-
It'd be really cool if the program had another form of "auto loop" under "Fast Editing" that automatically found the top 5 closest matching points in the waveform. For those samples that are particularly difficult, maybe you could choose to search for the 10 best loops, instead.
This would be a godsend for samples like this one I'm trying to find a good loop for that's 55,072 samples long at 32kHz. It likes adding this kind of bassy sound when it loops in the wrong spots, even if there's no warning triangle. I know there's a good loop in here somewhere but I won't find it easily.
- First, you'd pick "Find Best Loop" under Fast Editing. It runs a brief numbers check on the sample.
- Then, a window appears. At the top, to set the allowed number of contenders, it has one of those "type box that also has up/down arrows" so you can put in a # yourself, but nothing outside the limits.
- Below that, buttons or a drop-down asking if you want to search for loops in a short, medium, or long form (may also include very short or very long, based on how long the sample is). Each has a ballpark estimate for the resulting loop sizes, based on both the Sample Rate and Sample Size of your waveform.
- If the program knows my sample is 55,072 samples long at 32kHz, it then might give the options as "Short (<2000 samples) Medium (2000-8000 samples) Long (8k-12k samples) Very Long (>12k samples)." (These are just dumb numbers, IDK if those sound realistic or not.)
- Ideally, these limits will mitigate how much time the program will spend calculating. I figure, if it knows the sample length as well as the sample rate, it'd know the sizes to set top & bottom limits at, so that it doesn't waste time comparing innumerable tiny loops or try to make unfeasibly large loops.
- As a contingency plan, though, if the system notices "loops to compare" exceeds a certain amount (like if you accidentally chose the wrong sample window size), it can stop the process, notify the user of how many potential loops it has to sift through, and give the option to "continue (may be slow)" or "re-configure settings." Implementing the failsafe in this way would account both for human error and computers sometimes worrying over nothing.
- After this, a window for cycling through the resulting 3/5/10 best loops would appear. This window wouldn't disable the sample playback controls, so you could see in real-time where each of the "Found loops" place their start & end points. Since playback can keep going even when you change loop points, you could see what they sound like in real time as well.
- The window would also have the option to "search again," which would take you back to the "select sample size for finding loops" window.
This wouldn't replace the existing auto loop, as that has its place too... I assume. I do think that macro could use a tune-up, though. Something allowing for more user input, like a pop-up for fine-tuning the crossfade before it does anything. As is, it's kind of a crapshoot as to whether you'll like the result, and it often makes the fact that it's looping more obvious, ironically. -
that would be cool indeed, but i dont think you and i live long enough to see a machine making better loops than a pair of human hands, instructed by a human brain. for very difficult loops it is easier to edit the sample in a wave editor that has the feature CrossFadePasting (NOT CrossFadeLoop, that again is a machine trying to make a loop).
you copy a piece from the start and paste it at the end of the sample. try different overlaps, different places.
in a wave editor you can locally manipulate the volume too, many loops dont fit because of different volume levels.
regards bottrop
Sign in or register to take part in discussions.
Polyphone needs you!
Polyphone is free but there are costs associated with its website and development. A small donation will help a lot.
Donate
Learn the basics
Try a tutorial
Scroll to
top
top