Searching Equalization That's Digital in / Digital out

6495 Views 30 Replies 6 Participants Last post by Audioguy, Jan 9, 2016

tpointon

Discussion Starter · Dec 26, 2015 (Edited)

This will be a rather longer post to give some background.

In a nutshell, what I'm seeking is alternatives to the miniDSP DDRC-22D. I have nothing against the DDRC-22D, but usually I like more than one option when considering a purchase.

If you choose to read further, I would appreciate any comments on what I'm about to say. I need to know that I'm getting this more or less right and where there might be weak points in my pursuit of audiophile music reproduction on a Honda Civic budget.

My 2.1 audio setup has a NAD C390DD 'all-digital' amp at it's core. The signal path I'm currently using is:

Digital source > Raspberry Pi w/ MusicBox > HiFiBerry Digi+ > MiniDSP nanoDIGI > NAD C390DD > Axiom M80 x 2 + SVS 12 sub

I've been experimenting with REW to do room EQ with pretty good success. The C390 has 6 EQ filters at fixed frequencies ranging from 40 hz to 240 hz each -12 dB to +4 dB. Using REW, I set those filters as REW specified with a crossover at 70 hz (I think) sweeping 20-300. Then I did a second run of REW to calculate parameters for the nanoDIGI sweeping 80-8000. The nanoDIGI has 5 parametric EQ parameters. I used that sample to generate 4 configs for the nanoDIGI having house curves of -1.8, -2.0, -2.2 and -2.4 dB/octave.

Here's where I've arrived, showing the 2.2 dB/octave setting.

I like what I'm hearing, but I don't have much experience with this room EQ activity. I think if I am to pursue room EQ beyond where I am now, I will need more and/or better matched filter options than I have now.

From what can see, people who've been doing this for years would have been in the analog domain until the equalization stage where they would use say a BFD to go analog-digital_EQ-analog and on to analog amplification.

Obviously with my setup, I stay digital throughout and don't want to (nor can I) drop down to analog just to match the inputs to an EQ.

I think the miniDSP DDRC-22D is one device that would serve.

Does anyone have experience with the DDRC-22D that they would like to share?

Does anyone know of alternatives to the DDRC-22D?

Thanks for listening. I look forward to comments.

Happy Holidays All !
Tim

See less See more

Save

21 - 31 of 31 Posts

tpointon

Discussion Starter · #21 · Dec 29, 2015

tonyvdb said:
Not to steer this off topic but it seems that some room treatment would be a better option for you. Using as many filters as you are is not going to give you great results. In my experience the difference in most DAs is negligible.

Thanks, Tony. I will seriously take your advice into account.

Here's a little anecdote that emphasizes what you say. At one point I was trying out a listening position in my room. I could not seem to get the image to be centered ... everything seemed shifted right, at times further right than my right speaker. I began to wonder if I had a bad speaker so I switched L-R and tried a few other things. But there were no hardware problems. There is a patio door on the right wall of the (small-ish) room that has no window covering. I was sitting where a clear reflection was coming off that glass!

Save

bertadora

· #22 · Dec 29, 2015

Hi Wayne,
Regarding over equalizing, please note that I am doing full range equalization with floor standing speakers that goes down to 28Hz, I have a REL Storm III that I used before with my old speakers that went down to 45Hz. Currently the Sub-woofer is disconnected, I may consider later to ingrate it to my system to cover the range of 18-28Hz.
I am using the DEQ2496 parametric equalizer for creating both narrow notch filters to suppress room resonances and for smoothing the SPL curve for cases that needs narrower than 1/3oct bandwidth peaks that can not be treated with the 31 1/3oct graphic equalizer of the DEQ2496.

My listening room 3.3x3.3x2.4 (WDH) meters and its treated with Bass traps and absorption panels
Please see below predicted room resonance of my room from REW room simulation.

My goal is to suppress and bring them below ~300ms RT60 with priority to low frequency resonances, high gain and high RT60 resonances.
The parametric of the DEQ 2496 has 10 filters and should be used wisely to suppress both room resonances and for smoothing the SPL curve for cases that needs smaller than 1/3oct bandwidth peaks
see below my room SPL curve with DEQ2496 on "bypass" no equlization (Boomy bass, no High roll-off, bad sound)

See my room resonances with DEQ2496 on "bypass"

The DEQ2496 set manually using REW analysis data resolves most sound problems both resonances and SPL curve.

I can create a step by step procedure of how to use REW to manually set DEQ2496 GEQ (Graphic Equalizer) and its PEQ (Parametric equalizer) to achieve smooth SPL with roll-off and suppress room modes with narrow 1/10 Oct bandwidth and post it in REW and Equalizer forms for anyone that has DEQ2496 and wants to try my method.

Your feedback will be much appreciated
Regards
Issi (bertadora)

See less See more

Save

tpointon

Discussion Starter · #23 · Dec 29, 2015

It seems I started a lively conversation here. Thanks to everyone who contributed such useful advice!

For the immediate future, I have decided to concentrate on room treatments and working with the equipment I have.

My plan:

I have a laptop on which JRiver Media Center is running legitimately. When I started using it, I didn't want to dedicate my laptop to the listening room. That was a significant reason for my looking into the Raspberry Pi.

Recently I bought a used laptop to run Ubuntu Linux and see if I can work mostly with Linux, a browser (Firefox) and my data in 'the cloud'. I'm no fan of Windows.

So I moved my Windows laptop to the listening room to be my HTPC and the Linux laptop to my desk, from which I type as we speak.

JRiver works nicely with my NAD amp via USB for which NAD provides drivers. Also, I think I can work with JRiver to do filtering for house curves etc. A new learning curve, pardon the pun.

Since JRiver doesn't "talk to" Spotify as far as I can tell, I will use my Raspberry Pi to play Spotify. Perhaps at some future time JRiver will work with Spotify (if you know it does already, please tell me how).

Once again, thanks. You guys are awesome helpers. Hopefully I can contribute more in future as I learn more.

Cheers!
Tim

See less See more

Save

bertadora

· #24 · Dec 29, 2015

Hi Tim,
What is the size of your listening room?
Is it rectangular or shoes box type?
Your decision to first improve your room acoustic treatment is the right one.
DSP should be used only after it for additional improvement in the sound.
In my small listening room acoustic treatment helped but only the DEQ2496 made the breakthrough.
Regards,
Issi

Save

tpointon

Discussion Starter · #25 · Dec 29, 2015

bertadora said:
Hi Tim,
What is the size of your listening room?
Is it rectangular or shoes box type?

My room is 16' 5" x 13' 4" x 8' ( 5.1 x 4.1 x 2.4 metres ) = 1760 cubic feet. Here's the layout in REW Room Sim. There's a fireplace along the front wall that prevents moving the speakers closer together unless I come well into the room.

I've switched to JRiver MC to play music from my NAS through the NAD amp and it sounds very good. I was able to use REW's settings for a house curve to customize JRiver. I'll continue to work in that direction and to read up on and apply room acoustic treatments.

I appreciate your advice about the DEQ2496 ... if I decide to purchase a DSP, it could well be that one.

Thanks,
Tim

See less See more

Save

Wayne A. Pflughaupt

· #26 · Dec 30, 2015

bertadora said:
See my room resonances with DEQ2496 on "bypass"

Are you saying that your room resonances are where all the blue pointers are in the top graph?

Regards,
Wayne

Save

bertadora

· #27 · Dec 30, 2015

Hi Wayne,

The arrows up to 200hz are measured room resonance by REW, you can see the measured vs the predicted by REW room simulation.
The arrows from ~200hz up to 400hz we can relate them as room reverberations with decay above 300ms.
Recommended RT60 for small control/mastering room like my listening room is ~300ms, and it is also the default value in REW for "find resonances"

My goal is to suppress and bring them around ~300ms RT60 with priority with filters selection to low frequency resonances, high gain and high RT60 resonances as number of parametric filters is limited to 10.

See below spectrogram with DEQ2496 GEQ and PEQ filters "ON" done manually.

Regards,
Issi

See less See more

Save

Wayne A. Pflughaupt

· #28 · Dec 30, 2015

I’m no expert on this subject, but F. Alton Everest’s Master Handbook of Acoustics recommends 400-500 ms as the ideal reverb time for a studio control room, so a target of <300 ms seems pretty optimistic.

For starters, I wouldn’t be equalizing all those little ripples in response above 200 Hz that you’re showing. It isn’t going to give any audible benefits, and can actually be detrimental. Equalizing with filters as fine as the 1/10-octave type your talking about are tight enough to actually attenuate notes of the musical scale, if they happen to hit the fundamental of the note. So for a run of bass notes, for example, you could audibly notice a particular note at a lower volume than those before or after.

Second, if you take additional measurements some months down the road you will find that they are not perfectly identical to the previous ones, and REW will recommend a different set of filters. So – which measurement / recommended filter set is the correct one?

Fourth, it’s pretty well known that equalizers are not the proper tool for addressing acoustical issues. Absorption is the best method for reducing high Q resonances. An equalizer can only reduce the gain of a specific frequency; it has no effect on its decay time (other than what comes normally with a gain reduction). As such equalizers are mainly good only for smoothing peaks and troughs in response.

Again, I recommend reviewing the material mentioned previously in Post #19.

Regards,
Wayne

See less See more

Save

bertadora

· #29 · Dec 31, 2015 (Edited)

Hi Wayne,
I fully agree that first priority is Acoustic treatment that should be done in the small listening room as much as it can with the limitation of space and ascetics.
I am familiar with the F. Alton Everest’s Master Handbook of Acoustics from 1981, I have it and read it all.
I like to refer you to SOS magazine article from PAUL WHITE from 1985 that is title is "PRACTICAL ACOUSTIC TREATMENT" Part 3 https://www.soundonsound.com/sos/sep98/articles/acoustic_3.html

I quote from the T60 section "The ideal reverb time varies depending on the room size and the type of material being auditioned, though for a control room it's likely to be around 0.3 seconds"
I like to hear exactly how the material sounded in the control / mastering room.

Today, many of the young generation are listening to music material through smart phone and tablets and some of them with good quality head phones that does not have a reverberation as in a listening room.

Mastering engineer today have to make sure that the material they produce will sound good on both headphones and in a listening room. The result usually is a compromise with some extra reverb in the recording as there are no different versions published for headphones and listening room.

I listen to lots of opera material and to get the singers to sound natural the 300ms RT60 in my small room sounds right to me with comparison to high quality headphones.

The DEQ2496 parametric filters (in my case the 1/10 Oct filters) can be activated or deactivate.
When you detective the PEQ the GEQ that is set with its 1/3oct 31 sliders to target curve equalization is still active.
The real impact of the narrow PEQ filters can be evaluated easily by activating or deactivating the PEQ and listening to many quality recordings (My recommendation - Large drum impacts of symphonic orchestra, Pop music, Jazz, symphonic music, concertos of violin, piano, trumpets...)
Best regards,
Issi

See less See more

Save

AudiocRaver

· #30 · Jan 1, 2016

I agree that 0.3 seconds is a good T60 for a control room.

Save

Audioguy

· #31 · Jan 9, 2016

AudiocRaver said:
One more vote for minimal EQ. Ditch the graphic, minimize the parametric, or go with just 3 or 4 parametric - on the worst peaks - plus Dirac Live

Save

21 - 31 of 31 Posts

This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.

Top Contributors this Month