Is 12-bit-per-channel a sensible archival compromise for scanned 35mm negatives?
Asked 5/10/2025
3 views
2 answers
0
I’m digitizing several thousand 35mm color negatives over a period of years and need to choose an archival bit depth. My current workflow starts with high-resolution 16-bit-per-channel RGB linear scans, then I invert the negative, remove the orange mask by stretching each RGB histogram, and apply per-channel gamma corrections until the image looks close to final. At that point, only minor future edits are expected, such as small color corrections or modest shadow recovery.
Because storage and backups for thousands of large files are a concern, I’m wondering whether reducing the archived files from 16-bit to 12-bit per channel is a reasonable compromise. I want to preserve some editing headroom beyond 8-bit, but I may not need the full flexibility of 16-bit once the major tonal/color work has already been done.
Is 12-bit a practical middle ground for this kind of archive, or is it better to keep 16-bit and focus on compression or other size-saving measures instead?
Originally by Mephisto. Source · Licensed CC BY-SA 4.0
Mephisto
1y ago
2 Answers
2
This is just my opinion, not an informed one.
I do not think it is a good practice, but it is up to you.
Photography sometimes is about interpretation, so having more bits allows you to re-interpret the image, sometimes later.
Probably some other processes will have more impact on the file size than the bit depth, like the compression algorithm, and... ready? Noise reduction.
See if you can apply noise reduction early in the process and compare the results.
Noise basically adds unnecessary information to the file. Instead of assigning the same color to a big area, it assigns super small variations everywhere.
Originally by Rafael. Source · Licensed CC BY-SA 4.0
Rafael
1y ago
0
Generated from our catalog & community — verify before relying on it.
Based on the answers, 12-bit can be workable in some cases, but there isn’t enough certainty to call it a best-practice archival choice.
Whether the reduction is safe depends on factors like film grain, scanner noise, image resolution, and intended output. With high-resolution film scans, grain/noise can act a bit like dithering, which may make lower bit depth less visibly harmful. But archival files are often kept at higher bit depth specifically to preserve flexibility for future reinterpretation and editing.
The stronger guidance from the replies is: if long-term archive quality matters, keep 16-bit if you can. If storage is the problem, look first at other ways to reduce file size, especially efficient lossless compression, and possibly noise reduction earlier in the workflow if appropriate. Noise and grain can increase file size without adding useful image information.
So: 12-bit may be acceptable for near-final images needing only minor tweaks, but for archival masters, 16-bit remains the safer choice unless testing on your actual scans shows no meaningful downside.
Recommended products
UniqueBot
AI1y ago
Your Answer
Related Questions
How many bits of data does a digital camera sensor actually capture?
Why does Photoshop’s RGB picker still use 0–255 in a 16-bit image?
What’s the difference between color space and bit depth?
What does a fractional color bit depth like 23.5 bits mean on camera comparison sites?
How does editing in 16-bit Photoshop relate to 8-bit, 10-bit, and “16-bit” monitors?