Octatrack: 16bit or 24bit Mode?

Currently preparing content for use on the OT.

I don’t have an OT here though, so thought i would ask other users: Should I prepare the audio clips to be at 24bit quality?

Do you hear a difference in quality when using 24bit mode?

Is the benefit worthwhile the 50 percent extra memory usage for files?

Maybe a compromise approach to the OT modes could be 16 bit for preparation and performance, in 16bit mode … then if using the OT like a mastering unit, set to 24 bit and apply effects processing to a finished, imported 16bit stereo file, play it back and record it in 24bit … in 24bit mode.

i run 16 bit 44k, flex machines memory wont last long at 24 bit samples mate

1 Like

lol true and the flex machines are pretty much the apex of what the Octatrack is all about.

cool, glad to have made a slight yet fairly vital change to the content-in-preparation.

even though it is possible to go back and re-export 16bit versions of the audio loops, i would prefer it to be all ready-to-play. Shall keep to the 16bit format from here onwards.

To me, the quality difference between 16 and 24 bit, was the difference that made me decide to keep the OT. After trying it in 16bit, I was about ready to let it go.
Especially when you start to mess with Pitch and Rate, 24 sounds a lot better to my ears.

4 Likes

I do everything in 24-bit, but I don’t use Flex slots so I suppose that helps (I only use them for recording).

2 Likes

oh! cool, so there is a noticeable difference when changing Pitch … Hmmm, that is a favourable benefit.

To accomodate for the higher memory requirement, i guess the workflow with studio or performance approach to the Octatrack is:

load prepared loops to static slots … unlimited memory resources, works as stated in the manual, loops seamlessly without latency from card …

sample snippets as required from a larger file on a Static machine, into recording memory then load that to a Flex machine. for further rearrangement and one-shot sequencing

in this way the 80mb Flex Memory is best utilised, especially if in 24bit mode.

Actually there is a new Merlin manual pdf, i should read more of his fabulous thoughts on the subject of the OT.

Not really seeing how this makes sense - if we were talking about sample rate, then pitching down would indeed sound a lot cleaner because there are more samples per second to work with. But bit depth is related to noise floor, and allowing for more gain on quiet samples. I wonder, was this with samples that you copied in or recorded through the inputs? IIRC the OT has a weird gain structure where samples are recorded quiet then boosted up, so maybe 24bit had an impact there. But it shouldn’t be an issue if you’re copying over samples that are already normalised or close to it.

@previewlounge I recommend you do some A/B testing of your own to determine if 24bit is worth the RAM hit. Personally I choose the extra RAM because most of my stuff is gross and lofi anyway.

Also a lot of the stuff you’re doing probably won’t need to be on flex machines at all. Unless you’re jumping through slices with the fader/LFO or reversing stuff a lot, it can go static.

8 Likes

I hear you - I read up on the theory, after experiencing the difference, but found no logical explanation there. I’ve always recorded through the inputs, fwiw.

3 Likes

from wikipedia on bits…

"
A PCM signal is a sequence of digital audio samples containing the data providing the necessary information to reconstruct the original analog signal. Each sample represents the amplitude of the signal at a specific point in time, and the samples are uniformly spaced in time. The amplitude is the only information explicitly stored in the sample, and it is typically stored as either an integer or a floating point number, encoded as a binary number with a fixed number of digits: the sample’s bit depth.

"

PCM is pulse code modulation, the way digital audio is modelled in the digital environment.

so i guess in scenarios involving the information itself, such as time stretching, pitch changing, and rate, then there is more information involved in the process,

available to be used as finer grains in the algorithm’s adjustment of the data, or in the way the data is played back, how much and when…

leading to a perceivable difference in sound, most likely showing more harmonics and definition.

Interesting, there’s no point in denying your experience if you heard a difference, curious to test it myself!

2 Likes

What about audio files to compare, from people hearing it ?
Shouldn’t we compare with a sound card ?
I was born before CD quality, and my samples are 16 bit. :slight_smile:
If it sound really much better in 24 bit, I’d be very pleased to hear it. :wink:

I have no idea how the pitch-shift/timestretch algos on the OT work, so its not entirely impossible that increase in resolution, even in bitdepth, can yield a better end result. Something to test indeed?

OK so this also leads to the question: most sample packs are 24bit/44100 these days. I JUST got my OT last week’ish. Samples I plan on loading are 24 bit. I imagine I need to select 24bit playback, correct? The 16 bit pool of sample from elekton and loopmasters don’t particularly blow my mind, in terms of sonic fildelity. I’ve always been of the impression that 24 bit samples allow for greater headroom in the mix environment, and better offsets the inevitable “bit-loss” that happens over the streaming of multiple samples…then again, that was a concern in 2004 when I took a digital audio recording class at the local community college. What happens in the OT if you use 24 bit samples in a 16bit environment? It sounds like you should just select 24bit but, does it make a difference? In a purely subjective manner beyond the couple if replies above does 24bit sampling make a difference? Are you just making your own samples in the DAW of your choice, and exporting/dithering to 16bit/44.100?

There are no environments in the OT to speak of, it plays 16 and 24 bit files indiscriminately and at the same time if you tell it to. The only option you get is 16 or 24 bit for new files recorded into the machine. You can find this menu at project > control > memory.

In theory the only advantage 24bit gives you is a lower noise floor, so you can record quieter samples without noise issues. YMMV, do your own listening tests.

Thanks! I saw the 16 and 24 bit OT menu options while working my way thru the manual on another train of thought, and didn’t realize that is just for sampling input. I’m not at that point in the manual. Awesome!

Ohhh … cool! Okay so the 24 bit mode means nothing to the ability to play 16 bit or 24 bit, very cool, cheers for the illumination Anfim.

currently, even if the only advantage is a lower noise floor, i am actually going to keep rocking with the 24 bit preparations … i like making trippy effect preparations on audio files, then sampling just a tail of a sound here and there … the lower noise floor of 24 bit is helping me make abstract content explorations, remixing content a number of times in differing ways, without an accumulation of noise floor.

personally just using ableton to record and prepare the Machinedrum/Prophecy sounds/grooves … don’t have an Octatrack to test sound quality A/B comparison for the Pitch/Rate …

but if 24 bit does help ith lower noise floor, maybe less “noise floor” means better retiming, especially if re-timing to a slower tempo, as the Octatrack must “make-up” some audio content, so to speak, extending grains of audio to fit the new time.

an interesting a/b comparison would be the quality of re-pitching of an Octatrack recording at 24 bit and 16 bit of the same inputed audio, like a drum beat with a bass line.

record the input at 24 bit, then retime the captured loop and record how that sounds to a soundcard …

record the same input running 16 bit mode, assign the loop to a flex machine, retime and record the results

Here we go. Recorded a shitty drum loop and a shitty bassline on one track each in 16 bit, then duplicated the tracks, changed recorders to 24 bit and recorded again. In the recording I’m switching between the two pairs, while dropping the rate on all 4 tracks using the fader. The first one you hear is 16 bit, then every time you hear a bit of silence it goes to 24, then back and forth. At the end I also set timestretch to auto for all 4 tracks, and drop the bpm down to 30. There’s a tiny bit of mic noise because I forgot I had a 57 plugged into the mixer (doh) but it shouldn’t make a difference. Everything went through a mackie 1202 then a focusrite saffire pro 24. Exported as a 24bit wav, then normalised.

https://drive.google.com/file/d/0B_dkIH5xlcuPQUExQ0d0UWpTUU0/view?usp=sharing

3 Likes

The OT does have two settings for this. One for flex which when set to 24 will play samples at 24 or 16 bit depending on the sample, and when set to 16 will convert all samples to 16 when loading. The other setting affects whether recordings will be made at 16 or 24 bit…
Higher bit depth not only increases dynamic range but does play a role in digital processing as well, google around and you’ll find scientific articles about this…

9 Likes

Just found this…hard to keep track which is which but if I can trust my ears, there‘s definitely a difference…one version (I take it that‘s the 24bit) is less „grainy“, sounds like it has fuller bottom end and seems just a little cleaner, more defined to me…oh my.

EDIT: Funny enough, tried to replicate the experiment on my Octatrack Mk2 and for the life of me couldn‘t hear any difference lol.

1 Like