Home Theater Forum and Systems banner

1 - 17 of 17 Posts

·
Registered
Joined
·
41 Posts
Discussion Starter #1
Hi all.

I am having some real trouble with REW and getting my subwoofers in phase with my mains in my home theatre i was hoping to get some help with ?

I have my subwoofers crossed over at 90hz in my AVR. And with a distance setting of 6m. Mains have a distance setting of 3.4m. Measured with REW the FR looks alright. But it sounds wrong. Like the subwoofers sound it out of time with the mains by half a second. When listening to some music with a repetitive bass beat i can her the beat coming from the mains first then the subs second. So i went into my AVR and tried setting the subwoofer distance setting out further till it sounded right. So i ended up setting the subwoofers out to 18m and now it sounds exactly in time with the mains. Only problem is when checking with REW it has done some horrible things to the FR now with a massive dip at 70Hz and an overall messy looking FR.

Now i am really confused what's going on and have no idea where to go from here. If i adjust the sub distance to get the smoothest FR at the Xover i can clearly hear its out of time with the mains. If i set it up so it sounds in time with the mains it measures really badly with REW ?

Any help would be greatly appreciated.
Thanks,
Tim.
 

·
Premium Member
Joined
·
1,562 Posts
What are actual physical distances from the LP to the SW and mains?

Normally entering the measured distances into the AVR works pretty well. Often an optimized setting is found with the SW distance 0.3 to 0.5m greater, but that amount of error is not really significant. If there is a deep null at the XO freq with that setting it is necessary to invert the polarity of the SW so that it properly matches the mains.

It is often recommended here to use the REW RTA feature and adjust the SW distance if needed from that original distance until the most SPL in the XO range is created. You should not need to move the SW distance more than maybe 2-3m from the measured distance at the very most.

If that has not resulted in a good sound then there may be room modes that are causing the issue.

If you can get a relative flat SPL through the bass range then that's great. If not it may be necessary to move the SW, or the LP or use some EQ.

There is plenty of good advice to be found on setting up a SW so I assume you have seen some of those. If you have more specific questions we can probably help.
 

·
Registered
Joined
·
41 Posts
Discussion Starter #3
Actual physical distances from LP to SW is 3.2m. LP to mains is 3.4m.

The subwoofers are infinite baffle type mounted in a pair of manifolds in the floor directly in front of each main speaker.

I didn't think to try flipping the subwoofer phase 180. Might try that next. Only its not at the Xover that the null appears. It appears 20Hz below the Xover ?

Cant move the subs as they are permanently cut and mounted into the floor. LP can only be moved about 100mm at most back. Then it starts to get behind the side surround speakers. Cant go forward either as my projector screen would be too close. Its a small room 5m x 3.5m. So cant do much with object placement.

The thing is if i have my subwoofer set to say 2 or 3m more than the measured distance i can clearly HEAR and FEEL the subwoofers are delayed compared to the mains. Its not till i set its distance out a LONG way out to 15m more than the measured distance that the sound from the subs feels and sounds to be in time with mains. Its very strange.
 

·
Registered
Joined
·
41 Posts
Discussion Starter #4 (Edited)
A few pics to show what i am seeing in REW. This is with subs and the right main playing.
together
The first one is with Xover set at 90Hz. Sub distance at 8.7m and reverse phase. Thats the best FR i could get.

The two others are to show what happens to the FR when the sub distance is pushed out further. One at 16m distance with green at normal phase and brown at reverse phase. The next one is similar but at 18.10m distance.

Notice how when switching the phase its not just affecting things at the Xover at 90Hz. But its doing some nasty things around 60Hz and 70Hz ????? I thought it was only supposed to affect things at the Xover ? It seems to be messing with everything between 50Hz and 130Hz ???
 

Attachments

·
Premium Member
Joined
·
1,562 Posts
The SPL looks reasonably good in the charts above. It appears that there may be some smoothing applied to the graphs however and that is not recommended. The graph scaling is also non-standard so you should read the sticky thread on this forum about posting graphs.

For a 90Hz XO the "XO range" is probably about 50-180Hz with most of the impact nearer 90Hz where both the main and the SWs are playing at about the same SPL. As we move away from 90Hz the SPL is fading away for one of them so the impact is less.

There are peaks and dips in the SPL due to "room modes" that can be at any bass freq including in the XO range. It depends on the dimensions of the room and the LP and speaker positions.

The physical distances setup I described above is the "conventional" setup target that most all guides and automated setup systems target. This provides the best phase tracking of the 2 drivers through the XO range. Using that as the base setup there are also good alternate setups that can be found at SW distance settings of additional 1/2 wavelengths. So since the SW distance is 3.2m and the XO is 90Hz (3.8m wavelength) other good setup would be with a SW distance near 5.1m and 7.0m. Even these alternate settings will sound a little different even though the SPL is about the same.

Distances beyond that would be expected to deteriorate the sound further even though the SPL is still relatively smooth. If you prefer the resulting sound then there is nothing wrong with leaving it that way. We would suggest that you try a conventional setup for a while just to be sure it isn't something that just requires some time to get used to.
 

·
Premium Member
Joined
·
3,374 Posts
Did you try to use 80Hz as a crossover along with changing the phase. You might give that a try. Also I agree that setting the actual distance manually might help.
 

·
Registered
Joined
·
41 Posts
Discussion Starter #7 (Edited)
Yes i did try a 80Hz Xover and i think it was the same from memory. I could try it again but im pretty sure it didn't change things that much. Now setting it at 120Hz did change things a bit but i don't know if you would want the sub to Xover that high ?

As for the graph scaling do you mean its starting position at 10Hz instead of the standard 15Hz ? No smoothing unless there is some in REW as default ? Yeah i was just interested to see what information was down that low and forgot to re-size the graphs after. Basically all other settings in rew for the graph scaling are at default settings.

I have a SVS AS-EQ1 (Audyssey XT32) subwoofer EQ that i have been using to EQ the bass and it says to set a subwoofer distance of 6m.My AVR is a Yamaha so no Audyssey in that. I have had things like that for the past 3 or so years so am pretty used to it setup that way. But have always been upset with how the bass sounded out of time with the mains. So i have got back into REW to do some investigating to see what setting subwoofer distance settings has on SPL at the Xover to try and find the best spot for it. And its hard. So much is changing its almost impossible to find the best spot.

I ran the RTA feature of REW last night and it was next to impossible to see what was happening as it was jumping around all over the place and couldn't get a clear picture of the FR unless i ran a measurement instead.
 

·
Premium Member
Joined
·
1,562 Posts
Your SPL graphs are fine. I just am used to looking at 5dB grid lines about 1cm apart. It didn't even notice you were set at the recommended 105-45dB. I just assumed you were greater than that. That's my mistake. A slightly higher graph height and the 5dB grid lines would have appeared at the correct spacing.

You mentioned a manifold for the SWs. Are the drivers recessed significantly from the outlet at 3.2m? If so the distance should be the total distance to the drivers. Does each manifold containing 2 12" drivers? Are the 2 sets of 2 drivers driven as mono or L and R?

I would review a .mdat file if you want to attach one. It should contain 5 measurements assuming 2 SWs in a mono configuration:
1. SWL
2. SWR
3. SWs (SWL + SWR)
4. L
5. SWs + L

> If you have the capability to measure all these with REW "loopback timing" activated that would be the best. If you are using a USB mic that is not an option. I may not be able to give definitive answers without loopback timing, but I may still be able to provide some good direction.
> All sweeps should be from 10-20k Hz, even for the SWs. [10-3k is an option.]
> The AVR XO should be active (use your preferred XO freq).
> The AS-EQ1 should be active.
> If the SW are not mono driven, the "SWs" and SWR (measurements 2, 3) should be dropped from the measurements.
> I suggest you use the 6m SWs distance setting that the AS-EQ1 recommends. I would expect that unit to pretty accurate most of the time.

With this measurement file I would be able to confirm that the SW are properly working together and also working well with the L main. With loopback activated, I would be able to give a definitive distance setting for the conventional setup and compare that to the 6m setup. I would also be able to look at the case of the 15m distance setup. If the room setup is pretty symmetrical then the R main would be expected to be pretty similar. If there is significant asymmetry then after the L main is settings are determined we should also look at SWs, R and R+SWs to assure that all is well.

Just a minor note: The increased SW distance setting you prefer is actually making the SW sound arrive before the mains rather than later than the mains as you were thinking. More distance = less delay.
 

·
Registered
Joined
·
41 Posts
Discussion Starter #9
Ok got the .mdat with all the above measurements for you. Thanks for taking the time to look at it for me. I don't really know what all the results mean or what to do with them. Loopback enabled 6m subwoofer distance set. 90Hz Xover Subwoofers in normal phase.

Yes each manifold contains 2 subwoofer drivers. The top edge of the drivers are about 50mm from the top of the manifold. Too small an amount for the AVR to set.

The subwoofers are mono. So working as one. Room is totally symmetrical.

Yeah i realize increasing the subwoofer distance will mean the subs will play before the mains. That's what i was trying to do. As the subwoofers have always sounded to be behind the mains in time when they play music you can really notice it.

Thanks,
Tim.
 

Attachments

·
Registered
Joined
·
41 Posts
Discussion Starter #10
Couple pics to show subwoofer manifolds setup. Pictures are worth a thousand words and its eaiser to show than typing.



 

·
Premium Member
Joined
·
1,562 Posts
Very nice room and setup.

Basic Findings:
Your measuring system setup all looks very good and a nice job was done on the measurements.

SWL and SWR are ideally working together. The polarity and timing is identical between them and the SPL is properly reinforced across the entire active SW range. It doesn't get better than this. See below:
sws spl.png
sws ir.png

XO timing is ideal. The phase tracking through the XO range is as good as it gets. Any other distance setting will depart from this "conventional" setting.
Note that the "acoustic XO" is about 74Hz. The SPL across the "acoustic XO" range (40-160Hz) is well supported, i.e., the SW+L SPL fill in for the individual responses across the entire XO range. See Below:
xo spl.png
xo phase.png

Other Comments:
The total delay is about 240ms?! [I shifted the IRs 239.812ms to create some of the charts.] I am not sure why there is so much delay. That delay corresponds to an LP distance of 82.7m. There is no problem with that delay for music, but it could easily create a lip sync problem for any video material unless it is being compensated for.

The SW level is unusually high. Some increase over the midrange is normal. It looks to be about 13dB. A more typical amount is maybe 0-5 dB.

The Topt of the room is very low particularly at HF. This may make the sound a little dry. See Below:
topt.png

There is about 13% distortion at 52Hz!! This is very high and is present in the SWs and not the L main. I am not sure why this bad peak is there. The rest of range is excellent, very low. See Below:
distortion.png

There is a strong 14Hz resonance in the L channel - possibly the effect of the tactical transducer? See first SPL chart above and the Spectrogram below. There is also a 108 Hz noise. Like the 14Hz resonance, it starts immediately as the signal is applied (not delayed 240ms like the rest of the audio is. It is in the SWs channel and can be seen in the IR, spectrogram and windowed SPL. I measures about 62dB which is significant. I am guessing that this again is the tactile transducer. See Below:
spectrograph.png
100hz noise SPL.png
100hz noise IR.png

There is no reason to think that R, CC, SL, SR are not well timed as well as you probably have the appropriate distances entered into the AVR. The only factor might be if that unexplained extra delay that is in the L channel is not in all the others. This is easy to test if you like. Just take a measurement with loopback timing engaged of each channel without the SWs engaged. The location of the IR peaks should all be about the same timing; around 240ms just like the L channel measured.

I moved the SW distance an additional 15m (reduced the SW delay by 83.48ms) just for a quick look at something near your preferred setup. Below we can see there is a series of dips in the SPL through the XO range. The phase chart now shows that the SWs and L no longer track through the XO range. It crosses at a very steep rate. This results in that series of nulls in the XO range. The calculated trace is identified on the charts below as "A plus B". [These 2 charts are at the very bottom after the summary. I can seem to move them up to here.]

Summary:
The setup as tested is optimized as is for timing of SWs and timing of the XO handoff from SWs to mains.
The only major adjustment I would suggest is to reduce the SWs SPL level compared to the mains to see if that is an improvement. The other items mentioned above would probably not be a major impact on the sound. The high distortion at 52Hz is probably not very noticeable, but it would be good to try to find the cause and correct it.
 

Attachments

·
Registered
Joined
·
41 Posts
Discussion Starter #12 (Edited)
Wow thanks jtalden for all that. I had no idea how to interoperate or use any of that other than the SPL graph.

To confirm the audio delay. Yes there is delay set in the AVR of 240ms to help lip syc with the video. My video signal is delayed a fair bit by the time it makes ist way through the AVR then through the Lumagen video processor and then the display itself. So i had to add in audio delay in the AVR to get them to match up.

On the subwoofers high level. Yes i like to run my subwoofers a bit hot (ie boosted in level) as they seem to be lacking otherwise. I don't know why. Maybe my hearing is not very sensitive to low frequency's or something. When watching movies with the subs turned down to a more normal level i just don't hear much from them for some reason. That's also why i got some tactile transducers so i could get more feeling from the bass without having to run the subs so hot. I have the subs set +4 in the AVR and i also have +6 in my Behringer DEQ2496 from 20Hz to 70Hz to create a bit of a house curve.

With the 13% distortion at 52Hz. I have no idea what that might be. Do you know what can be a likely cause of this ? Is it the sub drivers themselves or the sub amp possibly ? Or something in the room itself rattling ?

And the 108 Hz noise in the subs and strong 14Hz resonance in the L channel. You mentioned they start immediately as the signal is applied not delayed 240ms like the rest of the audio is. That shouldn't be the tactile transducers as they would be delayed also as they get there signal from the SW preout of the AVR. And there amp has a 40Hz low pass filter set so the 108Hz couldn't be from them either. I am not sure what it could be ??? That's got me totally stumped. Because even room related things making nose would be delayed some as the sound from the speakers is delayed so the things in the room making noise should also be delayed i would have thought ? So yeah i cant think what it might be ?

So i guess the 6m sub distance setting as recommended from the AS-EQ1 was right after all going by your findings here. I wonder why it sounds as though the subwoofers are delayed compared to the rest of the speakers ? All the distance settings for the 5 surround speakers are set to actual measured distance from a tape measure. Sub is set to acoustic distance. Measures good. Sounds delayed for some reason ? Even my tactile transducers feel delayed and they are in parallel with the subwoofers so anything the subs get fed so too do the transducers at the same time.

If anything i thought i might feel the transducers first and have to delay them. But have found (or should i say felt) the opposite. They are the Crowson transducers which are the fastest quietest transducers you can get. Not to be confused with the likes of the Buttkicker type. Totally different animal.

I probably should have turned off the transducers when taking the measurements now i think about it.
 

·
Premium Member
Joined
·
1,562 Posts
Yes, then it cannot be the Crowson causing the 14Hz and 108Hz issues since it is delayed also. It is then maybe some sort of fault in the measuring system. I did not see anything else that looked wrong and it did not seem to be in all measurements so I was originally thinking that the measuring system was not the cause. If it actually is the cause, then those artifacts can just be ignored as they would not present themselves in normal use. I really have no good ideas for you.

As for the 52Hz distortion, yes, I would first suspect some sort of vibration in the room; maybe a window, wall panel, furniture, or ?? I would not think it would be the drivers or boxes themselves as the distortion shows up in both SW channels individually as well as together.

Also, I noticed you are using REW v5.00. You should change to REW v5.01 beta 19 as there are numerous bug corrections and several nice features added.
 

·
Registered
Joined
·
41 Posts
Discussion Starter #14
Ah right. Yeah i was wondering where you got the distortion graph from as my version of REW doesn't have those. I shall get the Beta.

The 52Hz distortion could be the floor shaking. Its a wooden floor sitting on cement legs under the floor joists and at least one of the joists vibrates on its supporting leg at certain subwoofer frequencies due the the subs shaking the flooring and the pressure waves under the floor causing it to jump up and down a bit. I cant get under there to fix it though as its sealed tight and is no room to physically fit under there anyway.

I bet that's what it is.

Yeah i was thinking the 14Hz and 108Hz problems might be soundcard or mic preamp related also. Or Laptop ? As if it was something in the subs or room it should show the delay.
 

·
Registered
Joined
·
41 Posts
Discussion Starter #15
The Topt of the room is very low particularly at HF. This may make the sound a little dry. See Below:
Slightly OT but I am wondering if i might have over treated my room in the HF range now after pondering over the images a little. ?

The entire lower half of the walls from floor to ear height is lined with 1" rigid fiberglass. The entire front wall is 100% 1" rigid fiberglass. And 2 corner bass traps. Thats possibly why my RT60 is so low in the HF. What dose dry sounding sound like exactly ? And is it a bad thing ? I have always thought of a low RT60 as a good thing but never really understood the whole concept of it.
 

·
Premium Member
Joined
·
1,562 Posts
I don't really know. I never experienced it. My acoustic experience is very limited. I am only pointing out that you are below the extreme of what I have seen recommended.

Everyone agrees that neither an anechoic chamber nor a very highly resonant room is a good space. The best place in between is probably a matter of taste.

A couple of links:

acoustic_measurement_standards.pdf

http://www.linkwitzlab.com/rooms.htm
 

·
Premium Member
Joined
·
324 Posts
Covering everything with 1" fiberglass is not the best idea as it throws the room out of balance. The bass will still be completely out of control and yet the high frequency will be over corrected giving you a very slanted waterfall.

You want the rt60 to be balanced not very low in the HF and no correction in the lf that 1" can't even come close to touching.

If you were to say cover the rear wall with 6"+ and then place say 8 mill plastic so to block the HF from being sucked out that could be a start.
 
1 - 17 of 17 Posts
Top