REW Graph - Before & After BFD - SVS PCI 20-39

**Dent** · **05-20-06, 01:22 AM**

Using the new RS digital SPL meter calibration file and REW, SVS PCI 20-39, crossover on my Yamaha receiver is fixed at 90 Hz, here are my results:

Before - Sub only:

start - sub only.jpg

After - Sub only:

end - sub only.jpg

After - Sub & Lft/Rt Mains:

end - sub & mains.jpg

All graphs together:

all together.jpg

**Sonnie** · **05-20-06, 11:33 AM**

Hey Dent... I think you did a great job.

How does it sound to you now?

You might want to look at setting up a house curve too and see how that sounds to you for movies. Save it on a different preset so that you can switch back and forth.

EDIT: You could almost have a house curve if you raise your measurement level a couple of db to say 82 and then drop your target response down to say 77db and leave your filtering in place as it is up to 30hz and pull from 31hz on down to the target line. You would be eliminating some of the boosting you're doing too. This wouldn't be quite as steep as most house curves but it might be one to try on one preset. You could load a house curve *.txt file too and play around with it on another preset.

If you wanna upload your *.mdat file for that first graph as an attachment here, I'll tinker with it and show you some examples of what I'm talking about.

**Dent** · **05-20-06, 02:14 PM**

Thanks for offer of help, Sonnie. I'll leave a link for my initial *.mdat file. I couldn't upload it here because you have a 500 KB limit for the .mdat file and mine is 751 KB. Maybe you could increase that?

The link is here: http://www3.telus.net/warner/initial.mdat

I would actually like some help if you could just getting this response more flat without all the undulations. I used the REW suggested filters which knocked down the peaks and then I manually used boosting filters to raise the dips but obviously not as flat as I wished. You seem to have the knack for it so I thought I would ask. You could also show me some examples of a house curve that you were talking about.

Thanks,

Warner

**brucek** · **05-20-06, 05:28 PM**

Quote:

help if you could just getting this response more flat without all the undulations

Looks like you don't have the Pre-Impulse response window in the Analysis pull-down set to Tukey 0.25 and after the measurement was taken did you set the Post-Ref Window Width (ms) from the default of 500ms to 125ms?

brucek

**Dent** · **05-20-06, 07:37 PM**

Quote:

brucek wrote:

Looks like you don't have the Pre-Impulse response window in the Analysis pull-down set to Tukey 0.25 and after the measurement was taken did you set the Post-Ref Window Width (ms) from the default of 500ms to 125ms?

brucek

Hi Bruce,

Actually, I did follow the instructions on this site and the Pre-Impulse response window was set to Tukey 0.25 before measurement and after the measurement I also did set the Pre-Ref Window Width to 125 ms. In your comment however, you talked about setting the Post-Ref Window Width from the default of 500 ms to 125 ms. The instructions on this site say to change the Pre-Ref but not the Post-Ref. Are the instructions on this site wrong?

Warner

**Sonnie** · **05-20-06, 07:38 PM**

Yeah... check out this post on your settings. It would be better if I had the raw *.mdat file of just your first measurement with the correct settings saved to it.

You can re-post your *.mdat file as an attachemtn after you get it setup correctly. I increased the limit for the .mdat extension.

**Dent** · **05-20-06, 07:46 PM**

Quote:

Sonnie wrote:

Yeah... check out this post on your settings. It would be better if I had the raw *.mdat file of just your first measurement with the correct settings saved to it.

You can re-post your *.mdat file as an attachemtn after you get it setup correctly. I increased the limit for the .mdat extension.

Hi Sonnie,

As I replied to Bruce, I'm sure those settings he mentioned were set correctly. I thought my *.mdat file to which I posted the link did contain the correct settings. When I reload it in my REW on my computer, the Pre-Impulse Window is set to Tukey 0.25, and the Pre-Ref window width is also set to 125 ms.

I'll also load the file on your site here but the link that I posted is valid.

P.S. I just tried to upload the *.mdat file again but since the file is 751 KB and your limit is only 750 KB, it said it was too big. But again, the link I posted should suffice.

**brucek** · **05-20-06, 08:43 PM**

Quote:

The instructions on this site say to change the Pre-Ref but not the Post-Ref. Are the instructions on this site wrong?

Oops, sorry Dent, my bad. I was posting from memory instead of looking it up. You're correct. Usually this setting takes some of the squiggles out......

brucek

**Sonnie** · **05-20-06, 10:54 PM**

Okay Dent... I fiddled with it using I believe brucek's house curve.

Here's the filters I used: dent.req

Here's the house curve text file: house2.txt

**Dent** · **05-21-06, 12:05 AM**

Sonnie,

When I try to load the filters that you used nothing comes up. Is it the correct file that you uploaded?

Warner

**Sonnie** · **05-21-06, 12:37 AM**

Try it now... it didn't work for me either intially, but I re-uploaded it as a binary file and downloaded it a second time and it worked.

**Dent** · **06-03-06, 01:27 PM**

Hello all, again,

I decided to try again from scratch and see what I could come up with. I still wanted to go for a flat response as opposed to a house curve as a flat response sounds fine for me. Also this time when I added in the mains I matched the level of the mains with the subwoofer before measuring unlike before when I think the mains were 3-4 dB above the level of the subwoofer.

Anyway here is what I came up with:

My initial response. Subwoofer only:

1.jpg

After 12 BFD filters later. Subwoofer only:

2.jpg

But then when I added in the level matched mains, a huge dip developed. Subwoofer and mains:

3.jpg

I took out the two filters that I had had above the crossover (90 Hz) and then remeasured, again with both subwoofer and mains, added two filters to flatten the response and came up with this. Subwoofer and mains:

4.jpg

I then wanted to see what the response was with just the subwoofer and came up with this. Subwoofer only:

5.jpg

Even though the response is boosted above 100 Hz well above the target line I figured that with most LFE content there probably is not much material above 100 Hz so the LFE content in movies won't sound out of wack. If I didn't boost that area above 100 Hz then my subwoofer + mains would have the huge dip as shown before.

Overall, everything sounds great!!

**brucek** · **06-03-06, 01:44 PM**

Yeah, you've done exactly what you're supposed to do.

You set up your sub and filters first and then added the mains to check for interaction.

More often than not, that new dip above 100Hz could be ameliorated with the phase adjustment on the sub itself. If not, then the BFD filters can be used (as you've done) to fix it.

I'm amazed you don't find your flat response....flat sounding.

Here's a picture of the boost between my mains and sub, and it doesn't seem overpowering at all. I suspect my crossover at 60Hz has something to do with this. If I tried that boost at 80Hz it might sound boomy.

Have you tried boosting the bass and discovered you didn't like the results?

brucek

**Dent** · **06-03-06, 02:47 PM**

On my sub I only have a 0 or 180 degree phase switch so one or the other. I have it set on 0. I tried switching it to 180 degrees and remeasured but it caused two other broad dips to form below the crossover (90 Hz) so I left it at 0. When I am listening to music I leave the response flat and when I listen to movies I just raise the subwoofer level on my receiver about 3 dB or so and everything sounds fine. Maybe I am creating a type of house curve just raising my subwoofer level on the receiver (or I could do the same level boost on the subwoofer gain control itself).

I guess that is a question I have then. Why don't you just make a flat response and then increase the level of the subwoofer over the mains using the receiver sub level control or the actual sub gain control when needed such as when watching movies but when listening to music, decrease the level so that it is flat, which I read that a lot of people do? Isn't this what most people do anyway, i.e. increase the level of the sub 3-4 dB hotter than the mains for watching movies?

I found that with the house curve it might sound fine at a certain volume but when I listened at a louder volume it was too much. I guess if someone always watches movies at a similar volume then that particular house curve would work but I find I am watching movies or listening to music at various volumes depending on who else is watching or who else is in the house at the time.

**brucek** · **06-03-06, 04:34 PM**

Quote:

Why don't you just make a flat response and then increase the level of the subwoofer over the mains using the receiver sub level control

Modifying the receiver subwoofer trim control is not really a good idea once you have the input level to the BFD set. See this section in the guide.
You could adjust the subwoofer amplifier level control though.

Most people simply have two BFD channels of filters set up for music or movies, so it's easy to switch between thetwo..

brucek

**Wayne A. Pflughaupt** · **06-03-06, 04:42 PM**

Quote:

I guess that is a question I have then. Why don't you just make a flat response and then increase the level of the subwoofer over the mains using the receiver sub level control or the actual sub gain control when needed such as when watching movies but when listening to music, decrease the level so that it is flat, which I read that a lot of people do?

I can’t speak for everyone, but I found with music, when the sub was flat, that the bass notes got weaker the lower they went. That’s why I curved response upward, so that the relative volume of bass notes remained constant as they went lower. Everyone’s room is different, however. If you’re getting equal fullness with both high and low bass notes with flat response, then you’ve achieved the objective.

Quote:

Isn't this what most people do anyway, i.e. increase the level of the sub 3-4 dB hotter than the mains for watching movies?

Typically so, often even more, depending on the size of the room.

Regards,
Wayne

**Josuah** · **06-03-06, 08:53 PM**

I think I read the house curve article posted here and it talked about how it used pink noise. Pink noise deals with the subjective difference in sound for us humans, but using that as the basis of your calibration is sort of like calibrating for what our brains tell us makes sense to hear, versus what a measurement mic tells us makes sense to hear. Sort of. That isn't very clear at all.

Anyway, the point is as frequency increases volume should decrease if you want your brain to think the volume is actually the same for all frequencies, when those frequencies are played together. That's why white noise, where all frequencies are played back with equal energy, sounds like there is too much high frequency information in it. Whereas pink noise sounds like all frequencies are represented equally.

Advertisement