Member
Joined 2009
I have a question about Physical Time Alignment. I want to find out the physical distance from the flange of the driver to the center of the voice coil. Can somebody post step by step instructions of how this is done with measuring software? (I use HOLMImpulse).
I know the classic method is to measure the Impulse Response of the tweeter, then lock the Time Zero option in the software and measure the Woofer. I can't seem to get a valid measurement doing that. Where should the microphone be in this case?
Is there a technique to measure a single driver when we know the distance from the microphone to the flange?
I know the classic method is to measure the Impulse Response of the tweeter, then lock the Time Zero option in the software and measure the Woofer. I can't seem to get a valid measurement doing that. Where should the microphone be in this case?
Is there a technique to measure a single driver when we know the distance from the microphone to the flange?
time coherence is false. I have studied this extensively and found that the woofer radiates sound from all different points on the cone. So you can't 'align' all points with one another. The other problem is you can only align it for one listening position.
You'll be in a better position if study is extended to understand where the approximation of an acoustically small driver as a point source is and is not valid. Try searching on acoustic size and radiation patterns.
This should also help you see where your understanding of polar behaviour is incorrect.
This should also help you see where your understanding of polar behaviour is incorrect.
lilun, probably out of scope (and date) for the OP, and using commercial DSP software, but you may find this article illuminating: Advanced Acourate Digital XO Time Alignment Driver Linearization Walkthrough
It would be very interesting to see how many members have fully active systems. (I suspect the ones who are the most satisfied with their speakers and, as a side-note, those who think amplifiers are secondary to the primary source of distortion.)
Despite not believing or understanding, probably, everything you have posted on this forum, I think you are right with this, and probably not for the reason you have.
Everything we hear is essentially based on the position of the listener's head. Turning or tilting the head changes the position of the ears in relation to the sources of the music and , therefore, changes the "time coherence". We all do it all of the time. When we can't locate a sound, we all turn and/or tilt our heads to try to better the source of the sound.
None of us stays in a perfectly static position when listening so the time alignment changes all of the time. I, for one, am unconcerned with this because it adds to the ever-changing experience of listening to good music. That said, I think alignment of the drivers (recessed into the baffle) is good enough on a flat baffle. Thereafter, I think the laws of diminishing returns and imagination (confirmation bias) take over.
Why did you start this thread; http://www.diyaudio.com/forums/multi-way/248130-hard-get-time-coherency.html
or this thread;
http://www.diyaudio.com/forums/multi-way/248086-crossovder-question.html
Both go hand in hand and if you had studied extensively, these Q's would be moot.
Greebster,
Explaining is futile. He is the Borg and speaker builders will
be assimilated.
In a normal time aligned 2 way the baffle can be slanted or drivers Z axis shifted to match the other driver at the crossover point and reference axis (toe in/out). The type of crossover used will also affect this decision. But none the less is always done at the crossover point whatever that may be.
The benefit of having them time alligned is that to obtain a phase-coherent system you won`t need to use a passive crossover that adds extra delay to the tweeter ( usually one order higher ) and allows the use of symmetrical 2nd order crossovers that would often comprise of a single cap on the tweeter. Negative is diffraction from the stepped baffle.
Measuring acoustic centers with a ruller or tape would easily produce an error that would be enough to ruin your phase tracking and you`d need to tilt the plane of the front baffle anyways. The project below, its 1 inch exactly so below the suggested values of 1 1/4 or 1 1/2 and was done after time-locked measurements of both drivers ( tweeter used as reference ).
As noted above by lilun, you`d achieve good phase tracking only on a limited spot, when you move around - polar picture would be different.
Measuring acoustic centers with a ruller or tape would easily produce an error that would be enough to ruin your phase tracking and you`d need to tilt the plane of the front baffle anyways. The project below, its 1 inch exactly so below the suggested values of 1 1/4 or 1 1/2 and was done after time-locked measurements of both drivers ( tweeter used as reference ).
As noted above by lilun, you`d achieve good phase tracking only on a limited spot, when you move around - polar picture would be different.
Attachments
This is true what you say to a degree - the cone surface is spread out over some area it is not a point source. I don't think it is a "true/false" kind of thing, however, so let me explain.
The amount that this "spreading out of sources" influences the relative phase angle (same issue as "time coherence") can be illustrated with this example: Let's say you have two drivers crossed over to each other with a second order crossover and the on axis response is "flat" and "in phase". It is very common to "check" the phase alignment by inverting the polarity on driver (e.g. the tweeter) and then looking for a deep null. If the arrival time, and therefore the phase, of the wavefront coming from different parts of the woofer cone was sufficiently spread out, nothing much would happen. But indeed you observe a null, often of around 20-40dB or more in some cases. This means that there is enough coherency in the combined wavefronts from the woofer to destructively interfere with the tweeter's wavefront, causing the null. This indicates that the variation across the woofer cone of the wavefront phase is not that significant. Is there perfect coherency: no. Is it totally incoherent: no. It's really much, much closer to a coherent source than not, however.
Similarly, if you move off axis it is well known that at high frequencies a lobing pattern will develop. This is simply due to the fact that the source (The cone) is spread out in space. Even if it was a perfectly rigid piston operating in perfect coherence, it is spatially spread out and thus off axis the wavefronts lose coherence. But remember that this only becomes an issue for frequencies having corresponding wavelengths that are "small" with respect to the diameter of the radiating surface. This is why people design multiway loudspeaker systems with smaller and smaller diameter drivers used to reproduce higher frequencies. Similarly, fullrange driver designers go to extreme lengths to tailor the cone so that only the central parts of the cone are radiating at higher frequencies. No one listens to their loudspeaker at 45+ degrees off axis, maybe +/-15-20 degrees, and it is easy to achieve sufficient coherence inside this window to make it a non-issue. So I say again: is it perfectly coherent everywhere: no. Is it sufficiently coherent: yes.
Minidsp users can use this easy method for time aligmnent (delay) setting, written by John Reekie here Refining a 4-way open-baffle speaker with the miniDSP 2×4
Time-alignment of drivers
You’ve probably seen speakers with a tilted-back baffle, advertised as being necessary for “time alignment” of the drivers. This is a more complex subject than simply getting the drivers’ acoustic centers on a vertical plane, but we can if we wish perform this kind of alignment digitally. Since the miniDSP has a time delay on each output channel, we can use that to effectively place the drivers on the same acoustic plane. (Note: this is effective only for the front radiation. I’m going to assume that getting it right out front is most important, and we will see how it looks out back once we’re all done.)
There are various ways of working out what to set the time delays to, but here’s a fairly quick and simple technique:
and AES E-Library The Acoustic Center: A New Concept for Loudspeakers at Low Frequencies
In practise, with minidsp or alike and a measurement system, we can optimize ta to any point in space. But the real issue is to find a compromise that works generally. With minidsp it is easy to "overcorrect" phase so that phase cycle is not the same any more!
We must remember that there are many factors influencing the phase match at the same time!
Time-alignment of drivers
You’ve probably seen speakers with a tilted-back baffle, advertised as being necessary for “time alignment” of the drivers. This is a more complex subject than simply getting the drivers’ acoustic centers on a vertical plane, but we can if we wish perform this kind of alignment digitally. Since the miniDSP has a time delay on each output channel, we can use that to effectively place the drivers on the same acoustic plane. (Note: this is effective only for the front radiation. I’m going to assume that getting it right out front is most important, and we will see how it looks out back once we’re all done.)
There are various ways of working out what to set the time delays to, but here’s a fairly quick and simple technique:
- Pick a pair of drivers, and place the microphone vertically mid-way between the centers of the drivers, and about a meter away.
- Turn off all channels except those two.
- Turn off the crossover filters for those two drivers.
- Set one driver to have a short (1 to a few ms, depending on the drivers) time delay.
- Run a sweep over a frequency range that is covered by both drivers.
- Look at the impulse response to determine how much the drivers differ from the specified time delay.
and AES E-Library The Acoustic Center: A New Concept for Loudspeakers at Low Frequencies
In practise, with minidsp or alike and a measurement system, we can optimize ta to any point in space. But the real issue is to find a compromise that works generally. With minidsp it is easy to "overcorrect" phase so that phase cycle is not the same any more!
We must remember that there are many factors influencing the phase match at the same time!
Tell me what is the movement of the voice coil divided by the wavelength then you will have your answer.
What I dont understand is that the wavelength at the crossover point say 3000hz, is a few cm. Yet even a few milimetres difference in positioning makes an audible difference. How do you explain this?
The biggest difference between physical time alignment and a digitally delayed time alignment is coherence off axis. The physical can work off axis, the digital will not. Tho I have pondered many an hour how to maintain on/off axis responses with this done.
In an experiment made an Orion like clone, but instead of a single driver up firing, used two side by side and the tweeter out infront of them ~5" on it's own baffle, 8"wide w/ 2" roundovers which roughly corresponds to the crossover frequency (slightly lower). This configuration allowed a far better time alignment off axis. As you move off axis the digital delay shifted from one driver to the other allowing a much broader listening axis. Dead on axis it is off to a small degree, but then not everyone chooses to listen only on axis and appeared to have little influence on the overall reproduction. The result of this experiment proved that a much improved off axis alignment could be achieved (within limits) digitally (MiniDSP) in an Orionesque design.
In all honesty, this alignment was a hit amoungst everyone that heard it. Room filled with music, all encompassing, regardless of listening position and we are talking far off axis, even at extremes, 60° or more. To pull that off diffraction must be a #1 priority. The only issue I have with this tho, is the same as the Orion has with it's non defined imaging at midrange and lower frequencies. Sometimes you've got to put down the pen and paper and do some real testing/listening. A little tweak here and there can make huge differences, sometimes not and other times prove you've face planted the ground.
In an experiment made an Orion like clone, but instead of a single driver up firing, used two side by side and the tweeter out infront of them ~5" on it's own baffle, 8"wide w/ 2" roundovers which roughly corresponds to the crossover frequency (slightly lower). This configuration allowed a far better time alignment off axis. As you move off axis the digital delay shifted from one driver to the other allowing a much broader listening axis. Dead on axis it is off to a small degree, but then not everyone chooses to listen only on axis and appeared to have little influence on the overall reproduction. The result of this experiment proved that a much improved off axis alignment could be achieved (within limits) digitally (MiniDSP) in an Orionesque design.
In all honesty, this alignment was a hit amoungst everyone that heard it. Room filled with music, all encompassing, regardless of listening position and we are talking far off axis, even at extremes, 60° or more. To pull that off diffraction must be a #1 priority. The only issue I have with this tho, is the same as the Orion has with it's non defined imaging at midrange and lower frequencies. Sometimes you've got to put down the pen and paper and do some real testing/listening. A little tweak here and there can make huge differences, sometimes not and other times prove you've face planted the ground.
Incurring an additional phase error is all. Use Edge (or similar) to model.
Time for you to hit the books. As I said before, read, read some more and read it again. Then build and measure a prototype. Something real experts do everyday.
- Status
- Not open for further replies.