TOPIC:

Jim wrote:

markus wrote:

sheraz,

Try to understand what bass management is and how it works and it should become very obvious why the only correct location for EQ is AFTER bass management

I think that is exactly the point sheraz is trying to make. :o) But is this always true? There is a fair amount of magnitude & phase correction that can be done upstream from bass management. RePhase has among other features a 'global' curve that would be an example of this, if I understand it correctly. Also, nanoAVR DL's Dirac is made aware of the speaker config during setup. If it's smart enough, it should still be able to blend things properly even though it can't address & test each speaker separately? How is this different from its inability to test mids separately from tweets, for example?

-- Jim

DL can't be made aware of bass management. For example, you can't tell Dirac Live that the signal for L is going through L and the subwoofer. Even if you could it wouldn't help Dirac Live to handle things properly. Dirac Live follows a single speaker optimization strategy. It needs to be able to "talk" to individual speakers to work properly.

markus wrote:

Jim wrote:

markus wrote:

sheraz,

Try to understand what bass management is and how it works and it should become very obvious why the only correct location for EQ is AFTER bass management

I think that is exactly the point sheraz is trying to make. :o) But is this always true? There is a fair amount of magnitude & phase correction that can be done upstream from bass management. RePhase has among other features a 'global' curve that would be an example of this, if I understand it correctly. Also, nanoAVR DL's Dirac is made aware of the speaker config during setup. If it's smart enough, it should still be able to blend things properly even though it can't address & test each speaker separately? How is this different from its inability to test mids separately from tweets, for example?

-- Jim

DL can't be made aware of bass management. For example, you can't tell Dirac Live that the signal for L is going through L and the subwoofer. Even if you could it wouldn't help Dirac Live to handle things properly. Dirac Live follows a single speaker optimization strategy. It needs to be able to "talk" to individual speakers to work properly.

This last paragraph Markus explains the issue in simplest term. I still can't get my head around completely how having bass before/after works but knowing that Dirac takes a signal and assumes that its is for one speaker clearly shows that it should be send to only one speaker not tow (sub is 2nd in case where bass management is after).

Dear All,

An interesting topic. Sheraz been already asking this question on few locations (here/avs/tech support line) so obviously a source of concern.. Let's see how we can answer this "hot" topic.. Let's start with a post we already made on AVS a while back when this topic came out. Here is the gist:

"
OPTION1: Source - > Dirac Live -> Bass management (BM) - > AVR (nanoAVR DL structure)
OPTION2: Source - > Bass management(BM) -> Dirac Live - > AVR (proposed as the ultimate solution by some of you)

Knowing Bass Management is just a glorified Matrix + Crossover network, OPTION 1 or OPTION2 is similar to EQ in a multiway active system . Say a 3way stereo system where you have EQ at the input side(for L&R) but also have EQ at the output side (for each specific driver: Low/Mid/High). Would you agree? My point being, this EQ argument isn't new! it's been there for many years. There will be 2 sides of the argument, both with very valid arguments, both with Pro's & Con's… The perfect system (to my mind) is actually a combination of both worlds, this way everybody is happy!

OPTION1: Source - > Dirac Live -> Bass management - > AVR (nanoAVR DL as of now)
This is similar to doing EQ at the input section of an active system (i.e. L&R overall EQ). When we run 8 measurements at 9 positions = 72 measurements, we're building a "model" of the system. In other words, we play a sweep on Front Left (as your bluray player will when content is played), it will indeed start sweeps @ Low Freq inside the SUB (routed thanks to BM) and then move to the Front left pass cut off frequency.
PRO's: That's the true response of the system (i.e. extended bass response inc potential crossover effect). Dirac will not only fix the speaker but also the transition to the SUB. It's easiest way to fix your content (i.e. 7.1 or Stereo) no matter the configuration you have.
CON's: The impulse responses may be hard to work with however as the sound is arriving at different time for different frequencies in different positions (the sound is not coming from roughly one point in the room) and if the delay and level differences between a sub and a "main" speaker is very large, it may be difficult to handle within a "combined" Dirac Live filter. Now, if you run Dirac Live trial on your PC/HTPC, you're most likely running under that configuration (AVR doing the BM). So have a trial at it. It's free to test and you'll see the results..

OPTION2: Source -> Bass management -> Dirac Live @ each output- > AVR
This is similar to doing EQ at the output section of an active system (i.e. each speaker/ each driver / so speaker EQ). You'll build the best response for each speaker. There are 2 possible locations where we could run the sweep signal for Dirac measurements in this configuration:
a) Sweep before the BM module to give the true color of where the signal is routed to = > i.e. sweep will go through Sub + Left speaker (thanks to BM) PRO's you do see the true colors of the entire system. CON's: you're now planning to apply correction to only one output since Dirac is "post BM" (i.e. DSP block at the end of the chain). It will not solve your problem in your room. (measurement and filters are not in sync of what's to fix). This solution isn't right
b) Sweep after the BM module to give the true color of how each speaker sound =>i.e. Sweep will only go through the speaker.
PRO you will do GREAT at solving your speaker. Produces the best possible speakers for the bass management to work with.
CON's. what about all the filters you may have in your BM section? They are not taken into account. i.e. once you do have BM enabled, the sound will change so your measurement isn't a representation of the system. Possible your bass management crossovers will create a dip/bump at the LFE transition. Dirac will unfortunately not be able to fix that. (post BM, can't see it).
The way we see it, Option2 will work very well if you know really well what you're doing upstream (i.e. at the BM level). Call it an "advanced mode". The flexibility of our BM module as shown by Kal's extensive review on Stereophile is a testimony of what you can do once in the right hand. With this said, we've got to be realistic that it's not as plug&play. REW, filter and acoustic interpretation of results will take time. That was the bottleneck which we evaluated at length, and with our objective of putting Dirac Live in the hands of the largest crowd (Affordable/Simple/easy to use), we tried to make it an accessory to an existing system that is most important plug&play. Don't get me wrong, Option2 can be a great fit too for complex setup and I'm sure that a lot of tinkerers with complex system would love that. I just hope that it makes sense to you that it's maybe for a different crowd. Maybe a different product. Different philosophy."

---

So down to what where we are few months later:
a) There are a lot more nanoaVR in the field so maybe some of your could pitch in? e.g. with some posts of your graphs.
b) There are some DDRC-88A that could easily be put before and after the AVR in a chain to see if indeed there is a huge issue? Or maybe just using the Dirac live trial on their PC (i.e. prevents any rewiring). Then just have their PC feed the AVR.. just a thought..
c) We still need to post some meaningful tests but here is an example on our test system of having Dirac before the bass management. Works fine,everybody is still alive in the room .. Would need to look a bit more extensively in the measurements though.. that was a long while ago.
dl.dropboxusercontent.com/u/84624505/REW...%2C%20no%20Dirac.png
dl.dropboxusercontent.com/u/84624505/REW...C%20WITH%20Dirac.png

@ Sheraz,

As mentioned, there is a free trial of Dirac. We'd recommend maybe that you have a trial at it. It would be a first learning experience and will be 100% exactly the same configuration as running Dirac on the nanoAVR DL.

DevTeam

devteam wrote:

OPTION1: Source - > Dirac Live -> Bass management (BM) - > AVR (nanoAVR DL structure)
OPTION2: Source - > Bass management(BM) -> Dirac Live - > AVR (proposed as the ultimate solution by some of you)

[...]

OPTION1: Source - > Dirac Live -> Bass management - > AVR (nanoAVR DL as of now)
This is similar to doing EQ at the input section of an active system (i.e. L&R overall EQ). When we run 8 measurements at 9 positions = 72 measurements, we're building a "model" of the system. In other words, we play a sweep on Front Left (as your bluray player will when content is played), it will indeed start sweeps @ Low Freq inside the SUB (routed thanks to BM) and then move to the Front left pass cut off frequency.
PRO's: That's the true response of the system (i.e. extended bass response inc potential crossover effect). Dirac will not only fix the speaker but also the transition to the SUB. It's easiest way to fix your content (i.e. 7.1 or Stereo) no matter the configuration you have.
CON's: The impulse responses may be hard to work with however as the sound is arriving at different time for different frequencies in different positions (the sound is not coming from roughly one point in the room) and if the delay and level differences between a sub and a "main" speaker is very large, it may be difficult to handle within a "combined" Dirac Live filter. Now, if you run Dirac Live trial on your PC/HTPC, you're most likely running under that configuration (AVR doing the BM). So have a trial at it. It's free to test and you'll see the results..

Option 1 is NOT similar to an active system because an active system has been optimized under free field conditions. The level and delay of each driver has been optimized BEFORE room EQ is applied. Home theater is different. Level and delay of satellites and subs is NOT optimized.
Furthermore, in an active system each transmission channel equals one specific speaker. In a bass managed scenario each transmission channel goes to multiple speakers, i.e. satellite AND sub.
Option 1 is simply wrong if bass management is involved.

devteam wrote:

OPTION2: Source -> Bass management -> Dirac Live @ each output- > AVR
This is similar to doing EQ at the output section of an active system (i.e. each speaker/ each driver / so speaker EQ). You'll build the best response for each speaker. There are 2 possible locations where we could run the sweep signal for Dirac measurements in this configuration:
a) Sweep before the BM module to give the true color of where the signal is routed to = > i.e. sweep will go through Sub + Left speaker (thanks to BM) PRO's you do see the true colors of the entire system. CON's: you're now planning to apply correction to only one output since Dirac is "post BM" (i.e. DSP block at the end of the chain). It will not solve your problem in your room. (measurement and filters are not in sync of what's to fix). This solution isn't right
b) Sweep after the BM module to give the true color of how each speaker sound =>i.e. Sweep will only go through the speaker.
PRO you will do GREAT at solving your speaker. Produces the best possible speakers for the bass management to work with.
CON's. what about all the filters you may have in your BM section? They are not taken into account. i.e. once you do have BM enabled, the sound will change so your measurement isn't a representation of the system. Possible your bass management crossovers will create a dip/bump at the LFE transition. Dirac will unfortunately not be able to fix that. (post BM, can't see it).
The way we see it, Option2 will work very well if you know really well what you're doing upstream (i.e. at the BM level). Call it an "advanced mode". The flexibility of our BM module as shown by Kal's extensive review on Stereophile is a testimony of what you can do once in the right hand. With this said, we've got to be realistic that it's not as plug&play. REW, filter and acoustic interpretation of results will take time. That was the bottleneck which we evaluated at length, and with our objective of putting Dirac Live in the hands of the largest crowd (Affordable/Simple/easy to use), we tried to make it an accessory to an existing system that is most important plug&play. Don't get me wrong, Option2 can be a great fit too for complex setup and I'm sure that a lot of tinkerers with complex system would love that. I just hope that it makes sense to you that it's maybe for a different crowd. Maybe a different product. Different philosophy."

Option 2 isn't "advanced", it's just the correct way of doing it. Dirac Live not being able to take bass management into account is a limitation of Dirac Live. Audyssey for example has the same limitation. Room correction and bass management should ideally be one integrated process.
Another limitation of Dirac Live and any other single speaker optimization approach is its inability to optimize combined speaker responses like L+R+SW and/or L+R+C+SW, etc.

markus wrote:

Home theater is different. Level and delay of satellites and subs is NOT optimized.

I think there would be an assumption that when you set up the bass management in your AVR, you also set the levels and delays/speaker distances.

While it's true that bass management isn't exactly the same as an active speaker, I can only think of one scenario in which this (Option 1) would be "wrong". Would you be able to provide specific examples of what errors will occur and under what circumstances?

john.reekie wrote:

markus wrote:

Home theater is different. Level and delay of satellites and subs is NOT optimized.

I think there would be an assumption that when you set up the bass management in your AVR, you also set the levels and delays/speaker distances.

I doubt that most people set level and delay correctly in their AVR. It requires measurement equipment they probably don't have. Owning such equipment somewhat defeats the purpose of buying an automated room correction system like Dirac Live They might have an AVR with some form of automated setup routine which might or might not get it right though.
The AVRs I know have rather coarse delay steps. Dirac Live seems to be capable of finer delay resolution, i.e. it will time align speakers even if speaker distance is already set in the AVR and you'll end up with potential comb filtering caused by having bass management downstreams.

john.reekie wrote:

While it's true that bass management isn't exactly the same as an active speaker, I can only think of one scenario in which this (Option 1) would be "wrong". Would you be able to provide specific examples of what errors will occur and under what circumstances?

I can only think of one scenario where it would be right Having all transmission channels match a dedicated speaker (all speakers set to large):
L channel > L speaker
R channel > R speaker
C channel > C speaker
LFE channel > subwoofer
LS channel > LS speaker
RS channel > RS speaker
LSB channel > LSB speaker
RSB channel > RSB speaker
This would effectively remove bass management from the equation - kind of counterproductive though.

Lots of "ifs". Having bass management upstreams makes all those complex considerations go away.

Thanks Markus. Could you be more specific on what you mean by "potential comb filtering"? If Dirac realigns the time delay on each speaker then I'm not seeing how this is an issue (any more than it is anyway with two separated sources). Is the potential issue that the two speakers had the wrong delay set in the AVR and when Dirac realigned the delay on the speaker channels it created an effective delay between the subwoofer signals sourced from the speaker channels? In that case you'd need to be off by say 3ms for a quarter-wavelength difference at 80 Hz. Is that what you mean?

^
Example: Send a coherent signal through L and R with one channel delayed. With bass management engaged you will measure the results of comb filtering at the subwoofer out.

devteam wrote:

Dear All,

@ Sheraz,

As mentioned, there is a free trial of Dirac. We'd recommend maybe that you have a trial at it. It would be a first learning experience and will be 100% exactly the same configuration as running Dirac on the nanoAVR DL.

DevTeam

The issue with evaluating the free version on pc is no way of testing it. Say we download Dirac on PC, configure it, apply filters on HDMI output. Now how do I test it using REW? Generally REW is attached to AVR and the sound is send thru avr. If avr has all filters, the sounds gets filtered out when recording in Mic with REW. With pc having the filter, there is no way to test the filters in REW.

With nanoAVR DL in chain, one can attach HDMI to 2nd input in nanoAVR and send signal thru it. This can be tested. So in short, free version is good to download, install and play something through PC. Its not good enough to get some testing done on it (unless there is a way you can share with us).

^
Dirac Live acts as a virtual sound card on the PC, so fingers crossed you should be able to send test signals from REW through it.

markus wrote:

^
Dirac Live acts as a virtual sound card on the PC, so fingers crossed you should be able to send test signals from REW through it.

How?
Configuration:
I will install Dirac on my laptop.
Same laptop is running REW.
Laptop is connect to avr using HDMI.

Now to your point, since the sound source is ultimately my PC (REW), the sound will go thru the virtual sound card where the filters are applied and then to AVR. So effectively, the filters will be applied. Is that correct?

^
Yes. If you want to measure without DL then switch them off.

Will try this hopefully tonight. Now Markus, the issue you have been pointing out about bass management, is there a way we can confirm it with REW reading?
I can take before/after Dirac reading of L/C/R channels. Will that show the issue? If not, can you please provide the steps and I'll happily perform them.
thx.

^
What do you want to measure?

markus wrote:

^
What do you want to measure?

Want to see if BM having before is in fact causing issue. Any way to capture that through taking REW Readings?