Latency test of 23 amp modelers

[PLysander]

Sorry but this analogy is totally wrong. A correct test would would be play two metronomes at the same bpm into the LR speakers respectfully and have one 6ms or 11ms ahead. Would you hear a difference?

*very loud sigh*

You would. But unless your ears are 10 feet apart, i don't see how that test would be related to latency at all.

 

[tribedescribe]

*very loud sigh*

You would. But unless your ears are 10 feet apart, i don't see how that test would be related to latency at all.

You right I should have said headphones instead of speakers, just updated. Not trying to get on your bad side or belittle you, but I'm curious to know if you own a studio and have dealt with interface latency? If so when do you start to notice latency? I only ask this because if you have every done any recording in a studio you would understand where Im coming from in regards to latency.

 

[Will Chen]

I wouldn’t say that someone it is to the point of being “thrown off” That’s very hyperbolic IMHO but I do think maybe some are more sensitive to some things and it’s what they are referring to when they say the feel is off. Again I don’t trust science for saying what every should be able to detect. People can and do at times baffle science. Humans are very complex beings with a wide number variations so it’s possible there are those like Steve who may be sensitive to the latency of it all amp, fx, ca, and mic work being in ALL the digital realm.

If someone is that sensitive to latency then:

1. Changing positions relative to their amp would be extremely hard to overcome due to changes in the analog latency due to changing the distance between performer and speaker.

2. A player highly sensitive to latency above all else can get the lowest latency by using a modeler with wired earphones unless they're only playing is sitting less than 5 feet directly on axis of an amp.

These are absolute facts, you keep challenging the science but these are absolutely proven and measurable situations regardless of whether you believe some superhumans exist which can detect 1 ms latency.

 

[three_eyed_otter]

Have a good one,

3EO

 

[PLysander]

You right I should have said headphones instead of speakers, just updated.

Sorry, it's the same thing. My point is that latency will never cause the phase interference issues between your ears that you're describing.

Not trying to get on your bad side or belittle you, but I'm curious to know if you own a studio and have dealt with interface latency? If so when do you start to notice latency? I only ask this because if you have every done any recording in a studio you would understand where Im coming from in regards to latency.

Oh, i haven't done studio time in a while. At home i'm using a cheap Beringher UMC202HD and a suite of plugins - mostly Helix Native these days. I don't have an exact latency figure for this setup, but it is low (~5ms), as i normally aim use to as small buffers sizes and as high sampling frequencies as my computer will allow. I think it was 96kHz at 32 samples to record with Native.

I can easily record and track using that. Feels no different than plugging headphones into a modeler.

 

[burningyen]

For perceptual purposes, zero. Worst-case, on the order of fractions of microseconds.

 

[Gojira1954]

In defence of poor Vai who's taken a lot of **** over this, in the few live clips I've seen of him he has his rig setup right behind him on stage. Given his fast and precise playing style I can see how a bit of added latency could nudge him out of his comfort zone, or at least possibly get his attention.

IDK, he seemed pretty much comfortable playing and shredding his ass off with Whitesnake at Hellfest with no amps in sight and joining Coverdale out front in the catwalk.

 

[GMGM]

I certainly felt the difference in Fractal FM and HX Stomp, especially at lower volumes. This is obviously exaggerated if you use external effects or pedals in the send/return loops, or if you add a latency heavy processor before or after.

Irridium seemed fast enough that it wouldn't both me, even at low volumes. I was hoping for similar results from the UA Dream, but that doesn't appear to be the case.

 

[Tito83]

I believe latency is part of it, so it must be adressed and be as low as possible - but I don't think it tells the whole story. Connecting feel with just latency is a mistake IMO.

IDK, he seemed pretty much comfortable playing and shredding his ass off with Whitesnake at Hellfest with no amps in sight and joining Coverdale out front in the catwalk.

Played for years with modeling. It bugged me, but I put up with it because it was a lot more convenient. Thing is he is Steve Vai. His threshold of where a big rig starts being inconvient is a lot higher then mine, starting with the fact that I have to carry my stuff myself. I doubt he had a full on modeling rig running, but I know he prefers real amps for monitoring.

When someone says "modeling feels weird" some take that as "it makes me unable to play". That's not it.

 

[mehegama]

In general most people cannot hear latency until about 15ms. But, musicians, singers, guitarists begin to feel latency around 5-10ms. At 7ms, latency starts to mess with a guitarist’s ability to play on top of the beat. Sound starts to feel sluggish at 10ms. With practice, you can somewhat get used to it.

When a guitarist says they can hear latency at 5ms, they mean they can perceive it. But, it’s actually more of a feel thing. It starts to mess with our ability to play in the pocket.

My old computer and DAW back in the day typically had 20ms round trip latency. It was not a subtle thing. It felt dramatic. But, with practice, I learned to deal with it. I could tweak the parameters and get down to 12ms which felt better but that would cause sound anomalies.

i know it is very a very personal thing, but i think 5-10ms is way too low to feel latency. Other wise you would have seen musicians standing frozen in one space on the stage in front of the monitors. That is not the case.
Take as an example a classic orchestra where there are "shredding" passages with the violins and the drum guy or the horns are like 20 meters away but they still adapt to the latency.
I think you are a bit too strict with these low values.

 

[mehegama]

Thing is he is Steve Vai. His threshold of where a big rig starts being inconvient is a lot higher then mine

Again, why??? Why would you think that? Just because he is a famous guitarist? He contradicts himself by introducing his setup with the ax fx in the loop.

 

[fretworn]

Y’all are so sensitive… Just not a problem for work (or fun)

Regular amps are often, plug in and “immediate”, in your face. That’s not necessarily awesome. A lot of us started out that way.

We purposefully add crap to the signal chain that end up modifying that characteristic as we add f/x. At least I do …

 

[Tito83]

Again, why??? Why would you think that? Just because he is a famous guitarist? He contradicts himself by introducing his setup with the ax fx in the loop.

The fact he and many others are wrong about latency being the issue, it doesn't mean the difference isn't there. There's a difference, latency is part of the problem, but it's not all of it. Do I know what it is? No.

 

[mehegama]

The fact he and many others are wrong about latency being the issue, it doesn't mean the difference isn't there. There's a difference, latency is part of the problem, but it's not all of it. Do I know what it is? No.

I agree every person perceives it differently. However referencing Vai's thresholds when the guy has no idea on what he is talking about, when he contradicts himself and countless live videos of him confirm his ignorance, is not a very reliable metric for sure.

 

[GMGM]

Hopefully most of us can agree that modelling sounds fantastic anymore. I'm glad to see those debates seem to be waning.

But I see latency as being the next frontier. I have already switched over to 100% modelling when I record. The sound quality, option and flexibility are just too great to ignore.

This should really be a focus. I'd rather they spend time getting this down as close to "0" as is mathematically possible. I'm really curious about the results of Fractal's AFX3 "turbo" option and it's impact on latency.

i know it is very a very personal thing, but i think 5-10ms is way too low to feel latency. Other wise you would have seen musicians standing frozen in one space on the stage in front of the monitors. That is not the case.
Take as an example a classic orchestra where there are "shredding" passages with the violins and the drum guy or the horns are like 20 meters away but they still adapt to the latency.
I think you are a bit too strict with these low values.

At least classical musicians follow a conductor .

Again, why??? Why would you think that? Just because he is a famous guitarist? He contradicts himself by introducing his setup with the ax fx in the loop.

You have to take Steve Vai's experience with a grain of salt. He's perfectly happy using the DigiTech Whammy pedal, and it has a very long latency time. If you add that to the front end latency of a Fractal? Nope, that would be miserable. IMHO, the reality is that if you want to use modelling, you should probably use a system like Fractal/HX/KP/NUX for all of your amps, cabs and effects. Eliminate as many A/D/A conversions as you can, and you'll be much happier.

 

[Digital Igloo]

Don't want to get into semantics here, but calling the time it takes a transducer to push sound waves to your ears "latency" is a bit... disingenuous within this conversation. We've all been in large halls or churches and our brains have ZERO PROBLEM resolving hearing things at a distance. It's been this way for tens of thousands of years. At one time some caveman smacked two rocks against each other and another caveman across the field thought "Huh, that's interesting" and that was it.

Latency within digital systems is absolutely, positively a thing, and companies have lived and died by it. Audio interface manufacturers go to massive lengths to minimize throughput latency, up to and including special control panels or entire DSP or FPGA-based mixers with DSP or FPGA-based effects built in. Digidesign/AVID built an entire empire in part on this; UA and others have taken up the mantle.

The question is and always has been "How much latency is too much latency? How much will begin to mess with your timing or feel?" That's an incredibly personal question, and I imagine the more experienced musician/producer/engineer one is, or the deeper in the pocket they're able to play, the more they may feel a disconnect due to throughput latency. But to discredit their experience because "Oh, it's only like being an extra X feet away from an amp so who cares?" doesn't get what latency is.

TL;DR:

• Playing an amp X feet away from you in a Y foot-sized room: TOTALLY 100% NORMAL, ZERO DISCONNECT
• Playing an amp X feet away from you in a Y foot-sized room but it sounds or feels like it's Z feet farther away than it really is: FOR SOME, YES THERE CAN BE A DISCONNECT

i know it is very a very personal thing, but i think 5-10ms is way too low to feel latency. Other wise you would have seen musicians standing frozen in one space on the stage in front of the monitors. That is not the case.

Imagine an amazing drummer and bass player playing in the tightest, funkiest pocket you've ever heard. If either one is hearing themselves 5-10ms later than they expect to be hearing themselves (again, they're likely able to easily compensate for being X feet away from a bass amp or floor wedge), that perfect pocket could definitely suffer.

It's not the same thing, but I've timeshifted tens of thousands of audio files by less than 5ms to attempt hitting a pocket.

But to be fair, I agree that 5-10ms is more than sufficient for most guitarists. But any producer who tells an artist they aren't hearing or feeling something isn't a very good producer.

 

[PLysander]

Don't want to get into semantics here, but calling the time it takes a transducer to push sound waves to your ears "latency" is a bit... disingenuous within this conversation.

Not trying to stir the pot, but, why? That's quite literally what latency is It doesn't really make any difference if the root cause is A/D-D/A conversion, a delay line, or being too far away from a speaker: you do something, and experience a delayed reaction to it.

In fact, for many audio applications, the kind of latency introduced by digital devices is actually better behaved than acoustic latency, because it will be pretty much frequency invariant. There're no weird pitch and phase artifacts to worry about with digital sampling delays.

TL;DR:

• Playing an amp X feet away from you in a Y foot-sized room: TOTALLY 100% NORMAL, ZERO DISCONNECT
• Playing an amp X feet away from you in a Y foot-sized room but it sounds like it's Z feet farther away than it really is: FOR SOME, YES THERE CAN BE A DISCONNECT

I do get your point regarding "natural" vs device-introduced latency, and how the brain interprets those (psychoacoustics strike again!), but remember: people are going back and forth over single digit milliseconds in this thread - and that's what's grinding my gears so bad. Frankly, it's a bit of a disservice to modern digital gear, which is for the most part exceptionally good at keeping these latencies as low as humanly possible.

Playing through digital gear might feel the same, better or way worse for different people, and that's perfectly fine; there're many possible reasons why that would be the case. But a couple extra ms between you plucking the low E string and your speaker blasting it it will not be the root cause.

But to be fair, I agree that 5-10ms is more than sufficient for most guitarists. But any producer who tells an artist they aren't hearing or feeling something isn't a very good producer.

PS: This, i fully agree with. I really liked @benadrian 's take on the subject.

 

[ejecta]

If someone is that sensitive to latency then:

1. Changing positions relative to their amp would be extremely hard to overcome due to changes in the analog latency due to changing the distance between performer and speaker.

2. A player highly sensitive to latency above all else can get the lowest latency by using a modeler with wired earphones unless they're only playing is sitting less than 5 feet directly on axis of an amp.

These are absolute facts, you keep challenging the science but these are absolutely proven and measurable situations regardless of whether you believe some superhumans exist which can detect 1 ms latency.

I’m not challenging facts… I’m saying it’s impossible for science to say specifically that every single person will experience latency the exact same way when it comes to playing an electric guitar rig due the huge variation in human anatomy and that every single person will have the exact same aural and feel result when it comes to what contributes to latency and how their anatomy will process it. We aren’t even getting into the difficulties in communicating what people are experiencing then top it off the subjective aspects which are many. You can try make me say something I’m not to try and argue your point but that doesn’t work either.

Lastly digital systems have a latency that’s built into the very system that’s a separate issue than latency of sound coming from a source at a distance and that “built in” latency is a noticeable thing to some and for some evidently in meaningful ways just like Digital Igloo said but all analog systems in and of themselves do not come close to having the same issues of latency in the system that digital ones do.

 

[TubeStack]

So what you saying is if we submerge them in water we should get less latency?

Or boil them in milk.

 

[PLysander]

I’m saying it’s impossible for science to say specifically that every single person will experience latency the exact same way when it comes to playing an electric guitar rig due the huge variation in human anatomy and that every single person will have the exact same aural and feel result when it comes to what contributes to latency and how their anatomy will process it.

There're many things your body just cannot do. I can state that you (and Steve Vai ) cannot discern 1ms audio latencies with the same confidence i state you cannot see infrared light.