made with a proper scanner ?

-- Q.-Tuan Luong (, January 06, 2001


I tried this recently using 6x6, 6x7 and 6x9 slides and a 50mm macro lens on a Minolta RD-3000 but was disappointed with the results. If you start with a larger transparency, though, I'm sure it would have worked much, much better.

I also tried "scanning" 35mm slides using the RD-3000 with a vintage Spiratone slide duplicator and found the results somewhat better but not as good as when the slides were scanned using my now-vintage and technologically obsolete (1950dpi, 3.0DMax) Microtek 35-t+ scanner.

My last experiment was using the RD-3000 as a pseudo-digital back on my Toyo 23G view camera and surprisingly, this actually worked quite well. The only downside is a multiplication factor of roughly 2.25:1, which I found a bit too high to be truly practical. For proofing or double-checking exposures, though, it might prove handy on occasion, especially if you already own a digital SLR for other reasons.

-- Jeffrey Goggin (, January 06, 2001.

The Leica Digilux 4.3 (manufactured by Fuji - and the same as its 4700) has an option called Digicopy. According to the Leica site: "Slide Copying Accessory for LEICA DIGILUX 4.3 LEICA digicopy 4.3, an optical attachement designed by Leica exclusively for the LEICA DIGILUX 4.3 is perfect for digitizing slide films from all manufacturers, but also black-and-white and color negative material. High quality reproduction and archiving of your valuable originals becomes very easy by this means."

One of the earlier threads of a few weeks back also mentioned this same item.

The Digilux site is at: digilux43/index_e.html

The camera itself is rather economical, as is the Digicopy accessory. By economical, I mean less than $1,000 for the two items. There is also a macro accessory.



-- Alan Agardi (, January 06, 2001.

Another thought: project the slide with a regular slide projector, and use your digital camera to photograph that image. It may be crude, but it could also be effective.



-- Alan Agardi (, January 06, 2001.

for most website applications, you generally dont want or need more than 100 dpi. Good quality monitors display at 72 to 96 dpi, so there is no reason to increase the resolution above 100 dpi if the image is not to be printed. for webpage purposes, it is easiest to simply have an 8x10 print made from the CT and scan the print. i maintain a website that features some 50-60 LF images scanned from both prints and direct CT scans, and trust me, the subtlety is completely lost on the internet. however, if your intent is to supply a user with high-resolution files which they can make high- quality prints from, you should take the CTs to a good local lab and have them scanned properly, at between 300-600 dpi, with 20-50MB file sizes.

-- jnorman (, January 06, 2001.

Tuan, I have not tried but I guess the problem is contrast. Unless you have a possibility to adjust it, the results might be of limited tonal range, just as when you duplicate slides on standard film without contrast control. I know Fuji has a "slide digitizer" for some years. Haven't seen the results.

-- Paul Schilliger (, January 07, 2001.

Olympus too, are advertising a slide duplicator attachment for some of their digital cameras. They claim it works with negatives as well as slides, but I doubt that 24 bits is sufficient to correct the orange mask which causes a severe offset in the RGB curves (Unless they have a cyan filter built into it. There's a thought!).
It would be my guess also that the contrast and brightness range of a slide is too high for a digital camera to cope with properly. A typical slide has a gamma of 2, and a brightness range of over 2000 to 1. Compare this with a sunny outdoor scene having a 'gamma' of 1 (naturally) and about a 1000 to 1 range.
24 bits just can't cut it. Dedicated scanners struggle with their 36 bit dynamic range.

BTW, and strictly off topic, isn't it about time that one of the big film manufacturers made a low contrast slide film, specifically for the purpose of scanning?

-- Pete Andrews (, January 08, 2001.

Pete's correct as the problem I had trying to copy slides off a light table is that the resulting images were dark, with very little shadow detail. This was less of a problem with the duplicator setup since I was using a flash as the light source and had the ability to increase or decrease its output as necessary but the subtle details were still lacking. It didn't help matters that most of the originals were shot with Velvia as the contrast level was quite high to begin with.

-- Jeffrey Goggin (, January 08, 2001.

Moderation questions? read the FAQ