How to test if an image was printed at 16 bit

measurements

#1

I’m a trust but verify kind of guy and I’ve been asking around and sometimes directly to the companies how I can print at a 16 bit depth and how I can verify that nothing in the path as truncated my image from 16 to whatever other bit depth along the way.

I’ve seen an image from Norman Koren to help us calibrate our monitor to the proper gamma (though old) and this gave me the idea that it may be possible to create an image that could create a visible difference when printed at 8 bit compared to when printed at 16 bit. Any idea how to do something like that?

Thanks


#2

Good day,

I doubt that there exist images that use the whole dynamic range possible with 16 bit. Best digital cameras provide up to 14 bit per channel.

On the display side there is no monitor that can display 16 bit. Presently, the maximum is about 12 bit per channel.

What do you mean by “printed”?

Print in the classical sense of print on paper uses rasters and commonly can’t reproduce 8 bit per channel.

Photo prints (i.e. what used to be called electro-photography) depend on the photographic paper and won’t allow much more than 8 bit per channel.

Remains the question, why 16 bit images are obviously around. The answer is that computers go with 8 bit, 16 bit, 32 bit, 64 bit. Consequently and if you have a camera delivering, say 12bit, you represent those as 16 bit images …

Does this help?

Regards

Herbie


#3

All you say is fine with me, true today we shoot with camera capable of 14 bit which is finally easier to handle using a 2 byte boundary or 16 bit and if you add any post processing to that, it’s very easy to lose a few bins in your histograms even if it’s much less visible then what we use to see at 8 bit. I’m talking about inkjet printing with modern printer, inks and papers. I just want to know if it’s possible to visually assess if a print was make with a 16 bit pipeline or an 8? If no one as an idea to create such an image, I’ll wait a bit and see if a spectrometer can but I don’t have one now.


#4

I’m talking about inkjet printing with modern printer […]

This technique won’t give you much more than 8 bit per channel, i.e. I see no real chance to distinguish between images coming either from an 8 bit or 12 bit source.

I’ll wait a bit and see if a spectrometer can but I don’t have one now.

I wouldn’t bother with such investigations.

Regards

Herbie