Taming the live room

We have constructed our tracking room/ live room and put down a solid ash floor. All walls and the ceiling are skimmed and painted gypsum board. Isolation levels are acceptable if not better than anticipated and the customised AC ducting is virtually silent :D However the room is extremely live as expected and needs some treatment to render it useful for recording. I would like to get the reverberation time down to around 1.5 seconds without loosing too much life in the room. I am thinking along the lines of building two suspended ceiling clouds and slatted absorber / diffuser panels on two of the opposing walls ( the ones without windows in the pic) We are also considering moveable "gobo" style panels if we need further control I would really appreciate some input and advice before starting the process and look forward to any comments this forum can put forward . . Matt I have attached a dimensioned drawing of the room

Have you done a REW test in there yet, Matt, to see where the problematic areas are? You need to do that to understand what types of treatment you need, and where you need them, and what frequencies you need to tune them to. Theoretically you need about about 800 sabins of absorption in there to get it down to ITU/EBU specs, but that would be too dry for your needs, by the sounds of it: You want it to stay very live and reverberant from what you say (RT60=1.5 seconds), so probably around 300 to 400 sabins would be good. However, the question is WHERE to put that, HOW to implement it, and how to TUNE it! :lol: That's the 64,000 dollar question... That's what REW will help you decide... So if you have done REW tests then please post them here. If not, then do them! :) - Stuart -

Thanks Stuart - planning to get some measurements taken this weekend. Any tips on mic and speaker positioning in the live room measurement context would be much appreciated... Do you feel that 1.5secs is too long in a room like this for recording live bands (guitars/keys/drums)? Perhaps I should be looking at some dryer but I am concerned about the space becoming too dead . . Matt

Do you feel that 1.5secs is too long in a room like this for recording live bands (guitars/keys/drums)?

That's 3 different things!! :) Drums generally like a "brighter" room, guitars seem to prefer a "cosier" sound. Vocals, for example, would be rather difficult in that room (unless you do a lot of Gregorian Chant recordings!) But in general, I would probably shoot for something around 1 sec overall for that room, and have some way of extending that or shortening that as the need arise: variable acoustic panels, gobos, etc.

but I am concerned about the space becoming too dead . .

Right! So do the treatment in stages, not all at once. Then you can stop when you have enough...

Any tips on mic and speaker positioning in the live room measurement context would be much appreciated...

For a tracking room, I like to do one test of the overall modal response of the room: Put a good full-range speaker down on the floor in one corner, and your measurement mic up almost against the diagonally opposite corner. That gives you a good indication of what the modal behavior is doing. Then do a second test with the speaker raised up above the floor in the typical location where you might normally have your drum kit, acoustic guitar, or whatever set up close to one end of the room, and the mic at ear height far away, towards the opposite end of the room, not centered, and reasonably far away from the walls (at least 3 feet in all directions, if possible). That gives you a reasonable idea of the overall room. - Stuart -

Thanks Stuart - still wrapping my brain around REW and hopefully will be all set to take some measurements this coming weekend . .

Took some preliminary measurements this weekend - in a bid to keep the graphs clearer and easier to follow I split the spectrum into two chunks: 20Hz to 400 Hz and 400 Hz - 20 KHz . As expected the RT 60 indicates a fairly long reverberation time at certain frequencies and the SPL response is somewhat uneven . . . Any observations/recommendations most welcome . .

Can you post the actual MDAT file here, so we can download it and analyze it in detail? If it is too big to upload to the forum, then put it some place like Dropbox or another file sharing service and just post the link here.

As expected the RT 60 indicates a fairly long reverberation time at certain frequencies

:shock: Ummmm YUP! Nearly 3 seconds at 120 Hz, and over a second and a half at 1.2 kHz is certainly what I'd call "fairly long"!!! :)

the SPL response is somewhat uneven . . .

That's to be expected, for sure, since you had the mic and speaker set up in opposite corners. Do a second test with them set up as I described. You can do both tests in the same MDAT file: Each time you hit the "measure" button in REW, it does another test and adds that to the data set, so when you save it, they all go into the same file. I think you can have something like 30 individual tests in one MDAT file. - Stuart -

Hi Stuart Wasn't aware of multiple positions in single MDAT so I did separate MDAT's for each position: 1. in corner on floor 2. 1 meter from corner on stands 3. 3 meters from walls on stands Cant seem to post MDAT on the forum so will upload onto a file sharing site - details TBC M

OK remembered my file swap password :? MDATS are here : http://bit.ly/1eaJzJj

Hi Stuart and others following this thread - any comments on the MDAT files I posted ? I am starting to build up some ceiling and wall panels and would appreciate some experienced input . .

Took some preliminary measurements this weekend - in a bid to keep the graphs clearer and easier to follow I split the spectrum into two chunks: 20Hz to 400 Hz and 400 Hz - 20 KHz .

It's better to do complete sweeps then just define which area o the graph you want to look at. With split sweeps, there's no way to look at the area from 300 to 600 Hz, for example, if you needed to do that. There's also no way to look at the complete RT-60 graph...

As expected the RT 60 indicates a fairly long reverberation time at certain frequencies ...

Right, and that's not what you want. Your decay times should be pretty much the same for all frequencies, with just a slight increase in the very low end, and a slight decrease in the very high end. The middle should be as flat as possible.

and the SPL response is somewhat uneven . . .

That's normal for small rooms. Especially untreated ones. It is practically impossible to get the response flat, except in an anechoic chamber... And in fact, your frequency response is not too bad at all. It could be better, and will be once the room is treated, but as a starting point it is OK. Your decay times are VERY long! :shock: That room would be great as an echo chamber, or for recording Gregorian chant, pipe organs, or orchestras...

I would like to get the reverberation time down to around 1.5 seconds without loosing too much life in the room.

Why so long? Normally, you want it under a second for a live room, unless you have specific reasons for making it longer. Typically 600 to 800 ms is a good goal, unless you want it really bright I'm looking at the "full sweep" data set, and seeing enormously long reverb times... That room is going to need some very serious treatment, to tame it. There's some obvious modal ringing going on, which is going to need serious bass trapping, but the overall decay times across the entire spectrum are pretty wild. What treatment do you have in there already? Also, photos would help... - Stuart -