• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Keith_W DSP system

OP
Keith_W

Keith_W

Major Contributor
Joined
Jun 26, 2016
Messages
2,846
Likes
6,551
Location
Melbourne, Australia
I can't say whether you would benefit from MSO but I can say your results from it look rather odd, level is nearly -30dB overall and loads of filters used. I suspect the inputs measurements not so good but hard to say.

I have all those measurements in storage, I can't remember which was which. Some of them were very early attempts with MSO when I did not know what I was doing. I grabbed the first one I saw and loaded it to confirm that Acourate's new MSO import feature works. I do not intend to use these old corrections again, I will make new ones. So don't judge those corrections ;)
 

OCA

Addicted to Fun and Learning
Forum Donor
Joined
Feb 2, 2020
Messages
730
Likes
581
Location
Germany
Where did you get that diagram from? Link please :)
It's a script written in Asymptote vector graphics language (attached).

I won't sacrifice access to my bedroom or out the house ;)

I doubt that about you sometimes :)
You could still place the speakers anywhere in the grey area. As long as you can keep 2.44m between speakers and form a triangle with the LP, you're fine. That kind of looks like ending up much like a 1/3rds rule placement in your case though.
 

Attachments

  • FirstRoomReflections.txt
    2.3 KB · Views: 45
Last edited:

OCA

Addicted to Fun and Learning
Forum Donor
Joined
Feb 2, 2020
Messages
730
Likes
581
Location
Germany
If you settle for 5ms delay of FWR's though (that's Dr. Toole's recommendation against 6ms by Linkwitz -most rooms will not ever satisfy neither due to limited size):

1706676457061.png


You can set up a 2.44m/side (that's the minimum recommended distance between speakers for proper stereo separation) equilateral triangle anywhere in the grey area.

For the rest of us with theoretically too small rooms [smallest room this can work: 5.2m (17ft) x 4.3m (14ft)], the question remains: Should one compromise stereo separation over hearing some of the first reflections too early? I went for 2.44m distance between my speakers and earlier side wall reflections in my system but I am not sure if that was the right thing to do. Without hijacking the thread much, any ideas welcome.
 
Last edited:

dualazmak

Major Contributor
Forum Donor
Joined
Feb 29, 2020
Messages
2,933
Likes
3,166
Location
Ichihara City, Chiba Prefecture, Japan
View attachment 346390

Hmmmm. I suppose it is possible. It would look really funny though. Not to mention a heck of a lot of work, not just moving the furniture, but also redoing the DSP. I am not keen :p

Really an interesting approach, indeed.
I am very much looking forward to hearing your result(s) of try(-and-error) on this thread. I am not in a hurry of course, please take your enough time for subjective and objective evaluations. :D
 
Last edited:

OCA

Addicted to Fun and Learning
Forum Donor
Joined
Feb 2, 2020
Messages
730
Likes
581
Location
Germany
View attachment 346390

Hmmmm. I suppose it is possible. It would look really funny though. Not to mention a heck of a lot of work, not just moving the furniture, but also redoing the DSP. I am not keen :p
How about some unorthodox approach?

1706683524116.png


Full symmetry front/rear/side walls! And you would defo need a new DSP :)

Edit:

I just checked, you can achieve perfect 1/3rds rule with LP and speakers, 6ms FWR delay and 2.44m separation simultaneously, DSP will be your slave!

1706686127798.png
 
Last edited:

ZolaIII

Major Contributor
Joined
Jul 28, 2019
Messages
4,291
Likes
2,540
View attachment 346390

Hmmmm. I suppose it is possible. It would look really funny though. Not to mention a heck of a lot of work, not just moving the furniture, but also redoing the DSP. I am not keen :p
No keep speakers as they are, move sofa a bit more from wall behind and put thick curtains to big glass surfaces there if that is acceptable and possible for you. Of course you would need to do DSP-ing from the beginning even it's not a big change regarding placement. Moving speakers out will make more phase inversion and holes while improving ISO 3382-1 is only wishful thinking while minimal phase simply works.
 
OP
Keith_W

Keith_W

Major Contributor
Joined
Jun 26, 2016
Messages
2,846
Likes
6,551
Location
Melbourne, Australia
I just checked, you can achieve perfect 1/3rds rule with LP and speakers, 6ms FWR delay and 2.44m separation simultaneously, DSP will be your slave!

View attachment 346400

The problem with that arrangement is ... it is 2.44m away from the speakers. I need to sit as far away from these speakers as possible. The closer to the speakers, the less "point source" they become. Thank you for your effort but I really am in no mood to reposition the speakers. It is waaaaaaaaay too much work. The cabling alone is complex and it would likely involve purchase of 4 pairs of new interconnects to make up the distance! But I will keep it in mind if one day I am bored and need a bit of a workout.

No keep speakers as they are, move sofa a bit more from wall behind and put thick curtains to big glass surfaces there if that is acceptable and possible for you. Of course you would need to do DSP-ing from the beginning even it's not a big change regarding placement. Moving speakers out will make more phase inversion and holes while improving ISO 3382-1 is only wishful thinking while minimal phase simply works.

Since I drew that diagram, the sofa has come closer to the speakers. The reason for doing that was to improve bass measurements. I am thinking of moving them back another 50cm or so to get further away from the speakers. There are already heavy drapes in front of those windows, you may be able to see them in some of the pictures on page 2.
 

ZolaIII

Major Contributor
Joined
Jul 28, 2019
Messages
4,291
Likes
2,540
@Keith_W yup that's very different from diagram. I also think you should move speakers back to what is shown on pictures. That's pretty much what you can do regarding placement tho find someone to help you move speakers they are heavy.
 

Mikig

Senior Member
Joined
Jan 12, 2024
Messages
481
Likes
546
Location
Italia
Hi Keith what exactly do you mean by “point source” .
Which sound characteristic, sitting in an armchair, best describes that aspect when listening?
 

AudioJester

Addicted to Fun and Learning
Forum Donor
Joined
Jan 7, 2020
Messages
978
Likes
1,316
There will be a minimum distance where all the drivers "converge", do you know what that is?

But it will never be a point source.

I say buy a second setup for the long wall - Kef LS60 AU$9k and run with your subs. Then be amazed
 
OP
Keith_W

Keith_W

Major Contributor
Joined
Jun 26, 2016
Messages
2,846
Likes
6,551
Location
Melbourne, Australia
Hi Keith what exactly do you mean by “point source” .
Which sound characteristic, sitting in an armchair, best describes that aspect when listening?

Hi Miki. A point source speaker means that the sound comes from a point. For example, KEF LS50 (or any KEF using their Uni-Q driver), Genelec 8351 (or any 1-series Genelec), or even something as large as a Tannoy Westminster or a Red Spade PSE-144. Small bookshelf speakers like the Rogers LS3/5a are not strictly speaking point source, but the drivers are so small and so close together that at normal distances they behave like point source speakers.

As Audiojester (below) mentions, large multi-driver speakers like mine have a minimum distance where the drivers "converge", although I personally do not know how to define where this distance is in terms of measurements.

There will be a minimum distance where all the drivers "converge", do you know what that is?

But it will never be a point source.

Do you know the definition of "converge" in terms of measurements? Help me out here, because I don't know.

I made up my own definition through reasoning that "the convergence point is the distance where the radiation cones of all the drivers mix". Therefore, if you start taking measurements in a straight line backwards from the speaker, the initial measurements closer to the speaker should show massive fluctuations because the drivers have not converged. As you get further away, you reach the convergence point, where any measurements further than this should show the same rate of change. I discussed this a little in post #32 of this thread where I took progressively further readings from the mic.

If you look carefully at my measurements, the drivers never converge. The horns and the cabinet never decay at the same rate, no matter how far I get from the speaker.

I remember looking at this measurement, coming to this realisation, and thinking "I need a beer".

I say buy a second setup for the long wall - Kef LS60 AU$9k and run with your subs. Then be amazed

I was complaining to my friend about that Besessen Zwanghaft Audiofil guy and he told me that he is right. I then stared a bit at my living room, and I realized that the width of my room is actually greater than the length of some people's living rooms (6m wide). So yes, I could get away by rotating the system and set it up against the long wall. Besides symmetry, there are other nice benefits, too. e.g. the speakers will be further away from the side walls and I get to look out the big picture window into my garden instead of having the window behind me. It means that the rear wall will be 1m behind the sofa, but I have a washroom, spare bedroom, staircase, etc. which will act as natural diffusers / places for sound to disappear into ... so it will be fine! Getting the speakers far away from the side walls will also help with BACCH.

In preparation for that move, I had to make sure that I could replicate Dr. Uli's tuning that he did for me over a Teamviewer session several months ago. It was the best sounding correction I have ever heard, and all my attempts to replicate it have failed. Tonight, I did a brute force replication by measuring the FR of his correction and setting that as my target. I can already see ASR's brows furrowing over this remark and the knives coming out. But I do what makes me happy, and Dr. Uli's tuning made me happy. The story of that correction is interesting in itself because the result was not what was intended, and I am sure that Dr. Uli himself has not heard this particularly weird looking target curve.

But before I go any further - YES I KNOW WHAT EVERYONE IS THINKING! No need to quote Toole to me, I know what he said about target curves. Just run with me on this one.

1706807345326.png


At the end of the Teamviewer session with Dr. Uli, the verification measurement looked like this (red/green = L/R, brown = target). It sounded great. However, over the months, little changes crept in. The sofa moved forward. Some other furniture was moved around. I made little adjustments to the volume of the tweeter and then adjusted it back. Who knows, maybe something happened to the system. The change in sound was too subtle for me to notice, until I decided to redo the DSP.

Even with the target curve saved on my PC, nothing I did could replicate that sound. In frustration, I repeated the measurements. This is what I saw:

1706807608808.png


Well i'll be a damned monkey's uncle. It looks nothing like the beautiful correction done a few months back. That's not even a "hot" tweeter, because I would expect to see a peak if the tweeter was "hot". Instead, what we get is this weird triangular from 7kHz up. Then it occurred to me ... Dr. Uli did that correction without the sofa in situ. So I removed the sofa and repeated the measurement.

1706807876337.png


By this stage I was struggling to understand what had happened. I tried moving the microphone back by about 1m, and that restored the original appearance of the measurement somewhat. But I had grown to like this particular sound. It does not even sound too bright, what is most interesting about it are the spatial qualities. It gives the impression of more open-ness, and the ASW is much wider. If I go back to Dr. Uli's target curve, it sounds ... "nice". But this sounds much nicer. That funny looking midrange hump actually adds some richness to the sound, and the top end adds "air" without sounding sibilant or sparkly. There is a kind of ethereal quality.

So out came Acourate. I took the above measurement, applied a lot of smoothing, and used it as a target curve (Acourate users: did you know that you can use any curve as a target curve? Save it as "target.dbl". Skip Room Macro 2 ("Target Curve Design") and proceed directly to Macro 3. Just make sure you adjust the gain of the target and save it, then indicate your new target in Macro 3's dialog box. You can even run different target curves for left and right speakers by ticking the "mono" option and running Macro 3 twice or even 16 different target curves for 16 speakers ... the possibilities are endless). After a full round of corrections, this is the result:

1706808501719.png


It sounds like a more refined version of that odd looking curve I showed earlier. It still has the midrange richness, the ethereal top end, but the image produced is more solid. While I was at it, I replicated Dr. Uli's original target. With Acourate Convolver, I can immediately switch between the two filters:

1706808650370.png


... so now that I know I can replicate that sound, I can now think about moving furniture around to reconfigure the room.
 

Mikig

Senior Member
Joined
Jan 12, 2024
Messages
481
Likes
546
Location
Italia
Hi Miki. A point source speaker means that the sound comes from a point. For example, KEF LS50 (or any KEF using their Uni-Q driver), Genelec 8351 (or any 1-series Genelec), or even something as large as a Tannoy Westminster or a Red Spade PSE-144. Small bookshelf speakers like the Rogers LS3/5a are not strictly speaking point source, but the drivers are so small and so close together that at normal distances they behave like point source speakers.

As Audiojester (below) mentions, large multi-driver speakers like mine have a minimum distance where the drivers "converge", although I personally do not know how to define where this distance is in terms of measurements.



Do you know the definition of "converge" in terms of measurements? Help me out here, because I don't know.

I made up my own definition through reasoning that "the convergence point is the distance where the radiation cones of all the drivers mix". Therefore, if you start taking measurements in a straight line backwards from the speaker, the initial measurements closer to the speaker should show massive fluctuations because the drivers have not converged. As you get further away, you reach the convergence point, where any measurements further than this should show the same rate of change. I discussed this a little in post #32 of this thread where I took progressively further readings from the mic.

If you look carefully at my measurements, the drivers never converge. The horns and the cabinet never decay at the same rate, no matter how far I get from the speaker.

I remember looking at this measurement, coming to this realisation, and thinking "I need a beer".



I was complaining to my friend about that Besessen Zwanghaft Audiofil guy and he told me that he is right. I then stared a bit at my living room, and I realized that the width of my room is actually greater than the length of some people's living rooms (6m wide). So yes, I could get away by rotating the system and set it up against the long wall. Besides symmetry, there are other nice benefits, too. e.g. the speakers will be further away from the side walls and I get to look out the big picture window into my garden instead of having the window behind me. It means that the rear wall will be 1m behind the sofa, but I have a washroom, spare bedroom, staircase, etc. which will act as natural diffusers / places for sound to disappear into ... so it will be fine! Getting the speakers far away from the side walls will also help with BACCH.

In preparation for that move, I had to make sure that I could replicate Dr. Uli's tuning that he did for me over a Teamviewer session several months ago. It was the best sounding correction I have ever heard, and all my attempts to replicate it have failed. Tonight, I did a brute force replication by measuring the FR of his correction and setting that as my target. I can already see ASR's brows furrowing over this remark and the knives coming out. But I do what makes me happy, and Dr. Uli's tuning made me happy. The story of that correction is interesting in itself because the result was not what was intended, and I am sure that Dr. Uli himself has not heard this particularly weird looking target curve.

But before I go any further - YES I KNOW WHAT EVERYONE IS THINKING! No need to quote Toole to me, I know what he said about target curves. Just run with me on this one.

View attachment 346614

At the end of the Teamviewer session with Dr. Uli, the verification measurement looked like this (red/green = L/R, brown = target). It sounded great. However, over the months, little changes crept in. The sofa moved forward. Some other furniture was moved around. I made little adjustments to the volume of the tweeter and then adjusted it back. Who knows, maybe something happened to the system. The change in sound was too subtle for me to notice, until I decided to redo the DSP.

Even with the target curve saved on my PC, nothing I did could replicate that sound. In frustration, I repeated the measurements. This is what I saw:

View attachment 346615

Well i'll be a damned monkey's uncle. It looks nothing like the beautiful correction done a few months back. That's not even a "hot" tweeter, because I would expect to see a peak if the tweeter was "hot". Instead, what we get is this weird triangular from 7kHz up. Then it occurred to me ... Dr. Uli did that correction without the sofa in situ. So I removed the sofa and repeated the measurement.

View attachment 346616

By this stage I was struggling to understand what had happened. I tried moving the microphone back by about 1m, and that restored the original appearance of the measurement somewhat. But I had grown to like this particular sound. It does not even sound too bright, what is most interesting about it are the spatial qualities. It gives the impression of more open-ness, and the ASW is much wider. If I go back to Dr. Uli's target curve, it sounds ... "nice". But this sounds much nicer. That funny looking midrange hump actually adds some richness to the sound, and the top end adds "air" without sounding sibilant or sparkly. There is a kind of ethereal quality.

So out came Acourate. I took the above measurement, applied a lot of smoothing, and used it as a target curve (Acourate users: did you know that you can use any curve as a target curve? Save it as "target.dbl". Skip Room Macro 2 ("Target Curve Design") and proceed directly to Macro 3. Just make sure you adjust the gain of the target and save it, then indicate your new target in Macro 3's dialog box. You can even run different target curves for left and right speakers by ticking the "mono" option and running Macro 3 twice or even 16 different target curves for 16 speakers ... the possibilities are endless). After a full round of corrections, this is the result:

View attachment 346617

It sounds like a more refined version of that odd looking curve I showed earlier. It still has the midrange richness, the ethereal top end, but the image produced is more solid. While I was at it, I replicated Dr. Uli's original target. With Acourate Convolver, I can immediately switch between the two filters:

View attachment 346618

... so now that I know I can replicate that sound, I can now think about moving furniture around to reconfigure the room.


Hi Keith,

thanks for the timely and precise answer, you are very kind as always,
this evening after work I will take my time to read it carefully and...I will definitely bother you with some other questions!! ;)
 

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
3,117
Likes
2,340
Location
Canada
Are those measurements displayed with no frequency dependent windowing? I'm just curious... As you already know, the frequency dependent reflection decay characteristics of the room will influence what you hear as well and so should not only rely in the apparent steady-state magnitude response smoothness and linearity.
 
OP
Keith_W

Keith_W

Major Contributor
Joined
Jun 26, 2016
Messages
2,846
Likes
6,551
Location
Melbourne, Australia
Are those measurements displayed with no frequency dependent windowing? I'm just curious... As you already know, the frequency dependent reflection decay characteristics of the room will influence what you hear as well and so should not only rely in the apparent steady-state magnitude response smoothness and linearity.

Of course there is FDW :) I prefer Acourate's 15/15 FDW when displaying graphs, it is detailed enough not to hide anything that is going on, but smooth enough for you to discern the trend. This is what it looks like without any FDW at all.

1706814430580.png


When I do corrections, I do not use 15/15 FDW. FDW 15/15 is for display only. I use 15/8 (refers to high resolution correction for bass, medium resolution for top).
 

3ll3d00d

Active Member
Joined
Aug 31, 2019
Messages
218
Likes
180
I use 15/8 (refers to high resolution correction for bass, medium resolution for top).
Fwiw, and if memory serves, it means 15 cycles at 16Hz, 8 at fs/2 then it interpolates between those two values at frequency rises (so the window is gradually getting smaller rather than there being a step change)
 

Sokel

Master Contributor
Joined
Sep 8, 2021
Messages
6,510
Likes
6,673
Hi Miki. A point source speaker means that the sound comes from a point. For example, KEF LS50 (or any KEF using their Uni-Q driver), Genelec 8351 (or any 1-series Genelec), or even something as large as a Tannoy Westminster or a Red Spade PSE-144. Small bookshelf speakers like the Rogers LS3/5a are not strictly speaking point source, but the drivers are so small and so close together that at normal distances they behave like point source speakers.

As Audiojester (below) mentions, large multi-driver speakers like mine have a minimum distance where the drivers "converge", although I personally do not know how to define where this distance is in terms of measurements.



Do you know the definition of "converge" in terms of measurements? Help me out here, because I don't know.

I made up my own definition through reasoning that "the convergence point is the distance where the radiation cones of all the drivers mix". Therefore, if you start taking measurements in a straight line backwards from the speaker, the initial measurements closer to the speaker should show massive fluctuations because the drivers have not converged. As you get further away, you reach the convergence point, where any measurements further than this should show the same rate of change. I discussed this a little in post #32 of this thread where I took progressively further readings from the mic.

If you look carefully at my measurements, the drivers never converge. The horns and the cabinet never decay at the same rate, no matter how far I get from the speaker.

I remember looking at this measurement, coming to this realisation, and thinking "I need a beer".



I was complaining to my friend about that Besessen Zwanghaft Audiofil guy and he told me that he is right. I then stared a bit at my living room, and I realized that the width of my room is actually greater than the length of some people's living rooms (6m wide). So yes, I could get away by rotating the system and set it up against the long wall. Besides symmetry, there are other nice benefits, too. e.g. the speakers will be further away from the side walls and I get to look out the big picture window into my garden instead of having the window behind me. It means that the rear wall will be 1m behind the sofa, but I have a washroom, spare bedroom, staircase, etc. which will act as natural diffusers / places for sound to disappear into ... so it will be fine! Getting the speakers far away from the side walls will also help with BACCH.

In preparation for that move, I had to make sure that I could replicate Dr. Uli's tuning that he did for me over a Teamviewer session several months ago. It was the best sounding correction I have ever heard, and all my attempts to replicate it have failed. Tonight, I did a brute force replication by measuring the FR of his correction and setting that as my target. I can already see ASR's brows furrowing over this remark and the knives coming out. But I do what makes me happy, and Dr. Uli's tuning made me happy. The story of that correction is interesting in itself because the result was not what was intended, and I am sure that Dr. Uli himself has not heard this particularly weird looking target curve.

But before I go any further - YES I KNOW WHAT EVERYONE IS THINKING! No need to quote Toole to me, I know what he said about target curves. Just run with me on this one.

View attachment 346614

At the end of the Teamviewer session with Dr. Uli, the verification measurement looked like this (red/green = L/R, brown = target). It sounded great. However, over the months, little changes crept in. The sofa moved forward. Some other furniture was moved around. I made little adjustments to the volume of the tweeter and then adjusted it back. Who knows, maybe something happened to the system. The change in sound was too subtle for me to notice, until I decided to redo the DSP.

Even with the target curve saved on my PC, nothing I did could replicate that sound. In frustration, I repeated the measurements. This is what I saw:

View attachment 346615

Well i'll be a damned monkey's uncle. It looks nothing like the beautiful correction done a few months back. That's not even a "hot" tweeter, because I would expect to see a peak if the tweeter was "hot". Instead, what we get is this weird triangular from 7kHz up. Then it occurred to me ... Dr. Uli did that correction without the sofa in situ. So I removed the sofa and repeated the measurement.

View attachment 346616

By this stage I was struggling to understand what had happened. I tried moving the microphone back by about 1m, and that restored the original appearance of the measurement somewhat. But I had grown to like this particular sound. It does not even sound too bright, what is most interesting about it are the spatial qualities. It gives the impression of more open-ness, and the ASW is much wider. If I go back to Dr. Uli's target curve, it sounds ... "nice". But this sounds much nicer. That funny looking midrange hump actually adds some richness to the sound, and the top end adds "air" without sounding sibilant or sparkly. There is a kind of ethereal quality.

So out came Acourate. I took the above measurement, applied a lot of smoothing, and used it as a target curve (Acourate users: did you know that you can use any curve as a target curve? Save it as "target.dbl". Skip Room Macro 2 ("Target Curve Design") and proceed directly to Macro 3. Just make sure you adjust the gain of the target and save it, then indicate your new target in Macro 3's dialog box. You can even run different target curves for left and right speakers by ticking the "mono" option and running Macro 3 twice or even 16 different target curves for 16 speakers ... the possibilities are endless). After a full round of corrections, this is the result:

View attachment 346617

It sounds like a more refined version of that odd looking curve I showed earlier. It still has the midrange richness, the ethereal top end, but the image produced is more solid. While I was at it, I replicated Dr. Uli's original target. With Acourate Convolver, I can immediately switch between the two filters:

View attachment 346618

... so now that I know I can replicate that sound, I can now think about moving furniture around to reconfigure the room.
Dear god,my dear Keith,is not about thinking what Toole says,is about a well established way (probably the only one) that things work.
Choosing to ignore something so basic will only lead you to build in sand.
It's a pity that so much effort has no solid foundations.

Things doesn't have to be complicated,sometimes we made so but at the same time we don't look at the big picture.

To help you about sum,I measured my friends Blumenhofer DV3,these are big 2 way horns.
They summed nicely at about 2.5 meters with the top horn inclined and the mic height between the top horn and the bass cabinet (MLP s at 4m)
Room is big of course,more than 150m² .

2-3 meters with your smaller ones should be plenty for summing.

From that distance and back,bass started taking over more
 

Mikig

Senior Member
Joined
Jan 12, 2024
Messages
481
Likes
546
Location
Italia
Hi Miki. A point source speaker means that the sound comes from a point. For example, KEF LS50 (or any KEF using their Uni-Q driver), Genelec 8351 (or any 1-series Genelec), or even something as large as a Tannoy Westminster or a Red Spade PSE-144. Small bookshelf speakers like the Rogers LS3/5a are not strictly speaking point source, but the drivers are so small and so close together that at normal distances they behave like point source speakers.

As Audiojester (below) mentions, large multi-driver speakers like mine have a minimum distance where the drivers "converge", although I personally do not know how to define where this distance is in terms of measurements.



Do you know the definition of "converge" in terms of measurements? Help me out here, because I don't know.

I made up my own definition through reasoning that "the convergence point is the distance where the radiation cones of all the drivers mix". Therefore, if you start taking measurements in a straight line backwards from the speaker, the initial measurements closer to the speaker should show massive fluctuations because the drivers have not converged. As you get further away, you reach the convergence point, where any measurements further than this should show the same rate of change. I discussed this a little in post #32 of this thread where I took progressively further readings from the mic.

If you look carefully at my measurements, the drivers never converge. The horns and the cabinet never decay at the same rate, no matter how far I get from the speaker.

I remember looking at this measurement, coming to this realisation, and thinking "I need a beer".



I was complaining to my friend about that Besessen Zwanghaft Audiofil guy and he told me that he is right. I then stared a bit at my living room, and I realized that the width of my room is actually greater than the length of some people's living rooms (6m wide). So yes, I could get away by rotating the system and set it up against the long wall. Besides symmetry, there are other nice benefits, too. e.g. the speakers will be further away from the side walls and I get to look out the big picture window into my garden instead of having the window behind me. It means that the rear wall will be 1m behind the sofa, but I have a washroom, spare bedroom, staircase, etc. which will act as natural diffusers / places for sound to disappear into ... so it will be fine! Getting the speakers far away from the side walls will also help with BACCH.

In preparation for that move, I had to make sure that I could replicate Dr. Uli's tuning that he did for me over a Teamviewer session several months ago. It was the best sounding correction I have ever heard, and all my attempts to replicate it have failed. Tonight, I did a brute force replication by measuring the FR of his correction and setting that as my target. I can already see ASR's brows furrowing over this remark and the knives coming out. But I do what makes me happy, and Dr. Uli's tuning made me happy. The story of that correction is interesting in itself because the result was not what was intended, and I am sure that Dr. Uli himself has not heard this particularly weird looking target curve.

But before I go any further - YES I KNOW WHAT EVERYONE IS THINKING! No need to quote Toole to me, I know what he said about target curves. Just run with me on this one.

View attachment 346614

At the end of the Teamviewer session with Dr. Uli, the verification measurement looked like this (red/green = L/R, brown = target). It sounded great. However, over the months, little changes crept in. The sofa moved forward. Some other furniture was moved around. I made little adjustments to the volume of the tweeter and then adjusted it back. Who knows, maybe something happened to the system. The change in sound was too subtle for me to notice, until I decided to redo the DSP.

Even with the target curve saved on my PC, nothing I did could replicate that sound. In frustration, I repeated the measurements. This is what I saw:

View attachment 346615

Well i'll be a damned monkey's uncle. It looks nothing like the beautiful correction done a few months back. That's not even a "hot" tweeter, because I would expect to see a peak if the tweeter was "hot". Instead, what we get is this weird triangular from 7kHz up. Then it occurred to me ... Dr. Uli did that correction without the sofa in situ. So I removed the sofa and repeated the measurement.

View attachment 346616

By this stage I was struggling to understand what had happened. I tried moving the microphone back by about 1m, and that restored the original appearance of the measurement somewhat. But I had grown to like this particular sound. It does not even sound too bright, what is most interesting about it are the spatial qualities. It gives the impression of more open-ness, and the ASW is much wider. If I go back to Dr. Uli's target curve, it sounds ... "nice". But this sounds much nicer. That funny looking midrange hump actually adds some richness to the sound, and the top end adds "air" without sounding sibilant or sparkly. There is a kind of ethereal quality.

So out came Acourate. I took the above measurement, applied a lot of smoothing, and used it as a target curve (Acourate users: did you know that you can use any curve as a target curve? Save it as "target.dbl". Skip Room Macro 2 ("Target Curve Design") and proceed directly to Macro 3. Just make sure you adjust the gain of the target and save it, then indicate your new target in Macro 3's dialog box. You can even run different target curves for left and right speakers by ticking the "mono" option and running Macro 3 twice or even 16 different target curves for 16 speakers ... the possibilities are endless). After a full round of corrections, this is the result:

View attachment 346617

It sounds like a more refined version of that odd looking curve I showed earlier. It still has the midrange richness, the ethereal top end, but the image produced is more solid. While I was at it, I replicated Dr. Uli's original target. With Acourate Convolver, I can immediately switch between the two filters:

View attachment 346618

... so now that I know I can replicate that sound, I can now think about moving furniture around to reconfigure the room.

hi Keith,
in simple words, you know I'm not a technician!!;) I would tell you that it is the vertex of the triangle that forms between the speakers and the listening position.
That triangle which, if too flattened, makes you hear the music directed by the speakers, therefore with the loss of the "illusion" of their disappearance, and if too elongated makes you lose the details while listening and risks putting you in knotted positions and resonances. I think that in simple words this is the result.
So we agree that the “visual” “source point”, when the configuration proportions are appropriate, focuses before your eyes, revealing the so-called stage.

My trivial question was asked because I have the same problem with the room, and I'm looking for points of reference on the fundamental themes, so I need to start from the most "trivial" definitions to at least compare on the basis.

I find myself struggling with a fairly large room, which complicated everything. and, finding an acceptable position for me without measuring instruments was, and is, a challenge. I've already turned it three times, now it seems like the best solution.
I have a Mac and if I also want a microphone, what type of program can be used to start giving a more scientific sense to this type of research that I have always had to do "by ear"??

I'll ask you the second question!! I noticed that you, and also many other users, use a piece of furniture between the speakers and the wall: have you ever measured whether this piece of furniture becomes a "fake speaker" that creates a sound box?? Thanks good evening!
 
OP
Keith_W

Keith_W

Major Contributor
Joined
Jun 26, 2016
Messages
2,846
Likes
6,551
Location
Melbourne, Australia
Fwiw, and if memory serves, it means 15 cycles at 16Hz, 8 at fs/2 then it interpolates between those two values at frequency rises (so the window is gradually getting smaller rather than there being a step change)



Dear god,my dear Keith,is not about thinking what Toole says,is about a well established way (probably the only one) that things work.
Choosing to ignore something so basic will only lead you to build in sand.
It's a pity that so much effort has no solid foundations.

My dear Sokel, I knew that people would come after me with that last post :) I don't ignore anything. But I do take my own measurements to check if my speakers are behaving according to the model. And if they don't, I have to explain why.

To help you about sum,I measured my friends Blumenhofer DV3,these are big 2 way horns.
They summed nicely at about 2.5 meters with the top horn inclined and the mic height between the top horn and the bass cabinet (MLP s at 4m)
Room is big of course,more than 150m² .

2-3 meters with your smaller ones should be plenty for summing.

From that distance and back,bass started taking over more

Okay, there is a difference between the Blumenhofer and my speakers. This is the Blumenhofer DV3:

1706847441186.png


Notice how all drivers are horn loaded. This means that the directivity of upper and lower frequencies shouldn't change too much. And more importantly, the overall behaviour of the speaker shouldn't change too much because all the drivers are of the same type - horn loaded.

And these are my speakers:

acapella_geschichte-Violon2.jpg


Horns from 500Hz up, bass cabinet for everything below. You mention that when you measured your friend's Blumenhofer, bass started to take over more, further away from the speaker? Look at my my measurements I posted earlier:

index.php


See what happens to the bass? Note in particular the difference between the red curve (measured 30cm from the speaker) and the bottom curve (measured at MLP). I extensively discussed this phenomenon earlier in the thread. Bass volume drops, anything horn loaded stays almost the same. So, if I managed to gain access to an anechoic chamber and correct the whole speaker to be flat under anechoic conditions, what I will observe at MLP is an upward tilting response - the opposite of a normal speaker.

So it is not a case of "ignoring Toole", it is more recognizing that a speaker with mixed driver types will behave differently and produce results that are not expected.

Now, I am certain that IF my speakers had horn loaded bass, they would not behave like that.
 
OP
Keith_W

Keith_W

Major Contributor
Joined
Jun 26, 2016
Messages
2,846
Likes
6,551
Location
Melbourne, Australia
Fwiw, and if memory serves, it means 15 cycles at 16Hz, 8 at fs/2 then it interpolates between those two values at frequency rises (so the window is gradually getting smaller rather than there being a step change)

Sorry, I meant to reply to this earlier but I got distracted :) My understanding of FDW is the same as yours. I would have thought that it would produce a smoother, lower res correction than 1 cycle, but it doesn't. It is a little beyond me, and I haven't bothered asking why. Here are a bunch of graphs with different FDW's.

1706855001490.png


FDW 15/15

1706855031714.png


FDW 8/8

1706855062807.png


FDW 1/1.

1706855367828.png


And this is FDW 15/1 (thick line above) compared with FDW 15/15 (lighter green line below). Notice that with the FDW 15/1 there is tight bass correction but it smoothens out at the top end, i.e. Toole's "broad tone control".
 
Top Bottom