Latency test of 23 amp modelers

12TameMen

Member
Messages
342
Model and source? Same testing methodology? Curious, I have one.

Quick Google only turned up this. Fractal Forum, six years ago.

Untitled.png
 

Jkist7

Member
Messages
193
Are you actually going to answer his question, or you don’t know?
According to my VERY ROUGH calculations, the latency of sound over a distance of 2 meters (just over 6ft) would be just under 6ms. If we add another meter (about 10ft), that delay time goes to 8.75ms. This is assuming average room temp.

So we can say most people will experience 9ms of audio latency per 10ft away (technically 10.13ft) from their speaker.

So if you already play at a decent distance from your speakers, and then add on top the latency from your modeler, and then perchance you have other digital pedals adding further latency, you will absolutely feel it. Whether it is crippling and inhibits your ability to play well, thats up to each individual, but I'd wager most of us would be able to compensate after an adjustment period, even in the absolute worst case.
 
Last edited:

MuzicToyz

Member
Messages
443
In all reality is there really any reason to be too concerned with latency in any modern modeler? Not like I have every played anything in the past 20 years that ever made me feel "disconnected" other than running direct with a bunch plugins. I would assume by the time you plug in a handful of pedals and a couple 20ft cords you have over 10 ms. No one has a meltdown if someone posts their new 30 pedal board, yet somehow a few ms difference between brand A and B somehow indicates something "better"? Some folks just have too much time on their hands...
 

Jynx

Silver Supporting Member
Messages
3,509
In all reality is there really any reason to be too concerned with latency in any modern modeler? Not like I have every played anything in the past 20 years that ever made me feel "disconnected" other than running direct with a bunch plugins. I would assume by the time you plug in a handful of pedals and a couple 20ft cords you have over 10 ms. No one has a meltdown if someone posts their new 30 pedal board, yet somehow a few ms difference between brand A and B somehow indicates something "better"? Some folks just have too much time on their hands...

Don't get it either. It's like someone boasting that their love appendage is one mm longer than someone else's.
 

PLysander

Member
Messages
589
In all reality is there really any reason to be too concerned with latency in any modern modeler? Not like I have every played anything in the past 20 years that ever made me feel "disconnected" other than running direct with a bunch plugins.

Guitar cord length will have zero measurable impact on latency, but other than that... no. There really is not.

It's hard to properly convey how insignificant sub-10ms latencies are for guitar stuff. What's funny is that Leo Gibson made a good job of highlighting the caveats for these figures in his own video, and no one seems to have registered that part.
 
Messages
213
Lots of the older Digitech rack units had that latency issue solved over 25 years ago.
And people wonder why I still gig with those older rack units.
 

snagged

Member
Messages
402
Lots of the older Digitech rack units had that latency issue solved over 25 years ago.
And people wonder why I still gig with those older rack units.
Two tin cans and a piece of string don't have latency either. That doesn't make it a good amp modeling rig.
 

Quad4

Member
Messages
1,891
I would assume by the time you plug in a handful of pedals and a couple 20ft cords you have over 10 ms.

Cables of just about any length are not going to matter for audio latency. Same for analog pedals.

It's hard to properly convey how insignificant sub-10ms latencies are for guitar stuff.
10ms is significant if you add one device that has a few ms in latency.


With what workload? [For Atomic]


Not sure the latency varies that much from preset to preset on the Atomic if you're not using the pitchshifter or FX loop.

Atomic Amplifire has a very optimized linear pipeline with a fixed set of blocks that can be on/off or reordered. Probably sounds scary to those coming from more flexible modelers, but it's never been an issue for me. The benefit of this design is that:
  • You cannot run out of CPU (DSP)
  • Preset changes are fast and gapless.
  • Presumably Mark has the DSP code very optimized and leverages a lot of tricks/assumptions due to the signal flow.
The main thing on the AmpliFire that would add more latency is use of the pitchshifter block (the physics of the low note waveform means you can't eliminate pitch shift latency). The other block that introduces some latency is the FX Loop (and that latency is pretty low but do not recall the actual number). Might be others but I don't think they add much if any effective latency (unless that's part of the effect like reverb).
 




Trending Topics

Top Bottom