First time with Dirac : sanity & expectations check ?

[chphph]

Hi everyone,

Long time lurker, first time poster !

I have recently bought a MiniDSP Flex with Dirac to improve my stereo system (WiiM Pro, Cambridge Audio 651A and Kef Q300). I have tested in the distant past the effect of Trinnov room correction on a demo Dolby Atmos system and my mind was blown, so I was coming to room correction with these expectations.

I have tried a few Dirac setups and filters with the MiniDSP, and besides the slight change of tonality that I requested with the target curve, the effect is barely audible. Now it has me second guessing everything. So far I'm quite disappointed of the result with 1k€ spent on the MiniDSP/UMIK/Dirac...

The room is fairly standard for a residential area with a wall on the left, (2m from MLP), 4m depth and the MLP 2.5m from the front wall. On the right there is no wall, 4m more of room. The Kef are front ported so I put them fairly close to the front wall.

The Dirac setup was done with the "intermediate" focus, a UMIK-1 and I did my best to respect the positions shown on the sketch in the app. This means I put the microphone outside of the couch for the furthest lateral measurements. I also tried to significantly change the Levels were good and the room was quite silent during the measurements.

I joined the before/after on REW, 1/12th octave smoothing (Second picture for R channel : yellow is OFF, green is ON). Does this look normal to you ? I saw a lot of very nice "after" FR over here with dips filled and bumps tamed, kinda jealous to be honest !

 

[alex-z]

Your system already had a good response before, I would not expect Dirac to accomplish much. You can't fix bass nulls with DSP, the mid-range ripple is probably inaudible, and the response above 1000Hz is nearly perfect.

The most substantial upgrade would be integrating a subwoofer to smooth out the bass below 100Hz. Some bass traps to smooth out the 100-400Hz region would not hurt either.

 

[TheZebraKilledDarwin]

Hi everyone,

Long time lurker, first time poster !

I have recently bought a MiniDSP Flex with Dirac to improve my stereo system (WiiM Pro, Cambridge Audio 651A and Kef Q300). I have tested in the distant past the effect of Trinnov room correction on a demo Dolby Atmos system and my mind was blown, so I was coming to room correction with these expectations.

I have tried a few Dirac setups and filters with the MiniDSP, and besides the slight change of tonality that I requested with the target curve, the effect is barely audible. Now it has me second guessing everything. So far I'm quite disappointed of the result with 1k€ spent on the MiniDSP/UMIK/Dirac...

The room is fairly standard for a residential area with a wall on the left, (2m from MLP), 4m depth and the MLP 2.5m from the front wall. On the right there is no wall, 4m more of room. The Kef are front ported so I put them fairly close to the front wall.

The Dirac setup was done with the "intermediate" focus, a UMIK-1 and I did my best to respect the positions shown on the sketch in the app. This means I put the microphone outside of the couch for the furthest lateral measurements. I also tried to significantly change the Levels were good and the room was quite silent during the measurements.

I joined the before/after on REW, 1/12th octave smoothing (Second picture for R channel : yellow is OFF, green is ON). Does this look normal to you ? I saw a lot of very nice "after" FR over here with dips filled and bumps tamed, kinda jealous to be honest !

There's a big problem with all these room correction systems nobody is talking about (because knowledge about it could hamper sales).

Think about this: before you start repairing your car, what is the pre-requisite. That you understand how the car works.
What is the aequivalent to listening setups?
To buy a room eq system, like lay people are made believe, because that's how things are being sold?

First you must be able to hear certain details. But not only that: you also must have a rough understanding, where in the frequency range these details are happening. Other wise you cannot make the connection of hearing and perception with a measurement.

So you need to know where dialogue of male or female voices are in the frequency spectrum. You must know where high notes, like a piccolo flute or screaming women or crowds are placed in the spectrum.


This is hugely important, because the automatic system is much less sensible, than the hearing.
YOU must be able to tweak the system and you can only do that, if you know what to listen for and know, where in the spectrum this detail is being measured.


Since you have spent the money already, you should do the following to become able to get something out of your system now:

On the computer for your music player get a parametric EQ and an analyzer (i.e. JRiver media center).
Use headphones and songs you know and with a sound you personally like.
But choose parts of songs with instruments, where you can barely hear these instruments in certain moments on the headphones.
Then use the parametric EQ with a sharp bandpaß (or, if it does not have a bandpaß, use a steep high and lopass filter and move them narrowly together, so that only a relatively small spectrum is getting through).
And then tune the frequency of the filter(s) until you can hear this instrument best. Then look at the frequency analyzer.
That's the frequency range of that instrument at that moment.
Memorize the range!
Compare the audibility of that instrument to your room setup.
Is it audible or is it lost?
Do this with several songs, instruments and movie soundtracks.

Do it with the midrange, forget the bass and treble for now (the foundation of the whole sound is always determined how the midrange sounds).

Also do it vice versa:
A sequence that does not sound good in the room. Lets say a screaming crowd that may hurt your ears.
Or a car crash and treble that hurt.

Play that back over headphones. If it sounds good on HP, find the hurting range you hear in your room on the headphones. This gives you a first indication about a problematic frequency range in the room.
Important: this does not mean, that the measurement shows a peak or a bump! It only tells you, that you hear something in that range, that is not correct - no matter what the measurement shows.

Why is that even possible? SHouldn't everything be refelcted in measurements?
Sure! But it's a matter of measurement resolution and if it is even showing up int he frequency domain. It could be a time domain problem.

A good sounding setup does not come automatically. This is impossible for a very simple reason:

What we perceive as good sounding, has much much finer resolution than any automatic room eq system can correct. It comes down to RELATIVE spectral balance levels. For example 1 dB in the 1.2k range in relation to the 600, may make a lot of difference in loud scenes in your room.

Forget all the talk about house curves as if they would tell the story. You personally may like to hear more details in the midrange, or like a brighter sound working better at lower levels.
All this cannot answer the auto EQ system.
And you can't answer that yourself, until you know, what to listen for. Only whne you know what to listne for, you can discover, what you want to hear and therefore the sound you prefer.

Frequency response measurements: one room may have a nasty 4k reflection that does not even show up in the overall spectrum. But once a loud scene happens and the mix contains a lot in that range, you may notice a harsh tone, that may even hurt.

But most people, if they do not have knowledge about frequency ranges, look at the measurements and are trying to guess, what is going on. They may be tweaking the air range, but the problem might be in the 4k or 8k range.

This can be avoided by either getting a true professional (who is also setting up professional studios) to tune your system, or by yourself starting to educate your ears.

 

[Old_School_Brad]

Looking at your before and after traces, it's clear why you didn't see a dramatic difference. Your initial response was already excellent before applying Dirac Live, and you evidently preferred that original response, given that the house curve only made minimal adjustments. Consequently, it's understandable why the changes weren't miraculous. Your system was performing well from the start, and Dirac Live's fine-tuning, while beneficial, couldn't drastically improve something that was already very good.

 

[GXAlan]

I have owned Trinnov and Dirac AV processors. Trinnov is on another level in terms of room processing even though it doesn’t try to flatten the entire response as aggressively as Dirac by default. The Trinnov Nova is an option if you just want 2 channel Trinnov but even then, it’s probably going to be subtle.

The way Dirac works is that the first listening position calibration is what it uses for delay. The rest is for averaging and you really should only move around where you intend to sit. It doesn’t know or attribute different algorithms for different positions. It’s an unweighted average of everything.

If you show your Dirac graphs *and* show us the spread of before and after, we can probably offer some added input.

 

[TheZebraKilledDarwin]

Looking at your before and after traces, it's clear why you didn't see a dramatic difference. Your initial response was already excellent before applying Dirac Live, and you evidently preferred that original response, given that the house curve only made minimal adjustments. Consequently, it's understandable why the changes weren't miraculous. Your system was performing well from the start, and Dirac Live's fine-tuning, while beneficial, couldn't drastically improve something that was already very good.

Nope, he's just another victim of the false information in hifi forums, where PR is being regurgitated to sell equipment to the uneducated masses.

One only has to look, how many people believe that the frequency response was a tell all measurement, while the more important time domain is being ignored.
And where nobody is even talking about hearing and education of the ears.

It's all just a regurgitation of easy to sell PR nonsense.

 

[Old_School_Brad]

Nope, he's just another victim of the false information in hifi forums, where PR is being regurgitated to sell equipment to the uneducated masses.

One only has to look, how many people believe that the frequency response was a tell all measurement, while the more important time domain is being ignored.
And where nobody is even talking about hearing and education of the ears.

It's all just a regurgitation of easy to sell PR nonsense.

I strongly disagree. For me, Dirac Live has been the most significant and transformative purchase I have ever made. It has completely revitalized and reignited my passion for music. The difference it has made in the quality and clarity of my listening experience is unparalleled. Dirac Live has elevated my entire audio setup, allowing me to rediscover and appreciate my favorite music in ways I never thought possible. And I have spent A LOT of time with REW and rePhase as well.

But I can appreciate different expectations.

 

[alvigg]

Old school - is that with the generic version or DLBC? The latter has also done wonders for my theater.

WIth MiniDSP downrezing another about 48kHZ, is that an audible difference for higher resolution files?

 

[Old_School_Brad]

Old school - is that with the generic version or DLBC? The latter has also done wonders for my theater.

WIth MiniDSP downrezing another about 48kHZ, is that an audible difference for higher resolution files?

Standard DL. With and without subwoofers until I got the full range speakers I wanted. Then it has been 2.0 for years.

No. There is definitely no audible degradation. For me.

 

[TheZebraKilledDarwin]

I strongly disagree. For me, Dirac Live has been the most significant and transformative purchase I have ever made. It has completely revitalized and reignited my passion for music. The difference it has made in the quality and clarity of my listening experience is unparalleled. Dirac Live has elevated my entire audio setup, allowing me to rediscover and appreciate my favorite music in ways I never thought possible. And I have spent A LOT of time with REW and rePhase as well.

But I can appreciate different expectations.

I believe you. But such personal experiences probably are the exact reason, why he was motivated to buy the equipment...

I find it strange that people believe, once they have bought speakers or an automatic room eq, to have now aquired educated ears and knowledge about sound quality.

 

[ozzy9832001]

I think your expectations were too high and you would benefit more from some passive room treatments to flesh things out a bit more.

It's not a bad result by any means -- but it's hard to improve upon a good response.

For DIRAC, I only use 3 measurements -- left, right and center.

Otherwise, it never, ever sounds right at all.

 

[dshreter]

I thought the whole point of Dirac was that it was not just a parametric EQ to smooth frequency response, but a full convolution filter based on impulse response. There are much less expensive and less sophisticated solutions for just PEQ.

Does looking at the frequency response even indicate if Dirac is “working”?

 

[Old_School_Brad]

I thought the whole point of Dirac was that it was not just a parametric EQ to smooth frequency response, but a full convolution filter based on impulse response. There are much less expensive and less sophisticated solutions for just PEQ.

Does looking at the frequency response even indicate if Dirac is “working”?

Agreed. It seems the OP has a strong setup, and with a good system in a good room, Dirac likely wouldn’t make a dramatic difference.

 

[chphph]

Nope, he's just another victim of the false information in hifi forums

Reading ASR reignited my interest for room correction system, but this interest comes from a mind blowing experience with Trinnov, a talk with the engineers and overall getting up to date on the effect of the room on audio reproduction. Ultimately, I made a choice informed to the best of my abilities to improve my hifi system. This is quite a hurtful comment

Before getting a UMIK and having a first experience with Dirac, it was impossible for me to know if my room/loudspeakers were well behaved. In the end, I have a pretty cheap setup (Kef Q300, 13 years old and 500€ at the time) and an untreated room. I set the target to be close to my loudspeakers on purpose in order to see first what kind of time domain correction Dirac was bringing. I am now experimenting different target curves, although I prefer to play around with EQ/PEQ because I can do that with the phone, no need to whip out the laptop and USB cable to the MiniDSP.

Now, two things made me feel less "meh" regarding this MiniDSP and Dirac :
- I got hold of a pair of Kef R300, far superior to my Q300 but tuned differently (quite medium heavy, deeper bass but quite subdued). In 15min I could get them to the tuning I like and (among other things) profit from the low distorsion bass with a boost in the target curve. These are a loan and will have to go back, but it definitely put in perspective the interest of Dirac.
- Realising I really like deeper bass, I now have a very good swiss knife of a device to perform the integration of one or two subs

I think your expectations were too high and you would benefit more from some passive room treatments to flesh things out a bit more.

This is a good point, the KEF R300 experience does push on that point : I like the overall tonality of my current Q300 and there's a big benefit in reducing the speaker distorsion. Room treatment is another big avenue. I'll investigate room treatment, although there will be fairly high WAF constraints

In the end, there's still a feeling of "this could do more" in terms of acoustic correction but maybe that's the gap with Trinnov, which is at least 4x more expensive for a whole system (Amethyst x10, Nova x4). The thing that struck me most with Trinnov was the pinpoint imaging in a 16.2 Atmos setup, almost surreal....!

In terms of louspeakers choice however I realise I now have more freedom of choice with Dirac and I can focus on low distorsion and a directivity that I like. Bass extension can be dealt "easily" with subs and MiniDSP now, as well as general tuning of the speaker.