or Connect
AVS › AVS Forum › Audio › Subwoofers, Bass, and Transducers › Seaton Sound SubMersive1
New Posts  All Forums:Forum Nav:

Seaton Sound SubMersive1 - Page 265

post #7921 of 9387
Quote:
Originally Posted by cdnbum88 View Post

Thanks for input. This is more of a feeler thought as I am very happy on what the AS-EQ1 did for my overall HT experience, but to consolidate to ONE unit or maybe AVR + AMP would free up some room in cabinet.

I should know this, but would the AS-EQ1 work if I indeed get another HP into the mix or would I then be better off or need a unit like the 4311 with dual sub outs?

 

The EQ1 will EQ 2 subs in exactly the same way as XT32 (IOW it has Sub EQ built in), so in that sense it is neutral to your decision. IMO you would be best off if you had XT32 rather than XT + EQ-1 - I went exactly that route myself so I am very familiar with what XT32 gives you over and above XT + EQ1. But there is a cost involved to move up to XT32 and only you can decide if it is worth it. Taking money out of the equation, the answer is 'yes'. Factoring money in, the answer is 'maybe'. 

post #7922 of 9387
Announcement of the Spring Iowa GTG.

http://www.avsforum.com/t/1460712/central-iowa-spring-audio-gtg-jtr-bamberg-seaton-salk-gr-research
post #7923 of 9387
Free HP Sub to all attendees?
post #7924 of 9387
Perhaps a Quiznos sub :-)
post #7925 of 9387
Quote:
Originally Posted by Newbie01 View Post

Free HP Sub to all attendees?

In that case, count me in!
post #7926 of 9387
Quote:
Quote:
Originally Posted by audioguy View Post

Quote:
Originally Posted by Newbie01 View Post

Free HP Sub to all attendees?

In that case, count me in!

Audioguy,
Don't kid yourself, the asymmetry of one more subwoofer in your room would drive you mad and you would be soon buying a sixth! tongue.gif
post #7927 of 9387
Quote:
Originally Posted by Mark Seaton View Post


Audioguy,
Don't kid yourself, the asymmetry of one more subwoofer in your room would drive you mad and you would be soon buying a sixth! tongue.gif

Probably THREE more!!
post #7928 of 9387
Quote:
Originally Posted by Mark Seaton View Post


Audioguy,
Don't kid yourself, the asymmetry of one more subwoofer in your room would drive you mad and you would be soon buying a sixth! tongue.gif

Probably THREE more!!
post #7929 of 9387
Quote:
Originally Posted by audioguy View Post

Probably THREE more!!
I don't see a problem with that...tongue.gif
post #7930 of 9387
Quote:
Originally Posted by JapanDave View Post

I don't see a problem with that...tongue.gif
No problem, but no need either.

(There Chuck, I just *saved* you some money!) biggrin.gif

Craig
post #7931 of 9387
And a Mod doesn't have to delete your post. Knew you could do it! biggrin.gif
post #7932 of 9387
So...what I got from that is... 3 free subs for all attendees?
post #7933 of 9387
Quote:
Originally Posted by Newbie01 View Post

So...what I got from that is... 3 free subs for all attendees?

You owe me half a cup of coffee! tongue.gif
post #7934 of 9387

Hi guys,

 

I have just bought a Behringer DSP1124P to use its Parametric EQ abilities to see if I can 'ice the cake' on my dual F2s in-room response and also to perfect the splice. I am using REW to create the filters.

 

Question: do you think it is better to run Audyssey first and then use the Behringer to put the finishing touches to what Audyssey has done, or to do it the other way around and do the PEQ filters first and then run Audyssey?

 

This is my first venture into PEQ and I am not sure I can improve on what I already have, but as the 1214P was a very cheap purchase (bought it used) it's worth a shot. If I can’t improve things any, I will just re-sell the 1214P and stay with my current setup as determined by XT32 and Audyssey Pro. 

 

Any advice gratefully received.  Thanks.

post #7935 of 9387
It won't hurt to try but conventional wisdom say use the PEQ first, then Audyssey.
post #7936 of 9387
Quote:
Originally Posted by kbarnes701 View Post

Hi guys,

I have just bought a Behringer DSP1124P to use its Parametric EQ abilities to see if I can 'ice the cake' on my dual F2s in-room response and also to perfect the splice. I am using REW to create the filters.

Question: do you think it is better to run Audyssey first and then use the Behringer to put the finishing touches to what Audyssey has done, or to do it the other way around and do the PEQ filters first and then run Audyssey?

This is my first venture into PEQ and I am not sure I can improve on what I already have, but as the 1214P was a very cheap purchase (bought it used) it's worth a shot. If I can’t improve things any, I will just re-sell the 1214P and stay with my current setup as determined by XT32 and Audyssey Pro. 

Any advice gratefully received.  Thanks.

Hi Keith,

I haven't used the Behringer device, but I have used the Velodyne SMS-1, (which is also an external PEQ), in conjunction with Audyssey. My experience in my own system was that running the PEQ post-Audyssey yielded the best results. (In fact, you have referenced that result in your FAQ.) That result was based on pretty decent pre-Audyssey FR, with no huge peaks or dips that Audyssey had to deal with. The SMS-1 was only used to tweak the final FR a few dB after running Audyssey. For that purpose, it worked extremely well.

However, in another system, I found that using the SMS-1 pre-Audyssey to reduce a huge 25+ dB peak was a significant benefit to the final Audyssey result, Without the huge cut implemented by the SMS-1, Audyssey was expending all it's resources to remove that peak and was therefore less effective at other frequencies. Audyssey has, (IIRC), about 9 dB of cut capability at any one frequency. If that is not close to enough cut for a huge 25+ peak, an external PEQ can be used effectively before running Audyssey to improve the response Audyssey "sees" prior to it's filters. The better the FR Audyssey is given, the better the final result.

In addition, and more importantly, the "Volume Normalization" process was negatively affected by this huge peak and lead to the well-documented ULF boosting. Reducing the peak with an external PEQ reduced the need for Audyssey to increase the overall level, which reduced the ULF boosting seen with the Volume Normalization process.

Bottom line, I think the best answer depends on the pre-Audyssey FR. If your system is pretty good to start with, Audyssey by itself can be highly effective, and the external PEQ can be used post-Audyssey to "tweak" the final result to good effect. However, if your pre-Audyssey FR has a single huge peak that dominates the response, the external PEQ can be used pre-Audyssey to great advantage to reduce that peak.

Craig
post #7937 of 9387
Quote:
Originally Posted by craig john View Post

Quote:
Originally Posted by kbarnes701 View Post

Hi guys,

I have just bought a Behringer DSP1124P to use its Parametric EQ abilities to see if I can 'ice the cake' on my dual F2s in-room response and also to perfect the splice. I am using REW to create the filters.

Question: do you think it is better to run Audyssey first and then use the Behringer to put the finishing touches to what Audyssey has done, or to do it the other way around and do the PEQ filters first and then run Audyssey?

This is my first venture into PEQ and I am not sure I can improve on what I already have, but as the 1214P was a very cheap purchase (bought it used) it's worth a shot. If I can’t improve things any, I will just re-sell the 1214P and stay with my current setup as determined by XT32 and Audyssey Pro. 

Any advice gratefully received.  Thanks.

Hi Keith,

I haven't used the Behringer device, but I have used the Velodyne SMS-1, (which is also an external PEQ), in conjunction with Audyssey. My experience in my own system was that running the PEQ post-Audyssey yielded the best results. (In fact, you have referenced that result in your FAQ.) That result was based on pretty decent pre-Audyssey FR, with no huge peaks or dips that Audyssey had to deal with. The SMS-1 was only used to tweak the final FR a few dB after running Audyssey. For that purpose, it worked extremely well.

However, in another system, I found that using the SMS-1 pre-Audyssey to reduce a huge 25+ dB peak was a significant benefit to the final Audyssey result, Without the huge cut implemented by the SMS-1, Audyssey was expending all it's resources to remove that peak and was therefore less effective at other frequencies. Audyssey has, (IIRC), about 9 dB of cut capability at any one frequency. If that is not close to enough cut for a huge 25+ peak, an external PEQ can be used effectively before running Audyssey to improve the response Audyssey "sees" prior to it's filters. The better the FR Audyssey is given, the better the final result.

In addition, and more importantly, the "Volume Normalization" process was negatively affected by this huge peak and lead to the well-documented ULF boosting. Reducing the peak with an external PEQ reduced the need for Audyssey to increase the overall level, which reduced the ULF boosting seen with the Volume Normalization process.

Bottom line, I think the best answer depends on the pre-Audyssey FR. If your system is pretty good to start with, Audyssey by itself can be highly effective, and the external PEQ can be used post-Audyssey to "tweak" the final result to good effect. However, if your pre-Audyssey FR has a single huge peak that dominates the response, the external PEQ can be used pre-Audyssey to great advantage to reduce that peak.

Craig

 

Craig - thanks for that I was hoping you might respond. That all makes sense and is very useful. In my system then I think the most benefit will be gained by running Audyssey first and then using PEQ as the icing on the cake afterwards. Audyssey does a pretty good job here and the room is well treated so it shouldn’t have to work too hard anyway. This is just an experiment which has come out of my continuing endeavours with REW (over at the 'REW Simplified' thread). If I cannot improve things I will re-sell the Behringer (bought for 70 bucks off ebay) and stick with Audyssey alone. My main endeavour ATM is to try to perfect the splice, which is the only thing that is not quite how I want it right now. I have much to learn about parametric EQing and HTS threads have been very helpful to me so far, as well as AVS of course. Deeper and deeper into the rabbit hole went Alice....

post #7938 of 9387
Quote:
Originally Posted by kbarnes701 View Post

Craig - thanks for that I was hoping you might respond. That all makes sense and is very useful. In my system then I think the most benefit will be gained by running Audyssey first and then using PEQ as the icing on the cake afterwards. Audyssey does a pretty good job here and the room is well treated so it shouldn’t have to work too hard anyway. This is just an experiment which has come out of my continuing endeavours with REW (over at the 'REW Simplified' thread). If I cannot improve things I will re-sell the Behringer (bought for 70 bucks off ebay) and stick with Audyssey alone. My main endeavour ATM is to try to perfect the splice, which is the only thing that is not quite how I want it right now. I have much to learn about parametric EQing and HTS threads have been very helpful to me so far, as well as AVS of course. Deeper and deeper into the rabbit hole went Alice....
Hi Keith,

If you are trying to improve the splice between the speakers ans sub(s), that presents a unique set of issues, and a parametric EQ may, (or may not) be the best tool. It depends on where the problem(s) are. If the problem(s) are in the time domain, then time aligning the speakers and sub(s) is the best approach for correcting those issues. I know you're already aware of the Subwoofer Distance adjustment technique to optimally time align the speakers and sub(s), and that is the first place you should start. However, if the issue(s) are in the frequency domain, (i.e., peaks and nulls in the FR from C & D interference), then the PEQ may be able to help... but only on the subwoofer channel. Remember that the PEQ is only applied to the subwoofer channel; it has no effect on the main channels. Therefore, if the problem(s) reside *above* the crossover, the PEQ will have little to no effect on them, (even if you set filters on the subwoofer channel that are above the crossover.) This may seem obvious, but it took me a while to recognize it with my SMS-1.

Nonetheless, if you are having trouble *above* the crossover, the first thing to try is a higher crossover. Re-routing that bass to the sub will change how it is "seen" by the room and how it is reinforced or cancelled by the room. It may be all you need to do. If that doesn't correct things, (or if it introduces localization problems), then the next thing to do is to find the offending speaker(s) and adjust their position(s) to alleviate the problems. Measure each speaker in isolation, with the crossover set, but without any subs. Then measure them again in pairs, (L + C, L + R, C + R), and then combined, (L + C + R). Look at the different roll-off points and the effects of the room on each speaker and each combination of speakers. Comparing those measurements may ferret out the problem cancellations. Once you have this info, you can use it to adjust the speaker placements to improve things as much as possible. However, if speaker placement adjustments are not possible, and the problems remain in the speaker channels, the PEQ won't help you, at least not much. (Re-running Audyssey using different mic placements *might* help as it will set different filter taps on the mains and subs.)

The only time the PEQ would be useful to fix things around the crossover splice is if the problems remain *below* the crossover. Then you can use the PEQ to adjust the subwoofer channel to try to improve things. Still, that's only one side of the splice and the PEQ can't do anything to the speakers, so you have limited adjustability there also. Bottom line, a PEQ is generally not the best tool for adjusting the splice between the speakers and sub(s), and should probably only be used a the "tool of last resort", when all else has been tried and failed.

Good luck. smile.gif

Craig
post #7939 of 9387
Quote:
Originally Posted by craig john View Post

Quote:
Originally Posted by kbarnes701 View Post

Craig - thanks for that I was hoping you might respond. That all makes sense and is very useful. In my system then I think the most benefit will be gained by running Audyssey first and then using PEQ as the icing on the cake afterwards. Audyssey does a pretty good job here and the room is well treated so it shouldn’t have to work too hard anyway. This is just an experiment which has come out of my continuing endeavours with REW (over at the 'REW Simplified' thread). If I cannot improve things I will re-sell the Behringer (bought for 70 bucks off ebay) and stick with Audyssey alone. My main endeavour ATM is to try to perfect the splice, which is the only thing that is not quite how I want it right now. I have much to learn about parametric EQing and HTS threads have been very helpful to me so far, as well as AVS of course. Deeper and deeper into the rabbit hole went Alice....
Hi Keith,

If you are trying to improve the splice between the speakers ans sub(s), that presents a unique set of issues, and a parametric EQ may, (or may not) be the best tool. It depends on where the problem(s) are. If the problem(s) are in the time domain, then time aligning the speakers and sub(s) is the best approach for correcting those issues. I know you're already aware of the Subwoofer Distance adjustment technique to optimally time align the speakers and sub(s), and that is the first place you should start. However, if the issue(s) are in the frequency domain, (i.e., peaks and nulls in the FR from C & D interference), then the PEQ may be able to help... but only on the subwoofer channel. Remember that the PEQ is only applied to the subwoofer channel; it has no effect on the main channels. Therefore, if the problem(s) reside *above* the crossover, the PEQ will have little to no effect on them, (even if you set filters on the subwoofer channel that are above the crossover.) This may seem obvious, but it took me a while to recognize it with my SMS-1.

Nonetheless, if you are having trouble *above* the crossover, the first thing to try is a higher crossover. Re-routing that bass to the sub will change how it is "seen" by the room and how it is reinforced or cancelled by the room. It may be all you need to do. If that doesn't correct things, (or if it introduces localization problems), then the next thing to do is to find the offending speaker(s) and adjust their position(s) to alleviate the problems. Measure each speaker in isolation, with the crossover set, but without any subs. Then measure them again in pairs, (L + C, L + R, C + R), and then combined, (L + C + R). Look at the different roll-off points and the effects of the room on each speaker and each combination of speakers. Comparing those measurements may ferret out the problem cancellations. Once you have this info, you can use it to adjust the speaker placements to improve things as much as possible. However, if speaker placement adjustments are not possible, and the problems remain in the speaker channels, the PEQ won't help you, at least not much. (Re-running Audyssey using different mic placements *might* help as it will set different filter taps on the mains and subs.)

The only time the PEQ would be useful to fix things around the crossover splice is if the problems remain *below* the crossover. Then you can use the PEQ to adjust the subwoofer channel to try to improve things. Still, that's only one side of the splice and the PEQ can't do anything to the speakers, so you have limited adjustability there also. Bottom line, a PEQ is generally not the best tool for adjusting the splice between the speakers and sub(s), and should probably only be used a the "tool of last resort", when all else has been tried and failed.

Good luck. smile.gif

Craig

 

Hi Craig - thanks for all that. The problem is below the XO. I have raised the XO already to 100Hz and am reluctant to go any higher, even if that would be a solution (if the problem was above the XO region). 

 

I have already done as much as I can with sub distance changes but it is not enough. WRT to speaker placement, there is so little opportunity for me there due to the very small size of the room. I realise PEQ might not be the answer, but for less than 100 dollars in total it seems like a worthwhile thing to try. I will keep the thread informed of my progress, or lack of it. I have a fairly steep learning curve to climb with the PEQ and REW's EQ section, but I am gradually, I think/hope, getting there.

post #7940 of 9387
Quote:
Originally Posted by kbarnes701 View Post

I have a fairly steep learning curve to climb with the PEQ and REW's EQ section, but I am gradually, I think/hope, getting there.
You're a fast climber, so I'm betting on you! smile.gif

Craig
post #7941 of 9387
Quote:
Originally Posted by craig john View Post

Quote:
Originally Posted by kbarnes701 View Post

I have a fairly steep learning curve to climb with the PEQ and REW's EQ section, but I am gradually, I think/hope, getting there.
You're a fast climber, so I'm betting on you! smile.gif

Craig

 

:)  Don't bet too much. And make it each way :)

post #7942 of 9387
Quote:
Originally Posted by kbarnes701 View Post


Hi Craig - thanks for all that. The problem is below the XO. I have raised the XO already to 100Hz and am reluctant to go any higher, even if that would be a solution (if the problem was above the XO region). 

I have already done as much as I can with sub distance changes but it is not enough. WRT to speaker placement, there is so little opportunity for me there due to the very small size of the room. I realise PEQ might not be the answer, but for less than 100 dollars in total it seems like a worthwhile thing to try. I will keep the thread informed of my progress, or lack of it. I have a fairly steep learning curve to climb with the PEQ and REW's EQ section, but I am gradually, I think/hope, getting there.

As Craig noted, the best tool and fix depends on the problem. If you post some measurements of the sub in isolation before Audyssey, post Audyssey, the center individually and the splice of the two that would be a great start. If you can add the results of distance adjustments, that would be even better. From your description it sounds like your issue might be dips that aren't quite filled in with Audyssey. This might be a case where PEQ before running could help a bit, especially with the headroom of the SubMersive above 40Hz.
post #7943 of 9387
Quote:
Originally Posted by Mark Seaton View Post

Quote:
Originally Posted by kbarnes701 View Post


Hi Craig - thanks for all that. The problem is below the XO. I have raised the XO already to 100Hz and am reluctant to go any higher, even if that would be a solution (if the problem was above the XO region). 

I have already done as much as I can with sub distance changes but it is not enough. WRT to speaker placement, there is so little opportunity for me there due to the very small size of the room. I realise PEQ might not be the answer, but for less than 100 dollars in total it seems like a worthwhile thing to try. I will keep the thread informed of my progress, or lack of it. I have a fairly steep learning curve to climb with the PEQ and REW's EQ section, but I am gradually, I think/hope, getting there.

As Craig noted, the best tool and fix depends on the problem. If you post some measurements of the sub in isolation before Audyssey, post Audyssey, the center individually and the splice of the two that would be a great start. If you can add the results of distance adjustments, that would be even better. From your description it sounds like your issue might be dips that aren't quite filled in with Audyssey. This might be a case where PEQ before running could help a bit, especially with the headroom of the SubMersive above 40Hz.

 

Hi Mark, 

 

Thanks - I will post the measurement graphs you ask for later today. I think you have summed it up well - dips that aren't filled in by Audyssey is what I believe the problem is. Unfortunately, the dip is over a crucial frequency range - about 50Hz to 150Hz. I am going to re-run Audyssey Pro this morning to get a 'clean' Audyssey calibration and I will be able to give you measurements of the Audyssey-set sub distances and also my tweaked distances to try to improve the splice. 

 

I'll get back to you later. Thanks for taking a personal interest in this. Much appreciated.

 

Keith

post #7944 of 9387

Mark - here are, I think, the graphs you want to see. If I have missed anything vital please let me know and I will find them or make new ones.

 

It has taken some time. I re-ran Audyssey Pro today, using my 'previous best' set of mic positions - mainly clustered around the one and only important seat but with a couple on the other seat too. I use 9 positions in all. I am confident I am using Audyssey properly and that that is not the problem.

 

Audyssey set my sub distances to 10.6ft and 9.6ft. All references to pre-tweak distances are these. I experimented for a long time with different distances but without a lot of success. The best alternative I could manage with my centre channel + subs was 15.0 and 14.0 feet (all references to post-tweak are these). This gave a marked measured improvement for the C + Subs but was very detrimental to the R+L+Subs measurement. The best compromise seemed to be to go back to 10.6ft and 9.6ft (subs 1 and 2 respectively).  See if you can make anything of my graphs - my head is spinning. :) All full range graphs are 1/6th smoothed and all LF graphs are unsmoothed (except for one which is easier to understand with 1/6th, but is covered elsewhere unsmoothed too (I hope).. I really don't know where to go next or what to do next. I am hoping that the Behringer PEQ will allow me to smooth the subs out and then that that will help everything, including Audyssey. All the graphs have a description in them at the top.

 

Here goes:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I did a good listening test after finishing the graphs (using the pre-tweak settings) and it sounds terrific. I hope I haven't caught Graphitis Nervosa ;)  But I am sure there is room for improvement. Thanks for your help with this - it is much appreciated.

post #7945 of 9387
Keith,

Whatever happened to this response:



???
post #7946 of 9387
Quote:
Originally Posted by craig john View Post

Keith,

Whatever happened to this response:



???

 

I have no idea Craig and it troubles me greatly. Either OmniMic speaks with forked tongue (unlikely) or I didn't use it properly (highly possible as this was my formative venture into measuring) or something odd is going on, I much prefer REW to OM and have taken the time to learn how to use REW properly (with much help from AVS-ers) so that isn't the problem. I am starting now to suspect that one or more of my mics is the problem but I don't see why. My Audyssey graphs look good (but as we know they are only predictions) but my Audyssey 'before' graphs look nothing like the REW graphs really (but then the Audyssey graphs are small and fairly useless and smoothed to hell and back). I have three mics - Audyssey Pro, OM and UMM6-1 (calibrated by CSL) for REW. They can't all have failed!

 

The only other thing I have done is add more treatments to the room since the above graph was made (to control some unwanted reflections). I ran Audyssey again of course after the new treatments went in, but it would amaze me if treatments could make the FR graphs worse!  And in any case, I have removed them and measured again and they make very little difference to the FR (as expected) but do make an improvement to the ETCs.

 

Something is odd and it worries me. HST, the system sounds fabulous and puts a huge smile on my face. I am just a perfectionist sort of person and if I think something is wrong I can't rest until I have identified it and fixed it. There is always room for improvement, so just because it sounds terrific is no reason (to me) to stop and rest on my laurels. Who knows, it may sound even better than terrific if I can get to the bottom of these issues. But the room sounds better than it measures ATM, and that also concerns me.

 

HST, I have to work with what I have. I know the room is not helping me, but I have treated it extensively and I am using high quality gear throughout the chain, so it isn't that. I have virtually zero placement options, which would help me no end if I had some choices there. I have Audyssey Pro, REW ... I will be interested to hear Mark's opinion when he has seen the graphs - I hope he can make some cogent suggestions for me to try. I will have the Behringer 1124P PEQ this week and am hoping that that can help me too. If I can sort out the bass, the rest will fall into place I am sure. If I can PEQ the bass to give Audyssey a better chance, then that should help too. I hope so anyway. Although there is a whole new learning curve there for me ;) Despite everything, I used Ratatouille this afternoon as one of my listening tests and suddenly realised I was watching the movie, so engrossed in the sound had I become, so it ain't all bad news. But I don't just want it to sound great, I want it to measure great too. To me, they go hand in hand.

 

All suggestions gratefully received, as always.

 

Keith

 

PS. Amazingly maybe, the system sounds better now (subjectively and memory infallibility aside etc etc) than it did when I made that OM graph - so WTF is going on?? 

post #7947 of 9387
Quote:
Originally Posted by kbarnes701 View Post

I have no idea Craig and it troubles me greatly. Either OmniMic speaks with forked tongue (unlikely) or I didn't use it properly (highly possible as this was my formative venture into measuring) or something odd is going on, I much prefer REW to OM and have taken the time to learn how to use REW properly (with much help from AVS-ers) so that isn't the problem. I am starting now to suspect that one or more of my mics is the problem but I don't see why. My Audyssey graphs look good (but as we know they are only predictions) but my Audyssey 'before' graphs look nothing like the REW graphs really (but then the Audyssey graphs are small and fairly useless and smoothed to hell and back). I have three mics - Audyssey Pro, OM and UMM6-1 (calibrated by CSL) for REW. They can't all have failed!
If you're using the calibrated mic for these graphs, I think that any mic issues can be excluded.

What smoothing was used with the Omnimic graph? Also, the scales of the Y axis seems compressed in the OmniMic graphs compared to the REW graphs, so you will visualize more detail in the REW graphs. Whether that is a good thing or not is discussed further below.
Quote:
Originally Posted by kbarnes701 View Post

The only other thing I have done is add more treatments to the room since the above graph was made (to control some unwanted reflections). I ran Audyssey again of course after the new treatments went in, but it would amaze me if treatments could make the FR graphs worse!  And in any case, I have removed them and measured again and they make very little difference to the FR (as expected) but do make an improvement to the ETCs.

Something is odd and it worries me. HST, the system sounds fabulous and puts a huge smile on my face. I am just a perfectionist sort of person and if I think something is wrong I can't rest until I have identified it and fixed it. There is always room for improvement, so just because it sounds terrific is no reason (to me) to stop and rest on my laurels. Who knows, it may sound even better than terrific if I can get to the bottom of these issues. But the room sounds better than it measures ATM, and that also concerns me.

HST, I have to work with what I have. I know the room is not helping me, but I have treated it extensively and I am using high quality gear throughout the chain, so it isn't that. I have virtually zero placement options, which would help me no end if I had some choices there. I have Audyssey Pro, REW ... I will be interested to hear Mark's opinion when he has seen the graphs - I hope he can make some cogent suggestions for me to try. I will have the Behringer 1124P PEQ this week and am hoping that that can help me too. If I can sort out the bass, the rest will fall into place I am sure. If I can PEQ the bass to give Audyssey a better chance, then that should help too. I hope so anyway. Although there is a whole new learning curve there for me wink.gif Despite everything, I used Ratatouille this afternoon as one of my listening tests and suddenly realised I was watching the movie, so engrossed in the sound had I become, so it ain't all bad news. But I don't just want it to sound great, I want it to measure great too. To me, they go hand in hand.

All suggestions gratefully received, as always.

Keith

PS. Amazingly maybe, the system sounds better now (subjectively and memory infallibility aside etc etc) than it did when I made that OM graph - so WTF is going on?? 
Well, those are very narrow nulls that may not have shown up as dramatically, (or at all), on a "smoothed" graph, such as your OmniMic graph. However, narrow nulls, while they look really bad on a graph, are much less bad sonically. A "null", by definition, is something you can't hear... something that is missing. It's far easier to "hear" a peak than to "hear" a null. A narrow, steep null is even harder to hear because it only affects a very narrow range of frequencies. You only miss it when the content is supposed to contain that narrow range of frequencies. In addition, if the harmonics fill in for the fundamental, you may not notice the absence of the fundamental at all, (the "Bose" philosophy, if you will.) And, if everything else comes through loud and clear, the totality of the content that you actually hear can still be very dramatic and good, and the articulation of individual notes can still be heard "through" the FR nulls. Frankly, that is what I think you're hearing.

In all honesty, I don't think a PEQ will solve these graphical issues for you. It is impossible to eliminate a narrow, steep null with a compensatory narrow band of PEQ boost; it simply doesn't work. You just end with with more cancellation, and a much bigger strain on your speakers and amps. Sure, the SubM's have massive headroom, but this use of it is just throwing energy into the abyss.

The only useful tools you have to affect these narrow, steep nulls are subwoofer placement and listening position adjustment. (For the nulls above the crossover, the PEQ will be even less effective and and positional adjustments will be the *only* tools at your disposal.) If you really want to "hear" what you're missing from these nulls, try leaning forward 6" or a foot. (Measure those positions, and I'll bet you'll find completely different null frequencies, and the nulls you have at your current measurement position will be completely gone.)

Bottom line, in a small to medium sized room, EQ can fix the peaks, but there is no way you can eliminate all the nulls, especially the very narrow, steep ones. Multiple subs can fill in some of them. Positional adjustments of the LP and subs can help with others. But complete elimination of all the nulls, especially the very narrow, steep ones, is almost impossible. Many people adamantly endorse these high resolution graphs with no smoothing. However, I don't think we HEAR that way. IME, 1/6th to 1/12th octave is how we "hear", and it's all the resolution one needs to optimize the sound at our LP's or in our listening areas. Using the high-resolution graphs just causes us to chase our tails trying to eliminate the un-eliminatable. (I know that's not a word.)

Where are your subs currently placed? Are they up front, between the center and the L/R mains? IME, that is usually one of the worst placement scenarios we have available. (It does *look* cool, but it usually measures poorly and doesn't sound as good as other placement possibilities.) If you can move them to opposing midwalls, or opposing 1/3 - 2/3 wall arrangement, I would expect better response. If that's not possible, do you have any adjustablity of your seating position? Sometimes moving the seating 6" or a foot off the wall can eliminate the huge severity of these nulls.

Finally, I suggest you spend more time enjoying the system than you do chasing these steep, narrow nulls that you can't eliminate anyway.

Good luck!

Craig
Edited by craig john - 3/12/13 at 5:54pm
post #7948 of 9387
Quote:
Originally Posted by craig john View Post

If you're using the calibrated mic for these graphs, I think that any mic issues can be excluded.
 

 

 

 

Hi Craig - thanks for your help and advice on this - it is very much appreciated.  Yes, the mic was individually calibrated by Herb at CSL and I am sure it is good. And of course the cal file is loaded into REW. 

 

 

BTW I should have mentioned that I am crossing over at 100Hz (on the advice of Mark). The XOs were suggested by Pro but not as first choices. The Pro first choice XO was 60Hz for LCR. I haven't re-run Pro and loaded the measurement file again to select the 60Hz XOs and then re-measured to see the impact though. I could do that if you thought it might make a difference, but I am very reluctant to set my LCR that low - they are THX Certified speakers (M&K S150s).

 
Quote:
What smoothing was used with the Omnimic graph? Also, the scales of the Y axis seems compressed in the OmniMic graphs compared to the REW graphs, so you will visualize more detail in the REW graphs. Whether that is a good thing or not is discussed further below.

 

 

The graph shown was 1/24th - I attach another below that actually has that info embedded into its file name. The compressed scale of the OM graphs does flatter I agree. I wonder if I can compress the scale like that in REW to get a better comparison...

 

Here is the OM with 1/24th smoothing - it looks very similar to the one you posted above.

 

 

 

Here is a REW graph of Centre + Subs, with the same 1/24th smoothing:

 

 

 

And here is the same graph smoothed to 1/6th:

 

 

 

The 1/24th smoothed one bears little comparison with the OM one, although these are months apart and all manner of changes may have taken place since then. I will dig out theOM kit later and do a new graph and see how it compares with the old one. The OM mic is not individually calibrated AIUI so that might explain some variations too. The area of the graph that concerns me the most is that dip from about 55Hz through to about 135Hz. I think that is an important range and as you can see it is quite wide. This is where I was hoping the PEQ would help me, by concentrating on that 55-135Hz region.

 

Here is the bass section showing 15-300Hz, with 1/24th smoothing. 

 

 

 

Here is what literally 30 seconds of work with REW can do to the graph above:

 

 

 

 

 

Here's the same graph with 1/6th smoothing:

 

 

 

 

 

 

 

And here is the predicted waterfall of that tweak:

 

 

 

 

 

I emphasis that this was 30 seconds of work and that I will be able to do better if I spend more time on it, as well as reducing the overall number of filters used - but the difference is impressive would you agree? This is why I am excited about getting the 1124P (in the mail ATM).

 

Quote:
PS. Amazingly maybe, the system sounds better now (subjectively and memory infallibility aside etc etc) than it did when I made that OM graph - so WTF is going on?? 
Well, those are very narrow nulls that may not have shown up as dramatically, (or at all), on a "smoothed" graph, such as your OmniMic graph. However, narrow nulls, while they look really bad on a graph, are much less bad sonically. A "null", by definition, is something you can't hear... something that is missing. It's far easier to "hear" a peak than to "hear" a null. A narrow, steep null is even harder to hear because it only affects a very narrow range of frequencies. You only miss it when the content is supposed to contain that narrow range of frequencies. In addition, if the harmonics fill in for the fundamental, you may not notice the absence of the fundamental at all, (the "Bose" philosophy, if you will.) And, if everything else comes through loud and clear, the totality of the content that you actually hear can still be very dramatic and good, and the articulation of individual notes can still be heard "through" the FR nulls. Frankly, that is what I think you're hearing.

In all honesty, I don't think a PEQ will solve these graphical issues for you. It is impossible to eliminate a narrow, steep null with a compensatory narrow band of PEQ boost; it simply doesn't work. You just end with with more cancellation, and a much bigger strain on your speakers and amps. Sure, the SubM's have massive headroom, but this use of it is just throwing energy into the abyss.

The only useful tools you have to affect these narrow, steep nulls are subwoofer placement and listening position adjustment. (For the nulls above the crossover, the PEQ will be even less effective and and positional adjustments will be the *only* tools at your disposal.) If you really want to "hear" what you're missing from these nulls, try leaning forward 6" or a foot. (Measure those positions, and I'll bet you'll find completely different null frequencies, and the nulls you have at your current measurement position will be completely gone.)

Bottom line, in a small to medium sized room, EQ can fix the peaks, but there is no way you can eliminate all the nulls, especially the very narrow, steep ones. Multiple subs can fill in some of them. Positional adjustments of the LP and subs can help with others. But complete elimination of all the nulls, especially the very narrow, steep ones, is almost impossible. Many people adamantly endorse these high resolution graphs with no smoothing. However, I don't think we HEAR that way. IME, 1/6th to 1/12th octave is how we "hear", and it's all the resolution one needs to optimize the sound at our LP's or in our listening areas. Using the high-resolution graphs just causes us to chase our tails trying to eliminate the un-eliminatable. (I know that's not a word.)

 

 

I agree with all of the above., other than whether PEQ can help. And I am not sure about that. So far anyway. 

 

The set of REW filters generated for the above tweak is below - the max boost I used was 9dB - that may be too much, but as you and mark say, the SubMs have massive headroom above 40Hz. If I was doing this for real I would experiment more of course and may limit the boost being applied to 6dB, for example. Mark can perhaps advise on a 9dB boost or not (remember the room is tiny so the subs have an easy time of it here).

 

 

 

Heck it's so easy I'll do it with a max 6dB boost just for fun:

 

 

 

That hump at 200Hz is irrelevant because I cross over well below that at 100Hz. Remember, 30 seconds work so it can be improved of course.

 

Here are the filters for the above graph:

 

 

The predicted waterfall for the above 6dB max boost:

 

 

 

 

Quote:
Where are your subs currently placed? Are they up front, between the center and the L/R mains? IME, that is usually one of the worst placement scenarios we have available. (It does *look* cool, but it usually measures poorly and doesn't sound as good as other placement possibilities.) If you can move them to opposing midwalls, or opposing 1/3 - 2/3 wall arrangement, I would expect better response. If that's not possible, do you have any adjustablity of your seating position? Sometimes moving the seating 6" or a foot off the wall can eliminate the huge severity of these nulls.

 

 

The subs are currently: left sub (called Sub 2 in all my graphs) is 1/3rd along left wall. Sub 1 is in front right corner. These are the only practical positions. I'd like to put Sub 1 also on 1/3rd on the right wall but if I did that I would have to exit the room through the window ;)  I can move the seats 6 inches forward though, but not much more. Might be worth a try. Easy to measure.

 

Quote:
Finally, I suggest you spend more time enjoying the system than you do chasing these steep, narrow nulls that you can't eliminate anyway.

Good luck!

Craig

 

It's not the narrow ones that bother me - I agree with you, they are probably inaudible. It's that 55-135Hz dip that bugs me. You might be right of course in what you say. Mrs Keith said to me last night "you know, before you got that mic you used to watch movies - now you seem to watch more graphs". :)  Women eh?  :)

 

I'd welcome your observations on the above (and Mark's of course too if he is reading along with us). Thanks again for your time Craig - it really is appreciated.

post #7949 of 9387
Hard to give a lot of certain input until you go back and take the separate measurements we recommended above. I assume you had already optimized the sub-speaker integration, but we can't see very well what is the result of what speaker or interaction.

I spotted a sub only measurement and post EQ in the other thread:


Is this direct into the Behringer or through the preamp? I believe you would be much happier by changing the target a bit and reducing the cut below 70Hz by about 6dB while not boosting above 100Hz.
post #7950 of 9387
Hi Kieth,



I am confused by this graph. ^ If this is just the subs, (no speakers), and the only difference between the tracings is the subwoofer Distance settings, and you kept the relative Distances the same, (1 ft.), then I would expect these graphs to be exactly the same. One would be delayed in time longer than the other, but the responses should be exactly the same. The FR's should only be different if the speakers are playing at the same time as the subs. In your case, it looks like your pre/pro is doing something different above and below the crossover than I would expect it to when changing the Distance settings by the same relative amount. (Either that, or it could be that I am misinterpreting what this graph depicts.)

Based on this graph, I plan to go home tonight and ensure my pre/pro is doing exactly what I expect it to with different subwoofer Distance settings. eek.gif I don't believe I've ever taken this exact measurement in my system.




Based on this graph, ^ it appears that Sub2 is set lower than Sub1. If you gain-matched before running Audyssey, you are no longer gain-matched after Audyssey. It looks like Sub1 is set 4 to 6 dB lower than Sub2. One thing I do after running Audyssey is "split the difference" of the trim settings so both are set to the same trim and the gain-matching is re-established. This reset generally has only a trivial impact on the combined FR, but it ensures both subs are working equally hard. Just a suggestion...




Based on this graph, ^ I'm surprised you're not one of those guys complaining that "Audyssey killed my bass!" You had a 6 dB peak at 50-55 Hz that Audyssey corrected, and from 40 Hz down, you've lost a bunch of bass. At 20 Hz, you're down about 6 dB. Losing the peak and the diminution below 40 Hz ought to make your bass sound much less powerful. Just sayin'.... What does it sound like with Audyssey off? Can you use the target curve editor in Pro to get some of that back, especially the >40 Hz stuff? IIRC, your amp doesn't have the Pgm2 setting, or you could just use that.

Craig
Edited by craig john - 3/13/13 at 9:34am
New Posts  All Forums:Forum Nav:
  Return Home
AVS › AVS Forum › Audio › Subwoofers, Bass, and Transducers › Seaton Sound SubMersive1