Why can't this be easier?

[Multicore]

As I sit here having measured the size and shape of my room and its acoustic response, wondering where the reflections are coming from in an L-shaped room with a desk and all manner of other furniture in it, getting to the stage where I have to start engaging trial and error and figuring out how to do that without sticking things irreversibly to the wall or ceiling, and I can't help thinking: it's 2024, this should be at least a bit less manual by now.

I have an iPad Pro with a built-in LiDAR which can generate a 3D scan of a room. I have a measurement microphone and software that allows me to analyse the frequency response of the room. Surely it wouldn't be beyond the wit of man to combine the two? Do a 3D scan of the room, do a frequency sweep measuring from the listening position, and when you put the crosshairs on a given frequency, indicate where in the room the reflections come from for peaks and nulls? Maybe even show the effect on frequency response of adding various kinds of panel in different places in the room?

As far as my Googling shows, there isn't anything that will do the whole thing joined up end-to-end. You'd have thought a manufacturer of acoustic treatment would have realised they could earn a good deal of business by showing people how to use their products. I'd pay good money to save my time and be confident of a good result. (And they'd get my money instead of me playing around with some spare Rockslab Sonic that I had lying around to see if it's worth making panels out of.)

I'm sorry I don't have a better suggestion but, and I mean this without any sarcasm, one option is to quit fussing and that makes things really a whole lot easier.

Some people really enjoy the extremely complex task of optimizing playback in their listening space. They find the process of learning, measuring, researching, asking, etc. rewarding in addition to the practical results. One of the reasons I enjoy reading ASR is because I enjoy reading the posts from this kind of contrinutor. I admire their work. And I admire the DIY ethos.

For the most part I don't have that patience or commitment. At least, not for our home stereo. I've invested a lot in other hobbies but this one doesn't interest me as much as those.

Why do I say this? You sound a bit frustrated. Maybe that's a sign that you're not cut out for the level of nerd commitment to the cause that upper tiers of ASR represent. I'm not. And if so then one way to make this be easier is to quit fussing and enjoy your music. It's what I and plenty others did.

 

[Verig]

What Zolalll said, it's pretty interesting to first take care of your bass with a good dsp (of course proper speaker placement before that). After that listen what problems remain. Not many.

Granted, last test drive in L-shaped room was with KEF LS50 Metas + Lyngdorf (with bass management) + KC62 sub. So kind of dream combo for this stuff.

I know this is a bit beside your original question, I'm just suggesting approaching this from different angle. Physical bass management is a nightmare unless you build the room ground up. However, if you decide to build resonators etc. then good luck and have fun, this is a hobby after all.

 

[ohnonotagain]

Why do I say this? You sound a bit frustrated. Maybe that's a sign that you're not cut out for the level of nerd commitment to the cause that upper tiers of ASR represent. I'm not. And if so then one way to make this be easier is to quit fussing and enjoy your music. It's what I and plenty others did.

Maybe. I don't mind committing time and money to things when it's a good use of time. We've reached a stage with equipment where, at pretty much any budget, it's not exactly "you can't go wrong" but "it's easy to go right". The room is where the biggest improvements to the experience can be made (yes, EQ works wonders, but it's not a panacea), so that's where it's worth investing the effort. It just feels like things could be easier than they are. Whereas I don't mind spending time on something worthwhile, it is a bit frustrating that investigating, calculating and then figuring out what to do about things should be something that can be automated.

 

[Verig]

I must ask in this point do you have Dirac or comparable dsp system to tackle this? "EQ" is not really the same. Because most of what you need was automated years ago. I did this fourteen years ago. It's just not free. This is an honest question, not trying to be annoying.

 

[ohnonotagain]

I must ask in this point do you have Dirac or comparable dsp system to tackle this? "EQ" is not really the same. Because most of what you need was automated years ago. I did this fourteen years ago. It's just not free. This is an honest question, not trying to be annoying.

I tried SoundID Reference and wasn't a fan of the output it produced. Currently I'm using an EQ curve that REW generated from my measurements with some manual tweaks in the low end.

 

[ZolaIII]

It can't be automated to great extent as the think tank behind it is bad. Closed system can't think out of the system (difference between deduction and induction). It will never be able to do physical adjustments (speaker placement, isolation and such) or really critically evaluating data (to some extent maybe one day). There are quite deacent (semi) automatic room correction systems and combined with trial and error by user they give rather good results. Like Dirac and ARC to name a few. Problem is either insane license fees, very expensive hardware if you really want full capabilities or both. I am capable enough to do it by my self and do it better for much less money. It's not for everyone and do include learning and experimenting but it's not theoretical physics. Hire at least you can get a crush course guidance how to get there. How much time it takes you to adopt it and what you will choose to implement is up to you. I started hire (for the X time) with basic room modes which every room has and that aren't easy to solve. Not hard either with just cuple of PEQ's on basic level.

 

[ZolaIII]

I tried SoundID Reference and wasn't a fan of the output it produced. Currently I'm using an EQ curve that REW generated from my measurements with some manual tweaks in the low end.

What software or hardware? If software which OS?

 

[Purité Audio]

I think you are talking in an idealised scenario where the right and left reflections arrive in a similar fashion and at similar times. And where the room is of a particular size and shape, probably not L shaped like the OP. There is no doubt for me that the absorption I have makes a massive difference. To my ears and rudimentary measurements. Would you insist that absorption is moot if the right wall is closer to both speakers than the left wall?

It would depend on the REW measurements for left and right channels, to be honest I haven’t used or measured speakers with a rough off-axis for a number of years.
I certainly wouldn’t recommend automatic use of absorption as much as the manufacturers who make the stuff would have you believe is necessary.
Keith

 

[Somafunk]

As I sit here having measured the size and shape of my room and its acoustic response, wondering where the reflections are coming from in an L-shaped room with a desk and all manner of other furniture in it, getting to the stage where I have to start engaging trial and error and figuring out how to do that without sticking things irreversibly to the wall or ceiling, and I can't help thinking: it's 2024, this should be at least a bit less manual by now.

I have an iPad Pro with a built-in LiDAR which can generate a 3D scan of a room. I have a measurement microphone and software that allows me to analyse the frequency response of the room. Surely it wouldn't be beyond the wit of man to combine the two? Do a 3D scan of the room, do a frequency sweep measuring from the listening position, and when you put the crosshairs on a given frequency, indicate where in the room the reflections come from for peaks and nulls? Maybe even show the effect on frequency response of adding various kinds of panel in different places in the room?

As far as my Googling shows, there isn't anything that will do the whole thing joined up end-to-end. You'd have thought a manufacturer of acoustic treatment would have realised they could earn a good deal of business by showing people how to use their products. I'd pay good money to save my time and be confident of a good result. (And they'd get my money instead of me playing around with some spare Rockslab Sonic that I had lying around to see if it's worth making panels out of.)

Room mode calculator, also does irregular sized rooms

 

[Audio Science Pal]

Maybe. I don't mind committing time and money to things when it's a good use of time. We've reached a stage with equipment where, at pretty much any budget, it's not exactly "you can't go wrong" but "it's easy to go right". The room is where the biggest improvements to the experience can be made (yes, EQ works wonders, but it's not a panacea), so that's where it's worth investing the effort. It just feels like things could be easier than they are. Whereas I don't mind spending time on something worthwhile, it is a bit frustrating that investigating, calculating and then figuring out what to do about things should be something that can be automated.

I like your idea. The basic apparatus would probably need to be similar in design to the Klippel NFS, but reconfigured to rotate the microphone around your listening position while analyzing the test tones from your speakers in their normal locations.

 

[Verig]

I tried SoundID Reference and wasn't a fan of the output it produced. Currently I'm using an EQ curve that REW generated from my measurements with some manual tweaks in the low end.

I'm not familiar with that one. Just wondering what was wrong with it.
I see many comments saying they don't like Lyngdorf and NAD / Dirac but it's always left a bit hanging in the air what was wrong. I've had excellent results with Dirac. The main sound and integration was great and it corrected all actual underlying problems. I just applied couple of +-3 tweaks. Personal preference that one, I like it a bit hot (that minus was next to area I wanted to boost but it was better taking a notch down from the neighbour to achieve that). After Dirac did the heavy lifting it was very easy to apply small adjustments in very basic sound engineering way without any issues.

 

[No. 5]

As I sit here having measured the size and shape of my room and its acoustic response, wondering where the reflections are coming from in an L-shaped room with a desk and all manner of other furniture in it, getting to the stage where I have to start engaging trial and error and figuring out how to do that without sticking things irreversibly to the wall or ceiling, and I can't help thinking: it's 2024, this should be at least a bit less manual by now.

I have an iPad Pro with a built-in LiDAR which can generate a 3D scan of a room. I have a measurement microphone and software that allows me to analyse the frequency response of the room. Surely it wouldn't be beyond the wit of man to combine the two? Do a 3D scan of the room, do a frequency sweep measuring from the listening position, and when you put the crosshairs on a given frequency, indicate where in the room the reflections come from for peaks and nulls? Maybe even show the effect on frequency response of adding various kinds of panel in different places in the room?

As far as my Googling shows, there isn't anything that will do the whole thing joined up end-to-end. You'd have thought a manufacturer of acoustic treatment would have realised they could earn a good deal of business by showing people how to use their products. I'd pay good money to save my time and be confident of a good result. (And they'd get my money instead of me playing around with some spare Rockslab Sonic that I had lying around to see if it's worth making panels out of.

If I am understanding what you want correctly, such software packages have existed in the world of acoustics for some time. ODEON and Ease, for example, and i-Sempa is one that I know of that's free. Some even let you listen to what changes in the computer model sound like. But bear in mind that in addition to accurate room dimensions, the acoustic properties of room boundaries and furnishings must be known with accuracy for the simulation to be anywhere close to correct, otherwise you will get very precise garbage as an output.

I like your idea. The basic apparatus would probably need to be similar in design to the Klippel NFS, but reconfigured to rotate the microphone around your listening position while analyzing the test tones from your speakers in their normal locations.

An even simpler way works too. A first-order ambisonics microphone can give you arrival direction to within a few degrees, and a higher-order microphone will do even better.

 

[Audio Science Pal]

An even simpler way works too. A first-order ambisonics microphone can give you arrival direction to within a few degrees, and a higher-order microphone will do even better.

Interesting. Will that function with REW, or is there another software package for the task in question here?

 

[No. 5]

Interesting. Will that function with REW, or is there another software package for the task in question here?

Not currently an option in REW (but the inclusion of multiple mic feeds is a step in the right direction), the aforementioned ODEON will do it, as will Catt Acoustics ReflPhinder package to name two options. If you are up for more math and programing, Matlab and Octave will do it too.

 

[levimax]

If you believe Dr Toole, room measurements above Schroeder are not useful for EQ as your ears / brain does not "hear" like a MIC and can separate direct sound from room reflections very effectively. The answer is buy speakers with flat anechoic on axis response and smooth directivity and measure them in your room at the LP and use EQ to knock down any room mode peaks below Schroeder, leave the nulls alone and you are done. Anything more than that and you are just as likely to cause more harm than good. The possible exception is if you have good anechoic measurements that show some on axis deviation and the directivity is smooth you can carefully try some EQ above Schroeder. In any case there is really no reason to try to measure or calculate room response above Schroeder as it is difficult and doesn't really matter.

 

[Basic Channel]

It would depend on the REW measurements for left and right channels, to be honest I haven’t used or measured speakers with a rough off-axis for a number of years.
I certainly wouldn’t recommend automatic use of absorption as much as the manufacturers who make the stuff would have you believe is necessary.
Keith

What would you look at in REW measurements, the difference in frequency response between channels or the impulse/ETC? If it's not my massively inappropriate speaker positioning and room, then I suppose maybe it's age or a personal thing. I don't hear the speakers clearly without absorption. It's not a frequency thing but a clarity and stereo imaging difference.

 

[Basic Channel]

If you believe Dr Toole, room measurements above Schroeder are not useful for EQ as your ears / brain does not "hear" like a MIC and can separate direct sound from room reflections very effectively. The answer is buy speakers with flat anechoic on axis response and smooth directivity and measure them in your room at the LP and use EQ to knock down any room mode peaks below Schroeder, leave the nulls alone and you are done. Anything more than that and you are just as likely to cause more harm than good. The possible exception is if you have good anechoic measurements that show some on axis deviation and the directivity is smooth you can carefully try some EQ above Schroeder. In any case there is really no reason to try to measure or calculate room response above Schroeder as it is difficult and doesn't really matter.

However the Haas effect would suggest that we cannot separate early reflections from the original sound if they are within 20ms and sufficiently audible. As far as I can tell Toole's research/advice doesn't account for suboptimal speaker positioning and small or weirdly shaped rooms.

 

[Newman]

However the Haas effect would suggest that we cannot separate early reflections from the original sound if they are within 20ms and sufficiently audible.

I don't think that is a good description of the Haas Effect, starting with 20 ms actually being 30 ms, but I won't get too pedantic. Suffice to say that it is an 'equal loudness' threshold rather than a detection threshold.

The detection threshold is the bottom line on the following chart, whereas Haas effect is the top dash line. The 'image shift' threshold, which I think you mean, is the second-to-bottom line.

© Toole, Sound Reproduction, 2006​

As far as I can tell Toole's research/advice doesn't account for suboptimal speaker positioning and small or weirdly shaped rooms.

Toole's work is generally relevant to small rooms, ie domestic spaces, and he has commented on non-rectangular rooms. For example in an L-shaped room, such as @ohnonotagain described, he noted that the following subwoofer layout (2&3) greatly assisted with smooth bass, both measured and perceived:-

© Toole, Loudspeakers and Rooms for Stereophonic Sound Reproduction, 1990​

 

[ohnonotagain]

If I am understanding what you want correctly, such software packages have existed in the world of acoustics for some time. ODEON and Ease, for example, and i-Sempa is one that I know of that's free. Some even let you listen to what changes in the computer model sound like. But bear in mind that in addition to accurate room dimensions, the acoustic properties of room boundaries and furnishings must be known with accuracy for the simulation to be anywhere close to correct, otherwise you will get very precise garbage as an output.

Wow, thank you! In particular i-Simpa (linking here for others) at least looks like something I can play with. It looks like it should be possible to scan the room using something like Scaniverse, import it into i-Simpa, and understand things a bit better.

 

[goat76]

If you believe Dr Toole, room measurements above Schroeder are not useful for EQ as your ears / brain does not "hear" like a MIC and can separate direct sound from room reflections very effectively. The answer is buy speakers with flat anechoic on axis response and smooth directivity and measure them in your room at the LP and use EQ to knock down any room mode peaks below Schroeder, leave the nulls alone and you are done. Anything more than that and you are just as likely to cause more harm than good. The possible exception is if you have good anechoic measurements that show some on axis deviation and the directivity is smooth you can carefully try some EQ above Schroeder. In any case there is really no reason to try to measure or calculate room response above Schroeder as it is difficult and doesn't really matter.

Full-range measurements are still useful for figuring out where the first reflection points are in the room, what acoustic treatment may be needed to tackle those possible problems, and maybe to balance out or lower the overall response in the time domain. I think that is the main focus point for this thread.