Does DSP belong in State of the Art Systems?
- Thread starter musicfirst1
- Start date
Ah, see the missspelllinging now 
guess our eyes just see W and M as 180 degrees out of phase
guess our eyes just see W and M as 180 degrees out of phase 
The mini DSP I don't know, but recently received an Accuphase DG-68 DAC/DSP/EQ, I'm using it as both a DAC (compared to my DCS Paganini, plays the same, just a different signature) and as a DSP, I'm really enjoying the results of the DSP in read and autocorrect mode from the room!
When I listen to analogue, the DG-68 is out of the system!
I advise you to experiment with a DSP!
When I listen to analogue, the DG-68 is out of the system!
I advise you to experiment with a DSP!
Last edited:
I've ordered a MiniDSP 2x4hd to play with. I'm going to play with only the crossover fuction as well as the timing/delay functions.
It will sit between my DAC/PRE and one pair of RCAs going to my mains and the other pair going to my subs.
I'm hoping/expecting better integration between my subs and mains, along with a audible insertion loss of the unit itself.
Has anyone gone down this road? Are there devices on the market that minimize the negative impact of this pursuit?
I posted this in the general forum because I know there are people who would never read the DSP forum...
If DSP is part of the design, like Meridian, yes.
If you have a shocking room in bass response - either too high or too low, yes
If one is to insert eq, the best way imo is to have a preamp that has a tape loop. Put the eq into the tape loop and with the press of a button you have eq in your system or completely out of it.
As someone who has extensively used DSP, I certainly appreciate what it can do to eliminate room modes and integrate subs seamlessly. But it’s worth noting things that DSP can’t fix. One such issue is loudspeaker coloration. For example, Harbeth has essentially staked its entire reputation on designing very low coloration midrange drivers that it calls RADIAL, which came out of an extensive research program with a UK university. Many cone materials were studied and a significant effort was put into understanding the coloration caused by different cone materials (polypropylene, metallic cones, Kevlar etc.). They eventually designed a new proprietary cone material that in their view had the ideal characteristics for which they invented a particular type of measuring device for measuring strength of materials (like Young’s modulus). Certainly, whether you buy into their rationale or not, Harbeth’s have a very distinctive sound (or as Alan Shaw would put it, lack of coloration). It’s almost like the speakers have no highs, which of course they do when you measure them. I think they demonstrated that a lot of the traditional brightness you hear with dynamic moving coils comes down to choice of cone materials.
The point here is that cone coloration is not a problem you can fix with DSP. There are many such issues (e.g., distortion, energy storage etc.). DSP can correct frequency imbalances at a location, yes, and do that very well. But all the other issues don’t go away.
The point here is that cone coloration is not a problem you can fix with DSP. There are many such issues (e.g., distortion, energy storage etc.). DSP can correct frequency imbalances at a location, yes, and do that very well. But all the other issues don’t go away.
Its interesting to see so many people opposed to the use of DSP/Room Correction but I'm not surprised of their disappointment given some of the mentioned examples of having only tried old hardware based implementations to come to that conclusion.
I don't recall seeing anyone mention that they have actually tried to use a modern form of properly sorted DSP/Room Correction software (Audiolense, Accurate) and after having spent the time to create a good set of filters for their room/gear they were still disappointed.
I wonder how many people opposed to the use of DSP/Room Correction are willing to post up their current room curves to give the fans of DSP here an idea of what your working with and how it might be so much better without the need for DSP/Room Correction? Maybe there are some rooms/systems that are that good that its not needed, but I have my doubts.
Could it be that a posted less than ideal room curve, without DSP/Room Correction, is contributing, at all, to colorations which the listener finds appealing leading them to the claim that "this" is how the original source is supposed to sound and that the filthy DSP stuff was just coloring the sound of my system?
Could it be that they unknowingly implemented DSP/Room Correction incorrectly when they tried it and now are of the belief that it sucks across the board?
Could it be that their gear lacks enough gain to make up for any headroom losses imposed by the use of DSP/Room Correction software and what they hear now, afterwards, is a muted version of what they used to play at -20dbfs but now requires -10dbfs but the gear is lacking enough gain to make up for it?
I don't recall seeing anyone mention that they have actually tried to use a modern form of properly sorted DSP/Room Correction software (Audiolense, Accurate) and after having spent the time to create a good set of filters for their room/gear they were still disappointed.
I wonder how many people opposed to the use of DSP/Room Correction are willing to post up their current room curves to give the fans of DSP here an idea of what your working with and how it might be so much better without the need for DSP/Room Correction? Maybe there are some rooms/systems that are that good that its not needed, but I have my doubts.
Could it be that a posted less than ideal room curve, without DSP/Room Correction, is contributing, at all, to colorations which the listener finds appealing leading them to the claim that "this" is how the original source is supposed to sound and that the filthy DSP stuff was just coloring the sound of my system?
Could it be that they unknowingly implemented DSP/Room Correction incorrectly when they tried it and now are of the belief that it sucks across the board?
Could it be that their gear lacks enough gain to make up for any headroom losses imposed by the use of DSP/Room Correction software and what they hear now, afterwards, is a muted version of what they used to play at -20dbfs but now requires -10dbfs but the gear is lacking enough gain to make up for it?
Bjorn, you are right, lots of things use DSP, but for audio I think most people associate it with DAC and EQ/X-over/DRC.
It´s found in many places in the recording chain, so I find it a bit strange, that some thinks its a bad thing to use.
But please give us your thoughts on DSP, instead of just poking around
It´s found in many places in the recording chain, so I find it a bit strange, that some thinks its a bad thing to use.
But please give us your thoughts on DSP, instead of just poking around
Last edited:
I am not anti-RC (room correction DSP), if it is applied at the right time and to the right frequencies.
The trouble with most RC systems is that they are built into full-range amps and this means that the entire frequency range has to pass through the DSP circuit, even though in many cases, only the bass range is actually adjusted.
Modern RC is incredibly complex, not just adjusting the relative levels of the individual frequencies, but time alignment and goodness knows what else. As only the bass (generally speaking) is adversely affected by the room's acoustic, there's no need for the delicate top end to be molested by this intensive signal processing, and it's this factor that those of us against this method of RC have concerns about.
My own amp has Dirac Live (one of the most respected RC DSPs available) that only adjusts sub 500 Hz frequencies, yet it is easy to demonstrate that (with first class speakers) there is audible loss of ultimate detail in the top end as it is delivered by the speakers. It's this aggressive (and totally unnecessary) signal processing of the top end that causes this reduction of the “goosebump factor” we crave - so I don't engage the filter within my full-range amp.
Fortunately my own system has an alternative RC method that is applied only to the bass frequencies. My Avantgarde speakers (and for that matter the Martin Logan hybrid electrostatics I owned for a short time) have RC built into their bass-only amps within the speaker's enclosures. This allows adjustment of the bass only while the rest of the frequency range goes totally un-affected by any RC processing. This can only be done with a bass amp and a separate DSP-free top end amp. Many active systems could also use this method, though sadly not old-design active speakers such as ATC that have no DSP within their bass amps. Modern actives such as Dutch & Dutch and a few others do have built-in RC DSP and DACs so get around this problem.
I hope the above explains why I and others here don't think that RC DSP used in the way it most often is has a place in State of the Art Systems.
Great for multi-speaker AV systems where audio finesse is of less concern and also great for the DIYer who cannot be expected to build an accurate (flat response) speaker from scratch.
The question should have been the other way around:
Can a passive loudspeaker be high-end?
And the answer to that is clearly no, since active (which implies DSP) will upgrade any passive speaker. Just removing the passive crossover and replace by an active with the same settings, amps driving each driver, and time alignment of drivers alone will clearly sound better.
However, "room correction" is something that generally don't work and the word is also misleading since you can't EQ the room. The room response basically isn't minimum phase and any correction will only work in one spot. You can only correct the speakers, though limited, and tailor the speaker response to the room. Any correction of the room itself must involve physical room treatment.
Because so many auto correction systems also corrects the room response which isn't minimum phase, the result is always mixed. Something may improve (typically the lows) but other areas will suffer. The latter mainly because one is introducing phase distortion and much of that sounds very unatural. Acoustic treatment on the other hand actually makes the room response to become more minimum phase. Meaning the amplitude and phase correlates.
Any correction should really be applied manually by someone who understands what can be corrected and what should be left alone. But the best speaker is one with an active crossover, which is a form of DSP.
Can a passive loudspeaker be high-end?
And the answer to that is clearly no, since active (which implies DSP) will upgrade any passive speaker. Just removing the passive crossover and replace by an active with the same settings, amps driving each driver, and time alignment of drivers alone will clearly sound better.
However, "room correction" is something that generally don't work and the word is also misleading since you can't EQ the room. The room response basically isn't minimum phase and any correction will only work in one spot. You can only correct the speakers, though limited, and tailor the speaker response to the room. Any correction of the room itself must involve physical room treatment.
Because so many auto correction systems also corrects the room response which isn't minimum phase, the result is always mixed. Something may improve (typically the lows) but other areas will suffer. The latter mainly because one is introducing phase distortion and much of that sounds very unatural. Acoustic treatment on the other hand actually makes the room response to become more minimum phase. Meaning the amplitude and phase correlates.
Any correction should really be applied manually by someone who understands what can be corrected and what should be left alone. But the best speaker is one with an active crossover, which is a form of DSP.
While I understand there are issues with DSP, as I have said before I think DSP can be SOTA if done right. I have refrained from asking a different question so far but I just couldn’t resist.
Can a system be called state of the art if people don’t show their measured frequency response?
I have to admit, I’ve listened to many audiophile systems whose owners are super proud of. And inevitably there are significant bass frequency response issues that they have simply gotten used to. Since they never measured, their own system has become their gold standard and they would criticize other systems as sounding wrong. Now while I agree that different people have different preferences for different target frequency responses, I simply don’t think that any system with a significant +10dB peak or -10dB dip in the bass frequency response to be state of the art.
To be honest, I didn’t ask the question to stir the pot. I fully acknowledge that we can’t have perfect rooms so if people love vinyl, it’s often extremely difficult to set up their system to avoid significant bass peaks or troughs. And I don’t believe that most systems need to be perfectly DSP’d to the desired target frequency response to be SOTA. But for those who think all DSPs are bad, maybe you just happen to have great frequency response and no major room acoustic issues without DSP. But then maybe not. After all, we don’t know what we don’t know.
Can a system be called state of the art if people don’t show their measured frequency response?
I have to admit, I’ve listened to many audiophile systems whose owners are super proud of. And inevitably there are significant bass frequency response issues that they have simply gotten used to. Since they never measured, their own system has become their gold standard and they would criticize other systems as sounding wrong. Now while I agree that different people have different preferences for different target frequency responses, I simply don’t think that any system with a significant +10dB peak or -10dB dip in the bass frequency response to be state of the art.
To be honest, I didn’t ask the question to stir the pot. I fully acknowledge that we can’t have perfect rooms so if people love vinyl, it’s often extremely difficult to set up their system to avoid significant bass peaks or troughs. And I don’t believe that most systems need to be perfectly DSP’d to the desired target frequency response to be SOTA. But for those who think all DSPs are bad, maybe you just happen to have great frequency response and no major room acoustic issues without DSP. But then maybe not. After all, we don’t know what we don’t know.
....and their spatial characteristics and their distortion vs. frequency/power and etc............ I am not saying that getting all of this will be determinative but it is fundamentally informative.
While the frequency response is certainly very important, there's also a fact one can achieve a super even one today with correction and the sound can be really bad due to correction of non mimimum phase behaviour and little or no improvement of the time domain.
So it depends on how one has achieved the result. But obviously a stellar system will have a very even response, and done the right way.
So it depends on how one has achieved the result. But obviously a stellar system will have a very even response, and done the right way.
I
I suspect the more important factor is whether it SOUNDS state of the art. Do we take our measuring gear into concert halls? We take our ears there and hope and expect that they will be thrilled by the performance. That's really all that matters when all is said and done - the sound and its effect on us.
Last edited:
I personally don’t define “state of the art” by flat frequency response. I define state of the art by “equal to the highest suspension of disbelief, and the greatest emotional engagement, I have experienced from an audio system.”
For me the challenge with that statement is that if somebody has a distorted system that they tuned to their favorite music, they may say their system is state of the art for playing their favorite music because the system creates “highest suspension of disbelief, and the greatest emotional engagement” for them. But they have no reference for a different system or even the same system in a different room setup which may create an even higher suspension of disbelief and even greater emotional engagement. A great example is that I know people who are used to their bookshelf speakers. So whenever they hear systems that fill in the bass from say 20-50Hz, they really dislike it, regardless of how well the low bass and mid-bass is integrated into their system.
My question never states that people need to have flat frequency response even though I am aware that some strong proponents of audio DSP believes systems should be tuned to the target frequency response as much as possible. But I was implying/saying that if you have major bass frequency peaks and troughs of say >5dB (and definitely >10dB), I just don't think you can say the system is state of the art. Chances are, you're going to get better sound by addressing room acoustics or simply moving your listening seat (and the speakers) if you don't want to use DSP to address the issue (and accept the compromises that DSP brings, if any). The issue is that if you don't measure, you don't always know.
That's why going back to the main topic, my personal feeling is that:
DSP can belong to a state of the art system. Although I'm well aware that many DSP implementations are suboptimal, making some systems sounding better without DSP than with. And often even optimal DSP can improve many aspects but lead to compromises in other sonic aspects.
But I also think that if people are trying to setup a state of the art system, they should at some point measure the frequency response (and maybe additional aspects of their room/speaker) because I'm not convinced that people can optimally address their room acoustic issues, position their listening position and speakers without measurements.
If you go this road you should write it clearly - the sound and its effect on ME, not on us!
As I say, when you are the lawyer, the jury and the judge thinks are simple.
Nice sounding definition, but too subjective, time related, recording dependent and vague to be of any real use in audio debates. However, it can be effectively used in mine is better than yours typical discussions.
BTW, no serious audiophile should refer anymore to frequency response when addressing speakers - at minimum they should address the power response.
Similar threads
 | Steve Williams Site Founder | Site Owner | Administrator |  | Ron Resnick Site Owner | Administrator |  | Julian (The Fixer) Website Build | Marketing Managersing |