BFD EQ with ETF - sub-mains integration (1 Viewer)

[Pete Mazz]

I will have to use a test CD for this one, as ETF signals don't seem to activate the crossover at the pre-amp.

I don't really understand the problem with this. The output from ETF and the soundcard is a simple analog signal. Just make sure your processor is not set up in direct or bypass mode.

Pete

 

[brucek]

Ranga,
Sorry I've been busy lately and haven't had a chance to do much reading here at HTF.
I have to say that I'm a bit confused about a few things in this thread, but I think BruceD's and Wayne's advice has been right on the money though, given the unusual setup.
Let me see if I have this correct.
You have a DVD player with built in Dolby processing and bass management with a 100Hz cross that outputs six analog signals (L/R/C/LR/RR/SUB). You "bypass" these six managed signals through your receiver which provides volume control only. The bypassed sub out of your receiver (which the DVD player managed at 100Hz) then passes through a BFD and then to your powered sub?
You have a CD player feeding an external DAC that feeds analog to your receiver that you run "Source Direct" and has an 80Hz bass management crossover that sometimes works to your sub? That's a bit confusing.........
A Source Direct usually doesn't have bass management associated with it - it just passes the analog signal direct, full range to the mains. Then depending on the method used by that manufacturer, there is either, no sub out, or full range sub out....
If it does have bass management associated with Source Direct, then you've likely done a double D to A conversion....It may be better to use the processors DAC in that case.. You'll have to clear these points up......
Your graph shows a black line of the "sub only" with no crossover (which looks to be easily equalizable).
Your other two lines pass through your processor obviously without a crossover enabled. Those are certainly full range signals I would presume because if I look at the red line, it's at the same level at 20Hz as it is at 80Hz. That's an awfully good main speaker? or the room is perhaps helping it out.....
Anyway, as you already know, an equalizer between a preamp and subwoofer can only modify the subwoofers response. It has no effect on your main speakers. The response of your mains can only be altered by positioning - as others, including yourself have said already. This was the reason BruceD advised to do each main speaker alone, so you get a better feel for each speakers interaction in the room. You may or may not be able to do something about undesirable room effects in this fashion. Once you do a response of both the mains individually and have moved them to their best positions or you've got it where you're willing to live with the response, then you can do a response of the mains together and see their interactions. You then may be able to fine tune some positioning at that point. That's all you can do for your mains..
Either way, I know you don't want to hear this, but you're having trouble with ETF - lose it.
Make a CDr of 1/6th octave tones and use that with your Radio Shack meter. Get the tones from Sonnie's site Link Removed or use a software signal generator at http://www.nch.com.au/action/index.html . Scroll down to Freeware and click on TONE GENERATOR......save the tones in five second files and burn them on a CDr. Limit your CD to sixth octaves that match the values in your graph.. This way you can put the CD into your player and for quick checking just let it play each tone and record the results - takes just minutes to do a response check. The CD will play in both your CD and DVD player and will feed through your system with all the appropriate crossovers and bass management engaged. You have the capability then of shutting off the sub for a mains only, or shutting off the mains and do a sub only check.
I know others recommend otherwise, but when I do a response of my "sub only" or a "single main only" or "mains only" or all speakers in concert, I want the processor bass management enabled. This allows me for example, to see the sub with its LPF dropoff, and shows exactly how the sub will be "in situation". This allows me to look at a peak that is perhaps at 120Hz and see how insignificant it is because when I have a crossover at 60Hz, its quite far down at 120Hz and I'll ignore it. If I didn't do the response with the crossover enabled and I saw the 120Hz peak, I might be overly concerned for nothing. This is the same reasoning I'm using when I do a response of a single main or both mains together with the bass management enabled.
Use your CD player and tones CD and pass this signal through your processor with the 80Hz crossover engaged and do your mains response until your satisfied with their positioning. If you have a 80Hz 2nd order HPF cross on your mains, then it will be down 12dB each octave.
Then shut off your mains and do a response with the sub only with the 80Hz cross in effect through your preamp. This will likely be a 4th order LPF and so will drop off very quickly and will show you where you can essentially ignore some peaks and valleys because of the cross. Now use your EQ and smooth the few peaks and that's it, unless you want to add some sort of house curve. This can be added after.
You can do a response with the mains combined with the sub now just to see how well they integrate and whether you want to wholesale adjust your sub level or not depending how it sounds......
Now that you have the mains in their optimum position and the sub equalized, you can try a response with the DVD player (using your tones CD) and its 100Hz cross and see what it looks like.. Perhaps you may have to make some "compromising" adjustments to the EQ filters at that point - but maybe not.....
brucek

 

[Ranga]

brucek

I don't really understand the problem with this. The output from ETF and the soundcard is a simple analog signal. Just make sure your processor is not set up in direct or bypass mode

I made sure that my pre-amp is not in bypass mode. I connected the output from the PC (ETF) to a Y cable that I then fed into the CD input on my preamp. The sub-out then goes to the BFD. What is baffling is the fact that the crossover is active when I play music through the CD input.

 

[brucek]

Ranga,

BTW, have you tried burning the ETF signals onto a CD (mentioned in the ETF website) and then measuring the response?

Nope. I've used ETF, and it works fairly well, after a bit of a learning curve, but the dragging of the PC into my HT area is not fun. I can do an entire response with my tones CD and SPL meter faster than you can boot Windows.......

brucek

 

[BruceD]

Ranga,

Sorry, I have been away at the Gigabit Ethernet Conference for the last 2 days and haven't been able to really get online.

Regarding ETF, again, I will say, use ETF with the low frequency (20Hz-200Hz) selection to only drive the sub, not in combination with your mains.

As an experiment you can also look at the difference between these two graphs;

Make sure the main speakers are disconnected for this

1) drive the sub's input jack directly from ETF with a single cable and 2) using just a single cable from ETF to the left CD input on your preamp with the 80Hz xover engaged. You shouldn't need a Y cable on the input jacks to the preamp.

You can also do each main speaker by itself (individually not together and not with the sub) with the 80Hz xover engaged on the preamp (use a single cable from ETF to the CD input for the right speaker, then do it for the CD input for the left speaker). No Y cables for this test either.

 

[Ranga]

BruceD
No worries.. I was checking for any of your posts.. so I knew you weren't around along with brucek.
BTW, I figured out that the mains had to be off for the sub to get the signal from the ETF output with the 80 Hz Xover engaged.
I'll post the EQed graph tomorrow when I get to use my laptop at work. I was impressed

 

[Ranga]

Ok
First, thank you all, esp. BruceD, brucek and Wayne for helping me out.
Here is the EQed graph -
Link Removed
Here are my BFD settings -
Link Removed
I must mention that setting the BANDWIDTH was trial and error more than the mathematical approach. So it would be nice if someone could take a stab at why these filters worked.
Two observations -
1. I had to set two filters quite close to 45Hz. Initially, I picked 45.5Hz. After setting the 80Hz and 100Hz filters, I noticed another hump close to 45Hz. The only way I could flatten it out was by setting another filter at 45Hz.
Any explanations for this?
2. The two nulls at 33Hz and 66Hz did not benefit from any increases in boost beyond what is currently set. Fortunately, I did not notice any reduction in overall output with these two boosts. I guess that's a good thing.

 

[brucek]

Ranga,
Your filters seem reasonable and the result is that your response looks very smooth.
My only comments would be that it appears there was a change of volume between the no-filters and filters measurement.
This makes it extremely difficult when looking at your graph to establish how effective each filter was in altering the un-equalized curve.
For example, look at 20Hz. there is about a 5 dB increase in level, yet none of your filters even remotely affects this frequency. You either moved your meter reading position or used an old unequalized graph or changed the volume while testing, between the unequalized and equalized readings.
Another example would be at 45Hz, where you have created about a -11dB filter and the value on the graph is the same both for equalized and unequalized.....See what I'm saying. It makes it difficult to use the unequalized as a "known reference" to compare the filters effect to...
Another observation might be that you have measured quite "hot" at around 100dB. You might want to do a remeasure of the unequalized and then equalized without touching the filters or volume between testing at around an 85dB level..
brucek

 

[Ranga]

brucek

Another observation might be that you have measured quite "hot" at around 100dB. You might want to do a remeasure of the unequalized and then equalized without touching the filters or volume between testing at around an 85dB level..

Hmmm.. I balanced the channels out at -1db in ETF and set the SPL meter at the 70db mark (although this should be irrelevant). The only way to reduce the volume would be to balance the channels out at say -10-20db in ETF so that the overall response is around 85db. Is there anything that I am missing here?

 

[brucek]

Ranga,

I balanced the channels out at -1db in ETF and set the SPL meter at the 70db mark (although this should be irrelevant). The only way to reduce the volume would be to balance the channels out at say -10-20db in ETF so that the overall response is around 85db. Is there anything that I am missing here?

Yes, I think you may have misunderstood. Stop using ETF!
Use your tones CD in your CD player, and play 1/6 octave tones from 20Hz to 160Hz through your system with your mains speakers shut off and your sub only turned on with a 80Hz crossover engaged and the BFD filters disabled. Record the frequency response from your listening position at a volume level around 85dB. Once the initial volume level on your sub and processor is established, begin the response test and "do not" touch the volume of anything until the test is complete. If you touch any volume control, start the test over.
Once you've completed the unequalized test, enable the BFD filters and do another response test ensuring that you have not touched any volume controls throughout all this testing. If you do touch the volumes of anything except a filter, you must begin all over again........ Post the two graphs and then it will be possible to comment on the results...
brucek

 

[Ranga]

brucek
Here is my updated frequency response and result of the EQ at a 80Hz Xover. I measured the response at the 100Hz Xover (using my DVD player's Xover) with the same filters as well. The volume seems higher as I used the 5.1 inputs on the pre-amp instead of the CD inputs for the 100Hz Xover. The volume on the pre-amp and sub remained the same for both sets of readings.
Link Removed
Here are the filters -
Link Removed

 

[brucek]

Ranga,

I think it looks real good, you didn't comment how it sounds.

You've successfully eliminated your major peaks at 38Hz-60Hz and at 80Hz and 100Hz. You have a small valley at 36Hz that is quite acceptable.

Others might tell you to also drop down the rise at 22Hz, but I suspect it gives it that extra window shaking power, so I'd probably leave it alone.

The filters seemed to work well for both the 80Hz and 100Hz crossover. The EQ lines look good..... See how easy it is now to observe on the graph the effect of your filters when there has been no volume change during the testing..

I find it's best to live with a change of your filter values for a while and get used to it before changing any more.

brucek

 

[Ranga]

brucek

Others might tell you to also drop down the rise at 22Hz, but I suspect it gives it that extra window shaking power, so I'd probably leave it alone.

I am not sure about the rise at 22Hz as it seems to be largely due to the SPL correction factor. I am certainly not able to hear/feel it at those frequencies when playing the test tones. I am yet to play some HT material to test it out, so I will find out if it's too overpowering.
Thanks again.

 

[brucek]

Ranga,

I suppose, at this point, it is ok to increase the sub's volume levels to calibrate for HT?

Yep...now you can adjust the subwoofer amp to trim up your overall level.

Usually after a large peak has been equalized down, you have to increase the wholesale level of the sub amp up slightly to compensate.

In your case, since your overall level was mostly affected by your peak at 38Hz-60Hz (which has been tamed by EQ), you'll now be able to wholesale raise your level a little to compensate, which will bring up those frequencies you had lost before...

brucek