or Connect
AVS › AVS Forum › Audio › DIY Speakers and Subs › Questions about my subs response.
New Posts  All Forums:Forum Nav:

Questions about my subs response.

post #1 of 26
Thread Starter 
I just finished measuring my subs response and I would like to know what you guys think of it. It was done with the sub calibrated with Avia to match the mains, the main volume at -25 and using the test tones by JMAC 0 dB 256 kbps MP3 from Realm of excursion and measured at the listening position. My sub is a RL-p 15 in a 8.5cf ported enclosure tuned to 17hz. Now these measurements have not been EQed but Im planing on getting a BFD soon to try to level it out but what I would like to know is how bad am I? Will this be tameable with a BFD? Does this look like a normal response for a small room (11.5' x 11' x 8') and an unEQed sub tuned to 17hz? Oh yea I used a digital radio shack spl and added the correction values of....

10Hz +20.5
12.5Hz +16.5
16Hz +11.5
20Hz +7.5
25Hz +5
31.5Hz +3
40Hz +2.5
50Hz +1.5
63Hz +1.5
80Hz +1.5
100Hz +2

Any feedback from you guys would be a big help.
Thanks, Jeremy
post #2 of 26
That's a pretty big drop from 45hz to 15hz. In a small room, you should have plenty of gain. Just as a means of double checking, you might wanna download RoomEQ Wizard and do a sweep - it's much faster and less prone to any accidents. Also, you can download Sonnie's digital meter correction file, which was tested against 4 didgital meters and averaged.
post #3 of 26
Thread Starter 
Ok I got REW working on my desktop, made a rca to 3.5mm wire for the Spl and figured out the soundcard and REW setup ( only took about 2hours I just couldent figure out how to get the "line in" to read the SPl). Now all is good, Ive been practicing on my PC speakers and been getting good measurements and feel confident that tomorrow I will do the same when I move this desktop upstairs and do the real system. Should'ent be too hard, all I need is the tower,keybord and mouse since Ill be using my tv (Sony xs955 HDMI to DVI). I actually move this PC around alot since about every mounth I take it to the garage and use the air compresser to blow out all the dust ( It builds up fast because of all the fans). Now all I need to do is figure out how to post the REW graphs?
post #4 of 26
The program allows you to save jpegs of the results really easily. Go to Graph and then at the bottom it says save Graph as jpeg. You can then upload those into AVS's Gallery and link to them from there.

Make sure the sound card FR is compensated for and use Sonnie's digital meter correction file, not the "common" corrections you find on the internet. Also, I'd encourage you to experiment with opening/closing various doors in your listening room, I had to figure out just the right "recipe" to get my really flat response
post #5 of 26
Thread Starter 
Ok heres my REW graphs. It looks pretty close to the manual one. So how bad of shape am in. Will I be able to fix this responce with a BFD? Does it look like its my room thats sucking out my low end or my hardware some where down the line? This graph was taken with the sub 5dB hotter than the mains by Avia (I have the receiver set with the sub 5dBhot at -10 on the sub out, which is what I used if it matters.)
post #6 of 26
Are these taken at the seat or near the driver? It looks like the port output isn't really coming into play. Did you stuff your enclosure or just line it? Are you using some kind of port cover - of so, which one?
post #7 of 26
Thread Starter 
There taken at the seat. You know I though that might be it. I have a grill over the port, its the speaker grill from my old Polk psw450 subwoofer. Its about 13" round and covered with grill cloth. I took the measurments with it on and when I started I was going to take some with it off and forgot all about it untill I moved the desktop back downstairs. I all ways wondered about the grill because if I blow at my hand than move the grill in between I can fell almost no air hitting my hand. Heres a pics of the inside......
post #8 of 26
Thread Starter 
I did stuff the very top chamber but it is nowhere near the port opening. Heres one of the build pic so you can see what I mean by the top chamber. I lined the walls after it was built and could not reach the very top with fiberglass.
post #9 of 26
That's a little more lining than usual with the top chamber stuffed, but like you said, it's not restricting air flow to the port in anyway, so that shouldn't be it. I guess the next step would be to try measuring with the port grill off.
post #10 of 26
Thread Starter 
Heres pictures with and with out the port grill. Theres only about 1dB - 1.5dB difference between the two. At 17hz its .8dB , at 16hz its 1.2dB and at 15hz theres a 1.6dB difference. I guess the port grill is not hurting as much as I thought
post #11 of 26
Thread Starter 
Im waiting on my BFD to get here but Ive been working on my EQ plain. What do you think of just cutting every thing down till its about flat with my 20hz dB mark? Heres some graphs. Am I on the right track? Does this look like the right plain to go about my problem? My filters are

Freq Gain BW/60
46hz -7 12
76.6 -7 8
51.95 -8 29
31.5 -2 4
post #12 of 26
Originally Posted by Jerm357 View Post

What do you think of just cutting every thing down till its about flat with my 20hz dB mark?... Does this look like the right plain to go about my problem?

Without knowing much about your design, my gut reaction is to say that there should be a root cause to your response that should be addressed before resorting to EQ. In-room response sloping down towards the tuning point of a ported sub isn't typical, and it'd be best to have a better understanding of what's going on before a brute-force fix via EQ. The second thing I'd throw at you is to be careful of misunderstanding the common "cut's only" guideline. Just because you've applied nothing but cuts doesn't mean you're good... in the situation as you've posted it, you are bringing ALL of your spectrum down to the lowest point in your graph. That simply means it's going to take more master gain to get up to your listening level, so the net end result would essentially be no different than if you had simply put a boost at the low point. The "no cuts" rule is really meant to be applied to the overall response strategy. Since deep dips in response are often a result of cancellations, boosting that point doesn't help. Trying to boost such points does more harm than good. But if you could cut every single point around that dip, you'd have the same net result of boosting only that dip, and that's not the intent of the "rule". The intent is: don't try to solve a dip with EQ. Either fix the issue that is causing it (placement, treatments, etc.), or just leave it be.

A single point dip is not your issue. FWIW, in your case, if it turns out that that just happens to be your sub's response, and there's no way other than EQ to fix it, then boosting may not be so much of a problem down to the tuning point... excursion is going to be lowest at that point anyway. But what you really have to be careful of is the fact that boosting that range is also likely going to boost the range below tune. And that could result in overexcursion problems below your tuning point.

EDIT: another obvious question: where is this sub? have you tried different locations in the room?
post #13 of 26
Agreed, something is definitely amiss. Agreed on the cuts as well, applying a lot of cuts up top is essentialy the exact same thing as boosting down low. If I recall correctly, you have the sub right in front of a very large iguana tank - based on where your seat is, that may be causing some problems? I'd try taking out some lining too.
post #14 of 26
Thread Starter 
With the graph shown I have the gain at only 1/4 (9 oclock) and 0db on the subout, I feel I have a lot of headroom on the amp.
When I play a 20hz sine wave the woofer barly moves. For this reason I feel that more gain would be a good thing but, the problem is when I calabrate the levels with Avia my level (on the amp) must stay very low (below 1/4)and between -10 to -5 on the subout to get it to match up at 85db. This is probly because of these high peaks at 45hz and 75hz, Dont you think? But I do bet your right about the iguana cage. Its got to be hurting me also.
post #15 of 26
The best thing to do is move the sub around. I know it's heavy but it can make all the difference. Or build another one!
post #16 of 26
Thread Starter 
Steve, I took your advice on the stuffing. If you look at the picture the stuffing is covering the four cutouts on the brace. When I did this I thought this was a good thing because I could stuff the top chamber of the box and the brace would keep it from falling out, So I started to think.... well its not restricting air flow to the port but, is it restricting air flow to the top of the box? So I then entered my box design in Winisd and subtracted 2.1cubes from my 8.5cube design and the graph looks pretty close to the REW graphs. So needless to say I removed all the polyfil from the top and even went one futher. I lowered the tune from 17hz to 15.45hz and I can tell an improvement. I only listened at low volume (like -40) to some music for a little while but I could feel more vibrations than I could before at that volume. I also did not hear lower output from lowering the tune which I was worried about and is why I tuned higher when I built it. You can probly measure a (slight) difference in over all output but I dont hear one Im still waiting for my BFD to get here and waiting to take measurements when it does. Thanks Steve for putting the stuffing idea in my head and I'll get some measurements posted soon so we can see the difference this has made.
post #17 of 26
Thread Starter 
A quick post to up date these graphs since I finally recived my BFD. Heres what I have now using four filters and with the mains off. What do you think? The filters are....

Freq Gain BW/60
46hz -14 18
75.75 -11 12
115 -9 19
178 -8 7

The first one is from 15hz to 200hz and the next is 13hz to 200hz.
post #18 of 26
Though that final FR looks great with EQ, it would be best to try and resolve the main problem first, as those are some pretty big cuts.
post #19 of 26
I know this would be a PITA, but can you take it outside and measure it in free space. That will give you the subs free air response, like an anechoic chamber. This is assuming you live in a relativly quiet area, and not next to highway or such. It will give you an idea of the subs response unrelated to what the room is doing, which will at least help you figure out if something is amiss with the sub. The way to do it is to first measure nearfield the cone, then the port, and combine the two in the graph, which will give you the combined response.

I really think this is a room mode problem. You clearly have cencelation causing the dip, and the peak on either side is implying a bad sub in that room. Its a very small room for a very large subwoofer. The answer IMO is going to be bass traps, along with moving the sub around. Where is it currently located in your room?
post #20 of 26
I just modeled the sub and noticed something else. Listen, I know all about the issues with room gain, how measurable it is, etc. however I have long noticed that EBS alignments in small rooms don't work as well as I would normally think they should. You designed your sub with an aggresive -6db extended bass shelf going with that large of an enclosure. I think the downward response you are seeing is following the free air response of the EBS alignment, showing no room gain in that area. The large peak dip is simply a room mode problem, which EQ will fix, though really own in a limited space. I really would look into some room acoustic treatment products.
post #21 of 26
Thread Starter 
I was reading the thread over at the shack called "Strange RS SPL meter calibration results" about how Ilkka was calibrating a newer type analog Radio Shack SPL meter and how is was so different from Sonnies correction file. I look at the graph Ilkka made and it looked like it had about the same drop in the low end that my sub's graph has. And since this response is just about the same as sonnies old analog meter file I decided to install it instead of the new digital file that I was using. Heres a pic of my old measurement with the old analog file installed, Now the measurement has not been adjusted by the calibration file yet but you can see the rolloff is about the same. Im going to take some more measurement tomorrow to see what happends, Im almost getting my hopes up that this might be my problem. What do you guys think of this?
post #22 of 26
Well Good luck, I will say, I hope its thats too. However, like I said, your response also follows a very steep bass shelf modeled on my computer. I would not have gone with such a large box tuned so low, I think that was a mistake. I do believe that was your problem. If room gain was more aggressive one might argue it would take care of that, but I seriously question if it could account for the over 6 decibles per octave you have below 30hz or so.

Also, the upper response of the Soundsplinter woofers, all of them, lumps up like yours, and then the inductance of the coil makes for a low pass filter in and of itself. I have been afraid to use these woofers because, even though they have lots of xmax for not a lot of money, the response doesn't look good. I felt that they were better suited to either a sealed box, car audio, or lots of EQ. I tried lots of different enclosures with the Sound Splinter woofers, and just couldn't ever get a response I liked the look of. I find it really interesting that your measured response looks very similar to the modeled response, with the acception of that dip, which I think is being caused by a room mode, but I'm not sure.

Maybe you already said this, but what is the crossover being used, and what is is it set too? A lot of what will fix the response comes from the crossover when I modeled it, so I wonder if that is being accounted for in your measurements or not.
post #23 of 26
Originally Posted by pjpoes View Post

I would not have gone with such a large box tuned so low, I think that was a mistake. I do believe that was your problem. If room gain was more aggressive one might argue it would take care of that, but I seriously question if it could account for the over 6 decibles per octave you have below 30hz or so

I gotta disagree. Several people have built ~260 liter, 15hz tuned subs using the RLp-15 and have had no such drooping low end. His sub modeled shows ~6db down from peak output stretching from ~17-30hz with the peak being ~80hz. So first off, he should be flat in the 17-30hz range, which he is not. Second, his actual response is down more like 15db in that range. Third, in his actual response, it looks as though the port output is being stifled....you can actually see a small dip where the tuning should be in each sweep, as if the measurement was taken almost close mic. Lastly, factor in a 4th order lowpass at 80hz in his design, and when level matching the sub to the mains, everything would flatten out pretty nicely. The problem isn't the design, it's something else.


however I have long noticed that EBS alignments in small rooms don't work as well as I would normally think they should

A small room will start yiedling room gain sooner than a larger room, so I'm not sure I follow.

Jerm, have you checked for leaks in your sub? How air tight is the sub - how did you seal it? Port output could be reduced if there are leaks.
post #24 of 26
Thread Starter 
Heres the new sub only measurements with the right calibration file and no EQ and the new 15.45hz port tune. Its 2.5 db hot over 75db with the port grill on and an 80hz crossover. These were taken at the LP with the receivers sub trim at -10 and the amp gain under the 1/4 or 9oclock position and the MV at -25 (I dont have it balanced yet with the mains, was just trying to get measurements) Thank god I dont have the severe problem I thought I did. The other pic is both with and without the port grill. Yes the sub is air tight.
post #25 of 26
That seems much more appropriate, hopefully it is the correct calibration
post #26 of 26
Ok I take back what I said, I had a completely incorrect set of numbers for the sound splinter woofer, that is the right sized box, and iwth an 80hz lowpass, it should be decent. Still has a room mode porblem, your lowest room mode is around 49hz, and that seems to be about where the peek is. Bass traps will go a long way towards fixing that, but so could EQ.

This is an incorrect graph in that it assumes a completely sealed room with no panel resonances. However, based on 49hz being the lowest room mode, 11.5 feet being the longest room dimension, I came up with this graph. It's actually closer to what you would see in a car than a room, but I was unable to get an accurate set of numbers. I tried changing the F0, and q values in my model to get a closer approximation and was unable to. I think you could basicly chop almost 12 db's off the low end gain based on typical leaky rooms, and potentially resonant walls. Anyway, if you want to see it, here it is.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: DIY Speakers and Subs
AVS › AVS Forum › Audio › DIY Speakers and Subs › Questions about my subs response.