Video Recorder

Which codec do I choose?

That depends on what kind of data you record. But here is a general rule of thumb:

• If your data analysis needs precise measurements in post-processing and you must store exactly what the recording device produced in a pixel-perfect way (e.g. for Miniscope recordings), choose the FFV1 codec.
• If you need accurate data, but pixel-perfection is not important, and you want great compatibility with applications that process video, choose the VP9 codec.
• If your data analysis pipeline is modern, you can try using the AV1 codec, which offers much better compression ratios and is suitable for lossless and non-lossless encoding.

Which container format do I choose?

Always choose MKV - it is widely supported, modern, and can hold newer codecs and additional metadata that Syntalos adds to the video files. The AVI option exists only for compatibility with older data analysis pipelines.

Should I use video file slicing?

You generally don't need to slice the video recording into multiple parts, especially not if you use the MKV container format.

You may will want to enable this feature if your offline data analysis pipeline does not have optimal memory management and wants to load the entire video into memory. This can become an issue if the individual video is large.

You may also want to enable this feature if your recording is extremely long (multiple hours) to increase robustness in case of an error that could corrupt the video file.

Should I use the "Encode after run" option?

The "Encode after run" feature is very powerful if your computer can not encode the video data while the experiment is running. It will cause the video to be stored as raw data, and encoded into the selected video file after the run. This also frees a lot of system resources for the experiment run. Before anebling it though, you should consider the following things:

• Do you actually need to store all of this data? You may be able to live-encode your video stream if you scale down the size of the frames or crop them before storing them to disk. Just put a Video Transformer in front of the Video Recorder to preprocess the frames.
• Can you select a different codec that your system may handle better? Maybe selecting a different bitrate or quality setting will help performance? Play around!
• If you choose to encode after the run, keep in mind that during the run, data is store raw. Unencoded video data is extremely large! Make sure that before using this option, you not only have sufficient amunts of free disk space for the raw recording available, but also enough space to store the encoded video alongside the raw video. Syntalos will only delete the original raw video copy after it was encoded successfully. Keep this in mind! Disk space has to be available in the data target directory (where the experiment data will be stored).