Tuning unusual rooms (practical advice wanted from pros)

[Marco Giustini]

This is a pretty famous room with a THX sound wall

308262.jpg

But you're not them so you might want to try a "bafflette" instead

308264.jpg

The concept behind the sound wall is this:
- Low frequency will play more or less in a spheric fashion from an LF cabinet. The energy going forward gets to you first. The energy going backwards, will bounce on the back of the screen (a long way in your case) and reach you very much out of phase. This is why the typical "LF cabinet in front of the wall" will usually show a dip around 100Hz: that's the out-of-phase energy coming back from the back wall.
The THX baffle wall solves the problem: the solid surface flush with the LF cabinet will immediately reflect back those frequencies which will join the rest of the sound IN PHASE instead of coming out of phase. In order to do so, you should consider the LF wavelength: the baffle should be long enough to be able to accomodate the length of the lowest frequency you want to bounce back. It's usually a few metres per side. 40Hz - the usual deepest frequency a cinema speaker can play - is approx. 8m long. I believe you want at least 1/4 of that length for a wall to be effective so you need no less than 2m wall around the speaker (I think it's 1m around the cabinet, but that is the bare minimum).
The wall should be very solid in order to be able to reflect LFs.

- HF. High frequencies would normally bounce on the screen and, again, reflect on the back wall and come back after a big delay - if you consider the very short length of a high frequency wave (8cm at 4KHz). That just makes your sound less clear. In your case, with a long distance, it might be heard as an echo. Now, because HFs are only a few cms long, you cannot "re-use" the energy that bounces from the screen. The only option is to kill it so it doesn't go anywhere and spoil your sound. 50mm of dense rockwool should be enough to almost completely kill anything above about 500Hz. You don't need a large wall here, just large enough to capture any reflections coming from the screen.

Reinforcing the LF is a serious experiment, you need a large, sturdy structure.
Stopping HF is simpler. Just add thik and dense rockwool around your speaker.

So if you have a few panels of dense rockwool (or equivalent sound absorbing material), I would give that a go - though you don't have a screen in front of the speaker so the effect might be limited.
IF you feel brave, a larger structure will make your LF and MF much cleaner. But then you'll have to store the structure somewhere.

Good luck

[Marcel Birgelen]

Edit: Marco was first to the race... so kudos to him. Also, I'm voting for bafflette as the word of the year and the word I was looking for.
It's far easier to spot the flaws in a mono mix than a complex multi-channel or even object mix, as there is usually far less going on.

I think this blog post describes pretty well what a baffle wall does, including some pictures. Now, it's obviously close to impossible for you to implement this, but you can add a "baffle bezel" around your center speaker, sticking out like 20 inch around the speaker, using some sturdy, non-reflectively coated material. It's know that most purist here will cry wolf, as a "partial baffle wall" is obviously by far not as effective as a full baffle wall, but I've found a "partial baffle" or "baffle bezel" to help.

When reconstructing the room, if possible, you should consider a baffle wall, obviously this only works with fixed screen setups. A baffle wall needs to be a sturdy construction. In our screening room, it's a double drywall of about 10 inches thick, filled with rock wool and the front is lined with mass-loaded vinyl with on top of it dark melamine foam with a rough texture, in order to diffuse both sound and light. Not cheap to build and also not very movable.

[Marco Giustini]

Edit: Marco was first to the race... so kudos to him. Also, I'm voting for bafflette as the word of the year and the word I was looking for.

I hope you didn't have to rewrite your message again!

I feel a full baffle wall is mostly for LFEs, those go down to 20Hz (if we're lucky) and the wavelength we'd be dealing with is now 17m.

BTW - I forgot you also have subs! But honestly the effort to reinforce LFE with a baffle wall is not trivial. Many years ago, I did improvised a baffle wall for 2x4642A, put together with salvaged wood. The owner didn't like it and said that unless I could show him a visible improvement on the RTA, the baffle wall would go. Thankfully there was a 6dB improvement at the lower end - though not down to 20Hz because of the size of the structure - so the "bafflette" stayed

Given the size of your speakers, it shouldn't be too hard to put something together.

Another thought on top of my head: would it be worth aiming the speaker up a bit? From the pics it looks like it's pointing a bit low. That is easy to check though, play PN and walk up and down the auditorium and confirm the HF doesn't suddenly drop towards the end of the room.

[Ryan Gallagher]

Another thought on top of my head: would it be worth aiming the speaker up a bit? From the pics it looks like it's pointing a bit low. That is easy to check though, play PN and walk up and down the auditorium and confirm the HF doesn't suddenly drop towards the end of the room.

The center pair is secured to that rolling platform with a slight rake, in an attempt to minimize the stage deck reflection and make sure the back seats are covered. If it was flat we would indeed lose a bunch of it in the bodies or seats and have even more intense floor reflections to deal with. But as far as SPL, the first few rows are definitely in a hot zone with this compromised setup. People only sit there when it's sold out.

An easy to install bafflette around the center channel is certainly a thought. It would have to store well and be mobile with that unit. Perhaps a facia combined with two winglets that lock in place. But purchases for projects like that will be unlikely to be in the cards unless I can do a proof of concept with quantifiable improvements, maybe some wool style packing blankets and pipe & drape would serve as a test.

Ferris Beuller sounded pretty great tonight. I think some proper time spent tuning with the whole mic array will get us to a pretty acceptable place. If i'm still noticing dialog frequencies that are constructively interfering and distracting I might try the bafflette idea.

I'm mostly ignoring LFE till we get the other cabinet back online. Although I did poke it's "center" and level today getting us closer to a coverage response in the x-curve target with the single sub.

Circling back to the wide vs narrow on those R+L mains cabinets... not sure. I gotta poke the TD for specs or get closer to the label. But one of them has definitely been bumped during some lighting focus (a pipe down lighting position used to be right behind them, alignment of the inner pair is not symmetric right now and you can identify some combing between the more splayed pair by ear).

[Tj Hopland]

Ya that qualifies as a unusual room. Room may not be too bad but the screen and speaker setup is pretty far from ideal for cinema. Normally I don't like the idea of a phantom center but when you can't get a decent speaker in a decent location.... Should be fairly easy to try since you have a 'board', just gotta get the center signal in there and mix it to the sides and see how it sounds. I would also be curious to try disconnecting one of the centers for at least some testing to see how things change. it may not fix anything but its easy to try and could give you useful information.

What are the screen speakers or other pa speakers in the other room? Since those are not being used at the moment maybe bring some of those over for testing? Again maybe not a fix but should give you some good information on what sorts of things work or don't.

[Marcel Birgelen]

I think it doesn't really matter what type of speaker you put there under that screen, the speaker will always have issues, as you have the stage in front. You can aim the speaker a bit more upward and I'm confident that will help a little, but your coverage will still be pretty bad.

It's a rather narrow and long room, so I've also been thinking about doing a ghost center. Didn't we actually talk about that before? Was that about the same room or do you have multiple Boss-level rooms over there?

Maybe an idea would be to still give the center channel their own speakers and put them underneath the L + R speakers on both sides... Just thinking. I wouldn't do that in any normal situation, but when fighting Bowser, Mario needs to pull out all the tricks...

I hope you didn't have to rewrite your message again!

Well, luckily, I didn't put in as much effort as you did. I simply put in a link to an external blog article doing the work.

[Tony Bandiera Jr]

Thanks Tony and Marcel. Just to continue the rambling and details:

The CENTER is a pair of Meyer UPQ-2P
The LFE is (was) a pair of Meyer X-800C powered subs.

Mains are Meyer too, but smaller full range cabinets as shown in attachments, I'd have to ask someone about the models, but similar era I expect.

Dumb EQ Question. In the CP750 are the channel "Bulk EQ" faders always having an effect, or are those one shot bulk adjustment tools? I did not touch them in my first pass effort but did notice our L and R have pretty different values I started from. I suppose in a full effort one would flatten, start with bulk EQ, and then tweak with 1/3 octave, and then further tweak with listening.\

If we are going by ear alone, where we started to where we are now is a definitive improvement, but as Tony pointed out, a tight X-Curve fit may not be what this room actually needs.

We show plenty of mono films, so there has been ample listening opportunity for center only. Even a handful of stereo films, which lacked the intelligibility issues prior. But in this round of investigation have not characterized anything specifically, those methods in Smaart are to be learned still.

EDIT: Ohh on upstage baffling, there is a full stage black traveller that plays closed for films about 10ft behind the screen. And a cyc behind that. But not much else other than the acoustic panels in the house.

Here are spectrum screenshots after my efforts with the 1/3 EQ for LCR (set to infinite averaging in the screenshots). Bear in mind this is the mic next to me at FOH.

I might re-load our old file and share the "before". And yes we intend to continue this learning opportunity with mics placed correctly instead of wall mounted next to me.

Left:

smaart_left.jpg750_left.jpg

Center:

smaart_center.jpg750_center.jpg

Right:

smaart_right.jpg750_right.jpg


The space for the curious:

IMG_7947.jpgIMG_7946.jpgIMG_7948.jpg

IMG_7952.jpgIMG_7951.jpgIMG_7950.jpg

IMG_7955.jpgIMG_7958.jpgIMG_7953.jpg

First off, that's a nice looking space!!

Yeah, you have the same issues I had with that University room I mentioned. Looks like your room acoustics are treated much better than HIB100 was, but....

Meyer Speakers are terrific for P.A. and High SPL concert sound, (and well worth the money for those purposes) but they, as Joe Redifer would say, sound like ass for cinema sound reproduction. Films are mixed using non line array speakers that are designed for the coverage of typical cinemas. (JBL, QSC, etc.) Using P.A specific speakers for cinema will NEVER work out properly for film sound, no matter how much EQ and other audio voodoo you try to pile on. I did some outdoor shows where I had no choice but to use P.A. gear (two of which were Meyers) and I just set everything by ear as I knew the x-curve would not work.

And having that center speaker at floor level whereas the L/R speakers are flown creates HUGE problems with EQ and time alignment (far worse in HIB 100 as the center was about FOUR FEET further back than the wall-mounted L/R speakers are!) You'd be better off with traditional JBL or QSCs set on movable platforms behind the screen (if it's perforated) placed with the HF horns around 2/3 up on the image area.

Finally, the FOH mic location is god-Awful. Too high up and under the balcony ceiling. It is never going to give you any accurate EQ readings for anything. About the only thing that mic, in it's current location, is good for is a rough idea of SPLs. And even that is compromised by the reflections from the ceiling and the open booth behind it.

One of the first things that could really help your case is to get a multiplexer with calibrated mics and set those mics up at the seats in the proper locations, THEN set the x-curve and work out from there. You can essentially ignore EQ with the current setup as the readings are nowhere near what's happening in the seats. (That same applies to the P.A. system as well.)

But bear in mind with the placement and type of speakers you have to work with, it will still not line up with a proper cinema B-Chain setup. That's why I said to trust your own ears based on known source material (That in my case, I had listened to in many rooms, both great and not so great) and try to get it to "sound right" based on that reference. The RTA is just a shiny toy in a sense in this situation. Nice to impress people with, but not too useful (and often misleading.)

I wish I could come to your venue and help out, but my current situation makes that impossible.

[Ryan Gallagher]

First off, that's a nice looking space!!

Yeah, you have the same issues I had with that University room I mentioned. Looks like your room acoustics are treated much better than HIB100 was, but....

Meyer Speakers are terrific for P.A. and High SPL concert sound, (and well worth the money for those purposes) but they, as Joe Redifer would say, sound like ass for cinema sound reproduction. Films are mixed using non line array speakers that are designed for the coverage of typical cinemas. (JBL, QSC, etc.) Using P.A specific speakers for cinema will NEVER work out properly for film sound, no matter how much EQ and other audio voodoo you try to pile on. I did some outdoor shows where I had no choice but to use P.A. gear (two of which were Meyers) and I just set everything by ear as I knew the x-curve would not work.

And having that center speaker at floor level whereas the L/R speakers are flown creates HUGE problems with EQ and time alignment (far worse in HIB 100 as the center was about FOUR FEET further back than the wall-mounted L/R speakers are!) You'd be better off with traditional JBL or QSCs set on movable platforms behind the screen (if it's perforated) placed with the HF horns around 2/3 up on the image area.

Finally, the FOH mic location is god-Awful. Too high up and under the balcony ceiling. It is never going to give you any accurate EQ readings for anything. About the only thing that mic, in it's current location, is good for is a rough idea of SPLs. And even that is compromised by the reflections from the ceiling and the open booth behind it.

One of the first things that could really help your case is to get a multiplexer with calibrated mics and set those mics up at the seats in the proper locations, THEN set the x-curve and work out from there. You can essentially ignore EQ with the current setup as the readings are nowhere near what's happening in the seats. (That same applies to the P.A. system as well.)

But bear in mind with the placement and type of speakers you have to work with, it will still not line up with a proper cinema B-Chain setup. That's why I said to trust your own ears based on known source material (That in my case, I had listened to in many rooms, both great and not so great) and try to get it to "sound right" based on that reference. The RTA is just a shiny toy in a sense in this situation. Nice to impress people with, but not too useful (and often misleading.)

I wish I could come to your venue and help out, but my current situation makes that impossible.

I wish you could too! lol. For what it's worth, there is no mode where changing our non-perforated screen or cabinet choices for the short period this room will still be operating. At least in that area we are stuck with what we got until the building gets rebuilt (they are only saving the historic facade in the plan). In that sense I might be attempting to polish a turd for a limited number of remaining film engagements. Worth it mostly as a learning excercise.

Yeah that FOH mic placement is pretty useless for anything but the crudest perspective. I'll work with the mic array Mon or Tues, but I'm outta town pretty quick after that. Presumably someone time-aligned this system way back when, but would be worth looking for those issues again.

Do you actually need a mic multiplexer? Can't the Octa-Capture and Smaart achieve the same properly configured? We can get each mic configured in the live RTA, but getting the averaging views going will be something to learn.

[Marco Giustini]

Indeed, a multiplexer is not needed anymore with an octacapture or similar sound input device. In fact, you get more data in less time as the multiplexer would send the software 1 second of audio from each microphone so after say 8 seconds and 4 mikes, you'd get 2 seconds of sound from each microphone. With Smaart and a multi-input device, in 8 seconds you get 8 seconds of audio from each input.

Getting the average in Smaart is as simple as adding an average reading in your "spectrum" configuration Latest versions of Smaart allow you to set a reference microphone so the levels of other mikes are normalised to the reference input before they're considered for the average. Something only WinRTA used to do (The D2 system) which has now been implemented in Smaart too.

BTW, WinRTA is now free as the developer has retired and he has decided to make all his work open source. Bless him. I still have my D2 Studio somewhere!

[Ryan Gallagher]

Ya that qualifies as a unusual room. Room may not be too bad but the screen and speaker setup is pretty far from ideal for cinema. Normally I don't like the idea of a phantom center but when you can't get a decent speaker in a decent location.... Should be fairly easy to try since you have a 'board', just gotta get the center signal in there and mix it to the sides and see how it sounds. I would also be curious to try disconnecting one of the centers for at least some testing to see how things change. it may not fix anything but its easy to try and could give you useful information.

What are the screen speakers or other pa speakers in the other room? Since those are not being used at the moment maybe bring some of those over for testing? Again maybe not a fix but should give you some good information on what sorts of things work or don't.

Thanks TJ. I missed that phantom center thought earlier. Also a fun experiment. Trying with a single active center is also worth giving a listen to.

As for all that potentially useful gear from the big room. It all got ripped out, piled, and shipped off in bulk to some Vegas consignment outfit. Including the model matching projector. Not what I would have done with it given other options, but storage is a pretty crazy subject in these two historic venues (one of which we had to completely empty). Only thing I managed to interject to save (outside of 35/70 equip) was the AP20 as a backup to the CP750, which seem to be prone to failure based on my time here.

[Jay Wyatt]

Ryan, I work at Meyer and I think have your venue in our database from a 2011 entry -- assuming this is the Paramount in Austin on Congress Ave. What you have is far different than what was proposed, so I'm not sure of the full backstory on how the venue arrived at that particular equipment list.

Judging from a glance, this is how I'd rank my concerns.

1. Center channel model & placement - I really, really don't love the splayed + ground stack deployment with the two UPQ-D2s.
1a. In general, any splay on cinema Center is a big no-no. A splay is a tradeoff for width of coverage at the sacrifice of intelligibility and intelligibility far outranks coverage on center channel in cinema playback. The splay will create lobes flanking the centerline and these lobes will vary in shape and size at every frequency above ~400Hz. See image UPQ-D2 splay at 2000Hz. A single UPQ-D3 or an Acheron100 would've been a far better choice, circa 2011, and half the cost to boot.
1b. A ground stack will yield a forward aspect ratio (distance ratio from Center to front seat and Center to last row) far too large. Any chance for uniformity front2back is doomed and it's not solvable with EQ. Also the groundstack will introduce the floor bounce reflection that's going to be problematic on intelligibility. If you could use that pick plate on the top and fly those you'll improve uniformity, hopefully overshoot to stage to mitigate the stage reflection of higher frequencies, and improve localization all at the same time. See the two vertical coverage predictions attached (4color vs 8color) for how a little lift can reduce high frequency variance by as much 12dB front to back!

Combining the issues from 1a and 1b (lobes from splay plus a wildly inconsistent front to back ratio) any multiplexed or averaged array of mics is going to be getting such a mixed bag of data that you are almost guaranteed to be making bad EQ choices that may solve one seat but break all others. Again, neither 1a. or 1b. is solvable with EQ so attempting to measure and make fixes with EQ is a lost cause in my opinion.

2. Processing - the left/right hangs look like they are splayed UPQ-D2 paired with either UMS or 500HP subwoofers. The CP-750 alone cannot give you the granularity needed to process this system correctly. You need something with ability for matrix mixing/ LPF+HPF/ and discrete channel EQ so the handshake between subs and point source can be managed in both frequency and time domain. Are you sure there is no Galileo Meyer processor somewhere in the system? I'd be very shocked if there is not.

3. L/R splay angles - With the Center channel deployment being a big head scratcher it makes me question if the tilts and splays for Left/Right were chosen (and deployed) correctly.

In short: Poor loudspeaker model selection, deployed in far from ideal positions, given the wrong processing. Give me a day or two to unpack the venue CAD from the database and bring the existing equipment you have into MAPP-3D prediction tool. From there I'm happy to walk thru some different scenarios and options. You'll get a million answers on a forum but doing realtime prediction gives real answers.

[Marcel Birgelen]

You'll get a million answers on a forum but doing realtime prediction gives real answers.

Look here, direct manufacturer support.

Yeah, a thousand people, a thousand opinions, but I guess that it helps that many of the people that hang out here secretly aren't amateurs (not speaking for myself, I'm the king of amateurs obviously...). But to be honest, I think that what most people here have been saying is pretty consistent with your own conclusion: "splayed" speakers will introduce combing, especially since those speakers are not line arrays. Placing the center so close to the ground will yield in reflections from the stage and bad coverage of the audience.

]Are you sure there is no Galileo Meyer processor somewhere in the system? I'd be very shocked if there is not.

Ryan mentions there are two dox Driveracks involved, probably one for each side.

Do you actually need a mic multiplexer? Can't the Octa-Capture and Smaart achieve the same properly configured? We can get each mic configured in the live RTA, but getting the averaging views going will be something to learn.

You don't need a multiplexer, the octa-capture is mostly fine. It's technically not calibrated, so someone could argue that the ADC in there could "tint" the measurements, but it gives you each microphone as an individual input, which is fine. Just don't use any EQ, De-esser, compressor or other fancy features on your capturing device.

I've only now seen the RTA420 on the picture and I agree with Tony... you're measuring how it sounds below the balcony, not where people are actually listening. While it's obviously impractical to put the microphone inside the crowd during a show, even just installing on top of the desk in the FOH, aimed at the screen should be far better.

[Ryan Gallagher]

Ryan, I work at Meyer and I think have your venue in our database from a 2011 entry -- assuming this is the Paramount in Austin on Congress Ave. What you have is far different than what was proposed, so I'm not sure of the full backstory on how the venue arrived at that particular equipment list.

Hey Jay. Thanks for taking the time. The venue is "Stateside at the Paramount" or "The State Theatre". But yes, Paramount or Austin Theatre Alliance would have been the original purchasing org.

As Marcel said, yes there are DBX 260s as the processors for the mains channels. Only the C,LFE, Ls,Rs go direct from the CP750
I'm just a staff projectionist here with enough trust/latitude to improve things that fall within my abilities. I'd have to poke our TD for original designed intentions. I think our roll-down screen got replaced once, perhaps the original was perforated, and the original plan was to suspend center, and either that changed with the new screen or it never worked out for the logistics of the mixed use space. But it's been this way my entire experience here.

The comments about the splayed cabinets are certainly interesting to follow up on.

[Ryan Gallagher]

Also why do I have this sensation of peering into a rabbit hole. lol. ;-)

[Steve Guttag]

Given the apparent width to length of the room, there is no justification for a splay either. Short and Fat...maybe but I tend to agree with Jay to keep to a point-source (single point, not a splayed one) as much as possible. Altec, on their Mantaray horns could do a splay without lobing, but that was designed into them. However, now obsolete, they are a moot point.

I've had good luck using JBL's PD series (program directivity) series like the PD6332 (available in various horn directivities) and others have used the EAW QX500 series (same concept, vastly different package, imagine working with a pyramid-ish shaped speaker). Naturally, I'm sure Jay and the folks at Meyer can come up with a solution too.