14 Mar

Does Record's Default Mixer alter the signal?

Written by Wendy Dunham (wendyou) and Giles Reaves (selig).

Q: Does Record's mixer alter the signal frequency in any way when using the default audio path? (i.e. does the mixer change or "color" the sound, making it better or worse?)

A: No. Tests have demonstrated that the basic audio path of Record is "ruler flat". Input equaled output with absolutely no variation.

Background: When RECORD was released, several users reported that their mixes sounded better, warmer or different than REASON, and they attributed this to the new Mixer. This test seeks to answer that question without bias using signal frequency analysis comparisons between an Audio Source test file and two exports – one through the Mixer and one through the Hardware Interface only. If there is no detectable difference between input and output, then the default Mixer is not altering the signal.

The Test:

A frequency-rich test file consisting of computer-generated Shepard Tones (Shepard-Risset glissando) was loaded into an NN-XT Advanced Sampler. This audio test file sweeps through the entire audio spectrum with 7 simultaneous fundamentals and their harmonics, making it an ideal and broad tonal test signal.

Download the original source Test File IMPORT (AIFF) for your own testing via an NN-XT Advanced Sampler:

ShepardTonesUpAIF.zip877.32 KB

Two tests were performed:

Audio signal path passed through the Record Mixer with no adjustments.
Audio signal path bypassed the Mixer and went straight to the Hardware Interface.

Each test output was exported to an AIFF file for direct comparison to the original audio file via spectrographic analysis.

iZotope RX was used for this examination, as it's multiresolution spectrogram shows more detail for improved visual analysis.

Download our 2 Test Result EXPORTS (AIFF) for your own analysis and comparison:
1) Audio through Record Mixer and 2) Audio through Record Hardware Interface only:

TestExports.zip2.27 MB

The Result:

There is absolutely no difference between the input versus output test signal, from 0 to over 20 KHz, even for weak harmonic levels below -110 dB. See the following animation or download the more detailed originals, below:

Spectrogram animation comparing Original Source versus Mixer and Hardware Interface exports.
No difference!

Download the 3 full-sized, detailed SPECTROGRAMS (Source, Mixer, and Hardware Interface):

Spectrogram_Results.zip2.58 MB

Inverted Phase Summation Test:

Selig confirmed my results, above, using an industry-accepted null test, below. He exported the Mixer/Hardware Interface output and summed it with the original test signal. The result was a perfect null, meaning that the two signals canceled each other out because they were perfectly identical: