Konnichiwa,
There is something which I note EACH AND EVERY comment on BDS misses. A speaker that within 1/2 wavelength of a wall (say 1m distance to the rear wall for 170hz) will not radiate into 4PI Seradian but into something close to 2PI Steradian. But because it actually stands on the floor it will radiate into 1PI Steradian and because it is usually also within 1m or so fo a sidewall it actually radiates into 1/2PI Steradian.
So the BDS effect on the speaker in a real room is NOT 6db shelf, but a shallow depression in the frequency resonse, which can be left untreated unless excessivey narrow baffles (< 50cm wide) are employed, which don't work acoustically anyway. So the whole BDS chestnut is simply a result of bad acoustic design (narrow baffles) and best addressed at the source. Look at Snell K or AN/E to see a Speaker sensibly designed with respect to BDS.
Alternatively the LF Alignment must be adjusted such that the LF rolls off at a suitable rate (1st order?) below the point where coupling to nearer surfaces increases the net SPL again.
Sayonara
Vikash said:
There is something which I note EACH AND EVERY comment on BDS misses. A speaker that within 1/2 wavelength of a wall (say 1m distance to the rear wall for 170hz) will not radiate into 4PI Seradian but into something close to 2PI Steradian. But because it actually stands on the floor it will radiate into 1PI Steradian and because it is usually also within 1m or so fo a sidewall it actually radiates into 1/2PI Steradian.
So the BDS effect on the speaker in a real room is NOT 6db shelf, but a shallow depression in the frequency resonse, which can be left untreated unless excessivey narrow baffles (< 50cm wide) are employed, which don't work acoustically anyway. So the whole BDS chestnut is simply a result of bad acoustic design (narrow baffles) and best addressed at the source. Look at Snell K or AN/E to see a Speaker sensibly designed with respect to BDS.
Alternatively the LF Alignment must be adjusted such that the LF rolls off at a suitable rate (1st order?) below the point where coupling to nearer surfaces increases the net SPL again.
Sayonara
Re: Re: Baffle step basics
Many people claim that narrow baffles offer better soundstage and imaging capability. Do you disagree with this view or do you feel that these things are less important than a flatter curve?
What do you mean by "which don't work acoustically anyway"? Something other than BDS?Kuei Yang Wang said:
Many people claim that narrow baffles offer better soundstage and imaging capability. Do you disagree with this view or do you feel that these things are less important than a flatter curve?
Re: Re: Re: Baffle step basics
Konnichiwa,
Yes, the phenomenae is called wavelaunch. It is related to BDS but not the same. A narrow baffle cannot "launch" a coherent wavefront once the waves bend around it. Subjectively this percieved as lack of impact in the upper bass/lower mid EVEN if the frequency response is corrected to flat by equalisation.
Narrow baffles have ANY (positive) impact on soundstaging only if inaproriate drivers are used. Small diameter domes are the worst offenders as they have a totally uncotrolled radiation pattern, naturally so as they where designed and invented for upward pointing operation in omnidirecrional speakers by Stu Hegeman.
For arguments sake, soffit mounting studio monitors maximises the Baffle Area to the size of the wall in which they are mounted. Yet they offer a spatial rendering that very closely approaces the Mike feed closely (I have had many occasions to test this).
Of course, excessively wide dispersion drivers on narrow baffles produce loads of unwanted radiation side and rearwards which will lead to severe combfiltering and increased reflected sound. In a studio we usually use a reverb unit to achieve similar effects. Yes, adding reverb makes the sound more spacious, by why do it with a loudspeaker?
Sayonara
Konnichiwa,
7V said:
Yes, the phenomenae is called wavelaunch. It is related to BDS but not the same. A narrow baffle cannot "launch" a coherent wavefront once the waves bend around it. Subjectively this percieved as lack of impact in the upper bass/lower mid EVEN if the frequency response is corrected to flat by equalisation.
7V said:
Narrow baffles have ANY (positive) impact on soundstaging only if inaproriate drivers are used. Small diameter domes are the worst offenders as they have a totally uncotrolled radiation pattern, naturally so as they where designed and invented for upward pointing operation in omnidirecrional speakers by Stu Hegeman.
For arguments sake, soffit mounting studio monitors maximises the Baffle Area to the size of the wall in which they are mounted. Yet they offer a spatial rendering that very closely approaces the Mike feed closely (I have had many occasions to test this).
Of course, excessively wide dispersion drivers on narrow baffles produce loads of unwanted radiation side and rearwards which will lead to severe combfiltering and increased reflected sound. In a studio we usually use a reverb unit to achieve similar effects. Yes, adding reverb makes the sound more spacious, by why do it with a loudspeaker?
Sayonara
The baffle loss step results in a 3-db change in POWER response due to the change of acoustic impedance. Ignoring other acoustic impedance issues, like the floor or nearby walls, this translates to a 6-db change in on-axis frequency response. Equalizing the on-axis frequency response will not fully correct the power response, so how the baffle step is dealt with can have different sonic results.
It is probably ideally dealt with by adjusting acoustical impedances, e.g., by designing in tradeoffs with baffle size, crossover frequency, plural drivers in the same band below step, and speaker placement relative to boundaries. (Two drivers working together will add acoustic impedance, if they are close enough together vs the wavelength.)
The sphere is the best external shape to avoid *edge* diffraction. (You HAVE to place a driver symmetrically in a sphere, by definition.) However, the baffle step will be at a higher frequency than a rectangular baffle, because the sphere starts narrowing sooner than its overall width. Placing a driver in the side of a cylinder will give you almost as smooth a response, edge-defraction wise, as a sphere, but provide baffle support to a lower frequency. It is also much easier to construct.
It is probably ideally dealt with by adjusting acoustical impedances, e.g., by designing in tradeoffs with baffle size, crossover frequency, plural drivers in the same band below step, and speaker placement relative to boundaries. (Two drivers working together will add acoustic impedance, if they are close enough together vs the wavelength.)
The sphere is the best external shape to avoid *edge* diffraction. (You HAVE to place a driver symmetrically in a sphere, by definition.) However, the baffle step will be at a higher frequency than a rectangular baffle, because the sphere starts narrowing sooner than its overall width. Placing a driver in the side of a cylinder will give you almost as smooth a response, edge-defraction wise, as a sphere, but provide baffle support to a lower frequency. It is also much easier to construct.
Well, this is pretty interesting:
So, for this example, sub 170hz freqs are actaully getting a boost, not a cut. Is this the same as room gain or domething slightly different?
Gee, is a 500mm baffle considered less than optimal? That's a wiiiide baffle.
Mos
So, for this example, sub 170hz freqs are actaully getting a boost, not a cut. Is this the same as room gain or domething slightly different?
Gee, is a 500mm baffle considered less than optimal? That's a wiiiide baffle.
Mos
Interesting concept, Vikash. Mind you, it might be simpler and give similar results to take a circular baffle and mount the driver off-centre.Vikash said:
I've knocked up a quick SVG page (you'll need a viewer ) to illustrate how a wave reaches the baffle and falls of gradually . Yes, I know it's all too much, but it was fun 🙂.
For comparisson, I've just added a second example on the link above that shows the fall-rate when just mounting the driver off-centre on a circular baffle. I don't dispute that there may be little perceived difference when all things are considered, but the fall-rate of my shell type curvy baffle thingamabob seems more appealing at first glance for a smooth baffle step.
Well I'm looking at all the possibilities Steve. I see numerous problems with such a baffle, but I'm looking at it purely with baffle-step in mind. I mean it wouldn't be the most practical design in the word in terms of size, and realistic options of enlcosing the driver will make the whole enclosure look quite daft. It could be a recommendation for those looking to build a dipole perhaps. And baffle-edge diffraction can be dealt with in the usual way.
In fact all these enclosure boundary effects are driving me nuts, so I'm starting to look into the dipole design. Perhaps it will give me a better speaker...
In fact all these enclosure boundary effects are driving me nuts, so I'm starting to look into the dipole design. Perhaps it will give me a better speaker...
I don't know. Think snail. 😎Vikash said:
Dipoles can be excellent. For me, I love the challenge of the enclosure, to try for the openness of a dipole with the room-friendliness of a small box.
I'm lucky with the Nonsuch 4 because the diffraction step is part compensated because I use four drive-units. There's a skew on the efficiency increase that I get this way and it's not a bad mirror of the baffle step.
I am working on a more radical approach to the diffraction question for future designs and I'll report back later, if there's any mileage in it.
But you're right. It does drive you nuts. Luckily I was there already.
- Status
- Not open for further replies.