pouët.net

Assembly 2016 PA impulse response

category: general [glöplog]
 
One of the sound guys at Assembly here.

I tried to capture the sound PA of this year's Assembly to an impulse response for people who might want to try out how their prod might sound on a largish system in a large hall. The impulse is captured on the center line right after the first block of chairs, around 15-20m from the stage. It turned out a little bit boomier than the real system is. The reverb seems to be somewhat believable though.

The files can be found at https://www.dropbox.com/sh/lesg835atqn0aqi/AADja-0tCmBWo8SFYW84wN6Ca?dl=0

The SIDR format can be inputted to Logic's Space designer to create a preset. The single WAV files can be inputted at least to http://www2.ika.rub.de/HybridReverb2/ which was the only free convolution reverb i could find that works on Windows and OSX and works with the impulses. The WAV format might work on some other convolution reverbs, but I don't know of any that actually load it.

The mic setup is a pair of Oktava MC 012's with cardioid capsules in an ORTF-ish setup at standing listening height. The result isn't perfect, but maybe we can improve it next year.
added on the 2016-08-04 02:48:15 by zarr zarr
Thanks, this is great! I suppose it might be a good idea to record also the short film compo ;)
added on the 2016-08-04 03:50:30 by cce cce
I wonder how much the existence of audience will affect the impulse response as compared to the empty hall =)
added on the 2016-08-04 08:03:13 by sol_hsa sol_hsa
Nice! Did you use sine sweep to generate the IR? Or did you playback an approximation of an impulse and just record that?
added on the 2016-08-04 08:26:33 by shaiggon shaiggon
Audience will of course affect the response. But it's not really reasonable to measure with the audience watching and listening. :)

Sine sweep was used. The software used is Logic's Impulse Response Utility.
added on the 2016-08-04 17:53:14 by zarr zarr
Nice :-) If nothing else, these impulse responses will probably make people think about mastering for PA and not only their headphones.
added on the 2016-08-04 22:15:45 by Sesse Sesse
Come on, there's still time! Put the sine sweep as the sound track to a demo and submit it to one of the demo compos. (and bribe the jury to qualify it) Then it will get played _with_ audience and everything.
added on the 2016-08-05 10:56:21 by yzi yzi
This also works with other audio editing software, e.g. in Audition/Cool Edit Pro you can use the built-in Convolution filter, just load the ".wav" file into it first.

But won't a crowded hall dampen the reverb quite a lot compared to an empty hall?
added on the 2016-08-05 13:38:47 by Kabuto Kabuto
It's not that crowded tho, the place is ridiculously oversized and every surface that is not a person or the screen is hard as concrete. Echoes like crazy.
added on the 2016-08-05 13:45:55 by msqrt msqrt
Quote:
Come on, there's still time! Put the sine sweep as the sound track to a demo and submit it to one of the demo compos. (and bribe the jury to qualify it) Then it will get played _with_ audience and everything.

I guess #3 in the 1k intro compo is a good start... but it only sweeps over the low frequencies. :D
#3 was when my neighbors must have thanked me for not buying a subwoofer.
added on the 2016-08-06 22:18:57 by Sesse Sesse
Quote:
If nothing else, these impulse responses will probably make people think about mastering for PA and not only their headphones.


It's the responsibility of those who setup the system in a hall to ensure music playback of regular mastered tracks for consumer home equipment sounds right at that specific place. That's definatly not the job of those who master the track in a (home) studio.
added on the 2016-08-06 22:49:35 by Salinga Salinga
You cannot simply remove the room's reverb for a room of this size. You can make it sound good, yes, but you cannot make a big party hall sound like your tiny cozy living room at home (and why should you)? Hence the impulse responses can help the artist in identifying problematic frequencies and work around them - naturally in such a way that it still sounds good on a home stereo.
In general: If you can remove it and it still sounds good afterwards, then do remove it. Chances are high that what you just removed might sound completely horrible on another PA. ;)
Salinga: You just can't get it to be the same. PA and home equipment is just too different, and it goes both ways—if you mix or master for PA and then push that directly out on a stream, it usually won't sound too good.

I agree the party should push for minimizing these effects whenever possible, though :-)
added on the 2016-08-06 23:13:11 by Sesse Sesse
> identifying problematic frequencies

And that cannot be done with the sound system at the event location? I have trouble to believe that in 2016 there is no technology to do that. Planes have technology to cancel out the engine noises in an aircraft cabin.
added on the 2016-08-07 03:36:03 by Salinga Salinga
You know the frequencies, the engine produces in advance. That's not the case with compo entries played in a large hall. Counter phase sound that doesn't arrive at the listerner's ears exactly at the same time the normal sound does, would sound horrible too. Cancelling out noise is one thing, adjusting for a hall's accoustics is an entirely other, Salinga.
added on the 2016-08-07 07:50:29 by Punqtured Punqtured
Well, noise-cancelling headsets don't get the sound in advance, really; they have to be prepared for any kind of noise spectrum. But noise is uncorrelated (so if you miss slightly on the phase, you'll just hear slightly more noise, not a super-annoying distorted sound), and the headset only needs to care about one specific listening position, whereas the PA has to care about everywhere there might possibly be people, which pretty much precludes sending any kind of counter-echo.

You _can_ adjust for frequency dips and spikes, though (as long as they're reasonably even in the listening area you care about), and a good PA engineer will. It's the phase that kills you.
added on the 2016-08-07 10:48:19 by Sesse Sesse
Quote:
Thanks, this is great! I suppose it might be a good idea to record also the short film compo ;)

http://archive.assembly.org/2016 with thanks to Limp Ninja for the link
added on the 2016-08-07 12:11:46 by numtek numtek
As hinted earlier I also had my own guerilla impulse response recording in the works. The idea was to have sine sweeps in a video entered to the short film compo. I managed to make a recording of the compo but I'm sad to inform you that the operation wasn't a great success for several reasons:

  • The sine sweeps weren't spaced with enough silence to fully capture the long tail of the hall reverb.
  • Even if they were, the noise floor was very high and a single yell could've ruined the recording anyway.
  • I didn't use a reference microphone.
  • The original audio in the video was rendered as 320 kbps MP3 by accident.
  • Sounds like the audio was downsampled from 48 to 44.1 kHz. At least the higher frequencies of the sweep sounded odd.

Because of the noisy environment some other technique such as MLS would've probably been a better choice.

I generated the sweep with Voxengo Deconvolver. It has a fully functional free version. The audio track had three and five second sweeps. I did the recording with Zoom H4's built-in mic at 48 Khz/24-bit.

The recording was done behind the audience, next to the mixing desk:
BB Image

Here's an attempt to salvage at least some impulse response from the recording: assembly_summer_2016_with_audience.wav

The reverb length sounds about right but high frequencies are seriously dampened and don't sound like the hall at all. To me, the IR posted by zarr is much more realistic.

Some files for reference:


At least the audience seemed to take the compo entry well, so no harm done! Keep on measuring. :)
added on the 2016-08-07 15:57:25 by cce cce
Excellent :D
Haha, fantastic. I sense a new Rob is jarig.
added on the 2016-08-08 00:09:19 by Sesse Sesse

login