AVS Forum banner

**miniDSP DDRC-88A Official Thread**8-channel AI/AO Dirac Live in a box

1M views 18K replies 427 participants last post by  halcyon_888 
#1 · (Edited)
Guys,

This is a dedicated thread on the MiniDSP product called DDRC-88A home theater in a box for those who are interested in Dirac Live Room Equalization solution. Only applicable with existing pre/pro/receivers that have an 8-channel analogue output to separate amps.

What is Dirac Live?
Dirac Live digital room correction technology optimizes the sound system both in terms of it's impulse response as well as the magnitude frequency response. Dirac accomplishes room correction through a technique called mixed phase. I.e. it is not IIR nor FIR, but something in between.
A technical write-up on Dirac Live by Mathias Johansson from Dirac's website is here: http://www.dirac.se/media/12044/on_room_correction.pdf

Details of the DDRC-88A are available here: http://www.minidsp.com/products/dirac-series/ddrc-88a

Wiring diagram hookup for the DDRC-88A:


Owner's manual is available.


Cost is under a grand USD - not including shipping.
  • For European AVS members, contact Homestudio with details provided by Keith Barnes in post number 522.
  • For UK AVS members, contact Kalibrate with details provided by Keith Barnes in post number 593.
  • The rest of the world will have to order from MiniDSP based in Hong Kong.

Again this is an external box solution for:
  1. those who have external amps to drive speakers; or
  2. those who have active speakers.
If you have a receiver with internal amps, sorry it's not going to help much.


The DDRC-88A has both balanced and un-balanced inputs via Phoenix terminal blocks. A solution for those with XLR pre-outs has been given by miniDSP in their forum thread post number 17019. MiniDSP recommends to:
  • buy a short XLR patch cable;
  • Cut it in the middle;
  • Use the XLR male to bare stripped wire for the DDRC OUT. You just "screw" the bare stripped wire to the pluggin phoenix connector we provide in the package (see user manual); and
  • Use the XLR female to bare stripped wire for the DDRC IN.
AVS member Jerry has provided step-by-step details for XLR cable connection in post number 729. Post up-dated to include heat-shrink for a more professional looking finish. Thanks Jerry!


Analogue input and output connection details.
  1. MiniDSP has confirmed that it's possible to have a mixture of balanced and unbalanced signals. One has to take care of the gain structure - see 2nd post for additional information.
  2. On the analogue input side; use either the balanced or the unbalanced connection - not both on a per channel input basis.
  3. It is also possible to have unbalanced input and balanced output or the other way around.
  4. On the output side, both the the balanced and unbalanced terminals are active; i.e. 'hot'. So both can be used at the same time.


Features:
  • Bypass of the DDRC-88A is available. Good for those who would like to compare their current flavor of room correction such as Audyssey, ARC, YPAO, MACC, AccuEQ, PEQ, etc. vs. Dirac Live. Note when Dirac Live is ON, the REQ within the pre/pro/receiver needs to be disabled with distances/levels set to zero. See Markus post number 611 for details.
  • Correction is adjustable for start and end frequency as well as tilting to a custom house curve. I.e. One can start from 20Hz up to 1kHz with a 10dB tilt and no correction below 20Hz or above 1kHz.
  • Can store up to 4 presets of adjustable room curves.:) Dirac Live has single seat positioning (often referred to as the money seat) or a wider sofa seating positioning measurement options. The single and sofa seat + 2 more can be saved into the presets. This would allow some flexibility in tuning the sound depending on listening situation. I.e. having family/friends over vs. critical listening by yourself. Note that an unlimited number of projects can be saved on the PC for loading into anyone of the presets.
  • Can do multi-zones. I.e. Up to four 2-channel stereo systems in the home, see post number 4 for additional details.

Current issues:
  1. Has a maximum of 7.1 channels.
  2. If you have height/wides/ceiling speakers (Dolby Atmos, Auro3D, DTS:X), you may want to consider an additional DDRC-88A solution to go beyond 8 channels. (Edit: AVS members westmd and Keith Barnes have successfully implemented a dual DDRC-88A setup. westmd in post number 4,900 and Keith Barnes in post number 8,675 with the tutorial in post number 8,683.)
  3. Sampling resolution is capped at 48kHz, might be a disappointment for those with multi-channel hi-rez files. I.e. 192kHz/24bit.
  4. Have to use MiniDSP's microphone for in room acoustical measurement with Dirac. Model number UMIK-1. Post number 18,754 in miniDSP forum explains the reasons for this.
  5. No remote trigger to start or IR code for ON/OFF operation. There could be a potential issue when the unit losses power - may cause a thump to the downstream speaker. Suggested to leave the unit on all the time and this may not please some.

Trouble shooting:
  • If the front light is OFF and the switch mode power adapter is ON (i.e. showing a green light. Edit: Keith Barnes advised that not all units power adapter have a light to indicate it is powered ON in post number 8,937) most likely the internal fuse needs replacing. Specs for the fuse is 2Amp 250v 20mm x 5mm Slow Blow. Revised fuse rating by miniDSP as advised by Markus in post number 9,186.
 
See less See more
1
#2,182 ·
I would divide my dream EQ into two parts: things that should be done automatically and things that the user can do manually.

I don't know anyone who finds creative enjoyment in figuring out the centre frequency, filter width and amount of boost or cut needed to address a problem. An automated program can deal with peaks & dips for all I care, hence my comment to Jerry. However, that algorithm is never going to know your personal taste, so shaping the overall sound should be user adjustable.
That's exactly how Dirac Live claims to work. You define the target curve and DL takes care of the heavy lifting, i.e. what can and should be corrected and what not.

In Jerry's example, if the automated part hasn't fixed a "nasty dip at 200Hz", I'd investigate why it hasn't before attempting to manually boost it by changing the target curve.
That's what I just suggested in a previous post. By the way, just because there's a dip that shows up in a 1/6 octave smoothed graph, this doesn't mean there's an actual or fixable problem.
 
#2,183 ·
Not sure why it would. PEQ is minimum phase only.
You have to read between the lines :)

Assumption is that user wants to fix a dip, dip not being fixed by Dirac because it is either unfixable or because it can't fix that without adversely affecting something else. Rightly or wrongly, user wants to "fix" dip anyway. In this context what is more likely to be successful? Gaming Dirac or PEQ? I would argue PEQ is more likely to be successful on these terms. Is it the right thing to do? i.e. will it improve sq? Well that depends on the specific situation :)
 
#2,184 ·
^

FWIW I find target curve tweaking (not in dirac but a target curve is a target curve) is more relevant to the LF. I prefer to start the LF rise around 120Hz as higher tends to muddy the upper bass and I shape it to pass underneath some room driven variation. There are some other room specific tweaks to further manually smooth (possibility what is referred to as repair above? but in my case I can directly control it) the result.

In the HF, I find it is just a question of how steep the slope should be and when to stop correcting, ie v much a broad brush.
One thing that is possible, since Dirac allows us to save target curve files, is that we could share them with other users. You could upload that target curve of yours, for example, and anyone with Dirac and the 88A could try it in their own room/system. Neat or nuts?
 
#2,186 ·
Probably the latter. One could look at the excess phase group delay in REW for example to further examine if that dip should be "repaired" or not.
I would divide my dream EQ into two parts: things that should be done automatically and things that the user can do manually.

I don't know anyone who finds creative enjoyment in figuring out the centre frequency, filter width and amount of boost or cut needed to address a problem. An automated program can deal with peaks & dips for all I care, hence my comment to Jerry. However, that algorithm is never going to know your personal taste, so shaping the overall sound should be user adjustable.

In Jerry's example, if the automated part hasn't fixed a "nasty dip at 200Hz", I'd investigate why it hasn't before attempting to manually boost it by changing the target curve.
You have to read between the lines :)

Assumption is that user wants to fix a dip, dip not being fixed by Dirac because it is either unfixable or because it can't fix that without adversely affecting something else. Rightly or wrongly, user wants to "fix" dip anyway. In this context what is more likely to be successful? Gaming Dirac or PEQ? I would argue PEQ is more likely to be successful on these terms. Is it the right thing to do? i.e. will it improve sq? Well that depends on the specific situation :)
Here is the "Nasty Dip". I am trying to figure out what is causing it and whether it is fixable. Markus, have you ever provided guidance regarding how to determine if the dip is minimum phase? Any other suggestions? Do we need to take this to the REW thread?

 
#2,187 ·
Guys,

Would it be worthwhile to provide links on the 1st page for impressions and graphs by owners of the DDRC-88A? The intent is for newbies to quickly find this out on the 1st page rather than asking repeatedly within the thread.

I've seen impression posts by Gooddoc, Jerry and Keith. How about Mike and dwaleke? Jerry is has posted the before and after ETC graphs, not sure about waterfalls/spectograms and frequency responses. Are others willing provide their before and after REW measurements too?
I think links to impressions posts would be useful, Steve, but I would not stay the same about measurements posts. Perhaps other opinions will be forthcoming.
 
#2,188 ·
Here is the "Nasty Dip". I am trying to figure out what is causing it and whether it is fixable. Markus, have you ever provided guidance regarding how to determine if the dip is minimum phase? Any other suggestions? Do we need to take this to the REW thread?
Anything reflective a little over two feet away from the speakers?
 
#2,189 ·
Has the dip always been there? i.e. with Audyssey and each measurement with DL? If so, then it may be the floor bounce effect or backwall bounce. May not necessarily need to be corrected. However, if it is "new" then maybe a remeasure would be of value. Most importantly, how does it sound? If it sounds balanced and proper tone/timbre/body then maybe it's worth ignoring. But then again, most of us here are kinda ocd so that's not going to happen, lol.
 
#2,190 ·
I have mains with extension to 20 Hz and surrounds with extension to 80 Hz, how is a single target curve going to deal with that?
Bass management. Using the same target curve should not be a problem. for the mains using 20hz crossover the curve for the mains carries down to 20hz and then crosses over to the sub/subs. For speakers with crossed higher, they just follow the target curve to their crossover point and then the sub follows the rest of the way down. If using a full target curve, I don't really see why you would use a different curve for different speakers.
 
#2,191 · (Edited)
Bass management. Using the same target curve should not be a problem. for the mains using 20hz crossover the curve for the mains carries down to 20hz and then crosses over to the sub/subs. For speakers with crossed higher, they just follow the target curve to their crossover point and then the sub follows the rest of the way down. If using a full target curve, I don't really see why you would use a different curve for different speakers.
I don't think Dirac will allow a target curve that could potentially result in a massive boost to a speaker on the low end. Jerry posted some scenarios and linked target curves end at the least capable speaker. So the scenario you're talking about is not possible as far as I can tell.
 
#2,193 · (Edited)
Here is the "Nasty Dip". I am trying to figure out what is causing it and whether it is fixable.

Here's one idea given where in the frequency range the dip is:

Try putting something obstructing in the way for the floor reflex, here's a guy testing it:


Some damping material on the speaker side of the obstructions would be a good idea too.

Those were made from IKEA shelves, I think:


If it makes a difference, then you know what it is. From there to actually keep those obstructions permanently - quite another matter, it can be hard to make it look ok...
 
#2,194 ·
One thing that is possible, since Dirac allows us to save target curve files, is that we could share them with other users. You could upload that target curve of yours, for example, and anyone with Dirac and the 88A could try it in their own room/system. Neat or nuts?
My curves are quite uninteresting but I can post later anyway. I find differences in the curve above 2kHz v subtle indeed, perhaps i am more deaf than i think :) I find the end result to be dominated by the input data and the way that is processed though rather than the target curve itself.

I think it would be interesting to see the measured data along with the subjective impressions though.
 
#2,196 ·
Here is the "Nasty Dip". I am trying to figure out what is causing it and whether it is fixable. Markus, have you ever provided guidance regarding how to determine if the dip is minimum phase? Any other suggestions? Do we need to take this to the REW thread?

[...]
The plot shows that it's not just a single dip but multiple. It's very unlikely they are all minimum phase.

John wrote an excellent article about how to identify minimum phase regions:
http://www.roomeqwizard.com/wizardhelpv5/help_en-GB/html/minimumphase.html
 
#2,198 ·
You'd probably want to use a different target curve if the radiation pattern between speakers varies.
When I get time, I will measure what I have and then try different target curves. When using Audyssey, my AVR always set my LCR's to full and surrounds to 40hz. My surrounds are pretty capable. They have 10" mid-bass drivers.
 
#2,199 ·
When I get time, I will measure what I have and then try different target curves. When using Audyssey, my AVR always set my LCR's to full and surrounds to 40hz. My surrounds are pretty capable. They have 10" mid-bass drivers.
I think your answer demonstrates what is true for the topic--whether you use the same targets for all speakers or not is highly dependent on your specific configuration. Which is also another reason why Keith's suggestion of sharing target files, while interesting, probably won't work very well.
 
#2,200 ·
Here's one idea given where in the frequency range the dip is:

Try putting something obstructing in the way for the floor reflex, here's a guy testing it:

If it makes a difference, then you know what it is. From there to actually keep those obstructions permanently - quite another matter, it can be hard to make it look ok...
I have tried this, what I call the "blocking method", but I was under the impression that it is used to find reflections, not frequency response anomalies, especially at frequencies as low as 250Hz.
 
Top