I bought the elektron house and techno sample pack.
they both at 16 bit, sounds like garbage.
i try to change to 24 bit, breaks most of the samples, not sure why, most samples (not all) won’t play?
can fix this?
appreciate.
also, is there some way to move samples in and out of this project within octa? is there a better video explaining how to copy/paste and/or move samples or patterns in the box?
i like SOME of their samples, but i want to start replacing ones i don’t like with my own, and slowly turn this into “my” sound, but i confuse trying to move samples around…
FYI Bit depth won’t change sound quality, bit depth is the available amount of possible levels.
whoever make the house set is funny person, use phazer instead of looper/studder delay on xfader, i mean, it’s not BAD, but lol, i change that right away…
you’d think that, but i ALSO have external gear routed thru octa, and when it looping more than one sample, it OBVIOUS that 16 bit sounds garbage, mate. i have a bunch of 16 bit samples from ableton plus my gear in my own set, and i experiment just for fun, because all people talking about sound quality in other thread. i dunno jack or shit about programming back end, but go ahead and make 2 sets 24/16 bit use same samples, listen. i could hear difference, high end transients start to get real “choppy” at 16 bit.
What you mean you are running out of headroom?
i mean, i’m constantly pushing the headroom on octa, because these LED aren’t the greatest sound meter are they? 
but trust me, you can tell difference between 24 and 16 bit set when you’re looping a bunch of samples, or at least I can. with the same samples.
but yeah, it gets pretty obvious with headroom from my TR-8S to Octa, distortion at 16 bit.
Did you try lowering the levels a bit on each sample, you can add the gain after on the master.
that’s not a bad idea. but i want all my set at 24 bit, pretty sure that setting also affects the d/a from external gear, and like i say, better clarity with external is obvious to at least me.
i just wandered why it’s destructive on the set to change from 16 to 24 bit internally. shouldn’t matter if samples all 16 bit, they just play back at that, but then my external gear is also getting I/O at 24 bit and not 16. this should be a “system” setting that doesn’t change when you load a new set…
If you set it to 24bit 16bit samples will still be 16bit, but any 24bit samples you load or record will be 24bit.
i’m aware.
the issue i’m having is that when i take the set that i bought from Elektron and change set from 16 to 24 bit, most of the samples aren’t playing back, and I don’t know why. It’s odd behavior and it should be fine, this should be easy internal thing for OT to do, but now most of the samples in the set won’t play back unless i go manually put trigs back in.
I have never bought any of the soundpacks, might be worth checking with support and see what they say?
Maybe try changing the project from 16 to 24 and saving, then converting the samples from 16/44 to 24/44 on a computer and overwriting the originals on your card with the 24 bit conversions. Reload the saved 24 bit version of the project and see if it works with the new 24 bit versions of the samples.
converting to 24 bit can actually theoretically make the 16 bit samples sound better in use. Even though the raw samples won’t be any different (they’re just 16 bit data with the least significant 8 bits of the 24 bit words filled with zeroes), the Octatrack is pretty much definitely working with fixed point math so for every 6dBFS of headroom your audio is losing one bit of resolution. So if you take a 16 bit sample that peaks at 0dBFS and lower its level 12dB so that it’s now peaking at -12dBFS (which is what the Octatrack does by default when sampling) you’re only actually getting 14 bits of resolution at 16 bit (which is starting to get low enough where it gets pretty noticeable) but at 24 bit you’ll still have 22 bits of resolution, which is more than enough to sound good. Because of the gain structure of the Octatrack, it’s pretty unlikely that audio is going to be in the sweet spot for 16 bit - consistently peaking above -6dBFS but not clipping, and with live sampling there’s the permanent in 12db pad so 14 bit is the highest possible resolution you can get when sampling in 16 bit mode. In most cases you’ll probably be getting more like 12 or 13 bit, if not less. If you aren’t sampling from the Octatrac’s inputs and only use them for thru machines, with your samples being carefully prepared on a computer before hand and careful attention to gainstaging on your live inputs, you could probably get a good sound at 16 bit but it would take some work. And even if you do that, it’s extremely unlikely that the Octatrack dithers audio after every calculation made to it, which would be necessary to avoid accumulating a lot of truncation distortion when dealing with fixed point 16 bit audio. Change the pan? It needs to be dithered. Adjust the volume? It needs to be dithered. Change the envelope? Needs to be dithered. At 24 bit the truncation distortion is so low that it’s negligible, at least in the context of a sampler.
At 24 bit you can pretty much ignore all of that and sound clean.
On the other hand, working in 16 bit gives the Octatrack more of a SOUND, and plenty of good music has been made with stuff that’s loaded with truncation distortion and compromised bitrates. For some stuff it might not matter; for some stuff it might sound better than 24 bit. For me, I never use 16 bit because I don’t want the OT to color my sound that much, if I want a low bitrate sound when I’m using the Octatrack I’ll get that upstream in my signal path, sample things that already have the sound I want, or use the lo-fi effect. But running the OT at 16 bit to get a little crunch is totally valid.
That seems like the famous bug. Whenever you change that setting, you have to go within the AED and preview them (FN+yes) so they can play again. There was another trick using the Card Tool IIRC but i forgot what it was exactly. It avoided to have to preview each slot one by one.
wow.
THANK YOU!!!
i appreciate someone explaining the technical details behind what I’m hearing.
i will forget most of this shortly, but at least i know where to look if i want to remind myself that there’s a reason i want everything 24.
is 24bit “default” “resolution” of the hardware D/A converters? so OT has to downsample in software? how does that work? not sure i using all the appropriate words for what OT is doing? like they could never add 32 as an option, because the hardware converters are limited, but they COULD add 12 bit as an option, if they wanted to throw that coding in, yes?
i thought i might be able to avoid that somehow, it’s odd how SOME of the samples still work, LoL…
It’s probably not the perfect description, but I think it gets the basic idea across.
EDIT: the flip side is that lower word lenths sound a lot better than most people think. The difference between 12 bit and even 8 bit isn’t as big as we tend to think, provided it’s recorded and dithered properly.
Well Bit depth does affect sound quality, a 4 Bit sample is not going to have the same quality as a 16 Bit sample.
Having said that, the audible difference between a 16 Bit sample and a 24 Bit sample is probably fairly negligible the main advantages being a lower noise floor and more room before distortion occurs (headroom).
I guess the automatic bit depth conversion for samples in RAM (flex machines) happens when they get loaded from the CF card. So the fastest way to get everything re-initialized correctly after changing the bit depth is saving the project and reloading it again.