Suite2P
Suite2P copied to clipboard
Load ScanImage File
Hi,
We are trying run our scan image files through the pipeline and need to create a new load function. We have adapted the Matlab version of suite2p to load the scan image files but are wondering if anyone has done this with the python version. If not, where can we find the current load file function within the python files? Thanks so much!
Can you give more details please? The Matlab version has always been able to load scanimage files. It also read the number of planes and channels from the headers. It is possible that with a different version of scanimage the headers are formatted differently.
As for the Python version, we have given up trying to load metadata from headers, because the header formatting can vary a lot between different types of tiffs, and no tiff loader that we tried worked reliably across types of tiff headers. Instead, you can put any metadata information you'd like in an ops variable that you can load in the GUI as either an npy array or a json file.
If this is about mROI scanimage tiffs, those are supported in the Python version only. Open an issue in the Python repository if you want to learn more about it.
We have to use an older version of scanimage which required us to change the load function for the Matlab version. When we use the current load function and load our scanimage tiff into the Python version, we are 800 frames short of our Matlab version. Since we had to change the functionality of the Matlab version and haven't done this with Python, I do not think that we are missing these 800 frames because they were considered to be bad frames. I think it is because our tiff isn't being read properly. Sorry, I hope this makes sense.
What data type are your tiffs? I'm thinking they might have fewer bytes... I'm not really sure. We expect int16 tiffs.
They are int16. Would it be helpful if we shared a sample movie with you?
Yeah that would help, thanks
Perfect, thank you! I sent an email to [email protected] with a sample movie attached.