Do we NEED subwoofers with full-range speakers, and how many?

[DWR]

Excellent post Frantz, and as a Geddes Abbey 12A owner I will echo your comments on the bass response in my room also being the best I have ever had. I have gone by the Geddes multi sub approach as much as possible, my system being in the living room rather limits my sub placement options, and my subs are in large enclosures being sealed subs. I have one in the corner at floor level behind my right speaker, another at the 36" level off the floor both firing out into the room. My third sub is behind the listening position about 6' and fires across the room at 53" above floor level. That 3rd sub seems to be the tough one to place especially in a living room setting. My 3rd sub is a 12" ported enclosure that is actually in the coat closet so only the grille shows on the wall like a picture. I run all 3 subs at 90 degrees out of phase with the main speakers. All of my subs were built by myself.

 

[DWR]

Nyal that is an excellent point on the delay factor. On the sub that is behind the listening position I have it delayed 5.82mS to make up for the difference in distance to the listening position from the 2 front subs. Have you ever tried the Thor subs with your Orions? The Linkwitz Orions are the speakers I had before I went with the Abbeys, I ran the Thor subs with them and they did blend very well with the Orions. My 12" sub is actually a Thor that I redesigned the enclosure on, when I sold the Orions I kept the Thors.

 

[Gregadd]

Frantz at least one article warns against mating a 10" with a 12" and so on. Mating different brands also present problems because they have different design approaches. As long as the design is the same and the cone is the same you should be okay.

 

[DWR]

Greg, Geddes multi sub approach states that the subs need not be the same, in size or type.

 

[Gregadd]

Greg, Geddes multi sub approach states that the subs need not be the same, in size or type.

I know .I take issue with that.

 

[DWR]

OK understood, just so you know, the Geddes method for placement is determined by frequency response of the subs. The lowest FR sub going in the corner being the foundation .

 

[Gregadd]

just so you know i'm not talking of the top of my head.

1: Choosing your subwoofers
We get a lot of questions about what types of subs to purchase for a multiple sub setup. When choosing the right subwoofers for your system, it’s a good idea to choose identical subs for optimal cancellation of room resonances, or ones of similar output level and design. Don’t for example mix and match a high quality 15” servo subwoofer with a cheap dual 6” bandpass sub that came with your “cubed” speaker system. Implementing an inferior subwoofer with a good one will limit your systems dynamic range and bandwidth to the weaker sub making your ears focus on the one that is breaking up and running out of steam. Mixing different subs will also reduce the success of canceling out room resonances since they will exhibit different amplitude and phase responses. Always select two well designed subwoofers (preferably the same) that are each in their own capable of filling your theater room with ample bass without bottoming out or running out of gas. If you can’t afford two subwoofers at the moment, buy one quality sub for now and add a similar capable one down the road when you’ve got the cash.




http://www.audioholics.com/tweaks/get-good-bass/subwoofer-connection-guide

 

[garylkoh]

Frantz at least one article warns against mating a 10" with a 12" and so on. Mating different brands also present problems because they different design approaches. As long as the design is the same and the cone is the sign you should be okay.

Depending on the design of the amplifiers in each of different brands of subwoofers, they may present different group delay. However, I find a great advantage to mixing up the cone sizes. Air has a fixed reactance, and the different cone areas couple to the air in different ways.

That is always assuming that you are using quality components to start with. Don't dig in the basement for the cheap subs that came with computer speakers, home theater in a box, etc. and expect a decent result. Throwing one of those in the mix could take you backwards.

 

[DWR]

Greg I never said or implied that you were talking off the top of your head and if you got that impression I am sorry I had no intent of that. Also I agree with the text you have posted in the fact that you don't want to use a "Best Buy" $79.99 with quality subs, that is not what I meant when I said they did not have to be the same size or type. The text is also aimed at subs in a home theater set-up. I do wish Dr Geddes would get involved in this thread.

 

[garylkoh]

http://www.audioholics.com/tweaks/get-good-bass/subwoofer-connection-guide

Greg, Sorry, but they are talking there specifically about subwoofers for LFE in a home theater. With proper set-up, I'm with Dr. Geddes that you don't need the same type or size subwoofers. Even in a home theater, I'm for multiple varied subs, just not in the positions that the audioholics guide suggests.

 

[Gregadd]

Just a few examples. Inertia means subs of different sizes may have different transient response, servo vs non -servo, different cone materials, different FR and crossover points, different crossover slopes. different Q, different enclosure approaches, etc. I suppose you can deal with all that but why go down that raod?

 

[FrantzM]

HI

The dealy factor is real and one can compensate with something like the cheap Behringer DXC 2496 ... However it seems that we, humans take a while to process low Frequencies.. Up to 50 ms for our brain to process a 50 Hz signal... so the slight variations in distance may not be much of an issue, our brains might have integrated it all.

Slight correction, my use of the 3 subs method was in a pure 2-Ch system.

I did mix the subs and the results were not audibly different ..

 

[microstrip]

You need subs with infinitely adjustable Level, Crossover and phase controls. The phase is particularly important, I am not sure and simple switch will do…

Frantz,
Very interesting post - I was not aware of Earl Geddes work. Can you advise on a separate crossover unit with these capabilities? Do you prefer analog or digital units?

 

[brianherlihy]

Frantz and DWR - what type of subs and what size cones are you using. this has been a very interesting thread as i am struggling to get rid of a node sitting at 65 Hz in my room

 

[Nyal Mellor]

, I find a great advantage to mixing up the cone sizes. Air has a fixed reactance, and the different cone areas couple to the air in different ways.

The Velodyne 1812, which I had the pleasure of listening to at a recent Bay Area Audiophile Society event, is actually two subwoofers in one box (one 18" and one 12"), with a crossover between the two. I talked to the head of engineering and he said they did that because they were able to optimize each sub for its purpose resulting in overall better sound quality.

 

[Nyal Mellor]

Nyal that is an excellent point on the delay factor. On the sub that is behind the listening position I have it delayed 5.82mS to make up for the difference in distance to the listening position from the 2 front subs. Have you ever tried the Thor subs with your Orions? The Linkwitz Orions are the speakers I had before I went with the Abbeys, I ran the Thor subs with them and they did blend very well with the Orions. My 12" sub is actually a Thor that I redesigned the enclosure on, when I sold the Orions I kept the Thors.

No I have not tried the Thors. I wanted to get some reference level subs that I could use in multiple systems, including my 'winter project' speaker system (an open baffle design).

 

[Nyal Mellor]

HI

The dealy factor is real and one can compensate with something like the cheap Behringer DXC 2496 ... However it seems that we, humans take a while to process low Frequencies.. Up to 50 ms for our brain to process a 50 Hz signal....

I agree, and if you think about the mathematics, a small time delay associated with different placement of a sub relative to a mains (e.g. 5ft ~ 5ms) is not a large phase error because the wavelengths are large.

But not time-synching a system means you have introduced a frequency dependent phase error, since the impact of 5ms time-synch error is not the same at all frequencies.

A very important related factor, and one that has been discussed a few times is the potential to localize subs if too high a crossover point is chosen. Whilst the commonly accepted figure is 80Hz for a sub crossover, AES research suggests that one can localize down to 50Hz. If this is true then using multiple subs does fix some problems but introduces some more into the mix. It just depends whether the problems you have fixed are greater than the problems you have created.

 

[MC352]

I put in a sub, didn't have an active crossover, it drove me nuts and i had it turned off more than on. Then I put in an active crossover, and bingo everything sounded much better. Now I am doing some tweaking to try to make it even better.

 

[FrantzM]

I agree, and if you think about the mathematics, a small time delay associated with different placement of a sub relative to a mains (e.g. 5ft ~ 5ms) is not a large phase error because the wavelengths are large.

But not time-synching a system means you have introduced a frequency dependent phase error, since the impact of 5ms time-synch error is not the same at all frequencies.

A very important related factor, and one that has been discussed a few times is the potential to localize subs if too high a crossover point is chosen. Whilst the commonly accepted figure is 80Hz for a sub crossover, AES research suggests that one can localize down to 50Hz. If this is true then using multiple subs does fix some problems but introduces some more into the mix. It just depends whether the problems you have fixed are greater than the problems you have created.

Nyal

Although the phase error is frequency dependent, it remains that the signal processing time remains high with respect to the largest module of phase error... I am not sure I understand what you mean however by System Time Synching.

 

[Nyal Mellor]

Nyal

Although the phase error is frequency dependent, it remains that the signal processing time remains high with respect to the largest module of phase error... I am not sure I understand what you mean however by System Time Synching.

I meant adjusting the speaker system so that an impulse send to each speaker would arrive at the listening position at the same point in time. I think the better term (at least more widely used) would be 'time alignment'...

Now I am not sure what you mean by 'the signal processing time remains high with respect to the largest module of phase error'?