How do you judge your system's neutrality?

Good link Shadorne, thanks.

Almarg wrote:

Since the source material will essentially always deviate to some degree and in some manner from being precisely accurate relative to the original event, then it can be expected that some deviation from accuracy in the system may in many cases be complementary to the inaccuracies of the recording (at least subjectively), resulting in a greater transparency into the music than a more precisely accurate system would provide.

Al - This is a fascinating idea. If I understand you correctly, you are saying two things:

(i) The target of the concept of ‘accuracy’ is the RECORDING, whereas the target of the concept of ‘transparency’ is the MUSICAL EVENT that the recording represents.

(ii) In some cases, sacrificing some accuracy (to the recording) may increase transparency (to the musical event).

Regarding (i), I am in complete agreement. I think your observation about the differences between the respective targets of the concepts of ‘accuracy’ and ‘transparency’ captures both the usage of the terms among audiophiles as well as the underlying metaphors that those terms invoke.

Regarding (ii), I am in agreement, but in a more tentative way. I have actually been giving this topic some thought over the last few days, in the context of mulling over Cbw’s Rube Goldberg machine, viz., an audio system that exaggerates contrasts. Cbw raised it as a possible challenge to my operationalization of ‘neutrality’ presented in the original post, but I’ve been thinking more about whether an audio system that exaggerates certain contrasts, while not being truthful to the recording, might be more truthful to the musical event that the recording represents. Or, to use your observation, the question is whether an audio system that exaggerates certain contrasts might sacrifice some accuracy for the sake of greater transparency.

The sonic characteristic I’ve been thinking about in particular is dynamics. In light of the fact that so much compression is used in music recording, a playback system that exaggerated dynamics swings, while not being strictly truthful to the recording, might be more truthful to the musical event that was recorded (prior to the use of compression). That idea makes complete sense to me.

Yet my agreement with (ii) above is tentative, for two reasons. The first is that there are other sonic characteristics, like timbre, that do not seem like they would benefit from exaggeration. The second is that my technical understanding of how exaggerations in contrast might be achieved is limited. For example, can dynamics really be exaggerated above those of the recording by the kind of equipment available to the consumer? If so, how does the exaggeration of dynamics affect, for example, transient information? If you have other ideas and/or experiences of how sacrificing accuracy can sometimes increase transparency, I would love to hear about them.

Muralman wrote:

How, with all the layers of components can one know if one component addition or change makes a difference on it's own or is it a lost in complex relationships with the other components.

I said a few things about this subject in an earlier post on this thread. Here is what I wrote:

It is, of course, impossible to hear a component individually. We can only hear it in the context of a system. Because of this inescapable fact, there is always a potential fallacy when we hear a characteristic of an audio system and then attribute that characteristic to an individual component. A system might sound bright. Is it the speakers? Is it the cd player? Is it an impedance matching issue? If we get this wrong, we’ve made the Fallacy of Division, i.e., the misattribution of a system characteristic to one of the system’s components.

Muralman – I quite agree with you that audio systems are HOLISTIC, in the sense that the system-level characteristics are a result of the complex interaction of all of the system’s components. But it seems to me that our understanding of audio systems is almost always MECHANISTIC, in the sense that we try to reduce system-level characteristics to component-level characteristics. That is, I believe, more of an artifact of our minds than of the audio systems we are trying to understand. Hence I also wrote:

… the attempt to reduce system characteristics to component characteristics is unavoidable. It is fallible. But it is what we have.

Muralman - I believe you are saying that, the more complex a system gets, the more difficult it will be to attribute system-level characteristics to individual components. I agree with this. I do not completely agree, however, with your conclusion:

It has been my experience less is more. My wires can't be more simple. The DAC lacks a filter chip. The preamp is spartan. Every little change proclaims itself loudly, training me to go simple.

I don’t have a problem with this idea, in theory. In practice, it has not always been my experience. I can give two examples from assembling my system.

Example #1: The use of Room EQ Wizard software in combination with Meridian Room Correction made a dramatic difference in the quality of bass I was able to achieve in my listening room. This is reflected in a much flatter in-room frequency response below 200Hz, as well as the subjective impression, confirmed by other listeners, that my EQ work significantly improved the timing and transparency of my system. Meridian Room Correction involves intensive real time computation. It is not simple.

Example #2: The addition of a reclocker between my transport and my dac improved my system in at least four areas: (1) increased perceived resolution; (2) better imaging focus; (3) less shrillness in high frequencies; and (4) lower noise floor. The reclocker discards the timing data of a digital audio stream and reclocks the audio data using a high precision clock, thereby reducing jitter. Also, not simple.

Muralman - You mentioned my system in your post, so I assume your comments about the drawbacks of system complexity were triggered by the relative complexity of my system. While I agree that complexity provides more opportunities to get things wrong, I think in some cases, it provides opportunities to correct things that are already going wrong. Having said that, it is probably significant that the two examples I mentioned are both in the digital domain. If you look at my system from the point at which it becomes analog, you will notice that it is quite simple: The preamp is in the same unit as the dac (the Meridian), followed by 1 meter of analog interconnects, followed by a Pass amp of very simple design, followed by 2 meters of speaker cable to the speakers. This reflects my partial agreement with you about the value of simplicity.

(i) The target of the concept of ‘accuracy’ is the RECORDING, whereas the target of the concept of ‘transparency’ is the MUSICAL EVENT that the recording represents.

(ii) In some cases, sacrificing some accuracy (to the recording) may increase transparency (to the musical event).

Bryon, yes that is an excellent restatement of what I was trying to express.

I'll try to present further thoughts tomorrow on your reservations concerning the second point. But my quick initial thought is that the inaccuracies I was referring to would not necessarily (or even typically) be "exaggerations." They would be either complementary colorations (a simple illustrative example being a frequency response inaccuracy in the system offsetting a complementary inaccuracy in the recording), or they would be inaccuracies in the system which smooth over, obscure, or homogenize inaccuracies in the recording.

Best regards,
-- Al

Wow. Fascinating posts, Bryon. However, ultimately I remain unconvinced by your argument, though it is an impressive statement of your case. Just to take a couple of for instances, first the EQ issue. Surely if a system is very "neutral," "transparent," or "accurate," it would not need any EQ? It seems to me that by applying EQ, you are not faithful or truthful either to the recording or the musical event that it represents. Again, IMO you are merely changing the sound of your system (and the recording, of course) so that it represents your personal sonic priorities. There is also the "flat frequency response" thing we have already discussed in this thread - no concert hall has a flat frequency response - this simply doesn't sound good for the live event, so to try to create it in your system seems pointless (and usually results in a very lifeless/soulless sound, IME). The room correction issue is interesting, too - again, what is the standard that you are trying to correct the room to? I don't think any two audiophiles would perfectly agree on this. To be clear, I am not suggesting that room correction is worthless, indeed these systems can make a big difference; I just wonder - how do you know when it has been corrected? Again, only you can answer that for yourself, and your answer may be very different from any other given audiophile's. The "lower noise floor" is also a topic of much division among audiophiles - take my brother's Nottingham turntable, a line which is well known for it's "black background." To my ears, the sacrifice made here to attain this is the removal of too much of what some call "low level information," for example a loss of much of the sense of the sound of the actual musical event/space that was recorded. Live music does not exist in a vacuum, even in a very dead recording studio - the sound of the space is an integral part of the music. This is particularly noticeable in moments of silence within the music. As John Cage famously demonstrated with his piece 4'30", silence is never really silence in a concert hall.

Sorry about the rambling - I'm under the weather, and never was the clearest writer - your posts are much better than mine in that regard! I just wanted to share the thoughts that came to me as I read your interesting posts. I would be particularly interested in your answer to my first question about the EQ issue, as it seems to me that if true "neutrality", "accuracy", and "transparency" could exist in a system, EQ would be completely unnecessary, and I was very surprised to see you mention it in this context.

Learsfool wrote:

…it seems to me that if true "neutrality", "accuracy", and "transparency" could exist in a system, EQ would be completely unnecessary, and I was very surprised to see you mention it in this context.

Learsfool – I did not mention EQ in the context of our ongoing discussion about neutrality, accuracy, or transparency. If you look again at my post, you will see that my mention of EQ was in the context of a reply to Muralman’s suggestion that it is difficult to identify the contribution of each component in an audio system, and that systems that are complex suffer from this more than ones that are simple. Nowhere in Muralman’s posts, or in my reply to him, is there mention of neutrality, accuracy, or transparency. Muralman’s question was not about neutrality, accuracy, or transparency. But I found his question valid and interesting in its own right, and so I took the time to answer it as best I could.

It is worth mentioning that I never suggested that my own system is especially resolving, neutral, accurate, or transparent. I value each of those characteristics, and I would like to believe that I have achieved some measure of each of them. But I have left the specifics of my own system out of the discussion up until my last post, because I didn’t want the conversation to become a comparison of different equipment, which there is an abundance of on Audiogon. I broke my silence about the specifics of my own system only because Muralman brought up my system in his post as an illustration of his concern about complexity. It was in an attempt to answer his concern that I described some of the particulars of my system. But I do not regard my system as a model of any particular sonic characteristic. I mention this because you wrote the following:

Surely if a system is very "neutral," "transparent," or "accurate," it would not need any EQ?

This makes it sound as though I have characterized my system as “very ‘neutral,’ ‘transparent,’ or ‘accurate.’” I have not.

This brings me to your comment about the incompatibility of (1) valuing system neutrality, accuracy, or transparency; and (2) the use of EQ in the context of room correction. You wrote:

It seems to me that by applying EQ, you are not faithful or truthful either to the recording or the musical event that it represents.

I disagree with this. The use of EQ for room correction is usually a matter of using notch filters to suppress room modes. As you no doubt know, room modes are frequencies that result in standing waves, the volume of which can easily be exaggerated by 10dB or more. If left uncorrected, standing waves make music sound bloated, uneven, and slow. Standing waves can be corrected either through room treatments or through EQ. For those who do not have a dedicated room (like me), the use of EQ for room correction can be a very effective option for controlling the destructive effects of room modes.

The use of EQ, in the way just described, is not a move away from system neutrality, accuracy, or transparency. In fact, it is the opposite, as anyone who has heard the effects can testify to. The result of suppressing the room modes in my system made the system more neutral, more accurate, and more transparent. The measure of this is not merely the flatter frequency response achieved under 200Hz. The measure of this is the perception of the listener.

Your comment that “Surely if a system is very ‘neutral,’ ‘transparent,’ or ‘accurate,’ it would not need any EQ?” seems to reveal an assumption that:

THE SYSTEM = THE EQUIPMENT

If we substitute “the equipment” for “system” in your question, then it would read: “Surely if the equipment is very ‘neutral,’ ‘transparent,’ or ‘accurate,’ it would not need any EQ?” I would agree with this rhetorical question IF I believed that the system = the equipment. But I believe that:

THE SYSTEM = THE EQUIPMENT + THE ROOM

That is why, in a previous post, I wrote:

…it is useful to conceptualize accuracy in terms of information, specifically, the information available at the source vs. the information presented at the ear. (I say “at the ear,” rather than “at the speaker,” since the room, and your listening position in it, are ultimately part of the system).

The alteration of the signal for the purposes of room correction seems like a deviation from system neutrality, transparency, and accuracy ONLY IF you believe that the system = the equipment. If the room is part of the audio system, then changes to the signal in the equipment do not necessarily result in less neutrality, transparency, or accuracy AT THE EAR. And that is where it counts.