6 replies to this topic

[Joe Reynolds]

Hey guys, just a quick question.

 

I've had my audio integrated into a game (always a good start) but I found myself being annoyed at a few things.

 

.The menu music is too loud relative to everything else

.The in-game sound effects are also too loud

 

Would I make these quieter before sending them of to the programmer or would it be the programmers job to adjust these levels?

 

Cheers guys!

 

[Mussi]

Games typically allow the user to set the master-, sound effect- and music volume in the option menu. I'm guessing it would be your job to determine what the default values should be and maintaining consistency between different sounds and music, but don't take me up on that.

[nsmadsen]

Instead of juggling multiple versions of assets, have the programmer lessen the volume via code. It's pretty simple to change the playback volume of an asset and keeps file management much more simple.

[TheChubu]

And please include different settings for volume for your users (ie, voice, music, effects, master), I hate when the gun shots are too loud or the voices too quiet and you only get master and music settings.

[bschmidt1962]

In general, you want to deliver sounds at a fairly 'hot' level, and set the volumes in software.

But first find out how exactly the programmer sets volumes.  Is it on a 0.0 - 1.0 scale?  (common for Flash), is a MIDI-like 0-127?  is it in dB?  Can it  "go to 11" (i.e. play a sound louder than it's default level).

Knowing how their system deals with volume is very important for you to remain in control of the mix, even indirectly.

 

You want to make sure that when you deliver the sound, you also deliver the volume level it should be.  Otherwise, you'll have the programmer either 1) guessing or 2) just playing everything full volume.

So, for example, your delivery might a spreadsheet that looks like this

Event       WaveFile          Volume

Select        SelFX3.mp3        0.7

DeSelect    DeSel.mp3         0.6

mainMusic  MyTheme.mp3   0.3

 

That way, the programmer knows what to set to what.

 

... or talk them into something like FMOD or Wwise that gives you the control

 

There are some exceptions-- if you have a very quiet wind rustling sound for example, it doesn't really make any sense to normalize (set to max volume) that sound.

Also, if you have a set of sounds that go together, it's good to have their  levels be correct relative to each other.

 

 

 

Brian

Brian Schmidt Studios

GameSoundCon

[MrDaaark]

If you have to make a chart with tons of volume adjustments, you just as easily make the adjustment yourself. It should all be consistent and normalized. You don't want everything adjusted with code for the same reason you don't hand over an over saturated texture and say "You can de-saturate it with code".

In that case, "adjust with code" will end up being someone else opening up the media editor and making the adjustment that the media creator should have in the first place.

[bschmidt1962]

If you have to make a chart with tons of volume adjustments, you just as easily make the adjustment yourself. It should all be consistent and normalized.

 

While you have a good point, that can be a very inefficient workflow.

When it comes time to tweak the sounds' volumes in-game, it can be extremely cumbersome to have to check out the asset, edit it, and then check it back in just to make it a bit louder or softer.  Also, if that wave asset is used in multiple places, then you've unintentionally changed the sound level elsewhere in the game of that sound, which might have been fine in that context.

 

it can be far easier to simply tweak a constant in the game code or a table somewhere.  When you're in the "iterate/try" loop, it's waaaay faster and is also less prone to errors (i.e. using the wrong version of the wav file).

 

So having to go back to the source asset and edit it whenever you want to tweak the volume is not a real good solution.

Short of a proper tool, I try to have games use a data driven table that maps GameEvent to [WaveFile,Volume] as a bare minimum.  And I (as sound designer/composer) own the wave files and the data table.

 

Ideally, you would deliver the sounds roughly where you want them, and then have the ability in code (or data-table) to tweak the volume levels up or down from there without having to change the wave assets themselves.

 

Btw--its for reasons like this that tools like XACT, Fmod, WWise, etc were created in the first place :-)