Enclosure resonances, not a big deal?

[gedlee]

Originally posted by planet10


I don't know about you, but i listen to music, and way down is where the subtle details of the music live, the subtle details that really help with verisimilitude. On the other side the removal of grunge that hides this stuff relieves our flight/fight reflexes from having to work as hard. We are more relaxed and enjoy the music more.

dave

Dave

I think this a mistatement of the issue. The issue is not listening to low level signals, say 50 dB down, its listening to aberations that are 50 dB BELOW the signal. Thats a vastly different thing. Its this masking that I contend makes the enclosure aspects inaudible despite the fact that you can do all kinds of things to make them audible, but only if you surpress the signal content!

[planet10]

Originally posted by dlr
Grunge isn't a very well defined term.

Grunge is anything that obscures/hides/masks information in the signal. Some is onvious, some (particularily for low level stuff) you don't even know is there until it is gone.

dave

[planet10]

Originally posted by gedlee
The issue is not listening to low level signals, say 50 dB down, its listening to aberations that are 50 dB BELOW the signal.

When i say 40 dB down (the figure i usually use) i mean downward dynamic range. Stuff that is that far down in the precense of 0 dB signal.

dave

[gedlee]

Originally posted by badman

Arguments of expertise are a fallacy. I have no 'credentials' per se (beyond some years of experience) but can assemble a decent argument on many of the points under discussion.

There are few things where credentials don't matter and experince is a credential. I would say that they are "necessary but not sufficient".

Originally posted by badman


Ignoring "Xpert" is the best route if you're bothered by his behavior.

Agreed, I will go back to doing that.

Originally posted by badman

Regarding level perceptibility, testing I've seen was done with non-music signals. That alone separates the test from the usage environment, like a wind-tunnel vs. a racetrack vs. a road with cars. I do a number of tests when testing small items. One of the most important is long-term behavioral. Do I still enjoy the item in question after the novelty has worn off? Do I find myself focusing on one aspect of the sound? Do I listen to more or less music?

Obviously it's not an easy test to perform with multiple subjects, and has many factors applied, but that's why it must be longer duration. If my longer-term musical satisfaction increases, more attention is spent on the whys and wherefores. If I find my rig less enjoyable, the change is reversed, or further investigated.

Blind testing is an important validation step... sometimes. Just as with the THD wars, the ears have it all over the microphones, as far as musical enjoyment goes.

I think that your procedure for listening evaluations is much like mine. Where we might disagree is on cause and effect. I never accept anything that I think that I can hear until I can measure it. That IMO is a necessary requirement to establishing cause and effect. I was never able to measure enclosure effects even though I too belived in them initially. Then when I reduced the enclosure structural aspects and the sound remained just as good, that kind of closed the loop for me. I do not say and have never said that ANY cabinet is fine, only that in extensive evaluations I find that my cabinet construction is sufficient. But its not up to many people standards - that, I claim, is unjustified.

[thoriated]

U may simply calculate it: a contrubution that is 20dB down and adds linear ....

....so on and so forth.

Yes, and what about a listening room aberration that happens to reduce the amplitude of the initial wavefront by, say, 15 db and another listening room aberration that emphasizes the contribution of the panel resonance by, say, 10 db at some hypothetical frequency? Suddenly, the delayed coloration is louder than the initial sound. It appears such considerations have so far been outside of your intellectual envelope.

[SY]

Deleted some OT posts. Keep it to the technical discussion, folks.

[badman]

Originally posted by gedlee

Dave

I think this a mistatement of the issue. The issue is not listening to low level signals, say 50 dB down, its listening to aberations that are 50 dB BELOW the signal. Thats a vastly different thing. Its this masking that I contend makes the enclosure aspects inaudible despite the fact that you can do all kinds of things to make them audible, but only if you surpress the signal content!

But, given a well-behaved time-domain from the drivers themselves, any enclosure ringing that remains when the excitation drops off suddenly will be dramatically higher in relative amplitude. It's not during the dynamic transients- it's immediately after, where the source material is relatively quiet and consists of resonance within the instruments and environment of the recording... resonances with (likely) similar characteristics to those of the enclosure, and with much closer amplitudes.

Originally posted by gedlee

There are few things where credentials don't matter and experince is a credential. I would say that they are "necessary but not sufficient".

<<Insert "We've discussed Xpert enough today" here>>

I think that your procedure for listening evaluations is much like mine. Where we might disagree is on cause and effect. I never accept anything that I think that I can hear until I can measure it. That IMO is a necessary requirement to establishing cause and effect. I was never able to measure enclosure effects even though I too belived in them initially. Then when I reduced the enclosure structural aspects and the sound remained just as good, that kind of closed the loop for me. I do not say and have never said that ANY cabinet is fine, only that in extensive evaluations I find that my cabinet construction is sufficient. But its not up to many people standards - that, I claim, is unjustified.

I think it's flipsides of the same coin. It's a subjective rather than an objective determination of which dataset is more valuable, data vs. hearing. When building something, it's very difficult to disassociate your prejudices from the evaluation. We're agreed that a 4" thick steel enclosure isn't necessary. But, an easily implemented set of shelf braces sure can go a long way.

But it's possible that you determined via (known flawed, there's no such thing as a 'perfect' measurement) measurements that the enclosure resonances were below an audibility threshold, and that influenced your listening to be willing to ignore them. This is just a flipside of the coin of hearing what you want.

I think that much of this hobby is in the finer points, and sometimes they're obscured via short-term measurements. I can't remember an album well after the first listen, or even (usually, if it's music that I deemed worth evaluating) give a good idea of whether I'll enjoy it longer-term.

Human auditory memory is known to be short.....

But that's based upon short-term testing! Perhaps auditory memory is longer with more time spent with a driver/speaker/system? I know I recognize the 'sound' of speakers and drivers I've long not-heard, if I spent enough time with them. Likewise, there are subtle bits of music that take time to pick up on, and certainly wouldn't happen in a DBT with unfamiliar everything, but once they're recognized, these (typically low-level) information pieces can be critical, and one can be aware that they're altered, even though they're very low level and the alterations even lower.

Maybe I'm just slow, but rapid fire evaluation is extremely difficult for me, but I am pretty confident in longer terms. When doing switch-based tests, I meditate before hand, to clear my mind and relax. I have to be very careful to not 'actively' listen as much as 'hear', for the low-level stuff to be clear. The correlation of these shorter-term tests to longer-term experience has only been useful when extreme care to be passive and comfortable and clear-headed are taken. However, the combination of the two gives a great deal of confidence in the results. If I listen to a specific device, swap it in and out a few times over a month or three, then do A/B/A (sometimes DBT, sometimes not) testings in the right fashion, the character I've come to know is much more apparent.

Enclosure resonances are tough to eval in this situation. My own experiment consisted of a pair of sealed 8"s, one with dowel braces much like your own bracing and 3/4" void-free ply of mediocre quality, and one with the same enclosed volume and wall thickness of better ply with a few corner chocks and shelf braces. Upon longer term swapping in and out the nicer cab was consistently preferred due to better 'clarity' and 'impact'. This included a few single-blind tests with buddies but nothing more formal.

None of this has been done with appropriate documentation, but I think you can relate to doing 'good' experimentation to satisfy your own curiousity without feeling the need to go further with it.

[gedlee]

Originally posted by badman

But, given a well-behaved time-domain from the drivers themselves, any enclosure ringing that remains when the excitation drops off suddenly will be dramatically higher in relative amplitude.

As I said before IF the enclosure is going to be audible this IS where it will most likely occur.

Originally posted by badman


I think it's flipsides of the same coin. It's a subjective rather than an objective determination of which dataset is more valuable, data vs. hearing. When building something, it's very difficult to disassociate your prejudices from the evaluation. We're agreed that a 4" thick steel enclosure isn't necessary. But, an easily implemented set of shelf braces sure can go a long way.

But it's possible that you determined via (known flawed, there's no such thing as a 'perfect' measurement) measurements that the enclosure resonances were below an audibility threshold, and that influenced your listening to be willing to ignore them. This is just a flipside of the coin of hearing what you want.

To me the point is that you have to do both, objective and subjective. Unless you can measuire what you hear and hear what you measure, you don;t know what you are doing. The loop has to be closed, but it makes no difference which direction you go arround the circle.

I too take a very long time to decide on anything subjectively regarding sound quality. This is precisely why I gave up going to shows. At one show some guys came into the room, sat down listening to what was probably an unknown piece of music and got up after 15-20 secconds stating "What crap!" One person wouldn't listen because the amp was so bad (although how they knew that was beyond me. I've measured that amp and it actually works quite well.) If thats the evaluation process, then I don't want to be part of it. Yes, there were several other people who spent some time and a couple of them actually bought speakers, but I still have serious concerns over that type of audition.

[badman]

Unfortunately, "The loop" has open sections on both sides- the hearing mechanism fails, but so does the measurement. And they fail differently...

We're agreed regarding shows. I'm generally able to judge when there's high-quality sound but am very careful to avoid negative judgements due to show conditions, since they are unfamiliar everything, usually.

I don't bother with rooms that are really packed, either. Too much distraction and bother. I usually make the rounds many times to try to catch the rooms I'm most interested in when there's a break. And speaking of breaks.... can show organizers PLEASE stop having shows in high-rise hotel venues? It's the worst of all worlds. The old CES and THE show venues were nice because you weren't getting dry-humped by fat audiophiles in the fire-hazard overcrowded hallways. There was outdoor walking area to let your ears and mind relax between exhibits. This was a nice aspect at VSAC 08 as well, there was plenty of opportunity to chill out in the lobby or outside between rooms. I don't think I'll ever bother going to a 'nomal' hotel show again.

I've found that at these shows, also, the best sounding rooms are often speakers I wouldn't like. Dynamically restrained "soft" speakers playing the music that suits them, with lots of acoustic treatment. A little break from the pains of overblown demos from many exhibitors.

[gedlee]

Originally posted by badman
Unfortunately, "The loop" has open sections on both sides- the hearing mechanism fails, but so does the measurement. And they fail differently...

The goal is a solid closed loop and recent studies at Harman and others are showing that this can be done very effectively. Many just don't accept this, but I have always believed that it was possible and even innevitable.