Equalizer update

[pbirkett]

A little comment regarding the onboard DAC, a member of zerogain has the DEQ2496 and he says its comparable to his Arcam Alpha 9 CD player in quality, which is no mean feat....

http://www.zerogain.com/forum/showth...hlight=deq2496

[jhenderson010759]

I think the DEQ2496 DAC is excellent also. I use it to drive my Stax 006t amp via the XLRs. Very engaging, musically.

The interesting thing about that particular setup is that I can't seem to find an eq setting that makes the Stax sound better to my ears than unmodified response. I have the frequency response measurement from Stax, and I've used it to create very subtle eq patterns which provide, for instance, flattened bass response. But upon critical listening sessions, I still prefer the unmodified response. Curious, because the measured Stax response cries out for a +3 dB shelving filter below 1 kHz. OTOH, 3 dB isn't much of an amplitude variation.

Now, this is not true for the DEQ -> Grace -> HD650 setup. In that case, I much prefer the flattening of the bass hump, as outlined in gerGs post (above), although I tend to prefer leaving the midrange and treble unmodified.

[Steve999]

Hey jhenderson, I'm a 5-month-plus DEQ2496 user, this was the piece of gear that made everything worth it for me. I use mine every day and do something a little different with it nearly every day. I'm really enjoying your posts. Are you a digital sound processing engineer? (I thought I read that or something like it in your profile.) I feel like the true lucky one, I get the benefit of the discoveries and expertise of folks like you and gerG. I would never have stumbled on this stuff or known how best to use it on my own.

You guys are so far over my head I really feel I have nothing to contribute! And since I got the DEQ2496, the great headphone quest has been over for me. If I want a new sound, I make one! What to say?!?!? Other than, thanks for all the great info, it's wonderful to read the new posts and have a way of sharing experiences with what for me is a very treasured piece of gear.

It's exciting to see more people trying this out and experiencing it for themselves.

All the best!!!

[jhenderson010759]

Steve -

Thanks for the kind words, but I am far less knowledgeable about audio as gerG and many others lurking around here.

Our DSP boards are primarily used for industrial control and data acquisition applications, not audio. But, sometimes the technology is similar. For example, we make a card with 32 channels of 192 kHz, 24-bit A/D for sonar apps. Developing a card like that is a humbling experience - the smallest layout or component selection error degrades performance a few dB. Maintaing 100 dB of SNR and 95+ dB of SINAD with minimal crosstalk across a large channel count is a hellacious task. That <$100 sound cards produce numbers greater than this (though with fewer channels) is... fantastic.

Being and engineer and an audiophile is oxymoronic. On the one hand, I have to dismiss as nonsense all of the posts I have read about being able to discern sound quality differences between two different TOSLink cables or someone discussing "jitter" over a USB cable. ABX test anyone?

I favor "auro-nihilism". But, I don't think we understand everything that must be measured yet.

I realize now that, I don't always want accuracy in my audio chain. I think I've probably attended over one hundred classical guitar concerts in my life. But, if my gear sounded exactly like some of those concerts, I'd sell it. No, what I am seeking is enjoyable, musical hyper-realism.

So, the quest continues....

Jim

[tomek]

hey gang, question about using the Behringer.

in the manual, it says to only flatten the room for 100hz and up when using the auto EQ feature with pink noise and a mike.

i've done this now, and have fine tuned my system to an amazing level. however, i notice that when the machine is adjusting pink noise levels, it does have everything flat at 100hz and up, but below that, there is a dip in my system at 80hz, a significant one, then it is back on level at 40hz.

my speakers are quite large, and are supposed to be flat to 32hz. is hole in the bass part of the room, and show i try to adjust for it with the behringer unit, or is there a reason the manual recommends against it. is this something a 'bass trap' solves? it seems quite a big dip when the rest of the speaker is now perfectly flat.

ps: i love this thing. i can't listen to my speakers with the EQ off now. i have trouble imagining i ever did in fact.

[kwkarth]

Originally Posted by tomek

hey gang, question about using the Behringer.

in the manual, it says to only flatten the room for 100hz and up when using the auto EQ feature with pink noise and a mike.

i've done this now, and have fine tuned my system to an amazing level. however, i notice that when the machine is adjusting pink noise levels, it does have everything flat at 100hz and up, but below that, there is a dip in my system at 80hz, a significant one, then it is back on level at 40hz.

my speakers are quite large, and are supposed to be flat to 32hz. is hole in the bass part of the room, and show i try to adjust for it with the behringer unit, or is there a reason the manual recommends against it. is this something a 'bass trap' solves? it seems quite a big dip when the rest of the speaker is now perfectly flat.

ps: i love this thing. i can't listen to my speakers with the EQ off now. i have trouble imagining i ever did in fact.

The reason you only want to auto calibrate 100Hz and up us because of the wavelengths of sound below 100Hz and how those frequencies interact with your room geometry to produce standing waves and rarefactions. Moving the mic or the speakers even one inch can produce drastically different results. Put on a 50Hz sine wave tone and walk around the room. It helps illustrate the problem if you also augment your listening journey with an audio sound level meter. Move your ear and the sound level meter near a side wall, move into the corners of the room. You will most likely notice drastic differences in sound level. You can't properly equalize nodal room modes. You have to treat the room to remove those problems.

Yes, this is where tube traps and the like come into play.

[ooheadsoo]

Hey tomek, what mic did you use with the behringer?

For some Jon Risch SQ&D (simple quick and dirty) bass traps, go down to whereever you can find bales of fiberglass for sale. Not rigid fiberglass, just the regular stuff stuffed into clear plastic bags. Buy 3 bags for each corner of your room. Leave the stuff inside the bags, stack them in the corners of your room. Presto. Decorate/dress them up as necessary. Most stuff won't stop the bass traps from being effective.

Or you could buy rigid fiberglass tubes. Jon Risch doesn't think it works ideally, but it does work according to many people who have tried it. You might even stuff the tubes with the fiberglass mentioned above Again, decorate the tubes as necessary.

[gerG]

tomek, which model of deq did you get?

Be careful with the auto-eq function. It is really intended for large spaces, and yields funky results in a normal sized room. You may end up with sliders yanked all over the place. Although such an adjustment may measure flat on the rta, it probably is not. Work for smoothness.

Man, lf irregularities are frustrating. You should get a cancellation trough and a reinforcement peak associated with each dimension of the room. The trough will be at a frequency associated with a wavelength twice the characteristic dimension. The associated peak will be at twice that frequency. The equation is simple, f = a/l, where f is frequency, a is the speed of sound, and l is wavelength. As an example, my room is 14' wide by 8.5' high, and I get this (subs only):



First trough is about 41 hz, speed of sound is 1134 ft/sec, which tells me that my room has a dimension of 13.8 ft. There is also an associated peak at 82 hz. Pretty cool, huh? I am lucky that I have no rear wall, which is the real beast in these situations. The next trough is the ceiling.

Anyway, you can eq the peaks down, but do not try to boost the cancellations. You may run out of amp before you get there.

Options for getting rid of the bass troughs are:

1) Bass traps on the offending walls (one of which is overhead)
2) Active control (too complex for most setups)
3) Get rid of the walls
4) Ignore the dips, eq the peaks, and get on with the music
5) Fair warning, I could be one of the dips

Moving speakers around and shifting listening location will move things a bit, but the basic modes will persist.


gerG

[tomek]

hi there. i got the deq2496. as for the mic, i'm not using the Behringer one that they recommend because the shop didn't have it, I'm using a substitute, another omnidirectional flat response mic.

i've had some differing results with the auto EQ function. on a trial to trial basis, with the same target curve i'll get significantly different adjustments.

however, i've got one setting now that is just fantastic. it really puts a big boost on at 8k+, but is a lot more subtle before that. the only thing that bothers me is the +8.5db adjustments in the 100-160hz range, which is a result of room interactions that i shouldn't really be EQing I guess.

tonight i'm going to fiddle more to give the vocals a little bit more 'fullness'. right now they're prominent in this setting, but they lack a certain smoothness that i get on other settings. i'm trying to blend the best of some of these settings to get it right where i want it, so i never have to touch it again.

[tomek]

Originally Posted by gerG

. Although such an adjustment may measure flat on the rta, it probably is not. Work for smoothness.

what do you mean, aim for smoothness? do you mean smooth out the sliders so they resemble a curve more than the random up and down positions they are in now?

[gerG]

Hi tomek. Sorry for the confusion, but for some reason I thought that you had the older 8024 unit.

At any rate, yes, that is exactly what I meant. There is some level of averaging going on in the rta, and I am not certain how the edges of the sample bins line up with the eq bins. I get suspicious when the auto-eq comes back with an up/down sequence in an area that is relatively flat to begin with. I like to use the geq for large offsets, and the peq for local stuff, if necessary.


gerG

[tomek]

hi again,

i'm looking for some help with this thing. when it auto eq's, it does it in that up and down fashion you mentioned. it will have one frequency at normal, and then the one on each side significantly altered.

the unit has removed some 'tinny' sound from my stereo. the imaging is better and it is smoother, but the vocals seem recessed. how would you suggest i alter the auto eq to remedy this. is the peq better, or should i smooth out the geq? and which frequencies will bring the vocals forward?

[jhenderson010759]

During the automatic RTA equalization process, clicking the Page button will toggle you between pages 2 and 3. On page 3, you can adjust the Delta and Max parameters, via the rotary knobs.

The Delta parameter controls the maximum permissible amplitude difference between adjacent graphic equalizer frequencies. If you make this a small number, say 2 dB, you are indicating that the resultant equalization curve must contain no "abrupt" adjustments greater than 2 dB. The output curve will be much "smoother".

The Max parameter specifies the largest permissible excursion between the highest and lowest equalizer adjustment in the resultant equalization curve. So, by setting this to, say 8 dB, you indicate that the automatic RTA may not make adjustments to the response whose total span exceeds 8 dB.

[Jahn]

EQ is really helping with these old recordings I have of the Beach boys. now if i only had some crossfeed...

[jhenderson010759]

If using a PC as a source, Foobar supports a number of crossfeed filters. The standard filter is subtle and musically unobtrusive. Others, such as the 4Front crossfeed filter are more obvious.