Same signal, one on span, one on bogaudio analyzer:
It seems span is around -38db, while bogaudio analyzer around -24. Why?
Try a higher quality on the analyzer.
very strange: 3 analyzers, 3 different peak levels
… but at the end, all that matters is the sound to my ears 
Db itself is always a relative number. You need to specify what 0 means to make it absolute. You will find the same thing withVU meters. This is not surprising at all. oh, Sry, didn’t intend to respond to @Ahornberg - just responding to post.
3 Likes
ah, I had forgotten there is a standard. and, btw, ± 10V is not 20v, it’s 10. for peak that is.
Huh?
I guess the confusion stems from the definition of sine wave amplitude. Some measure the displacement from 0 (I guess what your are calling peak amplitude), whereas others measure peak to peak. I prefer peak to peak because it implicitly forces you to compensate for DC offset. If measuring displacement from 0, a naïve person might mistakenly assign a 10V peak amplitude for a unipolar 0-10 signal, and 5V peak amplitude for bipolar +/- 5. But they have the same peak amplitude of 5V - the unipolar 0 point is 5V.
Measuring peak to peak seems more obvious / intuitive from a language perspective. I immediately know I must compute (max - min) to get the the correct peak to peak amplitude of 10V.
The phrase “peak amplitude” sounds too much like “peak to peak”, which I think leads to a lot of confusion.
2 Likes
oh, sure enough, you are correct. I did not know that! actually, I think my statement was technically correct, +/- 10 does seem to be called “10v peak”. But none the less, I did not know the definition of “peak to peak”.
Not really sure about this 
dB its relative to common value in Rack. As explained by Vortico here,
10V is 0db and 5V is -6.02 db.
In fact, if I feed a basic Sine (2khz) to scope and Span, it works as expected:
Why on bogaudio analyzer is 0db?
On MixMaster mix, there are even both
On single channel 0db, on Master -6db… 
By design :-). We figured 10V=0dB (which is the standard) would be good for the master (when soft-clipping is off technically, but almost the same when it’s on), but that 5V=0dB would be more appropriate for the tracks.
3 Likes
So, VCV proposes a VCV standard, not everyone follows it. Either intentionally to make a better product, or unintentionally because it doesn’t matter to most people. Is that really so difficult to believe?
Yes, I just find it very weird (even more within the same plugin) But I can see…
Thanks
1 Like
As squinky says, dB is a relative rather than absolute value - it depends on what voltage the developer decides 0dB is.
Oscillators (mostly) output +/- 5V by default. It made sense to us to make that 0db on the tracks on MM. It also means tracks have some headroom (ie room for user error) built in - a track on MM won’t actually clip until it hits +6dB. Most users who don’t fully understand all this level stuff will probably presume under 0dB is ‘good’ and over 0dB is ‘bad’.
The Audio module on the other hand expects +/- 10V. It therefore made sense for the Master on MM which connects to that to also be calibrated such that 10V = 0dB. If you go above 0dB on the master on MM, your audio will clip - you don’t get any extra headroom there (other than the soft clipping). But that’s one stereo bus you need to keep a careful eye on rather than 16 different tracks…
If MM tracks were calibrated such that 10V = 0db (rather than 5V = 0dB as they are) everyone would need to learn to mix each track to -6dB max and we would probably be responsible for thousands of clipped/ruined audio recordings…
All this is explained in the manual btw. As Marc said - it was a considered design choice at the time and 5 years later, I’d make the same choice again today.
We were not the first mixer to take this approach btw - iirc the Msc Hack mixer that came before MM did the same.
7 Likes
Absolutely! Given that that is actually the Rack voltage standards it’s the only thing that makes sense, IMHO (but always open to learn).
Which is what 90% of the producers does (i.e. mix tracks at -6db), more or less
Its a common choice to mix below 0db on each track, due to headroom.
That’s why basically (I believe) all DAW I’ve used have the same reference in both tracks/sends and master.
But I see your point
Sorry for bump this topic, but I don’t get this on bogaudio Spectrum:
Why feeding a sine of 5v peak-to-peak (yellow in Scope, Purple in Analyzer) its correctly 0db in Analyzer, while a noise higher than 10v peak-to-peak is -24 db on Analyzer? Shouldn’t be at least “higher”?
EDIT: got it while writing it!! I’m tired! Hehe of course the sum is higher than 5DB, but every single component is lower in DB. Thanks anyway.
3 Likes