Step By Step Sub Equalization using BFD and ETF (1 Viewer)

[jeff peterson]

I got my 1124 from www.lentines.com for $129

 

[Bill Catherall]

Thanks Jeff. I found it at Lintine's.
Also, with regard to ETF alternatives...I'm not necessarily interested in ETF equivelants...just something that will graph (and save) the audio input from a soundcard. Really...how difficult can this be? Is it really worth $150 to pay for just that type of functionality? What about Linux...is there anything like this available for Linux?

 

[BruceD]

Bill,

Sorry, the only program I know that does a similiar job is the Spectra+ program which I think is about $295+ vs. the ETF5 software for about $150.

The software is fairly complex as it is doing Fast Fourier Transform math (remember calculus in college?) to create the frequency graph and typically requires a PC with > 166Hz CPU (and even then the graph takes a couple minutes to compute and display on a 166Hz CPU).

BruceD

 

[Steve Morgan]

What a great thread. I got the ETF software for Xmas and a BFD so I'm waiting on my sub to do this but this really helps!

I have a question when you take your sub reading where should your cross-over be set? I have mine set at 70hz but can change it from 20hz to 200hz.Should I just get the cross-over out of the way and set at 200hz or set it where I will use for listening?

When taking readings other than the sub say for front L/R do I just run the cable to the channel on the amp that speaker is connected to or use the pre-amp (B&K Ref30)?

One thing I am going to do is lug the PC into the HT room.This would mean I only need 2 cables about 10ft. long rather than 35ft.Would the sheilded RS cable work in this case?

Great work,

Steve M.

 

[BruceD]

Steve,

For xover settings, I'd thoroughly read through the ETF demo-room info first.

Connecting cables can be at either place you specified (amp or preamp), but if you use the Ref30 analog inputs make sure all speakers are set to Large so no bass management is engaged for the initial speaker/room response tests.

You want to see how each speaker is interacting with the room first (maybe even move speakers around a little), before deciding about crossovers and how to best use the BFD EQ.

Shielded RS cable is just fine.

BruceD

 

[Bill Catherall]

Can't most spreadsheets do Fourier analysis? Here's my idea...write a program that grabs the data from the sound card and sends it to a file. Import the file into a spreadsheet, do the Fourier analysis, and plot the graph. The capture software could be free (or really cheap)...and most people already have some kind of spreadsheet program that came with their computers.

But even if you don't have a spreadsheet I'm sure it isn't such a difficult task to write a program that does a FFT. I'll have to review my text books, but I don't remember it being all that complicated.

 

[BruceD]

Bill,

There are many methods for windowing or gating the raw data input from the impulse response during the FFT analysis (Blackman, Hann, Bartlett, Blackman-Harris, etc.), and this is exactly what is so important when it comes to identifying and/or excluding frequency responses from first reflection points of walls, echo or RT60, and other frequency aberations of the speaker/room interface.

If you can account for all of those sliding variables and do it on the fly in an excel spreadsheet, be my guest. I'd rather have an expert in math and acoustics provide me with a flexible program that lets me see what I want, the way I want to see it.

I don't beleive you can get a reasonable result from an excel spreadsheet, in a reasonable amount of time, with any kind of meaningful accuracy. But you are sure welcome to try. Feel free to share.

I forgot, you will also need a calibrated MLS test signal, a form of white noise. MLS=Maximum Length Sequences

BruceD

 

[Pete Mazz]

My PC is also in another room. I hooked up using 50 ft runs of 75ohm Canare LV-77S cable with RCA jacks and it works fine. I had to adjust the RS meter to 60 dB setting to pump up the signal some, but it seems to work fine. The meter does roll off the upper freqs alot. I'm going to order the calibrated mic setup from ETF. I'll give you guys a heads up when I compare the two.

Pete

 

[Bill Catherall]

Well I found some freeware that does exactly what I'm wanting (but does it's own FFT instead of using a spreadsheet). The only problem is that it doesn't allow you to manually select the frequency range that you'd like to graph. So the resolution can't be custom set. It's called AnalFreq. I've been playing around with it and it's pretty nice for freeware. It's fast and easy to use. But like I said, because you can't get good resolution for the lower frequencies it might make it difficult to make EQ adjustments to a subwoofer.
But I wonder how difficult it would be to add that functionality. I mean it's just a matter of changing some of the parameters in the program. As it stands the lowest range you can graph is from 0 Hz up to 750 Hz (with a resolution of 1 Hz and .1 dB). It's not too bad really. It'd still be somewhat usable. Do you need a higher resolution than 1 Hz and .1 dB?
I'm wondering (assuming this program had the ability of higher resolution) what does ETF offer that AnalFreq doesn't that is important to getting accurate readings and making accurate EQ adjustments? (Now I'm not trying to bash ETF here. I'm simply trying to understand what it is that is most important when making measurements that this freeware package doesn't offer.) What is it that ETF has that I would need that would cost me $150 more?

 

[Rick Radford]

Ken,
Thanks for your encouragement to drag that pc into the HT room and go for it. I'm sure you're right how easy it is once you get everything setup. I'm not having much luck running the self calibration tests though.
I stopped by RS tonight and returned my speaker cable and mono 1/8 adapter in exchange for $20 worth of F to phono adapters, Stereo audio adapters and monaural audio adapters.
Bruce, I scrounged around the house and found 4 x 15 ft lengths of cable left from a satellite install. Only thing is, it doesn't say RG6 or RG59 on it. Can anyone interpret for me:
Belden-T YV45132 High Performance 2.25 GHz DBS cable Duobond (R) 11 75 ohm series 6.... yada yada yada. Hope that's enuf to tell what it is.
I also have some cable that says:
TFC-T10 File no. E86650 CATV (UL) 6 series.
Any idea what this is?
I also have some flat wire RCA cable with red/white/yellow for audio/video connections. Any way to tell whether or not it's shielded?
OK. Enough cable talk.
I'd love to buy this ETF software. I was sold with the customer service response time (and all the good response here, of course).
But the rub is that right now, I only have ready cash for either the BFD OR the software, not both.
I did not find the same Part #s as the ETF help file suggested. 42-2481D does not exist at the RS.com site.
I got p/n 274-883 1/8 stereo adapter with 2 rca female plugs. I attached an A/V cable with red/yellow/white RCA connectors.(connecting the red and white cables). I plugged one into the line out and one into the line in of the sound card.
Running the level check, I get -2 dB in green regardless of whether the adapters were plugged in or out. This makes no sense to me. The help file says to check it this way and that with the input disabled, it should read broken (-30dB or so).
My mixer was set:
Midi balance muted
CD Audio balance muted
Line in muted.
Microphone balance muted.
It's late and I'm tired. Enough for now. I'm probably missing something obvious.

 

[Pete Mazz]

Rick, did you set your recording controls to just Line In?

Pete

 

[BruceD]

Bill,

Yes, I think anything with at least 1/6 octave accuracy should be sufficient as seems to be the case with your 1Hz, 0.1dB example.

Added benefits of ETF for me:

1) 3D frequency waterfall graphs with time in the z dimension that show you the modal frequency peaks (for bass) and the higher frequency reflections?

2) Ability to use the software in pseudo-real time, i.e. keep the system constantly recording while you change variables and watch the changed results on the screen as they happen (speaker or sub placement).

3) Stack multiple results in the same graph, like main speaker and sub individually measured, but viewed together.

4) Settings menu for changing the gating window or time slice and method of FFT analysis for viewing the frequency graph. This also allows me to measure only direct sound from a speaker (like an anechoic chanmber) or the speaker+room sound.

5) Measure speaker distance exactly for 5.1 speaker setup

6) Measures phase changes over the frequency range

Just a few I find useful, and most valuable for me:

accuracy and results quickly, definately worth my $.

 

[Bruce N]

5) Measure speaker distance exactly for 5.1 speaker setup

Bruce, could you go into a little more detail as to how you perform this measurement? How do you direct the test signal to the center and surround channels? Also, how do you calibrate ETF so it displays the actual dB level instead of the graph just topping out at 100dB?
So far, I've only used ETF to EQ my SV subs with a Rane PE-17 EQ. What a difference, the subs were great before, now their just outrageous! Then the extra PC I was using in the theater room had to go to a new owner. I haven't be able to use ETF for several weeks. All I need is a cheap ATX case and I can assemble a dedicated ETF PC with some leftover parts I have laying around. Then I can start playing with ETF again.
Thanks to everyone who has posted their experiences with ETF. It's a huge help for us slow learners!
Bruce

 

[Rick Radford]

Pete,
I noticed you sucessfully ran your test without relocating your computer. I think I'm resigned to dragging my PC in the HT room. It'll give me an excuse to clean up around it anyhow.

did you set your recording controls to just Line In?

No, not at first. I didn't even know there were 2 mixers (a playback and a recording mixer) until I ran into these difficulties with ETF. Somewhere along the line between emails to ETF, I discovered the recording mixer. :b
But I also found another problem. I was using my SB Live! value card's black jack (line-out #2) and the light blue line-in. Once I got the recording mixer set properly, I found the black jack did not work (got -57 dB/-58 dB rec levels). When I changed the line-out to the lime jack (line-out #1) I got -2 dB on both left and right channels on the level check. And the broken test showed -57 dB/-58 dB for Ch1 and 2.
I think I'm in business now.
Thanks to all for your patience and advice. I have got to find a way to work this software into the budget. This is too good to pass up.
Notice: a new software testing company is being setup called Idiotproof Software Testing Services. Email me for details...

 

[BruceD]

BruceN,

There is a selectable icon at the bottom of the ETF program screen after you have captured an impulse response from one of the speakers.

With this feature, you can adjust the speaker distance (move the speaker) and adjust the channel delay of your DD/DTS processor until both are corretly set.

It displays a popup window that shows the distance in meters and feet from that speaker to the microphone (at listening position) and also shows the delay time in ms (milliseconds). Accuracy is to 0.05 ms.

To effectively use this feature, you do need to connect the programs output cable to the amp driving that specfic speaker and measure every speaker individually. You could also use a preamp input, but be careful because bass management can really mess this up.

 

[Greg Lee]

Rick I don't remember if you said you had used the demo version of ETT or not. If not, I strongly recommend it if money is tight for the purchase. I did buy it the other night, primarily for convenience - overlays,etc. But I think you could do fine with the demo. I'm sure it would beat the heck out of the manual charting method.

One obvious tip you probably know about is that to print a graph from the demo version, just press 'Printscreen' on your keyboard, then go into WORD and Edit->Paste. First set your 'page setup' to landscape, and you'll get a great full size image of the graph that you can save or print.

 

[Bill Catherall]

Those are some good points BruceD...thanks.

Another question...has computer fan noise been a problem for anybody when making these measurements with the computer in the room? If so, how did you over come it? Does ETF have built-in noise reduction/filtering?

 

[BruceD]

Bill,

Yes, that can be a problem. See notes in program help or ETF site.

Specifically, that's why I use my notebook with battery power and my mic-preamp with battery power---No extraneous sound.

I suppose putting the PC in a closet, around a door opening in the wall, etc. would help.

 

[Pete Mazz]

It displays a popup window that shows the distance in meters and feet from that speaker to the microphone (at listening position) and also shows the delay time in ms (milliseconds). Accuracy is to 0.05 ms.

This may be a problem when running long cables into another room. I was getting exaggerated distances, and I think it's because of the length of cable. I'm going to try to add an equal length of cable in between the two lines that loop between the soundcards input and output jacks to see if that straightens this out.

Pete

 

[Rick Radford]

Greg, yes, I do have the demo version.. which limits the measurements to 2 at a time.

Thanks for the heads up on the Printscreen command. I was using it with Paint.. I'll hafta try it with Word97.

I hope to purchase this software soon.