First Go at Dirac Live - Is My Response Normal?

mrrodge

Active Member
Hi all,

I have a newly set up Arcam SR250 with Focal 926.

More or less have an equilateral triangle for listening position/speaker distances and placement isn't (too) bad; right speaker is a bit closer to the corner than I'd like but isn't equidistant from the walls. Left channel is mid-wall and is a bit closer to it than I'd like, which sort of influenced my decision to go for the Focals (no rear ports). The room is carpeted wall to wall.

So I did the very wide soundstage measurements and filter with Dirac Live last night. I left Dirac's suggested (flat) curve and applied it to the receiver. Measurements/curve as below.

Dirac Curve First Attempt.png


I played a 96/24 copy of Abbey Road into the receiver (via optical from Raspberry Pi Squeezebox player, Qobuz) with the Dirac filter off and wasn't displeased with the sound; it was much better than my previous setup. Even though the room has been rearranged and positioning is nothing like my last setup, the sound exhibited many similar and familiar traits, though just seemed better controlled.

Half way through Here Comes the Sun I turned on the Dirac filter (expecting marginal difference) and was really, really shocked at how drastic the change was. It actually made my ears feel funny.

I'm no audiophile so not sure if I'm using the right words but the clarity went through the roof; particularly the vocals. I wouldn't call the sound bright, just clear. The thing is though, I listened for a while longer and couldn't really warm to the sound. It's like nothing I've ever heard before, but then again I've never had a setup like this before so don't know if it's right or not. Have I been conditioned to bad sounding systems previously, and has everything I've ever listened to sounded 'wrong'?

It took out the bass I'm used to, as is reflected on the graph. The lows aren't gone, they're definitely still there, just not taking over, which I think is probably a good thing. All in all I thought Abbey Road sounded really, really good.

The problem came with 'noisier' music. I put on my FLAC rip of (original and remastered, tried both) copies of Definitely Maybe and I also tried the Biffy Clyro single Tiny Indoor Fireworks (Lossless FLAC). I want to use the word 'hissy' but it wasn't hiss, I think it was silence/separation, even during noisy choruses. It's like the noise and impact (not the volume) of the music has been taken away and it's almost too clinical and isn't exciting. It's not tinny, it's sort of weak. Turning it up loud doesn't sound loud; yet I have to shout to hear myself. I found myself going much louder than I normally would, but it didn't give me what I was trying to find. Can anyone relate to this? Is this what they mean when they say the music isn't fatiguing any more? Was I looking for the fatigue? Really struggling to explain it.

The best metaphor I can give is at normal, low listening volumes. Imagine watching the BBC News channel via and off the shelf surround sound system in a stereo configuration with a subwoofer and the reporter reading the headlines. Usually the result is a boomy voice and a pretty loud, intrusive, way over the top sound just for watching the news. Then when you switch the speakers off and the TV's own speaker kicks in the sound goes quiet and neutral, almost like a feeling of relief, like when you take your socks off (or is that just me?!). You immediately perceive the relief as 'wow the TV's speaker is crap' but you soon adjust and realise it's actually better than sound hitting you from all sides just to watch the news. That's the effect I get when I turn on the Dirac filter mid song, but I don't think I really got the realisation 'this is actually better'.

All in all, not sure where to go next. Most things I've read on Dirac suggest that I should persevere and get used to the suggested curve, and that altering the curve is out of my league. What's the thoughts and experiences of others? Thanks and sorry for the lengthy post!
 
Last edited:

Nico72

Active Member
Thanks for sharing your results.
There are YouTube tutorials on how to tune your target curve View: https://m.youtube.com/watch?v=3NX53v-zK4A

You can drag up and down those yellow dots to reproduce some of that low frequency bump if you liked it. I think the video also suggest adding a target point in the deepest through and avoid correcting upwards by more than 4 dB, and also follow the natural drop of the speakers at the top end (though you don’t seem to have much of that).
The fact that you can tolerate higher volume without feeling uncomfortable is a good sign, as it generally means there is lower distortion.
Anyway I suggest you try to tune the target curve a it to suit your taste if you have time, and let us know if that works.
 

mrrodge

Active Member
Thanks for the quick reply - I'll definitely check out the video(s) when I get chance.

You can drag up and down those yellow dots to reproduce some of that low frequency bump if you liked it.

Thing is, I'm not sure I liked it - I know I referred to a lack of excitement in a way but I did like how the bass seemed better placed and less wobbly with the filter on, so I'm not sure I'm concerned over a lack of it as the lows are still there and don't overpower.

The fact that you can tolerate higher volume without feeling uncomfortable is a good sign, as it generally means there is lower distortion.

That's good to know, thanks. It's almost like empty space between the instruments; I don't know how else to describe it.

Are there many other Dirac Live users out there? If so, can you post your curves and the types of music you listen to? I don't know where to start; looking for a consensus really.
 

RBZ5416

Distinguished Member
Which version of Dirac are you using? Should be 3.0.11 from Dirac themselves rather than ye olde version that may still be lingering on Aram's site.

Which mic? The Arcam supplied one isn't great in either incarnation & most use a UMIK or XTZ.

Dirac has flattened some severe bass boost that you will have become accustomed to. As you suspect, it will take some time to get used to the new cleaner sound.

It's possible to reduce the frequency range so that Dirac only corrects the bass if you really don't like what's happening further up the frequency range. There are also numerous variations of curve you can load as well as designing your own. Having said all that, some people simply don't like Dirac for music.

I'd suggest posting here instead, as there's plenty more experience of Dirac with the multi-channel versions of the amp.

 

mrrodge

Active Member
Which version of Dirac are you using? Should be 3.0.11 from Dirac themselves rather than ye olde version that may still be lingering on Aram's site.

Which mic? The Arcam supplied one isn't great in either incarnation & most use a UMIK or XTZ.

Yup I got the latest direct from Dirac, used the integrated USB one that came with the receiver and the appropriate calibration file from Arcam.

Dirac has flattened some severe bass boost that you will have become accustomed to. As you suspect, it will take some time to get used to the new cleaner sound.

Was coming to that conclusion; may listen for a few more days before trying anything different I think.

There are also numerous variations of curve you can load as well as designing your own. Having said all that, some people simply don't like Dirac for music.

You mean there's presets within the Dirac software? I looked for something like that and couldn't find them.

I'd suggest posting here instead, as there's plenty more experience of Dirac with the multi-channel versions of the amp.

Will take a look, thanks.
 

mrrodge

Active Member

Hear Here

Active Member
I have Dirac Live with my NAD M33. Included is a calibrated mic but I'm not convinced of its quality and the "calibration" appears to be built into the NAD which makes me suspect that it’s a generic calibration representative of the type of mic provided and not for the individual mic.

I’ve not been convinced that there’s an IMPROVEMENT in sound, although certainly it’s DIFFERENT. I’ve found the same with all the DSP systems I’ve used in my system (Dirac, Anthem, RoomPerect, MARS) and all difficult to know how accurate the results are without access to post-application measurement with independent software and mic. With all DSPs I’ve found that some of the music's life and excitement is sucked out. This was worst by far with RoomPerfect (in a Lyngdorf amp) but to a certain extent with the others too.

I'll persevere with Dirac a little more, but longer term I want to use the DSP that's included in my Avantgarde speakers to do the job instead as this should be less inclined to spoil the music's excitement.

I find the instructions totally inadequate, for example, how should one adjust the microphone level on the second page of the measurement process?

Does the Arcam system recommend that the mic is pointed upwards towards the ceiling? NAD suggest this, but I'm not convinced that this should provide the best result. What about ceiling light fittings? Won't they adversely influence the readings? The trouble is that the calibration table assumes the mic is placed in the direction recommended. Commercial measurement mics should have 2 tables - one for forward facing an a 90 degree one for upward facing use. Unless I'm mistaken in my understanding!

Peter
 
Last edited:

Khazul

Well-known Member
TBH while I think dirac is the best of the non high end systems, I have never liked any of them for music and your description of your sense of it certainly rings some bells.

I am quite used to listening to and working with neutral near field monitoring, so it is definitely not a case of not being used to the resulting near flat response. To me there is always something wierd about (I suspect) how they attempt to compensate for first reflections - which is part what I think is responsible for feeling that the sound is being force injected into your head. The effect to me is always something that in the past would have set off my personal broken-mix-needs-fixing alarm bells from when I use to work with audio production.

While nowhere near perfect, I always prefer simple but high precision digital EQ based room correction (just because digital filters can so some tricks that analog filters cannot).

Oddly for TV/movies, I am not bothered about them at all - maybe because it is surround an the first reflections are more naturally present in 5.1/atmos etc mixes and so I dont get this drilling into your head sensation. Actually it reminds me more of when I have messed around with micro timing to setup precise sound positioning in a mix using time difference as well as level difference and got it badly wrong so that they conflict with the result that your brain just feels confuses due to conflicting perceptions.

(My distance timing for surround setup here is very precise at the ILP - not that either).
 
Last edited:

RBZ5416

Distinguished Member
Commercial measurement mics should have 2 tables - one for forward facing an a 90 degree one for upward facing use. Unless I'm mistaken in my understanding!
No, you're spot on. In a stereo environment the mic should point forwards. But unfortunately, DSP is still largely thought in a multi-channel environment where the mic should be pointed up to gather sound equally from the front, rear & sides. Consequently mics like my XTZ only have a 90° calibration file & not a 0° for forward facing.

You can use REW to check the actual results of what Dirac is doing rather than what the software would like you to believe it's done.
 

3rdignis

Active Member
When it is said the result is flat, compressed or not dynamic I believe this is a result of losing headroom.
As stated don't boost more than 4db (I don't boost at all). if you imagine a pipe with a graph of a musical waveform, if you boost the graph goes beyond the pipe so you have to compress to fit the signal down the pipe, thus losing headroom.
 

3rdignis

Active Member
When you have a graph, use this program to find better positions for your speakers, try to get the base close to 0db while maintaining extension and not doubling up on nulls.
Arqen.com excel wall bounce.
 

3rdignis

Active Member
I don't eq above 300hz, I use acoustic panels from mafia.com at 1st reflection points 4 ceiling, 3 behind and 1 left 1right. And toe in.
 
Last edited:

3rdignis

Active Member
Eq gives less but more accurate bass.
You have to lessen accuracy because nulls can't be completely removed.
Cut the modes less to balance bass and you should get used to the new sound.
 

Hear Here

Active Member
No, you're spot on. In a stereo environment the mic should point forwards. But unfortunately, DSP is still largely thought in a multi-channel environment where the mic should be pointed up to gather sound equally from the front, rear & sides. Consequently mics like my XTZ only have a 90° calibration file & not a 0° for forward facing.

You can use REW to check the actual results of what Dirac is doing rather than what the software would like you to believe it's done.

That's all helpful and encouraging. Since I'm only interested in 2 channel, I suspect that I'd get better Dirac results if I bought a mic that has the 2 calibration tables and I use it facing forwards using the appropriate table. I've had this rather costly miniDSP UMIK-1 mic recommended but happy if you could suggest an alternative -

Amazon productView: https://www.amazon.co.uk/dp/B00N4Q25R8/?coliid=I3GQH98W9EHUV8&colid=VLNA592THMDB&psc=0&ref_=lv_ov_lig_dp_it


It's very odd that the sound with Dirac filter engaged is so different, especially as I have the version of Dirac that only adjusts below 500 Hz. It produces full-frequency sweeps and the screen shows the results of the full sweep, but the "straight line" recommendation is only from 500 Hz downwards. Nevertheless the sound through the entire range appears to have been adjusted. This may be an acoustical illusion though!

Peter
 

RBZ5416

Distinguished Member
I've had this rather costly miniDSP UMIK-1 mic recommended but happy if you could suggest an alternative -
The UMIK is pretty much the tool of choice along with the XTZ Pro, but as I said XTZ doesn't have a 0° calibration. There is a version of the UMIK that comes with individual rather than generic calibration that isn't much more expensive. If wanting cheaper you may as well stick with the one supplied with the amp.

Although Dirac "light" supplied with NAD only correct low frequencies, I believe the impulse correction is still applied to whole range.
 

Paul7777x

Distinguished Member
I have Dirac Live with my NAD M33. Included is a calibrated mic but I'm not convinced of its quality and the "calibration" appears to be built into the NAD which makes me suspect that it’s a generic calibration representative of the type of mic provided and not for the individual mic.

I’ve not been convinced that there’s an IMPROVEMENT in sound, although certainly it’s DIFFERENT. I’ve found the same with all the DSP systems I’ve used in my system (Dirac, Anthem, RoomPerect, MARS) and all difficult to know how accurate the results are without access to post-application measurement with independent software and mic. With all DSPs I’ve found that some of the music's life and excitement is sucked out. This was worst by far with RoomPerfect (in a Lyngdorf amp) but to a certain extent with the others too.

I'll persevere with Dirac a little more, but longer term I want to use the DSP that's included in my Avantgarde speakers to do the job instead as this should be less inclined to spoil the music's excitement.

I find the instructions totally inadequate, for example, how should one adjust the microphone level on the second page of the measurement process?

Does the Arcam system recommend that the mic is pointed upwards towards the ceiling? NAD suggest this, but I'm not convinced that this should provide the best result. What about ceiling light fittings? Won't they adversely influence the readings? The trouble is that the calibration table assumes the mic is placed in the direction recommended. Commercial measurement mics should have 2 tables - one for forward facing an a 90 degree one for upward facing use. Unless I'm mistaken in my understanding!

Peter

For stereo music and 2.1 or 2.2 Dirac recommend pointing the mic to the speakers, that is 90 degrees from vertical, as do Minidsp in all of their Dirac incarnations.

I do too. I tried both options several times with several different speakers, and the sound, to me, is noticeably fuller that way.

Sorry for the crap description, what I mean is that it hangs together as the whole spectrum.

With the mic pointed to the ceiling I found the bass on some complex tracks was a little ‘seperate’ is the best way to describe it.

Ps, that was using the UMIK-1.

Pps, Arcam provide two calibration files for their mic, one for vertical and one at 90 degrees.
 

mrrodge

Active Member
What's this 500hz limitation? I haven't seen anything to suggest mine was limited in any way with the Arcam.
 

mrrodge

Active Member
Pps, Arcam provide two calibration files for their mic, one for vertical and one at 90 degrees
Can you direct me to those? I found two files on their site for the SR250 but one was for the older style mic. I did my measurements with the calibration file I could find and the mic pointed up.
 

Paul7777x

Distinguished Member
Can you direct me to those? I found two files on their site for the SR250 but one was for the older style mic. I did my measurements with the calibration file I could find and the mic pointed up.


The SR250 (discontued) page has a list of downloads if you scroll down to the bottom.

There you’ll find a pair of mic files and curves too. I’ve no idea how useful the curves are, I didn’t try them.

But they are worth a shot if only to calibtprate your ears to a couple of different curves so you can get a better idea of what the graphs mean in reality.
 

Paul7777x

Distinguished Member
Aha. Just checked again and those ones are for two seperate mics.

I’ll check again, I’ve sold my SR250 so I may have lost the source of the files.

In any case, use the one you used before and try pointing it at 90 degrees for another sweep.

There is also a Harman Kardon curve I found very useful, though I never found it on file.

I’ll find it again and send you a copy. It means fiddling with the curve manually but it is worth persevering.

I took me a couple of weeks of messing about until I had what I wanted and I loved it for years.
 

Paul7777x

Distinguished Member
Well, my mind seem s to be going off at pace...

It wasn’t the HK curve I used it was this one.

EE072039-238C-4924-AE1E-530A9F2F0D00.jpeg
 

Hear Here

Active Member
That
NAD only provide a "Light" version of Dirac unless you pay for a full range licence.
That's right. The full version is available from Dirac for a $99 if I remember correctly. I understand that DSP is most successful in the bass area as that's where room characteristics mess up the speakers' sound.
 

The latest video from AVForums

Oculus Quest 2 VR headset + Rotel A14 MkII Amp Reviews & Best of the Month
Subscribe to our YouTube channel

Latest News

Samsung QD-OLED TVs primed for 2022 launch?
  • By Andy Bassett
  • Published
Sky to add NBCUniversal's Peacock service
  • By Andy Bassett
  • Published
Bang & Olufsen announces Beoplay EQ TWS earphones
  • By Andy Bassett
  • Published
AVForums Podcast: 28th July 2021
  • By Phil Hinton
  • Published
Sky drops Sky One brand and introduces Showcase and Max
  • By Andy Bassett
  • Published

Full fat HDMI teeshirts

Support AVForums with Patreon

Top Bottom