I've seen some references to the idea that a bipole driver configuration can eliminate baffle step effects. I'm not seeing it.
Any driver putting out long-wavelength frequencies (compared to box dimensions) is going to radiate into 4*pi steradians. Put two drivers on the front of a box, and both are going to "creep" around in the lower frequencies and drive the whole hemisphere. Put one on the front and one on the back, and both are going to do the same thing.
Am I misunderstanding the argument or is it not grounded in reality? At the end of the day, it shouldn't matter where you place a driver if the wavelengths it is sourcing are large compared to the box (baffle) dimensions.
Any driver putting out long-wavelength frequencies (compared to box dimensions) is going to radiate into 4*pi steradians. Put two drivers on the front of a box, and both are going to "creep" around in the lower frequencies and drive the whole hemisphere. Put one on the front and one on the back, and both are going to do the same thing.
Am I misunderstanding the argument or is it not grounded in reality? At the end of the day, it shouldn't matter where you place a driver if the wavelengths it is sourcing are large compared to the box (baffle) dimensions.
I think the reason for this idea would be that if you have 2 identical mounted on identical baffles 180 degrees opposed, then above the baffle step frequency, you would just be hearing the driver pointing at you at its full sonic output and nothing from the one pointing away, and below the baffle step frequency you would hear half the sonic output of the driver pointing at you while half "creeped" around the back and would remain unheard, and you would hear half the output of the driver facing away from you, the half that "creeped" around the back relative its front.
Correct, my assumption is that (below the baffle step frequency) we would hear the "front" half of the front-firing driver and also the "back" half of the rear-firing driver. Both radiate into the full room, thereby knocking each driver's contribution down by 50% in terms of local acoustic power density. I just don't see how this is an improvement.
Saw a lot of interesting/good info in a quick browse, a pity he was banned [why?]: melonhead at diyaudio site:www.diyaudio.com - Google Search
GM
What creeps round the front from the back replaces what's creeped round the back from the front, why would the one knock the other down, unless you are thinking of the way a dipole works?
I'm wondering if the "theory" is more interpreted as going from one driver to two, with each summing on the opposite side of the other, thus mitigating the baffle step. What this misses is the fact that of course the SPL will get louder -- the driver count doubled. It's irrelevant where the drivers are located (disregarding any comb filtering alluded to above).
jimmyjazz, it all depends on wavelength, baffle dimensions and driver diameter.
We usually see simplified explanations and graphic presentations of radiation patterns, which is sad. 20-20000Hz means wavelength of 17,2 meters - 1,72 centimeters, thousandfold difference.
Basically baffle width (radius from drivers centerpoint to nearest edge) determines where front- and backside waves meet each other and interfere. When they have exatly same phase angle, they sum up (+6dB) When angles are different summation is milder until at 180¤ there is total cancellation (null dB)
So, if we look at a bipole loudspeaker's on-axis spl response, low frequencies get boost, then there is a dip, then bump, then dip,bump, dip over and over again.
Things get more difficult when we realize, that when the wavelength is shorter than the drivers radiating area diameter, the radiation pattern starts to get narrower and the spl at 90¤ (following the baffle surface) attenuates rapidly, the more the higher we go in frequency.
That was just thinking about "on-axis" sound pressure. Off-axis response gets first messy and then start to attenuate when we go more off-axis and higher in frequency.
And thinking gets even more difficult if we place that speaker in a room with some walls that induce reflections and standing waves. Our hearing system has two "mics" and a very clever analyzing system with many different ways to equalize, localize and interprete the sound. A measurement microphone at some randon spot in the room gives a very peculiar spl response and impulse presentation, when we look at the analysis.
Dr. Russell's animations are nice Dan Russell's Acoustics and Vibration Animations
We usually see simplified explanations and graphic presentations of radiation patterns, which is sad. 20-20000Hz means wavelength of 17,2 meters - 1,72 centimeters, thousandfold difference.
Basically baffle width (radius from drivers centerpoint to nearest edge) determines where front- and backside waves meet each other and interfere. When they have exatly same phase angle, they sum up (+6dB) When angles are different summation is milder until at 180¤ there is total cancellation (null dB)
So, if we look at a bipole loudspeaker's on-axis spl response, low frequencies get boost, then there is a dip, then bump, then dip,bump, dip over and over again.
Things get more difficult when we realize, that when the wavelength is shorter than the drivers radiating area diameter, the radiation pattern starts to get narrower and the spl at 90¤ (following the baffle surface) attenuates rapidly, the more the higher we go in frequency.
That was just thinking about "on-axis" sound pressure. Off-axis response gets first messy and then start to attenuate when we go more off-axis and higher in frequency.
And thinking gets even more difficult if we place that speaker in a room with some walls that induce reflections and standing waves. Our hearing system has two "mics" and a very clever analyzing system with many different ways to equalize, localize and interprete the sound. A measurement microphone at some randon spot in the room gives a very peculiar spl response and impulse presentation, when we look at the analysis.
Dr. Russell's animations are nice Dan Russell's Acoustics and Vibration Animations
Last edited:
Sure, but I'm trying to specifically address the claim that one can mitigate baffle step effects by putting woofers on opposite sides of a cabinet. This allegedly "cancels out" the (theoretical) 6 dB drop in response below the baffle step frequency.
I don't think it does. Either the benchmark needs to be those same two woofers on the front face of the cabinet, or it needs to be a single woofer on the front of the cabinet. In the former case, I don't think the listener would hear a difference, since both woofers would still be radiating low frequency content into 4*pi steradians of space. In the latter case, I *DO* think there would be an audible difference, because now there are two woofers playing instead of one. It shouldn't matter where those drivers are located -- BOTH will experience the baffle step effect at lower frequencies in a small box.
I don't think it does. Either the benchmark needs to be those same two woofers on the front face of the cabinet, or it needs to be a single woofer on the front of the cabinet. In the former case, I don't think the listener would hear a difference, since both woofers would still be radiating low frequency content into 4*pi steradians of space. In the latter case, I *DO* think there would be an audible difference, because now there are two woofers playing instead of one. It shouldn't matter where those drivers are located -- BOTH will experience the baffle step effect at lower frequencies in a small box.
It's been explained quite well here already. What you lose in forward radiation you gain in rear driver wraparound, and the wavelengths are long enough that it's constructive summation. With double forward woofers you still have a step you need to compensate for, you just have more woofer to deal with it, hence 2.5 way designs where the second woofer only plays up to baffle step. Bipole does the same thing but more elegantly.
Well, call me stupid then, because I'm still not seeing it. In EITHER case both woofers are radiating into 4*pi steradians, right? They both put half their sound power around to the other side of the (small) cabinet. Essentially, both are monopoles at those low frequencies. I don't see how it matters on which side of the cabinet they are mounted.
If the explanation is that such an arrangement doesn't compensate baffle step at very long wavelengths but DOES at some shorter wavelengths, then I can see how that would be true, but I've been trying to focus on bass response.
If you can point me to a link that explains it I would be grateful.
If the explanation is that such an arrangement doesn't compensate baffle step at very long wavelengths but DOES at some shorter wavelengths, then I can see how that would be true, but I've been trying to focus on bass response.
If you can point me to a link that explains it I would be grateful.
Baffle loss/step phenomenom is gradual, it depends on wavelength like I said, and takes 4 octaves to realize. With a real world loudspeaker construction, we get the 6dB benefit in low end no matter how drivers are located. Differencies start appearing at transition zone and above it.
Just study this more and please make a prototype speaker, do measurements yourself! It is easy and fun! Playing with a proto gives you totally different perspective and vision how different parameters affect soundfield pattern and spl.
Just study this more and please make a prototype speaker, do measurements yourself! It is easy and fun! Playing with a proto gives you totally different perspective and vision how different parameters affect soundfield pattern and spl.
Juhazi, why are you referring to a "benefit" in the low end? Are you referring to the benefit gained by adding the 2nd driver?
I completely understand the graph you just posted -- it shows a 6 dB drop in the bass because of the monopole nature of the driver once the wavelengths get much larger than the box dimensions.
I'm not trying to be pedantic. I'm trying to be precise and understand if my expectation that bipole woofers DON'T mitigate baffle step is correct, and if I'm wrong, why? Consider it a thought experiment. I appreciate everyone's efforts here. It's a great resource.
I completely understand the graph you just posted -- it shows a 6 dB drop in the bass because of the monopole nature of the driver once the wavelengths get much larger than the box dimensions.
I'm not trying to be pedantic. I'm trying to be precise and understand if my expectation that bipole woofers DON'T mitigate baffle step is correct, and if I'm wrong, why? Consider it a thought experiment. I appreciate everyone's efforts here. It's a great resource.
Was that comment for me? If so, yes, of course I see that. I have no problems understanding baffle step. My skepticism is related to the "solution" some have proposed wherein they move one woofer (of a pair) from the front of the box to the back of the box.
Again, I'm *NOT* trying to be pedantic here. I know it seems otherwise.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.