Scan Size for 35mm

whitecat

Lone Range(find)er
Local time
6:49 AM
Joined
Apr 7, 2006
Messages
1,345
I'd like to hear some opinions on what size files everyone uses for a 35mm BW frame. I have my own opinion but am curious on what everyone else dose. I scan in tiff format.
 
When I'm having them drumscanned, I get scans about 80mb (8-bit per channel, RGB) files from 35mm film. (Unless I need them larger for some reason).

But for shots I'm just playing with, and not specifically printing for a show (or my portfolio), I just scan to roughly 22mb files (2400dpi).

I rarely have 35mm film scanned 16-bit.
 
Last edited:
Don't trust my memory, but I think the 48-bit TIFF files from my HP scanner are in the mid-20Bb range. Something 24-25Mb. My old scanner doesn't do 16-bit grayscale. My new scanner will make 16-bit grayscale files. Stay tuned. I should have it in a week or two.
 
just a question, des 16 bit buy you anything in BW. I have read several times that it does not. Just wanted to confirm
 
The highest no fiddling resolution of my Epson V100 is 3200, so that is what I use for serious work. Files are about 35MB.
 
For the normal frame, I scan at 2400 DPI, 16 bit gray TIFF and 8 bit JPEG at the same time in Vuescan. After post-processing the TIFF, I overwrite the 8 bit JPEG with the new version (80% quality).

The 16 bit seem to give me more headroom for curve and level manipulations. The TIFF is about 7.5 megs, the resulting JPEG about 1.2 megs. When everything is done, I only keep the JPEGs, and of course, the negative ;).

Cheers

Ivo
 
just a question, des 16 bit buy you anything in BW. I have read several times that it does not. Just wanted to confirm

It buys a lot. The point being that you probably don't really need more than 8-bit when you're finished - although I always stay in 16-bit. However, B&W can often require some heavy lifting on the tonality side and that's where 16-bit really becomes your friend and avoids posterisation that 8-bit can easily leave you with. Even starting with an 8-bit image, you can gain in converting it to 16-bit before tonal manipulation then convert back after for the same reasons.
 
It may be overkill, but when I'm scanning a B&W negative for a nice print, I scan in color (search for previous threads on this topic with some tests), 3200 dpi, 16 bit. This makes a HUGE honking tiff file, 70 megs or so, but it captures everything up to and including film grain.

I then make a working copy, usually psd format, to do the actual clean-up and print.
 
there's a formula out there involving film size, output resolution and output size that gives you an exact number to scan at. Right now i forget it because it's pinned up in my office. You may be able to do a search for it, but tomorrow night when i get back from the office I can post it.
 
For casual stuff I scan at 2400dpi in 8-bit. My scanner will go to up 4800, and for more serious stuff I'll go either to 3200 or 4800 and sometimes 16-bit. Actually have some frames I need to rescan at the higher res soon.
 
found the formula.
(dpi(output)) x (printed image long side/original photo long side)=scan dpi

so if you want a 300dpi print at 8x10 from a 35mm neg you get
300x10/1.5=2000DPI scan
 
Last edited:
Back
Top Bottom