Dirac Live with Bass Control

lexicon

Well-known Member
It would be interesting to see if measured impulse response of your two sub configurations corroborates the subjective timing issue.
Here is the MDAT for the stacked sub measurements. Not sure where the 12hz hump has appeared from unless its been caused by physically placing one sub on the other. Probably shouldn't complain though at 12hz!

Looking at the REW room simulator I think that I should probably have tried inverting the sub where it was placed to the left rear of my sofa so I'll have to try that in due course.

I have included measurements of the centre and the combined subs from the centre channel as the centre is on a higher crossover lest you want to look at the timing of that.

Let me know what you conclude!
 

Attachments

  • sub1+2 Stacked May 9.zip
    8.2 MB · Views: 60

mb3195

Distinguished Member
Here is the MDAT for the stacked sub measurements. Not sure where the 12hz hump has appeared from unless its been caused by physically placing one sub on the other. Probably shouldn't complain though at 12hz!

Looking at the REW room simulator I think that I should probably have tried inverting the sub where it was placed to the left rear of my sofa so I'll have to try that in due course.

I have included measurements of the centre and the combined subs from the centre channel as the centre is on a higher crossover lest you want to look at the timing of that.

Let me know what you conclude!

Hi Jon,

Are you sure you cant get a flatter response with the L&R running them at a higher crossover? If you scale the graph more accurately you can see swings of around 10db between 60hz & 80hz which isn't great really. Whereas your sub response with the higher crossover look significantly better bar the null at 80hz.

The hump down at 12hz is surprising as well, although you wont notice this in real life.

2 subs located separately at higher crossovers should be able to get you better results IMO.

Your centre speaker also looks down on gain against your L&R, quite significantly across the whole frequency range, but very much so below 200hz - around 20db which i'd imagine would be really noticeable on certain scenes. Have you got anything in your room that is causing a null for your centre? I get one myself but nowhere at this sort of level.

One final note, are the two subs the same? If so, they are not gain matched correctly, sub1 is peaking at 87.1db and sub2 is at 92.2db. I'm assuming these are individual measurements of both subs with the peak of 95.8db when the subs are combined at the same time? Being co-located probably wont make a big difference, but something you should be aware of when running multiple subs in separate locations.
 
Last edited:

lexicon

Well-known Member
Hi Jon,

Are you sure you cant get a flatter response with the L&R running them at a higher crossover? If you scale the graph more accurately you can see swings of around 10db between 60hz & 80hz which isn't great really. Whereas your sub response with the higher crossover look significantly better bar the null at 80hz.

The hump down at 12hz is surprising as well, although you wont notice this in real life.

2 subs located separately at higher crossovers should be able to get you better results IMO.

Your centre speaker also looks down on gain against your L&R, quite significantly across the whole frequency range, but very much so below 200hz - around 20db which i'd imagine would be really noticeable on certain scenes. Have you got anything in your room that is causing a null for your centre? I get one myself but nowhere at this sort of level.

One final note, are the two subs the same? If so, they are not gain matched correctly, sub1 is peaking at 87.1db and sub2 is at 92.2db. I'm assuming these are individual measurements of both subs with the peak of 95.8db when the subs are combined at the same time? Being co-located probably wont make a big difference, but something you should be aware of when running multiple subs in separate locations.

Thanks Mark

I’m still experimenting with curves and Xovers so maybe.......early days really. You have to run Bass Control on every possible combination and then assess them all on demo material of course which can cause a few complaints 😤 . I have to wait for shopping trips.......

The centre (and surrounds) is crossed over at 100hz and mounted above screen angled down so quite close to ceiling - it certainly doesn’t sound low (if anything the opposite) was level measured the other day using an Atmos test disc so I don’t get that!

The subs are both identical and are set at default settings - I guess I could still level match them by measurement though. The top one would be more on a level with the mic so that may make a difference?

The additional sub gain by stacking is evident though and makes a big difference in practice.

So much to try - will keep me going through lockdown and beyond 😁 I’m even swapping amps about to find the best combo as well!
 

Ringnut

Distinguished Member
Also, I suggest you have a word with @Conrad. re time-aligning your subs with your mains, they look way off at the moment.
 

lexicon

Well-known Member
Also, I suggest you have a word with @Conrad. re time-aligning your subs with your mains, they look way off at the moment.
What are you looking at to determine that? Dirac Bass Control should be looking after that unless stacking the subs has altered the timing.
 

lexicon

Well-known Member
Tell you what though one’s timing always seems to improve after a few glasses of Tescos’s Campaneo Old Vines Garnacha 😂🤣😂
 

Ringnut

Distinguished Member
What are you looking at to determine that? Dirac Bass Control should be looking after that unless stacking the subs has altered the timing.

I'm looking at the boxes on the left. You have your timing reference set on your left channel (1). Compare the measured delay between your right channel and the centre channel as compared to your subs.
 

lexicon

Well-known Member
I'm looking at the boxes on the left. You have your timing reference set on your left channel (1). Compare the measured delay between your right channel and the centre channel as compared to your subs.

I’ll rerun Dirac + Bass Control and remeasure again to verify as a starting point when I get a chance. Will swap subs around to see if the apparent gain difference follows sub at the same time.
 

lexicon

Well-known Member
@mb3195

Morning Mark
I have just carried out sound tests on the two subs from the original mic (UMIK-1) position using the REW SPL Tool and an Atmos test signal. I checked these several times.

Top Sub : Max 77.7 Min 75.4 Average 76.55
Bottom Sub Max 78.5 Min 76.2 Average 77.35

That's an average variance of under 1db which is probably just positional.

That's the method I would normally use to check levels.

Now it may be that when REW does it's sweeps it calculates levels another way but at the moment I cannot account for the apparent difference in levels on those sub MDAT sweeps.
 

mb3195

Distinguished Member
@mb3195

Morning Mark
I have just carried out sound tests on the two subs from the original mic (UMIK-1) position using the REW SPL Tool and an Atmos test signal. I checked these several times.

Top Sub : Max 77.7 Min 75.4 Average 76.55
Bottom Sub Max 78.5 Min 76.2 Average 77.35

That's an average variance of under 1db which is probably just positional.

That's the method I would normally use to check levels.

Now it may be that when REW does it's sweeps it calculates levels another way but at the moment I cannot account for the apparent difference in levels on those sub MDAT sweeps.

That’s really strange.

What I think you should try just to be sure though is this.

Switch off sub 2.
Run a tone generator through REW with a Hz level that you know you don’t have any nulls, say 40hz.
Check you level.
Switch off sub 1, switch on sub 2 and repeat.

It might be worth double checking as I personally have stopped using Dolby test tones as I don’t think they are very accurate.
 

lexicon

Well-known Member
Ok - cheers, will add that to the to do list.

I’ve just rerun a nice clean dirac measurement and thought I’d have a play with Bass Control.

As regards Xovers I’m not sure the usual “set everything to say “80 or 100hz” applies or at least that’s not what I think Dirac are trying to achieve here.

From what I can ascertain, you run “BC calculate” and it is supposed to end up with the most Optimal Xover for each speaker group. I’m not sure that‘s working entirely correctly at the moment but I assume that’s the idea.

If you want to relocate the Xover on any group I believe you are supposed to find a point where both the sub and the speaker group you are looking at have good combined energy and choose that. Of course that could be different for each speaker group.

I‘ll see if I can improve the centre response (as that one looked the worst) by moving it‘s Xover around and then see if moving the Mains Xover from 40 to 80hz has any effect on the centre corrected trace. Every time you change a Xover level it takes a good 5 mins to recalculate.

Another point of note is you cannot change the original speaker groups it selects, well you can before you turn Bass Optimisation on (i.e old dirac) but as soon as you switch it on it puts all speakers back in the original auto calculated groups. It’s clearly on Lockdown 😁.

The other interesting question is can it make all these changes without altering any delays as I think the delays and levels you see loaded into the processor are fixed after the first mic measurement and never change.

It may be that it is changing timings but you never get to see what it has changed them to unless you measure it with REW.

Otherwise if it has fixed the original timings and levels it can only adjust other parameters such as phase presumably but I’ve always thought that If you change one it affects the other.

I think Dirac/Storm have some training planned on all this soon which may give us a bit more insight on how it all works.
 
Last edited:

lexicon

Well-known Member
Jon,

I don't have REW, but I get the gist of your results from the subsequent discussion.

Sounds confusing.
Yes - I’m trying to work through this in a logical manner. It may take a while and a number of repeated measurements to reach a conclusion I think. Bass Control gives one almost too many variables to choose from as regards crossover approach; an element of trial and error is difficult to avoid.

I’ve made a start by doing a second calibration with the subs stacked on one channel.

If I can verify the measurements with the sub in that position to my satisfaction and it still sounds as tight, I’ll then move back to relocating the sub on another channel.

Before I run Dirac Bass Control on that I will invert it and take a look at the Dirac sweep.

While obtaining nice looking measurements that conform to generally accepted norms is a good aim I don’t think this always tells the whole story. I swapped out the Arcams for the ATI at the front LCR recently and this adds a significant amount of speed and snap but with perhaps a slight loss of warmth and engagement. Significantly boosting the EQ between 40 and 100 hz for my mains also makes a positive difference to punch although no doubt this would be frowned upon in some quarters as upsetting the balance between all channels by keeping all curves the same.

But that’s the appeal of Dirac - the flexibility is there to try different approaches if you are happy to put in the time and effort.

All takes time but interesting for an AV nerd like me who enjoys playIng with these things...🤓.
 

nkatz

Active Member
Bass Control gives one almost too many variables to choose from as regards crossover approach; an element of trial and error is difficult to avoid.


It's too bad Dirac leaves a lot of the repetitive task of determining the optimal XO freq to the user, instead of letting Dirac do the grunt work of testing different ones.

My understanding is that BC uses an iterative process to try different combinations of the subs' level and phase to find which combination minimizes spatial response variation.

My question is, does it do this computationally and/or empirically?

Jon, when you run BC does sound come from the subs?
 

lexicon

Well-known Member
It's too bad Dirac leaves a lot of the repetitive task of determining the optimal XO freq to the user, instead of letting Dirac do the grunt work of testing different ones.

My understanding is that BC uses an iterative process to try different combinations of the subs' level and phase to find which combination minimizes spatial response variation.

My question is, does it do this computationally and/or empirically?

Jon, when you run BC does sound come from the subs?

Yes, I’m trying to establish what parameters such as delay, level, phase it’s actually adjusting when it does it’s calculate procedure. The answer to your question is computationally - after the final usual (however many you set) ordinary Dirac sweep measurements are taken that’s it. No more test tones/sweeps.

The displayed delays and levels loaded into the processor on filter export (calculated I understand from sweep no1) do not seem to change which to be honest I assumed they would need to!

However the timings on my last REW measurement were different so I will need to do a few more iterations to try to pin down what‘s occurring.

There’s an additional layer of complexity now and we are assuming of course that Dirac is doing exactly what it is designed to do 100% correctly on launch at version 1. That would also seen unlikely given the fact that the auto calculated crossover levels are clearly incorrect and that we are all aware that this software is under constant “development”.

Having said all that I listened to The Man from Uncle last night (great bluray for sound if you haven‘t tried that one) on the Storm and it sounded fantastic to me.
 
Last edited:

lexicon

Well-known Member
That’s really strange.

What I think you should try just to be sure though is this.

Switch off sub 2.
Run a tone generator through REW with a Hz level that you know you don’t have any nulls, say 40hz.
Check you level.
Switch off sub 1, switch on sub 2 and repeat.

It might be worth double checking as I personally have stopped using Dolby test tones as I don’t think they are very accurate.
Morning Mark
Very busy week with work so far so haven’t had a lot of time to play.
I reset both subs to factory default settings this morning first and then did the 40hz sine tone test and on that it is showing a 6db difference. You can easily hear that one is noticeably louder than the other.

I have gone back to my dealer - may need a swap out.
 

mb3195

Distinguished Member
Morning Mark
Very busy week with work so far so haven’t had a lot of time to play.
I reset both subs to factory default settings this morning first and then did the 40hz sine tone test and on that it is showing a 6db difference. You can easily hear that one is noticeably louder than the other.

I have gone back to my dealer - may need a swap out.

That certainly wouldn’t help when trying to integrate them in separate locations, but I wouldn’t worry too much. If you can change the gain on the back of the sub to match you should be ok.

My original sub3 has a +10db gain on the rest of my subs due to the original firmware they used, Ive got an -10db attenuator which sorts it out ok.
 

lexicon

Well-known Member
That certainly wouldn’t help when trying to integrate them in separate locations, but I wouldn’t worry too much. If you can change the gain on the back of the sub to match you should be ok.

My original sub3 has a +10db gain on the rest of my subs due to the original firmware they used, Ive got an -10db attenuator which sorts it out ok.

Thanks Mark but they should be within a 1-2 db tolerance at most and who knows what else may be wrong. I have just swapped them over in case it was room gain but pretty unlikely and no, the low one on the top instead of the bottom still reads 6db low. The low one was open box / ex demo so may be older. Yes, agree - could be firmware possibly.

Awaiting response from SVS via dealer but I’m clearly going to have to redo dirac yet again once this is sorted.

Going from one to two subs is proving a challenge it seems but just so glad I ordered two identical subs as I may never have found this out!
 

mb3195

Distinguished Member
Thanks Mark but they should be within a 1-2 db tolerance at most and who knows what else may be wrong. I have just swapped them over in case it was room gain but pretty unlikely and no, the low one on the top instead of the bottom still reads 6db low. The low one was open box / ex demo so may be older. Yes, agree - could be firmware possibly.

Awaiting response from SVS via dealer but I’m clearly going to have to redo dirac yet again once this is sorted.

Going from one to two subs is proving a challenge it seems but just so glad I ordered two identical subs as I may never have found this out!

Easy peezy with a minidsp mate 😉
 

mb3195

Distinguished Member
Only for someone of your technical calibre Mark 😁🤓
Nick swapping sub out tomorrow - how’s that for service?!

Really good service mate.

Being honest though, the minidsp isn’t too difficult and I’d imagine you’d be able to get better results than trying to do it through any bass management system. I logged into @steelman1991 laptop the other day, we managed to get it all EQ’d perfectly within an hour or so with me being a few hundred miles away.
 

lexicon

Well-known Member
Really good service mate.

Being honest though, the minidsp isn’t too difficult and I’d imagine you’d be able to get better results than trying to do it through any bass management system. I logged into @steelman1991 laptop the other day, we managed to get it all EQ’d perfectly within an hour or so with me being a few hundred miles away.

That sounds like a potentially sensible idea. So the subs go on one Storm channel via the minidsp and are timed first? Then Dirac sees them as one and does it’s cal from there?

Are we talking about a minidsp HD?

I’ll give Bass Control another go with the replacement sub first and if I’m not happy that seems worth a go if it’s not too costly.
 

steelman1991

Well-known Member
Really good service mate.

Being honest though, the minidsp isn’t too difficult and I’d imagine you’d be able to get better results than trying to do it through any bass management system. I logged into @steelman1991 laptop the other day, we managed to get it all EQ’d perfectly within an hour or so with me being a few hundred miles away.

Less of the "we". I only supplied the equipment, you brought the skill and expertise. But what a difference - stunning :clap: :D:D. Oh and yeah in around an hour, my system went from very good to exceptional. You'd have been burned at the stake in years gone by for what you did to my system, Albus :)
 

mb3195

Distinguished Member
That sounds like a potentially sensible idea. So the subs go on one Storm channel via the minidsp and are timed first? Then Dirac sees them as one and does it’s cal from there?

Are we talking about a minidsp HD?

I’ll give Bass Control another go with the replacement sub first and if I’m not happy that seems worth a go if it’s not too costly.

Yes, although I’d probably pull the curtains back on Dirac as results through the minidsp would probably be better, time aligning then takes a matter of minutes with the new alignment tool, it took us 2 measurements to get the crossover levels perfect. Happy to help you out Jon if needed, in some ways it’s easier remotely at a desk rather than crammed in my dark dingy cinema room where I can’t see anything!!

Less of the "we". I only supplied the equipment, you brought the skill and expertise. But what a difference - stunning :clap: :D:D. Oh and yeah in around an hour, my system went from very good to exceptional. You'd have been burned at the stake in years gone by for what you did to my system, Albus :)

Still impressed then Derek?
 

steelman1991

Well-known Member
Yes, although I’d probably pull the curtains back on Dirac as results through the minidsp would probably be better, time aligning then takes a matter of minutes with the new alignment tool, it took us 2 measurements to get the crossover levels perfect. Happy to help you out Jon if needed, in some ways it’s easier remotely at a desk rather than crammed in my dark dingy cinema room where I can’t see anything!!



Still impressed then Derek?

What gave you that impression Mark :thumbsup: You got your new "toy" yet?

@lexicon - let him do his magic I'm sure you'll be impressed :eek:
 

The latest video from AVForums

AVForums Movies Podcast: Which is the best decade for horror movies?
Subscribe to our YouTube channel

Latest News

LG rumoured to release 97-inch OLED TV in 2022
  • By Andy Bassett
  • Published
What's new on Blu-ray (UK) for Halloween 2021
  • By Mark Costello
  • Published
AVForums Movies Podcast: 20th October 2021
  • By Phil Hinton
  • Published
PSB announces M4U 8 MKII wireless headphone
  • By Andy Bassett
  • Published
Melco introduces N50-S38 music library
  • By Andy Bassett
  • Published

Full fat HDMI teeshirts

Support AVForums with Patreon

Top Bottom