TOPIC: Bi-Amping and Dirac

Hello Team,

As a starting note, I don’t intend this thread to change into a debate on the worth of Bi-Amping as those debates are already numerous. I just have a question I’d like some help with:

I currently bi-amp my 2-way speakers. If I use the SHD’s matrix function I can send each channel to 2 outputs.

Now, what happens when I run Dirac room correction? Does Dirac “see” 1 or 2 speakers on each channel? Is there a way to make it see 2 speakers per channel (one tweeter, one woofer) so that only the required frequency ranges are sent to my stereo amps?

Also, let’s say I was using balanced outputs for the stereo speakers, can I still use one of the unbalanced outputs to feed the sub? If not, that’s ok as the sub accepts high level signal from the amps. But in either config again what does Dirac see? 1 speaker per channel with extended bass or 1 speaker per channel with dedicated sub or something else?

Answers to the above very gratefully received and any advice on best way to set up while still bi-amping welcome.

Thanks in advance.

All Dirac 'sees' (hears?) will be the combined left channel output, and the combined right channel output.

However, I would not do what you're proposing and you'd lose the ability to have your subwoofer on a dedicated output, and therefore lose the ability to apply crossovers and delays to integrate this with your main speakers.

Yes that makes sense. Better to allow the correction to see the sub and speakers as discreet elements that it can manipulate separately. See I knew it made sense to ask!
Thanks

Just another thought (a.k.a question) wrt 2.1 correction:
Dirac1 set as: Channel 1 (Left) and Channel 3 (sub).
Dirac 2 set as: Channel 2 (Right) and Channel 3 (sub).
So now Dirac works it’s magic on each channel.
Firstly then, why is a second Dirac needed? If Dirac 1 is able to correct Ch1 and Ch3 individually, why could it not have done Ch2 also? Or are they not truly treated as separate channels?
Second, my speakers and sub overlap with the speakers going down to 60Hz and the Sub going up to 100Hz. Rather than me guess the best crossover point, why not leave high and low pass filters bypassed. This way Dirac can decide where to send the 70-100Hz signals based on whichever hardware handles it best. Hey, maybe the most even result is letting both speakers and sub handle those frequencies.
What do you think?

asx77 wrote:
Dirac doesn't treat Ch 1 and Ch 3 individually. It treats the combination as a single left channel.

asx77 wrote:
Dirac doesn't control the crossover between sub and main speaker. You need to set this yourself first and then run Dirac.

Dirac just sees the total left channel signal. It makes no difference if this is from a single speaker or a speaker plus a subwoofer.

Loving the forum. First one I’ve ever really used and great to bounce thoughts and ideas around. So here’s another one! I want to give Dirac the best possible chance of doing a fantastic correction. What if I set the wrong crossover point between speakers and sub? Then I’m hampering it.
So... just for identifying the best crossover point how about this:
Dirac 1, Channel 1 = Left, Channel 2 = Right. This means Dirac 1 sees Left and Right as one speaker.
Now, Dirac 2,Channel 3 = Sub.
Start room correction. Look at results. This will indicate where Dirac wants to roll the speakers and sub frequencies off. Great! Now I can set the crossover in the best place right? (Wrong?).
Next I can set things correctly, D1 to Ch1 L, Ch3 Sub and D2 to Ch2 R and Ch3 Sub. Crossover set and away we go with some more magic.

asx77 wrote:
What you proposed would not give you any useful information.

Dirac is simply not helpful for setting up the main speakers to subwoofer crossover. It is geared up to work on two full-range channels. Left and right. Nothing more. If you are looking for Dirac to help you with sub and main speaker integration then I'm afraid you are going to be disappointed. All t can realistically do in this regard is that if you get a peak in the combined response of of sub + mains then Dirac will give you the power to flatten this peak off. Small dips could also be reduced but big dips caused by cancellation of sub and main speaker signals in a poorly set up crossover region could not.

Possibly it's worth me saying that the Dirac software is exactly the same on miniDSPs with 2 inputs but 4 outputs (like the SHD or DDRC-24) as it is on simpler components with just 2 independent outputs.

Perhaps I should explain why what you proposed isn't helpful...

A sub-only measurement would just show you the maximum frequency range capability of the subwoofer but this will almost certainly go to much higher frequencies than you'd ever want to use on it.

Measuring the two main speakers playing together will give a result that is harder to interpret than measuring each individually, which is something you'd need to do manually with probably REW when you are manually deciding on how to set up the crossover.

Hi ok, appreciate your time giving feedback.

Note what you say about other Dirac enabled products but I’m looking to replace my streamer too potentially.

I understand your comments wrt sub integration limitations and I guess even an AV processor with a dedicated sub out still asks you where to set the crossover point.

Could I start correction off with just:
D1 Ch1 L
D2 Ch2 R

I can then see where the mains start struggling with low frequencies and bring the sub into the mix (D1&2,Ch3) with crossover set 10Hz higher than that.

I choose it but it’s based on measured speaker performance.

asx77 wrote:
Yes you could but I'd recommend that you learn how to use Room EQ Wizard (REW) and use that for making measurements to optimise the crossover rather than using Dirac for making these measurements.

Yes, I would confirm what Subsonic has just said. When I started on this journey, I really was struggling. The secret is to get through each stage successfully and not to jump ahead too soon, or you'll just have to redo things (over and over again like I did), because doubt creeps in that you may not have done it right previously.

REW is the first stage. And like all new pieces of complicated (really!!...) software, it's not obvious. Bear in mind there's lots in there that you probably don't need. Connect your mic up and do a sweep on your mains to see their frequency response. Then do the same on your subs. If there's a crossover on your subs, set it to max. And then move to stage 2. Armed with these 2 frequency responses you can put meaningful (and tested) numbers into the Xover section of the add-in - on the Output page. Get the Routing page right and you're almost there.

Edit: Sorry, I say subs in the above - because that's what I have. But with your 2 way speakers, it's exactly the same process.

Then, and only then, start playing with Dirac. And if you thought you were nearly there...... There's hours of tinkering to be had. But, honestly, you're through the worst.

Gents,

Thanks for the advice. I have used REW so should be able to run a sweep. What if however the crossover point is muddled with some modes/anti-modes? That's the only reason I was thinking about running some Dirac correction on my mains and seeing when Dirac rolled things off.

Alternatively i could just simplify things and set the crossover at 80Hz. My Sub (Rel Strata V) filter goes up to 99Hz, although I suppose to get a roll off it will effect frequencies above that. My Speakers according to the spec go down to 60Hz (+/-3dB). It just doesn't feel right. Like I'm limiting Dirac's potential!

Run the sweep with the mic 2 metres away from the speaker. Bring the speaker away from any walls. Job done.

I saw someone on YouTube measuring speakers outdoors and thought “nah! Too much hassle “ but I like your suggestion, thanks!

EdHowarth wrote:
I'm afraid that really won't give the same effect as measuring outside, if that's what you had in mind. I've made time-gated measurements indoors before, with speakers on a chair in the middle of the room but the boundary reflections limit how low frequency meaningful data can be achieved for. The bass region we're talking here will be too low for this approach to be viable, and I don't think excluding the room effects is desirable anyway really.