SUB PLOT - Would you add the filters?

[MakeFlat]

The attached graph shows the average curve for 3 listening positions. The effect of filters is shown in dashes. Would you add the filters as REW suggested or just lower the sub's output by 5 db?

[Sonnie]

That actually looks pretty good without filters, but I'd flatten it out anyway, just because I could. You could boost that 29Hz back up and it would still effectively be a cut. You could also bring that minor hump at 50Hz on down. Just enter a couple of filters yourself.

[Otto]

Yep, I agree with Sonnie.

[bossobass]

Just my opinion, based on many hundereds of graphs using many different subs in many different locations...

It makes little sense to EQ a response without the satellites being included in the sweep. If you EQ the sub only, you may be making the end result worse when the sats are turned on.

Using a standard 80Hz crossover point, the mains can have effect down into the 30Hz range. You should always follow the standard regimen before post EQ (with sats on):

1. Absolute phase.
2. Location.
3. Volume leveling of sats to sub(s).
4. Relative phase (or, delay settings).
5. Final volume leveling.
6. Post EQ.

IOW, the graph shown makes it impossible to suggest whether or not post EQ should be applied, and where to apply it, if at all, IMO.

Bosso

[Sonnie]

I'd say I've looked at a few graphs as well, but I can't remember one where 30Hz was affected by the mains crossed over at 80Hz. Not saying there aren't any... just saying that it's most likely be a very small percentage.

Either way, we've been making suggestions in this way since 2001, measure the sub only... equalize it and then measure with the mains to see the interaction, since most of us only have the option to eq the sub. Maybe we've been doing it backwards for 6 years ... but it's been effective.

[Wayne A. Pflughaupt]

Well, I can see Bosso’s point. I had some trouble with this myself. You EQ with the sub only, run a fresh sweep adding the mains, and see changes that need to be made. Problem is, your new sweep wants to use all-new filters - it won’t allow you to open up and modify your existing filters. You have to tweak Chart B with filters from Chart A, then sweep and check Chart C to see if the filter tweaks got what you wanted. (‘Course, it’s highly possible that REW will allow you do a second sweep using existing filters, and I just don’t know about it...)

Typically response changes when adding the mains will show about an octave above and/or below the crossover point, in my admittedly limited experience. With an 80 Hz crossover, response from the mains should be down more than 30 dB at 30 Hz compared to the sub – too low to register.

Still, I can imagine some scenarios where the mains might make an impact down there. For instance, if your mains were really too big and bass-heavy for your room, effectively flattening much of the crossover roll-out, and you had a room mode at 30 Hz on top of that. The EQ would eliminate the peak in the sub, but it would still be present (albeit to a lesser extent) in the mains, which would show up with a combined sweep. Unusual yes, but certainly not out of the realm of possibility.

Regards,
Wayne A. Pflughaupt

[bossobass]

Sorry guys, I'll be more specific...

I use amps that all have signal and clip indicator lights. When I used to use separate 1/12 octave sine wave tones, I would keep an eye on the indicator lights of the amps for the mains to see when the mains were getting enough signal from the crossover to light them up as the tones increased in frequency. That point was always somewhere in the 30sHz, not exactly 30Hz.

As far as only having EQ for the sub, the location, leveling and phase/delay tweaks all affect the interaction between the sats and subs. PEQ on the sub will also affect the interaction in that it introduces phase changes in many cases also.

Keep in mind that the stock pre/pro crossover is a 2nd order HP on the mains and a 4th order LP on the sub. Add to this the fact that the room gain and boundary gain effects are not influencing the sub up above cross, but certainly are influencing the mains below cross. You can imagine that the crossover ends up being nothing like the intended 4th order Linkwitz/Riley across 2 octaves.

It's also interesting to switch the pre/pro from stereo to 5 stereo when taking a measurement to see the effect of tweaks across the LCR, etc, especially right at the crossover point.

Maybe we could ask the OP to show a graph of the sub only, sub and mains, then sub only with EQ and finally, the sub with EQ with the mains added? Then, finally, sub with mains added with EQ applied to that combination (instead of applying EQ to the sub only first). Might be educational.

Bosso

[Sonnie]

I see your point to a point and it makes sense.

Here's where I see the challenge and I've run into this before with some of my previous systems from way back. You measure the sub only... then the sub + mains and see a 6db peak at 60Hz that appears with the mains, but wasn't there with the sub only. So you try to filter the peak at 60Hz but nothing moves... it just stays there. Of course you can try phase adjustment and it might help.

I found myself aggravated to death one day... I had a serious peak at 100Hz that I thought the sub was causing. I ultimately filtered it to -48db and it didn't budge. I was :scratching: ... then I figured it out... ahhh, my mains were still on and I wasn't paying any attention at all to that fact. I was trying to adjust something that couldn't be adjusted. If I remember correctly I was able to help that peak with the tone control on my pre-pro at the time.

But generally I've helped a lot of people adjust the sub only to flat or a house curve and the majority of the time when the sub + mains is played, it looks good. Honestly, I can't even remember the last time it was an issue. Again, I'm not gonna say that it's not an issue and I am definitely not gonna suggest to anyone that it's not necessary... they can proceed as they wish. It certainly will not hurt to measure both ways prior to eq'ing...

[Wayne A. Pflughaupt]

But generally I've helped a lot of people adjust the sub only to flat or a house curve and the majority of the time when the sub + mains is played, it looks good. Honestly, I can't even remember the last time it was an issue.

Yup – it’s not common, but it does indeed happen. The tell-tale complaint goes like this:” I cut that peak xx dB [insert ridiculously severe number] and it won’t budge!”

Regards,
Wayne A. Pflughaupt

[MakeFlat]

Thank you all for the recommendation on what to record. Okay, here is the FR of the sats and sub blended.

[Sonnie]

Actually you should still only post the graph from 15Hz to 200Hz, even when looking at the mains + sub... and post the sub only graph as well.

Btw... I just uploaded the latest RS correction values yesterday... make sure you use those.

[MakeFlat]

Sonnie wrote:

Actually you should still only post the graph from 15Hz to 200Hz, even when looking at the mains + sub... and post the sub only graph as well.

Btw... I just uploaded the latest RS correction values yesterday... make sure you use those.

Alright I have put the curves on on a graph from 15-200Hz. When I ran REW, the oldrsanalog.cal was in effect. The new calibration file RADIO SHACK 33-2050.cal was loaded just now and is now ready to use the next time I run REW.

The curves are attached:

[brucek]

When I ran REW, the oldrsanalog.cal was in effect. The new calibration file RADIO SHACK 33-2050.cal was loaded just now and is now ready to use the next time I run REW.

You don't need to redo any measurements because of the new mic calibration file though, because the soundcard and meter calibration data isn't included in the impulse response. The mic/meter and soundcard calibrations are only applied when calculating the frequency response.

So if you saved your mdat file, then to apply or remove a mic/meter or soundcard calibration for an REW measurement after it has been taken, simply load or clear the cal data as required and press the Apply Windows button (invoked when you click the IR Windows ICON) to recalculate the frequency response with a new calibration file.

brucek

[MakeFlat]

brucek, thank you very much for the time-saving tip.

[bossobass]

MakeFlat wrote:

Alright I have put the curves on on a graph from 15-200Hz. When I ran REW, the oldrsanalog.cal was in effect. The new calibration file RADIO SHACK 33-2050.cal was loaded just now and is now ready to use the next time I run REW.

The curves are attached:

Hi MakeFlat,

Thanks a lot for the effort

As I suggested earlier, and you can clearly see, the sats do indeed have influence down to the low 30sHz.

It looks as if you're running your sub about 6-8dB hot?

Is the crossover point still at 100Hz for these graphs?

Where do you have the distance set for the sub in relation to the mains?

Where is the relative phase control set?

Bosso

[Sonnie]

Actually in the real world it won't really look like that, will it? Isn't the LFE on DD and DTS tracks 10db hot? On top of that, if you add a house curve, which many people tend to do, the sub response is going to be considerably higher than the mains and thus the mains will not be effecting the response below probably 80Hz.

Either way, it still would not keep me from attempting to level the sub response as best I could.

[MakeFlat]

bossobass wrote:

Hi MakeFlat,

Thanks a lot for the effort

As I suggested earlier, and you can clearly see, the sats do indeed have influence down to the low 30sHz.

It looks as if you're running your sub about 6-8dB hot?

Is the crossover point still at 100Hz for these graphs?

Where do you have the distance set for the sub in relation to the mains?

Where is the relative phase control set?

Bosso

Bosso,

You are right - the sats' influence go down as low as the low 30's.

The sub is in phase with the sats. I checked that a while back using the THX test that is included with some DVD movies. The sub is 2 ft behind the right sat that is 9 ft from the middle listening position. As for the crossover, it is set at 100Hz.

The sub looks like it's about 6 db too hot compared to the midrange. The sub also appears to be 6db above the reference SPL of 75dB when tested by itself, with the sats disconnected (shown on my first post in the same thread). I am quite puzzled by this but perhaps this is normal - the graphs I have seen by others also have the curve hovering above the reference line, except where the sub takes a dive at 24dB/octave.

[Sonnie]

Here's an example of my response with all my levels set properly...

Red = Sub Only
Gold = Sub + Mains



In most cases with the levels set properly, this is what I see over and over and over... very little interaction below the crossover point.

If you are getting that much interaction down into the 30's, I'd suspect your levels are probably not set the same as most I've seen.

[MakeFlat]

Sonnie wrote:

Here's an example of my response ...

In most cases with the levels set properly, this is what I see over and over and over... very little interaction below the crossover point.

If you are getting that much interaction down into the 30's, I'd suspect your levels are probably not set the same as most I've seen.

Sonnie,

Perhaps it's a level issue. I am also seeing that my mains are not rolling off below 100Hz, the crossover point. They seem to roll off at 40Hz. Hmm... I wonder if it has something to do with the Denon 1604 "Direct Mode", or is it because it is a bottom of the line Denon. Question: Would it be better to use PLII Mode on the A/V Receiver when running REW? I am thinking that may be the "Direct Mode" would not have full bass management, on a low budget receiver like the 1604.

[Jerm357]

You dont have a stereo setting?

[Otto]

MakeFlat wrote:

I am thinking that may be the "Direct Mode" would not have full bass management, on a low budget receiver like the 1604.

I think you are correct that the Denon isn't doing any bass managemnt in Direct Mode, and it's probably not because of its price tag.

Generally, Direct Mode is for those that want a quick selection that will allow them to listen to audio directly without any digitizing or other overhead that could modify the signal from the original. I would imagine that your bass and treble controls are also disable in Direct Mode.

If that's the case, you should just find a mode that does apply bass management to a stereo signal (as Jerm357 suggested). That's the only way to make an accurate measurement of your mains/sub setup. If you are not seeing a rolloff in your mains at the crossover point, you are not high-pass filtering them.

[brucek]

I think you are correct that the Denon isn't doing any bass managemnt in Direct Mode, and it's probably not because of its price tag.

That's certainly apparent from the response curve supplied.

A 100Hz crossover will result in the mains having negligible effect on the response below ~50Hz. The mains would enjoy an HPF of -12db/octave from the bass management combined with the speakers natural 2nd order dropoff to result in the diagram you see below. Take your receiver out of Direct mode and use stereo to engage the bass management.



brucek

[woozy_one]

MakeFlat wrote:

The sub also appears to be 6db above the reference SPL of 75dB when tested by itself, with the sats disconnected (shown on my first post in the same thread). I am quite puzzled by this but perhaps this is normal - the graphs I have seen by others also have the curve hovering above the reference line, except where the sub takes a dive at 24dB/octave.

I've also wondered about this.

[bossobass]

brucek wrote:

That's certainly apparent from the response curve supplied.

A 100Hz crossover will result in the mains having negligible effect on the response below ~50Hz. The mains would enjoy an HPF of -12db/octave from the bass management combined with the speakers natural 2nd order dropoff to result in the diagram you see below. Take your receiver out of Direct mode and use stereo to engage the bass management.

Attachment 2931

brucek

Yes, it would be nice if the in-room roll off of the mains would look anything like the graph you posted, but the reality is that standing waves and room and boundary gain can grossly attenuate that HP filter-imposed roll off.

If you look at the graph MakeFlat posted, it seems that there is a LP filter on the sub that's much lower than 100Hz. It would be odd to me if the mains had no filter (because his receiver's 'direct' mode has no BM), but the sub did?? My pre's direct mode has BM, so I'm not familiar with one that imposes a LP but no HP.



Looking at Sonny's sub/sub+mains graph, the sub appears to be 10dB or more hot, which would swamp the crossover's intent for unity gain across the crossover region. If you properly level the sub and mains, at least leveling them at the crossover point with an 80Hz tone, first the sub only, then the mains only, then overlay a few graphs with the relative phase control at different positions, it might give a better picture of the mains influence.

I'll post a graph later in the week when I get some time to run a few sweeps.

Interesting stuff. Sorry if I'm boring anyone. Thanks again for the graphs.

Bosso

[Sonnie]

Again, the LFE, which will respond up to 120Hz, will be 10db hot on DD and DTS tracks, therefore my graph would be more indicative of real world response.