Does DSP belong in State of the Art Systems?

Hear Here · Feb 9, 2024

Chops said:
And as far as Jim Smith is concerned, you're once again incorrect. Jim uses a pair REL S-812 subwoofers in his own system, so... Not to mention he does the same with his clients systems if they have subs.

Did you see the long video of Jim setting up Mike's system (of Sun Coast Audio) in his own home? He set up Mike's Avantgarde Mezzo XD speakers (very similar to my own Duo XDs) and I'm 99% sure there were no subs involved apart from the bass enclosures that are part of the AG system. I'm also 99% confident that no DSP would be found in the system, apart, from the XD software built into the AG's bass enclosures. By all accounts, reported many times on the AudioShark forum, Mike's system delivered sound that most listeners considered the best they'd ever heard - anywhere.

Whatever you claim, I consider there should be no need for external subs in well treated rooms if carefully-chosen first-class full-range speakers are properly set up and room treatment (often only domestic features such as carpets, curtains, etc) is attended to.

I am no more interested in measuring the response at my listening position than I would when sitting in a live performance venue. For me, if the sound from my own system is as close to a live performance as it is (though no doubt a little less so than Mike's after Jim had worked his magic), I am more than happy not adding extra speakers and with no signal messing in a vain attempt to get them sounding better.

Others with less able speakers may well find that extra speakers need to be added, but that has extra complication and much more difficulty in setting up. I prefer a simple system that avoids this unnecessary complication.

We obviously have diametrically opposite views on this matter, so perhaps we should leave it as stalemate! Good to talk and compare thoughts though.

Analog Scott · Feb 9, 2024

Hear Here said:
I am no more interested in measuring the response at my listening position than I would when sitting in a live performance venue.

I would highly recomend you reconsider. Room EQ Wizard is a free program and the microphone that is used with REW is just over $100.00 bucks . The info you get from measuring is worth a lot more.

Hear Here said:
Others with less able speakers may well find that extra speakers need to be added, but that has extra complication and much more difficulty in setting up. I prefer a simple system that avoids this unnecessary complication.

It isn’t just about the speakers. It’s also about the room. With REW you may find issues you just didn’t know you had and often they can be addressed without complicating the existing system. Sometimes a simple acoustic treatment addition or small adjustment in speaker or listener position can give you substantial improvements that you would not have otherwise known could be done.

And on top of all that it really is interesting to see how our systems measure.

sigbergaudio · Feb 10, 2024

Hear Here said:
Whatever you claim, I consider there should be no need for external subs in well treated rooms if carefully-chosen first-class full-range speakers are properly set up and room treatment (often only domestic features such as carpets, curtains, etc) is attended to.

If two speakers can deliver even in-room bass response is completely dependent on room and placement. In some situations this is indeed possible at least for a limited sweetspot, in others it is not.

Carpets and curtains does nothing for the bass, but it is indeed helpful for the overall acoustics.

Hear Here said:
Others with less able speakers may well find that extra speakers need to be added, but that has extra complication and much more difficulty in setting up. I prefer a simple system that avoids this unnecessary complication.

This is true, it is not necessarily trivial. And you want them to play up to at least 80hz, which typically means you need capability to high pass your speakers. I fully understand that not everyone want to go down this route. But done well, it is an improvement in almost any system and situation.

Bjorn · Feb 10, 2024

Atmasphere said:
Put another way, as I pointed out earlier on in this thread, do the Distributed Bass Array first, which usually takes care of 95% of your problems, then use the DSP and bass traps to do the other 5%. If you do it the other way around you are wasting your money as the effects will be minimal.

That's a strong statement. Below is an example of only treatment being used before and after (no EQ applied either). Price of this treatment is in this case much lower than "distributed bass array".

before an after freq response at 4m distance.jpg

Waterfall before:

Waterfall after:

The drawback with something like double bass array is that it doesn't work in the mid and upper bass. And we are more sensitive to this bass area vs the sub frequencies. SBA (single bass array) is a better approach here IMO and can also be combined with high quality treatment for mids and highs on the rear wall.

Analog Scott · Feb 10, 2024

Bjorn said:
That's a strong statement. Below is an example of only treatment being used before and after (no EQ applied either). Price of this treatment is in this case much lower than "distributed bass array".

View attachment 125032

Waterfall before:
View attachment 125033

Waterfall after:
View attachment 125034

The drawback with something like double bass array is that it doesn't work in the mid and upper bass. And we are more sensitive to this bass area vs the sub frequencies. SBIR is a better approach here IMO and can also be combined with high quality treatment for mids and highs on the rear wall.

That is a night and day difference

Hear Here · Feb 10, 2024

Analog Scott said:
I would highly recomend you reconsider. Room EQ Wizard is a free program and the microphone that is used with REW is just over $100.00 bucks . The info you get from measuring is worth a lot more.

It isn’t just about the speakers. It’s also about the room. With REW you may find issues you just didn’t know you had and often they can be addressed without complicating the existing system. Sometimes a simple acoustic treatment addition or small adjustment in speaker or listener position can give you substantial improvements that you would not have otherwise known could be done.

And on top of all that it really is interesting to see how our systems measure.

Thanks. In fact I do have a UMIC-1 calibrated mic and REW and have done some test measurements, though I'm no expert at interpreting the results. I've also used this calibrated mic and the generic mics supplied with amps that feature Dirac Live, RoomPerfect and MARS, where an "auto adjust" filter of frequencies below about 200 Hz can be injected into the amp. In ALL cases, the sound without the filter has more top end sparkle than the sound with the filter, despite the fact that these "room correction" DSPs are only ADJUSTING low frequencies - the higher frequencies cannot avoid the filter in a single-amp system.

What these DSPs are doing if you think about it is not room correction at all (the room stays exactly the same), but messing with the nice flat signal that the amp should be delivering to the speakers in an attempt to iron out the response delivered by the speakers. In fact not even that - they attempt to mess further such that the frequency response at our individual listening position is flat. I don't approve of that method. The amp's output shouldn't be messed with, but other means taken to "correct" the sound reaching our ears as far as possible. Granted it may never be 100% free of anomalies, but careful choice of speaker type, positioning, room treatment, etc should be all that's needed. After all, we go to live music events and enjoy the performance without measuring its quality at our seat and every seat would have a different measurement. The less processing of the signal (the clue is in the abbreviation DSP), the better the potential sound. But it still needs effort and I believe that most of us don't buy the most suitable speaker and don't spend enough time and effort to get things as right as we could - doubtless myself included!

Analog Scott · Feb 10, 2024

Hear Here said:
Thanks. In fact I do have a UMIC-1 calibrated mic and REW and have done some test measurements, though I'm no expert at interpreting the results.

Cool! I think you can see a lot without being an expert. Pretty easy to pick out room modes and assess the extent of their presence. Also you can spot reflection issues above 100-150 hz.

Hear Here said:
I've also used this calibrated mic and the generic mics supplied with amps that feature Dirac Live, RoomPerfect and MARS, where an "auto adjust" filter of frequencies below about 200 Hz can be injected into the amp. In ALL cases, the sound without the filter has more top end sparkle than the sound with the filter, despite the fact that these "room correction" DSPs are only ADJUSTING low frequencies - the higher frequencies cannot avoid the filter in a single-amp system.

My philosophy is start with practical solutions first. Bass traps, targeted absorbers and speaker/listener position. I found the measurements to be very helpful for this.

Hear Here said:
What these DSPs are doing if you think about it is not room correction at all (the room stays exactly the same), but messing with the nice flat signal that the amp should be delivering to the speakers in an attempt to iron out the response delivered by the speakers.

Depends on the RC software. When I get back from China I will be auditioning the BACCH RC SP. it does actually do room correction much the same way it does cross talk cancellation. Should be interesting. But a very different beast.
And since I already have the BACCH and consider it essential for 2 channel stereo no reason not to use the RC too.

sigbergaudio · Feb 10, 2024

Hear Here said:
Thanks. In fact I do have a UMIC-1 calibrated mic and REW and have done some test measurements, though I'm no expert at interpreting the results. I've also used this calibrated mic and the generic mics supplied with amps that feature Dirac Live, RoomPerfect and MARS, where an "auto adjust" filter of frequencies below about 200 Hz can be injected into the amp. In ALL cases, the sound without the filter has more top end sparkle than the sound with the filter, despite the fact that these "room correction" DSPs are only ADJUSTING low frequencies - the higher frequencies cannot avoid the filter in a single-amp system.

A more careful approach is also possible. Using manual EQ to add 1-3 filters below 100-150hz goes a long way of evening out the response in most rooms, you don't have to employ an automatic calibration system like the ones you mention.

tony22 · Feb 10, 2024

Bjorn said:
That's a strong statement. Below is an example of only treatment being used before and after (no EQ applied either). Price of this treatment is in this case much lower than "distributed bass array".

View attachment 125032

Waterfall before:
View attachment 125033

Waterfall after:
View attachment 125034

The drawback with something like double bass array is that it doesn't work in the mid and upper bass. And we are more sensitive to this bass area vs the sub frequencies. SBIR is a better approach here IMO and can also be combined with high quality treatment for mids and highs on the rear wall.

Bjorn, that’s very impressive indeed. Would you consider (perhaps in a separate thread) going through the details of this before and after? Size of the room, what it looked like at the start, which treatments impacted each part of the observed improvement, etc? I for one would love to see that.

Bjorn · Feb 11, 2024

Atmasphere said:
The problem you have here is cancellation. No DSP or set of bass traps can fix that (been there done that). If you know of a way it can, I'm open to hearing about it.

The is wrong. Treatment can fix cancellations. I have already shown an example of that. If there's a cancellation somewhere, there's also pressure points with high level and resonances at the same frequency. When this is treated effectively, the cancellation will either be minimized or completely removed is the treatment is sufficiently effective.

Here's another example with a cancellation higher in frequency, at 120 Hz:

Frekvensrespons før og uten sidetiltak.jpg

Two Broadsorbor panels placed correctly did this:

Frekvensrespons med to stk Broads 150x80x10cm.jpg

Using four Broadsorbor panels improved it further and the cancellation was almost gone:

Frekvensrespons med fire stk Broads 150x80x10cm.jpg

Or here's before and after adding two Modex Edge bass traps where a cancellation between 50 H and 60 Hz is almost eliminated.

frequency response with and without at earheight_ blue is without.jpg

Bjorn · Feb 11, 2024

Brucemck2 said:
Agree 100%.

In reply to Analog Scott ...

The two rooms and systems were very different. My first dedicated room was designed by Russ Herschelman. We built a room within a room using a detached two car garage as the shell. There was extensive use of channels for the walls and ceiling with two layers of sheetrock of different dimensions. There was a false floor that was a foot high with channels that ran the length of the room that were tuned to various frequencies, with vent openings that could be adjusted. Most of the room treatments were RPG and approximately one third of the walls and ceilings were used. Speakers were Krell Lat2 with two Krell subwoofers. I also had custom built line arrays with Accuton mids and Fountek ribbons. The DSP was provided by early versions of Accourate and later a heavily modified Tact processor. (The analog stages were custom built, the clock and power supplies were heavily modified, etc. etc.) This was a combined analog (Basis, Graham, Koetsu, Aesthetix) and digital (Meridian 800) setup.

The second room was designed by Dennis Erskine and Keith Yates. There is extensive room treatment from RPG and custom bass absorbers built by Keith. The room was extensively modeled using fluid dynamics models which were then tested and confirmed in place via on-site measurements. Speakers were professional audio gear built into very specifically constructed baffle walls. There are five 18" subs in specific locations within the room. DSP processing is provided by Accourate and Trinnov. There is also a QSC unit that does processing below 100hz for the subs only. This is an all digital setup.

Do you have any before and after measurements of these room treatments you can show?

Bjorn · Feb 11, 2024

Atmasphere said:
I googled 'Distributed Bass array waterfall' and found this thread with measurements without digging. Perhaps you could show how Dr Toole is wrong. That should be easy, right?

No offense intended, but if its between Dr Toole or yourself (and given my own experience in this matter) I'll go with Dr. Toole.

Someone already pointed this out but for further information this type of bass array you linked to, which is called SBA (single bass array), does use acoustical treatment. The rear wall is basically completely absorped with treatment. The type of treatment can be of different type, depending on budget, but need to be sufficently broadband to work. When done correctly, it works a lot better than multple distributed subwoofers or DBA (double bass array) because the treatment will be effective above the sub frequencies as well.

FIY: You are using the word reverberant incorrectly. If it was truly reverberant in the correct sense, there wouldn't be standing waves! See below for more information.

Lyd & Akustikk | Små rom og etterklangstid

www.lydogakustikk.no

Bjorn · Feb 11, 2024

tony22 said:
Bjorn, that’s very impressive indeed. Would you consider (perhaps in a separate thread) going through the details of this before and after? Size of the room, what it looked like at the start, which treatments impacted each part of the observed improvement, etc? I for one would love to see that.

Thanks. This was also before the ceiling was treated. It was further improved with ceiling treatment, but unfortunately I don't have the final measurement at hand now.

I work with this comercially, thus I'm afraid I can't disclose in detail what I have done here. That's only shared with the customers. It was at a low cost in this case.

I'm not opposed to DSP at any way. But I don't believe in using DSP "as room correction" for what's not minimum phase behaviour. This will lead to phase distortion, something we are particular sensitive to above a certain frequency. That's why many DSP systems sound so unatural or plain weird.

But DSP for active speakers used correctly and to integrate subwoofers is IMO what high fidelity or high-end is truly about. Passive crossovers have limitations, and physical time alignment always bring in other compromises. Problem is that many have heard poor use of DSP and have made conclusions based on this.

Brucemck2 · Feb 11, 2024

Bjorn said:
The is wrong. Treatment can fix cancellations. I have already shown an example of that. If there's a cancellation somewhere, there's also pressure points with high level and resonances at the same frequency. When this is treated effectively, the cancellation will either be minimized or completely removed is the treatment is sufficiently effective.

Here's another example with a cancellation higher in frequency, at 120 Hz:
View attachment 125142

Two Broadsorbor panels placed correctly did this:
View attachment 125143

Using four Broadsorbor panels improved it further and the cancellation was almost gone:
View attachment 125144

Or here's before and after adding two Modex Edge bass traps where a cancellation between 50 H and 60 Hz is almost eliminated.
View attachment 125145

RPG’s Modex products are very effective. I’ve got six of their plates (three on the ceiling and three in one corner). The form factor works well.

Bjorn · Feb 12, 2024

There's a lot of treatment out there though that isn't very effective. And optimal placement is very important.

Below is another measurement that shows that a cancellation is effected. This is with only one single Modex Plate type 2 (35-100 Hz).

Green is without and red is with the single Modex Plate 2 (only sold in Europe). High resolution on the graph here. Notice espeically the area around 46 Hz.

While the overall response here isn't great, it does show that a low frequency cancellation can be treated and in this case with only one bass trap.

pjwd · Feb 12, 2024

Bjorn said:
There's a lot of treatment out there though that isn't very effective. And optimal placement is very important.

Below is another measurement that shows that a cancellation is effected. This is with only one single Modex Plate type 2 (35-100 Hz).

Green is without and red is with the single Modex Plate 2 (only sold in Europe). High resolution on the graph here. Notice espeically the area around 46 Hz.
View attachment 125204

While the overall response here isn't great, it does show that a low frequency cancellation can be treated and in this case with only one bass trap.

Bjorn
Are these measurements an average over a listening window are just a single spot .. that seems a lot fir a single panel
Phil

Bjorn · Feb 12, 2024

pjwd said:
Bjorn
Are these measurements an average over a listening window are just a single spot .. that seems a lot fir a single panel
Phil

At a single spot, at ear height. But the change isn't always as much as the latter one. It can vary.

Below is the before and after waterfall at ear height from the same.

Before:

After - one Modex Plate type 2:

Waterfall with one Modex plate type 2.jpg

tony22 · Feb 12, 2024

Bjorn said:
There's a lot of treatment out there though that isn't very effective. And optimal placement is very important.

Below is another measurement that shows that a cancellation is effected. This is with only one single Modex Plate type 2 (35-100 Hz).

Green is without and red is with the single Modex Plate 2 (only sold in Europe). High resolution on the graph here. Notice espeically the area around 46 Hz.
View attachment 125204

While the overall response here isn't great, it does show that a low frequency cancellation can be treated and in this case with only one bass trap.

Bjorn, if you can answer in a very simplistic sense, is the approach addressing a bass null like this to find the point in the room with the corresponding multiple of that null frequency, where there is excess energy, and use the right trap in that location?

Hear Here · Feb 12, 2024

sigbergaudio said:
A more careful approach is also possible. Using manual EQ to add 1-3 filters below 100-150hz goes a long way of evening out the response in most rooms, you don't have to employ an automatic calibration system like the ones you mention.

I'm sure that's true, but my reluctance to use DSP room correction, even for small adjustments as you suggest, is because (with a full-range amp system), the ENTIRE frequency range has to pass the DSP - there is no "bypass for higher frequencies" - and it's a signal processor that I don't want any more than other signal processing such as tone controls. I prefer not to process my incoming signal any more than is absolutely necessary.

Kingrex · Feb 12, 2024

I am not pro or con DSP. I don't know what it's impact really is. As in, can I hear it. I definitely heard rhe DEQX. Very audible digital sound. But I'm digressig from my point. I made a room change. I closed a wall with a barn door. It was a punch in the face reality for me how important room treatments are. I had a large amount of first reflection issues. A absorber cleaned a lot up. My point being, I always resisted DSP to some extent as I figured you have to deal with the room first. Otherwise your fighting obstacles DSP can not overcome. It may become frustratingly difficult to use DSP if your constantly chasing structural room issues that need pysical room correction. Issues such as reflactionsand echo. If the room is as ideal as possible, then maybe DSP can do fine tuning.

Does DSP belong in State of the Art Systems?

Well-Known Member

Well-Known Member

Industry Expert

VIP/Donor

Well-Known Member

Well-Known Member

Well-Known Member

Industry Expert

Well-Known Member

VIP/Donor

VIP/Donor

VIP/Donor

VIP/Donor

Member Sponsor

VIP/Donor

Well-Known Member

VIP/Donor

Well-Known Member

Well-Known Member

Well-Known Member

Similar threads