BTW if your getting a clutter on rendering out from detection you should transpose your source material up 1 or even 2 octaves. The lower you get in the frequency spectrum the more packed together notes/spacing become this can confuse (most if not all) modern detection algos ... So the higher the octave, the more spacing between notes = better detection.. Then just transpose back down (if you want to) after processing....
HTH
L
Audio to Midi Vst Plugin ...
-
- KVRAF
- 5629 posts since 22 Sep, 2005
Important!!
You want to try to get material with little or no pitch modulation and try to have a limited amount of velocity tremelo.. If you run into problems with detection you can also use a freeze buffer (like spectral monkeyage) to lengthen the parts so you can at least get the notes....
L
You want to try to get material with little or no pitch modulation and try to have a limited amount of velocity tremelo.. If you run into problems with detection you can also use a freeze buffer (like spectral monkeyage) to lengthen the parts so you can at least get the notes....
L
-
- KVRian
- 1030 posts since 15 Feb, 2005
i tried virtually everything...
Intelliscore
Midifier
Musciendoz
Digital Ear
Rephrase
Audiotune
Widi 3.0
Akoff Music Composer
I would link to the other threads i described my experiences but the search function is down....basically none of them work well for voice or singing ur synths or samplers,...realtime or offline....the latest melodyne works the best....would be a nice dream though.....fretted audio's audio driven fretted synths track very well to voice actually, but ur limited to their synths
Intelliscore
Midifier
Musciendoz
Digital Ear
Rephrase
Audiotune
Widi 3.0
Akoff Music Composer
I would link to the other threads i described my experiences but the search function is down....basically none of them work well for voice or singing ur synths or samplers,...realtime or offline....the latest melodyne works the best....would be a nice dream though.....fretted audio's audio driven fretted synths track very well to voice actually, but ur limited to their synths
-
- KVRAF
- 5629 posts since 22 Sep, 2005
It's really a matter of isolating the source material as much as you can. Honestly I always (even with Melodyne) normalize the target as much as possible. I think theres a PDF at the Celemony site that explains how to prepare your audio for detection. I think theres a blurb in the Manual also....
L
L