Obviously "Generated sound gain" de-couples the Simhub's Main Volume (Slider) from the Windows Audio Volume setting. Also, it seems to have the Simhub Main Volume work like a Gain Control on the input side, comparable to an audio mixing console where you can set the gain of incoming signals. (And final output signal level is then set separately by the dedicated Sound card's volume).
- Is this correct?
- What's the purpose / benefit of it?
- Does it make a difference in terms of how the "normalising" of the signal level(s) happens?
- Does Simhub utilise fmod limiter to normalise / scale down effect channel signals that summarise to levels >1?
I couldn't find any hints or documentation. Would be very happy to get some info. Thanks.
Hi, I can't answer the more technical questions but to provide some context, the "Generated sound gain" option was added to offer more flexibility with volume control.
https://github.com/SHWotever/SimHub/issues/421
@romainrob Thanks a lot! Your info helps me to understand the main reason (I assume) why the "Generated sound gain" option exists.
Still, I'm very interested to know more about how Simhub deals with level normalisation, since it seems to me that there are some misconceptions about "clipping". I posted my "more-in-depth" related question recently in the Simhub discord Shake-It-Discussion Channel (but couldn't find any answers.)
The reason why I'm asking: I was under the impression that people are concerned about clipped signals, while I came to the conclusion that Simhub's fmod sound engine and its limiter feature (if used by Simhub?) scales down "overshooting" levels linearly anyway. To me that looks like an automated gain control where "clipping" can't happen anyway. No matter if you "mix / stack up" effects levels that could add to summarised levels above 100% "within" Simhub's gain stages.
In other words, since i can't "force" Simhub to output levels that can overdrive the sound card input, I was assuming there must be some kind of automatic gain control at work (in Simhub's master out bus). I admit it's a bit of a nerdy question, but from an audio mixing perspective it's interesting, because it influences the way you compose your tactile-mix in terms of gain staging.
Though, I don't know if my assumptions a correct?
Obviously "Generated sound gain" de-couples the Simhub's Main Volume (Slider) from the Windows Audio Volume setting. Also, it seems to have the Simhub Main Volume work like a Gain Control on the input side, comparable to an audio mixing console where you can set the gain of incoming signals. (And final output signal level is then set separately by the dedicated Sound card's volume).
- Is this correct?
- What's the purpose / benefit of it?
- Does it make a difference in terms of how the "normalising" of the signal level(s) happens?
- Does Simhub utilise fmod limiter to normalise / scale down effect channel signals that summarise to levels >1?I couldn't find any hints or documentation. Would be very happy to get some info. Thanks.
The answer is in the documentation as usual:
All the gains are "multiplicative" :
Effect gain at one output = effect gain * effect per channel gain * channel gain * general gain
For instance :
-
effect gain = 100%*channel gain = 100%*effect per channel gain = 100%*general gain = 100%gives a 100% gain for the effect at the output. -
effect gain = 80%*channel gain = 100%*effect per channel gain = 100%*general gain = 50%gives a 40% gain for the effect at the output.
Which clearly states a number of things:
1) The extensive use of the word "GAIN" in SimHub is 100% WRONG. But used in FMOD documentation extensively, because it is geared towards the *game developer* who would be interested in GAIN to set the *game* output signal levels to match aa standardized *Full Scale*. So given the English -> French translations present... it is an honest mistake by the developer of SimHub.
2) The above statement from the simhub developer repeats 100%. He also says that the signal path is multiplicative by a maximum denomination of 1.0 (100%). Which means that the only option other that 100% is LESS THAN..... meaning it is a function of VOLUME... Not GAIN.
1*1*1*1=1
0.80*1*1*0.50=0.40
Subtractive = VOLUME.
3) Normalization in FMOD is an additive function. Not subtractive. Meaning, you would artificially raise the levels of the signal to match a desired output level (Like say... Full Scale Standard). Simhub does not do this (there is no reasonable case to need this, since the game signal already does this) and that is why there is no documentation to support this. Because it doesn't normalize the signal.
4) It does however compress the output signal. In fact it does so embedded in the Multi-Channel function of FMOD. I would not rejoice so soon though. The proof is again in the SimHub Documentation:
"When starting with ShakeIt, the main settings you will use are the General gain and the Effect gain, make sure to keep other settings at 100% when beginning (see below all gain settings purposes)."
The problem with the multiplicative model is that while it can be very flexible, the more flexible it is, the more complex it becomes. Which means the more easily you can lose track of where you are.
5) Back to compression... a life saver when you are not in control of the input (sudden and uncontrollable loud noises in a recording), or if the inputter is daft (doesn't understand gain/volume/levels/scales)
But rest too heavily on it.... you get "muddy" or "warbled" outputs. Even worse... as that compression kicks in an out spontaneously, it outputs transient spikes when that knee kicks in and out. Some might mistake it for resonance even.....
You could fix it... except you can't in Simhub. Just a standard FMOD compression parameter, with some knee options. Which means that you as the user is not in control of your signal anymore (no shade on the developer... compression is there for a reason... the trick is to not trigger it.... something the developer ASLO warned you about at the beginning of the documentation).
Thanks VOR for shedding a profound light on the topic!
@tennenbaum My apologies. I also forgot to mention the answer to your title question.
The difference between Generated Sound Gain and Windows Volume is actually profound in a sense.
When you install SimHub, it asks you whether you wish to link Windows and Simhub Master Volume.
As a person who runs his Rig in the basement away from my sweetie and anyone else who could hear it... I personally keep them linked. However... I only keep them linked because I know that I will not change it/them from 100%
If you wish to suppress the output in a linear fashion (Like when you live in an apartment, or have someone else who is trying to sleep while you race)..... This setting and the configuration becomes more important.
Generally speaking, you are trying to suppress the output globally.
But when you link the Global SimHub Output and the Windows Output... and present it with anything other than 100% (which is just a pass-through), you will encounter a two-fold suppressions globally.
This is where linking them can cause problems.
You drop your simhub output by 10% (90% Global Gain... shoot me on that Global Gain thing)
So:
Linking the windows and Global Gain will generate a 20% master mix reduction.
Unlinking them (Generate septate Gain) Setting will reduce it by 10% (providing you keep your Windows Volume at 100).
Why?
Because Simhub has a program will send its signal (in all of its suppressed and manipulated glory) to Windows, who will ALSO give you an opportunity to suppress it separately. Which an fools errand if you are the person creating the signal exactly the way you want it, customized for your setup, to begin with.
The developer says so in the documentation.