New Sub...2nd try at REW

**rafaelsmith** · **03-26-08, 06:07 PM**

I finally go my new sub....a HSU STF2. Its really my first "real" sub as the one I had before was an el-cheap 8" Sony.

I had posted a few weeks ago with some plots I did witht he old sub.

After hooking my new HSU up....I played around with REW and trying it out in 2 possible locations I can put in in my room.

Does it look like I am doing things correctly?

I am still learning what all this is telling me. I assume the 2nd plot(loc #2) is the better?

Location #1 in back Corner

Location #2 side near front right speaker.

**brucek** · **03-26-08, 06:18 PM**

No, I would say location #2 is poor with respect to extension. You have lots of low end extension in #1 . You also have a couple peaks that can be equalized away, and then the wholesale gain increased.

Is the graph through your receiver with the mains off and the crossover engaged?

brucek

**rafaelsmith** · **03-26-08, 06:31 PM**

Quote:

brucek wrote:

No, I would say location #2 is poor with respect to extension. You have lots of low end extension in #1 . You also have a couple peaks that can be equalized away, and then the wholesale gain increased.

Is the graph through your receiver with the mains off and the crossover engaged?

brucek

Those are through my receiver (H/K 247), with the mains(Klipsch KSF 10.5) on and the XO engaged at 80.

I need to redo the measurements with the mains off?.

But I think I understand what you are saying with regards to the low-end extension.. I guess my brain was thinking that less dips/peaks is best.

**brucek** · **03-26-08, 07:03 PM**

Quote:

I need to redo the measurements with the mains off?.

If you plan to equalize the sub, it's advisable to disconnect the mains and do a sub only measure first to see how it contributes by itself. Once the best position is found and then the sub equalized, you add the mains and see how the sub and mains integrate. The phase of the sub usually requires adjustment at that time to make the crossover transition smooth.

brucek.

**rafaelsmith** · **03-26-08, 07:13 PM**

Quote:

brucek wrote:

If you plan to equalize the sub, it's advisable to disconnect the mains and do a sub only measure first to see how it contributes by itself. Once the best position is found and then the sub equalized, you add the mains and see how the sub and mains integrate. The phase of the sub usually requires adjustment at that time to make the crossover transition smooth.

brucek.

Thanks for the info.

The boss is about to get home so I need to put away all the wires, PC in the middle of the room, tripods and such =P but will get back to taking proper measurements tomorrow.

**JohnM** · **03-27-08, 04:06 AM**

The extra low end in location 1 looks to be excitation of a couple of primary modal resonances. I'd be tempted by location 2, favouring a more uniform response over some output at the very lowest frequencies. Location 1 would probably be improved by some crossover phase adjustment though.

**mojogoes** · **03-27-08, 07:02 AM**

Not to highjack this thread....i would just like to ask if it were better to have both sub and main speakers connected to take some readings...or like i have done just the sub connected through the bfd set at bypass via s/b card and preamp and ecm8000....so that i'm only taking readings of the sub.

Can i then include the main speakers with my spl meter only for integration after the sub id eq'd.

**rafaelsmith** · **03-27-08, 10:22 AM**

Quote:

JohnM wrote:

The extra low end in location 1 looks to be excitation of a couple of primary modal resonances. I'd be tempted by location 2, favouring a more uniform response over some output at the very lowest frequencies. Location 1 would probably be improved by some crossover phase adjustment though.

Hopefully I will be able to do more measurements when I get home from work today.

Going on my ear alone...the sub in loc #2 sounds/feels less "localized" than in loc #2

I want to do 4 measurements....both locations with 180phase and 0phase.

Something that has me concerned is that I can run a measurement multiple times in a row without moving anything or changing any settings and get different plots...some of which differ quite abit. So I am not sure if I have something setup wrong or if the results I am getting can be trusted.

Here is a rough foreplan of my room/apartment. I am just not sure if its even possible to get a good plot in this room....

**brucek** · **03-27-08, 10:31 AM**

Quote:

i would just like to ask if it were better to have both sub and main speakers connected to take some readings

Yes, after you take a sub only reading with the mains disconnected. You want to have the sub tested through the receiver so that the standard crossover is in place that you use.

brucek

**rafaelsmith** · **03-27-08, 07:00 PM**

OK decided to start over from scratch...replaced some of the adapters cables I was using cause one of them was flacky. Reset REW and redid the calibrations for soundcard and mic/SPL.

Getting more consistent results...did 4 measurements....both locations @ 180 and 0 phase.

So it seems now that I have REW configred correctly and am getting consistent results...the results are not as nice =P

Location #2: Blue = Phase @ 0, Orange = Phase @ 180

Location #1: Purple = Phase @ 0, Aqua = Phase @ 180

Well not sure what this is really telling me....My room sucks? =P

Phase @ 180 seems the better at both locations?

Although my SPL and REW are calibrated...@75db...everythings seems high?

Really confused.

**mojogoes** · **03-27-08, 08:22 PM**

Cheers Brucek....apologies once again!!!

**rafaelsmith** · **03-28-08, 12:22 PM**

I am still learning what all this means and what REW is actually telling me.

Curious What does 1/3 octave smoothing applied? mean.. I mean what is it telling me? Ive seen some people post plots with it applied.

My plots are a mess and nowhere near a nice as others ive seen here.

I do not know how to interpret the data...is it even telling me what I need to know to improve my repsonse...if at all possible. Does my sub suck, my room suck, my AVR suck...all the above...Starting to think this probably was not a good idea

**brucek** · **03-28-08, 01:33 PM**

Quote:

Blue = Phase @ 0, Orange = Phase @ 180

If these are sub only frequency response plots, you do realize that phase has no effect until the mains are added? Phase has no effect on a single speaker. It affects the way two speakers interact at their common frequency areas (i.e. crossover).

The smoothing feature is for plots that are above the ~200hz range since comb filtering tends to make the plots unreadable. You require the smoothing to get a feel for the underlying trend of the signal. You don't use smoothing on subwoofer plots.

Quote:

My plots are a mess and nowhere near a nice as others ive seen here.

I think they look great. You could use a little equalization on the last plot to lower the reasonant peak, but other than that it's pretty good.

Quote:

Does my sub suck, my room suck, my AVR suck

They all look fine to me. It's usually never an AVR problem (other than not set up properly with respect to crossovers etc). To determine what the sub can do without the room you can drag the sub into an open area (middle of the room away from walls), and do a near-field measure. This will reveal what the sub is like without the aid or hinderance of the room. Then place the sub into its corner and take a measure at the listening position and see what that pesky room is doing.

brucek

**rafaelsmith** · **03-28-08, 01:38 PM**

Quote:

brucek wrote:

If these are sub only frequency response plots, you do realize that phase has no effect until the mains are added? Phase has no effect on a single speaker. It affects the way two speakers interact at their common frequency areas (i.e. crossover).

The smoothing feature is for plots that are above the ~200hz range since comb filtering tends to make the plots unreadable. You require the smoothing to get a feel for the underlying trend of the signal. You don't use smoothing on subwoofer plots.

I think they look great. You could use a little equalization on the last plot to lower the reasonant peak, but other than that it's pretty good.

They all look fine to me. It's usually never an AVR problem (other than not set up properly with respect to crossovers etc). To determine what the sub can do without the room you can drag the sub into an open area (middle of the room away from walls), and do a near-field measure. This will reveal what the sub is like without the aid or hinderance of the room. Then place the sub into its corner and take a measure at the listening position and see what that pesky room is doing.

brucek

Thanks....Was curious about the phase...I seemed to get alightly different plots...especially at the extrem low end with just the Sub and different phase.

Base on those last two plots I shows which location would seem to be the best for me? My gut tells me Location #1.

Do you think an investment in BFD to EQ would be something I should consider? I guess more toys to play with is always a good thing =P

I will try putting the sub in the middle of the room and doing the near-field test as you suggest...I assume that means placing the sub in middle of the room adn the pic on the floor nex to it? Probably have to recal things since the SPL reading is going to be much higher?

**brucek** · **03-28-08, 01:59 PM**

The location #1 looks pretty good.

Yeah, a BFD wouldn't hurt. Looks like you have a resonant peak around 45Hz.

As long as the sub is away from walls as much as possible and the microphone is within a few feet, you should get a better feel for the actual sub output with minimal room influence (of course outside is best, but you don't want your wife having you committed, so maybe avoid that).

brucek

**rafaelsmith** · **03-28-08, 06:15 PM**

Ok, finally picked what I think the best location for the sub. Not sure if I did the sub in the open with SPL near correct.

Sub In middle of room w/ SPL 1 foot away.

Sub Only

Main + Sub Phase @ 180

Am I doing ok?

**brucek** · **03-28-08, 06:24 PM**

Yeah, and the response isn't too bad when you add the mains.

I don't think the near-field worked out too good. Your room is overwhelming that situation.

Your crossover area looks a bit funky. If you only have a phase switch with two choices, then use the sub distance in the receiver as a proxy for a variable phase control. Play and get it a bit smoother if you can.

brucek

**rafaelsmith** · **03-28-08, 07:55 PM**

Quote:

brucek wrote:

Yeah, and the response isn't too bad when you add the mains.

I don't think the near-field worked out too good. Your room is overwhelming that situation.

Your crossover area looks a bit funky. If you only have a phase switch with two choices, then use the sub distance in the receiver as a proxy for a variable phase control. Play and get it a bit smoother if you can.

brucek

My room is rather small...only about 5-6 feet from center to love seat, sofa, table and entertainment center.

I was wondering about the area around the XO...seemed odd...my sub only has a 2 position switch for phase 0 & 180. 180 looked the best to me.

I will play around with receiver distance settings to see what it does.. My receiver has diff XO settings for main, center and surrounds...then the SUB can bet set to same as either main, center or surround, currently have all set to 80...but my mains can handle lower...might need to rethink my XO settings and see if that changes anythign with the plot

Not entirely sure a BFD/EQ would do me much good...but its rather cheap so might be fun to play with.

Thanks again for the help...Ive learned alot these past few days

**mojogoes** · **03-29-08, 12:50 PM**

Quote.
Not entirely sure a BFD/EQ would do me much good...but its rather cheap so might be fun to play with.

From what i've read and been told "a bfd will give you more than you think" and will make a very good sub better given its not in the perfect room in the perfect position that is.

**rafaelsmith** · **03-30-08, 04:04 PM**

Quote:

brucek wrote:

Yeah, and the response isn't too bad when you add the mains.

I don't think the near-field worked out too good. Your room is overwhelming that situation.

Your crossover area looks a bit funky. If you only have a phase switch with two choices, then use the sub distance in the receiver as a proxy for a variable phase control. Play and get it a bit smoother if you can.

brucek

Best ive been able to get by adjusting phase switch and sub distance.

**rafaelsmith** · **03-31-08, 08:44 AM**

I should be able to borrow a BFD to try out.

Assuming I can get the BFD hooked up and all...what sorta results should I expect..if any?

I went thru the filters section of the REW help but when I follow those steps with my measurment...It finds alot of peaks (about 11)...and from looking at my plot I can see why.

I was curious about my plot becuse it seems to have alot of peaks?

My understanding is that...at least from reading the forums and the BFD/REW guides is that its not meant for such a "bouncy" response? I do not have 1 or 2 peaks like the guides.

Also, I have a question about target level....when I look at my plot it seems to me that everything is too "HIGH"...i.e the whole graph is above the target level...I would have expected when I used "Find Target Level" that it would have move it up.

I am not sure I did the calibration for REW correctly.

I have to have my volume, input levels and such where they are at to achieve the results outlinded in the calibration section of the REW help guide(the part about Check Levels...to get input level to be ~-18db..i end up having to raise the AVR..which ends up being much higher than 75db on the SPL). Otherwise I get popup message telling me to raise the AVR volume when I run measurements with REW. Could my Meter/SPL be bad?

Still trying to make sense of what all this data is telling me.

**brucek** · **03-31-08, 09:13 AM**

Quote:

Also, I have a question about target level....when I look at my plot it seems to me that everything is too "HIGH"...

Correct. You simply need to manually lower the response line (or manually increase the target line). Also, since you are obviously using your mains + sub for that plot, you need to switch the target to full range or the REW program will suggest a ton of un-needed filters. Really, you should be using your sub only when you equalize and have the mains shut off.

brucek

**rafaelsmith** · **03-31-08, 09:19 AM**

Quote:

brucek wrote:

Correct. You simply need to manually lower the response line (or manually increase the target line). Also, since you are obviously using your mains + sub for that plot, you need to switch the target to full range or the REW program will suggest a ton of un-needed filters. Really, you should be using your sub only when you equalize and have the mains shut off.

brucek

Thanks,

And for yet another question =P

When I first hooked everything up..I had a HUM comming from both the mains and sub when I used the input on the AVR that the PC/Soundcard were hooked up...unplugging the PC/Soundcard input removed the HUM...so I knew it was being caused by the PC. I had one of these lying around so used it..which removed the HUM. (I have it hooked up to the line-out of the soundcard).

http://www.radioshack.com/product/in...entPage=search

Was that a bad thing to do? would it cause problems with REW and measurements?

**rafaelsmith** · **03-31-08, 03:09 PM**

Quote:

brucek wrote:

Correct. You simply need to manually lower the response line (or manually increase the target line). Also, since you are obviously using your mains + sub for that plot, you need to switch the target to full range or the REW program will suggest a ton of un-needed filters. Really, you should be using your sub only when you equalize and have the mains shut off.

brucek

OK my first shot at using REW EQing functions. I do not have access to BFD yet but I went ahead and tried playing around with REW to see if I have an understanding.

First plot is plot of just the SUB.

Then using REW first i did Set the Target Level then Find Peaks, Assign Filters, OPtimise PK Gain and Q. Target Level came out to be 74.2 which seems low to me based on the plot.

Then I redid the EQ by forcing a target level of 80db...Just picked that cause it looked more like the overall "center" of my sub plot.

Am I even on the right track? I have looked thru some of the REW and BFD help files and threads and as I understand things...large gain adjustments are considered bad? REW came up with some really high ones...17 and 10?

Before I bug my friend to let me borrow his BFD...I would like to know what I am doing...seems to me if I use the filters recommended by REW i will end up worse than I started.

Thanks again for any advice.

Advertisement