TOPIC:

Hello,
I know and understand why some people tend to say that bass management downstream is not optimal for PEQ.
My avr (Rotel 1562) permits to apply manually choosen filters PEQ for each channels including sub. I have 5 filters available between 20hz and 190hz for each channels. I can use REW and Umik-1 to design EQ for the 5 filters band available in Rotel.
Is it theoritically correct to do the following: to design 5 filters for low FR (20-190htz) to each channel (set to large) and for the sub. After that set the xo in the AVR at 80htz, set speakers to small and run Dirac via NanoAvr-DL. Then, in dirac, to use the left side curtain to cut the correction for each channels unto the xo point and keep the ful optimized target curve choosen by Dirac for the sub channel?

No offence but that doesn't make a lot of sense. What problem are you trying to solve?

None taken. The problem I tried to solve is what Kal Rubinson and others in others Forum say about the faulty chain: source-Dirac-Bass management-output or the called bass management downstream instead of beeing upstream .

But are you experiencing any issue due to this alleged "faulty chain"?

john.reekie wrote:

But are you experiencing any issue due to this alleged "faulty chain"?

Good question.
That is the problem between the theory and the practice and how to see (to ear) the difference for an amateur like me.
I read Forums to learn as much as I can but I am no expert. I understand enough the physics and the theory ofacoustic to know that I do not know enough!

So to answer to your question, as far as I know and as far as I can judge the sound of my system: I do not think so.

But, it was the same before I changed the placement of my fronts speakers and the sound improved. And the same when I began to treat my room with DIY bass traps and broadband absorbers : wow what a difference it makes even they are not optimal nor the most correct in theory. So I was telling me maybe it is the case for this allegibly "faulty chain". .

Good answer Let me get back to you again tomorrow.

There is a long thread at avsforum regarding Arcam with Dirac and BM downstreams, plus other bugs.
A couple of days ago i did some testing with Dirac on Pc while i did BM in NanoAvr and to set the curtain at the xo did not turn out well.
I turned of the Peq:s and only left BM, xo and timing On in Nano, my receiver only works as volume and all the ch ar set to large plus Lfe.

So without Dirac
i.imgur.com/MZPOJPF.png

With Dirac and here i moved the left curtain to 100hz.
i.imgur.com/e7ghBfu.png

And here you have Dirac working fullrange.
i.imgur.com/ZnTZfxr.png

I was a bit suprised at the results and i am gonna buy a second Nano with Dirac

smygolf wrote:

There is a long thread at avsforum regarding Arcam with Dirac and BM downstreams, plus other bugs.
A couple of days ago i did some testing with Dirac on Pc while i did BM in NanoAvr and to set the curtain at the xo did not turn out well.
I turned of the Peq:s and only left BM, xo and timing On in Nano, my receiver only works as volume and all the ch ar set to large plus Lfe.
(...)
I was a bit suprised at the results and i am gonna buy a second Nano with Dirac

I am not sure to understand correctly. Your chain is PC- NanoAvr-DL ? - receiver where is BM - - speakers. Then you set into your receiver all speakers to large no xo and LFE only through sub channel to get the result in your third graph? In Dirac, then, can you tailor a target curve?
I am eager to see what John Reekie would have to say about that?

The signal chain is like this.

Pc with Dirac ->
NanoAvr (only doing BM, delay and xo at 100hz) ->
Receiver (all ch set as Large plus Lfe to the subs, so no BM )
And the result is the third graph

smygolf wrote:

The signal chain is like this.

Pc with Dirac ->
NanoAvr (only doing BM, delay and xo at 100hz) ->
Receiver (all ch set as Large plus Lfe to the subs, so no BM )
And the result is the third graph

Thank you. I am quite surprise too!
How does it sound?

I only did one measurement in sweatspot to get a true curve so i could compare it without dirac.
If you do all the nine measurements you aint gonna get a "perfect" curve.
Also a chair has a better curve then couch and its up to you to decide how you will do it.
I always choose chair for music and couch for movies.

and my experiance with Dirac is Good

Sorry, I was supposed to reply earlier. In the case of absorption, there is pretty much incontestible theory as to why this makes a difference. In the case of speaker placement, YMMV and you haven't provided detail but there's solid theory as to why it could make a difference. In the case of BM before/after Dirac... not so much. Honestly, there is only one corner case (*) that I'm aware of so far, but in exchange for eliminating that corner case (i.e. which most likely won't apply to your situation) you lose a number of practical advantages.

HTH

(*) - en.wikipedia.org/wiki/Corner_case

john.reekie wrote:

Sorry, I was supposed to reply earlier. (...)In the case of BM before/after Dirac... not so much. Honestly, there is only one corner case (*) that I'm aware of so far, but in exchange for eliminating that corner case (i.e. which most likely won't apply to your situation) you lose a number of practical advantages.
HTH
(*) - en.wikipedia.org/wiki/Corner_case

Thank to come back!
Yes it helps.
My intuition (ouch!) tells me that in small room like mine (40 cubic meters) the differences in the measurements induced by the different distances signal from the sub and from the fronts are insignificant.

To John and/or Devteam,
I saw that Arcam recently updated the firmware of their 550 and 850 receivers to send the sweep signal during Dirac Live measurements procedure through a speaker and the sub together.
Thank God I did not bought those receivers, but I own a nanoAvr-DL.
Am I right to think that it is already the way that nanoAvr-DL works ? For all channels (front left for instance) the left side (from 20hz to 100hz) average response curve "before" in Dirac is always exactly the same that the sub channel curve. That told me that the signal begins at the sub to gradually going through the left speaker as the fr of the sweep rises; isn't it ? (I never put my hand on the sub driver during a measurement to check it).

It should alway go thru the subs if you use bass management (speakers as small)
So it dosent matter if BM is first or last the subs are always getting a signal and LFE signal goes seperate.
So whats up with Arcam im not sure!!

If you dont use BM (speakers as Large) the subs only gonna work with the LFE.
Like this graph, speakers as small and without BM (you cant do this in a receiver but you can in Nanoavr HD)
i.imgur.com/mIbHigz.png


From AVS forum....
In Arcam's implementation DL does not know that it measures sat & sub at the same time. It thinks it would see a single speaker. So what happens is that it calculates delays based on each measurement. This is done by looking at the impulse peak. The impulse peak is dominated by the speaker with higher frequency content, i.e. the sat and not the sub. As a result the sub in each sat measurement does not get the proper delay but is delayed based on the sats distance. Things get worse when DL now measures through the LFE. This will find the proper delay of the sub but now the sub gets delayed and in the end the earlier measured response for the sub/sat combination is no longer valid. In other words DL will base its correction on a response that is no longer valid.

DL is a single speaker solution. You have to provide it with single speaker responses. Arcam's implementation doesn't (unless all speakers are set to large).

All these problems go away when DL is implemented downstream of bass management.