Mike Chaney's Tech Corner

Sorry in advance if the has been discussed else where or if this is a RTFM subject.

When patch processing images and saving the output as newly created Tif format files, How can the bit depth be specified to be something other than 24 bits per pixel/8 bits per channel ?

For example, I just processed a set of images where originally 16 bit per channel and the saved Qimage output was 8.

Whether discussed before or not, always good to refresh certain topics.
It is my understanding that, at the present time, Qimage does all processing in 8 bit mode.
It is also my understanding that printing from Qimage is an 8 bit adventure as it conforms to Windows printing routines.
Plans are afoot to move to 16 bit for printing and perhaps batch filtering in the future.
Right now 16 bit images have yet to wow anyone with any visible detectable differences. It just makes bigger files which get converted back to 8 bit for printing anyway.
Best I can offer. Maybe Mike will have time to clarify.
Great question, though!

Fred

Fred-
I agree.  There are those that say they have printed both ways and can see the difference.  Kind of like those that see the difference in bi-directional printing and High Speed Off.  I think if one does it themselves they may want to see it.  At the normal viewing distance for a photograph (2-5x the diagonal) I doubt it.

That said, the paper and droplet sizes in different prints surely are a part of that equation.

As to the larger 16-bit files, I think they should be saved.  All I do is reduce to 8-bit and call it ###A or -8. That's what I print in QI. 
Adjusting 16-bit images in editing software has it's place for quality.

Thank you for your answers.

I do understand the limitations brought about by printed media, but my use of Qimage is more to perform in-process flow manipulations of the type that can't be achieved else where. It was my hope that Qimage had a flag to save files in 16 or 32 bit per channel which minimize banding should the Qimage output file need processing with another image manipulation package afterwards.

With the 8 bit limitation in mind, I need to plan for Qimage being the final step post processing.

I'm fairly sure Qimage will open 16 bit files; it converts to 8 bit automatically. So there is probably no need to make a special 8 bit file for printing.
Terry
Edit: just checked, Q opens 16 bit tif no problems. Image info' say's it's 48 bit (16x3 of course)
T

Yes, I do that if I don't PS it first.  I don't know WHAT Mike does with 14-bit NEF files--to 16 or 8 bit. 

The reason I create an 8-bit from a 16 is only for QI.  Not because it cannot handle it; it just speeds up the "processing" when actually printing.  I don't know that leaving it at 16-bit and letting QI reduce it would be any better quality or not.  Only Mike knows what he is doing there.

If I remember correctly from previous posts on Yahoo, raw files are processed as 16 bit, but in the batch filter it's 8 bit.

He sure does  :) He'll chime in, I'm sure, if what I said is incorrect.
Terry.