Equivalent of "transpose" for volume
The very very short answer would probably be "Gain".. Gain can be greater than 1 (make it louder) or less than 1 (make it softer). Multiply every sample by the "Gain" and you change the loudness of the sound appropriately.
the longer answer is somewhat more interesting (or boring, depending on your perspective
).
Why does this seem not as straightforward for volume as it is for pitch?
First, you have to understand the difference between a physical parameter and a perceptual parameter.
A physical parameter is something you can objectively measure-- like "how high is the voltage", or "what is the frequency of the sin wave". All you need is some kind of meter. In contrast, perceptual parameters are how we (humans) perceive things. Is it a high pitch or a low pitch? How loud is does it sound?
Pitch is the perceptual parameter usually liked with frequency. Sounds with higher frequencies have a higher pitch.
Loudness is the perceptual parameter usually linked with amplitude. Sounds with bigger amplitudes generally sound 'louder'
Pitch and loudness are actually very different.
If you take a frequency value and multiple it by a constant (let's say 2.0), it's up an octave. You can do that on every track and it's always up exactly one octave. Everybody will hear the sound as exactly one octave higher. The perceived change (pitch) has a one to one mapping with the physical parameter (frequency). (yes, there are some pathological cases, but we can ignore those)
Loudness (volume) isn't so easy... even the meaning of making something "twice as loud" is very hard to define. while "up an octave" is the same for everybody, "twice as loud" isn't. Further complicating matters is that when you change the amplitude of a sound, it's change in "loudness" depends on the frequency content of the sound. (See Fletcher Munson curves http://en.wikipedia....r–Munson_curves). So how do you deal with that?
There isn't a nice clean mapping from Amplitude to Loudness like there is from frequency to pitch. That's why there is no "loudness" equivalent of "Middle C".
So that's a very long-winded way of saying why we generally label the "loudness" slider to be "Gain" (the physical parameter adjustment), while using the perceptual parameter (pitch or semitones) when we talk about transposing.
Brian Schmidt
GameSoundCon2012, Oct 24/25 San Francisco (just before AES)
the longer answer is somewhat more interesting (or boring, depending on your perspective
).Why does this seem not as straightforward for volume as it is for pitch?
First, you have to understand the difference between a physical parameter and a perceptual parameter.
A physical parameter is something you can objectively measure-- like "how high is the voltage", or "what is the frequency of the sin wave". All you need is some kind of meter. In contrast, perceptual parameters are how we (humans) perceive things. Is it a high pitch or a low pitch? How loud is does it sound?
Pitch is the perceptual parameter usually liked with frequency. Sounds with higher frequencies have a higher pitch.
Loudness is the perceptual parameter usually linked with amplitude. Sounds with bigger amplitudes generally sound 'louder'
Pitch and loudness are actually very different.
If you take a frequency value and multiple it by a constant (let's say 2.0), it's up an octave. You can do that on every track and it's always up exactly one octave. Everybody will hear the sound as exactly one octave higher. The perceived change (pitch) has a one to one mapping with the physical parameter (frequency). (yes, there are some pathological cases, but we can ignore those)
Loudness (volume) isn't so easy... even the meaning of making something "twice as loud" is very hard to define. while "up an octave" is the same for everybody, "twice as loud" isn't. Further complicating matters is that when you change the amplitude of a sound, it's change in "loudness" depends on the frequency content of the sound. (See Fletcher Munson curves http://en.wikipedia....r–Munson_curves). So how do you deal with that?
There isn't a nice clean mapping from Amplitude to Loudness like there is from frequency to pitch. That's why there is no "loudness" equivalent of "Middle C".
So that's a very long-winded way of saying why we generally label the "loudness" slider to be "Gain" (the physical parameter adjustment), while using the perceptual parameter (pitch or semitones) when we talk about transposing.
Brian Schmidt
GameSoundCon2012, Oct 24/25 San Francisco (just before AES)
Well stated!
Author
So, if the term is "gain", one question then: what would be 100% gain? Would it mean the volume remains unchanged or would it mean the sound would be double as loud? (always talking about perceived sound, that is)
So, if the term is "gain", one question then: what would be 100% gain? Would it mean the volume remains unchanged or would it mean the sound would be double as loud? (always talking about perceived sound, that is)
Actually, it depends on whatever soft- or hardware you're using. Using percentages for something labeled "Gain" is pretty rare though, in my experience.
Either it's labeled with dB or - like on guitar amps and pedals - with a knob that just goes from a minimum to a maximum value.
When it's indeed labeled with a true relative value, it's like Brian said:
1 (or 100%) means the Gain stage isn't changing the volume.
Cheers,
Moritz
Author
Internet cable getting cut = one week without internet. Gah.
It's just to make up for the difference between MIDI and the FM instruments. The idea is to scale the original volume of the MIDI instrument to match that of the FM instrument (which in turn will affect proportionally all velocity parameters and such, which is why it must be relative - 0 must stay 0, for instance).
OK, so what I have right now. That's what I wanted to know.
Actually, it depends on whatever soft- or hardware you're using. Using percentages for something labeled "Gain" is pretty rare though, in my experience.
Either it's labeled with dB or - like on guitar amps and pedals - with a knob that just goes from a minimum to a maximum value.
It's just to make up for the difference between MIDI and the FM instruments. The idea is to scale the original volume of the MIDI instrument to match that of the FM instrument (which in turn will affect proportionally all velocity parameters and such, which is why it must be relative - 0 must stay 0, for instance).
When it's indeed labeled with a true relative value, it's like Brian said:
1 (or 100%) means the Gain stage isn't changing the volume.
OK, so what I have right now. That's what I wanted to know.
This topic is closed to new replies.
Advertisement
Popular Topics
Advertisement
