Print

[dysamoria]

i'm very frustrated. please excuse my messy structure in this message.

i'm used to working with samplers such as Akai S6000, Kontakt 3 and such. i'm familiar with the need to assign an accurate root note (the pitch of a recording) to the sample in it's file attributes/properties in order to place it accurately in a keymap. i've had no trouble with those other samplers, but i've been messing with MilkyTracker for hours, maybe over a day, and i cannot seem to "get it." it doesn't behave as i expect. i'm a tracker user from ages ago, so i think i'm familiar with tracker stuff. i never had trouble importing Ultrasound patches, tho i never edited them extensively, and i never built any multisampled instruments from scratch in a tracker (just on my Akai).

is a tracker, or just MilkyTracker, a very different beast from a traditional sampler?

 every time i set the root note on the sample edit page, it changes it on ALL samples. i have no idea what the extra note setting in the instrument editor is. it isn't ever the same as the root note in the sample editor page. neither note reference is the same as what is in the sample's file itself (seen in Awave Studio and Sound Forge).

 i get the impression that i'm just missing something really dumb and obvious like trackers being a completely different beast from the Akai sampler i'm used to. is the sample rate of the file a complication? i honestly never dealt with anything other than 44.1KHz samples in a traditional sampler and all the tracker stuff i did with varying recorded sample rates in the sample files never mattered much until now.

 the samples i'm working with right now are non-standard format. 12-bit at 30KHz from a Roland S-xx sound library. i'm converting instruments from "OUT" files (disk images from those hardware samplers) to FastTracker XI instrument files with Awave Studio 9.2. samples are getting converted to 16-bit to maintain the quality of the 12-bit samples (over 8-bit), but the sampling rate is staying the same. some of the exported/converted XI files load into MilkyTracker seemingly fine while others are insane (i assume Awave Studio is to blame for poorly converting them). i have been totally unable to manually create a multisampled instrument in Milky in cases where the conversion fails.

 i even resampled the individual samples to 44.1KHz in Sound Forge to make sure they're "standard"... no luck.

where is the root note set per sample in Milky? does Milky READ the root note from an XI, WAV or IFF file when loading/importing them? how does the root note apply to instrument creation? why can't i assign the correct base/root note to individual samples separately from each other (changing one changes all)? they import with anything BUT the correct root note. there's no "sane" state.

i've read all the docs i can find. that's to say, there's not much out there. FastTracker's manual and Renoise's manual both refer to the "feature" of multisampled instruments, but they don't explain/describe how/why it works. nothing about whether or not i can just treat these things like i did my Akai,

sorry for the disjointed nature of this posting. my language skills used to be excellent but i'm all scattered because of a medication i'm eliminating from my life (a benzodiazepine, which messes with cognitive abilities). i have PTSD and autism on top of that (or under it). i really have a hard time communicating lately but i needed to ask about this before pulling any more hair out. sorry. please excuse my tone/ramblyness...

as an aside... i'm new here. being on disability, i have lost all the equipment i used to have and am using an old PIII 500MHz computer for Renoise (can't afford to buy anything to replace what i lost since i have no money). since Renoise isn't meant to be a replacement for Fast Tracker, and my task is to make old Roland sample library content available to myself/others in the tracker community (classic old sounds in better-than-8-bit quality!), i followed user recommendations to check out Milky Tracker. Milky is obviously incomplete, but it is quite useful. it's the closest i can get to Fast Tracker on this limited computer. Milky is actually quite excellent in many ways and i'm getting used to it. i was a Fast Tracker user for years, way back. i'm also a MOD Plug Tracker user, but i've grown away from it for anything but utility. i've read as much as i can about Milky and searched the forum/website. there's a lot of stuff in Milky that is impossible to know about if a person isn't the developer, hasn't followed development from day one, hasn't been interacting with developers, etc... so... now i'm trying communication. otherwise, i'm just giving up because i reached my personal limit on trying to grok it alone.

thanks for reading.

[DasKreestof]

>is a tracker, or just MilkyTracker, a very different beast from a traditional sampler?

It’s not too different from a primitive sampler.

Some of the frustration you’re experiencing may have to do with something you’re not used to, which is working with a sound engine that is not a native multiple of 11khz, 22.05kz, 44.1, or 48 etc.  Milkytrackers native engine which is legacy based, works with a default 8363Hz. If you import a wav file sampled at 44.1khz, Milkytracker will automatically tune the relative note for you so that it will play back at the correct speed which means the relative note is F6 fine tune -28.

Your 30khz samples might need to have their root notes adjusted to play back at the correct frequency depending on how Awave is writing out the wav headers.  Awave may be adjusting root notes/frequencies during conversion. You may want to test that.

The other big difference from the samplers you’re used to, is that multi-sample instruments in Milkytracker cannot overlap layers. To layer, you must multiple tracks.
Make sure that your converted multisample instruments do not overlap layers or exceed xi specs.

I believe Milky adjusts the root note based on sample frequency in the header of a wav, so that a middle c will play back the sample at the rate indicated in the header. I could be totally wrong.  I don’t know how it interprets xi’s. I assume changes to Xi’s would be coming from awave instead of MT.

I’m not sure about the complications of multisampled instruments and tuning, but perhaps the above info will help you with trial and error to determine the behavior of MT if someone else doesn’t pipe in first.

Edited: replaced F#5 Fine tune +36 with F6 fine tune -28. I just did an import and that's what I got. I'm not sure where I got the F#5 Fine tune +36 from

[dysamoria]

this is useful info. thank you VERY much! 

i have had more success with importing WAV files by themselves, but that, obviously, loses the key mapping. as for overlapping zones, i'm pretty sure the Roland disk set didn't support overlapping key maps itself, but i could be wrong. i'm in the process of interacting with the Awave developer to see if i can track down specific variations between the source and the output so he can improve the conversion output.

thanks again, DasKreestof!!

[DasKreestof]

I just edited my previous reply. I replaced F#5 Fine tune +36. with F6 fine tune -28.
I just did an import and that's what I got.
Sorry for any confusion. I'm not sure where I got the F#5 ft36 tuning from.