How to understand vgmtrans loop points? by Dark_Ansem at 4:53 AM EDT on July 2, 2020
So, I'm looking at some PS1/2 audio files, and I can see that there are some loop points highlighted by VGM but, for the life of me, I have no idea how to understand them. Any help, please?
Most formats that are streamed (opposite of sequenced), use loop points based on samples instead of specific points in time. This makes it more accurate in terms of seamless looping.
I give you an example: You have a monophonic (single-channel) PCM wave file, that is exactly 10 seconds long, and was recorded with a sampling rate of 22,050 Hz. This can be a playing instrument for example. The total number of samples is playtime (in seconds) multiplied by the sampling rate (in Hz), which gives you 220,500 samples for this example. Decimals in seconds would give you a more arbitrary total sample count. Let's assume, that you need to loop a specific part of the sample, because it's a trumpet or a piano for example, and you need the loop, because otherwise it would be a one-shot on playback. You now have two options to set the start and end positions of the loop: 1) Using a time-based point in HH:MM:SS.ssss format or 2) the sample number. The accuracy of both options changes with the total length and sampling rate. Time-based looping is also dependent on the accuracy of the playback engine. If it only can understand parts of seconds of up to 1 or 2 decimals, then it isn't nearly as accurate as when sample-based or having more decimals for a seamless loop.