[Patrick C]

I've been happy with my theater - mostly. This week, however, I started playing around with ETF...big mistake.


Is this as bad as it seems:

Room.doc


NOTE IF A PASSWORD BOX POPS UP, JUST HIT CANCEL AND IT WILL SHOW YOUT HE DOCUMENT. SORRY FOR THE INCONVENIENCE.


The LFE's seem *reasonable*, but the overall freq. response (last graph) is bad. Am I using the software improperly, or is my room that bad (I'm using an RS mic with the .cal file).


I thought maybe my sub was just turned to high (causing the serious spike in the low end), but it calibrates great with test tones, and I've tried turning it all the way down....



I'm happy to provide more info if needed, and appreciate the input.

 

[BasementBob]

Patrick C

I don't suppose you could post nice safe .gif files instead of a document.

 

[Patrick C]

Sorry, I don't have the full version so I can't export the graphs. I have to copy and paste screenshots. And, since I just wiped my computer clean and reinstalled Win 2k, I haven't installed photoshop yet. I don't have any way to save the gifs.


However, YOU DO NOT NEED TO ENTER A PASSWORD to see the doc file. Just hit cancel and it will work fine.


Sorry for the inconvenience.


pc

 

[BasementBob]

Patrick C

Some versions of Windows 2000 have Microsoft Paint with the ability to save/edit .GIF and .JPG and .PNG files (as well as .BMP). You can PrintScreen and Paste into Microsoft Paint.

 

[Patrick C]

Thanks,


Here are 3 of them in jpeg (from word, to paint, to imaging)

6 slice xover at 80

Full Spectrum (xover at 120)

Low Freq., xover at 120

 

[BasementBob]

Although the amplitudes of 6slice.jpg and sub120.jpg are quite a bit different, the positions of the peaks and valleys are at the same frequencies.


What are the dimensions of the room?

(e.g. 19' x 17' x 11')

 

[Patrick C]

Bob,


First, thanks for the input. I am NOT an expert acoustitician by any means, so help is always appreciated.


My room is square 16x16x7'4" +/-.


To break up the room, I built a stage and soffit area which houses the L/C/R and sub:

See #17 for what the setup looks like


The entire stage area is covered in TS+


I also have a second row of seating on a raised platform. Finally, I have two columns (one on each side) made of MDF, and crammed with fiberglass.


Lower walls are TS+ and upper areas are poly batting.


Because the 6 slice doesn't seem to show any room modes (right???), and because the lower frequencies were consistently louder than the others, and because the LFE is moderately flat (take away the sharp spike near 75Hz), I just assumed the sub volume was too high. However, I kept dialing it down more and more with similar results.


Any thoughts on how I can be sure my readings or accurate and/or assess the problem?


Thanks,


Patrick

 

[BasementBob]

Patrick C

Your latest link has active X controls on it (probably flash), which I can't download either.

Can you draw your room, including dimensions, where your speakers are (including elevation of center of bass woofer), and where you took your ETF5 measurements.


I presume TS+ is TheaterShield Plus.

 

[Ethan Winer]

Patrick,


> Am I using the software improperly, or is my room that bad

 

[BasementBob]

 http://www.bobgolds.com/PatrickC.GIF

 

[Patrick C]

Bob,


Thanks for the chart.


here's the sketch:

Sketch


Don't laugh, I did it at work in 5 minutes. The stage is 4" high, and the matching soffit is 6" thick.


And here's the raw jpg with the photo I mentioned earlier (not sure why it's so shaky):

stage


here's 2 construction pics which give you the idea:

construction 1


Construction 2


The chart you sent makes some sense (lots of modes b/t 250 and 400 Hz - which could account for my sharp drop.


I'm still not convinced, though, that my measurements are right. a 30db linear drop from 20-4000 Hz seems insane to me.


Any suggestions to start corrective measures? Chime in Ethan...

 

[Gig103]

Can anyone respond or PM me with some learning information? I don't know what those charts mean but I would love to learn

 

[bpape]

Square is bad - unfortunate but true. You have a peak in the mid 70's. Your 2nd mode for both width and height are right at 70 while the 1st axial for height is at 77. That's the bad news.


Also, look at the number of modes listed for each 1/3 octave. According to one theory, it is desirable to have the number of modes the same or higher than the previous 1/3 octave, never less. You have several places where you have less. This results in 'bunching' in certain areas which can cause a multitude of problems.


The good news is that you can treat all of them simultaneously since they are relatively close without messing with other things too much. Also, since you're not trying to 'fix' anything down in the 20-30Hz range, corrective measures can be of relatively reasonable size.


For instance, placing a tuned trap in the center of the front wall where the wall meets the ceiling and the same on the side wall will hit that frequency range where it is strongest. A perforated absorber can easily be made that is a bit less that 6" deep.


I can't get to your pics right now for some reason. Going to your site keeps timing out. Will look again later.

 

[BasementBob]

Patrick C

Although I've run ETF5 a couple of times, and done one specific experiment with it, I'm learning here too.
Quote:

I'm still not convinced, though, that my measurements are right. a 30db linear drop from 20-4000 Hz seems insane to me.

Can you play mp3's through your speakers? If so then download a frequency sweep. Alternatively Avia and DVE DVDs have frequency sweeps on them. You may be able to hear a boost.


What does the ETF5 RT60 waterfall look like? Your sub120.jpg is gated at 100ms, so I'm wondering what the earlier response is. Perhaps what the graphs are showing is mostly a room ring. Sound travels about a foot a millisecond, so 100ms is about 100' to 113'. That's got to be bouncing around your room a few times.


Which speakers are driven using ETF5?

Have you done a close mike test (microphone less than 1" from the speaker) to guage the equipment response ?

Some of that fullspectrum.jpg chop (comb filter) may be room effects like SBIR, and some of it may be multiple speakers.

 

[bpape]

OK. Can see your plots now. Interesting stuff.


The plot with the sub crossed at 120 shows a much bigger peak at around 75 than at 38. The one with 80 xover shows them about equal. This tells me that part of the peak at 75 and 120 is due to sub position I believe. If you are not going to cross it over at 120, I wouldn't put too much stock in that. I'd set the crossover where you are going to run it and do your measurements there.


Also, based on your room dimensions and the severity of the modal overlap, that broad peak could in fact be true. Tame the 1st and 2nd modes of the width and length and the whole range would drop.


I understand what Bob is saying about using a shorter gate time but in reality, DVD's and music are going to have content longer than that by a ton and your ears are not going to chop it off after a few ms. so I would still consider this valid. It might be valuable in identifying what is speaker and what is room but ultimately, you're still gonna have to deal with it. Just a different way to look at it.


Once you flatten out the bass a bit and tame the room, I suspect that you'll be able to bring things more in line up to 4k by reducing the sub level a little.

 

[Ethan Winer]

Gig,


> Can anyone respond or PM me with some learning information? I don't know what those charts mean but I would love to learn

 

[Ethan Winer]

Patrick,


> a 30db linear drop from 20-4000 Hz seems insane to me. Any suggestions to start corrective measures? Chime in Ethan.

 

[Patrick C]

So here's the post with tone controls on, (-12db bass +12 db treble (all the way baby):

tone1.jpg


WOW! What a difference.


I tried a bunch of other things to:


1. I removed the sub (VTF-2) from the equation, by running the mains (PSb IMAGE 5ts) at full frequency. Same problems.

2. I bypassed the DSP's in my receiver (Denon 3802). Same problem.

3. Tried a sweep on AVIA and couldn't flip the dial on the SPL meter fast enough to keep up!


I'd say it's time for some bass traps and parabolic eq's...is it worth it?

 

[Patrick C]

Oh yeah, here's the RT60 with tone controls turned off completely:

rt60

 

[bpape]

OK. After looking at the pics, I see some more potential issues.


The 'alcove' that is your stage area is tough. The space between the speakers is away from a boundary but the outside is at a boundary. That could be causing some suckout due to cancellations.


The room is laid out sideways to normal configuration (front to back being the longest dimension). Sometimes you have to do this though. It just makes the rear wall reflections earlier.


Also, looking at the RT60, it appears as though you are already a bit on the dry side between 400 and 1600 but still 'wet' from 2.5k up and from 300 or so down. Trapping as discussed before will help the bottom end but you'll need to do it with 'hard' traps. You really don't want or need much more upperbass to upper mids absorbtion.


I see that the left side 'protrusion' (where the G'eye is) has a rack on the side so no trapping there . How about the other side? Any chance that is pretty open behind? Would make a great place for a bass trap for the length dimension. How much depth do you have there and is it free floor to ceiling? Are you willing to cut out the drywall and build something in and cover with cloth?


That's where I'd start looking.


The BFD could help you out I think. Peaks are more your problem (though it won't help with reverb times in the bass.) Parametrics can be put to good use to help with peaks. I think I'd try the trap route first and see what happens. Really up to you.