HT vs Musical Subs. Is this just snakeoil? (1 Viewer)

[Saurav]

Bruce,
I have absolutely no idea All I'm trying to say is, it is slightly lopsided to completely ignore parameters that we know are relevant, but have no easy way to measure.
One possible way to evaluate equipment is to go by gut feeling on all the parameters. On the face of it, this looks extremely unscientific. However, if you think about it, measuring one parameter and ignoring the others isn't really scientific either.
Here's an example - let's say you're walking by an apartment with an open window. You hear a piano playing. Can you tell whether it's a real piano, or a recording of a piano? Usually, you can. You can't see what's inside the apartment so there are no visual cues to either aid or distract you. However, there is something in the sound of a real instrument that most stereos do not recreate very well. I believe this has mostly to do with dynamics - real instruments have an attack that stereos just cannot duplicate.
Here's how I judge equipment, sometimes - if I close my eyes in front of the stereo, does a human voice sound like a real person in the room with me, or does it sound like a recording? Do the voices and instruments sound life-like, or does it sound like I have a band of 4" high Barbie dolls on a 3' high pedestal playing in front of me? If I'm in my bedroom, does it sound like a real piano or guitar playing in my living room? That's a rough way of estimating all the parameters together, combined into a total "accuracy" - after all, an "accurate" system's ultimate test would be to sound like the real thing, right? Once you have a system that can do that, it will measure pretty flat in frequency response, as well as any other parameter you can measure. However, frequency response by itself doesn't mean much, IMO.
FWIW, my stereo has startled me sometimes... I was in the kitchen once with a live recording playing, and it made me turn and look out the window because I thought I heard the voice outside. At the same time, my system is pretty inadequate at recreating "large" music - for instance, a live rock concert sounds nothing like actually being there.
Edit: I didn't answer your question at all.... anyway, how would you measure dynamics. Burn a test tone CD with tones at different volume levels. Play them back through your system and measure the SPL at your listening position. Make sure that the output SPL follows the input SPL. That's one crude way of doing this. However, this goes back to a comment in a different thread, this only tests steady-state dynamic response, not transient dynamic response. That would require a test CD with step function signals, and how would you measure the response... maybe with an acceleratometer attached to the speaker's driver, measuring its movement for the supplied input signal? That data could give some idea on the transient dynamic accuracy of a system.

 

[Michael R Price]

I don't mean to start any arguments by mentioning tube amps, etc. It was just meant as an analogy.

Yes, dynamics is an important element too. I'm not sure what constrained dynamics sounds like, though. Haven't listened to any 'constrained' speaker, to be honest. And while I have noticed a bit of harshness sometimes from my system, I think it's because my speakers are still breaking in, or because my room isn't the best setup. I don't think there are any recordings that are harsh by nature. But when I play a rock song at very high level, it does sound harsh. Could this be because of compression, distortion, lack of dynamics? Don't know.

Saurav, what is room overloading?

Dustin, I agree that if you have a good sub, it will be good for HT and music. If you have a poor output sub, it -might- be adequate for music but will be poor for HT. If you have a poor quality sub, it -might- be adequate for HT but will be poor for music.

 

[Saurav]

Saurav, what is room overloading?

I have a vague idea of what it means, not a technical one. Basically, smaller rooms can "support" smaller sound - if you're trying to create a rock concert's volume levels, you really need an acoustic space that's large enough to support it. I'm not totally sure how this plays out in room acoustics terms. Let's try an easier example - if you had a powerful enough sub, could you do, say, sustained 100 - 120 dB of 30 Hz bass in a room of any size? I would think that you would need a large enough room to handle that, wouldn't you? Otherwise the room's gain would combine with the sub's output to make everything sound very bloated, or something like that. Like I said, I'm not an expert in this area, so I'm not completely sure of how this works. Maybe one of the subwoofer experts could explain this, or correct me if I'm wrong.

 

[BruceD]

Saurav,

Well my current understanding, by no means complete, would lead me to believe that by using sophisticated acoustic measurement software you could actually observe the effects of any changes in the speaker-room interface. Changes like speaker placement, parametric EQ, room treatments, etc.

By this I mean using an MLS signal to excite the room/speaker inetrface, recording the impulse response, and then performing FFT on the captured data for additional analysis.

Viewing the impulse response, phase response, frequency response, and 3D frequency-time waterfall charts could give you a pretty clear picture of what is actually happening in your listening room. Also you would have a permanent record to track changes.

 

[Saurav]

I guess the question is this: is there an easy way to interpret that data in terms of the "accuracy" that someone is trying to achieve? You seem to know much more about this kind of testing than I do.

Also you would have a permanent record to track changes.

Good point.

 

[Michael R Price]

Interesting. Unfortunately, no one I know has 'dynamic' speakers like big horns and the like. I don't have a drum solo recording but I may see if I can get one... I can say that especially at high volume, music does seem to have a nice forward impact, but I'm not sure about 'exploding.' My speakers have a sensitivity of about 90 dB.

My speakers do reproduce a bit of the... 'meaty' impact of brass and drums, etc. But not as well as a friend's Klipsch RB-5s. And there was very little of such impact/detail on another friend's dad's old electrostatics. I'm not coming up with the right words to describe this though. Maybe such a feeling is partly due to the high frequency response and time coherency? (It would seem to me that a poorly time/phase-aligned system, like mine, would lose a little bit of the transient 'explosion'. Hmm.)

As far as room overloading, it seems as if you meant the bass response of a room providing too much gain, causing a peak in the bass and hence 'boominess.' At least according to computer simulations, I shouldn't have this problem. Tomorrow or Friday I'll be able to borrow a friend's SPL meter and test response. Someday I'll probably make a DIY microphone and buy a good soundcard/measuring system and the like. Very interesting stuff, but money tends to get in the way of interests.

Saurav, do you think dynamics might have something to do with distortion caused by compression? At least with woofers/subs, higher volume levels can cause the extremes of the signal to be 'flattened' out by a loss of driver strength... I think. If the system can't smoothly reproduce that 'last bit' of a sound, then all bets might be off. Just speculation.

This is a nice discussion though I think the original subject of the thread is lost in space by now.

 

[Saurav]

As far as room overloading, it seems as if you meant the bass response of a room providing too much gain, causing a peak in the bass and hence 'boominess.'

Something like that. I guess it should be possible to EQ the sub to compensate for the room's modes, and achieve flat response at high SPL levels.

 

[Steve_Ma]

Saurav, Michael, Bruce, et all,

Thanks for responding. No, the original question was not lost in space. The discussion merely raised more questions and issues...which I suspected it might.

FWIW, this is exactly the type of feedback I was hoping to hear/read. How do we quantify (assuming we can even identify - lol) some of these subjective priorities, and etc that are involved? I didn't expect an answer to be as simple as the original question might have indicated, but I think I got what I was looking for: More data and opinions from those with more exp than myself. More info to consider and digest.

Regards,

--Steve

 

[Michael R Price]

Ok, not lost in space but... 'morphed.'

The RB-5s are 2-way bookshelf- 8" woofer and 1" tweeter in a 6" 'tractrix' horn I believe. The rated efficiency is 96dB but it may be a bit lower. I like its sound a lot.

Transient accuracy and 'edge': I wonder if there are any speakers designed for high efficiency/dynamics AND time-phase alignment...?

Anyway, have to go now but this is opening up some interesting ideas (such as, can you measure dynamics, and is time-alignment the last step to getting music to 'explode'/snap?)

 

[BruceD]

Steve,

The only way I have found to correlate what I hear with what I like in a repeatable fashion (without taking 10 years to train my ears) is to use the acoustic software I was discussing in the earlier thread. With this sw it only took me 2 years.

With the ETF5 software I'm able to identify the sound character of things like RT-60 echo, first reflections, modal room peaks, phase anomolies, ringing at certain frequencies, etc by identifying them in acoustic graphs and them changing something like speaker position or a room treatment to eliminate it.

It has really helped me tune my enviroment and learn what I like or dislike.

 

[Alexis]

For good drum tracks, look at the Kodo CD's or the Sheffield Labs test cd with the drum tracks.

The sheffield lab is great at testing if your sub is truely musical.

 

[Mark Seaton]

Hi Michael,

Transient accuracy and 'edge': I wonder if there are any speakers designed for high efficiency/dynamics AND time-phase alignment...?

Was this a set-up for me?
The problem is that this is not a trivial task to accomplish. For conventional, multi-driver, direct radiating dome and cone loudspeakers, you will always have a spacial mis-alignement of the origin of the sound. The speakers marketed as "time-aligned" are in fact only correct at one point in space, or at best on one single plane due to the multiple origins of the sound.
The company I work for does indeed have a technology which enables, and actually delivers, exactly what you are asking about. Our Unity Summation Aperture allows the implementation of a multi-way speaker in a single, constant directivity horn. With the correct physical configuration and crossover, the speakers are time-correct over the entire coverage angle of the horn. While we only have two of the products on the website, we have two, more compact speakers which are already in use for professional applications, which I am using in my own home system, and which we will be introducing to the personal theater market. Unfortunately they are expensive to manufacture, and will not be priced for the high volume masses ($5,000 retail each), but they will deliver what you are asking about in terms of dynmaics, efficiency, low distortion, and time/phase accuracy... and no horn honks. Check out our admittedly lame website at www.soundphysics.com , and directly you can check out our Unity VFL. To answer your question though, indeed, large dynamics, low distortion, low compression and uniform response bring us MUCH closer to lifelike reproduction. While the importance of time/phase accuracy is a topic of debate, and is convoluted by other linked properties, I have only heard improvements when a system is brought closer to correct. With our large VFL and TD-1 loudspeakers combined with our professional bass horn, the BassTech7, we can experiment with correct phase and full-range response, although not practical in most home environments. Yes, it's quite an experience.
Hope this helps,

 

[Michael R Price]

Wow, that looks amazing. Unfortunately, I had enough trouble affording $450 speakers.
I had a thought a little while ago, about time alignment. If you had a speaker that wasn't time aligned, could you use a computer to 're-align' it with different delays at different frquencies? (Or, also, equalize the signal without phase errors.) I was thinking this could be done with a Fourier transform program which would be able to analyze the signal and figure out how to change the signal in order to correct it. Does anyone sell such a program? Of course, we'd have to deal with using a computer as a source, but the results would be interesting.
However, I don't know of any way to artificially fix problems with dynamics and compression.

 

[Saurav]

Here's the issue I see with time alignment of multiple drivers that are physically distant from each other - move your head (i.e., listening position) an inch, and you need to re-do your delay calculations. The time alignment only works at a point, or a line or a plane.

The only way to get multiple drivers to be "completely" time aligned is to have them physically be in the same location. Which is impossible to achieve, so the closest you can get is using concentric drivers (which have very little physical distance between the individual cones), or full-range single drivers.

 

[Mark Seaton]

Hi guys,

Surav wrote:

Here's the issue I see with time alignment of multiple drivers that are physically distant from each other - move your head (i.e., listening position) an inch, and you need to re-do your delay calculations. The time alignment only works at a point, or a line or a plane.

Surav answered your question Michael as the primary problems in "correcting" a speaker lies in the spacial/ 3 dimensional issue of things. This is what our Unity horn technology solves. Both Tact and ClarityEQ have products which can automatically correct the phase and amplitude response of your speakers using a measurement taken at the listening position. Implementing this is really a matter of DSP power and the understanding to implement the equations. By no means trivial, but not hard for someone versed in both DSP and loudspeaker operation. The problem comes from execution. Unless two acoustic sources producing the same frequency are less than ideally 1/4 wavelength, they will not behave as a single source. All of the loading points within our Unity horn are in fact within 1/4 wavelength of eachother at the frequencies they overlap.

Extending Surav's last point about the ability to only correct for 1 point or plane in space, you then have to consider that while you can correct the direct energy from the speaker at the listening position, the reflected energy will not have the same accuracy, where a real/live source will reflect off surfaces coherently. There are a lot of things to consider with respect to time/phase/amplitude accuracy, yet they are quite intertwined, so true consideration of the significance any one alone can be rather difficult.

Regards,

 

[Mario_C]

Some Interesting reading.
http://www.audioperfectionist.com/Ne.../journal1a.pdf
http://www.audioperfectionist.com/Ne...s/journal2.pdf

 

[Steve_Ma]

Great links. Very interesting reading.

Thanks Mario.