I'm using Bonsai to record and save calcium imaging video data. The data is 10 bit grayscale, which we package as RGB565 to avoid having to deal with 16 bit grayscale formats. I've been using the VideoWriter node with "DIB " as the FourCC to encode the data, but this results in enormous file sizes. I'm interested in using the FFV1 codec instead since it is lossless and should result in smaller file sizes. However, when I use two different VideoWriter nodes to save two copies of the same video stream using DIB and FFV1, the individual pixel values differ slightly between the two copies. Why is this happening, and is there anything I can do about it? Thanks!
Joe