Do the Elektron Octatrack AD/DA converters colour the sound?

I keep coming across forum posts or facebook questions from people worrying about the quality of the converters in the Octatrack. What usually follows tends to be quite heated discussions about whether the Octatrack affects the sound at all. I finally decided to do a bit of A/B testing so that I could have an informed opinion on the matter.

What I decided to do was to send various audio out of Cubase at 24bit/44kHz, via my RME Fireface UFX+ to my Ferrofish PULSE16mx digitally over MADI. Then record that audio back into Cubase via two parallell paths:

• One from the Ferrofish analog outs right back into the analog inputs.
• One from the Ferrofish analog outs through my Octatrack mk1 and then back into the Ferrofish analog inputs.

I then formed an opinion by comparing the recordings from the two paths. But MY opinion isn’t going to convince anyone one way or the other. So I then decided to make three 24bit 44kHz wav files for others to listen to.

Blind Test

Test1_Beat

Test2 Vocal

Test3 Synth

Each of the three audio files plays both versions and tells you which is which. Then they will be played 5 times in random order. Three will be one version, two will be the other. Can you tell the difference?

I’ve created the following email address: octatracktest@gmail.com in case people would like to report what they hear back to me. If some of you do, I’ll update this blog post with the data.

I’d like you to say in the email what you think is what for all three files. If you can’t tell, just guess. I’d also like you to state the following in the mail:

• When listening, did you consider your listening conditions poor, decent or very good?
• Could you tell them apart, and if so, did you consider the difference worth caring about?

EDIT: After some feedback, I reduced the level of the direct loop versions by 0.7dB so that both report same LUFS. No other processing has been done. I also re-upped the files to Google Drive instead of Dropbox, since Dropbox shows waveforms. Not detailed enough to tell the verions apart, but still possibly distracting.