Todd
Agreed, but I guess that begs the question of why one would want to spend a lot of time working on an atypical situation? I guess that I just assumed that you were looking at low damped rectangular rooms because you thought them more common. I can also see where you might have done this in the begining because it was the most numerically available (don't want to say "easiest"). I had stopped seriously looking at rectangular rooms more than 25 years ago - to wit my thesis ("Low Frequency Sampling Statistics in NON-Rectangular Rooms"). I find the facination with rectangular rooms (axial, tangential & oblique modes, symmetric sub and listener placement, etc.) as being a real drawback to understanding LFs in real rooms, which are never rectangular when you consider that they are filled with furniture.
Agreed, but I guess that begs the question of why one would want to spend a lot of time working on an atypical situation? I guess that I just assumed that you were looking at low damped rectangular rooms because you thought them more common. I can also see where you might have done this in the begining because it was the most numerically available (don't want to say "easiest"). I had stopped seriously looking at rectangular rooms more than 25 years ago - to wit my thesis ("Low Frequency Sampling Statistics in NON-Rectangular Rooms"). I find the facination with rectangular rooms (axial, tangential & oblique modes, symmetric sub and listener placement, etc.) as being a real drawback to understanding LFs in real rooms, which are never rectangular when you consider that they are filled with furniture.
gedlee said:
Well, I'm not sure I would say atypical. I mean, I think there are many rooms which are tending towards rectangular and low absorption, though not perfectly so. So, it may still be usefull to treat them as such, as a first approximation. and yes, it is easier to deal with, so a good place to start. I agree that non rectangular rooms are the most interesting, but particularly when ther is NOT a lot of absorption. that is where the worst problems are liekly to be, and thus more interesting to me. Anyway this is why SFM was developed.
breez said:
I created my own optimization, not Harmans. The customer needs to record a noise signal at several seats, with the signals coming from each source. The waveforms can be played on a CD or whatever, and the mic can be any old mic since this is only LF stuff and absolute calibration is not necessary. I takes this data and calculate the optimum settings for a DCX2496.
We looked at making our own box to do this, but in the end the DCX2496 option looked the best since there is no way that we could compete with that price. I've done several rooms like this now and it appears to work well.
Earl, could you walk through the way this sub implementation would be expected to work?
Customer buys 3 (or however many) bandpass subs from you.
You build them and send them.
Customer buys dcx2496 (from you?)
Customer sets them up somewhere in his room.
Customer plays a signal and records results with ? (what program?)
They send the results to you, you send back optimum settings for the dcx2496
Is that roughly how it is expected to go?
-Tony
Customer buys 3 (or however many) bandpass subs from you.
You build them and send them.
Customer buys dcx2496 (from you?)
Customer sets them up somewhere in his room.
Customer plays a signal and records results with ? (what program?)
They send the results to you, you send back optimum settings for the dcx2496
Is that roughly how it is expected to go?
-Tony
gedlee said:
I expect that people will record a lot of useless data. The problems with inadequate hardware (mics and soundcards), software and noisy environments are incalculable. There is a (good) reason why Audyssey or JBL deliver a black box that includes a microphone. Others sell loudspeaker calibration as a service.
Then there are problems with the DCX. Visit http://tech.groups.yahoo.com/group/DCX2496/
I don't know how much time you have, but I expect you'll have to spent all of it on explaining recording software, PC sound problems, how to set up an AVR and Behringer's DCX. Do you think that there is a chance to charge people for that?—they already paid for the subs.
Good luck, Markus
I don't know Earl, in one thread you argue that a sound field is 3 dimensional and therefore a single eq response can not correct the response to be optimum at more than a single point. Now here you are indicating that you can provide single eq function that will yield optimal results over several seating positions, and that eq is being applied to multiple sources as well. This can only translate to it not being optimal at any of the positions.
john k... said:
Are you thinking of the DEQ2496 perhaps? That unit is for EQ, the DCX is a digital crossover with 6 outputs and multiple filters and gain settings are available.
So this unit would mirror, digitally, what the control settings are on most subwoofer amplifiers for phase, crossover and gain etc. Just much more functional.
I had already bought one to help me do these adjustments, before I even heard that Geddes was going to offer this.
-Tony
Tony
Correct on the functional steps, and I do think John got mixed up.
Recording the signal is pretty easy as there are no end of software that can record wav files, Windows used to come with one, still does as far as I know. Nothing needs to be calibrated as long as nothing is changed in the setup between runs. Noise is not an issue because the software deals with that. It's all signal processing and as long as you know how to do it there aren't any problems.
Markus will just buy the JBL box I suppose - thats fine.
Correct on the functional steps, and I do think John got mixed up.
Recording the signal is pretty easy as there are no end of software that can record wav files, Windows used to come with one, still does as far as I know. Nothing needs to be calibrated as long as nothing is changed in the setup between runs. Noise is not an issue because the software deals with that. It's all signal processing and as long as you know how to do it there aren't any problems.
Markus will just buy the JBL box I suppose - thats fine.
Unfortunately the DCX cannot do what I really wanted, the three main channels all summed. It can only sum two. So I would use left and right summed with the receiver set on "No sub". From what I understand there is not much uncorrelated LF in the center channel so this would work fine. If the center is completely uncorrelated with the L&R then this is a potential problem for which I don't have an immediate solution. You could sum left and right electrically externally and then use the DCX to sum this sum with the center.