ProSoundWeb Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: [1] 2  All   Go Down

Author Topic: Effect of digital latency on phase in monitor solution  (Read 8489 times)

Chris Shaw

  • Newbie
  • *
  • Offline Offline
  • Posts: 19
Effect of digital latency on phase in monitor solution
« on: July 26, 2011, 08:21:02 AM »

I have recently proposed a new sound system for my church but suddenly realised that we could well run into problems with the proposed monitor setup. I just wanted to run this by the forum to check whether or not my concern is valid...

We typically have 6 to 10 musicisans/singers on stage for a normal worship set. Of course the ideal would be to provide each with their own monitor mix, however time, personell and equipment resourses wont allow for that. I wish to avoid providing more than 6 mixes from front of house (5 regular plus 1 spare). We're moving from floor wedges to IEM too.

The main method I have chosen to reduce the number of mixes was to use equipment like the Rolls PM351 (http://www.rolls.com/product.php?pid=PM351) for instruments like bass and keys. In theory this allows musicians to receive a more general monitor mix and then add a bit more of themselves to taste (I wouldn't necessarily use the mic in/thru part of it). For example I can send a drum and bass monitor mix and then the bassist can add a bit more bass themselves, and in the same way the keyboard player could share a mix with other musicians.

The mixer/multicore system I have proposed is the Roland M-300 mixer with two Roland S-1608 stage boxes (split via a standard ethernet switch since one mixer REAC socket will be used for multitrack recording). Digital systems of course have latency, and this is where I foresee the potential problem...

The quoted mixer/multicore system latency is ~375 microseconds, plus ~200 microseconds for the switch. If my maths is correct, this delay will mean ~870 Hz would effectively be phase shifted by 180 degrees on the monitor mix which would then be mixed with the original signal and subsequent cancellation would occur at that frequency (assuming same amplitude from both sources). (575 microseconds x2 = 1.15 miliseconds = period of a ~870 Hz sine wave). Other frequencies would be shifted by differing amounts.

Is this correct? Will I get phase problems?

I've tried to put this in context with a standard floor monitor and instrument amp setup. 1.15 miliseconds delay = ~0.5m delay. So for a typical musician with a monitor wedge and an amp they should get this effect too right, depending on the relative positioning of the floor monitor, the amp and the musician? Is it just that real life reflections, two ears and human ability that this generally isn't a problem for musicians?

Thanks in advance for your input.

An no we can't afford Roland's personal monitoring system!
Logged
Chris Shaw
Tech Team Lead
King's Community Church, Hatfield, England

g'bye, Dick Rees

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 7424
  • Duluth
Re: Effect of digital latency on phase in monitor solution
« Reply #1 on: July 26, 2011, 10:14:42 AM »


The quoted mixer/multicore system latency is ~375 microseconds, plus ~200 microseconds for the switch. If my maths is correct, this delay will mean ~870 Hz would effectively be phase shifted by 180 degrees on the monitor mix which would then be mixed with the original signal and subsequent cancellation would occur at that frequency (assuming same amplitude from both sources). (575 microseconds x2 = 1.15 miliseconds = period of a ~870 Hz sine wave). Other frequencies would be shifted by differing amounts.



Where and how is it re-mixed with the "original signal"?  Am I missing something?

Even if all your figuring is true (which it should be), delay, phase shift and cancellations are a natural phenomenon and our ears and brains are used to a certain amount.  Only if such cancellations are excessive in frequency/time will it be a problem.  If it were a problem, I doubt that any such systems would sell...... 

I wouldn't worry about it
Logged
Pay no attention to the man behind the curtain...

Matthias Heitzer

  • Newbie
  • *
  • Offline Offline
  • Posts: 39
Re: Effect of digital latency on phase in monitor solution
« Reply #2 on: July 26, 2011, 11:03:33 AM »

Generally, a combfilter (that's what you get if you mix two waves that are out of phase) is not very noticable as long as it doens't change.
Logged

Frank DeWitt

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1010
    • LBP DI Box
Re: Effect of digital latency on phase in monitor solution
« Reply #3 on: July 26, 2011, 10:02:33 PM »


The mixer/multicore system I have proposed is the Roland M-300 mixer with two Roland S-1608 stage boxes (split via a standard ethernet switch since one mixer REAC socket will be used for multitrack recording). Digital systems of course have latency, and this is where I foresee the potential problem...

An no we can't afford Roland's personal monitoring system!

I see why.  You are looking at about $10,000 for 32 ch
Logged
Not to Code

Tom Young

  • Global Moderator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 620
Re: Effect of digital latency on phase in monitor solution
« Reply #4 on: July 27, 2011, 07:48:47 AM »

I have recently proposed a new sound system for my church but suddenly realised that we could well run into problems with the proposed monitor setup. I just wanted to run this by the forum to check whether or not my concern is valid...

We typically have 6 to 10 musicisans/singers on stage for a normal worship set. Of course the ideal would be to provide each with their own monitor mix, however time, personell and equipment resourses wont allow for that. I wish to avoid providing more than 6 mixes from front of house (5 regular plus 1 spare). We're moving from floor wedges to IEM too.

The main method I have chosen to reduce the number of mixes was to use equipment like the Rolls PM351 (http://www.rolls.com/product.php?pid=PM351) for instruments like bass and keys. In theory this allows musicians to receive a more general monitor mix and then add a bit more of themselves to taste (I wouldn't necessarily use the mic in/thru part of it). For example I can send a drum and bass monitor mix and then the bassist can add a bit more bass themselves, and in the same way the keyboard player could share a mix with other musicians.

The mixer/multicore system I have proposed is the Roland M-300 mixer with two Roland S-1608 stage boxes (split via a standard ethernet switch since one mixer REAC socket will be used for multitrack recording). Digital systems of course have latency, and this is where I foresee the potential problem...

The quoted mixer/multicore system latency is ~375 microseconds, plus ~200 microseconds for the switch. If my maths is correct, this delay will mean ~870 Hz would effectively be phase shifted by 180 degrees on the monitor mix which would then be mixed with the original signal and subsequent cancellation would occur at that frequency (assuming same amplitude from both sources). (575 microseconds x2 = 1.15 miliseconds = period of a ~870 Hz sine wave). Other frequencies would be shifted by differing amounts.

Is this correct? Will I get phase problems?

I've tried to put this in context with a standard floor monitor and instrument amp setup. 1.15 miliseconds delay = ~0.5m delay. So for a typical musician with a monitor wedge and an amp they should get this effect too right, depending on the relative positioning of the floor monitor, the amp and the musician? Is it just that real life reflections, two ears and human ability that this generally isn't a problem for musicians?

Thanks in advance for your input.

An no we can't afford Roland's personal monitoring system!

A monitor wedge is usually 5'-6' away from the user's head. This represents far more delay and potential phase shift than the 1.15mS delay you reference.

The degree of phase shift (combfiltering) is also dependent on relative level. If the acoustic energy from the monitor at the user's ears is greater than the acoustic energy from the source (which it should be) the phase shift is less audible.
Logged
Tom Young
Electroacoustic Design Services
Oxford CT
203-888-6217

Chris Shaw

  • Newbie
  • *
  • Offline Offline
  • Posts: 19
Re: Effect of digital latency on phase in monitor solution
« Reply #5 on: July 27, 2011, 08:42:36 AM »

Thanks guys for all your responses. I have attached a block diagram which I hope explains the situation more clearly. The signal coming straight for the instrument via the split has no delay. The generic monitor mix signal coming from the mixer (which includes some of the instrument in the mix) has ~575 microsecond delay.

If it were a problem, I doubt that any such systems would sell...... 

I wouldn't worry about it

I forgot to add in my original post that this is not a problem in a normal analogue system, because electrons travel so fast and hence there is no signal delay. It's the use of such systems with a digital pa, which have inherent latency, that I am concerned about.

Generally, a combfilter (that's what you get if you mix two waves that are out of phase) is not very noticable as long as it doens't change.

When you say comb filtering is not noticible, are you referring to a speaker stack or are you referring to the electronic mixing of two signals? I agree that it is generally not noticible in the former case. I'll try and put a graphic together displaying the filter - what I do know is that the tooth centered at ~870 Hz (as I calculated previously) is very wide!

A monitor wedge is usually 5'-6' away from the user's head. This represents far more delay and potential phase shift than the 1.15mS delay you reference.

The degree of phase shift (combfiltering) is also dependent on relative level. If the acoustic energy from the monitor at the user's ears is greater than the acoustic energy from the source (which it should be) the phase shift is less audible.

I agree that this problem is present in the 3D envinroment of monitor speakers and amps and that we humans are able to cope with it.

However, the important difference in the case that I am querying is that the mixing of the two signals is electronic and is fed to the musician via headphones/earphones. There are no room reflections to create multiple phase paths and the signals are coming from the same place (the headphones), so the ear/brain cannot distinguish based on directionality.

With this in mind, is my concern valid???

Thanks again
« Last Edit: July 27, 2011, 09:03:38 AM by Chris Shaw »
Logged
Chris Shaw
Tech Team Lead
King's Community Church, Hatfield, England

TJ (Tom) Cornish

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 4317
  • St. Paul, MN
Re: Effect of digital latency on phase in monitor solution
« Reply #6 on: July 27, 2011, 08:47:35 AM »

Please resize the picture - it's huge!!!

Yes - your concern is valid - you can't use a "More-Me" situation like this with different path latencies - it sounds unusably terrible.  The modern way to handle this is to not have a "generic mix" that you add to with a more me box, but rather independent auxes for everyone.  This is generally not a big deal, since most digital boards have a pretty good number of aux sends.
Logged

Chris Shaw

  • Newbie
  • *
  • Offline Offline
  • Posts: 19
Re: Effect of digital latency on phase in monitor solution
« Reply #7 on: July 27, 2011, 09:07:26 AM »

Please resize the picture - it's huge!!!

Yes - your concern is valid - you can't use a "More-Me" situation like this with different path latencies - it sounds unusably terrible.  The modern way to handle this is to not have a "generic mix" that you add to with a more me box, but rather independent auxes for everyone.  This is generally not a big deal, since most digital boards have a pretty good number of aux sends.

Thanks Tom. Sounds like you've heard the effect first hand? I'll have a rethink. Will probably just have to use more mixes.

And I have just resized the picture! (sorry)
Logged
Chris Shaw
Tech Team Lead
King's Community Church, Hatfield, England

TJ (Tom) Cornish

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 4317
  • St. Paul, MN
Re: Effect of digital latency on phase in monitor solution
« Reply #8 on: July 27, 2011, 09:15:03 AM »

Please resize the picture - it's huge!!!

Yes - your concern is valid - you can't use a "More-Me" situation like this with different path latencies - it sounds unusably terrible.  The modern way to handle this is to not have a "generic mix" that you add to with a more me box, but rather independent auxes for everyone.  This is generally not a big deal, since most digital boards have a pretty good number of aux sends.

Thanks Tom. Sounds like you've heard the effect first hand? I'll have a rethink. Will probably just have to use more mixes.

And I have just resized the picture! (sorry)
I have indeed.  Your band will appreciate more control anyway!

We have a couple ways of monitoring at our church - the front row of the most mobile people have wired ears run off auxes from FOH.  The backline use ears powered by wired headphone amps, again from FOH auxes.  The bass rig has a "more me" setup, but this is done a little differently - there is a mixer/headphone amp in a little rack at the bass position, but the system feed is still a dedicated aux.  This works for us, because we make sure to not feed the bass signal back down the bass aux, thus avoiding multiple paths.

It's not as convenient to have wires, and we hope to go all wireless eventually, but it's a middle step that gets us going in the interim.

Logged

Matthias Heitzer

  • Newbie
  • *
  • Offline Offline
  • Posts: 39
Re: Effect of digital latency on phase in monitor solution
« Reply #9 on: July 27, 2011, 12:41:58 PM »

Quote
Quote
Generally, a combfilter (that's what you get if you mix two waves that are out of phase) is not very noticable as long as it doens't change.

When you say comb filtering is not noticible, are you referring to a speaker stack or are you referring to the electronic mixing of two signals? I agree that it is generally not noticible in the former case. I'll try and put a graphic together displaying the filter - what I do know is that the tooth centered at ~870 Hz (as I calculated previously) is very wide!

without an A/B check, it's hard to say wheter there is a combfilter or not.
If you combine the two bass signals in a headphone mixer, you will hear a big difference between between the single signals and the combined one. As long as there is just a constant delay, this may or may not be a problem for your bass player. It justs adds additional "eq" to the signal.
It would drive a vocalist crazy, but for bass it could work.

If the EQ in the mixer also affects the monitormix, there is no way to predict how the two signals combine.
If you need a cheap solution for "more me", you could send the two signals to different ears, most humans have two of them.

Another problem for the bass player will propably be IEM itself, most of them need to feel what they're playing.

IEM's not cheap, and the earbuds are the cheapest part of it. Best would be a dedicated monitorconsole and operator.
They gonna need almost everything on iem, including the congregation.

Is there a real need to use IEM? Unless stage volume is a really big problem, the musicians will propably be happier with 5-6 wedge mixes.
 

Logged

ProSoundWeb Community

Re: Effect of digital latency on phase in monitor solution
« Reply #9 on: July 27, 2011, 12:41:58 PM »


Pages: [1] 2  All   Go Up
 



Site Hosted By Ashdown Technologies, Inc.

Page created in 0.04 seconds with 19 queries.