Thanks, that RedEdit procedure seems to have worked. The only thing I struggled with there, as someone who only very rarely uses RegEdit, was that in RegEdit, Edit -> New -> Key produced a folder under Qimage-U; I did Edit -> New -> String Value, then made the name AppFolder and the data the correct path for the user account. After that Qimage started up fine.

Thanks for the prompt reply, and that may be an issue, but it is not the issue. I uninstalled Qimage, copied the installer file over to the user account, switched to the user account, installed, ran, and got the same or a very similar error message. Huh. So I uninstalled again, and reinstalled again under the user account (i.e., for the second time), watching it closely. What I noticed is that even though I was under the user account, of course I had to enter the admin password to start the installer, and then the Qimage installer tried to put the AppData under the admin account instead of the user account. I manually pointed it to the same location except under the user account, and continued with the installation. After installing, I tried to run it, and got the same or a very similar error. See attached screen captures.

I'm setting up a new computer, Windows 11 Home. I have an administrator account that I won't use for normal use, and a user account. I installed Qimage Ultimate 2023.112 from the administrator account, and under that account it works fine. But when I try to start it from my user account, I get the error message:

Qimage is unable to write to the folder:
C:\ProgramData\ddisoftware\Qimage

Please reinstall Qimage in a folder to
which you have write access.

(Screen capture also attached.) All the other software I've installed (MS Office, DxO, Affinity Photo, etc.) seems okay. Any suggestions? Thanks!

Thanks for all the thoughts, and to update my issue:

(1) I realize that printable L*min will never be 0, although with a few combinations the ICC profiles indicate it should get down to about 3. With the Pro-100, Canon Luster, and Canon's ICC profile (the file is CNBBDUA0.ICM), the profile indicates L* min = 5, and that's close to a* = b* = 0 (actually about at b* = -1). At precisely a* = b* = 0, profile-indicated L* min is about 5.5. I'm basing this on observing the 3D gamut plot, rotated and enlarged, at https://www.iccview.de/, a screen capture of which is the first attachment. But measured with the ColorMunki, a file printed in Qimage with rel. col. + BPC, the printed image patch of L* = a* = B* = 0 measures L* = 9.

Similarly, with the Pro-100, Canon Glossy II, and Canon's ICC profile (the file is CNBBDMA0.ICM), profile-indicated L* min = 5, but that's at about a* = 0, b* = -3. I had written to see the second attachment, but with it the board reported I'd exceeded my allocated 256 KB. At precisely a* = b* = 0, L* min is about 7. The printed patch of input values L* = a* = B* = 0 measures L* = 11.

So it's seems like there's an issue significantly beyond L* min not being neutral and/or being substantially above 0.

(2) I'm aware of, and use, the X-Rite software's ability to make special B&W profiles, and I intend to do more testing, partly around that. But to try to isolate variables and start with the most current, widely-used combinations, I decided to start with Canon's main papers and Canon's own profiles. My sense and limited testing indicates that X-Rite-built B&W profiles deliver slightly more neutral grays but no increase in Dmax / decrease in L* min.

(3) FWIW and as a comparison and control, I got two targets printed on Ilford RC silver halide paper, Pearl surface from Mpix (on which the target patch with input values L* = a* = B* = 0 measures L* = 7, a* = 1.0, b* = 2.4) and Glossy surface from Fromex (on which the target patch of L* = a* = B* = 0 measures L* = 8, a* = 0.8, b* = 2.2). Also, taking repeated measurements, over different days and after different calibrations, with the ColorMunki Photo shows that all measurements are within about +/- 0.3 of each other, and many are closer than that. In other words, I can't vouch for this spectrophotometer's accuracy (reporting the true value), but I can vouch for its precision (repeatedly reporting the same value).

For some reason I cannot get black to print maximum black using Qimage Ultimate 2023.103, and I think it is likely to be some setting issue I'm misunderstanding.

I'm trying to assess B&W printable range and linearity. I made a test image file with 10 patches from pure black to pure white, i.e., L* = 0 to L* = 100 in steps, all with a* = b* = 0. Lightroom's Develop module confirms that the file's black patch is L* = 0, and all the other patches are +/- 0.2 from the desired L* values. An RGB color-picker likewise confirms that the black patch is 0,0,0. ICC Profile Inspector reports that the ICC printing profile used has an L* minimum of 5, and a visual inspection of a plot of the profile at the ICC View website confirms this.

I printed to a Canon Pro-100, using the same driver settings I use to build ICC profiles with my X-Rite ColorMunki Photo spectrophotometer, color matching = none etc. In Qimage I have it set for relative colorimetric rendering intent and black point compensation--which I think should make the 'black' patch come out about L* = 5. By actual measurement with the ColorMunki, the 'black' patch is L* = 9. I got similar results with two standard papers, Canon Pro Luster and Canon Glossy II. I went back and re-measured some old profiling targets, and indeed the ColorMunki reports as expected the 'black' patch is about L* = 5.

Am I wrong in thinking that using relative colorimetric with black point compensation, an area of L* = 0 in the file should print as maximum printable black i.e. about L* = 5? Is there some other setting you think I might be missing?

Thanks!

Thanks for the quick reply, but that is definitely not how my copy (version 2022.120, so current) is behaving. I even wondered if somehow auto-cropping had been turned off when I saved the printer settings for the paper I was using, so I laid out a couple of images on a page, turned on auto-cropping for each, did a "Save current printer settings" over the older one for the paper I was using. When I go to layout the next page, auto-cropping is again off.

I'm not sure whether this is new or longstanding, but I thought auto-cropping used to stay how it was set until I changed it, and with recent versions it seems to revert to off after every print. I prefer to have the setting be durable, or at least that to be the default behavior. This is causing really irritating behavior like my choosing from the Prints tab 5.1x7.1" (my custom size, to trim to borderless 5x7" when printed on letter-size pages) and, despite my custom cropping, it gives me the original aspect ratio so 4.72x7.10". I guess I have to go in and manually turn on auto-cropping for every image I'm laying out.

So I think this is a change in Qimage's behavior and I find it very irritating. Hopefully there's an easy fix!

Thanks!

Roy: thanks for the generous offer, but my question was so I'd be better-prepared for the next time. In the particular case, I'd already done what I did and that ship had sailed. The guy called me on a Saturday night and needed a file to be printed and mounted to be used as a 'Sgt. Pepper'-style part of a group photo that Tuesday or Wednesday.

Mike: Doh! I don't know why I didn't think to try printing to a file. I've done that from Lightroom, but Qimage has a lot of options and capabilities and I've only scratched the surface in the maybe eighteen months since I licensed it and use it instead of LR for printing. I see that sharpening is adjustable in Qimage's print-to-file function. Does it use the user-selected interpolation from the Processing area, or does it always use Fusion, or what?

FWIW, considering the file I prepared was almost a 5x upscale, I was pleased with how smooth and non-weird/artifacty it was with Qimage's Fusion interpolation. My approach to sharpening was probably not the best, but all things considered, this was a good experience.

I'm looking for advice on what are the best techniques to use Qimage Ultimate to upsample a photo and export it as a JPEG. Recently I was asked for one of my photos to make a life-size cutout of the requestor's daughter. In my photo she is 3469 pixels tall, and I figure the print was going to run about 4'9" long / tall, so natively that's only about 61 ppi. Generally I'd assume and advise that anyone serious enough to have a 24" photo printer and roll paper should probably be left to do final file prep however is best, but based on the requests I got, the time urgency, and my sense from the discussion, the guy with the printer was just somebody with access to the printer who was printing for a friend as a last-minute thing. So I:
(1) opened the TIFF from the raw converter in Qimage Ultimate,
(2) in the HQ Sharpening field I set the radius to I think 4 and the % to I think 100, and
(3) did a Save As... and then in the Output Options box checked Resample (Fusion) box and set the length to 17100 pixels.

My most basic question is sharpening: is that radius applicable to the resampled 17100 pixel length (what I wanted) or the original 3469 pixel length (for which a radius of 4 would be really high)? Is there some way to do post-resampling sharpening, other than saving the file with resampling, then opening the resampled file and sharpen it, and save it as a third version?

Also, is there some recommended or preferred resampling procedure when the goal is to output a file?

Thanks!

P.S.
I don't have Gigapixel AI and it won't even run on my old computer. Also, Qimage's resampling is fine for my own needs and the sizes I can print. So really, my choice was between Qimage and Affinity Photo.

Thanks everyone for the suggestions, but I guess I was not clear. The rotation / geometric fix would have to be done before the image file gets opened in Qimage. Typically I do this in DxO PhotoLab with ViewPoint. Horizon rotation is most common, but sometimes multiple other adjustments are made. I used the horizon rotation example just because it is simple.

Attached is a more complex example. What I want is to be able to tell Qimage: 'I want to crop this photo to a 5:4 aspect ratio, and be able to increase or decrease the size of that 5:4 crop box and move it around the image area, but don't allow any of that black border region to be included in the area to be retained after the crop.'

IMO, Qimage would really benefit from a way to constrain the cropped subset of an image file to be printed to the actual image area. That can be tough with, e.g., images that were rotated a degree or two, and then rendered and outputted without a crop. To explain: as I do my raw conversions (most often in DxO), it's not unusual for me to rotate the horizon a degree or two, or sometimes make other geometric adjustments (e.g. simulating front rise on a view camera). The result is saved as a TIFF where the actual image area is some irregular (e.g., trapezoidal) shape within the rectangular TIFF. The simple case is a horizon rotation, where the file is mostly the image, but with black triangles along each edge. I don't crop it in the raw converter because I may choose different image areas depending on the aspect ratio of the final print. An example is attached.

When I go to print the image with these black triangles (or whatever), I have to adjust my crop so as not to print any part of then. Sometimes that's not easy, especially when you're trying to get as much image area as you can without getting any of the non-image (black irregular border) area. It would be great to have a 'constrain my crop to the image area' checkbox, which would then bump a crop that would be outside it just to the edge of it.