Black screen

[fepegar]

Hi, @apleynes. Thanks for sharing this, it looks great. I gave it a try with multiple image and I always get a black screen (I can only see the controls at the bottom).

My command is uvx pydcmview colin27_t1_tal_lin.nii.gz.

neofetch output:

                    'c.          fernando@hermes.local
                 ,xNMM.          ---------------------
               .OMMMMo           OS: macOS 14.7.4 23H420 arm64
               OMMM0,            Host: Mac14,15
     .;loddo:' loolloddol;.      Kernel: 23.6.0
   cKMMMMMMMMMMNWMMMMMMMMMM0:    Uptime: 76 days, 10 hours, 42 mins
 .KMMMMMMMMMMMMMMMMMMMMMMMWd.    Packages: 184 (brew)
 XMMMMMMMMMMMMMMMMMMMMMMMX.      Shell: zsh 5.9
;MMMMMMMMMMMMMMMMMMMMMMMM:       Resolution: 1710x1107
:MMMMMMMMMMMMMMMMMMMMMMMM:       DE: Aqua
.MMMMMMMMMMMMMMMMMMMMMMMMX.      WM: Spectacle
 kMMMMMMMMMMMMMMMMMMMMMMMMWd.    Terminal: iTerm2
 .XMMMMMMMMMMMMMMMMMMMMMMMMMMk   Terminal Font: MesloLGMNFM-Regular 14
  .XMMMMMMMMMMMMMMMMMMMMMMMMK.   CPU: Apple M2
    kMMMMMMMMMMMMMMMMMMMMMMd     GPU: Apple M2
     ;KMMMMMMMWXXWMMMMMMMk.      Memory: 1542MiB / 8192MiB
       .cooc,.    .,coo:.

[apleynes]

Hi @fepegar, thanks for posting the issue and sharing your neofetch! I can confirm the issue on my end with iTerm2. In the meantime, you can either use kitty terminal or use the unicode fallback. For the unicode fallback, specify COLORTERM="" in your environment or right before running the command like COLORTERM="" uvx pydcmview colin27_t1_tal_lin.nii.gz

The graphics backend I'm using claims that it should work with iTerm2. I'll investigate and keep this thread updated.

[fepegar]

Thanks for your reply. It works with kitty, but it's rotated and the other views don't work (I always see this sagittal view).

The unicode fallback works on iTerm2, but the resolution is very small (I guess that's expected?).

[apleynes]

Hi @fepegar , thanks for the patience.

I fixed the problem with other terminals and it should work with iTerm2 and other terminals where textual-image can run (see support matrix here https://github.com/lnqs/textual-image). I've only tested it locally though and can't guarantee remote SSH.

The new version with the terminal display fixes is pushed to PyPI (v0.1.6). Just do pip install --upgrade pydcmview

Regarding the views, to confirm if I understand correctly: specifying the dimensions (t) and changing what x/y are don't change the image? If you can share the file so that I can try reproducing the issue, that would help a lot.

For unicode fallback, the resolution is then controlled by the font size of the terminal and the zoom level. There's some experimentation to get the right mix but yes it is definitely not a great experience and really just a worst-case fallback to see something.

Thanks!

[fepegar]

Hey @apleynes. The file is the T1-weighted NIfTI from https://nist.mni.mcgill.ca/colin-27-average-brain/