Hi all,
Me revopoint metrox crashed during a save, leaving me with an empty .Revo file. All the data folder contents seem to be unharmed, but I can’t get them to show in a project.
I tried copying them to another new project and adding the nodes manually through editing, but can’t get them to show up. Are they tied to the project guid somewhere?
Any help to recover or add to a new project would be amazing, as I currently have the data of about two days of scanning inaccessible.
If you copy the data folder into a new project, you should only have to recreate the “nodes” in the .revo file. GUID is the folder name of the scan (the subfolders in the data folder). “type” probably matters as well but you can figure that out from other projects which one is which.
That is what I have tried.
I have added it to another project, copied the data, and appended the nodes section so all nodes should be included. However, every time I open the project in revoscan, only the nodes originally scanned in that project show up, and the Revo file reverts back as well.
Never change any names in the scan data folders beside the main project folder name, you should be able to open it again after recreating manually the *.revo files
Open Revoscan, create project, exit Revoscan. Copy the data folder of broken project into new project. Edit the new .revo file and add a “node” for each subfolder in the data folder (using folder name as guid). I just tried it, it works fine. Make sure you exit Revoscan before editing the .revo file.
Revoscan must see something it doesn’t like in the scan folder. Are you sure the scans themselves are intact? Is there a property.rvproj for each? Does the guid in that file match the folder name? In the cache folder, is there a frames.dataset file of not 0 size?
There is a log file in appdata/local/revoscanmetrox/logs that might tell you something. If I intentionally screw up a guid it logs the failure to “open sub project” but is not helpful enough to say why. It would at least tell you it’s trying to open the copied scans.
It seems i can fool it by copying those over from a working dataset, trying to figure out what impact the definitely wrong framecount has on processing them