Closed MounaSafiHarab closed 7 years ago
@MounaSafiHarab This is probably an HDF5 file using some new feature or version that the existing HDF5 reader doesn't support. Can you send me an example?
@rdvincent
Sure. I sent you a separate e-mail with details of where I placed some example files for you. Thanks!
@MounaSafiHarab Hi, I have looked at this problem and it seems to have been fixed after 2.5.2 was released. Perhaps @natacha-beck could make a 2.5.3? I think this commit is probably the one that fixed the issue: 662e23c4e4a5bcf111a2b6d0e9fc7d53626db31c
@rdvincent and @MounaSafiHarab just done a minor release 2.5.3 that include this fix.
But Mouna still have issue:
so the new brainbrowser fixed the viewing of almost all files except the DTI ones still issue the same error for me so all anatomical (t1, t2) and all fMRI are good
I have put 2 dti sample files that still have this issue in a directory that I will send you the path for in an email.
@MounaSafiHarab @natacha-beck Thanks I should be able to look at it soon.
@MounaSafiHarab @bert I star to take a look to this issue, Apparently it is a HDF5 file but BrainBrowser failed to find the xspace, ysapce and zspace (I can get this information with mincinfo). Then it try to read the file as a NetCDF and failed since it is a HDF5.
@natacha-beck @MounaSafiHarab I have added a commit to my fork that should fix these files, I will do a pull request soon.
thanks @rdvincent @natacha-beck
problem resolved in 2.5.4!
Tested on the brainbrowser website (volume viewer). Loading the minc file directly does not work and gives the error below: "Sorry, this does not look like a NetCDF file." Loading the header and raw files work without a problem. The file comes from a Philips scanner.
p.s: I am using a dcm2mnc binary provided to me by @rdvincent (that solves issues with GE and philips scanners we have encountered in the past) to generate the minc files (in case this info is of relevance).