Measuring time alignment

N

nzlowie

New member
Hi all

First time here so please let me know if this isn’t the correct place to ask this question…

I’m looking for a simple way to time align speakers and had a thought about 2 different frequency impulse’s on the same sound track, possible?

I thought if you could record 2 different frequency impulses at the frequencies to suit the speaker drivers you want to compare then play this over the system and measure the result with something like Room EQ Wizard. If we had a 100hz (will be played back through the woofer) and a 2.5khz (playback through the midrange) impulse on the same file and measured the playback result we’d see 1 impulse if the time alignment was perfect, 2 pulses if they were out of alignment.

Or can we record the 2 different frequency pulses starting at exactly the same time but on different channels of a stereo file. Woofer connected to one channel with the 100hz pulse and the other channel connected to the mid range with the 2.5khz pulse.

Will this work?

Many thanks.
 
Just re read my post and it doesn’t really read as a home recording question, if you forget about the why it sort of is…

So can we record the 2 impulse scenario’s I’ve talked about? I don’t have this sort of software installed so that was going to be my next request, if it is possible can someone please do this for me??

Once again many thanks for your thoughts and time.
 
You can generate the frequencies at almost any pulse length you desire in almost all DAWs. I would use Samplitude (Pro X Silver) but you could use Audacity, probably Reaper.

Paste the two pulses into a pair of mono tracks (not stereo IMHO) and then you can check (in Sam at least) that they start and stop exactly on the dot.

Now all you need it a pair of mics and an AI. I would say my AKG P150s would be PERfek!

But why? I thought all that 'time alignment, phase coherence' diddly do-dah was 80s hi fi bollocks now? Some fekking WEIRD looking speakers back then!

Dave.
 
The important thing with time alignment is that with a scope and a signal source, then you can see the delay, and by moving the speakers, you can align an HF driver with a LF driver - just record the two different pulses, and use whatever crossover system is used to split them out. You could record these on separate tracks and then route them with bi-amping to the two drive units - BUT - this would not reveal any delays introduced in the crossover.

As Dave says though, why bother. Tannoy went for it big time with their speaker drivers for years. Other speaker systems didn't, and the results were down to personal preference. With modern speaker systems where we are using many cabinets (line arrays, for example), the concept requires driver output to overlap, so any attempt at time alignment only works in one direction. Time alignment in terms of PA delays makes sense for the listeners, but in studio size spaces, I'm not sure there's really any benefit I have ever noticed.
 
Well, we recall how painful it was watching KEF and Bowers go through these gyrations : )
 

Attachments

  • Walsh_f.jpg
    Walsh_f.jpg
    348.2 KB · Views: 3
The only frequency that matters when electronically time aligning a woofer with the next bandpass up is the actual crossover frequency. Being out of phase at the crossover will yield a drop in level, a dip at that frequency. One driver having inverted polarity will cause a deep hole in the response, the details of which will depend on the phase (phase and polarity are related but different). The crossover filters themselves will be creating phase shifts that make it impossible to match the two drivers even a little off the crossover frequency, but the level will be diverging quickly so that's a minimal problem. So, one way to do it is to generate a sine wave at the crossover frequency and do your measurement there. You could also do a sweep through the crossover frequency. Each driver on its one should be 3dB down at that frequency and the response should be about equal to the 1kHz level with both drivers on.
 
To expand a bit more on why I want to look at this here is my project.
I have a folded bass horn below a mid-range horn, the bass horn has eq via a minidsp. Due to the nature of the folded bass horn and the latency of the minidsp the acoustic start point (?) is potentially a lot further back from where the acoustic start point of my mid horn is. I’d really like to know this distance….

If I played both frequencies mentioned, the 100hz would go through the minidsp, amp and horn – the 2500hz (or whatever) would go through the amp, crossover and horn. (Separate amps for bass & HF) This difference between arrival times would tell me how far I need to move one, the mid horn is movable.

Hopefully this makes sense.

Cheers
 
Go for it ! I'd set up any old MIC at the approximate listening position and run some pulse files. Room EQ used to be pay for resolution, and I suppose it could be worthwhile ? I use a software 'scope called Visual Analyser
 
nzlowie said:
To expand a bit more on why I want to look at this here is my project.
I have a folded bass horn below a mid-range horn, the bass horn has eq via a minidsp. Due to the nature of the folded bass horn and the latency of the minidsp the acoustic start point (?) is potentially a lot further back from where the acoustic start point of my mid horn is. I’d really like to know this distance….

If I played both frequencies mentioned, the 100hz would go through the minidsp, amp and horn – the 2500hz (or whatever) would go through the amp, crossover and horn. (Separate amps for bass & HF) This difference between arrival times would tell me how far I need to move one, the mid horn is movable.

Hopefully this makes sense.

Cheers
Click to expand...

Thing is, because of the nature of the eq, crossover, drivers and cabinets, the delay at those two different frequencies won't accurately represent the delay at the actual crossover frequency, and the actual crossover frequency is where it matters.

REW should be able to tell you what the delay is at that frequency from each source. Perhaps set up a two octave sweep with the crossover frequency in the middle, run it through one cabinet at a time with a mic at the same distance from each (fairly close to isolate the direct sound from room reflections). That should give you the information you need to figure how far to set back the high cab from the low one.

That said, I would be tempted to do it by ear, starting with the two cabinets on either side of me at the same distance, and running a short looping pulse through both cabinets at the crossover frequency. I'd move the high cabinet back until the pulse seemed to be arriving at the same time. At lower frequencies it may not be possible to separate the two cabinets, but I'd still give it a try.
 
Set up two mics* at say 1 mtr. Play a burst of noise through both speakers and record the result in a DAW on two tracks.

Observation of the two resultant tracks will give the time displacement. Bit 'O' maths turns mSecs to mm or (archaically!) inches.

*Don't even have to be the same type or even same principle so long as the diaphragms are vertically aligned.

Dave.
 
ecc83 said:
Set up two mics* at say 1 mtr. Play a burst of noise through both speakers and record the result in a DAW on two tracks.

Observation of the two resultant tracks will give the time displacement. Bit 'O' maths turns mSecs to mm or (archaically!) inches.

*Don't even have to be the same type or even same principle so long as the diaphragms are vertically aligned.

Dave.
Click to expand...

I was thinking of that, but then I'd just use my ear method to save me the trouble of setting up the mics. Either way would get it close, but REW would get it right (if you interpret the results correctly).

Personally, I'd rather delay the high cabinet electronically so they could be aligned physically. It would them phase coherent off axis, with the added bonus of being more space efficient.
 
Thanks guys
I'll try a few things over the weekend..

If possible can some please record a small .wav file with the 2 frequencies starting at the same time?

Cheers
 
You must log in or register to reply here.
Back
Top