Settings for archiving photos? TIFF, 48-bit?
 

I am trying to determine the best settings for archiving photos. There are from a Kodak Porta. I have read online that the color space should be sRGB and that TIFF with ZIP compression and PNG (which has ZIP compression) are equally good for scanning. I know ZIP takes longer than LZW but it always decreases file size. But some have wrote that PNG is better because it is a more standardized format than TIFF for long-term use. (Yes, TIFF has metadata, but that is not that important to me.) They also claim that PNG is more widely supported than it was a few years ago and there is also a version for 48-bit color (but I could find little mention of it).

Is it worth scanning in 48-bit? From what I read it does not hurt, but is not as essential if there will be no post-processing. These photos are in good condition, unlike my videos. I also do not know what DPI setting approaches the point of diminishing returns. Would it better to have, let's say, a 48 bit photo scanned at 800 dpi or a 24 bit photo scanned at 1500 dpi? Thanks for any input...

Let me put it this way for you:

Until recently, I was a professional photographer. Long story, lousy industry now, focusing on marketing and hosting these days. Excluding narcissistic kiddies that shoot too many selfies, I probably shot more images on some days than most people shoot in a year. Sometimes it was RAW, sometimes JPEG. My camera gear cost more than most cars.

I find the attitude of "only this" or "only that" to largely be clueless advice, when it comes to image formats. Again, I shot RAW and JPEG. TIFF can be better than JPEG (level 10-12 max quality compression), but how noticeable will it actually be? TIFF is comparable to Huffyuv for video, while JPEG is comparable to MPEG. But the important difference here is that the images are not in motion. Differences become more negligible. TIFF compared to RAW is laughable; it's NOT the same as was TIFF vs. JPEG in quality.

My cameras shoot anywhere from 12-bit to 16-bit, which is used for RAW, but JPEG is 8-bit. And I have JPEG images that won awards. I shot film year before digital, and won awards in that time as well. Quality images are not determined by compression.

sRGB vs. Adobe 1998 vs. others depends on the print workflow.

Most important aspect = intended use. So, what's your intended use?

If you're working at the Library of Congress, yeah, sure, scan those huge images at max bit depth and resolution. We need valuable photos to be retained for history. They're not scanning tons of images, and can afford the time, equipment and space.

But if you're scanning an old box of home photos, that's insane. I scan our old home photos as 8-bit sRGB JPEG in most cases, and only do TIFF or DNG for images in need of restoration. And I'm still not done with that project. It's taking forever. I've dedicated full days to it in the past, and barely dented a box of images among many.

Thank you. From what I have read Adobe RGB would be a better choice if the photo would be used for both web and print. Currently I only intend to digitally archive these photos, but something may change down the road, if the Library of Congress gets in touch....

The online consensus seems to support scanning in TIFF, especially if the photos will be further edited. If I find the negatives for the photos I scan I would re-save the TIFF as a high-quality JPEG.

Regarding DPI, some sites advocate that there is little benefit gained from scanning 1-hour type photos at a Q factor more than 2.0, or 300 dpi. Others state that since the claimed dpi of the scan is often less, and due to the benefit of oversampling, a dpi of 600 should be the target. (Then again, the last author wrote that he didn't "trust" lossless TIFF compression, and recommended uncompressed TIFF.)
https://the-print-guide.blogspot.com...ng-lpi-vs.html
https://blogs.loc.gov/thesignal/2013...i-is-too-much/
http://www.scanyourentirelife.com/dp...r-photographs/

The other issue is color bit depth. From what I have read scanning a color print at 48 bit is likely a waste because it cannot represent that range of colours. For scanning of negatives, it appears a dpi of around 3000 is recommended, though if more is better I would want to scan higher up to the optical resolution. I would also scan at 48 bit. But I do not know if scanning at a better dpi is preferred over going to a higher bit depth, if the photo was taken professionally.