Soft synths are great for quick and easy loading and control of a sound set.
 
One of the few downsides to them is how they affect output volume.
 
1. Load kickdrum.wav into audio track 1.
2. Insert Dimension, Rapture, and Drop Zone.
3. Load same kickdrum.wav into each one and set "element" outputs to 100%.
4. Insert MIDI track for soft synths.
5. Leave all track settings at default values.
6. Solo each track and play noting peak outputs.
 
Results:
 
Raw kickdrum.wav: -2.5 peak signal.
Drop Zone: -3.7 peak signal.
Dimension: -10.5 peak signal.
Rapture: -11.5 peak signal (master @ default 75%).
Rapture: -6.6 peak signal (master @ 100%).
 
Rapture is the only one with a master volume giving you more output control.
 
Conclusion: Each soft synth diminishes original sample output volume by a different amount. It seems the more sophisticated the synth, the more it cuts the output. Rapture cut output by as much as 9dBs, quite significant. Not knowing this can dramatically affect the final mix.
 
One way to increase soft synth output is to load the same sample into 2 or more "elements", though this will add to CPU load.  
 
Steve

soens
Rapture cut output by as much as 9dBs, quite significant. Not knowing this can dramatically affect the final mix.

Testing Rapture with the level set at 75% is not an accurate test, as the master level is an attenuator. I assume you did not attenuate the audio track by 75% as well. I also think you might have overlooked triggering the sample with velocity = 127?
 
I inserted a handclap sample normalized to 0.0 dB into an audio track and as an element in Rapture. With the Element and Master levels turned up to 100% (no attenuation) and a MIDI note velocity of 127, the Rapture output was -1.3 dB, not -9 dB. 
 
With DropZone the drop with output up full and velocity = 127 was -1.2 dB.
 
With Dimension Pro the drop with output up full and velocity = 127 was -4.3 dB.
 
I'm really not sure this matters much anyway, especially if the difference is around -1dB. Few patches consist solely of a single waveform. As soon as you add in more elements, effects, filtering, or whatever, you'll need to change the level to gain-stage properly.
 

I don't know if that's a bad thing, all of the peak levels you mentioned are still too hot to bounce (-12dbfs peak reference). I always have to turn my soft synths way down to get reasonable levels from them anyway. Remember digital ain't analog, 0db on an analog board is still around 24 db away from clipping. If you're running much hotter than -12 dbfs channel levels in the box you're going to have a hard time getting a good sounding mix.
 

TomHelvey
I don't know if that's a bad thing, all of the peak levels you mentioned are still too hot to bounce (-12dbfs peak reference). I always have to turn my soft synths way down to get reasonable levels from them anyway. Remember digital ain't analog, 0db on an analog board is still around 24 db away from clipping. If you're running much hotter than -12 dbfs channel levels in the box you're going to have a hard time getting a good sounding mix.

 
Exactly. You don't have to do gain-staging with a single audio sample except in the context of the project's mixer. With a soft synth, you have to gain-stage internally.

Anderton

soens
Rapture cut output by as much as 9dBs, quite significant. Not knowing this can dramatically affect the final mix.

 
I'm really not sure this matters much anyway, especially if the difference is around -1dB. Few patches consist solely of a single waveform. As soon as you add in more elements, effects, filtering, or whatever, you'll need to change the level to gain-stage properly.
 

 
You're right.  It doesn't matter.  When it comes to softsynths there are multiple ways of controlling volume:
 
1.  MIDI volume
2.  MIDI expression
3.  Velocity
4.  Audio master out from synth
5.  Audio gain in Sonar (sometimes connected to softsynth master out, sometimes not depending on synth)
6.  Mixing board volume (if you use one)
7.  Audio master out from synth for a particular patch (different from global master out, i.e. Kontakt)
 
To attempt to get matching levels based on numbers alone is, to my mind, an exercise in futility. 
 
Jerry
www.jerrygerber.com

The MIDI notes were set to 100. I then recorded a kick drum from an outboard synth into Sonar. Then I copied that sample and loaded it in to the soft synths with volumes set to 100.
 
Obviously your results will vary from mine. The point wasn't anything to do with what the levels were, but simply a comparison of how variant the outputs can be with all settings remaining constant. A person might believe that different soft synths have the same output levels when loading the same sample set and the settings are the same. This shows you that they aren't.
 
As far as the MIDI note levels go, no matter what they are set to the differences between the synths should remain fairly constant.
 
 
 
Another thing I discovered is Sonar plays some audio samples at a higher pitch than other audio players. (all set to 44,100/16bit). I do not understand that one at all.

soens
The point wasn't anything to do with what the levels were, but simply a comparison of how variant the outputs can be with all settings remaining constant.

 
When you emphasized how Rapture supposedly attenuated the signal by 9dB and that this would "affect the final mix," and presented as a solution that it was necessary to load multiple elements to raise the output level, it seemed you were focusing on the absolute change, not just a relative one. If you wanted a louder sample it seems it would make more sense to trigger it at full MIDI velocity instead of loading another sample into an element.
 

A person might believe that different soft synths have the same output levels when loading the same sample set and the settings are the same. This shows you that they aren't.

 
Okay, but I don't see any practical significance. You're loading a sample into an instrument, not an audio track. It will be processed in various ways, subject to various curves, played at differently velocities, and you'll eventually need to adjust the level anyway.
 

Another thing I discovered is Sonar plays some audio samples at a higher pitch than other audio players. (all set to 44,100/16bit). I do not understand that one at all.

 
That's not much data to go on, but if you're talking about samples, some "raw" samples are not perfectly in tune, so they're compensated for in the instrument. How this compensation occurs varies from instrument to instrument, so for example a sample that's slightly off pitch may be compensated for in Kontakt, but when loaded into Dimension Pro which doesn't have the same compensation, will remain slightly off pitch unless compensated for within Dimension Pro.
 

I just worry about the volume of the particular instrument I am working with, and couldn't care one whit about whether another synth would play the same samples at a different volume.
 
As long as the volume knobs work, and as long as midi supports velocity, I'm all set.
 
Bob Bone
 

Vaguely reassuring that soft synth volumes are all over the place. Just like the old days with the real things... Inconsistency made it fraustratingly fun.

I notice for instance NI's Moog emulator Monark can purposely default to all it's quirks...

It was just information. Take it or leave it. I'm not here to argue or anything like that. Time to move on....