i need help in muzys i want to use mixdown to file but don't know what the interpolation or resolution parameters are for, can anyone help please?also, when you mixdown to file, does this take into account any vst fx in the mixer channels?
ta
scott
Muzys mixdown
-
- KVRer
- 29 posts since 20 Aug, 2006 from Portsmouth, UK
i need help in muzys i want to use mixdown to file but don't know what the interpolation or resolution parameters are for, can anyone help please?also, when you mixdown to file, does this take into account any vst fx in the mixer channels?
ta
scott
Do not argue with what is; simply follow the progress of truth. -^-
-
- KVRAF
- 7414 posts since 8 Feb, 2003 from London, UK
Interpolation
Audio data in the computer is held as samples. In a low quality audio file, there might be 8192 samples per second. You might want to output at 48000 samples per second. Or you might be going from 48000 to 44100.
Easy way. Say we have a file with four samples per second but you want to output six samples per second. Either we run out of sample - very bad - or we have to stretch the samples out. Mmm, now a sample is just a number. You can't stretch a number... In one second, the four numbers might be 0.8, 0.6, 0.4 and 0.2. Remember, those are snapshots at four points in time within a second. We don't know what happens between them.
Interpolation really just means guessing. The settings in Muzys mean (AFAIR):
None: use the most recent value. We might get 0.8, 0.8, 0.6, 0.6, 0.4 and 0.2 as our six numbers for six samples per second.
Linear: take the average of the nearest values, taking into account how far from the target time we are. We might get something like 0.8, 0.633, 0.5, 0.367, 0.267 (I doubt I got that right!).
Cubic Hermite: more complex form of guessing that tries to fit a curve to more than two values at a time. I won't even try this one...
Often Cubic Hermite will be nearer what the original source sounded like. Almost always it will use more CPU time!
Resolution
This is how much information each sample stores. With only 8 bits, there are only 256 possible values. With 32bit floats... err, there's a lot of values.
Effects
Yes, they are.
Audio data in the computer is held as samples. In a low quality audio file, there might be 8192 samples per second. You might want to output at 48000 samples per second. Or you might be going from 48000 to 44100.
Easy way. Say we have a file with four samples per second but you want to output six samples per second. Either we run out of sample - very bad - or we have to stretch the samples out. Mmm, now a sample is just a number. You can't stretch a number... In one second, the four numbers might be 0.8, 0.6, 0.4 and 0.2. Remember, those are snapshots at four points in time within a second. We don't know what happens between them.
Interpolation really just means guessing. The settings in Muzys mean (AFAIR):
None: use the most recent value. We might get 0.8, 0.8, 0.6, 0.6, 0.4 and 0.2 as our six numbers for six samples per second.
Linear: take the average of the nearest values, taking into account how far from the target time we are. We might get something like 0.8, 0.633, 0.5, 0.367, 0.267 (I doubt I got that right!).
Cubic Hermite: more complex form of guessing that tries to fit a curve to more than two values at a time. I won't even try this one...
Often Cubic Hermite will be nearer what the original source sounded like. Almost always it will use more CPU time!
Resolution
This is how much information each sample stores. With only 8 bits, there are only 256 possible values. With 32bit floats... err, there's a lot of values.
Effects
Yes, they are.
-
- KVRer
- Topic Starter
- 29 posts since 20 Aug, 2006 from Portsmouth, UK
thanks for the info on interpolation and stuff, most grateful.
does it matter which settings are used to write this file to disc?
thank you oh knowledgeable one...
does it matter which settings are used to write this file to disc?
thank you oh knowledgeable one...
Do not argue with what is; simply follow the progress of truth. -^-
-
- KVRAF
- 7414 posts since 8 Feb, 2003 from London, UK
Depends on what you mean by "does it matter". If you don't care what the audio file you write sounds like, then no. Otherwise, yes. It may also matter if you want to use the file to master a CD (44100 samples/second) or DVD (48000 samples/second). For interpolation, use the one that sounds best.sadhaka wrote:does it matter which settings are used to write this file to disc?