Gain matching question - AVS Forum | Home Theater Discussions And Reviews
Forum Jump: 
 5Likes
  • 1 Post By craig john
  • 1 Post By Artzilla
  • 1 Post By craig john
  • 1 Post By Artzilla
  • 1 Post By craig john
 
Thread Tools
post #1 of 15 Old 05-16-2020, 05:28 AM - Thread Starter
Senior Member
 
Artzilla's Avatar
 
Join Date: Jul 2006
Posts: 316
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 92 Post(s)
Liked: 39
Question Gain matching question

I've always level matched my subs and achieved extremely satisfying results using dual cylinders. After moving to a new house and new room I was unable to achieve similar results, even after adding a 3rd sub. Obviously the room is a huge factor but without getting into details of that I'll just get to my question.

I want to try gain matching and understand the process but am at a loss as to how to measure the subs. I have 2 SVS cylinder subs that downfire onto a plate which serves as the base. The third newer sub is an SVS front firing box. How do I measure to gain match? The construction of the subs doesn't allow me to get the cylinder woofers in the same space and orientation as the front firing sub and vice versa. As I understand it placing the woofers in the same spot and the mic the same distance from the woofers is key.

Should I:

A) Just place all 3 subs in the same spot and measure from the side of the cylinders at the opening between the cylinder and the base plate. This would put the mic about 2" off the floor and whatever distance from the center of the woofer cone I find to be recommended. It's a 12" woofer so the edged of the cylinder is roughly 7" and I would back the mic off from that and use the same total distance to measure front the front center of the cone on the front firing sub. I just can't imagine this would result in accurate gain matching of all 3 subs. Plus the mic would be in the neighborhood of 13-14" off the floor. (I am generalizing measurements but you get the idea)

B) Remove the base plates from the cylinders, lay them on their sides (effectively making them front firing in this orientation) and raise the sub so that the woofer cone is in the same position as that of the box sub? Even if the positioning is the same I have to think that reattaching the base plate and returning them to their standing orientation this will be innacurate as well.

I have no other ideas and suspect I may not be able to accurately gain match so I welcome any suggestions.

FWIW The subs are mismatched but close enough in characteristics that Ed Mullen indicated they should play well together and 2 of them have for years. They are:

1. SVS 25-31 plus 500 watts 25 hz native tune - port plugged to 20 hz tune - downfiring
2. SVS 20-39 pc plus 500 watts 20 hz native tune - down firing
3. PB12 Plus (Recent single driver version not the older dual driver beast) 800 watts - 20 hz native tune - front firing

Last edited by Artzilla; 05-17-2020 at 05:29 AM.
Artzilla is offline  
Sponsored Links
Advertisement
 
post #2 of 15 Old 05-16-2020, 09:15 AM
AVS Forum Special Member
 
imureh's Avatar
 
Join Date: Mar 2009
Location: Houston, TX
Posts: 5,539
Mentioned: 286 Post(s)
Tagged: 0 Thread(s)
Quoted: 3499 Post(s)
Liked: 4715
Quote:
Originally Posted by Artzilla View Post
I've always level matched my subs and achieved extremely satisfying results using dual cylinders. After moving to a new house and new room I was unable to achieve similar results, even after adding a 3rd sub. Obviously the room is a huge factor but without getting into details of that I'll just get to my question.

I want to try gain matching and understand the process but am at a loss as to how to measure the subs. I have 2 SVS cylinder subs that downfire onto a plate which serves as the base. The third newer sub is an SVS front firing box. How do I measure to gain match? The construction of the subs doesn't allow me to get the cylinder woofers in the same space and orientation as the front firing sub and vice versa. As I understand it placing the woofers in the same spot and the mic the same distance from the woofers is key.

Should I:

A) Just place all 3 subs in the same spot and measure from the side of the cylinders at the opening between the cylinder and the base plate. This would put the mic about 2" off the floor and whatever distance from the center of the woofer cone I find to be recommended. It's a 12" woofer so the edged of the cylinder is roughly 7" and I would back the mic off from that and use the same total distance to measure front the front center of the cone on the front firing sub. I just can't imagine this would result in accurate gain matching of all 3 subs. Plus the mic would be in the neighborhood of 13-14" off the floor. (I am generalizing measurements but you get the idea)

B) Remove the base plates from the cylinders, lay them on their sides (effectively making them front firing in this orientation) and raise the sub so that the woofer cone is in the same position as that of the box sub? Even if the positioning is the same I have to think that reattaching the base plate and returning them to their standing orientation this will be innacurate as well.

I have no other ideas and suspect I may not be able to accurately gain match so I welcome any suggestions.

FWIW The subs are mismatched but close enough in characteristics that Ed Mullen indicated they should play well together and 2 of them have for years. They are:

1. SVS 25-31 plus 500 watts 25 hz native tune - port plugged to 20 hz tune - downfiring
2. SVS 20-39 pc plus 500 watts 20 hz native tune - down firing
3. PB12 Plus (Recent single driver version not the older dual driver beast) 800 watts - 20 hz native tune - front firing
What you are referring to is gain matching not level matching. Level matching would be a lot easier as you just place mic at MLP and adjust gain on sub to get the same levels at MLP. In gain matching you are effectively removing the impact of the room and measuring the subs with mic really close to the driver. So want to make sure you are referring to gain matching

Subwoofers: Dual Rythmik FV18s Rev2 with Paper cone, BOSS Platform mini riser, 2 BK LFE
Speakers: PSA MTM-210 L/C/R - Klipsch 250S Surrounds, PSA MT 110sr Rear Surrounds, 4 x SVS Prime Elevation for Atmos
Video: Epson PowerLite Home Cinema 6500 UB; Room: 2100 cuft sealed on suspended floor
Receiver & Players: Denon X4400H, Oppo UDP-203, Roku Ultra. Apple TV, Minidsp 2X4 HD
imureh is online now  
post #3 of 15 Old 05-16-2020, 01:01 PM
AVS Forum Club Gold
 
craig john's Avatar
 
Join Date: Oct 2003
Location: Lancaster, PA
Posts: 12,231
Mentioned: 80 Post(s)
Tagged: 0 Thread(s)
Quoted: 1634 Post(s)
Liked: 1937
Quote:
Originally Posted by Artzilla View Post
I've always level matched my subs and achieved extremely satisfying results using dual cylinders. After moving to a new house and new room I was unable to achieve similar results, even after adding a 3rd sub. Obviously the room is a huge factor but without getting into details of that I'll just get to my question.

I want to try gain matching and understand the process but am at a loss as to how to measure the subs. I have 2 SVS cylinder subs that downfire onto a plate which serves as the base. The third newer sub is an SVS front firing box. How do I measure to gain match? The construction of the subs doesn't allow me to get the cylinder woofers in the same space and orientation as the front firing sub and vice versa. As I understand it placing the woofers in the same spot and the mic the same distance from the woofers is key.

Should I:

A) Just place all 3 subs in the same spot and measure from the side of the cylinders at the opening between the cylinder and the base plate. This would put the mic about 2" off the floor and whatever distance from the center of the woofer cone I find to be recommended. It's a 12" woofer so the edged of the cylinder is roughly 7" and I would back the mic off from that and use the same total distance to measure front the front center of the cone on the front firing sub. I just can't imagine this would result in accurate gain matching of all 3 subs. Plus the mic would be in the neighborhood of 13-14" off the floor. (I am generalizing measurements but you get the idea)

B) Remove the base plates from the cylinders, lay them on their sides (effectively making them front firing in this orientation) and raise the sub so that the woofer cone is in the same position as that of the box sub? Even if the positioning is the same I have to think that reattaching the base plate and returning them to their standing orientation this will be innacurate as well.

I have no other ideas and suspect I may not be able to accurately gain match so I welcome any suggestions.

FWIW The subs are mismatched but close enough in characteristics that Ed Mullen indicated they should play well together and 2 of them have for years. They are:

1. SVS 25-31 plus 500 watts 25 hz native tune - port plugged to 20 hz tune - downfiring
2. SVS 20-39 pc plus 500 watts 20 hz native tune - down firing
3. PB12 Plus (Recent single driver version not the older dual driver beast) 800 watts - 20 hz native tune - front firing
Gain-matching is generally only used with identical subs. Gain-matching is used to ensure all subs are driving the same energy into the room, so that no one sub is set higher or lower than the others. This ensures all the subs have the same max output limits and that no single sub is the output limiter for the system as a whole. With non-identical subs, one runs into the problems you describe. Its hard to match the gains of subs with different drivers, different amps and different form factors. And, even if you do get a perfect gain-match at the level you use for the gain-matching technique, its no guarantee that when you place them back into their final positions, that the gain structures will still be identical, or that one sub won't limit the other two.

You don't describe the issues you're having with level-matching in your new room, or why you think gain-matching might solve them. More information about the exact issues you're having, as well as a detailed description of the room and listening position may provide the information we need to help you get things squared away.

If you *really* want to try gain-matching, I think your second technique would be the better option. However, I can't guarantee you'll get a correct gain-match even with that technique.

Just sayin'

Craig
yodog likes this.

Lombardi said it:
"Perfection is not attainable, but if we chase perfection we can catch excellence.
My System (Edited Feb. 2020 to add 4K and Atmos updates)

Last edited by craig john; 05-16-2020 at 01:08 PM.
craig john is offline  
Sponsored Links
Advertisement
 
post #4 of 15 Old 05-17-2020, 03:58 AM - Thread Starter
Senior Member
 
Artzilla's Avatar
 
Join Date: Jul 2006
Posts: 316
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 92 Post(s)
Liked: 39
Quote:
Originally Posted by imureh View Post
What you are referring to is gain matching not level matching. Level matching would be a lot easier as you just place mic at MLP and adjust gain on sub to get the same levels at MLP. In gain matching you are effectively removing the impact of the room and measuring the subs with mic really close to the driver. So want to make sure you are referring to gain matching
Correct. Level matching is what I have done in the past. Gain matching is what I want to do now. I think I stated it correctly (second paragraph) but to be clear yes, A measurement technique for gain matching is what I'm after.
Artzilla is offline  
post #5 of 15 Old 05-17-2020, 05:09 AM - Thread Starter
Senior Member
 
Artzilla's Avatar
 
Join Date: Jul 2006
Posts: 316
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 92 Post(s)
Liked: 39
Quote:
Originally Posted by craig john View Post
Gain-matching is generally only used with identical subs. Gain-matching is used to ensure all subs are driving the same energy into the room, so that no one sub is set higher or lower than the others. This ensures all the subs have the same max output limits and that no single sub is the output limiter for the system as a whole. With non-identical subs, one runs into the problems you describe. Its hard to match the gains of subs with different drivers, different amps and different form factors. And, even if you do get a perfect gain-match at the level you use for the gain-matching technique, its no guarantee that when you place them back into their final positions, that the gain structures will still be identical, or that one sub won't limit the other two.

You don't describe the issues you're having with level-matching in your new room, or why you think gain-matching might solve them. More information about the exact issues you're having, as well as a detailed description of the room and listening position may provide the information we need to help you get things squared away.

If you *really* want to try gain-matching, I think your second technique would be the better option. However, I can't guarantee you'll get a correct gain-match even with that technique.

Just sayin'

Craig
Thanks Craig. The issues I have will not be solved by gain matching but it's the first of other steps I planned to take. It is not intended to address what I list below, just a different foundation to start with vs level matching. It may turn out just to be a learning exercise but that's ok. I'll know what the results are and and will be satisfied with gaining the experience even if it is not an improvement.

***Midway through the next paragraph I realized a somewhat unique amp gain issue might make gain matching impossible or at least not optimal but I left the the info below in case you care to read it.

I agree about identical subs but if I recall a past conversation with SVS support they were not against it. Perhaps due to the similarity of the 2 cylinders, all 3 being ported and either native or tunable to 20 Hz, and all having 12" drivers. Gain structure however is different on all 3 units. Also, interesting fact I learned from SVS was that the amp in my pb12 plus will not drive that sub to it's full potential unless the gain on the amp is set to max. I was able to achieve that with a reasonable -10 or -12 trim for that sub in the receiver & keep it level matched with the cylinders which are running off the second sub channel then through an AS-EQ1. This did help with the dynamics but I'm getting a little off topic and that may render gain matching impossible if I want full performance from the pb12.


Issues:
1) Bass lacking unless I run my mains full range and double bass. I know - heresy - but a pro's youtube video indicated that solved a problem for a client that nothing else seemed to so I tried it. It was a significant improvement but I prefer not to run my gear that way and it is likely contributing to the next item.

2) Taughtness/attack/clarity of bass is not bad but not as good as I have achieved with mostly the same system in a different home with only the cylinder subs. Decay is also an issue but I want to try the steps below before looking into bass traps.

3) Pressurization. Inferior to the previous room running only the 2 cylinders.(much larger 23' x 28' x 8' - with tile and lots of glass-not exactly bass helpers) I think I should be able to achieve more than my current configuration is yielding.


My guess is I don't have the right time/phase alignment with the mains and likely have some cancellation going on. On the advice of Ed Mullen despite having 2 discrete sub channels and XT32 should run all the subs off a single output.

The plan was to gain match the subs. Replace the AS-EQ1 with a MiniDSP between the subs and receiver so Audyssey sees a "single sub". Then I can time align all as a single sub post Audyssey with REW. After typing through this however I think due to the PB12 amp Gain issue I will have to stick to level matching.

I may also play with the Geddes method for placement and try raising one of the subs off the floor. My best past sound in the previous home was achieved with the 2 cylinders colocated by stacking one on top of the other and running them sealed (since the top sub blocked the ports of the bottom sub). I initially thought it was just having both in the sweet spot for that room but recently considered the possibility that elevating the second sub addressed some room modes I could not affect with both subs on the floor. I did not have REW at the time so I cannot say one way or the other.

Thanks again for your input

Last edited by Artzilla; 05-17-2020 at 06:25 AM.
Artzilla is offline  
post #6 of 15 Old 05-17-2020, 05:28 AM - Thread Starter
Senior Member
 
Artzilla's Avatar
 
Join Date: Jul 2006
Posts: 316
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 92 Post(s)
Liked: 39
Quote:
Originally Posted by Artzilla View Post
Thanks Craig. The issues I have will not be solved by gain matching but it's the first of other steps I planned to take. It is not intended to address what I list below, just a different foundation to start with vs level matching. It may turn out just to be a learning exercise but that's ok. I'll know what the results are and and will be satisfied with gaining the experience even if it is not an improvement.

***Midway through the next paragraph I realized a somewhat unique amp gain issue might make gain matching impossible or at least not optimal but I left the the info below in case you care to read it.

I agree about identical subs but if I recall a past conversation with SVS support they were not against it. Perhaps due to the similarity of the 2 cylinders, all 3 being ported and either native or tunable to 20 Hz, and all having 12" drivers. Gain structure however is different at least in relation to the position of the gain adjustment on the cylinders. Also, interesting fact I learned from SVS was that the amp in my pb12 plus will not drive that sub to it's full potential unless the gain on the amp is set to max. I was able to achieve that with a reasonable -10 or -12 trim for that sub in the receiver & keep it level matched with the cylinders which are running off the second sub channel then through an AS-EQ1. This did help with the dynamics but I'm getting a little off topic but that may render gain matching impossible if I want full performance from the pb12.


Issues:
1) Bass lacking unless I run my mains full range and double bass. I know - heresy - but a pro's youtube video indicated that solved a problem for a client that nothing else seemed to so I tried it. It was a significant improvement but I prefer not to run my gear that way and it is likely contributing to the next item.

2) Taughtness/attack/clarity of bass is not bad but not as good as I have achieved with mostly the same system in a different home with only the cylinder subs. Decay is also an issue but I want to try the steps below before looking into bass traps.

3) Pressurization. Inferior to the previous room running only the 2 cylinders. I think I should be able to achieve more than my current configuration is yielding.


My guess is I don't have the right time/phase alignment with the mains and likely have some cancellation going on. On the advice of Ed Mullen despite having 2 discrete sub channels and XT32 should run all the subs off a single output.

The plan was to gain match the subs. Replace the AS-EQ1 with a MiniDSP between the subs and receiver so Audyssey sees a "single sub" and I can time align all as a single sub post Audyssey with REW. After typing through this however I think due to the PB12 amp Gain issue I may have to stick to level matching.

I may also play with the Geddes method for placement and try raising one of the subs off the floor. My best past sound in the previous home was achieved with the 2 cylinders colocated by stacking one on top of the other and running them sealed (since the top sub blocked the ports of the bottom sub). I initially thought it was just having both in the sweet spot for that room but recently considered the possibility that elevating the second sub addressed some room modes I could not affect repositioning the subs on the floor. I did not have REW at the time so I cannot say one way or the other.

Thanks again for your input
Regarding Craig's inquiry about the room:

This is a multi-use living, HT,music room - we do everything here except cook and sleep. MLP is just off center of the long wall of a 13' x 19' x 8 room. Mains B&W 803s towers, 6' & 7' or so from side walls on the opposite wall with baffles about 32" from the back wall. There is a low rise Sanus stand between them as well as a 65" tv set back as far as I can get it.

Windows and doors do not work well for a lengthwise configuration. To be sure the room is compromised, but fairly well damped with carpeting, furnishings & wall hangings. It does have 3 openings one being a stairwell (maybe that should be considered a 4th) so it is far from ideal.

I've attached an image as the room is and another with placement options I've considered "if" I decide to cover or seal some windows. I'd be curious if you have any input.
Attached Thumbnails
Click image for larger version

Name:	As Is.jpg
Views:	36
Size:	459.2 KB
ID:	2727740   Click image for larger version

Name:	New Plan Alternate-02.jpg
Views:	36
Size:	512.1 KB
ID:	2727742  

Last edited by Artzilla; 05-17-2020 at 05:36 AM.
Artzilla is offline  
post #7 of 15 Old 05-17-2020, 05:31 AM - Thread Starter
Senior Member
 
Artzilla's Avatar
 
Join Date: Jul 2006
Posts: 316
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 92 Post(s)
Liked: 39
Quote:
Originally Posted by imureh View Post
What you are referring to is gain matching not level matching. Level matching would be a lot easier as you just place mic at MLP and adjust gain on sub to get the same levels at MLP. In gain matching you are effectively removing the impact of the room and measuring the subs with mic really close to the driver. So want to make sure you are referring to gain matching
Imureh,
I just realized that the subject line read "level matching". I could have sworn I typed gain matching. I tried to change the header but it either won't allow me to or I'm doing it incorrectly.

***Maybe there is a moderator afoot who could change the Thread title to "Gain" matching question" for me?

Thanks.

Last edited by Artzilla; 05-17-2020 at 09:07 AM.
Artzilla is offline  
post #8 of 15 Old 05-18-2020, 08:49 AM
AVS Forum Club Gold
 
craig john's Avatar
 
Join Date: Oct 2003
Location: Lancaster, PA
Posts: 12,231
Mentioned: 80 Post(s)
Tagged: 0 Thread(s)
Quoted: 1634 Post(s)
Liked: 1937
There is a lot to unpack here, and there are a number of things that can cause the issues you describe. The best way for us to help you, (and/or for you to help yourself), is to take some measurements of the bass in your room. Until the root cause(s) can be delineated, we can only speculate.

Quote:
Originally Posted by Artzilla View Post
Thanks Craig. The issues I have will not be solved by gain matching but it's the first of other steps I planned to take. It is not intended to address what I list below, just a different foundation to start with vs level matching. It may turn out just to be a learning exercise but that's ok. I'll know what the results are and and will be satisfied with gaining the experience even if it is not an improvement.

***Midway through the next paragraph I realized a somewhat unique amp gain issue might make gain matching impossible or at least not optimal but I left the the info below in case you care to read it.

I agree about identical subs but if I recall a past conversation with SVS support they were not against it. Perhaps due to the similarity of the 2 cylinders, all 3 being ported and either native or tunable to 20 Hz, and all having 12" drivers. Gain structure however is different on all 3 units. Also, interesting fact I learned from SVS was that the amp in my pb12 plus will not drive that sub to it's full potential unless the gain on the amp is set to max. I was able to achieve that with a reasonable -10 or -12 trim for that sub in the receiver & keep it level matched with the cylinders which are running off the second sub channel then through an AS-EQ1. This did help with the dynamics but I'm getting a little off topic and that may render gain matching impossible if I want full performance from the pb12.
I can understand SVS saying these 3 subs will work together because they are all ported and they can all be tuned to have the same F3. However, I can't imagine them telling you the subs could be gain-matched when they all use dissimilar amps with significantly dissimilar gain structures. More importantly, I don't think gain-matching is the correct approach given your situation and it won't addrrss the problems you describe. It seems you have also come to the, (correct), conclusion that gain-matching won't work for your situation. In fact, with the significant dissimilarity of the gain structures of the amps in the cylinder subs vs. the PB12+, gain-matching would be contra-indicated.

The requirement that the PB12+ be set to its maximum gain setting might be really problematic for your setup. When you set the amp on the PB12+ to max output, and the receiver trim to 0, what SPL is read at the primary measurement position with just that sub connected? In general, you want this level to read between 75 and 80 dB, and you would use the subwoofer gain control to set the the measured level to this range. However, if the sub amp needs to be at its maximum setting to get full performance from the sub, it may not be possible to do this, and that sub will be set well above the other 2. More importantly, it may be set high enough that there is not enough adjustment capability in the receiver trim range to get it in line with the other two. Hopefully, that is not the case.

Quote:
Originally Posted by Artzilla View Post
Issues:
1) Bass lacking unless I run my mains full range and double bass. I know - heresy - but a pro's youtube video indicated that solved a problem for a client that nothing else seemed to so I tried it. It was a significant improvement but I prefer not to run my gear that way and it is likely contributing to the next item.

2) Taughtness/attack/clarity of bass is not bad but not as good as I have achieved with mostly the same system in a different home with only the cylinder subs. Decay is also an issue but I want to try the steps below before looking into bass traps.

3) Pressurization. Inferior to the previous room running only the 2 cylinders.(much larger 23' x 28' x 8' - with tile and lots of glass-not exactly bass helpers) I think I should be able to achieve more than my current configuration is yielding.

My guess is I don't have the right time/phase alignment with the mains and likely have some cancellation going on. On the advice of Ed Mullen despite having 2 discrete sub channels and XT32 should run all the subs off a single output.

The plan was to gain match the subs. Replace the AS-EQ1 with a MiniDSP between the subs and receiver so Audyssey sees a "single sub". Then I can time align all as a single sub post Audyssey with REW. After typing through this however I think due to the PB12 amp Gain issue I will have to stick to level matching.
I think your suspicion of a mis-timing of the subs could be the culprit. However, I'm not sure the solution of putting all 3 subs on the same subwoofer output is the optimal approach. My suggestion would be to place the two cylinder subs on the Sub 1 output and then physically place those two subs equidistant to the LP. This will allow you to use the single Sub 1 Distance setting to time these two subs identically. Then place the PB12+ sub on the Sub 2 output and set its Distance appropriately, (remember that the "physical" distance of the subs is not necessarily the same as the "acoustic" distance due to the inherent latency in the subwoofer amplifiers, and the "correct" Distance setting for most any sub will virtually always be longer than the physical distance to take the latency into account.) Re-run Audyssey after making these changes and Audyssey will set the Distances based on the arrival times it measures. Those settings may or may not be the optimal settings for the speaker/subwoofer(s) splice, which I'll address later, but they should be correct for the relative Distances of the subs to each other.

Quote:
Originally Posted by Artzilla View Post
I may also play with the Geddes method for placement and try raising one of the subs off the floor. My best past sound in the previous home was achieved with the 2 cylinders colocated by stacking one on top of the other and running them sealed (since the top sub blocked the ports of the bottom sub). I initially thought it was just having both in the sweet spot for that room but recently considered the possibility that elevating the second sub addressed some room modes I could not affect with both subs on the floor. I did not have REW at the time so I cannot say one way or the other.
The Geddes approach might be your best option, but you absolutely NEED to use measurements to make it work. Unfortunately, the description of the Geddes technique, ( https://mehlau.net/audio/multisub_geddes/ ), doesn't specifically address how to use the Distance settings in the Bass Management processor, or how to connect multiple subs to a processor that has dual subwoofer outputs. It assumes you'll connect them all to one output, which may or may not be the optimal way to use the processor. If you want to go this route, I would probably use the PB12+ as the "primary" sub. Set its gain at maximum, then place it where it is most optimal, (see below, but most likely in the front left corner.) Then use measurements sweeps of the subwoofer and the CC to optimize/maximize the SPL at the crossover point. Once that is done, then add the 2nd sub on the Sub2 output and measure the same sweeps as before. You're primarily looking to fill in the nulls that existed in the first set of measurements. Slowly turn up the gain on the 2nd sub until you start to see an improvement in the frequency response. You may need to play around with both Levels and Distances to optimize this. Then repeat the process with the 3rd sub. In this scenario, you may be better off with the 3rd sub in a position that is not symmetrical to the listening position, but try as many positions as possible.

Remember that the Geddes technique is not used to maximize system headroom. It is used to optimize the frequency response, and it assumes that the combined output of the three subs will provide all the "needed" headroom. Since it generally results in different gain settings for each sub, you may find that the highest set sub is the sub that hits its max output first and therefore limits the system output to that level. If that level is all the output you'll ever need, you're done. If not, you're back to the drawing board.

Since the issues you have described can all be explained by frequency response anomalies, the solutions are best handled by adjustments that affect the frequency response, and with frequency response measurements. Without measurements, any suggestion made by anyone are just based on speculation and guesswork.

Here's one such speculation: Your subs and speakers are not timed properly to each other. I have found that Audyssey often fails to get this parameter right, especially in systems with multiple subs, and even more frequently when the multiples subs are all run off of one subwoofer output but the subs are placed at different distances from the LP. The best way to correct this problem is to measure the response and adjust the subwoofer Distances accordingly using the measurements in real time. Here's an example of this phenomenon:



This is a system with 3 subs connected to one subwoofer output, and therefore just one setting for subwoofer Distance. The cyan trace is with the Audyssey set subwoofer Distance. Note the 20+ dB cancellation centered on the crossover frequency of 80 Hz. This is what happens with the speakers and subs are mistimed relative to each other. The green trace is the exact same system, the only difference being an addition of 3.6 feet to the subwoofer Distance setting. The speakers and sub are now correctly timed to each other and the cancellation is eliminated. On these fora, this process is often referred to as the "Subwoofer Distance Tweak." The difference in sound between these two measurements is stunning. The first setting is lifeless and dull with seemingly no bass energy. The second setting adds power and weight to the bass with pressurization and tactile impact, and articulation. It is certainly possible that this is your biggest, and possibly ONLY problem with your bass response. Frequency response measurements will definitely tell the tale.

Quote:
Originally Posted by Artzilla View Post
Regarding Craig's inquiry about the room:

This is a multi-use living, HT,music room - we do everything here except cook and sleep. MLP is just off center of the long wall of a 13' x 19' x 8 room. Mains B&W 803s towers, 6' & 7' or so from side walls on the opposite wall with baffles about 32" from the back wall. There is a low rise Sanus stand between them as well as a 65" tv set back as far as I can get it.

Windows and doors do not work well for a lengthwise configuration. To be sure the room is compromised, but fairly well damped with carpeting, furnishings & wall hangings. It does have 3 openings one being a stairwell (maybe that should be considered a 4th) so it is far from ideal.

I've attached an image as the room is and another with placement options I've considered "if" I decide to cover or seal some windows. I'd be curious if you have any input.






Neither of these configurations will result in an optimal listening position. The first has the seating against the back wall. This will generally result in large peaks or nulls, (mostly peaks), due to the close and high intensity reflections from the back wall. The second configuration places the LP in the virtual center of the room which is generally the weakest point for bass response in any room. A more "optimal" listening position is placed at one of the odd number divisors of the room length, (i.e., thirds, fifths, etc.) If you can move the seating in either drawing to 2/3's or 3/5's of the room length, you'll be in a much better position for your starting bass response.

In the first diagram, the PB12+ is placed on the right side next to an opening to a stairwell and a doorway to the kitchen. That sub will receive little, if any, room reinforcement. Can it be moved to the left corner in the spot where you currently have a small cabinet? That placement would provide nearly optimal room reinforcement. It will also actuate all the modes in the room, which can then be addressed by the other 2 subs.

You have the option of different tunings of your subwoofers. I suggest you tune them all the same, whichever tuning that is. All 3 subs in "sealed" mode may work for you, especially on the lowest bass, where room-gain can provide more output in the lowest frequencies, (below 20 Hz). OTOH, all 3 subs set to 20 Hz tune can also work well, with output optimized at 20 Hz and above, but sacrificing the output below port tune. Either way, avoid tuning one sub in sealed mode and the other sub or subs in ported mode. A sealed/ported configuration will be very difficult to optimize and is not recommended.

So, now you have a lot of work to do... If you want to post some measurements of your subs, please start with the baseline response of each sub in each possible placement within your room. We can go from there.

Craig

Lombardi said it:
"Perfection is not attainable, but if we chase perfection we can catch excellence.
My System (Edited Feb. 2020 to add 4K and Atmos updates)
craig john is offline  
post #9 of 15 Old 05-18-2020, 08:50 AM
AVS Forum Club Gold
 
craig john's Avatar
 
Join Date: Oct 2003
Location: Lancaster, PA
Posts: 12,231
Mentioned: 80 Post(s)
Tagged: 0 Thread(s)
Quoted: 1634 Post(s)
Liked: 1937
Quote:
Originally Posted by Artzilla View Post
Imureh,
I just realized that the subject line read "level matching". I could have sworn I typed gain matching. I tried to change the header but it either won't allow me to or I'm doing it incorrectly.

***Maybe there is a moderator afoot who could change the Thread title to "Gain" matching question" for me?

Thanks.
@rboster

Lombardi said it:
"Perfection is not attainable, but if we chase perfection we can catch excellence.
My System (Edited Feb. 2020 to add 4K and Atmos updates)
craig john is offline  
post #10 of 15 Old 05-18-2020, 08:52 AM
AVS Forum Addicted Member
Moderator
 
rboster's Avatar
 
Join Date: Jul 2000
Location: USA
Posts: 25,958
Mentioned: 78 Post(s)
Tagged: 0 Thread(s)
Quoted: 1807 Post(s)
Liked: 2708
Quote:
Originally Posted by craig john View Post
Done!

FS: Tannoy AMS DC8 Speakers: https://www.avsforum.com/forum/209-a...-speakers.html

FS: RUIPRO 8K Fiber DisplayPort Cable 33 feet : https://www.avsforum.com/forum/211-a...d-support.html
rboster is online now  
post #11 of 15 Old 05-18-2020, 10:09 AM - Thread Starter
Senior Member
 
Artzilla's Avatar
 
Join Date: Jul 2006
Posts: 316
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 92 Post(s)
Liked: 39
Wow - that's a heckuva reply. Many Thanks. I have to say I feel a bit guilty because much of what you suggested has already been done to get me to the point I am currently at but there is still helpful info there and it's always reassuring to get confirmation on decisions made. I will reply to you in detail when I'm not working and when I figure out how to use the multi-quote feature.

Thanks again!
Artzilla is offline  
post #12 of 15 Old 05-18-2020, 11:29 AM - Thread Starter
Senior Member
 
Artzilla's Avatar
 
Join Date: Jul 2006
Posts: 316
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 92 Post(s)
Liked: 39
Craig - I can't figure out multi quote so I'll cut & paste. So much for working

Based on a comment in your post about the necessity of measurements I should reinforce that I am using REW with a calibrated Umik focusing mostly on frequency response measurements but also noting the waterfalls and spectrograms. (Waterfalls are horrible thus the reference in the previous post about doing what I could before pursuing room treatments.) Also - we are in agreement that time alignment is the target now that gain matching has been eliminated. I'll just stick with level matching and proceed with the original plan - plus whatever this conversation yields. I should also point out that even running the subs off a single output I will have independent time/distance/phase control of each sub with the minidsp. Essentially the mini DSP should leave little for Audyssey to do - except mess up the sub vs mains distance in it's calculation until I do the distance tweak So, yes, I am familiar with and perform this with every system change. It yields definite and significant improvements.

I can understand SVS saying these 3 subs will work together because they are all ported and they can all be tuned to have the same F3. However, I can't imagine them telling you the subs could be gain-matched when they all use dissimilar amps with significantly dissimilar gain structures. More importantly, I don't think gain-matching is the correct approach given your situation and it won't address the problems you describe. It seems you have also come to the, (correct), conclusion that gain-matching won't work for your situation. In fact, with the significant dissimilarity of the gain structures of the amps in the cylinder subs vs. the PB12+, gain-matching would be contra-indicated.



I agree. I'll have to go back to my emails to review the context of the suggestion. As I think back I believe the amp gain issue was discussed well after the conversation of gain matching so I may have overlooked what was obvious to the SVS rep. Gain matching is definitely out.


The requirement that the PB12+ be set to its maximum gain setting might be really problematic for your setup. When you set the amp on the PB12+ to max output, and the receiver trim to 0, what SPL is read at the primary measurement position with just that sub connected? In general, you want this level to read between 75 and 80 dB, and you would use the subwoofer gain control to set the the measured level to this range. However, if the sub amp needs to be at its maximum setting to get full performance from the sub, it may not be possible to do this, and that sub will be set well above the other 2. More importantly, it may be set high enough that there is not enough adjustment capability in the receiver trim range to get it in line with the other two. Hopefully, that is not the case.

I can't tell you what the measurement would be with the trim set to 0 would be but it would be ridiculous. SVS was very clear to get as close to max gain on the sub as I could without maxing out the negative trim levels of my receiver. Fortunately I was able to do this. My pre/pro goes as low as -15 trim and at max gain on the sub and at a trim of -12 the result was 78db with the pro/pro's internal test tones iirc.

I think your suspicion of a mis-timing of the subs could be the culprit. However, I'm not sure the solution of putting all 3 subs on the same subwoofer output is the optimal approach.

I'm sure time alignment is a big piece if not the entire puzzle. Putting all 3 subs on the same output was Ed Mullen's suggestion. He took a good bit if time explaining that all of the gain and phase measurements and adjustments I made with REW to get the subs measuring as well as I could with each other and the mains before running Audyssey were basically rendered null & void when I ran Audyssey allowing it to adjust the subs independently. I can't explain it as well as he did but the object is to get the best results you can with manual adjustments on the subs, placement etc, then let Audyssey see and treat your multiple subs or sub group as some refer to them as a single subwoofer. If done properly all you should have to do after that is the distance tweak.


My suggestion would be to place the two cylinder subs on the Sub 1 output and then physically place those two subs equidistant to the LP. This will allow you to use the single Sub 1 Distance setting to time these two subs identically. Then place the PB12+ sub on the Sub 2 output and set its Distance appropriately, (remember that the "physical" distance of the subs is not necessarily the same as the "acoustic" distance due to the inherent latency in the subwoofer amplifiers, and the "correct" Distance setting for most any sub will virtually always be longer than the physical distance to take the latency into account.) Re-run Audyssey after making these changes and Audyssey will set the Distances based on the arrival times it measures. Those settings may or may not be the optimal settings for the speaker/subwoofer(s) splice, which I'll address later, but they should be correct for the relative Distances of the subs to each other.


If you look at the first diagram I sent I have the cylinders on each end of the couch so they are just as you suggested. Further, the are manually adjusted to play nice together, then I run the AS-EQ1 to get them eq-ed to each other. If you are not familiar with the AS-EQ1 it's the stand alone version of the bass section of Audyssey XT32 that SVS and Ausyssey collaborated on before XT32 was available.

After doing this I measured and adjusted the PB12 to get it to blend well with them. Then run Audyssey. The area that makes it troublesome is then trying to get all 3 subs on 2 outputs to blend well with the mains. This portion would be easier and (assuming the subs are properly integrated with each other) to time align(Distance tweak) them with the mains. FWIW my measurements after XT32 but before the distance tweak are just as you pictured - a big deep suckout at the xover point.

Regarding your comments in the "Geddes" section of the post:
It would certainly be a psuedo Geddes approach experimenting with placement. I agree with your suggestions particularly using the PB12 as the primary sub and being that it is the newest and most powerful of the group I would be surprised it it was a limiting factor if placed properly.

Regarding the room:
Yes, neither configuration is ideal. FWIW, in future iterations of the 2nd diagram I had moved the seating position forward to approximately 1/3 of the length of the room. This was an early draft focused on trying to get things to fit, get my rear surrounds back and hopefully a little more sense of space in the sound by getting away from the back wall. I am aware of the cancellations. I find the my starting points for mlp and speakers with a mode calcualtor then measure and adjust as much as the floor plan allows. Part of the reason for considering the Geddes approach is because I am considering the vertical dimension of those modes as well.

In the first diagram, the PB12+ is placed on the right side next to an opening to a stairwell and a doorway to the kitchen. That sub will receive little, if any, room reinforcement. Can it be moved to the left corner in the spot where you currently have a small cabinet? That placement would provide nearly optimal room reinforcement. It will also actuate all the modes in the room, which can then be addressed by the other 2 subs.

The PB 12 ended up in front of the stairwell because when I had only the 2 cylinders that was the best spot for a single sub as far as frequency response. I did try the left corner as you suggested (with a single cylinder) but the frequency response was poor by comparison so I gave up on that after 1 measurement and a recheck to ensure it wasn't an error. Too much low end gain - I could not tame it enough to be viable. I never tested that left corner with the pb 12 however. Might be worth a shot being newer/more powerful/front firing and spoorting the onboard EQ. Very likely that I'll pursue this in either room configuration.

You have the option of different tunings of your subwoofers. I suggest you tune them all the same, whichever tuning that is. All 3 subs in "sealed" mode may work for you, especially on the lowest bass, where room-gain can provide more output in the lowest frequencies, (below 20 Hz). OTOH, all 3 subs set to 20 Hz tune can also work well, with output optimized at 20 Hz and above, but sacrificing the output below port tune. Either way, avoid tuning one sub in sealed mode and the other sub or subs in ported mode. A sealed/ported configuration will be very difficult to optimize and is not recommended.

Sealed and the 20 Hz tune sound best. 2 of the subs are native 20 Hz so that's how they are set up currently. Sealed sounded good as well though. I spent weeks in the rabbit hole of different placements and tunings. So many variables with 3 subs and multiple tunes. I can only put my marriage through that every couple of years or with a gear change but I'm due. I also have had to be very careful of room gain and have spent more effort trying to avoid too much than not getting enough. This was the opening to the rabbit hole and honestly I was fairly happy with sealed. A sealed approach to the a front left corner placement might be just what the DR ordered.


Thanks again for the time and effort you've put into this. I thoroughly enjoyed your posts. I'll be re-reading them and plotting my dastardly deeds this week so I'm prepared when work allows me to dive in.
yodog likes this.

Last edited by Artzilla; 05-18-2020 at 11:47 AM.
Artzilla is offline  
post #13 of 15 Old 05-18-2020, 05:27 PM
AVS Forum Club Gold
 
craig john's Avatar
 
Join Date: Oct 2003
Location: Lancaster, PA
Posts: 12,231
Mentioned: 80 Post(s)
Tagged: 0 Thread(s)
Quoted: 1634 Post(s)
Liked: 1937
Well, it seems you have a good understanding of the principals and techniques involved, and you know how to take and interpret the measurements. You have all the tools you need to optimize your subwoofer system. Now it just a matter of doing your "dastardly deeds!"

My own journey is not exactly the same as yours, but not too dissimilar either. I have 3 identical sealed sub. Two are equidistant to, but not symmetrical to, the LP. They're both connected to the Sub1 output. The 3rd sub is connected to Sub2 and is a different distance to the LP. All 3 are gain-matched. I've measured and optimized the pre-Audyssey response and then let Audyssey do it's thing. After running Audyssey, Sub1 and Sub2 were always set to different levels... i.e., they were no longer gain matched. Therefore to re-establish the gain-matching, I split the difference between the two and set them both to the same, intermediate level. This is the finished result:





While it's not likely that you'll be able to hit 115 dB at 10 Hz, you should certainly be able to get smooth response with a target curve to your preference.

Good luck and enjoy the journey!

Craig
badtlc likes this.

Lombardi said it:
"Perfection is not attainable, but if we chase perfection we can catch excellence.
My System (Edited Feb. 2020 to add 4K and Atmos updates)
craig john is offline  
post #14 of 15 Old 05-19-2020, 04:04 AM - Thread Starter
Senior Member
 
Artzilla's Avatar
 
Join Date: Jul 2006
Posts: 316
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 92 Post(s)
Liked: 39
Quote:
Originally Posted by craig john View Post
Well, it seems you have a good understanding of the principals and techniques involved, and you know how to take and interpret the measurements. You have all the tools you need to optimize your subwoofer system. Now it just a matter of doing your "dastardly deeds!"

My own journey is not exactly the same as yours, but not too dissimilar either. I have 3 identical sealed sub. Two are equidistant to, but not symmetrical to, the LP. They're both connected to the Sub1 output. The 3rd sub is connected to Sub2 and is a different distance to the LP. All 3 are gain-matched. I've measured and optimized the pre-Audyssey response and then let Audyssey do it's thing. After running Audyssey, Sub1 and Sub2 were always set to different levels... i.e., they were no longer gain matched. Therefore to re-establish the gain-matching, I split the difference between the two and set them both to the same, intermediate level. This is the finished result:





While it's not likely that you'll be able to hit 115 dB at 10 Hz, you should certainly be able to get smooth response with a target curve to your preference.

Good luck and enjoy the journey!

Craig
Impressive graphs. You must have really maximized room gain to hit 115db at 10hz with sealed subs. That would be impressive for ported subs but you got the best of both worlds. I don't expect to rival your results but they do inspire me to revisit a sealed configuration to see if I can run sealed to maintain the taught tactile feel I had previously running sealed and also delve deeper into the low end.

Thanks again for sharing your insight. I'm going to see if I can squirrel away some time to do some baseline sealed/ported/placement measurements this weekend.

***OK - I went in to look at your system in your signature. Maybe your choice of subs had a little more to do with it than room gain. Nice.
badtlc likes this.

Last edited by Artzilla; 05-19-2020 at 04:19 AM.
Artzilla is offline  
post #15 of 15 Old 05-20-2020, 04:16 AM
AVS Forum Club Gold
 
craig john's Avatar
 
Join Date: Oct 2003
Location: Lancaster, PA
Posts: 12,231
Mentioned: 80 Post(s)
Tagged: 0 Thread(s)
Quoted: 1634 Post(s)
Liked: 1937
Quote:
Originally Posted by Artzilla View Post
Impressive graphs. You must have really maximized room gain to hit 115db at 10hz with sealed subs. That would be impressive for ported subs but you got the best of both worlds. I don't expect to rival your results but they do inspire me to revisit a sealed configuration to see if I can run sealed to maintain the taught tactile feel I had previously running sealed and also delve deeper into the low end.

Thanks again for sharing your insight. I'm going to see if I can squirrel away some time to do some baseline sealed/ported/placement measurements this weekend.

***OK - I went in to look at your system in your signature. Maybe your choice of subs had a little more to do with it than room gain. Nice.
Yes, my subs are a large part of the response I get. But not for the reason you might think, and room gain still has a LOT to do with it.

Here is a graph of the "native" response of the subs, measured outdoor, groundplane at 2 m, with my in-room response right below it.



You can see that the native response begins to roll-off at about 40 Hz. Yet the in-room response is rising below 40 Hz. Part of this is the target curve I've applied. But *most* of it is the consequences of "pressure vessel gain", aka room gain. How does this happen? Well, its a function of the inherent nature of the sealed subwoofer design along with the nature of small rooms.

In a "small" room, (which is virtually all domestic-sized rooms), the bass will transition from modal response to pressure vessel response at the point where the wavelengths of the bass frequencies exceed the length of the longest dimension of the room. A 40 Hz frequency has a wavelength of 28 ft. That wavelength will not fit inside a 20 ft, or even a 24 ft. room. So at 40 Hz or so, instead of the wave reflecting back on itself and causing peaks and nulls, the wave simply builds pressure inside the room. This is "pressure vessel gain" and it causes bass in small room to increase as frequencies go lower. It happens in all rooms, but it is more pronounced in sealed rooms because sealed rooms can pressurize more easily. It's also more pronounced the smaller the room gets. And it works better with sealed subwoofers!

A ported subwoofer will roll-off steeply below the port tune. A ported sub will roll-off naturally at 24 dB/octave. However, a driver in a ported sub is "unloaded" below port tune and distortion will increase dramatically, and the driver will quickly go into over-excursion and be damaged. Therefore, most ported subs have a high pass filter that rolls the response off even faster below port tune. Unless the port is tuned to a very low frequency, there won't be enough output to allow the pressure vessel gain to build and become useful.

A sealed sub will naturally roll-off at a much slower rate of 12 dB/octave. Also the air spring inside the box will help protect the driver from over-excursion, so a high pass filter is a lesser requirement. Sealed subs will therefore have more low frequency energy for the PVG to work with and there will be more augmentation the lower the frequencies go.

So that is why sealed subs in a sealed room can have greatly augmented low frequency response. However, it often takes 2, or 3, or even 4 sealed subs to make the process work, and to get enough output for humans to hear and feel the ultra-low bass. It also takes strong enough amplification and enough driver excursion to actually move the amount of air required for pressurization at very low frequencies. Remember that it takes 4x the Volume Displacement to go down one octave. So it takes 4x the Volume Displacement to go from 40 to 20 Hz and 16 x the Volume Displacement to go from 40 to 10 Hz. Then consider that a 6 dB increase in SPL requires a doubling of driver excursion and power, and you can see that to hit 115 dB at 10 Hz will require a significant number of high excursion drivers plus the power to push those drivers to full excursion. In my room, it takes (6) 15" drivers and 7,200 watts of power... AND a significant amount of pressure vessel gain... to hit 115 dB at 10 Hz.

The same thing can be done with ported subs, but it requires very low port tuning, which in turn means very large box size. And to hit 115 dB at 10 Hz, you may still need at least two large, low tuned subs to get there.

Anyway, I encourage you to try sealed mode on your subs. However, I don't recommend mixing sealed and ported subs... either all sealed or all ported, and set to the same F3.

Good luck!

Craig
LGWGM likes this.

Lombardi said it:
"Perfection is not attainable, but if we chase perfection we can catch excellence.
My System (Edited Feb. 2020 to add 4K and Atmos updates)
craig john is offline  
Sponsored Links
Advertisement
 
Reply Subwoofers, Bass, and Transducers

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off