ProSoundWeb Community

Sound Reinforcement - Forums for Live Sound Professionals - Your Displayed Name Must Be Your Real Full Name To Post In The Live Sound Forums => LAB Lounge => Topic started by: Scott Middleton on April 04, 2011, 03:46:02 PM

Title: Am I on the right track here? (time alignment)
Post by: Scott Middleton on April 04, 2011, 03:46:02 PM
This is a "bar PA" with (2x) JBL sr4733's hung, and (2x) 1x18 subs (JBL drivers, unknown boxes) ground stacked in the center, stage lip. 

So in trying some experiments the other night (no existing time alignment delay at all), I ballparked the subs being 5-7 feet in front of the mains, and so added 5 ms delay to the subs via the Crown XTI DSP.  This is while the band is playing BTW. 

Great, it sound better the low end is tightened up a bit!  So then I think, well maybe I can tweak it just a bit more by delaying the whole PA back to the drum kit.  Looks like the mains are roughly 10 ft ahead of the where the drum kit is usually set up, so I then add 10 ms to the main out on my 01V(and also the aux output that is driving the subs).  Now it sounds even better.  Drums are clearer and the low end is better still.

Granted, this is all guestimation (non scientific I know), but am I on the right track here?  Are there disadvantages to delaying the PA back to the drumkit?
Title: Re: Am I on the right track here? (time alignment)
Post by: Tim Weaver on April 04, 2011, 05:16:30 PM
am I on the right track here? 

Yes


Quote
Are there disadvantages to delaying the PA back to the drumkit?

Not until you get to really big stages where the mains delay wouldn't matter much anyway. 10-15 ms of delay on the mains won't hurt anything.
Title: Re: Am I on the right track here? (time alignment)
Post by: Mac Kerr on April 04, 2011, 05:24:18 PM
Not until you get to really big stages where the mains delay wouldn't matter much anyway. 10-15 ms of delay on the mains won't hurt anything.

Unless the band can hear the PA clearly, in which case you have just added 10-15ms of additional delay to the time of flight delay they already hear from the PA, perhaps making it appear to be as much as 30ms behind what they are playing.

Mac
Title: Re: Am I on the right track here? (time alignment)
Post by: Bennett Prescott on April 04, 2011, 05:50:27 PM
Scott,

I would also fiddle with that sub->top delay by ear, it is not as simple as physical distance.
Title: Re: Am I on the right track here? (time alignment)
Post by: Scott Middleton on April 04, 2011, 06:03:58 PM
Thanks guys. 

Will do Bennett.  I'll have the laptop w/ me next time and I can dial in to the XTI with SA and tweak to taste.
Title: Re: Am I on the right track here? (time alignment)
Post by: Marsellus Fariss on April 04, 2011, 06:23:08 PM
I agree with Bennett. I start by sending a sine wave at the crossover frequency through the rig with the phase of the sub (or the tops) inverted then tweak the delay time of the subs back and forth till you get to the point of most cancellation and the apparent volume is at it's lowest point. Then flip the phase back to normal and season to taste by ear.

I also always delay my rig to the back line unless it's going to cause latency problems like those described. Stand on stage yourself and talk in a vocal and see if you can hear any distracting latency. In my room it works out amazingly well. If your doing a show where the muso's are quiet and down stage you can kill the delay if you want.

And also consider that your DSP and console have some small amount of latency on their own. It's probably not enough to cause problems but if you've got other AD/DA in your chain these things can add up. Do a search on this and keep it in mind when programing.

Also I believe the standard feet to milliseconds multiplier is 1.125 milliseconds per foot.


One other thing to keep in mind is that your only time aligning the sources to people who are out front of the rig and the stage. People in other areas will still have arrival time issues. But that's life.   
Title: Re: Am I on the right track here? (time alignment)
Post by: Scott Middleton on April 04, 2011, 10:06:44 PM
Unfortunately I can't do the polarity flip trick until I can get in there w/ no patrons.  Its a bar and grill, so this is rare.  I hope to get this done eventually.

I'll keep latency of other gear in mind.

Yep, I was guestimating 1ms = 1ft.

If I can get the "dancefloor" area aligned and sounding good, that's my big concern.  Other areas, not so much.

This is down and dirty for sure.  Any improvement I can get is better than how it sounds now.
Title: Re: Am I on the right track here? (time alignment)
Post by: Brad Weber on April 04, 2011, 10:47:02 PM
If I can get the "dancefloor" area aligned and sounding good, that's my big concern.  Other areas, not so much.
Achieving alignment over an area can be difficult.  For example, on the centerline of the stage the subs may be around 5' in front of the mains but to listeners out to either side of the edge of the mains coverage the subs may be more than 5' ahead ahead of the more distant main but behind the closer main.
Title: Re: Am I on the right track here? (time alignment)
Post by: Dan Johnson on April 05, 2011, 12:05:59 AM
Also I believe the standard feet to milliseconds multiplier is 1.125 milliseconds per foot.
I think you have this backwards.  It would be 1.125 feet per millisecond which is equivalent to 0.889 milliseconds per foot.

If you want to be really picky, there are other factors that affect the speed of sound such as altitude and temperature but these numbers are good enough for estimating delay times.
Title: Re: Am I on the right track here? (time alignment)
Post by: luis Markson on April 05, 2011, 03:39:28 AM
In the case of a system like a Mackie S410s with S408 pole mounted on top, is there anything to be gained by experimenting with sub -> top delay?

I am assuming that Mackie would have considered this during manufacture? Or are unpowered speakers at this level often neglected in this area?

How would you go about listening for correct alignment?

I like the sound of using a pure sine wave, flipping the phase of either the top or sub and then listening for maximum cancellation. I am assuming that you would use a tone at the crossover frequency? My understanding of crossover types and their implications is sorely lacking though so I'm guessing there is going to be more to it.



Title: Re: Am I on the right track here? (time alignment)
Post by: Bennett Prescott on April 05, 2011, 03:59:41 AM
I start by sending a sine wave at the crossover frequency through the rig with the phase of the sub (or the tops) inverted then tweak the delay time of the subs back and forth till you get to the point of most cancellation and the apparent volume is at it's lowest point. Then flip the phase back to normal and season to taste by ear.

That technique works, of course, but the problem with it is... what is the crossover frequency?

This is a lot easier to do if you have made the phase slope of your tops and subs equal for an octave or so around what you want to be the crossover point. This usually means adding a few orders of XO to the sub LPF, or subtracting a few from the top HPF. My preference is to subtract from the tops HPF, which usually means a 2nd order filter. I find anything more than 4th order sounds like dog shit, which really means you're dealing with 6th-8th order acoustically if you're crossing near the tuning frequency of the box or the natural rolloff of a driver.
Title: Re: Am I on the right track here? (time alignment)
Post by: luis Markson on April 05, 2011, 04:07:40 AM
That technique works, of course, but the problem with it is... what is the crossover frequency?

This is a lot easier to do if you have made the phase slope of your tops and subs equal for an octave or so around what you want to be the crossover point. This usually means adding a few orders of XO to the sub LPF, or subtracting a few from the top HPF. My preference is to subtract from the tops HPF, which usually means a 2nd order filter. I find anything more than 4th order sounds like dog shit, which really means you're dealing with 6th-8th order acoustically if you're crossing near the tuning frequency of the box or the natural rolloff of a driver.

Rather than hassle you for a beginners lesson on XO, so I can understand your above post, could you refer me to a good resource on XO fundamentals...
Title: Re: Am I on the right track here? (time alignment)
Post by: Marsellus Fariss on April 05, 2011, 04:09:48 AM
In the case of a system like a Mackie S410s with S408 pole mounted on top, is there anything to be gain by experimenting with sub -> top delay?

I am assuming that Mackie would have considered this during manufacture? Or are unpowered speakers at this level often neglected in tis area?

Doubtful. If they picked a distance to allign the two they'd be wrong in a lot of other user scenarios. But having the tops right over the subs is a good scenario. Things will be pretty close depending on your distance to the rig but you can experiment with your DSP. I find it a great lesson in physics if you can delay the rig to a rather loud backline in real time and hear how it pops together while the shows going on.
Title: Re: Am I on the right track here? (time alignment)
Post by: luis Markson on April 05, 2011, 04:16:33 AM
Post deleted because it was unnecessary waffle
Title: Re: Am I on the right track here? (time alignment)
Post by: Marsellus Fariss on April 05, 2011, 04:38:57 AM
I think you have this backwards.  It would be 1.125 feet per millisecond which is equivalent to 0.889 milliseconds per foot.

If you want to be really picky, there are other factors that affect the speed of sound such as altitude and temperature but these numbers are good enough for estimating delay times.

Yea, thanks for correcting me on that. Your right.

Even though things get cloudy in the realm of time arrivals often enough if you get close with the math you'll hear a noticeable improvement under the right conditions. I think the more the backline volume effects your mix (like a small venue) the more you'll hear the difference. A few times I've forgotten to reset my delay to backline from the night before and have done it mid show and went WOW! Make sure your DSP or console can do this w/o audio dropouts though. 
Title: Re: Am I on the right track here? (time alignment)
Post by: Russel Murton on April 05, 2011, 11:18:29 AM
Yea, thanks for correcting me on that. Your right.

Even though things get cloudy in the realm of time arrivals often enough if you get close with the math you'll hear a noticeable improvement under the right conditions. I think the more the backline volume effects your mix (like a small venue) the more you'll hear the difference. A few times I've forgotten to reset my delay to backline from the night before and have done it mid show and went WOW! Make sure your DSP or console can do this w/o audio dropouts though.

I played with time aligning my QSC KW rig to the backline but I came into problems with the bleed from the monitors near the front of the crowd, also whilst the rest of the band's definition increased in clarity, the vocals started to cancel with the monitor bleed. Then when I had a band which didn't have a drumkit and lots of line level the cancellations between the monitors and the FOH were even more startling, resulting in it sounding a little off timing wise.

In the end I now time align the subs to the tops and leave the tops at 0msec as the vocal clarity is my main concern. I'd rather use a Radial Phazer 4 channel rack to align the guitars, bass DI and snare bottom than worry about aligning to the backline. It might be worth it in a very big room or outdoors, but when you are in a norrow room the side reflections mess with things a lot.
Title: Re: Am I on the right track here? (time alignment)
Post by: Bennett Prescott on April 05, 2011, 03:02:10 PM
Rather than hassle you for a beginners lesson on XO, so I can understand your above post, could you refer me to a good resource on XO fundamentals...
Luis,

Here is a fantastic resource, math heavy but if you're good at that (I am not) it can help you a lot. This guy is the L in the Linkwitz-Riley crossover filter set.

http://www.linkwitzlab.com/crossovers.htm
Title: Re: Am I on the right track here? (time alignment)
Post by: Scott Middleton on April 06, 2011, 08:12:51 PM
I played with time aligning my QSC KW rig to the backline but I came into problems with the bleed from the monitors near the front of the crowd, also whilst the rest of the band's definition increased in clarity, the vocals started to cancel with the monitor bleed.

Try flipping the polarity on the wedges.  I had to do this the other day for a gig.
Title: Re: Am I on the right track here? (time alignment)
Post by: Todd Huffines on April 07, 2011, 09:53:16 PM
Also I believe the standard feet to milliseconds multiplier is 1.125 milliseconds per foot.
I think you have this backwards.  It would be 1.125 feet per millisecond which is equivalent to 0.889 milliseconds per foot.

If you want to be really picky, there are other factors that affect the speed of sound such as altitude and temperature but these numbers are good enough for estimating delay times.

   I thought that sound traveled at 1130 or so feet per second.  That speed is constant.  Humidity has an affect on the thickness of the medium "air" which has a big deal with how far the sound carries from the source.  Since sound will not travel in a vacuum then the thicker the air the better sound moves. 

This thread has struck me as one of the most interesting for me.  When there is plenty of stage bleed, I am gathering that a little delay will help.  I am going to play with my dr260 on this.  Should be easy to do at the next soundcheck.   I am loving this!
Title: Re: Am I on the right track here? (time alignment)
Post by: Rob Spence on April 08, 2011, 12:51:50 AM
Also I believe the standard feet to milliseconds multiplier is 1.125 milliseconds per foot.
I think you have this backwards.  It would be 1.125 feet per millisecond which is equivalent to 0.889 milliseconds per foot.

If you want to be really picky, there are other factors that affect the speed of sound such as altitude and temperature but these numbers are good enough for estimating delay times.

   I thought that sound traveled at 1130 or so feet per second.  That speed is constant.  Humidity has an affect on the thickness of the medium "air" which has a big deal with how far the sound carries from the source.  Since sound will not travel in a vacuum then the thicker the air the better sound moves. 

This thread has struck me as one of the most interesting for me.  When there is plenty of stage bleed, I am gathering that a little delay will help.  I am going to play with my dr260 on this.  Should be easy to do at the next soundcheck.   I am loving this!
Sound travels at different speeds through air at differing humidities and temperatures.
In dry air at 20 °C (68 °F), the speed of sound is 343.2 metres per second (1,126 ft/s).
Title: Re: Am I on the right track here? (time alignment)
Post by: Nils Erickson on April 08, 2011, 03:34:00 AM
I think you have this backwards.  It would be 1.125 feet per millisecond which is equivalent to 0.889 milliseconds per foot.

If you want to be really picky, there are other factors that affect the speed of sound such as altitude and temperature but these numbers are good enough for estimating delay times.


   I thought that sound traveled at 1130 or so feet per second.  That speed is constant.  Humidity has an affect on the thickness of the medium "air" which has a big deal with how far the sound carries from the source.  Since sound will not travel in a vacuum then the thicker the air the better sound moves. 

This thread has struck me as one of the most interesting for me.  When there is plenty of stage bleed, I am gathering that a little delay will help.  I am going to play with my dr260 on this.  Should be easy to do at the next soundcheck.   I am loving this!

Sound travels at different speeds through air at differing humidities and temperatures.
In dry air at 20 °C (68 °F), the speed of sound is 343.2 metres per second (1,126 ft/s).

While I find this stuff really interesting to geek out on myself, I had to chuckle as I envisioned the BE adjusting the mains delay from 5 to 6 milliseconds as the humidity in the club raised due to more fans arriving and dancing.  ;D

Rather than doing those calculations then, I find it more productive to focus my brain on what is happening on the stage and making the music sound like what we are witnessing.  No sense in aligning the tops mid show if you have missed the fact that the keyboard player's solo is too quiet.

Not to discount that other crucial set up stuff, it really does help... Anyway, great info, thanks for some fun reading folks!

Cheers,
Nils
Title: Re: Am I on the right track here? (time alignment)
Post by: luis Markson on April 15, 2011, 07:37:51 PM
I agree with Bennett. I start by sending a sine wave at the crossover frequency through the rig with the phase of the sub (or the tops) inverted then tweak the delay time of the subs back and forth till you get to the point of most cancellation and the apparent volume is at it's lowest point. Then flip the phase back to normal and season to taste by ear.

I am going to give this a go tomorrow. Can anyone direct me to a 150Hz.wav?
Title: Re: Am I on the right track here? (time alignment)
Post by: Bennett Prescott on April 15, 2011, 08:27:43 PM
I am going to give this a go tomorrow. Can anyone direct me to a 150Hz.wav?

http://www.binkster.net/extras.shtml#cd
Title: Re: Am I on the right track here? (time alignment)
Post by: Dan Johnson on April 15, 2011, 08:29:37 PM
I am going to give this a go tomorrow. Can anyone direct me to a 150Hz.wav?
This link doesn't have exactly 150Hz but a lot of frequency tones as well as some other useful tools.  Scroll down to the test CD.

http://www.binkster.net/extras.shtml

Edit: Bennett beat me to it.
Title: Re: Am I on the right track here? (time alignment)
Post by: luis Markson on April 15, 2011, 08:33:57 PM
This link doesn't have exactly 150Hz but a lot of frequency tones as well as some other useful tools.  Scroll down to the test CD.

http://www.binkster.net/extras.shtml

Edit: Bennett beat me to it.

Already got that one..... As you say, no 150Hz
Title: Re: Am I on the right track here? (time alignment)
Post by: Bennett Prescott on April 15, 2011, 08:55:03 PM
Already got that one..... As you say, no 150Hz

In this case, close enough is close enough. I think I said earlier that you probably don't know what your actual crossover is anyway, so pick a frequency that matters to you and align it there.
Title: Re: Am I on the right track here? (time alignment)
Post by: luis Markson on April 15, 2011, 09:29:13 PM
In this case, close enough is close enough. I think I said earlier that you probably don't know what your actual crossover is anyway, so pick a frequency that matters to you and align it there.

Sorry Bennet, but could you explain why (if 150Hz is chosen as the Xo in DSP) that the actual XO frequency is unknown?
Title: Re: Am I on the right track here? (time alignment)
Post by: Brad Weber on April 16, 2011, 08:08:59 AM
Sorry Bennet, but could you explain why (if 150Hz is chosen as the Xo in DSP) that the actual XO frequency is unknown?
That crossover frequency is defining the corner frequency for the crossover filters and not at what frequency the acoustic signal from the two drivers is equal.  However, difference in levels between the high frequency and low frequencies portions of the signal also affects the frequency at which the level of the two signals is equal and thus the effective crossover point.  A simple example is that if everything else is the same but your speaker has a HF driver with greater sensitivity, then the HF acoustic output would be greater relative to the LF output and thus where the HF output level equals the LF signal in sound pressure level will shift down in frequency.
Title: Re: Am I on the right track here? (time alignment)
Post by: Dalton Brand on April 17, 2011, 07:32:51 PM
In the past I always added delay to compensate for the back line but recently I've run into issues with the increasing number of bands with IEMs. In a few of the brick lined rooms the drummer can end up in a weird zone of reflection from back of room back to the stage that would otherwise be less apparent were it not for the extra 10ft or so of delay plus digital board latency. Now I ask specifically during sound checks if there are issues. I can live without system delay if it means a happier rhythm section.