How to ACCURATELY integrat subs with mains (and a little challenge heh heh)

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
hope this is in the right subforum, I hope so cause it also has a lot to do with loudspeakers as opposed to just subs.

Bit of a ramble I'm afraid, feel it's needed tho in order to explain. sorry

OK, been mucking around a bit with distributed subs, as per Harmon Kardon and of course a bit closer to home it is something Earl is a champion of.

All that is fine. I run a splitter off my bass outputs of the deqx and run the bass signal thru a behringer dcx 2496, which I'm sure most are familiar with. So the actual frequency response side of things is well and truly covered.

The problem I find (and have mentioned before in a few threads) is that (well I assume) because the subs vary in distance from the LP, both to each other and the mains, if not integrated properly then I can tell that the arrival time of the signal from the various subs do not exactly coincide with the signal from the mains. It's a very subtle thing, but both annoying and distracting.

That may also be complicated by the dcx having it's own internal processing delay.

OK, THE QUESTION IS, (caps lock, sorry) what is a good method for measuring the arrival times fo the sub signals, so I can set the appropriate delay on each of the subs to have them arrive at the same time??

What have I tried? Well, (can't find the link) I tried the 'trick' of putting the subs into opposite phase, and adjusting the phase whilst playing pink noise and looking for the biggest dip. Seemed a bit inaccurate (and listening confirmed I didn't have it right). Maybe that only works when the subs are low passed and the mains are high passed? As I run it off a splitter, the woofer is not high passed, so I'm just trying to fill in underneath.

I tried a variation of that, and just played a sine wave at the frequency of the sub low pass, and adjusted the phase whilst watching the spl meter.

BTW, what is the correct term I'm looking for? I think some of the confusion in the past when I explained this is that I may have been using the wrong terms. Just now I've been using 'phase'. I tend personally to think of it in terms of time and distance differences, but are they just opposite sides of the same coin??

I have hooked up long leads to the behringer so i can have it on my lap and twiddle while listening, varying the time (in terms of distance) but after a while it gets really confusing and frustrating...did I go the right way, too much, too little. You get the idea.

In another thread shinobiwan mentioned using audacity, which I downloaded and somehow got working. A thought occurred to me, if I stick the mic in the LP and played a signal thru the mains and the subs in turn, hopefully the captured recording will allow the delay (difference) between when the woofer does the signal and the subs do the signal, and adjust the unit accordingly.

That too somehow does not seem accurate enough, Admittedly I was only using an audacity generated signal (pluck I think) and am wondering if another signal would produce better results, and if so where to find it. I guess we would need a 'sharp' snap type signal that encompassed all the frequencies in question, hopefully something with a clearly defined leading edge we could read from the recorded trace.

Is there such a signal? Maybe audacity is not the correct program?

Or what other tricks do people use?? I really want to get away from trying to do it by ear and trial and error, it bugs me greatly when they are not in time (nothing to do with FR as I say).

Ahh, I mentioned a little challenge didn't I? In the other thread where shin mentioned audacity, he also mentioned that he was umm, 'discussed' on another forum by s/one named terrorizer, aka thylantr (I think)...hope I'm not wrong. will edit it if I am.

Hahaha, anyway I browsed around a bit and found MUCH to my amusement and surprise, that I TOO was ''''famous''''' and had (kinda ) been discussed on that site. (well, not me personally but my project, tho there was a few comments made about me personally if you get my drift).

http://www.createforum.com/petereur...days=0&postorder=asc&start=0&mforum=petereuro

The site seems on the whole to be quite antagonistic and belittling of quite a few in audio and fora.

Naturally I signed up and had a bit of a chat, and to date there has been no real hostility or antagonism shown towards me, which is good.

Anyway, I kinda gave THEM the challenge of helping me integrate my subs properly. Admittedly I did not give them as much data as I just did here, but a slightly sarcastic comment was made about 'why are not my 'pals' on DIYA helping me with the problem'.

Well, in part that was because I had never put the problem before my pals, heh heh, but it did motivate me to put it here and see what help I get.

Then I can compare the help offered here vs the help offered over there, and over there they are quite scathing about DIYA.

So that's the 'challenge' heh heh. Where will I get the most help? This forum or the forum that bags this forum.

I was quite above board about it, and to be honest quite a simple yet very good suggestion was made by terrorizer which I tired and it has helped me advance quite a bit (silly silly dumb errors on my part).

I hope it stays fun and lighthearted tho. Forums are sposed to be fun right?
 
terry j said:
OK, THE QUESTION IS, (caps lock, sorry) what is a good method for measuring the arrival times fo the sub signals, so I can set the appropriate delay on each of the subs to have them arrive at the same time??

Calculating gives a good approximation. The speed of sound is approximately 343m/s [1125ft/s). That means for each meter distance you can assume about 2,9ms of delay (~0,9ms/ft).
 
Very interesting Terry. I was just drawing up some sub ideas for just such an arrangement and my plan was to use a DCX 2496 in a similar way to what you've mentioned - up to 6 subs. I haven't tried to do the integration, but can't see why it can't be done, so I'll watch this thread with interest.

See you next weekend.
 
Re: Re: How to ACCURATELY integrat subs with mains (and a little challenge heh heh)

pacificblue said:


Calculating gives a good approximation. The speed of sound is approximately 343m/s [1125ft/s). That means for each meter distance you can assume about 2,9ms of delay (~0,9ms/ft).

yeah, I got close by that sort of method quite a while back, tho I don't recall using such accurate figures as yours. that may help a bit.

on top if that, is there an accurate figure for the delay thru the dcx as well?? If we want the absolute best accuracy, that will need to be taken into account to?? Idly wondering as I type, does ntha figure vary according to how much processing it does?? ie a moving target of sorts?

Brett said:
I haven't tried to do the integration, but can't see why it can't be done, so I'll watch this thread with interest.

See you next weekend.

Yeah of course I'm convinced it can be done, but right now I think I'll go back to just two subs...haven't got a lot of time to spend getting it right you know! It's not next weekend is it?? think it's two weekends away isn't it/? (tho you are welcome anytime as you know)

I really think any slight discrepancy really stands out because I'm using the deqx and the way the bass has been measured with it. It does such a good job that anything not up to the same standard really stands out.

It could be that everyone else is waaaay better than me in integrating subs, or maybe I'm more sensitive to it than most, but I do think nthe absolute quality that the deqx (using the new technique) gives to the bass has quite a bit to do with it.
 
Re: Re: Re: How to ACCURATELY integrat subs with mains (and a little challenge heh heh)

terry j said:
on top if that, is there an accurate figure for the delay thru the dcx as well?? If we want the absolute best accuracy, that will need to be taken into account to?? Idly wondering as I type, does ntha figure vary according to how much processing it does?? ie a moving target of sorts?

I am probably the only person in this forum who does not yet have a DCX2496, so I can only guess. Usually DSPs give you either the possibility to set delays in mm or in ms. The better ones have both options.

How exact does it need to be? For one thing you won't be nailed to a place without moving, so it is probably not even practical to concentrate on settings up to the millimeter. And which setting is correct, if there are several listeners in the room at different positions?

Chapter F on this site gives hints about how accurate delay settings need to be.

Processing time should not affect the DSPs performance.
 
Try using the DCX "auto align" function with the subs lo-passed and the mains on the hi-pass side. You may not want to leave the DCX in the main speaker signal path permanently, but it may help you identify the appropriate relative delay (it's not just distance, as the filters also play a role). You may find you need to delay the mains rather than the subs.
Paul
 
Keep in mind that just because you have the mains crossed LR4 via the DCX, they may not actually be LR4 due to the natural response of the speaker. If you have a box modeling program that will allow you to apply active filters to sim the effect, then I would find a crossover frequency (and slope) for the mains that will correctly cross to the sub at it's given crossover frequency. So you will have different crossover frequencies and possibly slope for the main and subs, but the actual acoustic response will be closer to ideal.

For example I am building a 5.1 HT package. I have a sub that I want to cross at 50hz. My HT receiver will apply a LR4 filter @ 50hz for the sub, and the actual response should follow quite close since the sub is pretty much flat to 100hz. Now for the mains, the "standard" based off THX specs expects the mains to be sealed with Qtc=.7 (2nd order Butterworth) and the F3 at the expected crossover frequency. In my case, 50hz. The receiver will apply an additional BW2 filter to the mains, resulting in a net response of LR4 slope that is 6dB down at 50hz, in other words a perfect match to my sub. Now since my mains are ported-and yours probably are too- they don't have a response that looks like Qtc=.7 with F3 of 50hz. But I started applying BW2 filters at different frequencies (in the sim program) to find one that would most closely follow this ideal slope. I found that setting the filter to 60hz made the mains behave quite closely to the ideal sealed slope.

Now your setup is a little different, but the basic idea is the same. This will at least get you ballpark correct. Now you have to adjust delays based on the varying distances of speakers. Ideally you would want to plot the response of each speaker and sub by itself, then check response with all of them playing (sine wave, pink noise, whatever) and look to see if they appear to be summing correctly.
 
yeah the dcx allows delay in metric and imperial, and also in terms of time. I mentioned earlier that I feel I could have confused some people in that past as I always mentioned this in terms of distance (as that was how I was using the unit) but I feel others though of it in time or phase or whatever.

Are these terms interchangeable??? If not, what is the essential difference?

Bit of further background, this is 2 ch only. Not into HT. That also means that I don't need (or even want) low bass. I HATE being shaken when listening to music!!! That may be fine for movies, dunno hate movies and don't watch them. In any case it's not what I want. That's fine, between the deqx and dcx I can tailor the extension to suit me.

OK the mains are three way, and the deqx is only three way. So in order to use the subs I run a splitter from the bass output of the deqx and pass that thru the dcx and amps to the subs.

That means the woofer in the three way is NOT high passed with the subs low passed (I do low pass the subs of course with the dcx, but I can't high pass the mains).

So it is a 'classic' case I guess of just adding a sub to fill in below a set of mains. I only need to add the subs now because I only recently moved my system orientation around (trying the article on the decware site about aligning the system to a corner).

In my old setup I could get extension to 29 hz, which is all I want (hate the shaking remember), but due to the new layout both the speakers and the chair are much more in 'open space' and I have lost a lot of extension that I can't get back with the mains no matter how much boost I apply (the woofers are 18 inch PHLs, can take a lot of power and boost). On the other side the advantages of the new layout (imaging, much less 'mud' in the vocals and midrange etc) mean I'd rather keep it this way and go to the extra hassle and complication of adding the subs back in to keep the advantages yet get the old extension back.

I have not tried the auto align function, it could work but my gut feel is that at bass frequencies it will not do too much good?? Still, worth a try.

All boxes sealed.

As a thought experiment, this is what I think. Lets say we have an appropriate signal (what would that signal be and where do I get it) and play it thru the system with a mic in the LP being recorded in say audacity (correct program??)

Lets assume that all the subs are at different distances from each other and the mains. Let's say that we have four subs.

Ok, when a bass 'thump' is played we record the results. By my reasoning we should see a 'thump' in the recording which comes from the two aligned mains woofers, then a succession of thumps from each of the subs in turn according to their respective distances. Actually that needs to be reversed.

The subs need to come earlier than the mains signal (as they are piggy backed off the bass output) so that using the delay function in the dcx I can delay the early arrival of the sub signal back to the woofer. Ie, if I delay the woofer in the mains via the deqx, it will also delay the signal to the dcx.

Anyway, the essential point is that with the appropriate signal and measurement program I should be able to see on the screen the actual time arrival of the signal from each source shouldn't I? The it would be a trivial matter to adjust to make them arrive at the same time.

The question is what signal and program is best suited to doing this?

Maybe we need to look at impulse graphs for example. Dunno.

Does any of this make sense?
 
"OK, THE QUESTION IS, (caps lock, sorry) what is a good method for measuring the arrival times fo the sub signals, so I can set the appropriate delay on each of the subs to have them arrive at the same time??"

Hello Terry

Do they have a adjustable phase controls?? I have 3 subs that I use in my HT. I used the phase controls and a test tone to set them the subs up for max output. Just used an RTA to measure level. I don't know how close you can get them time-wise but a a minimum they should all be in phase and summing properly where the listening seats are.

Rob:)
 
yes rob it does have, and yes I have also done that too:)

In fact when I mentioned using pink noise before with t,he subs reversed and looking for the biggest dip, I got that from an article somewhere (talking BOUT INTEGRATING THE GOTHAM SUBS) ****, caps lock, and they used it to get the correct phase.

I did similar, but found the pink noise to not be as sharp and cler as I hoped. I ended up doing it as you did, and used a sine wave at the -xover frequency dialled in on the behringer (the mains just fall off naturally remember) and adjusted the phase for max spl (a quick test I think that you have the right phase is to then click reverse phase and see if you get a large drop in spl. sometimes the max spl can be shown thru a large range of phase and perhaps that last trick helps you find a more exact value?)

So why do we have phase adjustments on subs? I always assumed it was because of the differences distance wise between subs and mains to the LP. If so, does that mean that the terms phase and distance are talking about the same things??

I mean maybe it is, for example the wavelength of 50 hz is (guessing for illustration) say ten feet. Would that not mean being 180 degrees out of phase be the same as dialling in 5 feet???

That is the type of conceptual understanding I'm trying to get as well as (hopefully) a practical method of actually measuring the arrival times.
 
My understanding is that, yes, the terms are somewhat interchangeable, provided that the time delay you're talking about is insignificant compared to the impulse time of the signal you're playing, ie. since to travel 10 feet you're talking about all of roughly 9ms, if the music tends to be doing the same thing for more than 9ms at a time, then you're unlikely to notice. In bass content, you're particularly unlikely to notice, since it doesn't tend to be as distinct to start with.
 
some thoughts...

... the fact that you can hear different arrival times indicates that alot of high frequencies go through your subs. try a steeper and/or lower crossover. distortion could be an issue.

... since the dcx has only parametric eq, a fir eq based on a measurement of the rooms frequency response will yield better results. this is indicated in one of the toole papers, where he shows that 1/3rd oct eq is doesnt help to reduce ringing, because its resolution is too inaccurate. similar, i would guess its rather difficult to tune your parametric eq's amount and q exactly to the room modes, leaving you with little ringing spikes.

... using 4 subs, one on each wall midpoint gives the best frequency response with good low frequency support.

... you could try moving around. there could be a sweet spot, where the subs integrate well. then you can derive the delay from the distance to your normal listening spot.

as for my perspective, i use 4 subs and even while moving around the room i cannot hear integration issues. maybe i am blessed with bad enough ears :)
 
"I mean maybe it is, for example the wavelength of 50 hz is (guessing for illustration) say ten feet. Would that not mean being 180 degrees out of phase be the same as dialling in 5 feet???

That is the type of conceptual understanding I'm trying to get as well as (hopefully) a practical method of actually measuring the arrival times."

Hello Terry

I would be thinking along the same lines. My active set-up is in the analog domain as far as crossovers and EQ are concerned. My mains are active 4 ways. They are really sub assited 3 ways as I cross the subs in in the 50Hz range.

They are in my main set-up which does double dutty for stereo and HT. My HT preprocessor has Audyssey installed and when I run the set-up I get the same distance for L/R and Sub so I know that using the peak null as you discussed worked fine for me. The distance settings as bassed on the measured delay from the set-up microphone.

If I read the spec sheet on the Behringer right they are saying 1ms of latency. I not sure why you are having an issue integrating them. Why don't you throw up a picture of your room if you don't mind?? Might help. With the behringer you can certainly chose some very steep slopes. I use 24db LR and it works fine as far as keeping program content out of my subs.

Rob:)
 
OK the mains are three way, and the deqx is only three way. So in order to use the subs I run a splitter from the bass output of the deqx and pass that thru the dcx and amps to the subs. That means the woofer in the three way is NOT high passed with the subs low passed (I do low pass the subs of course with the dcx, but I can't high pass the mains).

Understood...so temporarily run the normal line level input of the deqx to the input of the dcx. Operate the sub out of the dcx normally. Take another output of the dcx to the input of the deqx for the mains (as I recall, dcx outputs can be used with no high or low pass filters).

With this temporary hook-up, run auto-align to establish the relative delay between subs and mains. Consider dcx internal delay (I recall someone measuring it at 1ms) when setting up your regular configuration and you should be good.

My system is in boxes at the moment, but I have used the dcx auto-align for sub/main integration and it worked well.
Paul
 
here is a post made by Earle Geddes from the 'setting up the nathan 10' thread.

It could very well apply to the question of accurately setting up subs, at the least it has data that could help others.

The main thing that you have to do now is to balance the subs with the room and the mains. I'll describe how I do this. The LF sound field in a small room has tro be considered as a steady state situation - there is no direct field in the time sense as the signals have arrived from multiple directions even before a few cycles of the LF signal have occured. The ear is simply going to react to this by considering this a steady state situation. Now in the steady state we have to deal with the random problem of sound measurements to get at the mean value of this steady state sound field. This is best done, or has to be done, by spatial averaging. So first setup the mains and the nearest sub. Set your spectrum analyzer to a very low bandwidth but not less than 200 Hz. and fairly long averaging time. This will yield a very long average of the sound signal. Take your mic and move it through large spatial positions near but arround the prefered listening position. Be carefull as small bumps of the cable can generate large eroneous signals into the mic. The sweeping has to be smooth. When the analyzer has completed its run you will have a plot of the frequency and spatial averaged LF sound field. Try adjusting the sub - never adjust the mains - to see if you can get a better response by changing the gain, the low pass point, the phase and or delay if you have it. The bass should be sagging slightly at this point since you will be adding in two more subs. Now add in the next closest sub and repeat the entire procedure again. You should be able to improve upon what you had before. If not try turning off the first sub and setting the second one optimal and then add in the first sub. If you still can't get a better response with two subs than with one then you will need to move one of them. You need to get an improvement from the second sub or something is wrong. Now repeat this process with the third sub. The third sub, when you are close to it should barely be audible. It it is loud, then once again, something is wrong. It should only be filling in holes at this point and not adding any actual level. The level after two subs should be about flat or possibly a slight rise - based on preference. I find a few dB rise at the low end is desirable for best effect. With the three subs things should be quite smooth when spatially averaged. You can now use any EQ that you have to make a final flattening of the response, but these bands should never be more than a few dB. When you are done, if things go as they usually do for me, your should have a spatial average of about +- 2-3 dB. This can take several hours so don't be impatient. But once you have it you will never change it. Make sure and write down all the settings!! Two people helps - one to sweep the mic and another to run the analyzer. Sweep the mike vertically as well as horizontally, but in a wide ellipse. Its not necessary to repeat the same pattern, its only necessary to NOT leave the mic in a stationary position. You can try wider sweeps or narrower ones, but the bigger sweep will likely be better. Hope this helps. Good luck.
 
All righty then. Have had a few interesting experiments lately.

I mentioned earlier that I had hooked the behringer up with long leads so I could fiddle on the fly whilst listening to music in the chair. Turns out a very wise and educational move.

I won't delay the suspense any longer.

I think I have been barking up the wrong tree, and for not a short while I might add.

Basically using the front panel you can (obviously..duhh terry) adjust the parameters of the subs.

Sp there I was changing the delay on the subs (and can turn on and off).....and I simply could not hear any difference between a sub(s) delayed or not. And I'm not talking a few measly milliseconds either.:xeye: I'm talking meters!

I first did a couple of nights ago, but thought I'd better check my results when I didn't have a few in me. Same result.

There you go everyone, normally I'd say sorry for 'wasting' your time, but I'm actually quite stoked I've learnt something and for all I know it may be useful to you too.

Funny how you get an idea in your head and run with it. Now that does NOT mean I don't hear/feel something that doesn't seem to gel, I still do. But I now know it has NOTHING to do with slight differences between the mains and subs and listening position.

So what in the heck is it??? Starting to think old house, wooden floors,.....structure bourne vibrations??? ie the subs in different positions cause changes in the floor vibrating (odd room layout too btw) and perhaps those vibrations reach me at different times?

all conjecture.

I bet now that I've ruled out that thought in my head it will cease to be an issue!!

No wonder people hear differences with mingpo discs and hallographs heh heh, the mind is a powerful thing.

cheers all, thank you very much for trying to help, hope at least there is now something here as a reward for your efforts.
 
For me, that is a comforting turn of events, since i never could hear subtle changes in sub delays.

As to noise from structural vibrations... Have you ever tried a slow sine sweep with your subs, say from 20-200hz in maybe 1-2minutes? This will clearly show every resonance in your furniture. Since they are triggered only above a certain SPL, do the sweep rather loud.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.