Home Theater Forum and Systems banner

1 - 20 of 67 Posts

·
Premium Member
Joined
·
348 Posts
Discussion Starter #1
I finally got around to filtering my SVS PB13-Ultras. I have them tuned to 15hz and applying filters with a BFD 1124p. Here's the graph of the pair of subs with Audyssey on but no BFD filters.
untreated.jpg
I created filters for a flat response and then 9 hard knee house curves in 1db steps from 2db up to 10 db. That way I have the flexibility of playing with all the curves while using the 10 presets in the BFD.
I accidentally erased my flat measurement so here are the 2db and 6 db house curves as examples of the results.
2db curve.jpg
6db curve.jpg
Ever wonder what Audyssey Dynamic EQ does to the curve? Here's a measurement of it while the BFD is bypassed. I have my target 10db house curve loaded as a comparison.
deq curve.jpg
 

·
Premium Member
Joined
·
348 Posts
Discussion Starter #4
Wayne,
I was just looking at the first graph again and noticed Audyssey is creating a hinge at 40hz. Would there be an advantage to hinging my house curves at 40hz and using less filtering, or should I leave it hinged at 30hz like I have it?
 

·
Super Moderator
Joined
·
9,252 Posts

As I understand, Audyssey's dynamic EQ function alters the curve with different volume settings. You'll probably want to keep the curve hinged in the 25-30 Hz range, but it's always best to experiment. You might like the Audyssey settings at low volumes.

By the way, it does look like you're over-equalizing. It's not necessary to "iron out" every little ripple in response. If you have filters with gains of only 1-2 dB and/or ultra-narrow bandwidth, like 4/60 or less (assuming you're using the BFD), they probably are not necessary.

Regards,
Wayne

Regards,
Wayne
 

·
Premium Member
Joined
·
348 Posts
Discussion Starter #6
Thanks Wayne,
Only the last graph has Dynamic EQ on. The first graph is purely Audyssey MultiEQ XT without DEQ engaged. In both scenarios Audyssey appears to be creating that 40hz hinge, which is fine, its easy enough to alter to 30hz. I see now its because thats where a reference line flattens out. Sometimes the obvious escapes those looking for complicated answers. You're right about my attempts to "iron out" the response. I have up to 7 filters on some of the curves, ones like 58hz and 66hz are only 4/60 wide. Are they doing any harm, or should I go back and remove them? Which frequencies would you go after, 80hz and 38hz? Or would you also do 31hz and 24hz. I'm refering to the first graph which is not using DEQ and has a flat target line.
 

·
Premium Member
Joined
·
15,054 Posts
Excuse my dumbness but what do you guys mean by "Hinge" looking at the graphs it seems odd that you loose so much SPL above 40Hz, What do you have the crossover set at? Dual 13 ultras should do much better than that.
 

·
Premium Member
Joined
·
348 Posts
Discussion Starter #8
I have my crossover at 80hz. Audyssey wanted to crossover the mains at 60hz and the surrounds at 40hz. But my speakers, both mains and surrounds are Paradigm Millenia 200's, which have very small drivers. therefore, I changed the crossover to 80hz. I'm sure Wayne can explain a hinge 1-million times better than I. But from my understanding, it is the frequency where you stop the climb of a house curve and everything below that is flat.
 

·
Super Moderator
Joined
·
9,252 Posts

In both scenarios Audyssey appears to be creating that 40hz hinge,
Hard to say whether or not Audyssey’s doing it, or that’s just the way your response looks right about there. It’d take a baseline “pre-Audyssey” graph to tell.

I have up to 7 filters on some of the curves, ones like 58hz and 66hz are only 4/60 wide. Are they doing any harm, or should I go back and remove them?
Hard to tell whether or not they’re doing any harm, but they probably aren’t doing any good either. It’s generally best to use EQ only to the extent that it makes an audible improvement.

Which frequencies would you go after, 80hz and 38hz? Or would you also do 31hz and 24hz. I'm refering to the first graph which is not using DEQ and has a flat target line.
It’s actually a pretty decent graph. The only thing really wrong is that everything below ~45 Hz is at a good bit higher level than what’s below 45 Hz (note that everything below 45 Hz is above the Target, while everything above 45 Hz is below the Target).

But even that might not be too bad, depending on if you’re using a hard-knee curve or not – you mentioned it, but only one of your graphs had it.

With the top graph, which appears to have no house curve, all it needs is a single broad filter, either to push up everything above 45 Hz, or pull down everything below 45 Hz (your choice).

Regards,
Wayne A. Pflughaupt

 

·
Premium Member
Joined
·
15,054 Posts
I have one pb13 and its graph is much more level until the 80Hz crossover, It just seems odd that yours slope downhill at 40Hz and is almost 1/2 gone by the time you get to 80Hz Something is not right in my limited knowledge. Have you played with the phase control on one of them?
 

·
Premium Member
Joined
·
15,054 Posts
Tony, as far as so much SPL being lost above 40 Hz, keep in mind that the graph is not a baseline reading; it's after-Audyssey EQ.

Regards,
Wayne
Hmmm, still seems odd as even after Audyssey it my levels still remained much more flat but you know this better than I do so It must be ok. I personally would want more SPL between 40 and 60Hz than what he is getting.
 

·
Premium Member
Joined
·
348 Posts
Discussion Starter #13
Its that sharp decline at 40hz that I was refering to as the "Audyssey hinge" as it seems to create a steep slope up to 40 hz and then flatten it after that. In my treatment of the response, Ive brought the over/under 40hz back in line with each other as indicated in graphs 2 and 3. I didn't measure the response without Audyssey on. I'll do that tonight and post it for a comparison to the Audyssey treatment.
 

·
Premium Member
Joined
·
348 Posts
Discussion Starter #14
Wayne,

I would have never tried to adjust that response with a single filter. Obviously my notions of equalizing are skewed. I lean towards a bass heavy sound, so I'll adjust my base line up and increase everything below 45hz. I thought of 10/60 as a wide bandwidth (that belief was formed because BFD's auto filters never seemed to go that high). I'll probably need to go a lot wider to raise that whole area. You'll have to influence my perception, how wide will I have to go in your estimation and where should I center it (I understand I'll have to experiment to see what works best).
 

·
Premium Member
Joined
·
15,054 Posts
Just for comparison sake this is my graph
The green line is my A/D/S ms3u sub and the yellow is the PB13U
Both readings were taken after Audyssey was done
 

·
Premium Member
Joined
·
348 Posts
Discussion Starter #16
Tony,
My subs are both corner loaded in a living room that's about 4500 cubic feet which is wide open to about another 5000 cubic feet of living area. Is it possible the corner loading is giving support to the lower frequencies while the rest is being lost to the large area volume? By the way, how do you overlay a FR graph? I spent about 10 minutes trying to figure it out yesterday and finally gave up.
 

·
Premium Member
Joined
·
15,054 Posts
My room is 4950 cubic feet, both subs are co-located in the same corner.
Corner loading will in most cases boost the frequencies but not just the low octaves even the 40-60Hz range should be up more. I dont know what to think. Do you have bass traps set up in the corner?
 

·
Premium Member
Joined
·
15,054 Posts
I should also note that playing with the phase on the subs moved the dip that you see at 70Hz to above the 90Hz range where my mains carry the range very well ( have not posted the new graph). It was actually fun playing with REW as moving the phase I could watch the dip move up or down depending on where I had the knob turned.
 

·
Super Moderator
Joined
·
9,252 Posts

By the way, how do you overlay a FR graph? I spent about 10 minutes trying to figure it out yesterday and finally gave up.
Look for the “Measured” icon. It’ll allow you to chose which measurements to overlay.

I lean towards a bass heavy sound, so I'll adjust my base line up and increase everything below 45hz.
It’s all the same. After bringing down the <45 Hz area you could just turn up the sub’s volume to get the “bass heavy sound” back.

It’s more practical to decide which route will get the best results, as far as smoother response goes. In this case I think boosting the >40Hz area will work better.

I thought of 10/60 as a wide bandwidth (that belief was formed because BFD's auto filters never seemed to go that high).
That’s because of the way REW is programmed. It’s programmed for reducing ringing caused by room modes, not response-smoothing. The former typically requires tighter filters.
I'll probably need to go a lot wider to raise that whole area. You'll have to influence my perception, how wide will I have to go in your estimation and where should I center it (I understand I'll have to experiment to see what works best).
Impossible to say. It’s best to just play with it in REW. Start with a 1/3-octave filter (that’s about 10/60 in BFD-speak) boosted 4-5 dB. Play with the bandwidth, frequency center and gain until you get things looking like you want. It’s possible it’ll require two filters spaced at something like 60 and 80 Hz, if you’re going for more of a straight-line hard knee house curve.

Regards,
Wayne
 

·
Premium Member
Joined
·
348 Posts
Discussion Starter #20
I should also note that playing with the phase on the subs moved the dip that you see at 70Hz to above the 90Hz range where my mains carry the range very well ( have not posted the new graph). It was actually fun playing with REW as moving the phase I could watch the dip move up or down depending on where I had the knob turned.
I didn't take any measurements with REW when I adjusted my phase. I just used an RS SPL meter to find the best boost. I noticed a slight turn in the phase knob gave me a 2 db boost but turning it any further would bring the dbs down.
 
1 - 20 of 67 Posts
Top