[dayneger]

Well, I looked around for a while and apparently there already are some interesting solutions for my Level III wish.

I'm considering trying out one of the Behringer Ultracurve DEQ2496 units.  Some people are using them in the digital domain, between the transport and DAC, which should hopefully keep any sonic impact to a minimum since it's not going through an additional ADC and DAC section.

Basically it gives you a 31-band equalizer plus 3 parametric filters to handle nasty room peaks.  With their microphone in use, it has an automatic pink noise sweep that accelerates the initial setup substantially.

Sounds like a very useful device to have in front of a DAKSA!

Anyone have experience using this guy or the predecessor, the 8024?

Cheers,

Dayne

[mb]

I'm considering trying out one of the Behringer Ultracurve DEQ2496 units.  Some people are using them in the digital domain, between the transport and DAC, which should hopefully keep any sonic impact to a minimum since it's not going through an additional ADC and DAC section.

Basically it gives you a 31-band equalizer plus 3 parametric filters to handle nasty room peaks.  With their microphone in use, it has an automatic pink noise sweep that accelerates the initial setup substantially.

Yes, me... I have a little review at AA. It is invaluable for me in controlling my floor/ceiling mode, and with some considerable modding, I've brought its dac section almost up to my upgraded CI-Audio VDA-1. FYI, the AutoEQ function is ok for midrange and treble eq (if you want it), but way too broadband (1/3 octave) for LF modes.

When I get my hands on a DACSA. it will certainly get it's signal post DEQ! It will be very interesting to check if the reclocker will bring benefit to a chain such as this.

[dayneger]

Hi mb

Yes, me... I have a little review at AA. It is invaluable for me in controlling my floor/ceiling mode, and with some considerable modding, I've brought its dac section almost up to my upgraded CI-Audio VDA-1.

That's funny that you replied, it was your review that got me very interested in the unit in the first place, especially using it just in the digital path!  Have you been tweaking the DAC part just for fun?

When I get my hands on a DAKSA. it will certainly get it's signal post DEQ! It will be very interesting to check if the reclocker will bring benefit to a chain such as this.

Most definitely!  Hugh, do you have any thoughts on that?  I'm not really sure what's involved with things in the time/phase domains when EQing in the digital realm, either.

Do you have any tips on how best to dial things in with the DEQ?  And what do you need for cables?

Dayne

[AKSA]

Hi Dayne,

I'm aware of the difficulties with rooms, but unable to offer advice on the use of digital EQ.  This is Ben's domain, and he's unfortunately not available right at this moment.

My gut feeling (substantial, at my age.... ) is that we should strive to get the room as good as acoustically possible, and THEN use digital EQ to correct the remaining anomalies.  This, to me, is analogous to designing an amp to be as linear as possible, THEN applying global feedback to scotch any remaining non-linearities.

I'm sorry I can't be more informative than this, Dayne;  it's not my area of expertise and I'd be a fool to overstate things here.

Cheers,

Hugh

[JohnR]

I had the 8024, but I didn't use it for actual listening. Apparently it could be modded to sound good but I couldn't see that it would be worth the effort.

One thing to remember about these things is that the digital processing assumes a certain sample rate. I understand that these units, being from the pro market, use 48 kHz internally. If so, then when using digital in and out, the unit has to perform a 44.1 to 48 kHz sample rate conversion, and vice versa, a process that introduces its own noise products. I'd be curious to learn more about this.

As Hugh indicates, room modes are in fact resonances. While you can EQ out the peak in the steady-state, you can't remove the resonance that way.

[mb]

My gut feeling (substantial, at my age.... ) is that we should strive to get the room as good as acoustically possible, and THEN use digital EQ to correct the remaining anomalies.  This, to me, is analogous to designing an amp to be as linear as possible, THEN applying global feedback to scotch any remaining non-linearities.

There's no doubt that I would prefer allowing room treatment to keep my layback chain as "pure" as possible, but here was my scenario:

- (large) main mode at 57Hz, due to 10" ceiling, hard floor
- limited space / acceptability of large traps required
- digital-only source (vinyl would have meant an additional A/D conversion in the chain)

My main mode requires a 1/10 octave -14dB notch. Yes, horrors! This was verified by measurement, but most importantly, fine-tuned by ear. I can't imagine what sort of trap would have worked similarly, without sucking out the rest of the bass. This notch is not noticeable, except for the few pieces of music that excite the mode (a rare note, as it's not precisely on the modern tuning system A=440Hz; appears on some authentic instrument recordings, where A=415 or 430??).

After the DEQ, I've purchased some Eighth Nerve devices for additional tuning. They really don't address the same room issues as a DEQ on LF modes.

Re: quality of the analog section of the Behringer? You get what you pay for (and more). A very fine dac section for US$300 + free EQ, or a fine digital EQ + free dac section... Some tweaking on the DEQ brings it up to the level of a very good dac (DIO level III, better than most US$500 cdps). Otherwise go for a high end dac.

[Tinker]

Hi mb

Most definitely!  Hugh, do you have any thoughts on that?  I'm not really sure what's involved with things in the time/phase domains when EQing in the digital realm, either.

Do you have any tips on how best to dial things in with the DEQ?  And what do you need for cables?

Dayne

Hello All,
   an interesting discussion brewing here.

Warning: some opinion ahead!

I should start by saying that a number of new HT and car audio developments make extensive use of EQ to try and achieve certain effects. Used judiciously these can be decent solutions to bad room problems.

These digital units usually perform filtering using one of two methods: IIR or FIR. BOTH methods introduce time domain anomalies into the signal, in particular ringing and degraded transients. Both types affect the phase of the singal although FIR filters can be made linear phase by using symmetric filter structures and can be extremely transparent if enough DSP is thrown at them. IIR filters are more efficient computationally, so many graphic EQ systems use these, although some of the better ones use FIR - especially now that DSP is becoming cheaper.   The Klark Technic systems for instance, but if I had $11K to spend I'd spend it on the room.  


I used to do a bit a live sound stuff where this kind of EQ is vital. However, in the home you need to ask yourself if radical EQ to flatten out the room is worth the sonic degradation in other domains (impulse, phase etc). High frequency room problems can often be remedied with some cheap, and high SAF wall hangings and concealed foam, but I would have to concur that short of rebuilding the room from scratch, taming bass modes is really only feasible with EQ, and great results can be got this way which will often outweigh the sonic penalty of an EQ.

A good way to get the best of both worlds is to use a parametric EQ and attack only the problem frequencies, leaving the small dips and bumps elsewhere alone. Parametric EQs are often linear phase FIR so less harm is done. You'll need to check that the filters are capable of fairly high Q. Of the order 50 or more to get anything like a sharp notch. Of course a lot of these digital eqs do this no problem. I have a DEQ with a Q of 100.

If starting from scratch wth room taming it is worth downloading one of the many free realtime analyser packages from the web and buying a cheap measurement mic from a local electronics or audio shop and measuring your room. This way you'll get a feel for what is going on. A flat room isn't necesarily needed for great sound. Sure you want it *pretty* flat, but having a smooth and short RT60 (time for a test pulse to drop by 60dB) at all frequencies (so smooth REVERB) is much better for sound quality (vis a vis bass booming which is a kind of reverb I suppose).

One final comment, I think JohnR has a point. I don't know the specs of the 8024 or the DEQ 2496 (but I'll chase them up). Many units do use sample rate convertors and this came up in the AES circles about 2 months ago. jury is still out, but general vibe is that multiple sample rate conversion steps introduce some real nastyness, which can be acceptale with "good" systems and downright vile with "bad" ones.

My 2c. Grab a book on acoustics. Measure the room first. EQ the bass modes (unless you hav dipole speakers!) treat the treble with wall hangings. This approach has the biggest impact on quality, for the least cash with the least deleterous impact on sound.

Cheers,
   T.

[mb]

...
One final comment, I think JohnR has a point. I don't know the specs of the 8024 or the DEQ 2496 (but I'll chase them up). Many units do use sample rate convertors and this came up in the AES circles about 2 months ago. jury is still out, but general vibe is that multiple sample rate conversion steps introduce some real nastyness, which can be acceptale with "good" systems and downright vile with "bad" ones.   ...

I'm very keen to get the final verdict on whether there is internal resampling. I hope not! Seems like having an Aspen Level III reclocker after all the DEQ stages should be a Good Thing, no matter whether there is reclocking or not.

FYI, my eq is limited to a few notch filters in the 32-130Hz range. Only as many as seem to be needed to control room artifacts. The mid/treble ranges are essentially pass-thru'.

[dayneger]

Thanks for chiming in, Tinker.  Your suggestion of using the parametric notch filters for the bass sounds like a good way to avoid some of these other issues.  This seems to be what mb has done with his system, apparently to good effect.  I'm interested to see what kind of measurements my room will cough up!

Whether there is internal resampling, and what impact it has on the sound (if any)?

MB, have you tried using the EQ at higher frequencies, or doesn't your room pose any problems?

[mb]

The AutoEQ function on the mid/upper freq gave me GEQ which was more or less within +/- dB throughout. The resultant sound was ok, but something was missing. I bypassed the GEQ, keeping only the PEQ for the bass, and things improved a lot. The target contour I used was slightly downtilted, so it would seem that the inroom mid/treble on my speakers is not bad . Stereophile (John Atkinson) had some inroom measurements with the speakers (Gradient Revolutions), "... astonishing ±1.3dB limits in-room, from the 32Hz 1/3-octave band to the 10kHz band", so I really am using the DEQ for room modes, not contouring.

[Tinker]

Seems like having an Aspen Level III reclocker after all the DEQ stages should be a Good Thing, no matter whether there is reclocking or not.

FYI, my eq is limited to a few notch filters in the 32-130Hz range. Only as many as seem to be needed to control room artifacts. The mid/treble ranges are essentially pass-thru'.

I am not afraid to apply the Technology where needed, but I do endorse the general idea of the less is more principle. Resamplers are used a LOT nowadays, because it is the easiest way to avoid sample rate compatibility problems. Sad but true, we will see more of these in gear yet to come. Some of them are VERY good, however. The latest AD device is a good example of the evolution of this technology.
I'm sure the AKSADAC will be a good thing, even with some resampling or filtering before it.  The final D/A and all the black magic in it is very important.

I would be pretty confident that the 32-130Hz range would be sufficient to tame the bad stuff in the bass. The baffle step and boundary effects kick in just above this, probably swamping other effects, and bey then we would be close to the schroeder frequency where foam and carpet will work wonders.

[Raj]

Hi Tinker,

I'm assuming that the aksadac will upsample? To what frquency range, sorry if this has already been stated........
Thanks
Raja

[Tinker]

Hi Tinker,

I'm assuming that the aksadac will upsample? To what frquency range, sorry if this has already been stated........
Thanks
Raja

An early proto used an upsampler, but we were never satisfied with this for a number of reasons and it wasn't clear what target sample rate to use.

The AKSADAC does not use upsampling. It is meant to be a pure digital path technology, so the samples go through unmolested. I feel this can be quite important for recordings that make heavy use of noise shaping. It does use oversampling, so of course there is a hint of processing in the interpolation filter, but this is several dB below the theoretical maximum CD noise floor.

There are other reasons for this choice, which I will discuss (or have already discussed) in due course. The main considerations which lead to these design choices are certain properties (like linearity) of the analogue stage and low jitter in the digital stage.

Cheers,   
   T.