• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). There are daily reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Optimizing Speaker Placement and Listening Position using REW

tallbeardedone

Active Member
Joined
Sep 3, 2022
Messages
102
Likes
164
Hey everyone,

Last night I hooked up my Macbook Pro to my Hegel h390 integrated amp and ran frequency sweeps in REW to determine the best placement of my Monitor Audio gx300's in my listening room.

It's super easy to do and really helped me nail the perfect placement in my room resulting in the flattest frequency response. It affirmed what my ears had already determined, mainly that some rear wall reinforcement brings up the mid-bass and tames the highs in my room, and that listening slightly near-field is audio nirvana with my Monitor Audio gx300’s!

If you want to hear how I did it so you can try it for yourself, read on.

Equipment:
Umik-1 microphone ($78) (kindly loaned to me by a buddy)
Macbook Pro 13”
usb-c to usb-b cable ($20)
usb-a to usb-c adapter ($5)
tripod
furniture sliders

Software:
REW (free download)

Amplifier:
Hegel h390

Speakers:
Monitor Audio gx300

Connection:
I connected my Macbook Pro to my amplifier with a usb-c to usb-b cable. The Hegel h390 is plug-n-play so all I had to do was set it as the audio output device on my macbook and I was ready to go. I downloaded REW and opened up the application. When I connected the Umik-1 microphone to my Macbook via usb, REW found it immediately and connected.

Room:
9QTZQlfBBE8qnMCRYcL6UvvhGbmChs8pNKczRWDgUir-B9TW1A6Ealbjhake6MZRw62Y35SnjtNwKBq9B1WXcB1vXfo5fLr1GT1hCoKPx9LLg3682xUyNbvn56A8TOzIPubFTJK7Lxd_Gr-LfAWP16NFU14Kk-nEuOBKum-77QcElCQY6mPNwyy_Mg


My listening room is 5.6 m wide by 4.0 m deep with a 3.0 m opening at the back leading to my open-plan kitchen and living area. The left and front walls are covered in floor-to-ceiling bookshelves. The right wall contains a window seat and bay window covered with a velvet curtain. The floor is carpeted and the fireplace is covered with a fur rug during listening sessions.

Speaker Positioning:

Speaker distance apart: 8’7” (2.66m)
Speaker distance from side walls: 4’10” (1.47m)

The distance between speakers was selected as the widest I could get them while still maintaining a sharp and focused center image. Toe-in is around 15 degrees.

Speaker distance from the front wall was adjusted using REW sweeps to find ideal placement. Results below.

1.01m from wall (FINAL mic position):
ReEnEJmrVTiKw9ADhhBJ1-aD-el_4LnhUHbSdliKJ8Cs7Zb3qa58z3M_Rw9O2EY3NPAn4z5J419ervXQB9dJj5bB109B7lnfB3inA1SoVKX0olOe8FgD_254o6HzPjNHDae_p6tBUciK5j4RSy0prPu9XXEx4Ed1JnsZRLLjOHCfH4O7GY-1_P8SHA

This is an absolutely excellent response. The peak at 65hz is a room anti-node that cannot be attenuated without EQ, but to be honest I rather like the slam and meat it adds to the bass. The key to getting this flat response was both determining the correct distance of the speakers from the front wall AND the correct listening position relative to the speakers, which ended up being more near-field than is often suggested. The final listening position was around 30cm inside the tip of the equilateral triangle formed by the tweeters and my ears. All distortion is well below the 50dB threshold of audibility.

This exactly matches the experience I was having of leaning forward in my chair to “enter the room” of the recording as shown in the following photo.
PGpK-gyv1qgwEifn77mXBgZHn84me8sFEd5jcqMDj7PzZ6TqWjIXZXcBg8lJ5uAwJbpP8H7SeJjKvMNZN37sWyFjwzZXhUNQ0cJaKHBbWlL2MgC5-k4eg7sA5mV2y6fSiMx9BVkFkLkvPplojX2px4WvhwXSobzcJxTkQP4fSPdNeLWsP9FegC0YUg

Note how a difference of only ~30cm steps me “through the wall” and “into the room” of the recording. It really is an incredible experience.

Here’s how I got there.

1.31m from wall (START POSITION):
9xk6MS0EBZLTNZeCToanPuYUZkyj3DgqLT4jV3j0ibHocY5TnEd9vZz6vz8_A_8qKPng9MJqqLAhP7FRRk8syruQlZkIZFyEXfjtsxFPhFksJ3dH9LiGcwHyoqFBUO78g89LHBfnMyROOZOlJaB2ksIZtOG-Fo103BqSgKVhuVuCvWiGxqi-4KLQRA

I started with my speakers 1.3m from the front wall and got the above measurement.
Note that the high frequency sound pressure from 2900-10,000Hz is all above the line. At this position the speakers sound bright due to this. There is also a reasonably strong bass node (destructive interference) at 90Hz which I wanted to attenuate. The bass anti-node at 68Hz is from the room and cannot be attenuated unless I use EQ, but to be honest I quite like it as it adds meat to the vocals and some slam which helps soothe the highs.

1.21m from wall:
yVWX9PWJdEJkbqgUrIdpuUe5by6CczupeZImdx_iIZeNaOF2mIKl37DlUqtJKnZ_e6oCt6JYfiQiybvp43n96hhkbjYwOjoJY-rc6gcnTK6yjway7yJmApT9tcOWGb4TN4Brwcq_Zxua27upXy5t6_tsVzlYeoNHgYE69zRzqkYA_2zMtxvzVLCYyw

Bringing the speakers in 10cm slightly attenuated the bass suck-out, but with it came other problems. Mainly a drop in the bass and mid-bass as well as another dipped region in the mid-range. Again, this resulted in the speakers sounding a bit bright. Moving on . . .

1.11m from wall:
HO5g2mrCCrhMx6dhYH5PhO3QEN4ip_TVqZoNHFrIO8Ncf9jhgPiRvNSHGX6oWxBFand7DRlbijGTNqeP_bMlxM6MicsQ27QOVzCUFHrUf2cocNeiivNaGcH3sVYPsUlJrGAdlvBk2tklgSW5IvMO0ELssAET4u3U2ZDtLx2F8MEUPHmUNkgh7feJZA

Not much better here. Bass cancellation node is slightly attenuated but a little wider. Bass and mids still below highs.

After this I decided to go as far into the wall as possible and see how that changed things.

0.41cm from wall (as close as I could get due to bookshelves)
REHTxZgWkeGJAH9PnXhp2mRennPOhhqr7pBHBApajYVxbIpiYGJ11ppsyRFjZXTjnWD7XygmkZZDzTTpI9HIg31gfDSu-ZfuP57iXoxsJsbw4sB_-7hsnBD7YrxchqML4ovl6D5dI6tMeTQB2_5wFo_B7Vi4cxhdh72iH2Bw0uZgXri8DSXmhXnRDA


Yikes. Here there was a major base node suck-out at 140Hz. This note is at -25dB from the rest of the frequencies and would result in that note being markedly less audible. Can’t have that. Due to the rear-wall reinforcement however, the upper bass and mids were slightly higher and more evened out with the highs. Speakers sound less bright here, but I’m completely missing a 140Hz bass note.

So where is the happy medium?

0.91m from wall
x5r6EITqYZHAOhTSgd7ln3-KgHTg4BFMeYUPqX9tjxxv_ZoRWyrSIHYD_DkwdMMfHM7wqMrUq2YMjqsdNnaG9lTmVLVaUjQ_mLu0FvqVeql1Ndv3ZfCECmyZy12zJjA0_jnvhOia2tFXS0SZF9MySiBhSQPkBZ-dsb4OKECnXvoVi8DbRMkTWBTsMg

That’s better. The bass node at 130Hz isn’t as severe and the bass and mids are rising! We’re getting closer to the right spot.

1.01m from wall
ptsuacoochmpL_5WK7Kxt9jOuZaDiFoSCEUiYpODXLyN3Ut9kEsmMJIMZCrST9OX_cYyPWm_QpSsz4nbYpG_OpmMAKuB45vcuph2jCs1Qiza3JHJQ-VUNhcbZKAG6d7-e3ouAnSf7s8ozCgqN-naRhE606qIiIshf4yvZmS-JLzfJMs_kQ6iWkUN6w

Again at this distance the bass node (at 110Hz this time) is attenuated. I am still tossing up whether here or the previous measurement is the best. I have to decide which note I want the suck-out to be centered at (110Hz or 130Hz) and where the upper bass and mids sound better. 0.91m may still be better but for now I have my speakers set here.

BUT WE’RE NOT QUITE DONE!!!

Up until now I’d been measuring from a listening position centered at the exact tip of the equilateral triangle created between the tweeters and the mic. But in my listening experience, the imagine, soundstage, and realism are always best when I lean forward in my chair and put my ears more near-field to the speakers. So I moved the mic here and did a sweep.

The results, as seen at the start, speak for themselves.

1.01m from wall (sitting 30cm nearfield from tip of equilateral triangle):
VmfvzH6d7h4hIHBQ553vSSg7MBy5aS0cqUq2HpD2ggQJgaCiLOjjy3Zh4bpkXn0UgMDzY3s8MR6v4LaMw1zDj0jTWjX1cl86Do-Uh7MJNq9K3cxlhrqntkj3t2UPD71Pn-RYE2X5E4CR6HOavizLNNIyRu-Osd-IzZUHyqgZe_kyuOTNVXhHtew4cA


At the moment this is where I have my speakers positioned. It sounds absolutely incredible from the near-field listening position. I can listen at any volume without fatigue and the soundstage is incredible. I can “see” in three-dimentional space every single instrument in well-recorded material and am surrounded on all sides by ambience and spatial cues. It’s so immersive it really has to be heard to be believed so if you’re in the area, come over for a demo!

What next?
This is a continual tinker, so next I’m going to adjust by the centimeter from 1.01m to 0.91m, measuring both from tip of equilateral triangle and nearfield, to find the absolute best spot. I’m very close but there are still minute improvements to be made. There always are. That’s the joy of this hobby.

If you have any questions or comments I’d love to hear from you. Thanks for reading!

Nick
 
Last edited:

tmuikku

Member
Joined
May 27, 2022
Messages
73
Likes
72
The distance between speakers was selected as the widest I could get them while still maintaining a sharp and focused center image. Toe-in is around 15 degrees.

Hi, nice!, thank sharing it!

There is probably many things at play to get the final excellent results, so anyone doing similar experiments with their space and speakers could perhaps benefit from extra information.

How did you arrive to the 2.66m distance and sharp image, what was the procedure here? Was this also a small difference when it happens and when it doesn't? Have you found out getting this sharp phantom image is first thing on the complete procedure?

How did you arrive to the toe-in? Was it part of the phantom image procedure or just the frequency balance? This might be different for some other set of speakers/directivity. While it relates to the frequency balance it also affects first reflections. Have you tried this on other speakers? did you get similar results?

How about height of the room? Do you have insight how it might affect the setup/procedure? You have quite long distances to boundaries which is probably mandatory to get the results (with so big listening triangle), what do you think? Front wall is about 1m behind speakers makes ~6ms delay, sidewalls are ~10ms out, ceiling looks high, probably also +6ms.

And one more, have you tried the psychoacoustic smoothing on rew?

Thanks! :)
 
Last edited:
OP
T

tallbeardedone

Active Member
Joined
Sep 3, 2022
Messages
102
Likes
164
Hi, nice!, thank sharing it!

There is probably many things at play to get the final excellent results, so anyone doing similar experiments with their space and speakers could perhaps benefit from extra information.

How did you arrive to the 2.66m distance and sharp image, what was the procedure here? Was this also a small difference when it happens and when it doesn't? Have you found out getting this sharp phantom image is first thing on the complete procedure?

How did you arrive to the toe-in? Was it part of the phantom image procedure or just the frequency balance? This might be different for some other set of speakers/directivity. While it relates to the frequency balance it also affects first reflections. Have you tried this on other speakers? did you get similar results?

How about height of the room? Do you have insight how it might affect the setup/procedure? You have quite long distances to boundaries which is probably mandatory to get the results (with so big listening triangle), what do you think? Front wall is about 1m behind speakers makes ~6ms delay, sidewalls are ~10ms out, ceiling looks high, probably also +6ms.

And one more, have you tried the psychoacoustic smoothing on rew?

Thanks! :)
Hey mate,
Thanks for reading. The long answers to all your questions are in two detailed documents I've created, Optimizing Sounstage in your Listening Room, and Optimizing Speaker Placement using REW. If you're keen on reading message me your email address and I'll add you to the docs.

Short answers are:

Speaker Distance Apart (and toe-in)
Speaker distance apart was carefully determined using "Little Room" by Norah Jones, which has a very centered vocal. I want my speakers as WIDE apart as possible to give me the widest possible soundstage, but too wide and the center image disappears. I started 1/5 room width from side walls (nearly 11ft apart) but this was too wide. I think of the sound emanating from the speakers as a widening cone, and both cones need to overlap, just slightly, at the listening position to have a sharp center image, as demonstrated in the photos below:
Screen Shot 2022-10-23 at 8.26.03 AM.png
Screen Shot 2022-10-23 at 8.28.29 AM.png

Note that with proper toe-in you can get the cones to overlap with the speakers wider apart, thus giving you the best of both worlds, WIDE soundstage and FOCUSED center image. This is how I ended up at my final width of 2.66m (8'8"), and a toe-in of roughly 15 degrees. Once I found the exact point where the image comes apart (about 9ft apart in my room) I used toe-in to get it back.

Height of Room:
My ceilings are flat and 2.7m high (9ft). This is the furthest hard reflective surface from the tweeters so time delay of first reflections is roughly 10ms, as you stated, so my ears can easily distinguish the source from reflected sound. I currently DO NOT have ceiling treatment, but my summer project is to build some acoustic panels similar to what I've installed in my theater room. This should improve the sound even further and I'll update the doc and all the REW graphs once they're installed!

The real take-away should be the MASSIVE difference in sound quality depending on speaker placement and room treatment. So many audiophiles spend thousands of dollars to, for example, tame the highs in their room, when the real and FREE answer is determining the optimal room placement for your speakers and taming reflections.

Again, if you want to go full nerd and read/comment on my docs, send me your email. Or if you have any more questions ask. I really enjoy nerding out on this stuff.
 

Attachments

  • Screen Shot 2022-10-23 at 8.26.31 AM.png
    Screen Shot 2022-10-23 at 8.26.31 AM.png
    201.9 KB · Views: 42
Last edited:

tmuikku

Member
Joined
May 27, 2022
Messages
73
Likes
72
Hi, thanks for the docs!

I got stereo setup going few months ago for the longest time, was few years in single speaker mono :) Short story, initially setup the stereo system too wide apart and center image was blurred. Disappointed to the performance having used to the single mono speaker sharp "image" started bringing speakers closer together and center phantom image got sharper. Further experimenting noticed the same thing as you, listening too far out (roughly tip of equilateral triangle) and the phantom image was kind of behind a veil, 2D soundscape as in your descriptions. Moving along the center normal between the speakers, straight forward from tip of equilateral triangle closer to speakers like you, I also got this "step into the sound" effect, and it was quite obvious when it happens.

Well, I'm very interested about what happens and why and how it all relates to written descriptions of perceived sound qualities in forums and papers. I have listening setup in family living room and the very good stereo image you describe happens with setup that is not practical for the situation (family living room) and thus my current task is try and figure out how to extend the effect further into the room so that it would happen at practical listening position. Hence trying to get more insight what makes the effect to be able to develop speakers so that they fit to my application better :) I thought this is very close to what you have written here, we have the same goal but bit different premises. I set out some quick experiments and got some observations that might be relevant for the topic so posting them here if anyone has had similar observations, or comments. I'm hoping we'd gather more info on this so that I, or anyone, could have enough insight to get the effect happen in their systems and listening applications.

About experiments and observations:
I varied spacing of the speakers some, as well as their relation to room boundaries but the step inside the room effect stayed pretty much the same in a way that it seemed to happen when distance from head to speaker was pretty much bang on 2 meters in my case. So, when the speakers were further apart I needed to go closer in, and when they were closer together the effect extended a bit further into the room. Still, could not get the effect to extend to practical main listening position (family livingroom) without moving the speakers middle of the floor. So, the effect appeared to be somewhat independent of the system placement in room but have not experimented further yet. I suspect height of the room and vertical directivity matters here. My speakers are DIY narrowish ~constant directivity, DI past 6 < 1kHz and frequency balance seemed to affect quite much, toe-in affected so that sharpest image was always happening when ear was on-axis and my takeaway from this is that a good sharp image requires proper frequency balance (enough highs) to happen.

I also noticed that the image/soundstage could be too wide, I didn't like to have it too wide, perhaps 45 degrees or so to each speaker seemed to be natural for me. This might be function of the directivity, too wide was just too weird and good was found experimenting. Do you have insight into this, ever got too wide image or have you paid attention how wide it is?

Contrary to your setup my setup is asymmetric in room, its huge open plan living room / kitchen, also the speakers are butt against the (front) wall as they are designed so. Still similar perceived stereo image thing happened, so I think appearance of the good stereo image is function of early reflections time and level, perhaps their frequency spectrum and direction as well.

Also, thinking what happens when moving head in and out from the "room of sound" and how small the change in physical location is in relation to your and my speakers smooth directivity, it seems that its not frequency spectrum changes that make the effect appear but most probably timing of early reflections, perhaps their level and spectrum accounted in. Toe-in affects spectra and level of early reflections as well, but not their timing.

Anyway, Its quite astonishing how small change (in listening position) makes some process in brain have very different results for perceived sound quality. I think its just some threshold that gets passed and suddenly there is clarity for the phantom images and envelopment and what have we adjectives. I think that what happens is the local room localization cues get perceived low enough so brain is able to lock into the localization cues in the source material the image happens. All this is speculation though, haven't had a chance to make more experiments yet.

Well, all the details are still missing :D How to translate any of this into practical sound system, how to choose or build a system that matches our listening situation so that the image happens. I think method like you have described could help almost anyone to get similar results, downside being the positions end up being middle of room and require dedicated listening space or movable setup. I suspect and hope it is possible to have speakers with such directivity that the effect would happen far enough so that speaker position and listening position could be relaxed some, to be practical in a family living room.

What do you think? Have you experimented with asymmetric setup, or narrower directivity systems? Have you experienced the sound anywhere else, like in hifi shows or live events?

Thanks!
 
Last edited:

tmuikku

Member
Joined
May 27, 2022
Messages
73
Likes
72
It would be interesting to know how common this is, or rare. Anyone else have this kind of very good stereo setup happening?:)

I guess many people haven't heard it, or consider the 2D image fine enough, or blurred kind of 3D envelopment. I haven't been into Hifi shows so I don't know how common it is among the other enthusiast and professionals and the likes. To me this kind of very good imaging is the main goal with hifi stereo setup, its The thing and very happy to have got it with DIY setup.
 

Alexium

Active Member
Joined
Sep 22, 2019
Messages
201
Likes
113
Location
Ukraine
Its quite astonishing how small change (in listening position) makes some process in brain have very different results for perceived sound quality. I think its just some threshold that gets passed and suddenly there is clarity for the phantom images and envelopment and what have we adjectives.
This description fits my experiments with the left channel delay in my car. I have upgraded the audiosystem in my old Audi sedan and added some DSP. As you know, in a left-hand drive car you're sitting close to the left and far from the right speaker. You can (and should) set the balance (levels) so that you're evenly submerged into the bubble of sound, but that's not enough for a sharp center image - you need to also delay the left channel so that it appears at the same "virtual" distance. And as you're adding the delay in 1 ms or 0.5 ms steps, there is a pretty sharp transition to when it "clicks" and the blurry center image becomes localized. The delay required is 5-6 ms in my case, and the width of the sharpness window is ~2 ms. That's equivalent to 60 cm distance mismatch over the total of 1.7 meters.

I guess many people haven't heard it
I know I haven't! Not even in the car, I can't say the soundstage is anything special even with all the delays set. Probably due to suboptimal placement of the speakers. And I have followed one guide at positioning speakers at my home, which describes this same effect of instant sharpness, but I didn't get no matter how I placed the speakers. Perhaps, the choice of the music material was wrong.
 
Last edited:

tmuikku

Member
Joined
May 27, 2022
Messages
73
Likes
72
Perhaps, there could be many things like ability to hear it, perhaps its something that is learned? Perhaps its already there but its just hard to connect descriptions of others to actual perceived sounds you hear. Many things mysterious and interesting :)

I think, it requires the step in/step out situation for anyone to realize its there. If there is no stark contrast being in or out of the sound while moving along center normal I can imagine its hard to know if the phenomenon is there or not. Its quite apparent when it is, literally one step closer to speakers and I was inside the performance or out of it with one step back.

One thing I forgot to write earlier, I got new waveguides shortly after got the stereo system running and did the above experiments and observations and of course the distinct step in/out phenomenon somewhat disappeared :D So, when the effect appears dramatic perhaps its function of frequency spectrum and early reflections playing well together so that multiple things snap in place at the same moment which makes big enough difference combined its easy to hear.
 
Last edited:

tmuikku

Member
Joined
May 27, 2022
Messages
73
Likes
72
What I would do, if never heard the effect, I would try find the setup other way around as well, set some of the first reflection delay times as anchor. Place speakers ~1.5m from sidewalls, ~1m from front wall as tallbeardedone has done, to get most early reflections past many milliseconds. Temporarily move the furniture around to allow it happen, if necessary. This might yield very small stereo triangle if the room is small. If room is big make the triangle no bigger than say 2 meters. Then start moving back and forth on the center normal, in other words line thats exactly between the speakers. Experiment with toe in. Try to concentrate to hear the effect of 2D and 3D sound in a way, that when you are far out the image is kind of 2D, sounds appear coming from the front of you like watchin a movie, and when getting closer to speakers concentrate if "envelopment" happens, in other words listen if sound seems to come also from sides / back as well, not just on the front, envelopment. You might need to get between the speakers even :D If its always 3D, sound coming all around, your listening triangle is probably too big and its just the reverberation of room that you always hear. Need to get the ratio of direct sound to reverberant sound higher by getting speakers closer. Well, not sure whats the best way to experiment with this, but start with very small stereo triangle and expand if the effect is not there. Perhaps some systems / rooms or combination of just aren't up to the task.

As for source material one could use some favorite music that has some kind of nice roomy sound to it on headphones for example. Usually singing voice is on the middle and human voice is also very familiar to us so listening to the singing would be first I'd concentrate to, sharpness/blurriness of it. Even if the full mind blowing effect would not appear one would still learn to listen how the sound changes when closer or further out from the speakers.

Another thing to test is how good the symmetry is in room and that the speakers match up. Play mono pink or white noise to both speakers, same exact signal to both. Listen if the phantom center is good. The noise should appear to be exactly between the speakers and the speakers should appear to be silent, even if you look at one of the speakers it should appear to be silent, or almost silent. If this is not happening, perhaps there is excess reflections / diffraction happening, or differences on the speakers spectra that reveal their position and which could prevent the imaging happening. Perhaps again the listening triangle is too big for the room. Do this test standing quite close to speakers, not at the tip of listening triangle but closer to or almost between the speakers. When you get strong phantom center with this test I see no reason the good 3D imaging would not happen as long as your hearing system is capable of comprehending it and the listening triangle is not too big for the room so that the early reflections get out of way and brain could lock into the cues on the recording. I guess this achieves similar thing as tallbeardedone Norah Jones test.

It could be this is so common phenomena people just can't connect the words here to perceived sound :) Or it is so rare not many has heard it, which also makes it hard to connect words to perceived sound. Only way to get there is to experiment with positioning I think, to hear it yourself. Take free weekend and spend it to critical listening and moving the setup around until something good happens :) If nothing happens, hopefully it was fun time with the hobby at least.
 
Last edited:

tmuikku

Member
Joined
May 27, 2022
Messages
73
Likes
72
Well, anyway, if someone that has never heard the effect and then hears it, there is no returning to lessor setup :D So, if Hifi is important hobby I accompany tallbeardedone and suggest everyone to spend time with the placement and try find the magical setup to get a chance to step into the performance, even if it wasn't long sustained like completely turning the furnishing around for a weekend. Then, join my quest to find properties of speakers that would make the thing work in a practical room and listening setup :)
 

Alexium

Active Member
Joined
Sep 22, 2019
Messages
201
Likes
113
Location
Ukraine
I very much endorse the use of pink / white noise for tests like this. It's much better than music for some calibration / measurement purposes.
 
OP
T

tallbeardedone

Active Member
Joined
Sep 3, 2022
Messages
102
Likes
164
Hi, thanks for the docs!

I got stereo setup going few months ago for the longest time, was few years in single speaker mono :) Short story, initially setup the stereo system too wide apart and center image was blurred. Disappointed to the performance having used to the single mono speaker sharp "image" started bringing speakers closer together and center phantom image got sharper. Further experimenting noticed the same thing as you, listening too far out (roughly tip of equilateral triangle) and the phantom image was kind of behind a veil, 2D soundscape as in your descriptions. Moving along the center normal between the speakers, straight forward from tip of equilateral triangle closer to speakers like you, I also got this "step into the sound" effect, and it was quite obvious when it happens.

Well, I'm very interested about what happens and why and how it all relates to written descriptions of perceived sound qualities in forums and papers. I have listening setup in family living room and the very good stereo image you describe happens with setup that is not practical for the situation (family living room) and thus my current task is try and figure out how to extend the effect further into the room so that it would happen at practical listening position. Hence trying to get more insight what makes the effect to be able to develop speakers so that they fit to my application better :) I thought this is very close to what you have written here, we have the same goal but bit different premises. I set out some quick experiments and got some observations that might be relevant for the topic so posting them here if anyone has had similar observations, or comments. I'm hoping we'd gather more info on this so that I, or anyone, could have enough insight to get the effect happen in their systems and listening applications.

About experiments and observations:
I varied spacing of the speakers some, as well as their relation to room boundaries but the step inside the room effect stayed pretty much the same in a way that it seemed to happen when distance from head to speaker was pretty much bang on 2 meters in my case. So, when the speakers were further apart I needed to go closer in, and when they were closer together the effect extended a bit further into the room. Still, could not get the effect to extend to practical main listening position (family livingroom) without moving the speakers middle of the floor. So, the effect appeared to be somewhat independent of the system placement in room but have not experimented further yet. I suspect height of the room and vertical directivity matters here. My speakers are DIY narrowish ~constant directivity, DI past 6 < 1kHz and frequency balance seemed to affect quite much, toe-in affected so that sharpest image was always happening when ear was on-axis and my takeaway from this is that a good sharp image requires proper frequency balance (enough highs) to happen.

I also noticed that the image/soundstage could be too wide, I didn't like to have it too wide, perhaps 45 degrees or so to each speaker seemed to be natural for me. This might be function of the directivity, too wide was just too weird and good was found experimenting. Do you have insight into this, ever got too wide image or have you paid attention how wide it is?

Contrary to your setup my setup is asymmetric in room, its huge open plan living room / kitchen, also the speakers are butt against the (front) wall as they are designed so. Still similar perceived stereo image thing happened, so I think appearance of the good stereo image is function of early reflections time and level, perhaps their frequency spectrum and direction as well.

Also, thinking what happens when moving head in and out from the "room of sound" and how small the change in physical location is in relation to your and my speakers smooth directivity, it seems that its not frequency spectrum changes that make the effect appear but most probably timing of early reflections, perhaps their level and spectrum accounted in. Toe-in affects spectra and level of early reflections as well, but not their timing.

Anyway, Its quite astonishing how small change (in listening position) makes some process in brain have very different results for perceived sound quality. I think its just some threshold that gets passed and suddenly there is clarity for the phantom images and envelopment and what have we adjectives. I think that what happens is the local room localization cues get perceived low enough so brain is able to lock into the localization cues in the source material the image happens. All this is speculation though, haven't had a chance to make more experiments yet.

Well, all the details are still missing :D How to translate any of this into practical sound system, how to choose or build a system that matches our listening situation so that the image happens. I think method like you have described could help almost anyone to get similar results, downside being the positions end up being middle of room and require dedicated listening space or movable setup. I suspect and hope it is possible to have speakers with such directivity that the effect would happen far enough so that speaker position and listening position could be relaxed some, to be practical in a family living room.

What do you think? Have you experimented with asymmetric setup, or narrower directivity systems? Have you experienced the sound anywhere else, like in hifi shows or live events?

Thanks!
Very cool to hear of your experiences! Thanks for sharing. At one stage I had my room set up along the left wall so that the side walls were asymmetrical. The right side was bookshelves and the left open into the living area. This was not ideal. I could still get a very focused center image and soundstage placement was good (I think this has much to do with the excellent horizontal dispersion of my particular speakers, among other things) but I didn't get quite the same ambient/enveloping experience. I think that was do with uneven reflections off the side walls but I can't be sure.

By moving the speakers to the wall they are currently on, I now get sound all the way down the length of the room. As far as you wanting to "aim" your stereo image further into the room, this can be done with toe-in, but there will still only be ONE position where you get the perfect enveloping 3-D soundstage. This is due to the physics of stereo imaging. The cones of sound radiating from the speakers need to precisely overlap, just slightly, at the listening position and ideally the side walls and ceiling would be equi-distant and diffused. At least that's what I've found in my room. If I stand along the center line of my room anywhere along it's length I can "see" the center image and the horizontal placement of performers, but only at the sweet spot do I get the 3-D enveloping soundstage where I can point out every instrument in 3-D auditory space and hear the ambient noise surrounding me on all sides. I think of it as positioning my head exactly where the performer was relative to the stereo mics. The speakers are now radiating the exact sound the mics recorded and I am hearing the exact sound as if I was "in the room". Very very cool.

As far as sounstage width, the limiting factor in speaker width was determined by how far I could get them apart and still get a focused center image. I like the WIDEST possible soundstage and never got to a point where I had a focused center image but felt the soundstage was too wide. The wider apart the speakers the more separation between performers and the easier it is to exactly locate them in 3-D sound space. My speaker are now 8'9" (2.66m) wide.

Did you build your speakers? If they are very directional then I would say you'll have to point them more-or-less directly at the listening position and only at that point where they intersect will you get the ideal sound. My 3-D soundstage has a lot to do with the excellent horizontal dispersion of my speakers I think. In the example in the doc it's very easy to hear depth from the sweet spot slightly near-field. I can exactly place each instrument in good live recordings. It really is very cool and easily noticeable to anyone who sits in the chair. It's all about the physics!

Again, thanks for your post!
 
Last edited:
OP
T

tallbeardedone

Active Member
Joined
Sep 3, 2022
Messages
102
Likes
164
It would be interesting to know how common this is, or rare. Anyone else have this kind of very good stereo setup happening?:)

I guess many people haven't heard it, or consider the 2D image fine enough, or blurred kind of 3D envelopment. I haven't been into Hifi shows so I don't know how common it is among the other enthusiast and professionals and the likes. To me this kind of very good imaging is the main goal with hifi stereo setup, its The thing and very happy to have got it with DIY setup.
Not many people have heard this. My local hi-fi dealer came around (he has $200k of equipment in his apartment) and he had never heard 3-D imaging like this. Speaker placement within the room, acoustic treatment of the room, dimensions of the room, and listening position relative to speakers and walls all need to be taken into account. But to be honest it's not that hard. I'm thinking of providing a service here in nz to help optimize their rooms for the 3-D experience.
 
OP
T

tallbeardedone

Active Member
Joined
Sep 3, 2022
Messages
102
Likes
164
I very much endorse the use of pink / white noise for tests like this. It's much better than music for some calibration / measurement purposes.
I used frequency sweeps. Can you detail how you would use pink noise to determine ideal placement? Or do you just mean for center imaging?
 
Last edited:

tmuikku

Member
Joined
May 27, 2022
Messages
73
Likes
72
Very cool to hear of your experiences! Thanks for sharing. At one stage I had my room set up along the left wall so that the side walls were asymmetrical. The right side was bookshelves and the left open into the living area. This was not ideal. I could still get a very focused center image and soundstage placement was good (I think this has much to do with the excellent horizontal dispersion of my particular speakers, among other things) but I didn't get quite the same ambient/enveloping experience. I think that was do with uneven reflections off the side walls but I can't be sure.
Yeah asymmetric setup is not perfect I think, but nevertheless got the 2D / 3D transition effect going. I'm not sure of course if my 3D effect goes to extent as you have achieved, this is something thats really hard to convey through words and relate to perceived sound, especially if there is no such experience where connect to :)

By moving the speakers to the wall they are currently on, I now get sound all the way down the length of the room. As far as you wanting to "aim" your stereo image further into the room, this can be done with toe-in, but there will still only be ONE position where you get the perfect enveloping 3-D soundstage. This is due to the physics of stereo imaging. The cones of sound radiating from the speakers need to precisely overlap, just slightly, at the listening position and ideally the side walls and ceiling would be equi-distant and diffused. At least that's what I've found in my room. If I stand along the center line of my room anywhere along it's length I can "see" the center image and the horizontal placement of performers, but only at the sweet spot do I get the 3-D enveloping soundstage where I can point out every instrument in 3-D auditory space and hear the ambient noise surrounding me on all sides. I think of it as positioning my head exactly where the performer was relative to the stereo mics. The speakers are now radiating the exact sound the mics recorded and I am hearing the exact sound as if I was "in the room". Very very cool.

As far as sounstage width, the limiting factor in speaker width was determined by how far I could get them apart and still get a focused center image. I like the WIDEST possible soundstage and never got to a point where I had a focused center image but felt the soundstage was too wide. The wider apart the speakers the more separation between performers and the easier it is to exactly locate them in 3-D sound space. My speaker are now 8'9" (2.66m) wide.
This is the part that lacks some detail, how the speaker directivity actually is involved. Directivity and toe-in does not change any of the reflection times/delays involved but affects their frequency response and level. The simplification of thinking with radiating cones is very nice as its easy to imagine and work with and seems to yield results at least in your room and speakers. But in reality there is no such cones and the speaker emits sound to all directions "seamlessly", its just that frequency response changes to any angle, low bass being about constant to all directions and usually with higher frequency attenuation happens the greater the angle is. Basically the top octave goes first when moving from on-axis to greater angles and when behind the speaker there is not much wavelengths shorter than the baffle. Then there can be some undulations on the mids depending on how good the speaker construct is acoustically, could be smooth, could be quite dramatic. It gets quite compelx quick, what the sound is toward what specular reflection and how they would affect. We can perhaps simplify and assume that if the walls are far enough, or the sound toward the wall is low enough in level and similar enough in frequency response to the direct sound, brain has easier time identifying the reflections from direct sound and filtering them out allowing the cues from source material to be perceived clearly.

Perhaps the cones represent usable listening window, where there is enough of the highest highs? Thats easy to work with, listening window.

I think you are right about the envelopment part, system being symmetric in room and head in particular position the envelopment/reflections snap in place. Also image width you are getting makes sense for wide coverage angle speakers I guess, or different kinds of speakers in combination with room might have various issues. For example, I got better and better envelopment the further out to my sides the speakers were, or the closer I was on the center normal, but now the image went too wide, speakers too wide apart. I guess as my room placement is asymmetric and the speakers have relatively narrow directivity there just isn't enough sound in the room to make the envelopment "loud enough" when the speakers are on front and image is just wide enough and I'm lacking some of the envelopment part.

On the other hand you cannot get the speakers far apart enough? and I think its because you seem to get too much envelopment if they were any wider. But as you achieve the phenomenon you are able to make the speakers just the right amount apart. Reasoning from this I think my system would benefit wider horizontal coverage to increase sound in room and have more envelopment without taking them speakers too wide apart. The closer the speakers were together the further out into room the effect seemed to extended I remember. Perhaps I should widen horizontal coverage and reduce vertical to get the effect further into the room, at least I suspect vertical reflections / directivity matters how far the sharp image and thus 3D effect extends.

Similar reasoning applied to your situation, if you wanted wider soundstage, push speakers further apart, you might benefit from narrower coverage angle speakers, or some more acoustic treatment I think. Perhaps the ceiling acoustic treatment you are planning achieves this, very interesting to see how it affects the system if it extends your listening triangle.

And reasoning from thoughts and data here so far it looks to me there is three major contributors for the 3D image that need to happen: proper direct sound frequency balance (enough highs), sharp phantom image (direct / reverberant sound ratio, size and symmetry of the listening triangle, good speakers that are alike), and fine envelopment. These are just my speculation so feel free to comment :D By fine envelopment I mean nice late reverberation and symmetry of early reflections, which might mean suppressed early reflections or delayed enough early reflections, so that early reflections are not a mess. My intuition says when early reflections are benign the brain is not focusing to them but to the source material cues instead.

Did you build your speakers? If they are very directional then I would say you'll have to point them more-or-less directly at the listening position and only at that point where they intersect will you get the ideal sound. My 3-D soundstage has a lot to do with the excellent horizontal dispersion of my speakers I think. In the example in the doc it's very easy to hear depth from the sweet spot slightly near-field. I can exactly place each instrument in good live recordings. It really is very cool and easily noticeable to anyone who sits in the chair. It's all about the physics!

Again, thanks for your post!
Yeah, they are not very directional but some, cardioid mid and ~90deg waveguide so DI around 6db at ~250Hz and smooth slope up to 10db at 20kHz. Its DSP system so I can EQ/optimize for any axis to be of any slope so can still experiment with toe-in, perhaps my EQ and toe-in is not optimal now but need to experiment further some day when there is time for critical listening. A grand mother and two active kids under same roof makes it almost impossible to have critical listening more than 30 seconds unless everyone else is out :D

ps. sorry for long posts but I feel the stuff is in heart of any Hifi setup and there is not too much this kind of indepth descriptive texts, that would be easy to translate into my own setup / system. Stuff is often either marketing or too shallow so I cannot relate it to my situation, or perhaps too deep and just cannot understand :D Or perhaps there is and I've been just TL;DR :D I was excited seeing you write about the 2D/3D transition as its something I've never saw others write about earlier, so thanks writing it and letting me outlet my thoughts on it.
 
Last edited:
OP
T

tallbeardedone

Active Member
Joined
Sep 3, 2022
Messages
102
Likes
164
Yeah asymmetric setup is not perfect I think, but nevertheless got the 2D / 3D transition effect going. I'm not sure of course if my 3D effect goes to extent as you have achieved, this is something thats really hard to convey through words and relate to perceived sound, especially if there is no such experience where connect to :)
Don't apologize I'm enjoying nerding out with someone who is as into this as I am, and who, like you said, shares a similar experience. It's so freaking cool and there's so much BS and wasted money in hi-fi it's great to find a like mind. I love learning and exploring the science the same way you do.

This is the part that lacks some detail, how the speaker directivity actually is involved. Directivity and toe-in does not change any of the reflection times/delays involved but affects their frequency response and level. The simplification of thinking with radiating cones is very nice as its easy to imagine and work with and seems to yield results at least in your room and speakers. But in reality there is no such cones and the speaker emits sound to all directions "seamlessly", its just that frequency response changes to any angle, low bass being about constant to all directions and usually with higher frequency attenuation happens the greater the angle is. Basically the top octave goes first when moving from on-axis to greater angles and when behind the speaker there is not much wavelengths shorter than the baffle. Then there can be some undulations on the mids depending on how good the speaker construct is acoustically, could be smooth, could be quite dramatic. It gets quite compelx quick, what the sound is toward what specular reflection and how they would affect. We can perhaps simplify and assume that if the walls are far enough, or the sound toward the wall is low enough in level and similar enough in frequency response to the direct sound, brain has easier time identifying the reflections from direct sound and filtering them out allowing the cues from source material to be perceived clearly.

Perhaps the cones represent usable listening window, where there is enough of the highest highs? Thats easy to work with, listening window.
You're 100% correct about the 3-D sound wave propagation and attenuation of high frequencies before low. The cone analogy is exactly that, where the frequency response is still flat enough that you're hearing all the frequencies. And for my speakers in particular, attenuating the highs so the frequency response tails off as you climb is actually where MY speakers sound very good. This may (will) not be the case with every speaker. That's what I meant by speakers with less ideal horizontal dispersion might have to be toed in more, as the "listening window cones" will be narrower, and thus more toe-in will be needed to get that stereo overlap we're after. Not sure if that makes sense I can do a diagram to explain if need be.
I think you are right about the envelopment part, system being symmetric in room and head in particular position the envelopment/reflections snap in place. Also image width you are getting makes sense for wide coverage angle speakers I guess, or different kinds of speakers in combination with room might have various issues. For example, I got better and better envelopment the further out to my sides the speakers were, or the closer I was on the center normal, but now the image went too wide, speakers too wide apart. I guess as my room placement is asymmetric and the speakers have relatively narrow directivity there just isn't enough sound in the room to make the envelopment "loud enough" when the speakers are on front and image is just wide enough and I'm lacking some of the envelopment part.
I'm not quite sure how the physics of the ambience works yet, but you've experience it too. There's a definite spot in the room where it feels like I'm sticking my head "through the wall" and "into the room" where the recording is taking place. It's uncanny. I can heard ambience BEHIND me right now as I'm listening. No idea how that works but it makes sense that reflected sound is attenuated so maybe it only adds that "hush" of space or ambience? I can't explain it yet and would like to hear more on what you think may cause it.

One possibility I've thought of is phase difference. When sound waves reflection off surfaces they reflect 180 out pf phase with the incident wave. I know Roger Waters uses out of phase sounds to throw images around the sound stage. In "The Ballad of Bill Hubbard" that radio is way out to my left in my room, and that dog is barking BEHIND my right shoulder in the back yard. So maybe the phase difference of reflected sound when it reaches your ears at exactly even times somehow adds to the 3-D ambience effect. Not sure just throwing it out there.

But I'm pretty sure the even distance to side walls helps in this regard as it's even more convincing now that I have symmetrical walls on both sides. I'm not sure about that though. I also don't feel like rearranging my room again to check.:rolleyes:
On the other hand you cannot get the speakers far apart enough? and I think its because you seem to get too much envelopment if they were any wider. But as you achieve the phenomenon you are able to make the speakers just the right amount apart. Reasoning from this I think my system would benefit wider horizontal coverage to increase sound in room and have more envelopment without taking them speakers too wide apart. The closer the speakers were together the further out into room the effect seemed to extended I remember. Perhaps I should widen horizontal coverage and reduce vertical to get the effect further into the room, at least I suspect vertical reflections / directivity matters how far the sharp image and thus 3D effect extends.
I wasn't super clear when I said they can't be far enough apart. They can. As I said in the doc I tried 11ft between them (1/5 room width from side walls) and there was no stereo image. Where they are now 8'9" is as far apart as I could get them while still maintaining a super sharp center image. This spreads out horizontal sounstage as wide as it can be in my rem and still be focused. For example right now I'm listening to George Winston's album Autumn, which is close mic'd solo piano. Right now when I close my eyes I can "see" the exact positioning or each key in space. The higher octave keys are all the way on my right, and bass keys on my far left, and I can hear the reverb of the piano cabinet from the left. So I can visualize where the mic was above this piano (where my head is). Long short the large distance between my speakers is what helps separate all the keys in space and makes the 3-D experience so much more enveloping.
Similar reasoning applied to your situation, if you wanted wider soundstage, push speakers further apart, you might benefit from narrower coverage angle speakers, or some more acoustic treatment I think. Perhaps the ceiling acoustic treatment you are planning achieves this, very interesting to see how it affects the system if it extends your listening triangle.
I was looking into acoustic insulation today. There are some nice panels here in nz that specify the exact frequencies they attenuate. Pretty cool. Hopefully they're not too expensive. I will most definitely re-run REW and see how they change things once I've installed them.
And reasoning from thoughts and data here so far it looks to me there is three major contributors for the 3D image that need to happen: proper direct sound frequency balance (enough highs), sharp phantom image (direct / reverberant sound ratio, size and symmetry of the listening triangle, good speakers that are alike), and fine envelopment. These are just my speculation so feel free to comment :D By fine envelopment I mean nice late reverberation and symmetry of early reflections, which might mean suppressed early reflections or delayed enough early reflections, so that early reflections are not a mess. My intuition says when early reflections are benign the brain is not focusing to them but to the source material cues instead.

This is excellently stated. From what I can gather flat frequency response is vital so we're hearing every sound of the recording at the correct volume, and then all reflected sound must be either attenuated or arrive late enough for us to clearly differentiate it from the source. Bass will reflect off walls so speakers need to be moved around to find the spot where the bass reflections interfere with the incident waves in the right way so we don't get any nasty node or bass suck out, or huge boomy anti-nodes.

Something I'll add to the docs is when I ROTATE MY HEAD, keeping it exactly in the center line between the speakers, I can "look" at each instrument and it comes into greater focus, as if I've actually turned my head to look at it in the room. In this case right now I can turn and look at the high keys and see them exactly in space floating there as they're struck, then turn my head to the left and do the same with the lower octave keys and they come into sharper focus. So damn cool! But not exactly sure of the physics behind it. I think it has to do with getting the direct sound from each speaker at the exact correct time, so each image appears in the correct place, and the reflected sound off my walls is much quieter as to be inaudible/clearly distinguishable from the direct sound.
Yeah, they are not very directional but some, cardioid mid and ~90deg waveguide so DI around 6db at ~250Hz and smooth slope up to 10db at 20kHz. Its DSP system so I can EQ/optimize for any axis to be of any slope so can still experiment with toe-in, perhaps my EQ and toe-in is not optimal now but need to experiment further some day when there is time for critical listening. A grand mother and two active kids under same roof makes it almost impossible to have critical listening more than 30 seconds unless everyone else is out :D
Do you have any frequency response measurements of your speakers? I'd love to have a look they sound v cool.
And can you explain more how the DSP system of you speakers works? Thanks.
ps. sorry for long posts but I feel the stuff is in heart of any Hifi setup and there is not too much this kind of indepth descriptive texts, that would be easy to translate into my own setup / system. Stuff is often either marketing or too shallow so I cannot relate it to my situation, or perhaps too deep and just cannot understand :D Or perhaps there is and I've been just TL;DR :D I was excited seeing you write about the 2D/3D transition as its something I've never saw others write about earlier, so thanks writing it and letting me outlet my thoughts on it.
Yeah man again it's so cool so have a shared experience like this and I'm hoping to help others enjoy it in their rooms was well. It's pretty damn cool. Reading/writing/learning about it is great fun. I'm a physics teacher so all of this is research for next years lessons on Waves and Electromagnetism as well. Really enjoying it. Thanks.
 
Last edited:

tmuikku

Member
Joined
May 27, 2022
Messages
73
Likes
72
You're 100% correct about the 3-D sound wave propagation and attenuation of high frequencies before low. The cone analogy is exactly that, where the frequency response is still flat enough that you're hearing all the frequencies. And for my speakers in particular, attenuating the highs so the frequency response tails off as you climb is actually where MY speakers sound very good. This may (will) not be the case with every speaker. That's what I meant by speakers with less ideal horizontal dispersion might have to be toed in more, as the "listening window cones" will be narrower, and thus more toe-in will be needed to get that stereo overlap we're after. Not sure if that makes sense I can do a diagram to explain if need be.
Ah yes, it makes sense. You suggest something I forgot to even think about, the highs attenuate also by distance, I was thinking only radiation angle, toe-in. So, when speaker is that far from you the highs droop some due to distance, highs attenuating more than lows through air, which would need some equalization to get them higher again. As the tweeter (as well as any other transducers) has physical dimensions to it which makes the response to narrow when wavelength gets small compared to the transducer size. Basically speaker toe-in works as EQ to some particular direction. For example, on-axis frequency response might have slight boost on high treble and it straightens out to some angle off-axis, so toe in works as high frequency EQ for most part. One just uses the toe-in to balance out the highs. It appears that you get nice balance with the 15 deg toe in. Thats roughly 15 degree off-axis angle of the speaker you listen to with equilateral listening triangle. Ever tried toe-in more, 15 deg in front of you so 45deg toe-in?:D This would dramatically change sound towards early reflections, but maintain response of the direct sound. Not sure how this would affect the 3D image, delays of reflections and balance of direct sound would stay intact, only frequency response of lateral early reflections would change, most probably to worse but never know :) This is what is suggested for constant directivity speakers, 45deg toe in reduces early reflections a lot, while increasing later reflections (no acoustic treatment!) which should increase clarity while maintain some envelopment. My asymmetric setup doesn't seem to benefit from this though, perhaps I'd need smaller room or at least symmetrical setup for this to work better.

If you had a separate EQ the toe-in could be anything (sensible) and one could flatten out the direct sound, which ever angle that was. Problem is the EQ affects to every other direction as well, which affects sound of the reflections, sound of the room sound :) This is what I need to experiment more, boost highs and listen more off-axis to balance direct sound out while having more highs in the room.
This is excellently stated. From what I can gather flat frequency response is vital so we're hearing every sound of the recording at the correct volume, and then all reflected sound must be either attenuated or arrive late enough for us to clearly differentiate it from the source. Bass will reflect off walls so speakers need to be moved around to find the spot where the bass reflections interfere with the incident waves in the right way so we don't get any nasty node or bass suck out, or huge boomy anti-nodes.
These probably all affect the "quality". Is good loud SPL important for the 3D effect quality? What happens if you turn down volume some? I'd imagine overall sensation is not as impactul, but the phantom images probably stays intact, sharp and in place. What happens to envelopment and the 3D quality, do they degrade somehow?

System capable of good comfortably loud playback volume is essential for good sound regardless of stereo image, everything sounds better just by turning it up until the system or hearing gives :)
Something I'll add to the docs is when I ROTATE MY HEAD, keeping it exactly in the center line between the speakers, I can "look" at each instrument and it comes into greater focus, as if I've actually turned my head to look at it in the room. In this case right now I can turn and look at the high keys and see them exactly in space floating there as they're struck, then turn my head to the left and do the same with the lower octave keys and they come into sharper focus. So damn cool! But not exactly sure of the physics behind it. I think it has to do with getting the direct sound from each speaker at the exact correct time, so each image appears in the correct place, and the reflected sound off my walls is much quieter as to be inaudible/clearly distinguishable from the direct sound.
Yeah something like that, not knowing the details whats actually happening at least I'm mostly speculating. Brain doing good work on the data and probably the data is good, or your brain is good, probably both. I have no idea what kind of things need to be aligned in the sound field around us for this to happen, or what kind of stuff needs to be circulating in the brain and whether its similar for all humans or just some, perhaps to all. Anyway, very cool stuff :) Perhaps its just matter of rendering the early reflections harmless and enhancing the late reflections and of course having good setting for direct sound.
Do you have any frequency response measurements of your speakers? I'd love to have a look they sound v cool.
And can you explain more how the DSP system of you speakers works? Thanks.
No room measurements with the new waveguides, some simulated responses only :) I need free weekend to revisit all measurements, its modular system and the parts being developed different times and iterations spanning few years already and my measurements are all from different surroundings and dates so its not very consistent I'm afraid. All measurements done on single day would be needed to get more accurate real response, then do some in room measurements. There is still some things I need to build and then book a weekend, perhaps during the winter time (finland) :) Its three way system so six channels DSP and amplifiers, pretty basic IIR functionality but sufficient enough to make pretty much any "room curve" and crossovers of course. Nothing too special on this front, novelty is adjustability and that its got reduced sound with increasing off-axis angle compared to typical speakers with direct radiating tweeters. So, its just good measurements loaded into VituixCAD and all kinds of outcomes can be tuned pretty easily, then just transfer the settings to DSP. I'm also able to change c-c spacings to some extent, put it close or futher to wall behind. Only lack the time for critical listening to be able to progress any faster :D Slow process but I'm not hurry, its fun hobby.
Yeah man again it's so cool so have a shared experience like this and I'm hoping to help others enjoy it in their rooms was well. It's pretty damn cool. Reading/writing/learning about it is great fun. I'm a physics teacher so all of this is research for next years lessons on Waves and Electromagnetism as well. Really enjoying it. Thanks.
For sure :D have fun! I'll return with thoughs should more come to my mind. Definitely after I've managed to do some more experiments.
 
Last edited:

tmuikku

Member
Joined
May 27, 2022
Messages
73
Likes
72
Hey it would be really cool if you had time and sketched your setup with this tool https://amcoustics.com/tools/amray
We'd get some data about the path length differences of early reflections, perhaps useful perhaps not.

My setup is more into attenuating the early reflections than making them long so not sure if my room data was useful as its definitely very asymmetric, early reflections all over the place, shortest ones being wall behind speakers (great attenuation +10db), floor, one sidewall perhaps 1m out, height is 2.5m so vertical reflections similar to many I think.
 

Attachments

  • right.png
    right.png
    123 KB · Views: 29
  • left.png
    left.png
    147.6 KB · Views: 29
Last edited:
OP
T

tallbeardedone

Active Member
Joined
Sep 3, 2022
Messages
102
Likes
164
A Finn! Very cool. I'm in New Zealand but grew up in Minnesota where there is a large Scandinavian population. I actually have "sisu" tattooed on my hand and a Finnish sauna in my back yard! Also my novels star a hulking Finn named Einarr "Sledge" Laukkanen. :cool:
Ever tried toe-in more, 15 deg in front of you so 45deg toe-in?:D This would dramatically change sound towards early reflections, but maintain response of the direct sound. Not sure how this would affect the 3D image, delays of reflections and balance of direct sound would stay intact, only frequency response of lateral early reflections would change, most probably to worse but never know :) This is what is suggested for constant directivity speakers, 45deg toe in reduces early reflections a lot, while increasing later reflections (no acoustic treatment!) which should increase clarity while maintain some envelopment. My asymmetric setup doesn't seem to benefit from this though, perhaps I'd need smaller room or at least symmetrical setup for this to work better.
I have experimented with more extreme toe-in but not recently. I will monkey with it one of these days and report back. I wonder if with more toe-in if the sound and imaging will project down the room as well as it does now. At the moment I can stand along the midline of the room 10m away from the speakers and the horizontal imaging is still great. This morning I was listening to the new Taylor Swift album while making coffee and standing at the far end of my kitchen she was was right there in between my speakers singing to me. Pretty cool. With more extreme toe-in I wonder if it will sharpen that image from further away or lessen it due to the point of intersection of the speakers getting closer to the rear wall. It will be interesting to test.

If you had a separate EQ the toe-in could be anything (sensible) and one could flatten out the direct sound, which ever angle that was. Problem is the EQ affects to every other direction as well, which affects sound of the reflections, sound of the room sound :) This is what I need to experiment more, boost highs and listen more off-axis to balance direct sound out while having more highs in the room.
I've been pondering getting a miniDSP or the like to add EQ to my signal. Not sure I want it but I could attenuate that 65hz anti-node and flatten response even more across the entire band. The thing about EQ is it's flattening the sound at the listening position which is the addition of direct and reflected sound (it can't differentiate the two) so maybe it's taking away too much of the direct sound or not enough? I would have to invest the money to try it out and atm I love my sound so much it seems foolish to gamble $1000+ on something that I might not even end up using. I know it's "old school" but I like the idea of having to use old fashioned speaker placement and room treatments to find the best sound in my room, not digital EQ. And that's not even considering my all-analog chain!

No room measurements with the new waveguides, some simulated responses only :) I need free weekend to revisit all measurements, its modular system and the parts being developed different times and iterations spanning few years already and my measurements are all from different surroundings and dates so its not very consistent I'm afraid. All measurements done on single day would be needed to get more accurate real response, then do some in room measurements. There is still some things I need to build and then book a weekend, perhaps during the winter time (finland) :) Its three way system so six channels DSP and amplifiers, pretty basic IIR functionality but sufficient enough to make pretty much any "room curve" and crossovers of course. Nothing too special on this front, novelty is adjustability and that its got reduced sound with increasing off-axis angle compared to typical speakers with direct radiating tweeters. So, its just good measurements loaded into VituixCAD and all kinds of outcomes can be tuned pretty easily, then just transfer the settings to DSP. I'm also able to change c-c spacings to some extent, put it close or futher to wall behind. Only lack the time for critical listening to be able to progress any faster :D Slow process but I'm not hurry, its fun hobby.
Sound sooo fun. I'd love to come tinker. My only speaker building experience is with paper, wire and neodymium magnets in my physics class. :p
For sure :D have fun! I'll return with thoughs should more come to my mind. Definitely after I've managed to do some more experiments.
Yeah man I'll do the same. Happy listening!
 
OP
T

tallbeardedone

Active Member
Joined
Sep 3, 2022
Messages
102
Likes
164
Hey it would be really cool if you had time and sketched your setup with this tool https://amcoustics.com/tools/amray
We'd get some data about the path length differences of early reflections, perhaps useful perhaps not.

My setup is more into attenuating the early reflections than making them long so not sure if my room data was useful as its definitely very asymmetric, early reflections all over the place, shortest ones being wall behind speakers (great attenuation +10db), floor, one sidewall perhaps 1m out, height is 2.5m so vertical reflections similar to many I think.
Yo this is super cool I'm playing with it now!
 

stvnharr

Member
Joined
Apr 1, 2022
Messages
24
Likes
9
Hello,
I have been following your thread and I like what you are doing. A future project of mine is to do something similar, once I get set up to use REW, etc.
I have my speakers set up a lot different than you do. And I have never ever done anything in/with/etc. REW. In fact I still play discs, but I have the capability to run REW once I get a microphone. It would be a great way to see what I have.

Here is how I have my speakers set:
https://www.audioaficionado.org/showthread.php?t=46634 . I follow the instructions in Post #2, to the letter.
This is a process, called a few different things which do not matter. It is a great way to place speakers in a room and lets one free to sit in a much larger area than the single spot still as a statue way that I read you are doing.
I should add that I first heard speakers set this way back in 2007 at an exhibit room at RMAF. Over time I learned how to do this and became successful in doing it. The above instructions allowed me to fine tune and tweak what I already had. The instructions themselves are the way to go about this, and they give very precise descriptions was what to do and what to listen for. I recommend the Ballad of Runaway Horse song for doing this as it works very very well, and better than anything else.

You may want to try this and see how it goes. I would think the use of REW in the process could be quite useful, and you are well versed in this.
And if you don't like the results you can always just move your speaker back to their current position, as you have them well marked and documented in the room.

Keep up the experimenting and posting of results.

Steve
 
Last edited:
Top Bottom