DSP & Latency In Powered Speakers and FRFRs
- Thread starter la szum
- Start date
Stone
Rock Star
- Messages
- 5,777
Not sure if my PowerCabs have any DSP other than engaging or defeating the speaker sims
But if any Latency I certainly can't her anything
jay mitchell
Roadie
- Messages
- 620
Yes.
It doesn't have to be a lot, but every little bit matters. My powered monitor adds ~.6ms latency. IIRC, the CLR adds just over 1ms.
- Messages
- 14,784
That's why I am curious, Jay. Not that the DSP in a powered speaker adds too much, but that
together with a modeller, and maybe a wireless, or a digital pedal with it''s own A/D/A conversion,
there is a cumulative latency that accrues.
Do you know where I can source specs on the latency of various speakers/monitors?
together with a modeller, and maybe a wireless, or a digital pedal with it''s own A/D/A conversion,
there is a cumulative latency that accrues.
Do you know where I can source specs on the latency of various speakers/monitors?
jay mitchell
Roadie
- Messages
- 620
It would be up to the manufacturers to disclose that information.
unix-guy
Roadie
- Messages
- 797
If you're using AES/EBU I don't think there will be any latency because there's no conversion happening.
The 4CM will get extra latency because of multiple conversions.
- Messages
- 3,702
From what I've read from non-official sources, the QSC K.2 series' latency depends on which EQ profile is loaded, with a range of 1.5 to 2ms
Sascha Franck
Rock Star
- Messages
- 6,071
Well, there's also processing latency. No idea how much that might be, though.
All I know is that some years back I got a Bose Soundlink Mini as a birthday present (from my sister who possibly didn't even know any other brands...) - and I was like "yay, maybe a great little practice speaker!" (along with, say, an MS-50G, which is operating for around 6 hours on batteries). But uhm. Latency through the analog mini TRS input is almost as bad as via bluetooth, go figure.
Btw. (just in case anyone should ever stumble over that turd), in addition, BT is pretty weak (very short range), the thing's volume isn't even remotely party-ready (and we're talking pretty silent parties already) and it has the most horrible smiley EQ of all times built in.
OrganicZed
Roadie
- Messages
- 598
It all adds up. Here's a hypothetical signal chain.
Shure GLXD+ Wireless guitar transmitter: 4.53 - 8 ms (dependent on frequency band used)
Fractal FM3 : 3.2 ms
Presonus Studiolive 16.0.2 Digital Mixing Board: 2 ms
Sennheiser DW-D Wireless IEM transmitter: 1.9 ms
Total : 11.63 - 15.1 ms
According to the Google, most people can't perceive less than 10-15 ms of latency.
Shure GLXD+ Wireless guitar transmitter: 4.53 - 8 ms (dependent on frequency band used)
Fractal FM3 : 3.2 ms
Presonus Studiolive 16.0.2 Digital Mixing Board: 2 ms
Sennheiser DW-D Wireless IEM transmitter: 1.9 ms
Total : 11.63 - 15.1 ms
According to the Google, most people can't perceive less than 10-15 ms of latency.
Jarick
Rock Star
- Messages
- 3,355
I'm with @OrganicZed - I don't think you're going to notice ~10ms of TOTAL latency. Of course when you stack wireless plus digital modeler plus interface plus other pedals that may add up to more.
But I'm currently running my guitar through a cheap wireless into a digital modeler into an interface into monitors and don't really notice enough latency to bother me. It's probably not as immediate as plugging directly into an amp but it's not noticeable. And I HATED playing plugins several years ago when there was high single digit nominal latency with my old computer and interface.
But I'm currently running my guitar through a cheap wireless into a digital modeler into an interface into monitors and don't really notice enough latency to bother me. It's probably not as immediate as plugging directly into an amp but it's not noticeable. And I HATED playing plugins several years ago when there was high single digit nominal latency with my old computer and interface.
nicolasrivera
Roadie
- Messages
- 407
It's all about math. Sum in all the DSPs involve in you chain and that it's the true number you need to evaluate and feel if its an issue under your fingers.
Sascha Franck
Rock Star
- Messages
- 6,071
That is not exactly true (and no, let's not get into the Steve Vai <1ms debate again). Those latency perception tests are usually done as kinda "timing tests". Hence, you are to differentiate between a signal being dead on and whatever amount of ms late/early.
But it's a very, very different story once it comes to haptic vs. audible feedback.
And along with that there seems to be a very narrow range of "oh yeah, that feels tight" and "ok, that's somewhat distracted". And, to make matters worse, the start and end points of that very range seem to be quite different for different folks. In addition, apparently it's got not all that much to do with latency tolerance.
As an example: anything up to 7ms might feel tight for you (fwiw, let's suggest we're using headphones - outside of that world, there's yet some more variables kicking in). 9ms however might already feel distracted, with 7-9 being a sort of grey area. And you might then be fine with anything up to, say, 15ms (which is quite a lot under headphones, but these are all just arbitrary numbers).
Another person however might think that anything up to 9ms is fine, but anything over 10 is not even acceptable anymore.
For a while, I've done quite some tests regarding the matter, and once picking (or striking, as with a drum stick) comes in, there seem to be no truly reliable statements regarding exact numbers.
My suspicion would be that there's different ways to perceive sounds (or rather their initial impact/transient). As a guitar player, you could for example basically feel the moment your pick attacks the strings, make that your timing reference and pretty much ignore the resulting audio. That'd possibly a good way to ignore latency. Or you could do it the other way around, as in only listening to the audio output, pretty much trying to line it up against the rest of the music. That'd be a good way to compensate latency. If we could instantly switch between the two, that'd possibly be the best, but life ain't easy, plus there's a lot more variables, especially once we're outside of any controlled environment.
Anyhow, in a nutshell: Whether it matters or not is entirely up to you. But it's certainly the best to avoid additional latency whenever possible. One thing I'm doing to get there is bringing my own little mixer for IEM gigs, simply because I don't trust whatever possibly unknown FOH folks and their methods to tweak my guitar signal anymore. Even if the baseline latency of nowadays digital consoles is extremely small, I had it happening more than once that it was very noticeable, likely because some FOH tool decided to slap a lookahead limiter or whatever onto my signal (or anwhere else and the monitoring wasn't compensating, whatever...).
nicolasrivera
Roadie
- Messages
- 407
Thank god im not one of those people!
ian_dissonance
Roadie
- Messages
- 561
In this instance I don’t think they intended this to be used with real time tracking. If you’re just mixing playback you’d never even know it’s there.
- Messages
- 14,784
This is precisely why I am so curious. Latency is not a monolith. It is a cumulative process built up
by all the A/D/A conversions going on in our ENTIRE signal path. Adding 10-15ms is massive for me.
It's like playing 30 feet from your source instead of 15 feet.
5-10ms latency is pretty natural separation/distance when playing in a room with others, or on a stage.
Add 10-15ms and that can easily make you feel distant from the band and your own sound sources.
JiveTurkey
Goatlord
- Messages
- 16,492
Grab a Fender FR.
Similar threads
