nzlowie Posted July 13, 2016 Posted July 13, 2016 Well.... thought I'd worked out how to measure time offsets on my horn project, but I'm not sure its right. The distance from the bass driver through the horn looks about right at 3ms or 103cm but the tweeter diaphragm measures around 60mm in front of the mid driver diaphragm but when I did the impulse alignment it needed 2.5ms or 86cm. I know physical measurements don't always align with acoustic measurements but this seems like a big difference.... I would find it easier to understand if it was measured from the mouth of the mid horn not the driver diaphragm. The mouth's of the bass horn and mid horns are in close alignment which would go against the above . Have I got it right? By the way...... sounded OK! Cheers Dave Sent from my SM-T800 using Tapatalk
leo Posted July 13, 2016 Posted July 13, 2016 I would use the FM hiss test.Slide the "tweeter" back and forth until the hiss sounds like one uniform homogeneous presence.That will be the correct distance.The risk is you could go mad.
davewantsmoore Posted July 14, 2016 Posted July 14, 2016 The time delay of the sound is affected by the frequency response shape. Measuring the distance to calculate the delay only works if the frequency reponse is flat .... if the speaker driver (all of them do) has a high pass and a low pass response then it doesn't work. Horns have more of a rolloff than regular speaker dirvers, so it's more of an issue. First: Apply the EQ and crossover filters to each of the bass and treble so they each seperately have the frequency response you want Do this before you consider aligning the drivers. Otherwise, if you adsjust the frequency response later, you may affect the "time alignment". High level process: Record a measurement of the bass Record a measurement of the tweeter (without moving the microphone) View the two responses overlaid together and work out what delay is required. There's a few different ways to go about that. Plenty of tutes out there.
Happy Posted July 14, 2016 Posted July 14, 2016 (edited) What happens when you move your head forward or sideways by say 30 cm? The time alignment gets spoiled? Sorry sidetracking Sent from my iPhone using Tapatalk Edited July 14, 2016 by davidro
Newman Posted July 14, 2016 Posted July 14, 2016 Have you posted the impulse alignment plot elsewhere?
davewantsmoore Posted July 14, 2016 Posted July 14, 2016 (edited) What happens when you move your head forward or sideways by say 30 cm? The time alignment gets spoiled? It is an issue for most multi-driver speakers. The damage is based on number of wavelengths .... so is more of an issue at higher frequencies, where small distances = larger # of wavelengths.... and more of an issues the larger the number of wavelngth the sources are separated by. Depends on how much the amplitude and phase of each driver changes when you move. THIS is why the speaker response at different angles (the "directivity") is usually said to be the most important indicator of performance. If you move, and the response of the speaker doesn't distort (then there's no problem) ..... if moving the listener, causes the response to change - then not only might you have an undesriable frequency response, but like you've noted, the time alignment may also go out of whack. Edited July 14, 2016 by davewantsmoore 1
Happy Posted July 14, 2016 Posted July 14, 2016 @@davewantsmoore thanks....know if Dirac Live provides something similar when it does the EQ?
davewantsmoore Posted July 14, 2016 Posted July 14, 2016 No. It operates on a "finished speaker". For something to affect the delay between two driver.... it must be sending a seperate signal to each driver (so it can delay one of them). This is a potetnial problem with "room correction" devices which operate on a "finished speaker". They can't easily tell if there is a time alignment issue between two drivers (becuase they're only measurgin the combined response) .... and even if they could, they have no way to fix it.
Happy Posted July 14, 2016 Posted July 14, 2016 Ok that makes sense thanks Sent from my iPhone using Tapatalk
Doncentric Posted July 14, 2016 Posted July 14, 2016 Hi Dave, Could it be there is a time delay caused by the tweeter amplifier?
nzlowie Posted July 14, 2016 Author Posted July 14, 2016 Hi Don, both the mid and tweeter amps are the same..... I did the eq with some time delay set, maybe I'll set it to zero the redo the eq. As this time alignment changes the eq i thought you would have the drivers aligned before it was eq'd. This would be the physical just like building a stepped baffle. Yes sound does change a bit when you move, but then a few horns I've listened to do as well. What a learning curve I've ventured into........ close to saying enough! Sent from my SM-T800 using Tapatalk
davewantsmoore Posted July 14, 2016 Posted July 14, 2016 As this time alignment changes the eq It's the reverse. The EQ (may) change the time alignment ..... by bending the frequency response, you are also bending the time response (phase). If you have an 'active multiway speaker' (like you do) .... where you have the ability to add delay to a driver. Then the time alignment process is basically the last thing you would do. In a speaker with a passive crossover filter .... any delay you want for a single driver, needs to be considered from the very start.... either by physically locating the driver (like on a stepped baffle as you mentioned) ..... or by designing a crossover filters that bends the frequency AND phase (time) response of the driver to just where you need it.
Recommended Posts