Room Correction: How, why, where?

[jimfisheye]

I have the Dolby software from the mouth of the beast itself. The rendered and encoded file matches the original mix exactly. It's 1:1. There's no "Atmos curve". Or... flat, no eq, would be the "curve".

It's easier to correct resonance issues with eq than bass traps and dampening only in that it's clicking software controls vs physical activity and light construction. Results based though, it's the opposite. Cursory treatment is a LOT more bang for the buck.

This is if you're interested in reproducing a mix. If you just want something to sound good to you regardless of accuracy, well you can do what you want and if it sounds right to you then it is.

My 2c on shitty rooms that you will NOT be treating physically:
Don't go for perfect. It's not going to happen no matter what you try. Just go for 'not fucked'. Keep the eq moves minimized. Just go after the biggest offenders. Back off on extreme moves. Let the room be weird and color the music. You'll end up hearing more in the end. Severe eq leads to phase mayhem and choked sound. I'm not afraid of being ruthless with eq! Diminishing returns are diminishing returns though.

Source: Running live sound in what could more accurately be called caves than rooms in some cases.

Similar advice for behind the mixing board. Up to 8 bands of adjustment on the eq on some track? Going for band #9 now? Yeah... Start over!

 

[AudiophileStyle]

I have the Dolby software from the mouth of the beast itself. The rendered and encoded file matches the original mix exactly. It's 1:1. There's no "Atmos curve". Or... flat, no eq, would be the "curve".

It's easier to correct resonance issues with eq than bass traps and dampening only in that it's clicking software controls vs physical activity and light construction. Results based though, it's the opposite. Cursory treatment is a LOT more bang for the buck.

This is if you're interested in reproducing a mix. If you just want something to sound good to you regardless of accuracy, well you can do what you want and if it sounds right to you then it is.

My 2c on shitty rooms that you will NOT be treating physically:
Don't go for perfect. It's not going to happen no matter what you try. Just go for 'not fucked'. Keep the eq moves minimized. Just go after the biggest offenders. Back off on extreme moves. Let the room be weird and color the music. You'll end up hearing more in the end. Severe eq leads to phase mayhem and choked sound. I'm not afraid of being ruthless with eq! Diminishing returns are diminishing returns though.

Source: Running live sound in what could more accurately be called caves than rooms in some cases.

Similar advice for behind the mixing board. Up to 8 bands of adjustment on the eq on some track? Going for band #9 now? Yeah... Start over!

The curve has nothing to do with what’s baked into the mix or Dolby software. It’s how Dolby recommends calibrating the room in which the mix is done. Without a common curve, there is no common sound.

 

[AudiophileStyle]

I do up to 13 mic placements in my audio room, although since I sit in one stationary position 5 would probably be plenty and I alone am the listener

I’ve been experimenting with this lately with Dirac. Everything from a single measurement to tons of measurements. The sonic difference are quite large. For a single listening position, I really like 9 measurements, all taken very tight to the listening position.

 

[jimfisheye]

The curve has nothing to do with what’s baked into the mix or Dolby software. It’s how Dolby recommends calibrating the room in which the mix is done. Without a common curve, there is no common sound.

Dolby recommends no such thing. The delivery format expects a flat system. Achieving a flat system is up to the listener as always.

My guess is this is more innocent telephone game than any intentional misdirection.

 

[mkt]

Interesting reading here https://professionalsupport.dolby.c...yuDGAQ/room-design-calibration?language=en_US including the FAQ with "What target should I use for calibration?" and this pdf
with the Dolby target cure plot https://dolby.my.salesforce.com/sfc...V/YHgXgrJVKAcelcXkD.DnlRUBZU1CqEMGDKweLRpuu98

 

[mandrix]

I’ve been experimenting with this lately with Dirac. Everything from a single measurement to tons of measurements. The sonic difference are quite large. For a single listening position, I really like 9 measurements, all taken very tight to the listening position.

Yes. I tend to favor 9 positions, for a very focused listening position. Honestly I can't really hear any difference with 13 now that I've been through all this multiple times.
I have done only 5 positions in the past, but that was more because of a time crunch because my sub volume control was acting up and I was running out of "quiet time" in the house.

 

[AudiophileStyle]

Dolby recommends no such thing. The delivery format expects a flat system. Achieving a flat system is up to the listener as always.

My guess is this is more innocent telephone game than any intentional misdirection.

They have recommended this for years.

https://dolby.my.salesforce.com/sfc...V/YHgXgrJVKAcelcXkD.DnlRUBZU1CqEMGDKweLRpuu98

 

[AudiophileStyle]

Yes. I tend to favor 9 positions, for a very focused listening position. Honestly I can't really hear any difference with 13 now that I've been through all this multiple times.
I have done only 5 positions in the past, but that was more because of a time crunch because my sub volume control was acting up and I was running out of "quiet time" in the house.

It’s kind of fun to try the different options. Plus, it’s free to mess around.

Comparing the tightly focused listening position measurements to the wide couch style is also ear opening.

All fun though. No right or wrongs.

 

[mandrix]

It’s kind of fun to try the different options. Plus, it’s free to mess around.

Comparing the tightly focused listening position measurements to the wide couch style is also ear opening.

All fun though. No right or wrongs.

Yes. My problem is usually the wife goes to town and is back in an hour and half. If you run into problems (planes/mowers/thunder/dogs) lol it extends the time you need.

 

[ar surround]

Yes. My problem is usually the wife goes to town and is back in an hour and half. If you run into problems (planes/mowers/thunder/dogs) lol it extends the time you need.

LMAO!! Boonie. I was just thinking about this dilemma a few minutes ago. I have to get my wife out of the house for a period of time, turn off the dehumidifier and AC, turn off the ringer on the phones, and hope that there are no significant extraneous noises...sneezing, coughing, puking, 4.8 magnitude earthquakes, etc. I like how my rig sounds right now with Audyssey room correction cut off at 500 Hz; and any attempts at further tweeks are more likely to screw things up.

 

[MTB Vince]

I totally agree with you 100%.
My fear as I have flush ceiling mount (4) speakers would be that the cloud panel would take away from the direct line signal to my seating area.
I built my Atmos room basically at the start of Atmos rooms being built, if I had to do over again I absolutely would do as you did with hang down speakers.
As I suspect and as you say my celling flush mount speakers are off the table when it comes to cloud.
However I am in process of doing a revamp of certain equipment, in research phase now, and possibly this year I will change from my flush mounts to hanging mounts like yours.
My flush mounts at tome of purchase was B&W's top of line. None of my other speakers are flush mount all cabinets.
What is your brand of hanging, if I may ask.
Thanks for the solid input.

Sorry @marpow,

I totally missed your response to my post way back in January and the question about the specific speaker mounts I employ for my ceiling channels. So in hopes that you still find that info useful 7+ months later (Doooh!), here is the link to these K&M (Konig & Meyer) professional studio monitor mounts. If you scroll to the bottom of the linked page you'll see where these mounts are paired with a specific interface plate that is compatible with whichever "install variant" studio monitor you want to hang. In my case that was the ATC SCM12i studio monitors which have a steel plate-reinforced back panel with threaded inserts the interface plate is bolted to. This ensures a rock solid, safe, and secure mounting even for quite heavy loudspeakers. My (large-ish) shoebox sized ATC's weigh 24 pounds each.

 

[jimfisheye]

They have recommended this for years.

https://dolby.my.salesforce.com/sfc...V/YHgXgrJVKAcelcXkD.DnlRUBZU1CqEMGDKweLRpuu98

This is getting into skewing a near-field setup to translate better to a full setup. Otherwise you might tend to back off on the highs and boost the lows too much on the mixing board. Near-field is a tight concise way to monitor and maybe removes some of the 'ear candy' factor for some no bs mix balance work but it's unnatural and can lead to backing off on highs and boosting lows. Because the speakers are right near you and the soundstage is shrunk down. Earlier in the document you will note their recommendations for room size and to calibrate for balanced audio across the whole audio band.

It's a recommendation for setting up to create a room neutral mix... even if you are mixing near-field.

Atmos is not intended to use a pre-emphasis curve akin to the RIAA curve for vinyl where it would have to be normalized on playback. I interpreted the "curve" comments here as "pre-emphasis curve". My bad! It wasn't meant to mean that. I imagine others will do the same since "eq curve" and "room analyzer software" tend to mean that on consumer stuff.

I think we're arguing at cross purposes. Dolby never intended this to read like it was a pre-emphasis curve. Their recommendations for handling calibrations to normalize between rooms all have solid advice. Some of the writing might be worded a little ambiguously... only because all of this is a mouthful to describe!

 

[jimfisheye]

Reeling back to the point I was going for.

Trust your ears! There's a working range that some room eq adjustment delivers good results in. Too strong of cuts or too many cuts crosses the line and starts mutilating the signal. ie. Don't try to make a big church sound like a small closet.

Or grab some Owens-Corning 703 and get some bang for the buck with a genuinely deadened room.

 

[AudiophileStyle]

This is getting into skewing a near-field setup to translate better to a full setup.

I’m sure a near field system can be configured in a room of the recommended size, but I’m guessing this is a “regular” size space.

 

[mandrix]

LMAO!! Boonie. I was just thinking about this dilemma a few minutes ago. I have to get my wife out of the house for a period of time, turn off the dehumidifier and AC, turn off the ringer on the phones, and hope that there are no significant extraneous noises...sneezing, coughing, puking, 4.8 magnitude earthquakes, etc. I like how my rig sounds right now with Audyssey room correction cut off at 500 Hz; and any attempts at further tweeks are more likely to screw things up.

lol. I hear you Brother. I had a good laugh from that.

 

[mandrix]

I’m sure a near field system can be configured in a room of the recommended size, but I’m guessing this is a “regular” size space.

View attachment 108053

Well don't freak out, but my audio room is 8.5 x 12.5 with a slanted roof. Thank all that is for room correction software.
I was running 7.1.4 but due partially to the fact that my AVR likes to extend DTS-HD to all available speakers, I decided to switch for now to 5.1.4., see how it goes.
Plus my base 5 speakers are a matched set.

 

[Albert Dagger]

Target curves can work for a single speaker or a group. e.g. If I make adjustments, I can group the fronts/rears/ tops or anyway I see fit. Or a single speaker.
When I add in the Harman curves I actually include all speakers, as the Dirac Live software seems fairly conservative on the low end. I can't recall offhand in this case exactly where the curtain is, in regards frequency, without looking at my calibrations but suffice to say I get the low end boost I desire.

I know you can group front speakers or all speakers if you want. Still, a target curve is used to calibrate each speaker individually, and that’s what I meant.

Technically, you shouldn’t get a boost if you only add a low shelf of +5db below 150Hz. That should be the natural response of a good speaker, when room modes are managed correctly.

 

[Albert Dagger]

Based on my own research, experience using several room correction products, and talking to experts in the field, I believe some of your points are a bit incorrect if one is pursuing accurate sound reproduction. If one is shooting for what pleases them personally, not necessarily accurate, then all bets are off (absolutely nothing wrong with that).

Current state of the art room correction using 65,000 tap filters at 44.1/48, and scaling up with sample rate, created with Audiolense, work stunningly well when correcting the entire frequency spectrum. It sounds much better than partial correction in my experience.

What I wrote is based on Floyd Toole’s scientific writings. I haven’t added anything of my own. He said that changing the natural response of your speaker above the transition frequency should be avoided (assuming you have good speakers)! But I don’t see any issues with experimenting and using room correction as one sees fit.

 

[barfle]

My room has 4” of open denim insulation behind the front speakers, which have ports in the back. One wall is fairly plain, but it’s where the seating is basically against that wall. The other side is generally open shelving, so that surface is fairly well broken up. It’s a “bonus room,” so there are slanted surfaces between the walls and ceilings. I floated the floor over two layers of carpet pad.

I haven’t done any other acoustic treatments so far, although I’m having a quilt made to put in the first reflection point on that plain wall. We’ll see what that effect is.

I use the built-in room correction in my Marantz pre-pro (Audessy?), with four measurement points, because that’s how many seats are in the room. I haven’t searched out resonances, but I haven’t detected any with normal music and movies.

Again, I’ll stop tweaking the room when I’m carried out.

 

[Sal1950]

Current state of the art room correction using 65,000 tap filters at 44.1/48, and scaling up with sample rate, created with Audiolense, work stunningly well when correcting the entire frequency spectrum. It sounds much better than partial correction in my experience.

Quite true but not many have that level of resolution.
Without that SOTA level of taps the app can cause more harm than good above the bass range.

For a single listening position, I really like 9 measurements, all taken very tight to the listening position.

How tight? I've found about 18" between positions to work well here.

They have recommended this for years.

Personally I find that curve quite extreme, specially the top end rolloff.
But then my ears are 74yo.