Printing with Piezo Pro inks in a 9800 to Hahnemuhle Photo Rag 308gsm…
With regard to image sizing - I have usually followed the advice in the manual and by Jon and Walker on these forums - i.e. do not resample image, just set the dimensions to whatever you want the file to print as, let the ppi fall where they do and send to QTR.
But my understanding is that QTR will always print at 720ppi ? And therefore I have never fully understood the reasoning behind not resampling in Photoshop - seeing as the files are going to be resampled (by an unknow algorithm?) in the print pipeline (by QTR).
Now I am working with some files shot by a client on a Leica Monochrome.
High frequency image with lots of edges.
The supplied file has a pixel dimension of 5976x3992.
I want to print at 55cm wide image size (about 21 inches I think)
That gives me a ppi of about 276 without resampling (i.e. native size)
I decided to do some tests. I printed one version at native size (55cm at 276ppi -no resampling), a second version at 55cm at 360ppi and a third version at 55cm at 720ppi (2nd and 3rd versions resampled in PS using Preserve Details)
The ‘native size’ versions seems to my tired eyes to smoother and softer
The 2x resampled versions do seem to have sharper and have more resolved detail.
To be honest not sure which I prefer for these particular high frequency images.
My point is that there does seem to be a difference.
I would be very interested to hear what others experience and advice is to with regards setting size/ppi/resampling - especially for prints that need to be ‘up’ sampled?
e.g. if your file PPI is less than 720 at print-size then think about up-res in PS but if that makes the image too huge or too noisy, strike a middle ground.
If your file is very high res (1200 ppi) at print size you don’t really need to worry about downsizing first.
If you file is 1.6GB + and you are printing small, downsize so-as not to choke the printer driver.
It would be interesting to know if you are printing on macOS using Print-Tool (and the QTR driver) or on Windows using QTRGui (which uses several underlying QTR Windows executables to rasterize and print). I think it is safe to say that the print pipeline (and resampling) is different for macOS vs. Windows. Roy has indeed confirmed in the past that resampling to 720ppi occurs in either case.
I am a Windows user (mostly) and print using QTRGui to a 7900 using Piezo Pro ink. After some testing (as you have done), I have settled on preparing a 720ppi TIFF file before I print using QTRGui. This allows me to control how (and when) the resampling is done. In my case, I am usually printing from “master” drum-scanned 4x5 negatives that result in large files and I am actually downsizing to 720ppi using a custom ImageMagick script that lets me control the deconvolution sharpening and minimize halos as part of the resampling.
To some extent, I believe the advice normally given on this topic is driven by a desire to keep the printing process from appearing “too complex” for the average user. Nothing wrong with that. For many users, I’m sure there is no obviously discernible difference in print output when just letting QTR (and the underlying OS) handling everything. But if, on the other hand, you can see the difference, then I would say “take control” and handle the resampling yourself.
I also believe the health and tuning/maintenance of your printer can make the difference on whether you can see the difference or not. If alignment, paper thickness, and paper feed settings are not matching the paper being used then it could well be that you won’t see any differences.
Walker, this has confirmed my understanding and makes more sense to me now. Appreciate your insight as always.
Dave, yes I should have mentioned, I am on macOS and using PrintTool and QTR. What you say about taking control of the process makes a lot of sense to me and the idea of sizing to 720ppi before sending to print has always felt like the ‘right’ way to approach it. Having said that, my testing so far (between sending an ‘un-resized’ and ‘resized’ file) has shown only very slight differences in sharpness and noise. The kind of differences that are very hard to see and require careful examination with a loupe and even then are inconclusive. And your point about the health of machine and correct paper settings is an excellent one and something I should definitely pay more attention to.