preamp vs amp in digital: meaning?

DSP, Plugin and Host development discussion.
Post Reply New Topic
RELATED
PRODUCTS

Post

hi there,

for what i know, in "analogic", preamp take a mic signal (in milli-volts) and raise the voltage to line level (½ to 1 volt). than, the amp take that line level and raise to speaker level (10 volts or above). it makes sense.

but what's the meaning of those controls in a digital plugin?
there isn't any concept of mic and/or line level, neither volts: its all between [0, 1]. does it models the circuit scaling (in the case of amps) from "½ to 1 volt" to "1 to 10 volts" within the [0, 1] range? or (in the case of preamp) from "0 to milli-volts" to "1 volt" within the same [0, 1] range?

example, guitar rig, which have both preamp and amp (master):

Image

Post

In a guitar amplifier, the pre-amp is usually responsible for most of the distortion and "character", while the power-amp brings it to speaker-level and usually has a more subtle distortion effect.

In this case, "boost" will be a clean digital boost before the preamp stage, "gain" will set a gain parameter in the main distortion stage, and "master" will be a clean volume fader, which might come before or after a modelled power-amp stage.

Post

Derozer wrote: Wed May 12, 2021 1:58 pm for what i know, in "analogic", preamp take a mic signal (in milli-volts) and raise the voltage to line level (½ to 1 volt). than, the amp take that line level and raise to speaker level (10 volts or above). it makes sense.
That's the theory.

In practice, your average "musical" preamp is quite "bad" in the engineering sense (although that doesn't mean that the engineers designing didn't intentionally make it that way; after all, we're interested in "musical" rather than "ideal" behaviour) and usually doubles as something varying from subtle saturation (eg. a "clean" vocal preamp) to full-blown distortion (eg. your average high-gain guitar preamp). It is this "non-ideal" behaviour that a digital "preamp" tries to model.

Post

mystran wrote: Wed May 12, 2021 3:53 pm
Derozer wrote: Wed May 12, 2021 1:58 pm for what i know, in "analogic", preamp take a mic signal (in milli-volts) and raise the voltage to line level (½ to 1 volt). than, the amp take that line level and raise to speaker level (10 volts or above). it makes sense.
That's the theory.

In practice, your average "musical" preamp is quite "bad" in the engineering sense (although that doesn't mean that the engineers designing didn't intentionally make it that way; after all, we're interested in "musical" rather than "ideal" behaviour) and usually doubles as something varying from subtle saturation (eg. a "clean" vocal preamp) to full-blown distortion (eg. your average high-gain guitar preamp). It is this "non-ideal" behaviour that a digital "preamp" tries to model.
thats quite clear. whats not clear is how it scale in different range.

example, in vcvrack V10pp is 0db (thus 1.0).

if i want to model a preamp, it would act (as analogy) between milli-volts to 1 volts - i.e. 1.0 / 10 / 1000 = 0.0001f to 0.1f).

so basically it should start to saturate for any signal above 0.1 :o which makes no sense for a digital signal :) such as weak signal saturate early...

hope the dubt is well expressed now?

Post

Derozer wrote: Wed May 12, 2021 4:41 pm thats quite clear. whats not clear is how it scale in different range.

example, in vcvrack V10pp is 0db (thus 1.0).

if i want to model a preamp, it would act (as analogy) between milli-volts to 1 volts - i.e. 1.0 / 10 / 1000 = 0.0001f to 0.1f).

so basically it should start to saturate for any signal above 0.1 :o which makes no sense for a digital signal :) such as weak signal saturate early...
Well, in digital it's all just numbers (and usually floating point numbers) so any choice of scaling for voltages is arbitrary. In my opinion, these questions you ask are largely UX questions: you should choose input and output levels that "feel good" in terms of how you intend your model to be used.

If you're doing actual simulation, then scale the "feels good" input level to some nominal voltage level internally, then compute the simulation, then scale the output to a "feels good" output level at the output. Remember that even in analog, lowering the voltage swing is just a matter of a resistor or a potentiometer, so your model isn't necessarily any less valid if you give the user a knob to adjust (within the range of sensible levels, as defined by the "feels good" metric).

Post Reply

Return to “DSP and Plugin Development”