• We would like to remind our members that this is a privately owned, run and supported forum. You are here at the invitation and discretion of the owners. As such, rules and standards of conduct will be applied that help keep this forum functioning as the owners desire. These include, but are not limited to, removing content and even access to the forum.

    Please give yourself a refresher on the forum rules you agreed to follow when you signed up.

Latency of amp modelers

sprint

Fractal Fanatic
confirms a reason why strymon iridium feels so great to me in the loop of Ax3 compared to other outboard digital pedals I've tried.
 

mr_fender

Axe-Master
A/D and D/A conversions add a small amount. The rest is DSP processing latency. The speed of the processor and memory and the size/complexity of the code it's running determines that. Things like oversampling and sample rate conversions affect it too.
 

GlennO

Fractal Fanatic
I’m not an expert, so I want to learn, what are the factors that affect latency?
Mostly A/D and D/A conversion. The processing inside the modeler usually contributes less to the latency than than the conversion. For example, the latency for a simple Axe-FX preset is usually more than 50% due to the conversion.
 

GlennO

Fractal Fanatic
the size/complexity of the code it's running determines that. Things like oversampling and sample rate conversions affect it too.
This. Before anybody gets too hung up on latency numbers, the important thing is what value you are getting for your latency, not the raw latency number itself. A crummy amp model will have low latency.
 

JoKeR III

Fractal Fanatic
In a direct comparison between two sounds, most people can detect a one millisecond delay. Without a direct comparison, it is possible to notice a delay or latency of 10 milliseconds, detecting anything less than 5 milliseconds is getting into the realm of not humanly possible. 3 milliseconds of delay is approximately the difference between speakers 3 feet vs 10 feet away from a listening position.
 

mr_fender

Axe-Master
So a less accurate modeler will have less latency? Or size/complexity doesn’t directly relate with quality?
Not necessarily. Depends on how well the code is written and optimized as well. Poorly written code will have more latency. It's kind of like any language. One person might talk about something for 20 minutes while another might be able to convey the same information in 2. Good code is not just about getting the results you expect. It's about getting the results you expect as efficiently as possible. Fewer instructions means less processing time.
 

GreatGreen

Power User
In a direct comparison between two sounds, most people can detect a one millisecond delay. Without a direct comparison, it is possible to notice a delay or latency of 10 milliseconds, detecting anything less than 5 milliseconds is getting into the realm of not humanly possible. 3 milliseconds of delay is approximately the difference between speakers 3 feet vs 10 feet away from a listening position.

If you layer two identical sounds on top of each other, 1 ms apart, you cannot detect a delay so much as phase cancellation.

If you took two sounds that don’t overlap in frequency, like for example a bass drum and a high hat, it is almost impossible to detect any difference under 10 ms. Even 10 ms itself is something you probably would not notice unless you were in a lab environment directly trying to determine differences under a microscope with close concentration.
 

Greg Ferguson

Axe-Master
Not necessarily. Depends on how well the code is written and optimized as well. Poorly written code will have more latency. It's kind of like any language. One person might talk about something for 20 minutes while another might be able to convey the same information in 2. Good code is not just about getting the results you expect. It's about getting the results you expect as efficiently as possible. Fewer instructions means less processing time.
And to cloud the issue is the number of samples of the signal that have to be done to provide accuracy. Fractal won’t say how many they do because that’s one of their secrets, but it’s one of the reasons for using such powerful chips.

Each sample pass improves the accuracy, so they loop through that code n times and consume x milliseconds each time then move on. The size/amount of the code, its efficiency, and the desired accuracy combine to determine the latency.

And that chart is terribly misleading and is basically worthless because it doesn’t reflect the quality of the output, only how long a unit took to achieve output. That a product did it faster according to the chart, but the sound sucked in actual use, doesn’t help anyone pick the better product.
 

sprint

Fractal Fanatic
And to cloud the issue is the number of samples of the signal that have to be done to provide accuracy. Fractal won’t say how many they do because that’s one of their secrets, but it’s one of the reasons for using such powerful chips.

Each sample pass improves the accuracy, so they loop through that code n times and consume x milliseconds each time then move on. The size/amount of the code, its efficiency, and the desired accuracy combine to determine the latency.

And that chart is terribly misleading and is basically worthless because it doesn’t reflect the quality of the output, only how long a unit took to achieve output. That a product did it faster according to the chart, but the sound sucked in actual use, doesn’t help anyone pick the better product.
not completely worthless as it tells you the ones whose
latency is hopelessly bad - so even if they sounded good (which I'm not sure they do), the poor latency would override any advantage in tone given that, imo, 10+ms of minimum latency starts to suck the life out of any tone.
 

Greg Ferguson

Axe-Master
not completely worthless as it tells you the ones whose
latency is hopelessly bad - so even if they sounded good (which I'm not sure they do), the poor latency would override any advantage in tone given that, imo, 10+ms of minimum latency starts to suck the life out of any tone.
True, not completely worthless but few people will understand that distinction.
 

sprint

Fractal Fanatic
3 milliseconds of delay is approximately the difference between speakers 3 feet vs 10 feet away from a listening position.
Physical distance related latency between listening position and sound source is often cited in discussions about latency perception, sometimes to suggest less than commonly understood impact on feel, but its not often highlighted that distance between listening position and sound source, as well as other factors impacting latency, needs to be added on to total rig latency - ie: I buy my Mooer thinking 9ms will be good enough but I sit 5 feet from my cab (+5ms) and run my fav pedal in the Mooer's loop (+9ms+) > now I'm at an intollerable 23ms.
 
  • Like
Reactions: Rex
Top Bottom