brianpowers27 said:
This is an interesting theory, and as someone who had done alot! of drywall work recently the resonant frequency of 5/8" drywall on 16" centers is right in the upper bass range.
Dr. Geddes has some extensive knowledge in this area, maybe he can address this concern?
-David
gedlee said:
That's the odd thing though, I've heard it with a pair of monopoles where it not only is NOT a disaster, it's actually *very* similar.
I've also heard and "cobbled together" both very good AND bad multi-subs configurations, where both results (good and bad) were *irrespective* of how much or little the freq. uniformity was ("flat") at lower freq.s.
On the best monopoles be it, single "subs", "non-uniform" multiple "subs", or a stereo pair, I often come away with this impression:
"I notice modal problems, especially when entering/exiting a high/low amplitude deviation area, and yet despite this the sound is fantastic." (..or something like that.)
Of course a *good* multi-sub setup can give you both uniformity AND excellent sound (that is not specifically accredited to uniformity).
The one thing that I've found where a Dipole can be superior is with respect to apparent channel separation at lower freq.s. Those side-nulls between sources seem to make a difference in the mid-bass region, providing a marginal yet distinct enhancement in clarity with respect to lateral image placement.
I'll present this link for those interested. http://www.musicanddesign.com/Dipole_modesA.html
We have been through all this before so i don't care to enter the discussion again. However, I will say, as shown in the analysis, that the major difference between dipoles, monopoles and cardioids are that dipoles will not excite the 0,0,0 mode (not room pressurization). Monopoles and cardioids will. With regard to excitation of modes, dipoles may excite fewer, depending on position, but dipoles are also the most sensitive to placement and orientation in that regard. Cardioids are least sensitive to position. Lastly, cardioids and dipoles radiate less power into the room assuming they have the same on axis free field response.
At this time I find that the major differences between the sound of different woofer systems is due the radiated power and the the room pressurization effects, and room pressurization may be the biggest issue, positive or negative, as the case may be.
Also consider: http://www.musicanddesign.com/DP_woofer_room.html
We have been through all this before so i don't care to enter the discussion again. However, I will say, as shown in the analysis, that the major difference between dipoles, monopoles and cardioids are that dipoles will not excite the 0,0,0 mode (not room pressurization). Monopoles and cardioids will. With regard to excitation of modes, dipoles may excite fewer, depending on position, but dipoles are also the most sensitive to placement and orientation in that regard. Cardioids are least sensitive to position. Lastly, cardioids and dipoles radiate less power into the room assuming they have the same on axis free field response.
At this time I find that the major differences between the sound of different woofer systems is due the radiated power and the the room pressurization effects, and room pressurization may be the biggest issue, positive or negative, as the case may be.
Also consider: http://www.musicanddesign.com/DP_woofer_room.html
"(not room pressurization)." was a typo. No "not". Room pressurization is the same old same old. Monopoles and cardioids have a finite volume displacement. Dipoles have a net volume displacement of zero. I know, rooms leak, but put a dipole in a small room and the bass will not be over loaded below the room's first non-zero mode, or room fundamental, what every you want to call it.
Frankly I think this is one of the basic reasons people find dipole bass so "different". I have been in a lot of rooms where the woofers are just overloading the room below the fundamental and the bass is mud. Put a dipole woofer in there with the same free field on axis response and things sound immediately better, cleaner, better defined. But correct the overloading with a monopole and there isn't a lot of difference. That's why I said room pressurization can be a plus or minus.
Frankly I think this is one of the basic reasons people find dipole bass so "different". I have been in a lot of rooms where the woofers are just overloading the room below the fundamental and the bass is mud. Put a dipole woofer in there with the same free field on axis response and things sound immediately better, cleaner, better defined. But correct the overloading with a monopole and there isn't a lot of difference. That's why I said room pressurization can be a plus or minus.
Thanks to the OP for the measurements - I think they are very telling of typical low frequency driver implementations. I recently been doing in room measurements, and have been pretty shocked to see what happens in the bass. I'm going to do a little write up soon, but my opinion now is that in room measurements are very necessary.
Thanks John for this series of articles - reading them was my impetus for doing measurements - one of the things I found was that, measured at my listening position of about 3m, my h-frame woofers response changed little with substantial rotation about its center (+/- 30deg), and I had to move things many feet to get largely different interference patterns at the listening position (I might add my 'room' is not rectangular and relatively large). But more to follow later...
john k... said:
Thanks John for this series of articles - reading them was my impetus for doing measurements - one of the things I found was that, measured at my listening position of about 3m, my h-frame woofers response changed little with substantial rotation about its center (+/- 30deg), and I had to move things many feet to get largely different interference patterns at the listening position (I might add my 'room' is not rectangular and relatively large). But more to follow later...
john k... said:
Hi John
I knew something was wrong. (Editing can prevent a lot of arguments down the road - I should have attacked you on this one!!
)As I have said, I have never been able to measure this "pressurization mode" in a real room (at least not at a significant level, i.e. comparable to a room mode, like 1,0,0 in my room) and I can only assume that leakage is the issue. That dipole bass sounds better in many rooms than the typical single monopole is, I think, fairly well established, but I am not sure that I'd jump to "pressurization" as the answer. I think that it might be more complicated than that. I do know that if you get multiple monopole subs setup properly in any room then the bass will sound as good or better than the typical dipole setup. This would tend to discount the "pressurization" explaination.
I did measurements several years ago when developing the NaO II and I certainly saw differences between monopole, diopole and cardioid below the room fundamental in my room. The dipole pretty much followed the anechoic roll off, the monopole showed a definate lack of such roll off and the cardioid was somewhere in between. The monopole was a sealed box. All three sources had the same free field on axis response, Q = 0.5, Fs = 25 Hz.
Don't care to argue about it. If you see diferently in your room then it must be so there.
Don't care to argue about it. If you see diferently in your room then it must be so there.
Hello,
I'm using sinusoidal bursts of 5 to 10 cycles with a raised cosine envelope (Hanning window). I think it is very suitable and informative for measuring speakers in a room.
Yes, burst response can be calculated, but I prefer using the bursts as real excitation signals because I can listen at the same time how the speakers (and the room!) sounds like.
One can also observe and hear for example room structural resonances in a very interesting way with the bursts. When one turns up the volume a bit, window glasses, room doors and some furnitures start to make noise. It is interesting to hear how a resonanse slowly builds up when burst begins and then continues ringing after the burst has died out. Very informative I think, because the music will do the same thing. Something that cannot be heard in the same way with steady state sinusoids, nor with MLS noise.
It is a good idea to try bursts of different number of cycles as the bandwidth depends on it. A very short burst may not excite a (structural) resonance above the hearing threshold but a longer burst can.
Music is a temporal experience! Thus we should optimise the system keeping that in mind. I think
- Elias
jeepee said:
jason_watkins said:
I'm using sinusoidal bursts of 5 to 10 cycles with a raised cosine envelope (Hanning window). I think it is very suitable and informative for measuring speakers in a room.
catapult said:
Yes, burst response can be calculated, but I prefer using the bursts as real excitation signals because I can listen at the same time how the speakers (and the room!) sounds like.
One can also observe and hear for example room structural resonances in a very interesting way with the bursts. When one turns up the volume a bit, window glasses, room doors and some furnitures start to make noise. It is interesting to hear how a resonanse slowly builds up when burst begins and then continues ringing after the burst has died out. Very informative I think, because the music will do the same thing. Something that cannot be heard in the same way with steady state sinusoids, nor with MLS noise.
It is a good idea to try bursts of different number of cycles as the bandwidth depends on it. A very short burst may not excite a (structural) resonance above the hearing threshold but a longer burst can.
Music is a temporal experience! Thus we should optimise the system keeping that in mind. I think
- Elias
Re: Re: Measured monopole and dipole room responses
Hello,
I don't know what is your definition of a steady state, but for me it means a state where signal variation is over and signal is continuous or zero. That situation will never happen during music where signal envelope is constantly varying.
If we could hear only steady state bass, we could not hear the bass at all from most of the records. There is integration time indeed, but it does not mean the sound will appear like a snap shot after the time has passed, rather the perception immediately begins and only the final touch will be set after integration is over. It makes things sound 'smoother' so to say. But it does not prevent us from hearing the temporal effects.
Bass transients are easy to hear and can be seen as envelope modulations at the ear signals. This low frequency modulation is one good thing to increase the perception of envelopment as Griesinger states it.
It's easy to try out and listen with your own ears. Linkwitz has it:
http://www.linkwitzlab.com/images/graphics/mlt-bst1.gif
When I do this test comparing monopole and dipole bass, monopole seems (hears?) to lose the ability to reproduce the envelope variation, and dipole does it better. As I see why this is so, it's because the monopole sparays more waste energy into the room in a wider radiation angle than a source which has some directivity like a dipole.
- Elias
Hello,
gedlee said:
I don't know what is your definition of a steady state, but for me it means a state where signal variation is over and signal is continuous or zero. That situation will never happen during music where signal envelope is constantly varying.
If we could hear only steady state bass, we could not hear the bass at all from most of the records. There is integration time indeed, but it does not mean the sound will appear like a snap shot after the time has passed, rather the perception immediately begins and only the final touch will be set after integration is over. It makes things sound 'smoother' so to say. But it does not prevent us from hearing the temporal effects.
Bass transients are easy to hear and can be seen as envelope modulations at the ear signals. This low frequency modulation is one good thing to increase the perception of envelopment as Griesinger states it.
It's easy to try out and listen with your own ears. Linkwitz has it:
http://www.linkwitzlab.com/images/graphics/mlt-bst1.gif
When I do this test comparing monopole and dipole bass, monopole seems (hears?) to lose the ability to reproduce the envelope variation, and dipole does it better. As I see why this is so, it's because the monopole sparays more waste energy into the room in a wider radiation angle than a source which has some directivity like a dipole.
- Elias
gedlee said:
Would you have some temporal (not steady state please) measurement of the multiple bass configuration comparing it's performance with a single monopole or perhaps a dipole? That would be very interesting comparison I think.
I don't have multiple bass boxes at hand to do the test myself because evolution brought me away from boxes
- Elias
Hi Elias,
Thanks for this post. I've been following it but I'm not up to your level with data etc. to contribute.
Even SL can only subjectively say that dipole bass is better (and he admits so), your measurements take this dipole-monopole perception to different level!
I've had monopole and dipole bass and definitely the dipole sound more realistic. In fact I can't live with monopole sub. It's signature is everywhere. It's good to understand better why.
If there's suggestion that there is something "better" than dipole bass, let's see the data
Keep it coming!
Thanks for this post. I've been following it but I'm not up to your level with data etc. to contribute.
Even SL can only subjectively say that dipole bass is better (and he admits so), your measurements take this dipole-monopole perception to different level!
I've had monopole and dipole bass and definitely the dipole sound more realistic. In fact I can't live with monopole sub. It's signature is everywhere. It's good to understand better why.
If there's suggestion that there is something "better" than dipole bass, let's see the data
Keep it coming!
Hello John,
Things are getting most interesting when some mathematics is involved
I agree with your statement you made in your link that one dimensional room analysis is pretty useless to see the real efects. About 10 years ago I wrote a code to calculate 3D room responses using the ray tracing method.
Here's some old results of comparing monopole, dipole and cardioid burst temporal efects:
http://www.diyaudio.com/forums/showthread.php?postid=1629620#post1629620
And here's the simulated impulse responses:
http://www.diyaudio.com/forums/showthread.php?postid=1631119#post1631119
Now I find it interesting for myself when comparing the 3D simulations with the room measurements and to the perceived responses.
All the modal analysis and eigenfunctions you present are interesting, but how are they perceived? A room mode, as I see it, is a steady state response and does not occur in most of the real music signals.
- Elias
Things are getting most interesting when some mathematics is involved
I agree with your statement you made in your link that one dimensional room analysis is pretty useless to see the real efects. About 10 years ago I wrote a code to calculate 3D room responses using the ray tracing method.
Here's some old results of comparing monopole, dipole and cardioid burst temporal efects:
http://www.diyaudio.com/forums/showthread.php?postid=1629620#post1629620
And here's the simulated impulse responses:
http://www.diyaudio.com/forums/showthread.php?postid=1631119#post1631119
Now I find it interesting for myself when comparing the 3D simulations with the room measurements and to the perceived responses.
All the modal analysis and eigenfunctions you present are interesting, but how are they perceived? A room mode, as I see it, is a steady state response and does not occur in most of the real music signals.
- Elias
john k... said:
As you can see, the resonse at any frequency is composed of the sum of contributions for all modes. (However, at low frequency only the first 4 or 5 modes make a significiant contribution.) Each individual mode has the form of a 2nd order low pass filter with fc = the modal frequency and Q determined by the damping. As such, the impulse response is the sum of the impulse response from each mode. The Fourier transform relationship between frequency response and impulse response still holds.
Of course, the impulse response is dependent on source and listening position. But for a given source position the impulse at any point in the room can be obtained through an IFFt which could be convolved with any signal applied to the source to to get the time response at the listening position.
Of course, the impulse response is dependent on source and listening position. But for a given source position the impulse at any point in the room can be obtained through an IFFt which could be convolved with any signal applied to the source to to get the time response at the listening position.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.