Dirac New Spatial Room Correction Software

gene

gene

Audioholics Master Chief
Administrator
Dirac Research is looking to the future. Dr. Lars-Johan Brännmark, Research Fellow and Chief Scientist at Dirac, has given a glimpse into what he believes will be the next generation of room correction. Spatial room correction is a concept that goes beyond correcting each speaker individually in the frequency and time domains. Instead, all speakers work together to correct each other, allowing the user to manipulate the spatial qualities of a system and overcome acoustical problems that are usually left unaddressed.

Imagine having a room correction system that truly optimizes every seat for perfect sound by using ALL of the speakers in tandem to do it? Dirac is on the cusp of making this a reality for you.

dirac.jpg


Read: Dirac New Spatial Room Correction Software
 
TLS Guy

TLS Guy

Seriously, I have no life.
Count me Skeptical Gene!

These schemes for vast numbers of speakers in home situations is an impractical piece of nonsense.

Dirac will not alter the speed of sound. So no matter how many speakers "cooperate" there will still be only one spot where the responses from all speakers can arrive at the same time. In any event if you are going to have that vast number of speakers, which I am certain is not required, you can afford to build a room of the correct dimensions. That will probably be cheaper than all those amps and speakers.

The fact is, that the vast majority of speakers are not very good. The advance has to be better and more powerful speakers of smaller size and footprint.

I believe the Sigberg Audio approach, of much improved small high powered speakers is the way forward to improved performance in the home.

I can tell you 11 good speakers are more than enough to give a good immersive experience throughout the listening area, and without fancy Eq. I know because I listen to it every day!
 
ryanosaur

ryanosaur

Audioholic Overlord
Another great article! ;) Thank you Gene… thank you Jacob!
I’ve seen some pretty surprising results from Dirac posted by others. I’m eager to get to play with Dirac and hope it comes to my rig next year!
While I’m mostly interested in working below Schroeder Frequency, the interesting thing is that many are using it full range because it makes a positive difference.
Add to that what have been raves for their newer Bass Control module… people are ditching their old miniDSP and MSO for DLBC…

This may be a ways off and is an intriguing development. Am interested to learn what comes of it!
 
CreoleDC

CreoleDC

Junior Audioholic
Move your main sitting position, than every settings is out the window.
 
gene

gene

Audioholics Master Chief
Administrator
Count me Skeptical Gene!

These schemes for vast numbers of speakers in home situations is an impractical piece of nonsense.

Dirac will not alter the speed of sound. So no matter how many speakers "cooperate" there will still be only one spot where the responses from all speakers can arrive at the same time. In any event if you are going to have that vast number of speakers, which I am certain is not required, you can afford to build a room of the correct dimensions. That will probably be cheaper than all those amps and speakers.

The fact is, that the vast majority of speakers are not very good. The advance has to be better and more powerful speakers of smaller size and footprint.

I believe the Sigberg Audio approach, of much improved small high powered speakers is the way forward to improved performance in the home.

I can tell you 11 good speakers are more than enough to give a good immersive experience throughout the listening area, and without fancy Eq. I know because I listen to it every day!
This has nothing to do with changing the speed of sound. This changes the mix to speakers so each speaker gets its normal on time signal but it also feedbacks a cancellation signal to the specific speaker that is located where a bad reflection happens. This signal is filtered and delayed unrelated to the main signal. This is full spatial bass optimization, no summing in mono needed like we've been doing.
 
CreoleDC

CreoleDC

Junior Audioholic
You missed the point. This system will allow every seat to be a good seat.
@gene, I know you told me that in a inbox message the other night when I had inbox you for recommendation on a AVR that doesn't run hot.

Gotta say though, you where spot on with your assessment of Yamahas YPAO, I did a hard reset I Did Not run YPAO 2nd time around. Sounds like the very impressive AVR when I first took it out the box. Going read the Whole article over this time around. Nothing gets pass you I of all AH members know that. "Walks away slow whistling.".
 
Last edited:
TLS Guy

TLS Guy

Seriously, I have no life.
This has nothing to do with changing the speed of sound. This changes the mix to speakers so each speaker gets its normal on time signal but it also feedbacks a cancellation signal to the specific speaker that is located where a bad reflection happens. This signal is filtered and delayed unrelated to the main signal. This is full spatial bass optimization, no summing in mono needed like we've been doing.
I will have to look into this more, but the fact remains that multiple speakers can only be timed to all give a single impulse transient at one location.

As far as reflections go, I can see how that could be ameliorated to an extent. But sound waves have vector, so no matter how much cancellation or other programs you choose, everyone in the room can not possibly get exactly the same balance of direct and reflected sound. What I do know is that the severe problems with issues of variability of sound from seat to seat are at their core caused by poor speaker performance. It is that old issue of variance of axis and off axis sound that is at the root cause of a good portion of this problem. My experience is that getting the direct and reflected sound to track closely is key here.

The other issue is that the speaker and the room are at best only 50% responsible. What I mean by this is that the recoding techniques have a huge part to play in this.
Poor microphone placement, selection and technique are a huge root cause of this problem.

This has really hit home to me this last couple of months. This is the Proms season and there have been daily or more concerts from the Royal Albert Hall since July 9 and continue until September 10. The sound, at least in my room has been incredible. All are on radio and online on iPlayer. Two each week are on TV. The BBC say they do separate mixes of the audio for TV and for audio only. Both have been superb, but the audio only stream a bit more expansive. For this year the BBC have gone back to the old Decca Tree above the conductors head, with minimal use of spot mics. I note this spot is often a mini Decca Tree. The Decca tree is central omni mic flanked by two Cardioid mics facing left and right, all about two feet or so apart. The bass is deep and natural just like the Decca recordings of years ago when they used this technique.

The point of all this is that the sound is very natural throughout the room. With the Dolby upmixer, imaging is incredibly good, not only left right but depth.

A few days ago I listened to Mahler No.2 (The Resurrection) under Sir Simon Rattle and the LSO with vocal soloists and two large choirs tiered either side of the organ, "The Voice of Jupiter".

The soloists were clearly in front, and realistically balanced. The orchestra and chorus stretched wide of the speakers, and the strings were in front, the brass and woodwind behind and percussion at the back. The choir were clearly behind the orchestra and and you could detect they were tiered way up higher than the orchestra. The sound stage was thrown way beyond the front wall of the room. The off stage trumpet, French horns and Flugel horns high up in distant high balconies were clearly reproduced and located. I think the ceiling speakers played a big part in this, and also giving that sense of the dome.

When the "Voice of Jupiter" added a huge foundation to the massive final chorus the effect was uncanny and complete.

So this convinces me that speakers, room and recording technique are the key to realistic spatial or immersion audio. This was incredibly close to the real live experience, with no boom or room affects upsetting the apple cart.

I have also been testing the new BPO Atmos immersive streams. They have not got this right at all yet. The bass does not seem right and for some odd reason the streams lack perspective and seem a bit top end heavy. The two channel stream with the Dolby upmixer is significantly more realistic with a better sense of space.
The BPO engineers need to go back to the drawing board on this.

I think we are only scratching the surface of the optimal mic techniques for Atmos concert hall recordings. I have read some research papers on mic techniques for Atmos concert hall recordings. There is no agreement. But a University study, from Sheffield, if I remember correctly, using blinded listening panels, showed a preference for a modified Decca tree mic arrangement, which I find interesting. I have to wonder if this is why, the BBC resurrected the old Decca tree for this years Proms.

My instinct is that room dimensions and treatment in the broadest sense to include furnishings carpeting, other décor, and perhaps judicious use of custom treatments as well as improved speakers coupled with better and more intelligently applied recording techniques will have the greatest impacts. Certainly my experience would suggest this is a very good and crucial foundation to progress to more realistic reproduction in the home.

I have no experience with Dirac, but for me Audyssey is a disaster and a major downgrade to quality and realism however employed. I regard that as a total dead end. However, I think properly used what we have currently at our disposal gets us a good deal of the way to hearing the recording environment imposed on out rooms, rather than the converse.
 
D

Danzilla31

Audioholic Spartan
I think an important point needs to made here.

This sounds pretty friggin cool man

That's all I gots to say :cool:
 
B

buckchester

Junior Audioholic
This article reads like an advertisement. I am sceptical of the claim to make an EQ profile that will improve every seat. Also, shouldn't these things simply focus on the frequencies that are controlled by the room (Below the room transition frequency)?

And how is one room correction tech better than another? Audyssey, ARC, Dirac, etc. As long as they have enough filters to smooth out the frequency response for the main seat below the room transition frequency, what would be the difference from one to another?
 
CreoleDC

CreoleDC

Junior Audioholic
@Danzilla31 @TLS Guy, wondering as I do think a lot about those ARC, "Auto room correction" if gravity below or above sea level would effect sound pressure.:cool:

How low a thermostat is set, I do know there's No gravity in Space so the Speed of Sound shouldn't effect time delay and reflections at all. DSP adjustments manually or done by a smart chip say, hypersonic chip plus the software encoded by a coder. Maybe @OldAndSlowDev could help, he codes all kinds of software.:)
 
Replicant 7

Replicant 7

Audioholic Samurai
@AcuDefTechGuy, at least on caught my little joke, open one window everything goes astray. Not much of anything as good as setting up your HT room right the first time around. Move, add or change out one speaker you'll have to rerun ARC all over again. I'm in Doc's conner, setup you "ROOM" first. Put the best driver's in your set-up. Andrew, isn't that what you've done with your dedicated HT room in the middle of your home. completely dimensional, no "window's" nothing on your walls. I'd bet you have a sliding door that leads to your private bathroom.
 
AcuDefTechGuy

AcuDefTechGuy

Audioholic Jedi
@AcuDefTechGuy, at least on caught my little joke, open one window everything goes astray. Not much of anything as good as setting up your HT room right the first time around. Move, add or change out one speaker you'll have to rerun ARC all over again. I'm in Doc's conner, setup you "ROOM" first. Put the best driver's in your set-up. Andrew, isn't that what you've done with your dedicated HT room in the middle of your home. completely dimensional, no "window's" nothing on your walls. I'd bet you have a sliding door that leads to your private bathroom.
No bathroom inside my HT. But just right outside of my HT. :D

Room Corrections and especially Dirac have gotten many audiophiles’ attentions especially in recent years.

Even Denon/Marantz have been rumored to offering Dirac.

My position on this is, tell everyone to try it for themselves just like everything else in life.

Personally, there’s no way I’m spending $5K for any of these features (no matter what they are) if I can do it for free or for less. :D
 
CreoleDC

CreoleDC

Junior Audioholic
No bathroom inside my HT. But just right outside of my HT. :D

Room Corrections and especially Dirac have gotten many audiophiles’ attentions especially in recent years.

Even Denon/Marantz have been rumored to offering Dirac.

My position on this is, tell everyone to try it for themselves just like everything else in life.

Personally, there’s no way I’m spending $5K for any of these features (no matter what they are) if I can do it for free or for less. :D
it sure hasn't gotten TLS guy's attention either he ain't buying into it.
But for a few dollars more huh.
 
CreoleDC

CreoleDC

Junior Audioholic
Ummmm for me the bar can be set pretty low. ;) oh sh$t we were talking about audio not my life :oops:
:D
so you believe in ARC? Do you use any of the audio room corrections? huh? or you a YPAO guys? or do you still use measuring tape and sound pressure level meter?
 
D

Danzilla31

Audioholic Spartan
so you believe in ARC? Do you use any of the audio room corrections? huh? or you a YPAO guys? or do you still use measuring tape and sound pressure level meter?
I'm really impressed with Audyssey for bass management I think all of the big boys Audyssey ARC Dirac are really good for bass management purposes
 
Matthew J Poes

Matthew J Poes

Audioholic Chief
Staff member
Count me Skeptical Gene!

These schemes for vast numbers of speakers in home situations is an impractical piece of nonsense.

Dirac will not alter the speed of sound. So no matter how many speakers "cooperate" there will still be only one spot where the responses from all speakers can arrive at the same time. In any event if you are going to have that vast number of speakers, which I am certain is not required, you can afford to build a room of the correct dimensions. That will probably be cheaper than all those amps and speakers.

The fact is, that the vast majority of speakers are not very good. The advance has to be better and more powerful speakers of smaller size and footprint.

I believe the Sigberg Audio approach, of much improved small high powered speakers is the way forward to improved performance in the home.

I can tell you 11 good speakers are more than enough to give a good immersive experience throughout the listening area, and without fancy Eq. I know because I listen to it every day!
This may work differently than you realize. I think that Dirac's use of terms is a little unfortunate and so may be causing confusion.

The core of this correction approach is not novel to DIRAC or unknown to researchers. Its a commercialization of a kind of correction known as a multi in multi out or MIMO correction. What that means is that we need to change how we think about audio going in and going out.

As a correction scheme, its a feed forward correction using minimization of the RMSE of the spatial measurements of the room against a target curve. So that is how Dirac live works as well, but Dirac live is a SIMO system and cannot redirect a reflection cancelation signal to specific speakers. Instead, response is corrected for each individual speaker and the routing in to out is essentially direct.

This system is different, you have your standard routing of the rendered channels to the appropriate output channels. You have the same kind of basic correction that Dirac LIVE applies. It is fair to say this is only correct at one location, or really, because we are minimizing RMSE for all spatial locations, its correct at no location, but most correct on the average. However, what we haven't done yet, and this new system can do, is stop the root cause of room modes.

So let's start with that, we aren't talking about the stochastic zone. The unique nature of this correction is not what is happening in the stochastic zone. It's how we handle the modal and transition zone. Evidence for that is seen in their corrections. Note that they show tons of spatial variance in the corrections above ~500hz or so, but absolutely no variance below that point. Why? Because they are feeding a cancelation signal back into a speaker that is in the right location for that. In a low channel count or stereo system, this isn't possible. But in a high channel count system in which the surround speakers are sufficiently full range and for which they have enough dynamic range, you can send a sufficiently delayed and inverted version of the signal that will fully cancel its reflection. Since you canceled the reflection at the source of the reflection, it's now correct at all locations.

So again, don't think about this in the stochastic region, think only in the modal region where canceling modes is pretty trivial. Imagine high channel count, high dynamic range, and bandwidth down to at least 50hz. Add in 4 subwoofers front and rear, as is not uncommon in high end home theaters. You can see how you can use a feed forward algo to cancel bass reflections off the appropriate boundary by using those surround speakers. You can also imagine that this works OK with 1 side surround (for lateral reflections), but better with 2, even better with 3. You can also imagine in a very high end install, you could have "surround" subwoofers along the wall. So variance is reduced at a rate of 1/n per barrier down to the unimodal region where the total of the LF sources in the room collectively can adjust for that mode.

What this will do, which is really pretty impressive, is allow you to have a vast listening area for which there is almost no variance at all. In theory, with enough speakers, you could have no appreciable variance. It's really pretty impressive. This tech has been on the horizon for quite a long time, but there are real challenges in bringing it to market.

I have the AES paper that Dirac wrote about this technology a while back and I will see if I can share it here.

For now, you should be able to access this one here:

For me, this is game changer valuable. It allows us to maintain multichannel bass. We don't have to sum to mono anymore. Any advantages of multichannel bass, such as those I noted in my article on Stereo bass (research I cited showed that our perception of spaciousness comes, in part, from lateral reflections at low frequencies) ensuring that we maintain lateral separation and remove room reflections, replacing them with only source produced LF reflections means we can accurately recreate any environment. We get flat smooth bass at every listening position, even for mother-in-laws. We maintain stereo bass improving perception of spaciousness. We have an excuse to install even more subwoofers.
 
Matthew J Poes

Matthew J Poes

Audioholic Chief
Staff member
This article reads like an advertisement. I am sceptical of the claim to make an EQ profile that will improve every seat. Also, shouldn't these things simply focus on the frequencies that are controlled by the room (Below the room transition frequency)?

And how is one room correction tech better than another? Audyssey, ARC, Dirac, etc. As long as they have enough filters to smooth out the frequency response for the main seat below the room transition frequency, what would be the difference from one to another?
Read my other post for a better characterization of what is going on. The article is an advertisement, it is just a news blip about a new product that isn't yet available to consumers. We haven't reviewed it yet, nobody has. I am the most familiar with Dirac Unison as I have contacts with Dirac and have been aware of it for some time. However, the home version has been on the horizon in their product amp forever. It's finally coming out soon to a few select pieces of hardware. I am still unclear what it will be in it's final form. It's not an easy concept to allow consumers to use. Its about the speaker locations, bandwidth, dynamic capability, etc. It was one of my criticisms to Dirac about bringing Unison to market, how do they make sure that consumers don't blow up a side surround by trying to have a 6.5" woofer in an SVS speaker produce 50hz bass at 115dB. Ain't gonna happen. But...they assure me they have fixes for this.

correct in the stochastic zone is handled differently. It's still very good. I explained it elsewhere and you can look at their white papers to understand further. They address the typical criticism by being a mixed phase system with robust handling of spatial variance.
 
newsletter

  • RBHsound.com
  • BlueJeansCable.com
  • SVS Sound Subwoofers
  • Experience the Martin Logan Montis
Top