• Welcome to ASR. 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!

How high is the distortion of your systems and what would you consider "clean" Output? - Distortion optimization and Listening Levels

Plompudu

Member
Joined
Jun 13, 2023
Messages
33
Likes
18
I'm currently looking into optimizing the Distortion / output capability of my System (2x Swissonic A306; 1x SVS SB-1000; 1x SVS SB-2000) and would love some feedback on my results!

The adjustments I've done so far via a MiniDSP 2x4HD are:
  • SB-1000 HP at 36Hz (LR 48dB/oct) ; LP at 80Hz (24dB/oct Butterworth)
  • SB-2000 no HP ; LP at 80Hz (24dB/oct Butterworth)
  • A306 HP at 80Hz (24dB/oct Butterworth)
My main Question is: how high is the distortion of your systems and what would you consider "clean" Output? (please tell your listening levels as well)
And would you go away from the Dolby recommended +10dB of the LFE-Channel or using a Low-Shelf Filter to have less distortion?


Here is for example the RTA of a random song in my playlist (Vianova - Melancronic) played at -20dB (using Dolby Reference Level, +10dB LFE) compared to a ~81dB sweep (-20dB sweep of L+R):
1752919565374.png
1752919505230.png

1752919593313.png


And here is the Intro of Kaiju no.8 at -13dB (Dolby reference; +10dB LFE) compared to a 10dB higher sweep:
1752921292604.png

1752921345621.png

1752921380439.png


And since the peaks reached up to ~100dB(Z) during the music test - I've also done a 95dB and 100dB:
1752921516256.png

1752920100741.png


I didn't do a test for the ~106dB of the show, since there was noticeable compression (or rather limiting) during the sweep:
1752921553143.png

1752922591151.png


Ohh and here is a response copy of the show, with the SPL adjusted to reference Level => I couldn't watch it at reference due to hitting the compression of the Subwoofers, but reference is too loud in my experience so yeah:
1752923034197.png
 
Last edited:
There's nothing to optimize with distortion other than not playing at a level where it becomes subjectively objectionable, using a sub for the LF and having an infrasonic filter so your sub doesn't dig too deep below 20Hz or so, if it can even produce 20Hz. We have no real control beyond that.

Distortion sounds, in my experience, outside of obviously scratchy and clipped territory, like the music is too loud. Really good, expensive speakers with low distortion sound clean even at high SPLs. It's an important distinction, loudness (subjective) and SPL (objective).

Try this: https://www.klippel.de/listeningtest/
 
There's nothing to optimize with distortion other than not playing at a level where it becomes subjectively objectionable, using a sub for the LF and having an infrasonic filter so your sub doesn't dig too deep below 20Hz or so, if it can even produce 20Hz. We have no real control beyond that.

Distortion sounds, in my experience, outside of obviously scratchy and clipped territory, like the music is too loud. Really good, expensive speakers with low distortion sound clean even at high SPLs. It's an important distinction, loudness (subjective) and SPL (objective).

Try this: https://www.klippel.de/listeningtest/
You're right - what I meant was looking at finding a good tradeoff between linearity, distortion, Decay time and loudness.

This is just turning off the HP of the SB-1000 (without re-aligning the Phase) - the increase in the range of the 27Hz dip looks nice, but I would assume that it's not worth it looking at the distortion and based on it sounding very clear compared to before.
1753003555025.png

1753003717836.png

1753003729885.png

1753003761266.png

1753003772113.png



Thanks for the link!


When doing the Test with the HD600 EQed to a Harman-ish response at 94dB at 1kHz can I get this:

1753001214981.png

1752928551657.png



When using my speakers at -24dB:
1753002380125.png


1753002398840.png

At -20dB when concentrating i get:
1753002608933.png
1753002621504.png


So I should aim for roughly less then -30dB / ~3% THD as a Threshold?
Would you say based on that that I should try a higher Crossover to the mains with my front Subwoofer (SVS SB-2000) for my listening at more then -20dB / ~101dB peaks?

1753003073208.png


Would you recommend a steep HP to filter Infrasonics and at what frequency in my case?
And again would you go away from the Dolby recommended +10dB LFE / adding a Low shelf Filter to be able to play louder.
What is your experience with using a Compressor?
 
It looks like you did a lot of work to understand the limits of your system - at least in graphs.

The subs you have are pretty good but would not expect infrasonic performance from them. That would be the the larger drivers and amps in the SVS line up. How far you want to push your subs is really a matter of preference. Graphs are helpful, but ultimately you need to decide which acceptable graphs sound best to you. Starting to roll off subs at 30 or even 40hz is not unusual I would say.

I would generally follow +10dB guideline for LFE, as that is how material was mastered. It was meant to be played at those levels.

Going to Compressor is kind of Inception level 2. Various titles have been mastered with various levels of bass for different scenes. Some deploy what I call real life resemblance, so that when galactic starship pans across the screen the whole hell breaks loose, but then if you calibrate for that peak, you end up with very little excitement for the rest 99.99% of the movie when it comes to bass. For such titles, I would find Compressor useful. Did not try it yet, but planning to do so once summer season is over and not much to do outside when November rains come.

But theoretically this could work, just not sure how it would actually sound. Probably OK, because many titles do not deploy such "real life" dynamics in bass response, recognising that people will still get the "huge" moments if 10dB louder, rather than 20dB louder.

Otherwise, the biggest issue with bass IMO is decay. That is really difficult to deal with as it is all room. I have managed to tame it down to some 30hz but since my 2x2 DBA is not perfect, it starts to ring quite a bit below that. Don't like it so I roll it off.

Finally, Dirac ART is apparently a miracle maker for all bass from 20-150hz. Lots of positive feedback thus far and it is rolling out by end of this year on D&M gear. Previously available only on Storm which is quite pricy. Now a $1K Denon 3800H with another $900 for full Dirac package will be apparently able to do wonders in taming the bass beyond what DLBC could ever do.
 
I’m believe it’s been stated here before that distortion cannot be accurately determined with in-room measurements.
Well, for subs you have compression tests that will show you when sub is no longer able to play louder at given frequency. This limit will then imply distortion.

Not good to push the hardware to its edge, but ultimately it is up to the owner to decide what is acceptable to them. Loud peaks usually don't last long, but still could permanently damage the sub.
 
You're right - what I meant was looking at finding a good tradeoff between linearity, distortion, Decay time and loudness.

This is just turning off the HP of the SB-1000 (without re-aligning the Phase) - the increase in the range of the 27Hz dip looks nice, but I would assume that it's not worth it looking at the distortion and based on it sounding very clear compared to before.View attachment 464410
View attachment 464411
View attachment 464412
View attachment 464413
View attachment 464414


Thanks for the link!

When doing the Test with the HD600 EQed to a Harman-ish response at 94dB at 1kHz can I get this:

View attachment 464398
View attachment 464215


When using my speakers at -24dB:
View attachment 464401

View attachment 464402
At -20dB when concentrating i get:
View attachment 464403View attachment 464404

So I should aim for roughly less then -30dB / ~3% THD as a Threshold?
Would you say based on that that I should try a higher Crossover to the mains with my front Subwoofer (SVS SB-2000) for my listening at more then -20dB / ~101dB peaks?

View attachment 464407

Would you recommend a steep HP to filter Infrasonics and at what frequency in my case?
And again would you go away from the Dolby recommended +10dB LFE / adding a Low shelf Filter to be able to play louder.
What is your experience with using a Compressor?
This is how I would approach this issue.

First, I would not stress the speakers and subs by deliberately forcing them to play at levels in which they show compression. That could easily damage them.

These are the speakers and subs:
  • Swissonic A306: https://www.erinsaudiocorner.com/loudspeakers/thomann_swissonic_a306/ When looking at THD at 86dB vs. 96dB SPL, we see that for a 10dB increase, the distortion in the LF jumps from -50 to -30 below 300Hz, including significant levels of 3rd, 4th and 5th harmonics. This is at 1meter. So I would avoid using these outside of a bedroom or any other setting where I would sit farther than 1 or 2 meters away. I would set the crossover to the subs more or less conventionally at 80-100Hz with a standard 24dB slope.
  • SVS SB-1000 & 2000: https://www.audiosciencereview.com/forum/index.php?threads/subwoofer-comparison.20494/ I plotted the burst tone test results in REW. These results represent the absolute maximum output of the subs tied to somewhat coarse perceptual limits, and it seems like there is 10dB of SPL over the minimum hearing threshold at 20Hz. So there a limited amount of headroom and you can count on a little boost from boundary coupling. But it's obvious based on the shape of the resulting curves that below 40-50Hz these subs will not produce much.
8.png


When using my speakers at -24dB:
You aren't providing any information with that number. -24dB of what? A dB is a ratio, not an absolute value, unless there is a defined reference.

Would you recommend a steep HP to filter Infrasonics and at what frequency in my case?
I would use a high pass filter at 20Hz.

And again would you go away from the Dolby recommended +10dB LFE / adding a Low shelf Filter to be able to play louder.
I would ignore this.
What is your experience with using a Compressor?
I would not use a compressor or limiter (same thing but with an infinite ratio). Both compression and limiting inject a lot of distortion into the signal. They are unnecessary at home. These are studio or live sound devices that are used either for musical effect or for protecting the gear.

I get the feeling you aren't convinced, but this is really not worth your effort in a home setting. What you're doing is done in live settings where you have kilowatt amplifiers powering very sensitive speakers and unpredictable performers that could do something that would send an insanely loud signal through the mixing board. The most important aspect of having a measurement microphone is to correctly integrate the sub with the speakers and then EQ to even out the frequency response. Beyond that, this is done by ear. If there is nothing obviously distorting, the setup is working correctly and within its operating range.

Play with the shape of the bass curve. Some people like a boost, others like a tilt. Once you have EQed to flat, this can be done in software without need of measurement other than to check the curve once in a while.
 
This is how I would approach this issue.

First, I would not stress the speakers and subs by deliberately forcing them to play at levels in which they show compression. That could easily damage them.

These are the speakers and subs:
  • Swissonic A306: https://www.erinsaudiocorner.com/loudspeakers/thomann_swissonic_a306/ When looking at THD at 86dB vs. 96dB SPL, we see that for a 10dB increase, the distortion in the LF jumps from -50 to -30 below 300Hz, including significant levels of 3rd, 4th and 5th harmonics. This is at 1meter. So I would avoid using these outside of a bedroom or any other setting where I would sit farther than 1 or 2 meters away. I would set the crossover to the subs more or less conventionally at 80-100Hz with a standard 24dB slope.
  • SVS SB-1000 & 2000: https://www.audiosciencereview.com/forum/index.php?threads/subwoofer-comparison.20494/ I plotted the burst tone test results in REW. These results represent the absolute maximum output of the subs tied to somewhat coarse perceptual limits, and it seems like there is 10dB of SPL over the minimum hearing threshold at 20Hz. So there a limited amount of headroom and you can count on a little boost from boundary coupling. But it's obvious based on the shape of the resulting curves that below 40-50Hz these subs will not produce much.
View attachment 464484


You aren't providing any information with that number. -24dB of what? A dB is a ratio, not an absolute value, unless there is a defined reference.


I would use a high pass filter at 20Hz.


I would ignore this.

I would not use a compressor or limiter (same thing but with an infinite ratio). Both compression and limiting inject a lot of distortion into the signal. They are unnecessary at home. These are studio or live sound devices that are used either for musical effect or for protecting the gear.

I get the feeling you aren't convinced, but this is really not worth your effort in a home setting. What you're doing is done in live settings where you have kilowatt amplifiers powering very sensitive speakers and unpredictable performers that could do something that would send an insanely loud signal through the mixing board. The most important aspect of having a measurement microphone is to correctly integrate the sub with the speakers and then EQ to even out the frequency response. Beyond that, this is done by ear. If there is nothing obviously distorting, the setup is working correctly and within its operating range.

Play with the shape of the bass curve. Some people like a boost, others like a tilt. Once you have EQed to flat, this can be done in software without need of measurement other than to check the curve once in a while.
As I noted in my previous post, I did not try compressor, but it was recommended by number of members to deal with the issue of certain mixes. And sounds like a great idea in theory. Not sure what distortion are you talking about?

Given how differently the content is mixed, I do use 4 different presets with different curves, settings, etc. And still find myself adjusting manually each of them based on the content. There is unfortunately no "reference" when it comes to soundtracks.
 
There's a similar threads here:


 
It looks like you did a lot of work to understand the limits of your system - at least in graphs.

The subs you have are pretty good but would not expect infrasonic performance from them. That would be the the larger drivers and amps in the SVS line up. How far you want to push your subs is really a matter of preference. Graphs are helpful, but ultimately you need to decide which acceptable graphs sound best to you. Starting to roll off subs at 30 or even 40hz is not unusual I would say.

I would generally follow +10dB guideline for LFE, as that is how material was mastered. It was meant to be played at those levels.

Going to Compressor is kind of Inception level 2. Various titles have been mastered with various levels of bass for different scenes. Some deploy what I call real life resemblance, so that when galactic starship pans across the screen the whole hell breaks loose, but then if you calibrate for that peak, you end up with very little excitement for the rest 99.99% of the movie when it comes to bass. For such titles, I would find Compressor useful. Did not try it yet, but planning to do so once summer season is over and not much to do outside when November rains come.

But theoretically this could work, just not sure how it would actually sound. Probably OK, because many titles do not deploy such "real life" dynamics in bass response, recognising that people will still get the "huge" moments if 10dB louder, rather than 20dB louder.

Otherwise, the biggest issue with bass IMO is decay. That is really difficult to deal with as it is all room. I have managed to tame it down to some 30hz but since my 2x2 DBA is not perfect, it starts to ring quite a bit below that. Don't like it so I roll it off.

Finally, Dirac ART is apparently a miracle maker for all bass from 20-150hz. Lots of positive feedback thus far and it is rolling out by end of this year on D&M gear. Previously available only on Storm which is quite pricy. Now a $1K Denon 3800H with another $900 for full Dirac package will be apparently able to do wonders in taming the bass beyond what DLBC could ever do.
I'm actually a bit surprised by their Infrasonic capabilities due to using them in a near perfectly sealed room and therefore getting very good room gain (UMIK Calibration only goes down to 10Hz, so a roll off below it has to be assumed) and likely having a ~10 and ~20Hz room mode from the top to the bottom floor of the house (or maybe floor resonance, since I feel it even at levels that should be way below the tactile threshold?):
1753039931270.png


I'll try a preset with a HP filter and see if that's what I'm looking for!
I also took a look at the Dolby Spec and they only ask for 31.5Hz, so I'll try a steep 31.5 HP-Filter first.


https://professional.dolby.com/siteassets/cinema-products---documents/dolby-atmos-specifications.pdf
1753040676591.png


Regarding the Decay it's mostly a Issue in the Midbass for my System (that I'm still unsure about how to treat due to placement constraints) and the nasty ~28Hz and~20Hz mode that I can't treat with the SB-1000 due to it's limited output capabilities:
1753041434353.png
1753041274705.png

1753041640910.png



That definitely sounds interesting, but spending ~2000€ is outside of my budget range for now.
Next upgrade I've been considering is a replacement for the SB-1000 if I find something good on the used market or getting Ascilab A6B Speakers if I get to save a lot during the next semesters working part time in the battery research department of my university as a programmer.
 
This is how I would approach this issue.

First, I would not stress the speakers and subs by deliberately forcing them to play at levels in which they show compression. That could easily damage them.

These are the speakers and subs:
  • Swissonic A306: https://www.erinsaudiocorner.com/loudspeakers/thomann_swissonic_a306/ When looking at THD at 86dB vs. 96dB SPL, we see that for a 10dB increase, the distortion in the LF jumps from -50 to -30 below 300Hz, including significant levels of 3rd, 4th and 5th harmonics. This is at 1meter. So I would avoid using these outside of a bedroom or any other setting where I would sit farther than 1 or 2 meters away. I would set the crossover to the subs more or less conventionally at 80-100Hz with a standard 24dB slope.
  • SVS SB-1000 & 2000: https://www.audiosciencereview.com/forum/index.php?threads/subwoofer-comparison.20494/ I plotted the burst tone test results in REW. These results represent the absolute maximum output of the subs tied to somewhat coarse perceptual limits, and it seems like there is 10dB of SPL over the minimum hearing threshold at 20Hz. So there a limited amount of headroom and you can count on a little boost from boundary coupling. But it's obvious based on the shape of the resulting curves that below 40-50Hz these subs will not produce much.
View attachment 464484


You aren't providing any information with that number. -24dB of what? A dB is a ratio, not an absolute value, unless there is a defined reference.


I would use a high pass filter at 20Hz.


I would ignore this.

I would not use a compressor or limiter (same thing but with an infinite ratio). Both compression and limiting inject a lot of distortion into the signal. They are unnecessary at home. These are studio or live sound devices that are used either for musical effect or for protecting the gear.

I get the feeling you aren't convinced, but this is really not worth your effort in a home setting. What you're doing is done in live settings where you have kilowatt amplifiers powering very sensitive speakers and unpredictable performers that could do something that would send an insanely loud signal through the mixing board. The most important aspect of having a measurement microphone is to correctly integrate the sub with the speakers and then EQ to even out the frequency response. Beyond that, this is done by ear. If there is nothing obviously distorting, the setup is working correctly and within its operating range.

Play with the shape of the bass curve. Some people like a boost, others like a tilt. Once you have EQed to flat, this can be done in software without need of measurement other than to check the curve once in a while.
I sit at 1.8m distance at my MLP or sometimes at arms length when working or gaming with keyboard and mouse with a different preset.
I'm already using a 80Hz Crossover, but I could definitely try a 100Hz crossover, but last time I did I had some localization Issues and stopped trying early on.

I'm referencing Dolby reference Level like in the comments before:
-24 => -24dB + 105dB +10dB(LFE) +6dB(Stereo Bass management) => up to 97dB peaks

Thanks for your Input!
 
Well infrasonic capabilities are relative to SPL. These subs definitively landed in a good room, but compressing at 100dB is not a great sign. Referencd level for LFE is 115dB, but 100dB would be still good if listening to -15dB master volume. Also, you decay seems to be going up from 35hz, which is also quite normal. Really difficult to control theses frequencies.

You should choose the option that sounds best for you below 35hz. I don't think that any graph will tell you that with your current gear. When replacing the sub, I would go with something close to the one your are keeping, which is probably 2000.
 
Well infrasonic capabilities are relative to SPL. These subs definitively landed in a good room, but compressing at 100dB is not a great sign. Referencd level for LFE is 115dB, but 100dB would be still good if listening to -15dB master volume. Also, you decay seems to be going up from 35hz, which is also quite normal. Really difficult to control theses frequencies.

You should choose the option that sounds best for you below 35hz. I don't think that any graph will tell you that with your current gear. When replacing the sub, I would go with something close to the one your are keeping, which is probably 2000.
The 115dB are only for the F3 at 31.5Hz Hz in the Dolby Spec - wanting it for Infrasonics seems to be way above the specification, so it would just be a nice to have?
But I guess since most people want a linear response it's gotten more known that Infrasonics should reach these levels as well in a HT?

1753046655156.png

1753046619154.png


I also think that the approach of basing the reference level on the room size like for the TV-standard for example seems reasonable based on my experience, since using a standard for giant theaters in a home setting seems problematic, but that's a separate discussion - https://www.audiosciencereview.com/...ference-sound-pressure-level-flowchart.11069/ :
1753047508211.png
 
Well, Dolby specs recommend 115dB peak limit. It is not the law, and some tracks might easily pass that mark. Roll-off of the bass has more to do with artistic intent, or with THX hat was agreeable to rolling off LFE beyond 30hz. Can't remember the whole story but probably not materially wrong.

Lots of movies are rolled off at 30hz and nurtured even at 50hz, and people use BEQ that is a process to restore the "lost" content. But that is a practice that is relying on credible information, but if anything goes wrong, your will be frying your subs for sure. Not that it happens, but just not a fan of it as know that if I let my subs loose, they can go really wild.
 
I'm actually a bit surprised by their Infrasonic capabilities
I would be a little skeptical of what exactly you are measuring and hearing below 30 Hz. LF measurements in-room are fraught with complications. Those subs have limited output below 30 Hz and very high distortion even at 20 Hz. Both are tested here https://www.audiosciencereview.com/forum/index.php?threads/i-measured-10-subwoofers.49042/ When distortion starts to get very high and the fletcher-muson effect comes into play (our hearing is ~20dB more sensitive to 40 Hz than to 20 Hz) you may be hearing and feeling the distortion rather than the fundamental. If you are serious about low distortion (<10% ) and high SPL below 40 Hz you need some serious large subs which are expensive and most likely won't add as much to the listening experience as you think. I just went through this (upgraded from 2 SVS SB3000 subs to some very large 18" subs) and while I can now play with low distortion and high SPL below 20 Hz it is not as useful as I had hoped. First off there is little content in most music below 40 Hz and second my room can not handle it. At about 22 Hz the drywall starts pounding against the studs and it ruins the effect. It is OK for brief transients like drums but for a 17 Hz pipe organ recording it just rattles the whole room.

In your case I would experiment with some high pass filters on the low bass to see what sounds better to you. You may like it full range but you might find rolling off the bass below 40 Hz or 30 Hz sounds better to you. It will probably change with the type of content and you may prefer something different for movies and music.
 
As I noted in my previous post, I did not try compressor, but it was recommended by number of members to deal with the issue of certain mixes. And sounds like a great idea in theory. Not sure what distortion are you talking about?
Compression changes the waveform when the signal exceeds the threshold by cutting the peak. A consequence of that, nonlinearities are introduced.

You can use compression to reduce dynamic range, but there a lot of complex consequences to the sound.
 
I just went through this (upgraded from 2 SVS SB3000 subs to some very large 18" subs) and while I can now play with low distortion and high SPL below 20 Hz it is not as useful as I had hoped. First off there is little content in most music below 40 Hz and second my room can not handle it. At about 22 Hz the drywall starts pounding against the studs and it ruins the effect. It is OK for brief transients like drums but for a 17 Hz pipe organ recording it just rattles the whole room.
I had similar experience in my living/dining/kitchen/hallways room. My subs are only 13.8" but 4 of them (6 drivers, 4.4kW of amps) and can raise hell even in relatively large 6,000+ cft space.

While the construction is reinforced concrete, the room still rattles from sub 25hz frequencies above 100dB. The large 3 m window/door aluminium construction starts resonating like crazy so I am rolling off subs at 30hz to prevent that.

At one point, I suspect that one of my subs broke the window as well. Happened 2 days after I placed it in that position. The window company that replaced it said it was probably due to the fact that glass was not tightly sealed any more.

IMG_7280.jpeg
IMG_7279.jpeg


EDIT: Just added the second photo showing the corner loaded Arendal 1723 2V that is the main suspect in the case o_O
 
Last edited:
I had similar experience in my living/dining/kitchen/hallways room. My subs are only 13.8" but 4 of them (6 drivers, 4.4kW of amps) and can raise hell even in relatively large 6,000+ cft space.

While the construction is reinforced concrete, the room still rattles from sub 25hz frequencies above 100dB. The large 3 m window/door aluminium construction starts resonating like crazy so I am rolling off subs at 30hz to prevent that.

At one point, I suspect that one of my subs broke the window as well. Happened 2 days after I placed it in that position. The window company that replaced it said it was probably due to the fact that glass was not tightly sealed any more.

View attachment 464708View attachment 464715

EDIT: Just added the second photo showing the corner loaded Arendal 1723 2V that is the main suspect in the case o_O
Wow broken glass!
 
Yes, but as the contractors noted, it was likely not sealed well either from the start of after 15 years in use. The rubber between the glass and the frame ages as well. We have couple of flats in the same building, and nothing like this happened in others.

Subs don't really, usually break windows, but this is a suspect case as no other force could really do it.

Just a reminder as how powerful subs can be if you assemble a large fleet of them. And how not all structures are well equipped to deal with such power.

I found this review interesting in many ways, the one that I would call in this case was rebuilding the whole HT room/walls as they could not withstand the power of subs.

 
Back
Top Bottom