• 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!

"The secret of big speakers"

oivavoi

Major Contributor
Forum Donor
Joined
Jan 12, 2017
Messages
1,721
Likes
1,934
Location
Oslo, Norway
Of course I agree that one should get as much right possible with the acoustic design ;) But I'm not sure I agree with the rest. What digital artefacts are we talking about, for example?

To be clear, I would never buy a speaker system that wasnt DSP-based these days. And I’m no expert at all. Still, my limited understanding is that there is a potential for DSP to create audible artifacts, if it’s not done right? I read a couple of studies on this some time back, but forgot the precise details. Will see if I can find them.
 

Purité Audio

Master Contributor
Industry Insider
Barrowmaster
Forum Donor
Joined
Feb 29, 2016
Messages
9,050
Likes
12,148
Location
London
Latency doesn’t make the slightest difference for music layback of course, when you choose a low latency mode in the Kiis you only ‘lose’ the total phase coherence , everything else is untouched.
Keith
 

maverickronin

Major Contributor
Forum Donor
Joined
Jul 19, 2018
Messages
2,527
Likes
3,308
Location
Midwest, USA
How much latency are we talking about? I can't imagine it being any more than TV/monitors these days...
 

Fitzcaraldo215

Major Contributor
Joined
Mar 4, 2016
Messages
1,440
Likes
632
To be clear, I would never buy a speaker system that wasnt DSP-based these days. And I’m no expert at all. Still, my limited understanding is that there is a potential for DSP to create audible artifacts, if it’s not done right? I read a couple of studies on this some time back, but forgot the precise details. Will see if I can find them.
I have used DSP Room EQ for over a decade now, as have numerous friends. I have heard absolutely no unwelcome artifacts whatsoever. Of course, DSP changes the sound, but that is what it is designed to do in a controlled way to provide its algorithmic benefits to improve the sound. Don’t want the improvements? You usually can just bypass it.
 

Fitzcaraldo215

Major Contributor
Joined
Mar 4, 2016
Messages
1,440
Likes
632
One obvious artefact of certain digital processes is latency ;)

In some applications, easily noticed.

Apart from that, a simple bypass button would reveal the potential win or loss of additional digital processing.

And in the end, KISS always wins, doesn’t it :)
But, really, in normal home listening situations, latency is a total non-issue, even in those involving lip syncing audio with video.

I use Dirac Live, which, no doubt, introduces up to milliseconds of latency. There is no problem syncing with video with Dirac on or bypassed, which is done automatically with no intervention. And, yes as you mention, Dirac has an on/bypass switch.

As to simple music playback, the listener cannot perceive the added latency. Pushing Play and having the music start even many milliseconds later via DSP is just not detectable.
 

Fitzcaraldo215

Major Contributor
Joined
Mar 4, 2016
Messages
1,440
Likes
632
It depends on the design. Check out with manufacturers. It could easily be tens of milliseconds.
Many subwoofers introduce latency because they use DSP input filter networks. That might introduce problems in some stereo setups, especially with subs positioned behind the speakers and if sub Phase controls lack sufficient adjustability. Sub positioning in front of the speakers can help compensate for this.

But, in a Mch setup like mine, distance is set acoustically via mic and all channel delays adjusted accordingly. So, there is no latency issue.
 

andreasmaaan

Master Contributor
Forum Donor
Joined
Jun 19, 2018
Messages
6,652
Likes
9,399
To be clear, I would never buy a speaker system that wasnt DSP-based these days. And I’m no expert at all. Still, my limited understanding is that there is a potential for DSP to create audible artifacts, if it’s not done right? I read a couple of studies on this some time back, but forgot the precise details. Will see if I can find them.

I guess you could stuff it up and create audible artefacts :)

A well-executed system should not do this however, and it should be easier to execute the system well (at least in terms of power, flexibility and available alternatives) with digital than without.

Would be interested to see if you can turn up any of those articles you read though, thx
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,522
Likes
37,050
I guess you could stuff it up and create audible artefacts :)

A well-executed system should not do this however, and it should be easier to execute the system well (at least in terms of power, flexibility and available alternatives) with digital than without.

Would be interested to see if you can turn up any of those articles you read though, thx
FIR filters for crossovers that are too steep have the potential to create the dreaded ringing, which is in fact audible. It also can put transient signals spread out over a time window that is longer than our ear's filtering time which would make it audible. How steep is too steep gets to be moderately complex as it depends upon the design of the filter and how many taps it uses. Low tap counts might need to stay at 36db/octave or less. Better filter design with more taps could manage much steeper crossovers with no unintended ill effects.
 

maverickronin

Major Contributor
Forum Donor
Joined
Jul 19, 2018
Messages
2,527
Likes
3,308
Location
Midwest, USA
It depends on the design. Check out with manufacturers. It could easily be tens of milliseconds.

That's actually in line with many displays. Even 50ms isn't uncommon at all, especially if you don't put it in game mode. I have an over the top 165Hz gaming monitor and its absolute input latency is still 24ms worst case. Most people here aren't gamers anyway so while it's something I would certainly check on I don't think typical ASR users would find it a concern.
 

Cosmik

Major Contributor
Joined
Apr 24, 2016
Messages
3,075
Likes
2,180
Location
UK
FIR filters for crossovers that are too steep have the potential to create the dreaded ringing, which is in fact audible. It also can put transient signals spread out over a time window that is longer than our ear's filtering time which would make it audible. How steep is too steep gets to be moderately complex as it depends upon the design of the filter and how many taps it uses. Low tap counts might need to stay at 36db/octave or less. Better filter design with more taps could manage much steeper crossovers with no unintended ill effects.
But don't forget that the ringing in the adjacent driver (it's crossing over to) is complementary, resulting in... perfection - unlike the non-DSP equivalent. Of course, in practice it may not be perfect at every location around the speaker, but I'll take my chances with my 4th order FIR filters.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,522
Likes
37,050
But don't forget that the ringing in the adjacent driver (it's crossing over to) is complementary, resulting in... perfection - unlike the non-DSP equivalent. Of course, in practice it may not be perfect at every location around the speaker, but I'll take my chances with my 4th order FIR filters.

If the sound source were from the same location it would be perfect. But since tweeters, midranges and woofers are usually offset physically that won't be the case. How audible it is would vary upon other factors, but it is potentially rather audible. 4th order filters probably aren't a problem.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,522
Likes
37,050
Spent a little time listening to big speakers with the bottom end rolled off. They don't sound like small speakers. They sound like big speakers playing a recording with a lightweight low end. Haha! This was 12 db/octave roll off at 100, 125 and 150 hz.

So I tried 24 db/octave, and still sounds like a big speaker with a chopped off bottom on the recording.

Finally some progress. 24 db/octave roll off at 150 hz. Followed by a 6 db peak above flat at 80 hz (not uncommon in some small ported speakers), which creates a slight dip between those two points. Then rolled off the response an extra amount (about 36 db/octave) below 60 hz. Now it does begin to sound like a small speaker. And doing this digital EQ on a music file likely doesn't cause all the phase nasties you get with a real ported speaker made with a Q of 1.1 or 1.2 or 1.5 even.

This reminded me of Hales Signature Two's I've had. Heavy inert cabinet, M-T-M driver arrangement in a sealed box. With a Q of slightly below .7 for the speakers. Good response to only 50 hz really, but with the sealed box and a sturdy amp it had a big speaker sound. It also was uncolored enough I was happy with them even though I'm a lifelong panel speaker fan.
 
OP
H

hvbias

Addicted to Fun and Learning
Joined
Apr 28, 2016
Messages
574
Likes
411
Location
US
Spent a little time listening to big speakers with the bottom end rolled off. They don't sound like small speakers. They sound like big speakers playing a recording with a lightweight low end. Haha! This was 12 db/octave roll off at 100, 125 and 150 hz.

So I tried 24 db/octave, and still sounds like a big speaker with a chopped off bottom on the recording.

Finally some progress. 24 db/octave roll off at 150 hz. Followed by a 6 db peak above flat at 80 hz (not uncommon in some small ported speakers), which creates a slight dip between those two points. Then rolled off the response an extra amount (about 36 db/octave) below 60 hz. Now it does begin to sound like a small speaker. And doing this digital EQ on a music file likely doesn't cause all the phase nasties you get with a real ported speaker made with a Q of 1.1 or 1.2 or 1.5 even.

This reminded me of Hales Signature Two's I've had. Heavy inert cabinet, M-T-M driver arrangement in a sealed box. With a Q of slightly below .7 for the speakers. Good response to only 50 hz really, but with the sealed box and a sturdy amp it had a big speaker sound. It also was uncolored enough I was happy with them even though I'm a lifelong panel speaker fan.

Nice experiment :) So not purely down to bass/frequency response in your instance.

Like I said in the OP the bass in those bass horns from the thread I linked (I need to fix that picture link) sounded like truly big speakers with the full force of the orchestra (Budapest Festival Orchestra recording.

edit: not exactly sure what the bass radiation pattern would be on them compared to the Kii. I removed one statement until I can dig up some measurements.
 
Last edited:

March Audio

Master Contributor
Audio Company
Joined
Mar 1, 2016
Messages
6,378
Likes
9,317
Location
Albany Western Australia
That's actually in line with many displays. Even 50ms isn't uncommon at all, especially if you don't put it in game mode. I have an over the top 165Hz gaming monitor and its absolute input latency is still 24ms worst case. Most people here aren't gamers anyway so while it's something I would certainly check on I don't think typical ASR users would find it a concern.
The latency can be significant. With my own dsp speakers I have the xo performed externally by acourate software. With a 24 core xeon bashing away at it the latency is still 27ms. High resolution Fir filters cause long delays, often unacceptable for video use with sensible levels of computing power.

Screenshot_20180911-083505.jpg
 

maverickronin

Major Contributor
Forum Donor
Joined
Jul 19, 2018
Messages
2,527
Likes
3,308
Location
Midwest, USA
The latency can be significant. With my own dsp speakers I have the xo performed externally by acourate software. With a 24 core xeon bashing away at it the latency is still 27ms. High resolution Fir filters cause long delays, often unacceptable for video use with sensible levels of computing power.

That is getting into the range where it can matter. Do stand alone DSP boxes with ASICs or FPGAs do any better?
 

andreasmaaan

Master Contributor
Forum Donor
Joined
Jun 19, 2018
Messages
6,652
Likes
9,399
Spent a little time listening to big speakers with the bottom end rolled off. They don't sound like small speakers. They sound like big speakers playing a recording with a lightweight low end. Haha! This was 12 db/octave roll off at 100, 125 and 150 hz.

So I tried 24 db/octave, and still sounds like a big speaker with a chopped off bottom on the recording.

Finally some progress. 24 db/octave roll off at 150 hz. Followed by a 6 db peak above flat at 80 hz (not uncommon in some small ported speakers), which creates a slight dip between those two points. Then rolled off the response an extra amount (about 36 db/octave) below 60 hz. Now it does begin to sound like a small speaker. And doing this digital EQ on a music file likely doesn't cause all the phase nasties you get with a real ported speaker made with a Q of 1.1 or 1.2 or 1.5 even.

This reminded me of Hales Signature Two's I've had. Heavy inert cabinet, M-T-M driver arrangement in a sealed box. With a Q of slightly below .7 for the speakers. Good response to only 50 hz really, but with the sealed box and a sturdy amp it had a big speaker sound. It also was uncolored enough I was happy with them even though I'm a lifelong panel speaker fan.

Interesting post - I've never tried this kind of experiment.

Are you sure that the "nasties" (by which I assume you mean audible artefacts) you're referring to in relation to high-Q ported speakers are really phase nasties and not corresponding frequency response nasties? All the research evidence to date would point very much to the latter.
 

watchnerd

Grand Contributor
Joined
Dec 8, 2016
Messages
12,449
Likes
10,408
Location
Seattle Area, USA
Unlike intrusive analog processing that adds its own set of artifacts, many much worse than digital...

But those analog artifacts are natural and smooth like waves from a tube amp.....or the ocean....or the hair of an angel.

Unlike those rigid, fascist, sharp, inhuman digital artifacts that would only sound nice to mammal-exploiting AI machines.

2hnqob.jpg
 

watchnerd

Grand Contributor
Joined
Dec 8, 2016
Messages
12,449
Likes
10,408
Location
Seattle Area, USA
Spent a little time listening to big speakers with the bottom end rolled off. They don't sound like small speakers. They sound like big speakers playing a recording with a lightweight low end. Haha! This was 12 db/octave roll off at 100, 125 and 150 hz.

So I tried 24 db/octave, and still sounds like a big speaker with a chopped off bottom on the recording.

Finally some progress. 24 db/octave roll off at 150 hz. Followed by a 6 db peak above flat at 80 hz (not uncommon in some small ported speakers), which creates a slight dip between those two points. Then rolled off the response an extra amount (about 36 db/octave) below 60 hz. Now it does begin to sound like a small speaker. And doing this digital EQ on a music file likely doesn't cause all the phase nasties you get with a real ported speaker made with a Q of 1.1 or 1.2 or 1.5 even.

This reminded me of Hales Signature Two's I've had. Heavy inert cabinet, M-T-M driver arrangement in a sealed box. With a Q of slightly below .7 for the speakers. Good response to only 50 hz really, but with the sealed box and a sturdy amp it had a big speaker sound. It also was uncolored enough I was happy with them even though I'm a lifelong panel speaker fan.

I'm trying to sum the math and mentally draw a frequency response graph to simulate what you did....

Something kind of like this?

66BA25fig3.jpg



Apropos of nothing, really, one of the best 2 ways I've ever owned / heard, within their limits, were the NHT Super Zeros.

1835956-pair-nht-super-zero-bookshelf-speakers-piano-black-100w.jpg



Super tiny, solid like a brick, sealed enclosure. Tiny 4.5" mid/"woofer" (for a pitiful definition of "woofer"), with a teensy 4.5" driver, laughable -3 dB point at 85 Hz, 2nd order crossover. But I swear to Max Planck, from the midrange up, and at low-modest volumes they crushed the vaunted LS 3/5A, and even gave some ML electrostats I owned a run for the money in the realm of detail.

I believe they had a bass Q of 0.5.

I tried many many many times to integrate them with a subwoofer (analog crossovers, class A/B amps), and never could get the system to gel. The differences in transfer functions was too obvious. So I sold them.

Should try to find another pair and do it with more modern DSP? Hmmm...
 
Last edited:

March Audio

Master Contributor
Audio Company
Joined
Mar 1, 2016
Messages
6,378
Likes
9,317
Location
Albany Western Australia
That is getting into the range where it can matter. Do stand alone DSP boxes with ASICs or FPGAs do any better?
It depends on what type of filtering they use. IIR filters cause minimal latency but won't be phase linear. FIR filters can be PL but cause delay proportional to filter resolution and sample rate.

Units like minidsp can do FIR filtering but simply don't have the processing power (sharc DSP) to achieve decent filter resolution, so realistically they are IIR only and have manageable latency.

I recently asked Uli (acourates designer) about latency in his software

Alan,

a lot of questions :)

1. Asio buffer
This simply means the number of samples handled by the Asio driver in the buffer. So the input samples are collected until the buffer is full. Then Asio switches to a second buffer for fill-up. In the meantime the first buffer can be read and processed. The same happens with the output buffers. The player writes into a buffer until it is full and stops writing. In the meantime Asio sends the content of a second buffer to the output sequentually. If the buffer is empty Asio switches to the first buffer. So the buffer switch time is buffersize/samplerate. With small buffers the time is shorter and thus the computer has to process more switches. Too many switches rise the CPU load.

2. FFT partition
The FFT calculation and the required CPU load or power depends on the partition size. Samller partitions increase the CPU load. Usually AC uses a partition of 32768 samples. This means that every partitionsize/samplerate a new FFT calculation is done. So to fill one partition there are 32 Asio buffer switches (with Asio buffersize 1024).

3. Filter delay
With the Acourate (near) linearphase filters the filter delay is half filterlength/samplerate. This is independent of the Asio buffersize or FFT partitionsize.

4. Total delay
The total delay can be assumed by sum of filter delay + 2*partitionsize + 2*buffersize. The delay can be reduced by using a minimumphase filter (downside = no phase correction, amplitude correction only. This is anyway ok with video as the brain is busy by picture processing). Furthermore the minphase filterlength can be reduced, this allows to also reduce the FFT partition and the Asio buffersize

5. Non-uniform FFT partitioning
This is the high end of FFT processing. But AC does not use it because up to now I have not fully understood how to implement it including multi-threading. It starts with short partitions and thus reduces the latency very much. It requires minphase filters anyway.
As Acourate basically intends to improve the udio playback it uses (near) linearphase filters and thus the actual implementation with uniform partition sizes is IMHO ok.

6. CPU load
This is simply the load index as also shown by the task manager. A CPU can be very busy but still do nearly nothing e.g. by a bad algorithm. If the buffers are too small the CPU load goes higher because of a lot of buffer switches and thus program interruptions.

7. Realtime index
The index is an information about how quick the calculation is done. A simple example: a CD player takes 30 minutes to read and play a CD. Realtime index = 1x. Now a CD ripper does it with a speadup of factor 10. The index is 10x.You can also define the index in this case as 1/10 = 0.1 or 10%. In your given example (picture) the convolution time is 60% of the Asio bufferswitch time. The buffer is pretty small with 256 samples and thus also the CPU load is high. In case of (near) linearphase filters this does not really make sense as the filter latency is much higher. You still recognize a big latency.

8. Multicore CPU
AC uses multi-threading. Thus you get the best behaviour if you have a CPU core for each channel as each convolution is carried out on its own core. Hyperthreading does not help for this case as hayperthreaded cores do not contain a numeric processor, they simply cannot process the convolution maths.

9. Latency optimization
You can use powerful CPUs with many cores. This helps to lower the CPU load and to do the calculations quickly. But the main brake is the phase correction. So if you can dispense with phase correction use minimumphase filters. Then shorten the filter if necessary and reduce the buffer sizes.

10: Video optimization
Some video players like JRiver allow to delay the picture. So JRiver even allows to use linearphase filters because it knows about picture+audio+filter delay. Wheras AC does not know anything about the picture.

- Uli
 
Last edited:
Top Bottom