using the DEQ 2496 with rew, warning, mainly graphics.

[terry j]

hi everybody

I am quite new to this game, but I must say with the excellent help available on this forum I'm very happy and just a little suprised with my progress. It really is an excellent program.

I have posted before regarding poor soundcard response using the onboard soundcard. As poor as they were, it nonetheless was quite adequate for low frequency use.

Since then, I have been able to borrow a friends external soundcard, which has enabled a larger frequency range to be looked at. The graph drops off heavily when measuring full range responses, this I attribute to the radio shack meter not really being of much use above 10k ( is that true?).

Anyway, after having 'mastered' the program sufficiently to feel the purchase of a BFD would be justified, I also managed to 'score' a subwoofer with which to play with. Things are just a little bit tight right now, so I often gazed longingly at the DEQ 2496 sitting next to me, and knowing that it has parametric functions wished I could use it instead. But of course, it was unsuitable as the Q etc was totally incompatible with the REW program.

Browsing (again) thru the various posts on the forum I came across one where someone wanted to use a Rane (?) with REW, which was supposedly no prob. One of the resident wise ones ( Brucek in this case) mentioned that it might need to have the choice of unit in the equalizer tab changed. Intrigued, I had a look and blow me down, selecting FBQ 2496 made the units in the parametric table compatible with the DEQ 2496!! Admittedly using that equalizer choice puts a BW OCT unit as .25 for example, but it doesn't take a rocket engineer to work out that that is equivalent to 1/4 in the DEQ.

So off I went.

raw sub response cutoff 50 hz.jpg

this shows my raw sub response. After using the recommend filter options ( and they are very accurate predictions by the way ) I ended up with this

pretty close just using parametric on sub..jpg

this only used three or four para's, so for the sake of illustration I artificially lowered the target curve in order to 'need' more para's.

for the sake of illustration, artificially lowered target.jpg

Then off I went using the recommend settings and ended up using nine para's of cut boost, and manually added one with positive boost to end up with

10 filters used, one manual boost 40.83,5.5,0.167.jpg

I did this deliberately because the DEQ has only go ten bands of parametric, and I wanted to explore it's limits, and it seems as if we've maxed out the unit.

However, the DEQ has also got a 1/3 octave GEQ so I incorporated that into the equation. Back to the raw response

raw sub response cutoff 50 hz.jpg

i then used the GEQ to pre filter the response before using any of my ten PEQ's.

graphic only no para yet. 20,-5.5;25,-10.5;50,-11;63,-12..jpg

In this graph I've used the GEQ to dial in -5.5@20,-10@25,-11@50 and-12@63.

Then using the recommended paras I got this

sorry, I must have loaded the picture incorrectly because it's not in the graph format like the others, so I can't show you. Bugger. Rest assured it was as close to the target response as before, but the point I'm making is that by using the GEQ prior to any paras, I only needed one parametric filter this time instead of the previous nine.

I've had a few beers now so it's a safe bet that any more illustrations for tonight are 'not on'! so that's all for now.

I thought this might be of interest because the DEQ is often used higher in the chain ( ie perceived to be of higher sound quality ) than the BFD's and we still have the auto eq functions to play with for the response well above 200 hz. (eg, I've still got nine more paras to play with to help sort out speaker response anomolies in the entire bandwidth).

Tomorrow I will get on to incorporating the auto eq with the sub eq, and see how it goes.

here is the raw full range response

raw full range response.jpg

and here is the response after fixing the sub

full range response after g and p equalizers added..jpg

What is of interest here is that perhaps (?) the use of this unit can go a long way towards fulfilling the wish list Sonnie has got going for a tailor made bfd?

For example it does it's own RTA with the Behringer mic, but unfortunately I can't get it to replace the radio shack mic yet. Hoping that it is my lack of knowledge and expertise, but unfortunately I'm beginning to think that I can't actually get the mic measurement to the analog out and hence into the computer. Bummer. If we could then that would be a major pluspoint.

Hope this is of interest, at the very least I can muck about with it for a while more and maybe not need a BFD. By the way, the number of times I checked on the Behringer website for a BFD 2496 !! ha ha.

lots of love

terry

[brucek]

selecting FBQ 2496 made the units in the parametric table compatible with the DEQ 2496!

Yep.

so for the sake of illustration I artificially lowered the target curve in order to 'need' more para's

Yep, you've certainly figured it out.....

What is of interest here is that perhaps (?) the use of this unit can go a long way towards fulfilling the wish list Sonnie has got going for a tailor made bfd?

The problem is that the DEQ is quite expensive compared to the 1124. The use of mutiple programs is also quite important to many people - that's why they don't like the FBQ.

I've still got nine more paras to play with to help sort out speaker response anomolies in the entire bandwidth

The anomolies at the higher frequencies are usually handled better with room treatment. You do have to be careful trusting the RS meter at the higher frequencies. Some say 5K is the upper limit.

You'll have to try and shelf up the bottom end a bit to add a little house curve. Usually a flat response sounds kinda 'flat'...

Interesting post.....

brucek

[terry j]

thanks bruce

I can't really up the bottom because the gain on my subwoofer is already on maximum! At the moment I'm just mucking about not really trying to get 'absolute' results just yet.

As far as multiple programs go, the DEQ has up to 64 memories, think that should just about cover the most obsessive amongst us!

Over here, it is only 20-30% more than the dearer of the two FBD's ( just guessing, will check it more exactly later). If I can somehow get the mic measurement from the DEQ into REW then that will go a long way towards ameliorating the extra cost, as then you could do away with one of the mixers that has phantom power. I was hoping someone with more knowledge than me could contribute to that because it would be fantastic.

Will post some more when I know some more. At the very least I am doing something whilst waiting for the BFD

lots of love

terry

[terry j]

glad I could correct my error before someone else pointed it out!

of course as all of you know, the DEQ is two to three times the prices of the bfd's!

sorry about that. If we could get the mic output into the REW then that would save us the cost of the mixer (if thats what we wanted).

Must say, the sound before and after auto eq cannot be compared! (leaving aside simply bass correction). That is enough for me to justify the extra cost, I'm just really happy to be able to use this marvellous program with it.

lots of love

terry

[terry j]

hi all

silly me, I've been hunting around and asking ( even queried Behringer, no answer yet but they very quickly acknowledged my question ) trying to get the following question answered, realized I should have asked the ones who know, and they live in the Home Theatre Shack!!

All jokes aside, as you can see from this thread that for now I'm using the DEQ 2496 in place of either of the BFD's, and very nice it is too. Am really getting the hang of the program now ( I think!), always learning with it.

Another attractive feature of the DEQ is it has the ECM 8000 ( optional, what I should have said is that it supports it) and so phantom power etc. The DEQ uses it for RTA and speaker/room correction. The RTA is quite neat, though of course the resolution isn't great, limited to the bands displayed, but more than useful if you didn't have access to the REW.

The query to Behringer was to see if there was a way to route the mic readings to the analog outputs, so I could then replace the R/S meter with the (presumably ) more accurate ECM 8000.

If that can't be done, as I suspect cause I've tried bloody well nearly everything, do any of you guys know if it would be possible to 'tap' into the mic cable and extract the mic readings that way, then input them to the line in?

Only three pins, one must be +15v, another must be 'ground', and hence the third must be the mic signal.

The pins labeled in the manual are 1= ground/shield

2=hot (+ve)

3=cold (-ve)

That is under the heading balanced XLR connectors, which I hope and am sure those of you who can answer this would already know.

Of course, if it is going to blow up the computer or soundcard then it's a non-starter, but I don't really think it should.

Anyway, it's obvious that I don't know anything about this, so please ignore my last sentence.

Thanks very much, will post back my experiences using REW and the DEQ later, it all adds to the knowledge base.

lots of love

terry

[brucek]

do any of you guys know if it would be possible to 'tap' into the mic cable and extract the mic readings that way, then input them to the line in?

No, it's not possible. This balanced output microphone requires a preamplifier to boost the signal to a line level to make it useful. It also requres a phantom bias voltage to make it work.

The simplest solution is to use the UB802 preamp that is ridiculously cheap to purchase and a very nice unit.

brucek

[terry j]

thanks Bruce, understand. Is the Ub 502 also acceptable?

Really starting to get the hang of the program, it really is a very good system. My interest is only in 2 channel, so am exploring the use of REW for more than simply sub EQ, and the more I learn the more uses I find. Of course, the R/S meter is fast becoming the limiting factor in the higher frequencies. Hence the desire to use the above units.

As far as I know, the UB series has been discontinued, but have found someone who is still listing both of the above, so hopefully I can grab one before they are all gone. Not that the replacment units are that much more expensive mind!

Will hopefully have some updates soon on using the DEQ and would you believe, finding it extremely useful with the DEQX unit as well!

lots of love

terry

[brucek]

As far as I know, the UB series has been discontinued

I don't think so. Here's the UB802 from their site.

The UB502 doesn't have a mic preamp. You need the UB802. The mic plugs directly into it, if you don't want to buy a cable.

brucek

[terry j]

OK can finally wrap this thread up, hope that it is/was of some interest to some.

My Xenyx 802 arrived yesterday, able to put the behringer mic to good use. Of course as it's not calibrated, it can't be relied upon to be totally accurate, but I'm sure any room anomolies more than outweigh any mic innacuraccies! My supplier is adamant that the UB 802 has been discontinued, so maybe behringer keep older units on their site for a while after being discontinued? Also, the main reason the 502 is unsuitable is that it doesn't have phantom power.

The anomolies at the higher frequencies are usually handled better with room treatment. You do have to be careful trusting the RS meter at the higher frequencies. Some say 5K is the upper limit.

Yes you are totally correct Bruce. However, what I did was to set up the speakers outside to measure them (boy doesn't the response clean up when away from the room!) so I could get as anechoic a response as possible. Then using the DEQ's 'target response' feature I eq'd the raw speaker response to flat. This leaves me with a flat speaker that I can then put into a real world ( listening room ) situation.As you rightly suggest, it is not the best idea to try and get flat 'fighting' the room.

This by the way is exactly the methodology used by DEQX. Flatten the speaker to a 'perfect' flat anechoic response then let the natural response in-room be as it may be, as any room anomolies up in the frequency range are best handled by room treatment, but it is always preferable to be starting with 'flat' speakers, which was the purpose of this little exercise.

The differences in the mic responses really take off after 5k

This is the plot with the Radioshack meter



outside measurement radioshack meter.jpg


And this is with the behringer



outside using ecm 8000.jpg

the HUGE rise in the response above ten k was not even hinted at by the radio shack meter, what is a bit suprising is how close the graphs are below that point.

Anyway, the following graph should be fairly self-explanatory, and the difference in sound IS as significant as the graphs might suggest!

The graphs (seperated) are of left and right before and after the auto eq function on the DEQ. As mentioned, done outdoors for as 'pure' a response as possible.



deq before and after.jpg


It was suprising just how much ambient noise there is outside that we become accustomed to, and I live on sixteen acres in the bush! No way you could do it outside in the city.


Nearly there now, have shown sub before and after eq previously, so straight to full response in room


full response.jpg

Can't do much about the dip at around 140, as an experiment threw a parametric at it, hardly changed so seems definately room related, would need treatment rather than eq.

Have not got any boost in the bass, I'm only interested in music so no need for house curve. But the room curve now shows what is supposed to be desirable, that is 20k down a few db, and that is a natural result of a flat speaker when placed in the room environment.

That's enough, hope it was of interest. I can tell you that the difference between DEQ in and out of circuit is dramatic, and I'm afraid that it is basically unlistenable without it now, like an old tinny wireless with BOOM ( as could be guessed from the raw curves?). The bass correction you are all aware of, it's the rest of the frequencies that get tamed with this unit.

Of course any tailoring to personal preferences are possible eg up the response at the top end to be flat in the listening position etc etc.

lots of love

terry

[brucek]

Really interesting stuff terry.

I have a vision of you out in the woods. Just you, your speaker and REW.......

Anyway, I don't want to send you back out there, but my feeling is that it would have been a good idea to use Sonnies calibration file for the ECM8000.

When I bought my ECM8000, I did some searching on the internet regarding the accuracy.
I found that when an ECM8000 was professionally calibrated, its calibration file was very close between units. A comparison of various calibration files I found on the internet revealed this to be true. They all drop off slightly and slowly below about 40Hz (and is only down about 1.8dB at 20Hz). There is also a consistent hump above 10Khz (consistent with your measurement).

I decided that if I used Sonnies ECM8000 calibration file, that it would therefore be close enough for me and as you say, at that point any room FR effects would be greater than any microphone inaccuracy.

I would use the calibration file for the ECM8000 I have attached next time you're out in the woods.

Attached also is the graph of the file.


ecm8000.jpg

ecm8000.cal


brucek

[terry j]

never even thought to check for a calibration file...duuhhh. It certainly does look like my response had quite a bit of 'behringer flavor' thrown in doesn't it! Have downloaded it so will re-do the measurements for comparison. Will do them slightly differently as a check of another methodology. The results won't be directly comparable unless the earlier ones are done again with the new cal file, well we will see.

I guess bush might conjure up visions of unspoilt natural environment, the reality is we are flat bang in the middle of farming territory, and previous farmers over the years have cleared the area for sheep and cattle. Out come us city slickers ( read NOT farmers) who start planting hundreds of trees ( in the worst drought in living memory mind aarggh). What was suprising about the unconscious outside noise was the content the mic picked up down to 20 hz and less! There is so much of it that at times the 'set volume' command would warn me to check I wasn't playing music, yet without the mic most of us would comment on how quiet and peaceful it is!

lots of love

terry

[terry j]

I know I've said this before, but if I keep saying it then logically ONE day it must become true, but think I can wrap this thread up!!!

Thanks Bruce, as you can see the Behringer Mic cal, even tho 'generic' has resulted in very different measurements, but surely in all likelihood vastly superior ones.


This graph shows left and right measured with and without the mic cal file, as you can see the huge hump after ten k has disappeared

mic cal comparison external .jpg


Funny how you can talk yourself into things, I distinctly remember thinking when I first saw the hump after 10k " Gee, didn't think it was that bad" !! but carried on regardless. Just goes to show that we certainly need measurement gear to prevent all manner of 'wishful thinking'.

OK, this time, now armed with accurate measurements, my procedure was to measure in the room with the mic on tweeter level about one meter away, then set a flat target level and let it rip. After completion, a bit of back and forward fine tuning and tweaking to get the response pretty **** flat.


This graph shows left and right, before and after tweaking the response.



left and right before and after, using mic cal, done in room.jpg


It may look a little cluttered, but I think you would agree that it is a pretty flat end response.

Then of course did the sub, it's a roughie but at least it's here and able to be experimented on, no need to show that cause I do think you lot already know all about eq'ing subs ha ha. This thread is about something else!

Anyway, here is the final in room response, both mains driven plus sub. You may recall the sub has some pretty nasty resonances between 120 and 200, so with a bit of judicious back and forth to help clean that up here it is, viola


final in room response, mic cal, done indoors.jpg



The room it's set up in is quite small, the computer room ( funnily enough, cause thats where all this learning has taken place) so no attempt for any room treatment, but if I were to do so I reckon I've got a pretty good place to start from. And indeed, the beauty of this is as I make changes to the room response with treatment, it is no bother to go back and fine tune as I go, an iterative kind of thing. Try that with a passive network!

Thankyou all for your help and interest, it is another card thrown into the pile for consideration in addition to the 'simpler' BFD's. My main system is tri-amped and I will be using the DEQX in that one, so the DEQ has no place for me there. However, if I had a 'normal' system ( branded name with passive network ) then I simply would not hesitate to grab a DEQ over the others, all of the abilities of the BFD's but with much more, as hopefully I've been able to show. Apologies if I've been dense in my explanations.

After all, what speaker with a passive network is actually flat? Not many I'd wager

lots of love

terry

[terry j]

tried to edit but couldn't work it out. Must be thick.

Maybe a BFD might not have worked with my sub. The response is pretty crap, so I had to boost it to the max to get it way over the target response, in order to give me something to cut back. This meant that in places the response was WAY over the target, and the PEQ only has 15 db of cut. By using the unused bands in the GEQ I was able to get rid of the majority of the signal leaving the PEQ to clean up the rest.

Maybe the BFD's have more cutthan that and so it wouldn't have been a problem, but what I'm saying is that having the GEQ as well could solve some probs that a PEQ could only partially do???

lots of love

terry