After reading about DSP crossing I started thinking more about time alignment. What sort of equipment and software can be used to test time alignment of a multiway design? Honestly I was kind of surprised that even with DSP assisistance that they still were not perfectly aligned and made me question some of my practices as well as commercial designs.
Figured proper placement of the drivers was how to get the time alignment correct and kinda had my own way of doing it. I would like to build a set up where I can run tests on time alignment and experiment with it more other then just using my ears.
Is perfect time alignment something achievable in a multiway design or it is only achieved with the use of a full range driver?
Figured proper placement of the drivers was how to get the time alignment correct and kinda had my own way of doing it. I would like to build a set up where I can run tests on time alignment and experiment with it more other then just using my ears.
Is perfect time alignment something achievable in a multiway design or it is only achieved with the use of a full range driver?
This done when you develop the crossover. Phase and "time delay" are related. Both the raw drivers AND the crossover filters have a phase response. You need to measure all of these to know what is going on, and then accommodate that in the crossover design to the extent that you would like.
All that stuff about "time aligned' driver placement is helpful, but I have seen too many miss the point that the crossover filters and the drivers (because the act like a bandpass) also have some influence on the time alignment. You have to approach it as a system.
In any multiway system you can only have time alignment in (at best) one plane (parallel to the floor), which designers typically locate as passing thru the tweeter or in between tweeter and midrange. At worst that is only true at one point in space. As you move up above that plane or down below it the distances to the various drivers in the system are changing relative to one another and therefore by definition the time delay due to travel of the sound thru air is changing - some reach the listening position sooner and some later. There is no getting around that.
Anyway, I would not get too overly obsessed with time alignment. The ear and brain are not like a microphone. The brain has evolved to make sense out of lots of confusing sound environments and as long as you stay within some reasonable bounds for time coherence it's all more or less the same.
All that stuff about "time aligned' driver placement is helpful, but I have seen too many miss the point that the crossover filters and the drivers (because the act like a bandpass) also have some influence on the time alignment. You have to approach it as a system.
In any multiway system you can only have time alignment in (at best) one plane (parallel to the floor), which designers typically locate as passing thru the tweeter or in between tweeter and midrange. At worst that is only true at one point in space. As you move up above that plane or down below it the distances to the various drivers in the system are changing relative to one another and therefore by definition the time delay due to travel of the sound thru air is changing - some reach the listening position sooner and some later. There is no getting around that.
Anyway, I would not get too overly obsessed with time alignment. The ear and brain are not like a microphone. The brain has evolved to make sense out of lots of confusing sound environments and as long as you stay within some reasonable bounds for time coherence it's all more or less the same.
Thanks bud. What would be a good set up for testing? Would you say the best design for an "array" would have a single point? I get what you are saying about the planes - so I am thinking something like a unity horn or synergy horn may be the best way to accomplish fixing the issues with a vertical plane since all sound would be radiating from a single source.
If you want to go that way, the unity horn concept is hard to beat. They are not especially small and are moderately deep.
For testing I make all my measurements using ARTA on a laptop with a USB pro-audio interface but there are many ways to get audio in and out of the computer these days.
I wrote my own tools for active crossover design that are freeware. Others are likely available.
Hardware: ECM8000 Behringer mic, with a calibration file downloaded from their site. Behringer UCA202 sound card, using it's Line In. And a mixer, with phantom power, needed for ECM8000 mic.
Software: I used HOLMImpulse to measure the delay between drivers, looking at impulse response.
Drivers with no crossover applied, and sweep range (I did not used sweep, but MSL) was overlapping in the crossover region
Then used REW, looking at Spectrogram window, and SPL, and fine-tuned the delay.
System is active 4 way, crossover using EqAPO.
The difference between a 'so so' and an aligned system is substantial. I can't believe what I was missing.. On good records, the stage is there, in front your eyes
Software: I used HOLMImpulse to measure the delay between drivers, looking at impulse response.
Drivers with no crossover applied, and sweep range (I did not used sweep, but MSL) was overlapping in the crossover region
Then used REW, looking at Spectrogram window, and SPL, and fine-tuned the delay.
System is active 4 way, crossover using EqAPO.
The difference between a 'so so' and an aligned system is substantial. I can't believe what I was missing.. On good records, the stage is there, in front your eyes
Last edited:
I am on your website and will be downloading your tools soon. Thank you so much for your contributions. I also checked out the link for the CPU based XO. Have you had any experience with it on the RPI? What is the best ADDA available for SBCs preferably with balanced connections? I know what I paid for my pro audio interfaces and the cost for some of the i2s applications seem to be so much cheaper I cant imagine the ADDA conversion being that stellar.
You need to look up relative acoustic offset.
Here is a link to one of the best guides on how to do it. It is written by Jeff Bagby. He references his own passive crossover design spreadsheet, but it works just as well with other capable software.
How To Find The Relative Acoustic Offset In PCD
And here's one by David Ralph Finding Relative Acoustic Offsets Empirically
Here is a link to one of the best guides on how to do it. It is written by Jeff Bagby. He references his own passive crossover design spreadsheet, but it works just as well with other capable software.
How To Find The Relative Acoustic Offset In PCD
And here's one by David Ralph Finding Relative Acoustic Offsets Empirically
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.