Rational Acoustics



Arthur Skudra
June 17th, 2008, 02:36 PM
While I'm stuck home sick and unable to attend Infocomm :( , and sticking my head over a portable medicinal mister, I thought it might be good to get some discussion stirred up about how the effects of temperature and humidity have impact on what you do to optimize a sound system. What strategies to you employ, if any? Outdoor venues certainly present some real challenges, but how about indoor venues where you might optimize a system in the summer when things are relatively humid. Will winter make a bunch of difference, particularly in northern climates, when the relative humidity generally takes a dive?

For a good read on the subject, I highly recommend this excellent article written by Dennis Bohn of Rane:
http://www.rane.com/pdf/ranenotes/Enviromental%20Effects%20on%20the%20Speed%20of%20S ound.pdf

Interesting when the Omnidrive first came out they made a big deal of the meteorology probe, but for some reason it never really caught on with the rest of the processor world. However with the recent release of the UX8800, you needed to enter the temperature and humidity in manually to get optimal results. For the years in between, what did you do?

Would love to hear your insights for both indoors and outdoors. Excuse me while I go cough up more crap from my lungs!

Harry Brill Jr.
June 17th, 2008, 03:15 PM
Some DSPs have the capability to enter temp and humidity with the additional capability to enter in distance for HF correction. It can work well. It's not predictable by intuition.

When I use a zoned PA with one of these DSPs, I will change the atmospheric settings from setup through show as needed. If you wait several hours without making a change, the change is very noticeable (an improvement). I usually check it about every hour through the day and of course right after the show begins.

During system optimization, be sure the HVAC is set for human comfort. That's how they will have it set when people are there. Some day I'd like to optimize a PA with a full house.

Jens Brewer
June 19th, 2008, 10:37 PM
Some day I'd like to optimize the system with a full house as long as they are reasonably quiet.

That reminds me of a comment/insult a lighting buddy threw my way awhile ago. After a particularly long and noisy measurement session (while he was trying to focus conventionals) he mentioned how much faster I used to be when I just did, 'check 1, check 2'. A day or two later during show, while I'm hunched over Smaart looking at a PFL'd presenter in Spectrograph, he commented, 'Are you going to be done tweaking by load-out?' For some of us, the show is just a better dressed soundcheck. :p

I'm waiting for the day when I can phone it in over a VPN, a few webcams, and a robot mic handler.

Harry Brill Jr.
June 20th, 2008, 11:37 AM
Someone asked me recently how much time I put into optimizing a system on show site. I said as much as they give me. In reality though it's a time saver.

We should talk about the circumstances of your noisy measurement. You know, you only need it loud enough to push the coherence to the point where it doesn't get any higher.

Calvert Dayton
June 21st, 2008, 12:18 PM
That makes sense. Dave Gunness made a study of air loss while back and found you really can't get air loss compensation right without knowing the relative humidity. Temperature turns out to be a factor in air loss mostly because of its effect on the capacity of air to hold water.

But temperature does have a pretty dramatic effect on the speed of sound of course, and I'm pretty sure the original Omnidrive had a provision for a temperature probe. For sure the later models did. One would think that this might be a popular feature amongst folks who do a lot of big outdoor events, since a change of just a few degrees can really muck with delay systems.

One other interesting aspect of the great outdoors is having no boundaries to speak of and fewer things in genedral for sound to bounce off, so ambient noise tends to be more of a problem than reverberation. I know that some people use cardioid mic's when measuring outdoors for that reason.

Oh and as regards full house measurements, David Grissinger came up with a pretty slick method for that. What he would do is station people around the room with mic's and tape recorders and then play a series of three sweeps and record them. So he only needed a total of about 10 seconds of quiet, then he could convolve the recordings with the original signal and analyze the transfer function and impulse response from each recording position at his leisure. You can do that easily with acoustic tools.

Arthur Skudra
June 21st, 2008, 12:58 PM
One other interesting aspect of the great outdoors is having no boundaries to speak of and fewer things in genedral for sound to bounce off, so ambient noise tends to be more of a problem than reverberation. I know that some people use cardioid mic's when measuring outdoors for that reason.Interesting you mention that, I always thought the use of cardioid mics was more applicable for indoor measurements where you want to try to minimize the effect of a certain reflection in the measurement. Wouldn't wind noise be a detrimental factor in using a cardioid test mic outdoors?

Arthur Skudra
June 21st, 2008, 01:01 PM
That reminds me of a comment/insult a lighting buddy threw my way awhile ago. After a particularly long and noisy measurement session (while he was trying to focus conventionals) he mentioned how much faster I used to be when I just did, 'check 1, check 2'. A day or two later during show, while I'm hunched over Smaart looking at a PFL'd presenter in Spectrograph, he commented, 'Are you going to be done tweaking by load-out?' For some of us, the show is just a better dressed soundcheck. :p

I'm waiting for the day when I can phone it in over a VPN, a few webcams, and a robot mic handler.Funny how the squints need hours of setup time, yet in many circumstances I've been involved in it was asking too much to even have an hour of system optimization time, even during lunch break! :rolleyes:

Calvert Dayton
June 23rd, 2008, 11:27 AM
Shhhh! Don't say things like that in front of the children.

Let's back up a little. In general of course, the main reason for using omnidirectional microphones indoors is that the reverberant characteristics of a space, in all their reflectiveness, are a very significant factor in how a given system will sound in a given room. So the last thing anyone normally wants to do is block out half the reverberant field and proportionally overemphasize the direct sound from the system in their measurements.

When there is some issue with, for example, a discrete reflection causing a comb filter in the analyzer trace that would not actually be an audible artifact for a human listener (since our ears tend to integrate short reflections with the direct sound), the preferred response is therefore to deal directly with that reflection if possible -- also saves you the trouble of carrying multiple microphones.

Some common strategies include:


moving the microphone out of the path of the reflection if possible,
placing some obstruction such as a case lid or a piece of insulating material in the path of the bounce,
placing the microphone very close to the surface causing the problem so that the reflection becomes so short that the resulting comb filter is moved out of the audible spectrum

Of course the world is large and lots of crazy things can happen and I suppose there might arise cases here and there, where one might want to break down and resort to using a cardioid indoors if they happened to have one handy. But I haven't heard of many people who routinely carry one for that purpose.

Outdoors of course, without significant boundary effects or reverberant fields to deal with in most cases, it's a much different set of problems. One is higher levels of ambient noise in many cases – one of the downsides of not having walls around you – and that's really what I was talking about.

Wind of course is a problem for any microphone. I've heard that it's more of a problem for cardioids than omnis in general because of their velocity sensing (vs. pressure sensing?) nature but I have only the vaguest notion of what that means. I can also only guess at what the practical implications might actually be in our case. For example I have an idea that the problem might be more pronounced in hypercardioid and shotgun mic's than in mic's with less radical directionallty. It's also possible that a certain amount of wind noise might tend to be somewhat less of a problem for measurement than for recording, since you might be able to knock it down significantly with enough averaging. And of course there are such things as wind screens.

I guess if someone did enough of that kind of work, in might even pay to break down and invest in a microphone blimp (http://www.imagewest.tv/index.asp?PageAction=VIEWPROD&ProdID=59).

Or if you needed a good rainy day project you might try making your own out of a shock mount...

http://www.microfilmmaker.com/tipstrick/Issue8/shotwind.html

or a bird feeder! :eek:

http://www.instructables.com/id/Microphone-Blimp/
http://www.metacafe.com/watch/963275/diy_microphone_blimp/

Harry Brill Jr.
June 27th, 2008, 03:19 PM
Excerpted from Sound Systems: Design and Optimization by Bob McCarthy

Cardioid Microphones

There are experienced users in the field of system optimization that use multiple cardioid microphones for testing. There are several important considerations worth bearing in mind when substituting cardioid mics for this application.


Omni vs. cardioid measurement microphones:


The cardioid microphones will give superior coherence readings, reduced frequency response ripple and generally easier to read data.
The cardioid mics are more directional than the human ear, and therefore the optimistic readings may be misleading. This is somewhat mitigated by the fact that the omni mics are less directional than the human hearing, thus giving a pessimistic response.
The cardioid mic must be very precisely aimed at the sound source for accurate results. This required vigilant attention to prevent false conclusions regarding speaker high-frequency response.
There is an additional complication when measuring the interaction between two sources. If the sources originate from different angles to the mic they cannot be both simultaneously on-axis. If the mic is placed on axis to one it rejects the other, if placed at the center axis between the sources then the error is spread to both.
The proximity effect, while well known for its effect in the near field of cardioid mics, continues in the far field. As the mic moves away from the source, the low-frequency response continues to roll off. Distant free-field measurements will show a reduction for each doubling distance. Indoors there is a natural tendency for low-frequency response to rise in the far field due to strong early reflections. This audible effect will be removed from the measured data by the proximity effect of the cardioid mic.


Buy the book HERE (http://www.rationalacoustics.com/store/books/sound-systems-design-and-optimization.html).;)

Jamie
June 27th, 2008, 04:00 PM
In general, I would NOT recommend using cardiod microphones for measurements. There are many issues that make their use particularly non-intuitive, the data they produce often misleading, and their construction less optimal/rugged. Also, there are very few cardiod mics out there that one would even consider "measurement quality", and those are pretty pricey. (the DPA 4011 is a fine example). Instead of listing off all of the issues here, let me point you in the direction of some good info on this matter.

From the "Microphone University (http://www.dpamicrophones.com/page.php?PID=1)" section of the DPA website (http://www.dpamicrophones.com/):
"The omnidirectional microphone (a pressure type transducer) is in its working principle a more simple capsule construction than a directional microphone (pressure gradient transducer). Simplicity can result in a cleaner and more dynamic sound with a flatter frequency response. Due to the nature of a single diaphragm omnidirectional microphone (the sound only excites the diaphragm from the front) the construction can also be more rugged and therefore offers even better reliability and thermal stability."

From Earthworks' (http://www.earthworksaudio.com/50.html) "Omni Application Guide":
"Why omni? – Omnis have some inherent advantages over directional mics. They sample sound at a single point in space, capturing it exactly – directional mics use a combination of sounds sampled at several different points to achieve rear rejection, creating phase problems in the process. The omnis do not exhibit any proximity effect or off-axis coloration. They have extended accurate low frequency response."

I would also direct you to a few good papers from Sage Technologies who are involved in industrial vibration measurement - and particularly this article which was published in S&V - "Differences in Measurement and Studio Microphones (http://www.sagetechnologies.com/library_documents/dave_qa/q&afeb0.pdf)"

And of course 6o6 McCarthy's book, "Sound Systems: Design and Optimization (http://www.rationalacoustics.com/store/books/sound-systems-design-and-optimization.html)" has this nice passage (http://books.google.com/books?id=w9B2kgjxl4cC&pg=PA180&lpg=PA180&dq=cardioid+measurement+mic&source=web&ots=05SqUvA9si&sig=eOB6cKOmGK-5qKB4AjKUuT7kFik&hl=en&sa=X&oi=book_result&resnum=9&ct=result#PPA180%2cM1) regarding this very topic.

Using cardiods for measurement does have some validity, but omni's should be considered the standard and best option.

-j

Harry Brill Jr.
June 27th, 2008, 06:17 PM
For those considering grabbing a spare SM58 out of a the work box, this should shed some light on some of the problems with that idea.

Arthur Skudra
June 30th, 2008, 11:37 AM
I think it should also be pointed out that directional mics vary in directionality according to frequency as well. At low frequencies, they take on an omnidirectional characteristic, which can be very misleading for measurement purposes, and of course at higher frequencies they take on a directional characteristic.

No doubt the best mic to use is a omnidirectional microphone and as flat of one you can afford. When faced with a situation where you have extremely poor coherency and few options to improve it, perhaps a very flat cardioid could be used to improve coherency, provided you understand the weaknesses in doing so, particularly in the LF. Then again, any "tweaks" in the LF done indoors should be done with a lot of caution regardless of measurement method.

Where I have found cardioid microphones extremely useful in test and measurement is localizing where a reflection is coming from using the "Continuous" option in the IR Analysis window (one of my favourite added features in Smaart v6). While watching the intensity of the "peak" in the impulse response, line up a laser distance measuring device beside the microphone, which allows you to pinpoint the offending surface and calculate distance as well. Kind of a poor man's polar ETC, without all the bells and whistles.

Bob Hulme
June 30th, 2008, 12:51 PM
While watching the intensity of the "peak" in the impulse response, line up a laser distance measuring device beside the microphone, which allows you to pinpoint the offending surface and calculate distance as well. Kind of a poor man's polar ETC, without all the bells and whistles.

Could be something to add to the Rational Store. Most of the laser distance devices I've looked at are only good for about 18 metres.

Thanks, Bob H

PS Great site guys!

Arthur Skudra
June 30th, 2008, 04:28 PM
Hey Bob,

I just bought myself a Hilti PD4, a super compact laser measuring device that is accurate to 2mm to just over 200 ft! Fits in my shirt pocket, extremely lightweight. Been rather pleased with it, though I'm wishing I spent some more money and got the PD40 or PD42 for being able to calculate area, volume, and even do pythagoras for wall heights (though I could do the same with my pocket calculator), maybe I'll trade it in or sell it. The Leica Disto are also a great laser measuring device offering equal capabilities for similar prices. Harry mentioned that green lasers are easier to see outdoors, if that's a consideration for you. Note that these devices are anywhere from $250 and up to $1000 depending on the features you get.

I do know that Hilti and Leica are pretty exclusive for dealerships, but perhaps something could be worked out for the store here, I see them on eBay quite a bit.

Arthur

Michael Häck
June 30th, 2008, 05:23 PM
using the "Continuous" option in the IR Analysis window (one of my favourite added features in Smaart v6).

It would be a much better feature if it would show the calculated delay for every runthrough in "Continuous" mode!

Michael

Bob Hulme
July 1st, 2008, 06:23 PM
Been rather pleased with it, though I'm wishing I spent some more money and got the PD40 or PD42 for being able to calculate area, volume, and even do pythagoras for wall heights (though I could do the same with my pocket calculator), maybe I'll trade it in or sell it.
Arthur


Thanks Arthur,
The PD4 is exactly what I need. If you want to sell yours to get a PD40 or PD42 let me know and I'll take it off your hands. I think Home Depot sells Hilti so if you want to hang onto yours I'll check it out there.

Thanks again, Arthur, I'm going out to my front porch to have a glass of wine with my wife and enjoy this beautiful Canada Day. Happy Canada Day all you Canucks, Bob H

Calvert Dayton
July 1st, 2008, 10:55 PM
Let's try to keep it on-topic guys. We want to try and keep the S/N ratio as high as possible so if you want to talk about something else entirely, please start a new thread. A discussion of laser range finders might be a lovely addition to our gear forum, I note.

Thanks very much.

Calvert Dayton
July 1st, 2008, 11:20 PM
Interesting idea, Arthur. Seems like you could effectively accomplish the same thing by placing some obstruction, such as a small piece of sound absorbing material, in the path between a suspected reflection and an omnidirectional mic though.

Calvert Dayton
July 2nd, 2008, 12:14 AM
In reference to your earlier post, Harry, I would probably have to take off my shoes to count the number of people who have asked me in complete seriousness if the SM58 would make a good measurement mic. Seriously. But I've also heard from people here and there who made it through the night OK with a borrowed SM81 after a dog ate their measurement mic. Not an ideal replacement for a good, flat omni for the lion's share of general measurement applications of course, but in a pinch it might beat the heck out of no microphone at all, particularly outdoors.

Inside or out, one would want to remain acutely aware that even a DP4011 (http://www.dpamicrophones.com/Images/DM02424.pdf) is down a good 6 dB by 20k at 45º off axis and about 12 dB by 90º, relative to its response from 10k on down. That would seem to make the 4011 (or any lesser cardioid, of course) a rather dubious choice in general in any situation where any significant amount of energy was likely to be arriving from any direction other than more or less straight ahead.

Arthur Skudra
July 2nd, 2008, 10:28 AM
Interesting idea, Arthur. Seems like you could effectively accomplish the same thing by placing some obstruction, such as a small piece of sound absorbing material, in the path between a suspected reflection and an omnidirectional mic though.
Murphy's law of precision: "Measure with a micrometer, mark with a chalk, cut with an axe." :D

I've done that as well using a wide variety of absorbing materials such as scrap sonex, leather jacket, etc, and while it's perfectly fine for a simple, boxed shaped room, I found that my method using a cardioid mic, laser (even a simple laser pointer will do), and continuous impulse mode allows one to pinpoint an offending surface easier (by simply looking at the intensity of the reflection), especially in a more complex shaped room, gets me in the ballpark of what/where I need acoustical treatment. I will agree with Michael that a continuous updating of distance would have been nice.

Jamie
July 3rd, 2008, 04:03 PM
When faced with a situation where you have extremely poor coherency and few options to improve it, perhaps a very flat cardioid could be used to improve coherency

&

using the "Continuous" option in the IR Analysis window (one of my favourite added features in Smaart v6).

Two comments:
First, low coherence is not necessarily an indicator of poor measurement quality. Often, it is simply an indicator of poor direct/reverberent energy. If you were just trying to improve the coherence trace, you can also open up the FFT time window by going to single FFT transfer function (instead of FPPO). This basically creates a measurement that considers reverberent energy as part of the signal because the FFT TC at high frequencies is significantly large (@341ms for a 16K FFT at 48K sample rate).

While sometimes annoying to have lots of blood on the screen (a Coh trace that is all over the place) getting in the way of the data, it is a useful indicator of the reverberant environment and system intelligibility. I am not convinced that a cardiod mic's rejection of off-axis mid-band energy really translates/correlates to a human's listening experience and perception of direct/reverb and system intelligibility. Then again, an omni mic isn't a perfect model either. . .

Second. Continuous IR was added back into Smaart with v6.1. It was there in SmaartLivev4 and v5.