Do effect devices constantly process the incoming signal?

Have an urge to learn, or a calling to teach? Want to share some useful Youtube videos? Do it here!
deeplink
Posts: 294
Joined: 08 Jul 2020
Location: Dubai

Post 21 Mar 2021

Hi RT

As the title suggests, do devices such as distortion, delay, reverb etc. constantly process the incoming signal - even when there is no signal?

To put in another way,
If I had 200 reverb units wired up,
Would they still use the same DSP/processing power if they were connected but no sound was passing through, as they would if the instruments were all generating a sound?

I would imagine in bypass mode, or if the device isn't connected - then no processing is happening. But do they process silence?

User avatar
Loque
Moderator
Posts: 8780
Joined: 28 Dec 2015

Post 21 Mar 2021

Depends on implementation. Some fx constantly produce a signal, also on silence. Others recognize silence and stop processing.
:reason: 11, Win10 64Bit.

User avatar
jam-s
Posts: 1313
Joined: 17 Apr 2015
Location: Aachen, Germany

Post 21 Mar 2021

It is up to the developer to implement silence detection and then to stop the processing of audio to save CPU cycles.
If you're in Aachen, come and visit us at the Voidspace. ... Pool's closed due to corona.

User avatar
pongasoft
RE Developer
Posts: 291
Joined: 21 Apr 2016
Location: Las Vegas

Post 21 Mar 2021

I concur that it is entirely implementation dependent. From the DAW point of view, the RE is a black box, so at a minimum it must must make a call to the RE even if the RE is doing nothing because it has no way to know that the RE will do nothing (for obvious reasons, if the device is bypassed, then it will actually skip it entirely, or in other words, bypass is implemented at the DAW level, not at the RE level). But the RE is never doing "nothing". At a bare minimum, and if properly implemented, it would have to detect that the input is silent (or not connected) and return as early as possible. REs use properties (like the amount of wet/dry signal, the volume of the device, an audio socket, etc...) and I have seen code where the very first part is to read all those properties at every call (prior to detecting silence). For example, detecting that the device is connected or not require a property to be read. There is a way to be notified of property changes instead of reading them, but then it is the job of the DAW to maintain this state and generate the appropriate list of changes to pass down to the DAW, or in other words, the work has to be done somewhere (even if no changes, I am sure that Reason developers are smart enough to do optimizations, but there has to be some flag to check that nothing has changed).

So yes adding devices that are unused is not free (+ there is the memory required for each of them no matter what) and is highly dependent of how careful the code has been implemented.

  • Information
  • Who is online

    Users browsing this forum: CommonCrawl [Bot] and 1 guest