Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Ryan O John]

So til now I've been tuning systems by ear, and for the most part I end up with a reasonably good sounding rig, and have a good show...  I've also never used any tools on the system outside of levels of processor outputs (Sub, LF, LMF, HMF, HF, etc), Graphic EQ, and time alignment between subs and tops. 

In the past few months, I've learned all I can in the time I've had about the technical side of system tuning, be it from the system tech on the tour, friends that know a lot, manuals, or articles...  I'm at the point that now I'm trying to integrate technical readings into my system tuning on a daily basis. 


Right now I'm on a club tour, a lot of the rooms we see on this tour are pretty rough, and we're not carrying a console, so sometimes my only options for system control is simply a (or a few) 31 band graphic, and leveling sub/mains, or sometimes re aiming or unplugging boxes...


Today I did a show in a club with a ground stacked array of 650 boxes, and took some screenshots throughout. 
Measured from a DBX RTA-m up against (not touching, but under 1mm) a large flat surface, and a mono Matrix output.  The microphone was placed in the position at which the delay from the left and the right sides of the rig were exactly the same (turns out it was pretty easy to do that...).  All of these are with music playing as opposed to pink, the first being a reference track, everything after being a band...

The first of these is after/during tuning by ear, the room rang a bit around 350 or so...  The bump at 10-12k made the system sound strange when cut enough to 'flatten that out,' sounded fine with it in... (maybe I'm losing my mind)

Second is during the opener.

Third is during the my set (some silent moment), at which point the room was pretty packed.



Show sounded good, and got plenty of compliments, so even though the graph looks a bit absurd to me, apparently it worked out.  The phase trace looks absurd to me, actually really, none of it LOOKS good to me, I really don't know enough about it all yet to really understand what I should consider acceptable, as I'm not hanging my own rig, nor carrying much processing power...

[Timo Beckman]

So first : it took me a couple off years to get used to the phase display and to make sense out of it so keep practicing .....and reading specialy the book by Bob mcCarthy about system design and optimalization .

Always remember phase is time over frequency . Flat phase would mean all frequencies get to the mic at the same time (so a flat line (wishfull thinking)) . If your phase trace goes down @ some point it means these frequencies are late/delayed . If your phase trace has a up worth angle @ some point the frequencies at that point are ahead in time compared to the point of synchronization of your analyzer .
If you're running multiple systems (like a L/R/C/S system with delay's) with all channels on your phase display becomes a bit callengeing to look at ...... So get a onax trace of just 1 channel and start comparing it with the rest of the system at multiple positions and combinations .

You might take a look @ the blog of 606
http://bobmccarthy.wordpress.com/
a lot about measurements with a FFT analyzer but there's a lot more in his book .
or take a look at my blog also on wordpress .
 

[Ryan O John]

So first : it took me a couple off years to get used to the phase display and to make sense out of it so keep practicing .....and reading specialy the book by Bob mcCarthy about system design and optimalization .

Always remember phase is time over frequency . Flat phase would mean all frequencies get to the mic at the same time (so a flat line (wishfull thinking)) . If your phase trace goes down @ some point it means these frequencies are late/delayed . If your phase trace has a up worth angle @ some point the frequencies at that point are ahead in time compared to the point of synchronization of your analyzer .
If you're running multiple systems (like a L/R/C/S system with delay's) with all channels on your phase display becomes a bit callengeing to look at ...... So get a onax trace of just 1 channel and start comparing it with the rest of the system at multiple positions and combinations .

You might take a look @ the blog of 606
http://bobmccarthy.wordpress.com/
a lot about measurements with a FFT analyzer but there's a lot more in his book .
or take a look at my blog also on wordpress .

I'll definitely take a look at that blog, and see if I can pick up that book again (I read parts of it years ago, and didn't understand most of it then...), what's the link to your blog?

From the great sounding systems I've mixed on, the phase trace has looked like something of an inverse/downward logarithmic curve, but wrapped of course.  So I always figured some goal near there would be ideal, but some of the places we've been in on this run, I would be excited if everything worked, let alone got a good phase trace.

...and by the way, I realize a TF mid show is for the most part useless...

[Timo Beckman]

http://timobeckmangeluid.wordpress.com/

have fun . it gets interesting from the meyersound system design/optimalization level2 @ mallorca posts . Before that i did not post to much in english (i'm from the netherlands so posting in english is a bit challenging sometimes when things get technical)

[Doug Fowler]

From the great sounding systems I've mixed on, the phase trace has looked like something of an inverse/downward logarithmic curve, but wrapped of course.  So I always figured some goal near there would be ideal, but some of the places we've been in on this run, I would be excited if everything worked, let alone got a good phase trace.

Ryan -

Don't get too 'wrapped' up in the overall phase response of a system. Aside from troubleshooting (polarity problems), it's not that useful. One could make a case for examining individual component timing, but chances are you won't get to adjust that anyway.

In fact, you could flatten the "phase response" of a smallish range of frequencies merely adjusting the receive delay. So it begs the question "what is the correct delay offset to use?".  Since, as Timo pointed out, different frequencies are arriving at different times, there is no correct answer.

Nice phase response displays on those great sounding systems you mixed on are indicative of well engineered systems.  Today, with advanced DSP filters, it's possible to get nearly flat phase response from midrange up.   

Or, the box design could be so good it needs little or no correction.  I suppose the bottom line is "don't get too hung up on overall system phase response". You can't do anything about it, normally. I see Timo is doing some stuff with allpass filters, this might be an exception.

[Timo Beckman]

Or, the box design could be so good it needs little or no correction.  I suppose the bottom line is "don't get too hung up on overall system phase response". You can't do anything about it, normally. I see Timo is doing some stuff with allpass filters, this might be an exception.

A lot also depends on the time you get on the job for tuning the system . If it's 10 to 20 min Make the most of it within that time frame . If it's 2 or 3 day's/nights do all that's needed to get the best result possible but never loose sight of the bigger/overall picture .
A lot can be improved by just the re-aiming of speakers or re-angling an array and most important walk the room and listen instead of putting in a bunch of mic's and blast away with pink noize and looking at a screen for a couple of hours .

[Dan Mortensen]

The microphone was placed in the position at which the delay from the left and the right sides of the rig were exactly the same (turns out it was pretty easy to do that...). 

Hi Ryan,

I'm definitely a novice at this, too, but one thing I think I've learned so far is that our ears are more forgiving about hearing multiple sources and ignoring (if not enjoying) the interference than is an analysis microphone.

As Timo says, be sure to only listen to one speaker source at a time, even to the point of turning off all but one box in an array, then adding array elements to see how the response changes as the boxes interact either helpfully or destructively. Otherwise you really don't know what you are listening to (which, in your case, is whatever PA configuration is left for you by all the previous operators).

Again, as Timo says, time available is the limiting factor, and the 10 or 20 min that we are often lucky to get is not enough to do a full component analysis. It is time enough to turn off one side and get a reading and then compare it to the other side, usually.

Good luck,
Dan

[Timo Beckman]

I see Timo is doing some stuff with allpass filters, this might be an exception.

The use of all-pass filters allows you to match different systems regarding the phase response but this is something that will take a lot of time and you really need to know what you're doing .
Also a lot of work i do matching the phase response of different systems is done at the company shop that builds the speaker systems so when i have to tune a room the phase response allready matches between the systems .
I'm aiming for a phase difference between systems of a maximum of 120° between all cabinets used 90° = even better . If i exceed the 120° there will be no addition but reduction and comb filtering between them .
Still tuning a room can get difficult but with matched systems you can concentrate more at re-angling/positioning/delaying and the overall job instead of trying to match the phase response of the systems to eachother . 
(let me know when you're up for a broodje haring if you coming to the netherlands)

[Dan Richardson]

The phase trace looks absurd to me, actually really, none of it LOOKS good to me

I'm not as far as looking at the phase trace. Is that coherence trace acceptable?

[Doug Fowler]

(let me know when you're up for a broodje haring if you coming to the netherlands)

Twee broodjes haring, met ui en augurk, alstublief.

;-)

[Timo Beckman]

Twee broodjes haring, met ui en augurk, alstublief.

;-)

In real dutch it should be Alstublieft
But if you were in Utrecht they'll understand it anayway .

[Dan Richardson]

I'm not as far as looking at the phase trace. Is that coherence trace acceptable?

Still wondering, from those as actually know what they're looking at.

[Jay Barracato]

Still wondering, from those as actually know what they're looking at.

Dan,

I think of coherence frequency by frequency, rather than looking at the coherence trace as a whole. If there is a change in the response at a frequency that is accompanied by a corresponding dip in the coherence, that suggests that the response measurement is being contaminated by some factor of the measurement process such as mic placement, primary reflection, employee with vacuum, etc.

For me, I seem to look at the section of the trace that has good coherence that helps me answer a particular question. I am not producing traces for publication, or review. A common problem is to have bad coherence in the upper HF region, but if what I am working on is aligning the subs, that doesn't really matter.

Another factor that I have not seen mentioned, because possibly people who own their own rig don't get to work on a variety of systems, is to make sure you get good measurements of rigs that you like what you hear. When I first started with a measurement system, I got to measure a couple of nice systems that were great sounded and properly aligned by someone else. Matching what you hear to what you see is a great calibration tool.

[Ryan O John]

Another factor that I have not seen mentioned, because possibly people who own their own rig don't get to work on a variety of systems, is to make sure you get good measurements of rigs that you like what you hear. When I first started with a measurement system, I got to measure a couple of nice systems that were great sounded and properly aligned by someone else. Matching what you hear to what you see is a great calibration tool.

That's a great idea!

[ProSoundWeb Community]