Re: Re: Re: Re: Multiple Small Subs - Geddes Approach
However, if the mains sub does not have sufficient power on the low frequency end, then the issue becomes complicated by using front subs to enhance low frequency output and reflection point subs to cancel.
Now, this must sound like I'm selling subs, which might not be a bad idea.
Subs with built-in adjustable time delay and polarity switch.
Just talking about subs and placing in general seems to simplify the issue. As a personal preference, I would want the mains to go down as low as the subs. This is necessary to preseverve the initial tranient impact before the room modes are excited. Then if the overal feel is not satisfactory, subs would be used at specific frequencies that are the most annoying, which means taylored frequency response depending on location. The "listening area" would be measred to identify the problematic frequencies, and the subs to fix the problem would be located at appropriate reflection points with delayed inverted phase of appropriate amplitude. The purpose would be to allow the subs to be used as absorbers to minimize reflection waves, and thus the room modes would not be excited. Care should be taken to limit the upper frequency as low as possible with sharp cutoffs for remote subs.markus76 said:
However, if the mains sub does not have sufficient power on the low frequency end, then the issue becomes complicated by using front subs to enhance low frequency output and reflection point subs to cancel.
Now, this must sound like I'm selling subs, which might not be a bad idea.
Subs with built-in adjustable time delay and polarity switch.
youngho said:
Youngho
This isn't quite true. In all my work I use spatial averaging and have done so ever since my PhD thesis (which clearly showed the need). So even if it was not mentioned, any results, or conclusions that I have drawn, were based on a spatial average arround the center of the room and are valid for any listener in that space. It was precisely this point that I complained about in the Welti paper in my letter to JAES. His sampling technique was not satisfactory since it only used symmetrical points from symmetrical sources. This creates a statistical bias in the results. I averaged over random points with random source locations because this yields a more reliable set of statistics. It's a complex point but if you want to get into the theory more then you need to understand it. Symmetrical sampling will always favor symmetric locations because of the statistical bias, but it isn't as statistically significant as random samples. When you do random, you find out that random comes out better. Welti, as I have pointed out, never tried random. I did both. I have only published my conclusions not my work.
I focus on spatial averaging in the setup because so many people will want to skip this technique because it is difficult. But it must not be skiped - thats critical. So I emphasize it more and more. But I have always done it myself.
Re: Re: Re: Re: Re: Re: Multiple Small Subs - Geddes Approach
Markus,
Is Earl's solution written up somewhere or is it 'hidden' in the posts?
Jan Didden
markus76 said:"Transient response", "reflection point"
Earls method IS the most simple DIY solution. Only the JBL unit is simpler to use.
Markus,
Is Earl's solution written up somewhere or is it 'hidden' in the posts?
Jan Didden
breez said:
I am interested in doing this too and have some ideas of what would be optimum. I'd be interested in working with you on this. If you don't copy software directly from JBL then you can publish it for free. In fact, unless you are making money off of it, it is hard for anyone to sue for infringment.
On random (or not) positioning: Isn't it so that the room modes are quite predictable, and that the max velocity occurs at precisely the middle (or 1/4 etc) between opposite walls, and that max pressure occurs at the wall boundary?
If that is so, why should the subs be placed 'at random'? If the subs are meant to counter or equalize nodal standing waves, they should then be placed precisely depending on the room dimensions. Or am I missing something?
Jan Didden
If that is so, why should the subs be placed 'at random'? If the subs are meant to counter or equalize nodal standing waves, they should then be placed precisely depending on the room dimensions. Or am I missing something?
Jan Didden
youngho said:
Yes, very mistaken. You may want to reread my stuff as you have a lot of it wrong.
In general, my approach results in the closest woofer being the lowest in level.
It is true that all approaches will yield the same results, but with a $1200 signal processing box you can get more data and more complex correction. Hoeever, my whole set of three subs and amps didn't cost me $1200. It's all about "value".
If I were doing a LF setup for some rich client I would use the JBL box. But if I'm trying to keep the costs down this box is overkill.
gedlee said:
I'm sorry, I'm not trying to misrepresent your views.
I had read this white paper here, which did not specifically discuss spatial averaging (except for three points all near the center of the room, which didn't seem to reflect your currently described method of spatial averaging), which as you note, you have been emphasizing more. Because this white paper was so abbreviated, I had significant problems accepting the conclusions, but I don't have the body of experience that you do. I hope you can understand why someone reading this white paper would not conclude that significant spatial averaging had been done, certainly not as I've seen it described more recently. I was referring to this original white paper in my post above: "Two cases were studied 1) subs in four corners and 2) one sub in a corner, and three
subs located randomly but one of them was placed closer to the ceiling. The later
placement is what I recommend. " Obviously, your approach has changed since then, but that is why I had referred to the original papers.
On his webpage, Markus included a quote attributed to you: "First setup the mains and the nearest sub...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...Now repeat this process with the third sub. The third sub, when you are close to it should barely be audible..." I hope you can understand why someone reading these words allegedly by you would conclude that the nearest subwoofer would be the loudest and the furthest one being the softest. Clearly, this quote was not written by you, right?
otto88 said:
Hi Otto
Yes, the mains should overlap the subs - I do believe that this is different than Welti. In my systems the mains have plenty of output capability so limiting power to them is less effective for sound quality than overlaping them. If you had mains that did not have enough power to play cleanly down low then you have no choice but to use a crossover. But I don't recommend this. I always use big woofers for their HF directivity, but that gives me the freedom to also take them low. And "Yes" I do favor "sealed", but sealed bandpass, not direct radiating.
Mixed sealed and ported is fine, because details like this don't have a big effect. The big effect is number and location.
And to be clear - my recommendation IS NOT for completely random placement, only non-symmetrical placement which allows for an almost random placement. But one sub should always be in a corner. This is not random, only which corner you choose. So maybe we should call my approach versus Welti as symmetrical versus non-symmetrical.
Re: Re: Re: Re: Re: Re: Multiple Small Subs - Geddes Approach
Heh, I'm sure there are different tradeoffs and preferences. You'll get no argument from me on this.markus76 said:"Transient response", "reflection points"
Earls method IS the most simple DIY solution. Only the JBL unit is simpler to use.
You'll still have to fix those loose doors and windows and anything that might rattle.markus76 said:
gedlee said:
markus76 said:
gedlee said:
"First setup the mains and the NEAREST (emphasis mine) sub...The bass should be sagging slightly at this point since you will be adding in two more subs...Now add in the NEXT CLOSEST (emphasis mine) sub...Now repeat this process with the third sub. The third sub, when you are close to it should barely be audible..."
Okay, the first sub is the closest. According to Markus and the quote, it's the loudest. The second subwoofer is the next closest, and it's the next loudest. The third subwoofer, because it's not the NEAREST or the NEXT CLOSEST, by process of elimination must be the FARTHEST, and this is the one that is barely audible when you're close to it.
(lightbulb) You must be saying the subwoofer NEAREST THE MAIN SPEAKERS, not the listener, right? Got it.
Re: Re: Re: Re: Re: Re: Re: Multiple Small Subs - Geddes Approach
The preference to have huge level variations (±15dB) at low frequencies?
Multiple subs do nothing to the low frequency reverberation time, that's true. Only active (not available as a product) or passive absorption can help. But what multiple subs do is to smooth the frequency response in a way it can't be done with any other method (I know of). That IS the number one goal for low frequency reproduction.
soongsc said:
The preference to have huge level variations (±15dB) at low frequencies?
Multiple subs do nothing to the low frequency reverberation time, that's true. Only active (not available as a product) or passive absorption can help. But what multiple subs do is to smooth the frequency response in a way it can't be done with any other method (I know of). That IS the number one goal for low frequency reproduction.