I took 3 measurements at various positions across my listening couch after calibrating the umik-1 using a 10 Hz to 22kHz sweep, both speakers and averaged them:
Think I might back off the active woofers a shade, but pleasantly surprised.
Here’s 10 Hz to 7.5kHz:
Those graphs would be better if you kept the Y scale the same in both images. It’s hard to compare them, but perhaps that’s not the point?
Same graph, just zoomed in. No EQ applied yet.
These plots are so good that you can limit correction to frequencies below 250 Hz. Above that the variations are small enough (within ±5 dB or so) to be left untouched.
I’m no expert (a keen beginner, still learning) but isn’t the desired output curve one that’s sloping downwards - B&K style? In that case wouldn’t a full range EQ be better?
As I understand it the problem with full spectrum EQ is that as the frequency rises, the response of the room becomes more local. So correcting a 5 dB hole in one location can create a 5 dB peak mere centimetres away. Averaging measurements can show up persistent features, but correcting for them can turn into a game of whack a mole. Lower frequencies have a less localised response (leaving aside nodes) and so are more amenable to global EQ.
I’d imagine that full spectrum EQ for headphones could work very well.
I’m going to start by adjusting my active woofers. I have them boosted a bit because I generally listen at lower volumes than the test (80 dB) and the bass response falls away a bit as volume decreases. I’ll make some changes and also try some tests at listening volume.
Before measuring I was mainly concerned about a hole in the mid-bass, which I can see at 60-100 Hz. I hope to straighten that out significantly by changing the active woofers and then applying some EQ.
@andybob These are impressive curves. Equalization or no. What speakers do you use? I have always been convinced that speakers, and their interaction with the room, are vastly more important than any electronics. Imagine upgrading the DAC to a femto-clock and seeing 5 or 10 dB response changes.
They’re Australian speakers made by Mike Kontor, Noteperfect Maestro. They use Morel drivers and the cross-over between mids and tweeters is really gentle and indistinguishable by the listener. I’ve substituted in some 8AWG North Creek inductors which hang off the bottom of the crossover box like the proverbial.
Edit: Here’s the said inductors installed under the crossover, with the speaker cables soldered on to the inside binding posts. The inductors are the only component in the midrange crossover.
And here’s the system as it is looking atm:
The room it is facing into is L shaped, so there is no reflection from the left, the first reflection off the right wall has a long stretched canvas artwork on it, the carpet is on the floor and there is a blind behind the listening couch which is about 6 metres from the speakers.
I have an L-shaped room and always figured that it was killing my sound having reflections on the right but not the left (as in your case). Do you notice any significant impact?
Can’t really compare without building a wall, and even my long suffering daughters would probably draw the line at that !
But I understand that the fewer reflections the better and asymmetrical reflections are better than symmetrical ones. So I think an L shape is an advantage. One interesting aspect is that at the far end of the L the bass goes to die in my kitchen pantry. If I stand in the pantry cupboard with the door open there’s a lot of bass floating around (If I get caught doing this I have to pretend to be cooking something).
LOL:joy:
Hahah nice. I’ve always read in the home theater mags (SGHT, etc) that if you have an L-shaped room, you might as well just give up trying to do anything to it. I think you’re the first to endorse the opposite 
-mike
It may well be a disadvantage, but given nothing can be done about it I propose we band together and pity those poor fools who don’t have access to an L shaped listening room. How do they find the strength to carry on ?
My living room is 15,5 x 5,5 meters square.
Feeling left out and lonely.
Reaches for headphones…
No longer left out and lonely: after careful deliberation, I got myself a MiniDSP DDRC-22D with Dirac Live. The DDRC-22D is limited to 96kHz, but then again – so are my speakers (Meridian DSP5200). Time to take on the elephant that is the room.
First measurements are in, showing a nice little bass-hole where I expected it to be (60-110Hz) as well as a slight bump at 800-2000hz. For now, I followed the target curve with a gentle roll off after 10kHz, following the measurements of the speakers.
Results are subtle but nothing short of revelatory: a nicely cleaned up bass department and a generally smooth, solid and ‘in control’ soundstage. It’s really nice to finally hear every trick Gary Peacock has up his sleeve while jamming it out with Keith Jarrett and Jack DeJohnette. 
Since my setup is the family room with a lively toddler running about, there is not much I can (or want to) optimise in placement or acoustic treatment, but this fills in the gaps quite nicely. I generally shy away from tweaks and cables and tend to rely on what I hear (I replaced a nice, expensive Meridian streamer with a Raspberry Pi since I couldn’t hear a meaningful difference). This though, is the real deal.
And since it works independently (the Dirac filters are stored in the hardware), it works with Roon perfectly.
I’m receiving a DDR-88A today, looking forward to a similar experience to you and a nice gentle intro into target curve editing.
I toyed with the idea of just running Dirac on my PC an adding the filters to HQP but felt an offboard solution would be more flexible.
Does anyone knows how to load the Dirac generated filters into HQP? I have both and would like to give it a try.
Does anyone knows how to integrate the filters generated by Dirac into HQPlayer. I would love to give it a try.
You can’t… But you can use HQPlayer to upsample to highest rate supported by Dirac and select Dirac as output device.
Thank You Jussi, that’s what I am doing right now with the MiniDSP hardware after HQPlayer, but thought that could be a more integrated way.
Best,
