• 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!

MQA, DSP and "sound quality"

svart-hvitt

Major Contributor
Joined
Aug 31, 2017
Messages
2,375
Likes
1,253
Olive was comparing commercial room EQ devices. The "best one" won, but it may or may not be the best possible. As I think is well explained in my book, if you begin with a well designed loudspeaker you will end up with a preferred room curve. But, as Cosmic said, if you equalize a Brand X loudspeaker to match that curve there is no assurance that it will sound like a winner. Good sound begins at the source - either a superb musical instrument well played or a neutral loudspeaker reproducing a good recording of that instrument.

Still, doesn’t Olive suggest that a specific speaker can be enhanced by DSP based room compensation?

So, according to Olive, if you have a good speaker, like BW802n, you should invest in room compensation for frequencies up to at least 500 Hz?

In other words, a good speaker is, good, while a good speaker plus good DSP based room compensation is better (still according to Olive)?
 

Floyd Toole

Senior Member
Audio Luminary
Technical Expert
Industry Insider
Forum Donor
Joined
Mar 12, 2018
Messages
367
Likes
3,909
Location
Ottawa,Canada
Still, doesn’t Olive suggest that a specific speaker can be enhanced by DSP based room compensation?

So, according to Olive, if you have a good speaker, like BW802n, you should invest in room compensation for frequencies up to at least 500 Hz?

In other words, a good speaker is, good, while a good speaker plus good DSP based room compensation is better (still according to Olive)?


For any loudspeaker it is highly probable that equalization below about 500 Hz will be beneficial - that is where the room is the dominant factor. However, equalization alone is beneficial only at a single seat. To share the benefits with other listeners numerous large bass traps or multiple subs are necessary. Above that frequency, equalization can benefit some loudspeakers more than others. The 802N has a directivity problem that EQ cannot directly address. A loudspeaker with relatively constant directivity but poor frequency response can be improved, but they are rare.

BTW, Olive and I don't disagree about this. I am still on consulting retainer to Harman, and we see each other frequently. Besides, he also likes good food and wine :)
 

Scott Borduin

Member
Forum Donor
Joined
Jan 16, 2018
Messages
55
Likes
136
Location
Portland OR USA
For any loudspeaker it is highly probable that equalization below about 500 Hz will be beneficial - that is where the room is the dominant factor. However, equalization alone is beneficial only at a single seat. To share the benefits with other listeners numerous large bass traps or multiple subs are necessary. Above that frequency, equalization can benefit some loudspeakers more than others. The 802N has a directivity problem that EQ cannot directly address. A loudspeaker with relatively constant directivity but poor frequency response can be improved, but they are rare.

BTW, Olive and I don't disagree about this. I am still on consulting retainer to Harman, and we see each other frequently. Besides, he also likes good food and wine :)

A question about the room affected and transition regions: when trying to correct, should one aggressively EQ up to the chosen frequency, e.g. 500 Hz, or gradually reduce the correction as one approaches that frequency? As I interpret the documentation for the Trinnov/JBL SDP-75, for instance, it seems that those units might use the latter approach?

Scott
 

Floyd Toole

Senior Member
Audio Luminary
Technical Expert
Industry Insider
Forum Donor
Joined
Mar 12, 2018
Messages
367
Likes
3,909
Location
Ottawa,Canada
A question about the room affected and transition regions: when trying to correct, should one aggressively EQ up to the chosen frequency, e.g. 500 Hz, or gradually reduce the correction as one approaches that frequency? As I interpret the documentation for the Trinnov/JBL SDP-75, for instance, it seems that those units might use the latter approach?

Scott
The most prudent approach is to employ some spatial averaging in the "adjacent boundary" region, and use wider bandwidth (lower Q) filters. This is explained in the relevant chapter in my book. Only in the room mode region should one design matching-Q filters to attenuate single resonances. One can equalize for a single seat or, with multiple subs or massive bass traps, at one or more seats. Obviously, never try to fill narrow dips - they are destructive interference dips.

So, you are right, reduce the resolution, and thereby the amount, of equalization at higher frequencies.
 

Fitzcaraldo215

Major Contributor
Joined
Mar 4, 2016
Messages
1,440
Likes
634
Dr. Toole - thank you for the excellent summary, insights into and amplification of your previous published remarks. It all makes perfect sense to me, and it is as definitive as it gets.

EQ in the bass typically results, as you say, in a huge payoff, and it seems to work noticeably "better" than no EQ even in systems constrained to a single subwoofer or no subwoofers. But, there is no question in my mind that multiple subs along with EQ are potentially better still.

Full range EQ is much more controversial, as you have highlighted. I do not question that it is, as you say, not necessary above about 500 Hz with the right, properly engineered speakers in a typical room. I agree, full range EQ is no panacea. Still, with less than ideal speakers, there may be some upside in applying it as a "band aid" in some cases, in my experience. That might have been the case to some extent in Olive's testing using the B&Ws. It may not achieve the higher perfection that better speakers might in a better room. But, I admit to being a pragmatist in my own imperfect system. I might be willing to accept a perceived improvement to a given system and room, even if not to the ultimate.

I do use the comparatively inexpensive "upgrade" of full range DSP EQ with very satisfactory results, a net improvement in my subjective opinion. Used properly and with care, the balance of upsides vs. downsides may favor its use from a personal preference standpoint in many cases, again in my experience. So, it might, as I see it, be worthy of some experimentation by the listener in some circumstances.
 

svart-hvitt

Major Contributor
Joined
Aug 31, 2017
Messages
2,375
Likes
1,253
Dr. Toole - thank you for the excellent summary, insights into and amplification of your previous published remarks. It all makes perfect sense to me, and it is as definitive as it gets.

EQ in the bass typically results, as you say, in a huge payoff, and it seems to work noticeably "better" than no EQ even in systems constrained to a single subwoofer or no subwoofers. But, there is no question in my mind that multiple subs along with EQ are potentially better still.

Full range EQ is much more controversial, as you have highlighted. I do not question that it is, as you say, not necessary above about 500 Hz with the right, properly engineered speakers in a typical room. I agree, full range EQ is no panacea. Still, with less than ideal speakers, there may be some upside in applying it as a "band aid" in some cases, in my experience. That might have been the case to some extent in Olive's testing using the B&Ws. It may not achieve the higher perfection that better speakers might in a better room. But, I admit to being a pragmatist in my own imperfect system. I might be willing to accept a perceived improvement to a given system and room, even if not to the ultimate.

I do use the comparatively inexpensive "upgrade" of full range DSP EQ with very satisfactory results, a net improvement in my subjective opinion. Used properly and with care, the balance of upsides vs. downsides may favor its use from a personal preference standpoint in many cases, again in my experience. So, it might, as I see it, be worthy of some experimentation by the listener in some circumstances.

I like this and @Floyd Toole ’s comments on DSP because they add colour to our previous discussion on objectivism and subjectivism.

Placing of speakers in room, implementing acoustical measures in room, and finally DSP for room compensation - at the end we use our ears in addition to measurements and rules of thumb given to us by empirical testing, psychoacoustical research.
 

JohnPM

Senior Member
Technical Expert
Joined
Apr 9, 2018
Messages
344
Likes
920
Location
UK
Seems MQA and DSP are not mutually exclusive. From the Roon 1.5 release notes: We have also covered some new ground with MQA: Roon will not only be able to perform DSP on unfolded MQA content, it will do so without destroying the MQA Signalling information–meaning: it is now possible to use features like EQ, Room Correction, and Volume Leveling while still taking advantage of the rendering capabilities of your MQA DAC.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,720
Likes
241,542
Location
Seattle Area
Seems MQA and DSP are not mutually exclusive. From the Roon 1.5 release notes: We have also covered some new ground with MQA: Roon will not only be able to perform DSP on unfolded MQA content, it will do so without destroying the MQA Signalling information–meaning: it is now possible to use features like EQ, Room Correction, and Volume Leveling while still taking advantage of the rendering capabilities of your MQA DAC.
Wow, that is a pretty significant confirmation of being able to apply signal processing. Given the history of Roon folks and Meridian, if anyone could break this ordeal it would have been them.
 

Werner

Active Member
Joined
Mar 2, 2016
Messages
109
Likes
135
Location
Europe
What wow about that?

I has been abundantly clear for a long long time that MQA supports DSP, when that DSP is encapsulated in a full MQA system.

Meaning that the DSP vendor needs an MQA license, and the DAC vendor as well.

Still does not help the huge installed base of DSP systems out there ...
 
OP
pirad

pirad

Active Member
Joined
Apr 22, 2018
Messages
178
Likes
61
It has been possible from the beginning to apply DRC to the first unfold of MQA. Eg. if you have both Tidal app and Dirac on the same PC they work together. 96/24 wasn’t really proven to be inferior to higher rate codecs in listening tests.
Whether the controversial “deblurring” is taking place here seems to be secondary.
 
Last edited:

NorthSky

Major Contributor
Joined
Feb 28, 2016
Messages
4,998
Likes
946
Location
Canada West Coast/Vancouver Island/Victoria area

Audio Flynn

Member
Joined
Oct 31, 2019
Messages
16
Likes
11
check out NAD M32.
48 hours into Tidal service though my M32 I bought one month ago. It replaced my Conrad Johnson tube separates.

MQA of recordings I have listened to for over 40 years is quite exceptional. First amplifier change in 13 years for my Klipsch Belles has been a successful life change.

All analog system will have its own room in mid 2021.
 

John Dyson

Active Member
Joined
Feb 18, 2020
Messages
172
Likes
90
MQA fixes or benefits very little on most recordings. Basically, if they use the same sliced and diced compressed mess that is available on most pop, even classical (just found some Telarc has the same problem) releases -- then MQA is meaningless... It is just more of the same garbaged up recordings.It is wonderful when one can find one of the precious few properly mastered, not chopped up recordings, but I have run into very few.

Until the listening public demands properly mastered materials, we'll keep on spending on $10k systems, and have no good material to listen to (unless direct from the engineer/non commodity high quality recordings.)

I know that this sounds like heresy -- but it is true, the damaged recordings have been sold since CDs came out. Since I used to do recordings myself, I knew what a recording really sounds like and then walked away from the HiFi hobby back in approx 1990. The CDs weren't getting any better, even though their theoretical capability was great.

MQA does NOT fix the problem, unless the recording is mastered correctly first (that means, no excess compression), then MQA encoded. This compression technique is very stealthy, and is less effective at compression than it is scrambling the signal. It is a layered/EQed compression scheme, approx 10dB at a time (with a carefully speced compressor capable of only 10dB of compression.) The scheme IS reversible, and it makes me wonder if there wasn't an original reason for the scheme -- you can get varying quality levels from the already existent scheme -- SIMILAR TO MQA. The difference is that there is no encryption with the current (since middle 1980s') analog based scheme, it is just damned detailed.

John
 

DChenery

Active Member
Forum Donor
Joined
Feb 6, 2020
Messages
205
Likes
453
Location
Nanoose Bay
Wow, that is a pretty significant confirmation of being able to apply signal processing. Given the history of Roon folks and Meridian, if anyone could break this ordeal it would have been them.

While it may be awesome programming, I think it is really "Polishing a Turd" as MQA still seems to be more "rent seeking" than value added.

 
Top Bottom