ARTICLE: What is RoomPerfect Room Correction Software?

Apollo

Well-known Member
Getting back to the subject at hand.

Can any owners or someone more familiar with the Lyngdorf, confirm what type of calibration mic is supplied please? I am assuming Omni-directional similar to the UMIK, XTZ and every other calibration mic I’ve seen/used, or is it something else?

If so, I don’t really understand the excerpt below from the article. The difference between using the mic pointing at the ceiling (90 degrees) and directly at the speakers (0 degrees) is usually relatively small.

“A conventional room correction system places an upward-firing mic at the main listening position and takes a measurement, before repeating this process across the seating area. Firing a mic at the ceiling from the seating position will typically measure less than 10% of the sound coming directly from the speakers themselves.”
 

DT79

Well-known Member
Getting back to the subject at hand.

Can any owners or someone more familiar with the Lyngdorf, confirm what type of calibration mic is supplied please? I am assuming Omni-directional similar to the UMIK, XTZ and every other calibration mic I’ve seen/used, or is it something else?

If so, I don’t really understand the excerpt below from the article. The difference between using the mic pointing at the ceiling (90 degrees) and directly at the speakers (0 degrees) is usually relatively small.

“A conventional room correction system places an upward-firing mic at the main listening position and takes a measurement, before repeating this process across the seating area. Firing a mic at the ceiling from the seating position will typically measure less than 10% of the sound coming directly from the speakers themselves.”
It appears pretty much identical to my XTZ.

Pointing it directly forward when taking the initial ’focus’ measurement makes sense to me, you’re ensuring that the first thing that hits it is the directly radiated sound from the speakers, not myriad reflections. It may be a small difference, but it makes sense to me that it is highly significant.
 

larkone

Distinguished Member
It appears pretty much identical to my XTZ.

Pointing it directly forward when taking the initial ’focus’ measurement makes sense to me, you’re ensuring that the first thing that hits it is the directly radiated sound from the speakers, not myriad reflections. It may be a small difference, but it makes sense to me that it is highly significant.
And more simply the company that designed and wrote the software to work with the mike they have specified and supplied recommend this is the best way to measure the main focus position for their product. I would not argue with this because Lyngdorf are far more qualified to tell me how their software and hardware works to best effect.

@Apollo The best people to answer your question is Lyngdorf so my recommendation would be to email their support and ask the question
 

DT79

Well-known Member
And more simply the company that designed and wrote the software to work with the mike they have specified and supplied recommend this is the best way to measure the main focus position for their product. I would not argue with this because Lyngdorf are far more qualified to tell me how their software and hardware works.
Quite
 

jfinnie

Distinguished Member
In a similar vein to the previous questions; in Dirac you have the option of target curves, which might be accused of robbing you of the character of the speakers. What's to stop you (other than it not being an automated process) from making a target curve based on a direct measurement of the speaker in order to retain the character to some greater or lesser extent? Is this an overly simplistic view of what RP is doing?
 

DT79

Well-known Member
In a similar vein to the previous questions; in Dirac you have the option of target curves, which might be accused of robbing you of the character of the speakers. What's to stop you (other than it not being an automated process) from making a target curve based on a direct measurement of the speaker in order to retain the character to some greater or lesser extent? Is this an overly simplistic view of what RP is doing?
Nothing prevents you fro determining your own target curve and configuring Dirac to apply filters based on that. It would probably be quite a lengthy process to achieve the desired end result though, and you can't replicate the proprietary analyses and determination that RP is performing. It's not as simple as it removing the room effect entirely, as the room is necessary. It's judging what the egregious detrimental elements are and the extent to which they should be mitigated that is the difficult part to do for yourself.
 

StephanG

Active Member
In a similar vein to the previous questions; in Dirac you have the option of target curves, which might be accused of robbing you of the character of the speakers. What's to stop you (other than it not being an automated process) from making a target curve based on a direct measurement of the speaker in order to retain the character to some greater or lesser extent?
As said above, nothing is keeping you away from doing it. Does it make sense though? I'd say it's a pretty safe bet that you have never heard what your speaker sounds like and therefore you can't be robbed of the character of the speaker. What a speaker sounds like is always the result of the speaker itself in combination with the room. If you want to hear the character of your speaker, the only way to do it is either in a very, very large room (no reflections from walls/ceiling), somewhere in a desert or in a anechoic chamber.

By putting a speaker in a room, you've already changed the character of the speaker and will change again if you put the speaker in another room. That is true for every speaker, every room and every room correction system out there. Here's a very basic, non-technical article about it: Does Room Correction remove a Loudspeaker's sound characteristics? - Trinnov Audio
 

Rob Sinden

Distinguished Member
AVForums Sponsor
I assume this first in-depth article on room correction started with RoomPerfect because it is the editorial teams preferred system and the one used in the Lyngdorf MP40 and MP60 which are their highest rated Processors.

All other correction systems start with a pre set curve that will change the sound of your speakers. RoomPerfect is the only system that creates a target curve unique to your speakers and your room.

This is why it is the only correction system that preserves the character of your speakers.

It's great that this Forum will soon be putting the other systems on the market to the same scrutiny and that there will finally be some expert, independent feedback on whether this assertion is true.

:clap:
 

jfinnie

Distinguished Member
I assume this first in-depth article on room correction started with RoomPerfect because it is the editorial teams preferred system and the one used in the Lyngdorf MP40 and MP60 which are their highest rated Processors.

All other correction systems start with a pre set curve that will change the sound of your speakers. RoomPerfect is the only system that creates a target curve unique to your speakers and your room.

This is why it is the only correction system that preserves the character of your speakers.

It's great that this Forum will soon be putting the other systems on the market to the same scrutiny and that there will finally be some expert, independent feedback on whether this assertion is true.
It sounds like form your post that at a base level it is working as I was suggesting - that a target curve is extracted from the speakers and then correction is applied based on the room knowledge?

It sounds like a pro on one side and a con on the other, I can see as many reasons why you would want to keep the speaker character as why you'd want to lose it.

Drawing a parallel from video calibration the only "Character" of a display you'd want to keep are its extremes of capability - black levels, peak whites, maximum gamut, etc - everything else you'd want to align as closely as possible to a reference standard.
 

Rob Sinden

Distinguished Member
AVForums Sponsor
If you have an affordable system, I can see why changing the sound of the speakers may be a good thing.

If you have great speakers, I’d definitely not want any correction system to change the way they sound.

This is the feedback we have had from hi end hifi dealers and explains why they haven’t embraced other correction systems for hifi.
 

jfinnie

Distinguished Member
If you have an affordable system, I can see why changing the sound of the speakers may be a good thing.

If you have great speakers, I’d definitely not want any correction system to change the way they sound.

This is the feedback we have had from hi end hifi dealers and explains why they haven’t embraced other correction systems for hifi.
I guess the clear advantage is that this feature is built in to the RoomPerfect proposal, so if that is what you want to achieve then you're golden as it is built in without thinking about it. Working in pro/T&M audio mostly it's not something that stands out to me as being that useful (actually the opposite) but I can see the sales argument for it in hifi, which is by all extents a pretty odd market.

Would it be fair to say though that a calibrator who knew their way around the software would probably be able to do similar with other solutions subject to the extra effort of divining that target curve in the first place?
 

Rob Sinden

Distinguished Member
AVForums Sponsor
You are 100% right about the hifi world being an odd market, but that’s another story.

You can’t remove the room and preserve the sound of the speaker but creating a nice-looking curve. Its way more complex than you would imagine.

There is easy way to put this to the test if you have a Lyngdorf stereo amp or Processor. Run RoomPerfect as normal and measure the response with REW or similar. Now create your own Voicing to produce the curve you are looking for.

Listen to your music with RoomPerfect on and toggle between no Voicing and your new Voicing that looks great. You’ll hear it doesn’t sound great. A good looking curve doesn't sound great which is why Lyngdorf don't provide them.
 

markymiles

Distinguished Member
I actually think the voicings are a welcome addition. RP by all accounts and from all the measurements I have seen pretty much leaves the higher frequencies alone. It focuses mostly on bass integration and the lower frequencies which it is normally very good at.

Thus the maketing is indeed correct it does leave the sound signature mostly intact as it doesn't hardly touch the frequencies where you would notice.

Not dissimilar to a treble dial, albeit much more configurable with PEQ filters you can use the voicings to tweak to taste if you so wish. You could for instance have a very damped room where you naturally have a lot of roll off which robs the sounds of a bit of sparkle at the top end. Just use the voicings with a very gentle hand of course to bring it up a bit. I do agree it would be easy to mess it up, so you do need to know what you are doing.

Similar to some processors with Dirac where you can also use PEQ before or after EQ or Trinnov where it is ridiculously configurable. Best to have as much available to the end user as possible in my opinion..
 
Last edited:

StephanG

Active Member
It sounds like form your post that at a base level it is working as I was suggesting - that a target curve is extracted from the speakers and then correction is applied based on the room knowledge?
Doesn't work unfortunately. If you have a speaker and a room and you don't know the physical/acoustical characteristics of either, then you can't gain the knowledge of only one, it's always a combination of both. It could only work if you measure the speaker in a anechoic chamber first and store this information in the processor per speaker. The closest you can get with this approach is JBL Synthesis where you have to select a speaker profile in the processor. In the end, it's all just physics and physics always works and wins. Please note that just because marketing claims nonsense, that one can't be happy with the results.
 

jfinnie

Distinguished Member
Doesn't work unfortunately. If you have a speaker and a room and you don't know the physical/acoustical characteristics of either, then you can't gain the knowledge of only one, it's always a combination of both. It could only work if you measure the speaker in a anechoic chamber first and store this information in the processor per speaker. The closest you can get with this approach is JBL Synthesis where you have to select a speaker profile in the processor. In the end, it's all just physics and physics always works and wins. Please note that just because marketing claims nonsense, that one can't be happy with the results.
There are relatively affordable ways of measuring speakers repeatably outside of an anechoic chamber - I do some work for a firm who has one such solution - but RP obviously isn't using that kind of approach.
 

StephanG

Active Member
@jfinnie Yes, systems like the Klippel NFS are used to avoid the cost and complexity of a anechoic chamber by taking multiple measurements around the speakers and approximate the speaker. These spherical wave expansion functions (essentially solutions to the Helmholtz equation) can then be used to calculate the sound pressure and phase at any given frequency in space based on the approximation. Measuring the speaker at a very short distance allows to eliminate the room acoustically, as long as the room is large enough. Also, the spatial relation of speaker and microphone are exactly known. You're correct that RP or other RC systems are not using that type of approach.

For those not familiar with these systems, here's a brief (non-technical) description: https://www.klippel.de/fileadmin/kl...NFS/Logan_Klippel Near Field Scanner_2015.pdf
And a short video showing it in action:
 

jfinnie

Distinguished Member
@jfinnie Yes, systems like the Klippel NFS are used to avoid the cost and complexity of a anechoic chamber by taking multiple measurements around the speakers and approximate the speaker. These spherical wave expansion functions (essentially solutions to the Helmholtz equation) can then be used to calculate the sound pressure and phase at any given frequency in space based on the approximation. Measuring the speaker at a very short distance allows to eliminate the room acoustically, as long as the room is large enough. Also, the spatial relation of speaker and microphone are exactly known. You're correct that RP or other RC systems are not using that type of approach.

For those not familiar with these systems, here's a brief (non-technical) description: https://www.klippel.de/fileadmin/klippel/Bilder/Our_Products/R-D_System/NFS/Logan_Klippel Near Field Scanner_2015.pdf
And a short video showing it in action:
Yes, seen the Klippel gear (I think Amir at audiosciencereview has featured it if I recall). There are simpler options that also work; I won't link as I do have some relation to them; but search "Tetrahedral Test Chambers".
 

IWC Dopplel

Distinguished Member
If we avoid all of the marketing and 'positioning'

A few things that IMO need to be thought through if a deeper understanding is required

  1. Everyone who has heard their speakers has heard them in their room and potentially others, so they know the sound of their speakers in their room, not in an anechoic chamber and they may or may not like them, hopefully they know what they like/dislike them so they can be a little more strategic about how they improve the sound
  2. if your room is bad or needs some assistance to improve the sound (all/most) then paying attention to the room first is obvious and has no draw backs if done correctly.
  3. Every time processing is undertaken it will or runs the risk of degrading the signal quality and adding unwanted aberrations, thats life, nothing is perfect, so a minimal approach makes sense
  4. Solving poor bass isn't possible in 90% of situations with room treatment, so this is where DSP/EQ is the most advantageous
  5. The performance of your system- dynamics, resolution, scale, tonal accuracy etc, etc have nothing or at least very little to do with EQ/DSP.
  6. Time aligning and phase is very important but all apart from your subs include mechanical solutions to this within speaker crossovers that cant be digitally adjusted, nothing new nothing magical possible with EQ and phase, it just needs to be correct, by definition this is a compromise unless you are targeting one MLP
I suspect its more about what DIRAC/RP do well an what they dont do so will which will determine which is better and the scale of the adjustments required/desired
 

Conrad.

Moderator
Everyone who has heard their speakers has heard them in their room and potentially others, so they know the sound of their speakers in their room, not in an anechoic chamber and they may or may not like them, hopefully they know what they like/dislike them so they can be a little more strategic about how they improve the sound

This is one of the questions I had. I don't know if I like the sound of my speakers in an anechoic chamber. I know I like the sound of them in every room I've heard them in but if you remove any and all room effects they might be terrible, I wouldn't know. I know I prefer them with no EQ (manual, Audyssey, Dirac).

And while I agree that a response that's over-processed in a bid to get to a flat response can sound terrible, I think it's fairly universal that a minimally EQd, flat response is preferred. Certainly that was the result of the fairly large Harman/Toole experiment where the listeners (both skilled and unskilled, industry and non-industry) preferred the speakers with the flattest response. In that experiment it was Revel Salon IIs.


Admittedly, that was a flat starting response which would then have been affected by the room. All speakers would have had the same affect applied on them by the same room though so, in theory, the revels should have been overall flatter than the other speakers even once the room was factored in.
 

Conrad.

Moderator
Doesn't work unfortunately. If you have a speaker and a room and you don't know the physical/acoustical characteristics of either, then you can't gain the knowledge of only one, it's always a combination of both. It could only work if you measure the speaker in a anechoic chamber first and store this information in the processor per speaker. The closest you can get with this approach is JBL Synthesis where you have to select a speaker profile in the processor. In the end, it's all just physics and physics always works and wins. Please note that just because marketing claims nonsense, that one can't be happy with the results.

I always imagined it working like the Adobe Photoshop Stack function. That's able to take multiple images and remove anything that's not statistically average between them. So you can take photos of a consistent view with some moving parts (people moving in front of monuments is the common example) and it can determine the non-common items and remove them.


Not sure if you can do the same with sound, but by sampling enough places in the room you might be able to determine what's statistically probably coming from the speakers and what's probably coming from the room. Apply an inverse filter to the room effects and you can probably get close to the native speaker response.
 

IWC Dopplel

Distinguished Member
One other point that is clearly obvious is an old one. - Take a human voice or a musical instrument they are very easily distinguished and appreciated in all rooms, and that at least in part is due to how our brain and ear works, its not linear, its sophisticated and evolved, we all know that. We assess the room and adjust accordingly to some and as I understand quite a large degree, its taken millions of years for this to happen.

If the Lyngdorf solution is genuinely different in its approach seemingly trying to appreciate more of what the speaker sounds like alone and what the influence of the room might be or at least what the majority of the room influence might be, great (I think). Personally I believe this a real challenge as its a multitude of factors not easily measured or interpreted. The results however, may well be better than an auto EQ solution that might have a slightly different approach. Be good to hear more about how and why this is 'so' different

Interesting that it doesn't show the finished SPL results. This is for me of interest because it DOES tell you what SPL you have in the seating position across the frequency range and that is useful. Its useful enough in an anechoic chamber and is the basis of EQ and speaker design, so whilst it wont tell you how good a speaker is it will tell you if its starting point is limited or off. I wish there was a better measure of some other attributes that we can hear but have no real way of measuring or at least I have never seen any measurements of (real dynamics, tonal accuracy, sound steering, imaging, etc)

One thing I am uncomfortable with is the proposition that all EQ systems apart from RP 'damage' the sound of your beloved (or not) speakers. I know its marketing, with a view that someone spending £10k on a processor is dumb enough to swallow that, but that for me is dumb marketeers. Be great to have more colour and content on the RP system and evaluation against the other alternatives
 

larkone

Distinguished Member
@IWC Dopplel The issue of trying to review RoomPerfect against other systems is how do you decide that the more manual setup systems are setup to the best effect with those speakers in that room, who decides? That is the fundamental difference with RoomPerfect, it is an automatic system, with the ability to tweak voicings after setup.

At the end of the day it is all theory and conjecture until you actually have real experience of hearing what RoomPerfect can achieve in your room with your speakers, it may be next to nothing or it could be revelatory. I know that in one of my rooms with my 2170 the effect is very subtle but significant, but in the other room (a known nightmare for sound) the 1120 produces an absolutely stunning transformation - from a total mess of noise to a clear, defined soundstage with exceptional clarity and the missing bass restored.

When I chose Lyngdorfs it was because their room correction was automatic and the reviews and user reports were enough to convince me to try it and have not looked back since. What I didn't want was a system that was going to mean a high learning curve, a laptop with REW and hours trying to decide if I had finally hit the sweet spot for my speakers and room. Would rather get on and actually listen to some music. The upshot of all this rambling is, in my opinion, RoomPerfect is superb and I would recommend anyone to at least consider a trial as they may just be surprised and delighted.
 
Last edited:

StephanG

Active Member
Everyone who has heard their speakers has heard them in their room and potentially others, so they know the sound of their speakers in their room, not in an anechoic chamber and they may or may not like them, hopefully they know what they like/dislike them so they can be a little more strategic about how they improve the sound
Well, yes. In their room, but as soon as you change the room, the speaker can sound different. So it's always a fixed scenario. That's why demos in dealer rooms are a double edged sword. It tells little about how it will perform in your own setup, it could be much better or much worse. However if it's a really good demo, it at least tells you what you can expect, if you go all the way and fully optimize the room.

The performance of your system- dynamics, resolution, scale, tonal accuracy etc, etc have nothing or at least very little to do with EQ/DSP.
That is not correct, EQ/DSP can influence these things. You can also control off-axis response, shift the soundfield up/down, left right and so on. It depends on your speakers how far you can go. If your speaker is a laser beam when it comes to dispersion, no amount of EQ/DSP is going to fix that.

Time aligning and phase is very important but all apart from your subs include mechanical solutions to this within speaker crossovers that cant be digitally adjusted, nothing new nothing magical possible with EQ and phase, it just needs to be correct, by definition this is a compromise unless you are targeting one MLP
Again, this can be controlled, depending on what your speaker allows you to do. Take Ascendo CCRM speakers for example. They have active versions of these and highly recommend using these over the passive ones, because they come with a DSP that allows maximum tweakability. When you're using a Trinnov for example, you'll be fine with the passive in most cases as well. That allows you to adjust everything to a degree where the DSP is of virtually no benefit. With other processors, the internal DSP of the speakers allows it to be "lifted" to a Trinnov level. So pick your poison.


I always imagined it working like the Adobe Photoshop Stack function.
Do you mean RP? Then not really, because you can't take multiple pictures without the room. The room is always there and doesn't allow to remove parts of like with with pictures. It's more like a mix of averaging the response from multiple positions and a SLAM algorithms. There's more detail in the original paper.
 

Rock Danger

Distinguished Member
I knocked up a new version of it today in my lunch hour.
 

The latest video from AVForums

Podcast: Samsung TV Launch & QN95A Neo QLED Review, plus Film & TV news & Reviews
Subscribe to our YouTube channel

Latest News

Tidal streaming service acquired by Twitter/Square's Jack Dorsey
  • By Andy Bassett
  • Published
Netflix adds support for reduced theatrical release window
  • By Andy Bassett
  • Published
Paradigm launches Founder speaker series
  • By Andy Bassett
  • Published
AVForums Podcast: 3rd March 2021
  • By Phil Hinton
  • Published
Music revenue in U.S. sees vinyl sales overtake CD
  • By Andy Bassett
  • Published
Support AVForums with Patreon

Top Bottom