Export best settings

[Tehillim]

Good day!!

 

I'm preparing some photos for huge print 300x240 cm. I start from a RAW file from my Canon 5D mark III. I develop it using canon software end I end up with a TIFF file that usually weights 120mb. I open this file in Affinity photo and when I export it (alway in TIFF rgb 16bit) the file only weights max 30mb...(the AF PHOTO project weights at least 350mb). This looks kind of strange to me and I fear I'm losing quality.

 

I need to export at max quality for printing...I'm I doing something wrong?

 

Please help

[Peterkaosa]

Not really helpful, but I find it interesting that you're going for 16bit – how come? I've once tried and sought it a bit odd and not very useful. But I like learning ;-)

[barninga]

AP compresses file with ztiff algorithm when saving. it's a non-lossy method, so it retains quality and give advantages in size.

[Tehillim]

Yesterday I've found that when you apply the adjustment layer "selective color" for some reason the file size changes from 120mb to 30 mb...this is weird...nobody reported something similar?

[R C-R]

File size is by itself not a good indicator of image quality. You also have to consider how much metadata the file contains, how compressible it is (& if the compression is lossy or not), any file format specific constraints that might be involved, & even what "quality" means in real world terms.

 

For example, the TIFF file format specification allows for the inclusion of preview images (thumbnails) in JPEG format, a wide variety of optional metadata types, & other things that do not contribute to image quality at all.

 

Also consider that a D50 MK II outputs in 14 bit RAW resolution, so the 2 least significant bits in a 16 bit data stream contain no image info at all & can be discarded by compression algorithms with zero impact on image quality. Processing applied during development & editing can "fill in" the two extra bits with finer resolution details, so it is worth working in 16 bit color in the app, but outputting to 16 bits doesn't really increase real world quality because we can't distinguish more than around 16 million shades, & 8 bits is enough for that: 8 bits can encode 256 shades per color, so with 3 color channels you can combine them in 256x256x256 or 16777216 ways. 16 bits can encode 65536 shades per color, so that equates to over 280 billion shades, millions of times as many as we can perceive. Even if we could, good luck finding a printing process that could actually print that many shades!

 

Besides that, no real world digital photo taken with a D50 MK II will have all those billions of shades -- how could a 21 million pixel image do that? That's why compression can reduce file size dramatically: all the file needs to store is information about the shades of the pixels actually in the photo, & there are several ways to do that without any loss of quality. That's also why any process that reduces the number of shades or changes their relative placement (like selective color replacement) can also reduce file size.

 

So really, there is nothing weird about it.