Delay help please |
Post Reply | Page 12> |
Author | ||
Semtek_soundsystem
Registered User Joined: 24 February 2016 Status: Offline Points: 11 |
Post Options
Thanks(0)
Posted: 12 September 2020 at 6:52pm |
|
Hi, I've recently purchased some labhorns and I own some usbs and mt1281's, I'd like to set the delays as best I can without smart etc, I used to just use horn length when using the usbs and Mt's, but I can remember about using a test tone at the crossover frequency and inverting phase and addingg delay untill you cause a cancellation,
My query is, when using a test tone, it is a sound wave which has hundreds of peaks and troughs (if you look at it as a wave) so how do I know I have caused phase cancellation at exactly the right point and not two differnet Peak Points in the WAVE? IF YA GET Me? Or Is there a trick I'm missing? Thanks, please don't make me try to elaborate, my paint skills are whack :) Cheers Edited by Semtek_soundsystem - 12 September 2020 at 6:53pm |
||
Darkstar
Registered User Joined: 08 October 2014 Location: Italy Status: Offline Points: 326 |
Post Options
Thanks(0)
|
|
Test tone must be a pure sinewave at the crossover frequency.
You will need a microphone you can trust for the measurement. This will hardly ever result in a correct phase alignment across all frequencies, however, you will only be sure to be phase aligned at the crossover's. |
||
Bass =/= Enough
|
||
Earplug
Old Croc Joined: 03 January 2012 Location: Europe Status: Offline Points: 7199 |
Post Options
Thanks(0)
|
|
"You will need a microphone you can trust for the measurement." With the above method, you don't actually need a microphone! Just a decent signal generator. |
||
Earplugs Are For Wimps!
|
||
Semtek_soundsystem
Registered User Joined: 24 February 2016 Status: Offline Points: 11 |
Post Options
Thanks(0)
|
|
Thanks darkster, but although it would be phase aligned at crossover it could be poorly time aligned still, it just doesn't make sense in my head. Imagine these arrows are up and down cycles of a sinewave <><><><><><>
As long as any of those positive or negative cycles are allinged it will cause phase cancellation, but they may not be the same positive or negative peak so will not be time aligned correctly,?
|
||
Semtek_soundsystem
Registered User Joined: 24 February 2016 Status: Offline Points: 11 |
Post Options
Thanks(0)
|
|
Thanks earplug I could use a dB meter to see the best results, maybe over thinking and just need to try it, maybe using a song as source and only playing a narrow band at crossover point would be better suited than a single constant sine wave? Cheers
|
||
Sonic the hedge
Registered User Joined: 12 May 2020 Status: Offline Points: 161 |
Post Options
Thanks(0)
|
|
You are absolutely correct, so do the time alignment FIRST, using the same tone at crossover/invert method. You do need a dB meter IMHO but it doesn't need to be anything calibrated because it's all relative. Even a phone app will do the job fine. Use a DB meter and adjust the time delay to the point of maximum cancellation, which you will find to be quite a wide band. Find the edges of the maximum cancellation band i.e. the points where the dB level starts to rise, then set the delay to the point halfway in between. Once this is done the phase match point is much easier to find again using a dB meter. If you have speakers with very nonlinear phase response e.g. anything that does not have infinite baffle - you can also then try shifting the test tone a few Hz either side of the crossover to see if you can find a better phase match. Ultimately phase matching in this situation is a compromise so just try to find the point that sounds best. Obviously using Smart or similar you can compare the actual phase plots and find the best (frequency) point to align. Not sure if this is a correct method but it works for me anyway.
Edited by Sonic the hedge - 13 September 2020 at 4:32pm |
||
DMorison
Old Croc Joined: 14 March 2007 Location: Aberdeen Status: Offline Points: 1647 |
Post Options
Thanks(0)
|
|
Sonic has answered this in part, but the other thing to bear in mind when using this technique is that you are looking for the shortest delay time that achieves the cancellation. Say for example, we're using a 100Hz crossover. That means each wave takes 10 milliseconds to complete. So, if we do the "invert polarity > find delay that gives max cancellation" trick and the delay time is 13 milliseconds, what does that mean? Well, it means that as you've worked out, we're already at least one extra wavelength out. So, reduce the delay to 3 milliseconds, the cancellation should still be as effective, but we're now in closer overall time alignment. It's also worth doing a reality check or two to help ensure the best results as well. Definitely do the check at multiple frequencies eg 80, 100 & 120Hz assuming a target 100Hz crossover - see how close they are; you may find better overall sound with a delay time that is the average rather than just the one from the exact crossover freq. Also, even though you may have a strong idea of which cab needs the additional delay, try it the other way round just in case. Again, because the sine wave allows the possibility of being a whole number of wavelengths out, you may find delaying the other cab allows you to achieve a lower delay time, which would get you in closer overall time alignment. HTH, David.
|
||
Sonic the hedge
Registered User Joined: 12 May 2020 Status: Offline Points: 161 |
Post Options
Thanks(0)
|
|
@David
Interesting approach and entirely valid. Shifting the delay by one period will afford a workable alignment with less overall delay. My concern with this would be the possible effects on the impulse response. By shifting the alignment deliberately one period out, any impulse e.g. a kick drum, will be left overall one cycle wider, and the critical leading cycle will be reduced by around 3dB as the cycle will only be produced by one driver (Sub/Kick etc) instead of summing correctly. The amplitude of each cycle across the entire note envelope will be similarly mismatched. IMHO this discernably reduces percussive impact. I think perhaps this is one reason why reflex (or any open rear loading) does not sound as 'tight' or fast as IB because the enclosure is achieving a phase invert acoustically to align & combine the energy from both sides of the driver. This has the effect of smearing the overall waveform presented to the listener. Not sure if I have that correct but it makes sense and experience tells me that's how it works. Of course longer delays can cause other issues e.g. with DJ or artists timing, so as ever it's a case of deciding the best compromise between exact alignment and minimum delay.
Edited by Sonic the hedge - 15 September 2020 at 12:27am |
||
Darkstar
Registered User Joined: 08 October 2014 Location: Italy Status: Offline Points: 326 |
Post Options
Thanks(0)
|
|
How are you going to find out if you're canceling? Db meters are microphones and you do need one, the more consistent the better.
|
||
Bass =/= Enough
|
||
Earplug
Old Croc Joined: 03 January 2012 Location: Europe Status: Offline Points: 7199 |
Post Options
Thanks(0)
|
|
Ok, point taken - but you are taking a relative measurement, so best quality mic/meter not needed - even one of those signal analyser apps should do in this case. If you were doing it properly, ie SMAART, etc, then yes - a good mic is essencial. |
||
Earplugs Are For Wimps!
|
||
Darkstar
Registered User Joined: 08 October 2014 Location: Italy Status: Offline Points: 326 |
Post Options
Thanks(0)
|
|
I said it as, in my experience, if you use a generic microphone without time-averaging the measurement it is really hard to understand wether you're cancelling or not due to the wildly inconsistent numbers. Although, now that I think about it, being able to average out the numbers alone should get the job done even with a cheap mic. Edited by Darkstar - 15 September 2020 at 11:12am |
||
Bass =/= Enough
|
||
Darkstar
Registered User Joined: 08 October 2014 Location: Italy Status: Offline Points: 326 |
Post Options
Thanks(0)
|
|
As Sonic said that is correct, you will be time and phase aligned at crossover frequency but it won't guarantee full band phase alignment (unless you have a linear system). That was my point. You also have to keep in mind that any changes in the EQ or output level will alter the crossover point and everything that goes with it. |
||
Bass =/= Enough
|
||
Post Reply | Page 12> |
Tweet |
Forum Jump | Forum Permissions You cannot post new topics in this forum You cannot reply to topics in this forum You cannot delete your posts in this forum You cannot edit your posts in this forum You cannot create polls in this forum You cannot vote in polls in this forum |