Great test. Number 1, miles ahead.
The ot mud slide gain, ha, is pretty easy to spot.
I didn’t know the dt sampled or played back at 48. Thats interesting.
How significant are the tech specs on the Octa, MK1 vs MK2? Do they translate to any real-world difference?
(respectively:)
Unbalanced audio inputs:
Input level: +8 dBu maximum
Audio input impedance: 9 kΩ
SNR inputs: 99 dBFS (20-20.000 Hz)
Balanced audio inputs
Input level: +17 dBu peak
Audio input impedance: 21 kΩ
Digital S/N ratio: 106 dB (20–20,000 Hz)
1 Like
I have to completely agree with @GirTheRobot on this one. 16bit output is lesser in quality compared to 24bit only because of noise floor/headroom considerations. You shouldn’t be able to tell the difference in an a-b blind test of well recorded sources, and it should certainly not be the reason the octatrack sounds obviously different in this test. If anything, it might be the gain staging setup on the ot, or some differences which arise out of converting stereo to mono, or just simply the difference in sound between both converters.
And about all the myths surrounding sound and audio quality: yes theres a lot of them, and yes, companies purposely exploit the fact that these myths exist to continually sell you new ‘better’ , or more ‘analog sounding’ stuff. The funny thing about this is that people now seem to be attracted by old gear, because of the ‘bad’ converters in them: a sp1200 is now a priceless treasure because of its ‘warm’ sound caused by its 12 bit (which is audible compared to 16 bit btw) converters. Dont be fooled into believing you actually need any fancy gear to make good music.
And to those that say: i trust my ears more than i trust science: you should! Everyone has some favorite pieces of gear which they think sound magical, and this is more important than what specs they have. Just be aware of the fact that paychoacoustics play a large role in this: the way a piece of gear looks or how much it cost might effect your perception of its sound more than the actual sound of the unit.
7 Likes
The Model Samples is even crispier. 
1 Like
I felt like maybe some dynamic compression was bringing up the ambience on the DT.
Would have been good to have heard the original wavs through neither machine. Eliminate the ‘faithful’ vs ‘enhanced’ question
Don’t mind me just out here confirming my bias
http://www.aes.org/e-lib/browse.cfm?elib=18296
“Overall, there was a small but statistically significant ability to discrimi- nate between standard quality audio (44.1 or 48 kHz, 16 bit) and high resolution audio (beyond standard quality). When subjects were trained, the ability to discriminate was far more significant.”
1 Like
Sample rate is one thing, bit depth is another.
1 Like
Yes. But the “argument against 192 is the same argument against 24”. I’m only contributing to the conversation by bringing attention to a knowledged study showing high res files can be discerned.
The study does not say anything on bit rate either way, only that it wasn’t tested.
Because it’s 25 bit. 
1 Like
Alrightalrightalright. While it pains me to admit it (mainly cause I have to readjust my head), I concede that there might be no difference between recording 16bit vs 24bit on the OT. I recorded the same source twice without changing any settings, loaded the files to my DAW and had a listen. 24 might be closer to the source material but not enough for me to say absolutely either way.
Def not a gamechanger and def not worth the extra RAM. I’m glad to have done this for the latter alone.
I really am curious about the test above now. I’d like to hear the original file, whether timestretch was on, ect.
However I still contend there’s a difference in highres audio versus standard. Ive done enough tests to be sure on that. However this means nothing to audio over Apple earbuds nor to people who don’t care enough (keep in mind audiophiles are often classical / jazz / classic rock aficionados…mics and instruments and rooms; nuances and subtleties matter to them). I’m not convinced the worth of a thousand pound strand of speaker cable. But a highres file vs a standard file over a good system is def a different experience.
K I’m out. Way too far down an empty rabbit hole.
4 Likes
High bit depth decreases rounding errors when mathematically computing signal processing…
Modern DAWs actually mix and process in 32 bit float, 64bit float is starting to show up now.
There are audio interfaces that output 32bit D/A…
When all sorts of crazy math computations are going on you need higher resolution to not introduce rounding errors, in the end you can drop it down to lower resolution after the math is done…
2 Likes
Before someone attacks me, even Digitone knows this… 
It’s a 24/48 device but it uses 32/96 for internal computation for higher sound quality…
2 Likes
There is a good video on YouTube where the guy bounces a complex acoustic recording (from a daw) to 24bit, 8bit dithered and 8bit not dithered.
The 24bit and 8bit dithered perfectly cancel each other out, leaving only the added dithering noise. All of the music is retained at 8bit but with some added noise. It makes perfect sense when you really think about it but also some how seems counterintuitive.
Obviously the 8bit undithered sounds much worse, as the detail is lost to rounding.
2 Likes
Having a higher bit depth when processing is obviously handy though but that isn’t what is happening in the test.
1 Like
We all have our anecdotes, and our own needs and use cases.
I have some compressed stems I use in OT for my live set. Did a lot of renderings and tests to see if 24 bit was necessary for my needs.
4 stems per track. They’re rendered at -6dB and playing back through OT (4 static machines plus 1 neighbor track each = 4 neighbor tracks, FX for each stem are eq, reverb, delay)
The difference between 16 bit stems and 24 bit stems was not noticeable on my OT MKII in this case.
That said, I’m surprised more folks in this thread aren’t noting the high frequency “bump”, “enhancement”, whatever you want to call it, that has been observed on the Digitakt’s output since its release.
I find the OT’s output to be rather flat (“even”, not “unexciting”), personally.
2 Likes
It’s 8 better.
This is my dream when I have kids.
1 Like
Would it be out of the question to do a comparison of a sample, straight up vs. run through the OTMK2’s inputs? You’re one guy who’d probably get it right
1 Like
yea seriously someone do it QUICK.
I’m on holidays n it’s driving me nuts 
1 Like