This is a failed attempt at sampling a rock drumkit on 6 tracks. The channels are as follows:. 0: oh l1: oh r2: kick3: snare4: room l5: room r. I've captured this into ardour 5. 12 using 3 different audio interfaces:. Behringer umc202hd - overheads (dynamic mics)line 6 pod studio ux2 - kick and snare (condenser + dynamic)zoom h2 - room ambience (built-in xy condenser mics). This file is a 6-channel 24-bit flac file encoded using ffmpeg from the raw wav files exported from the original ardour session. There are several issues with this recording however:. 1. The tracks seem to drift, because the individual audio interface clocks were not in sync. The proper way to record multitrack audio is using a single multichannel audio interface - but i didn't have one. 2. There's either x-runs or some usb transfer issues creating small glitches and dropouts in various tracks her and there. Don't know why did this happen, as we've been tracking the real drummer's performance without these issues. Now - fixing these issues manually would be an insane amount of work, but i hope maybe someone has means to either solve them with programming a special tool, or know a tool that could fix these, and make this recorded session ready to be sliced as a drumkit for say - drumgizmo. There's some really good stuff in here - an i was able to cut and mix some really nice drum samples, that i've been using for years, but it's not ready to be fully sliced for maximum flixibility. The instrument was played by myself - it's a drumset by pearl (don't remember the details), owned by the drummer of a band i recorded this with. The band was called small hint - hence the drumkit name. We were recording an ep, and i used some free time left to capture this as well. The ep was never finished and we disbanded soon after. Regarding fixing the issues - here's what i think needs to be done:. 1. I think each hit would have to be automatically phase-aligned on all 6 channels, to correct for the drift. 2. I think it should be possible to automatically detect clicks by simply watching for a sudden change in amplitude between adjacent samples - marking bad areas and then using something like audacity's repair effect to interpolate the waveforms. I think the glitches have much steeper changes in amplitude than even the drum transients, so it should be possible to differentiate between those automatically. If you found a way to fix at least some of these problems - please let me know!. If you've made some "remixes" on freesound - i'd also love to know that. Apart from that - sample what you can out of this and make some sick drum tracks!.
著者: Unfa
