Q: Does Record's basic audio path sound different from Reason's? Is there any overall saturation or EQ curve being applied by the SSL mixer in Record that may make Record sound inherently different from Reason?
A: There is no "overall effect" like saturation or EQ curve being applied by any of the mixers in Reason or Record, the NNXT, or ReWire itself. Reason and Record's basic audio path is not just ruler flat, but pancake flat as well.
For this test, we want to see if a signal passing through Reason's 14:2 mixer measures any different than if it were passing through Record's SSL mixer. I'll also test ReWire since I'll need it for the Record parts of the test (using Record in demo mode - yea, I'm cheap!).
This software uses a "swept sine" approach rather than a pink noise approach for spectral analysis, which is the same approach used in convolution reverbs interestingly enough. I begin by using FuzzMeasure to create a test sweep .wav file. I'm looking for worst case here (to reveal any issues quickly), so I'm choosing the common sample rate of 44.1kHz and a bit depth of 16 bits.
- Next I move to Reason where I create an NNXT and load the .wav test sweep file I just created. In the NNXT I turn off the filter and keyboard pitch control, but leave the other settings at their defaults. Moving to Reason's sequencer, I "paint" a MIDI note on C3 for 3 bars at 120 BPM to match the length of my test sweep file. I set the loop points for a three bar loop, and then I "export" the audio at 44.1 kHz, 16 bit. Now I have my first test file!
- Next I launch Pro Tools and "ReWire" it to Reason. I open the same Reason song file and this time I use Pro Tools to create the audio file, bussing the ReWired signal to an audio track and recording in real time. I export this track (again 44.1 kHz, 16 bit). Now I have two test files.
- Now it's on to Record, which I "ReWire" to Pro Tools, and after loading the test file in a newly created NNXT, I create another audio file in PT and export it.
- For fun, I import the test file to an audio track in Record, and (via ReWire) use PT to create one last audio file, which I then (predictably) export.
Now I have four test files: a Reason "stand alone" export, a Reason "ReWire" export, a Record "ReWire "export, and the "imported" test sweep audio file in Record, which I used to create the four graphs below.
What to Look For:
Notice that the overall level in all graphs of the NNXT is almost -9 dB down from the original sample - this is the NNXT's way of handling multiple notes played on a keyboard without overloading. Since our sample is only using a single voice, it reveals what I'm guessing is the amount of static gain reduction the NNXT uses.
Also notice that the level divisions on these graphs are 1/10 of a decibel, and the the line is STILL totally flat!
If you superimpose all these graphs on top of each other, you will notice that there is no variation at all. Each system is identical under these circumstances. This means that the Mixers AND ReWire are totally clean, at least when no gain is involved, although one could argue that using an NNXT for these tests already involves gain change. And that means that the gain change we see is not altering the frequency response in any way which is good! By including ReWire in the tests, I hope to have cleared up any uncertainty to it's ability to cleanly pass digital audio thereby making it useful for testing purposes.
Below is the final graph in which I imported the test sweep directly into an audio track in Record and ReWired it into Pro Tools. Notice that the level is at 0 dB EXACTLY, meaning that an audio file imported into an audio track in Record will play exactly as it was recorded. Which is what one would expect. Kinda resembles a pancake, don't it?
- There is no "overall effect" like saturation or EQ curve being applied by any of the mixers in Reason or Record, the NNXT, or ReWire itself.
- Reason and Record's basic audio path is not just ruler flat, but pancake flat as well.
Calling this one "Busted".