Raw & Post Processing, Printing > Digital Image Processing
Is there any disadvantage to saving TIFFs in 8bit instead of 16bit?
Dinarius:
The question is in the title.
And, while I can get many answers to the benefits of editing in 16bit (banding, colour range, etc.), I can find no answer to my question regarding saving TIFFs.
Thanks.
D.
kers:
--- Quote from: Dinarius on December 01, 2023, 04:01:01 am ---The question is in the title.
And, while I can get many answers to the benefits of editing in 16bit (banding, colour range, etc.), I can find no answer to my question regarding saving TIFFs.
Thanks.
D.
--- End quote ---
Well in 16 bit you have more information about the image. As an end product for printing 8 bit will do fine. Information is also the amount of pixels and the colourspace.
I always keep my photos as flat 16 bit prophoto as large as possible after they are ready.
In that way i can still change them in future without any penalty. If you use prophoto colour space it is also advised to keep them 16 bit because of the enormous size of that colorspace.
data storage is very cheap nowadays and even the harddisks get smaller...
Dinarius:
Thanks.
Both Metamorfoze and FADGI advocate 16bit as archival standard without stating, as far as I can see, why it's better than 8bit.
And I totally agree that the cost of storage now makes the question redundant. I just thought I'd ask! :)d
D.
digitaldog:
The advantages would be any further editing of that data. So, for example, you had a TIFF with output-specific edits based on a soft proof to a fixed output need no; you're not going to edit that iteration again, 8 bits per color is acceptable for all output needs (yes, even to print).
Master, layered data: Keep high bit. Here's why:
http://digitaldog.net/files/TheHighBitdepthDebate.pdf
Dinarius:
--- Quote from: digitaldog on December 01, 2023, 12:39:36 pm ---The advantages would be any further editing of that data. So, for example, you had a TIFF with output-specific edits based on a soft proof to a fixed output need no; you're not going to edit that iteration again, 8 bits per color is acceptable for all output needs (yes, even to print).
Master, layered data: Keep high bit. Here's why:
http://digitaldog.net/files/TheHighBitdepthDebate.pdf
--- End quote ---
Thanks!
D.
Navigation
[0] Message Index
[#] Next page
Go to full version