Room correction on OS X thread

[outofthebox]

The "logic" is dictated by economics. Presumably someone has done a calculation that says they would be better off selling X copies of the software at $600 than 10X copies of the software at $60. The implicit assumption is that they don't foresee selling many copies of the software.

.

 

yes i think that's pretty much it. you've nailed Dirac's pricing logic here. it's a bit of a niche product as of now.

 

in the hardware boxes world, the closest example would be the logic of spending about $2 to 3k each in the digital and D/A source chain - on a transport, DAC, a DEQX, by that logic it's definitely cheaper.

[Bob Stern]

Dirac make a big deal about their filters being "mixed phase". That in itself is not unique. iZotope SRC (used in Audirvana) inherently is mixed phase if you set the pre-ringing parameter to any value other than zero or one.

 

More specifically, if you look at Dirac's patent applications, they are directed to the problem of how to reduce the computing power to implement digital crossovers and EQ for car audio where the power of a full-blown computer is not available. The patents describe schemes for splitting the computation between FIR and IIR filters so as to reduce the required length of the FIR filter.

 

This is not a problem relevant to desktop computer audio. For example, iZotope SRC permits the use of FIR filters with 2 million taps.

[wgscott]

In fairness, they also put it in cars I can't afford: Dimensions for automotive | Dirac Research

[flak]

Dirac make a big deal about their filters being "mixed phase". That in itself is not unique.

 

Hello Bob,

 

you are right, as I said in my previuos post mixed phase is not proprietary to Dirac and is used by some other solutions.. it is nonetheless a prerequisite to be considered a modern DRC solution.

 

What is unique is the way it is implemented in its proprietary algorithm based on multipoint measurements... you can check its performance yourself in real life situation by listening and by measuring impulse and frequency response in different positions as well as by comparing Dirac Live to other products.

 

Dirac Research has many patents and has conducted extensive research in many fields like active noise control, sound field synthesis and room correction... a show of its technological leadership is currently present at the CES

 

Ciao, Flavio

[flak]

In fairness, they also put it in cars I can't afford: Dimensions for automotive | Dirac Research

 

Hello Wgscott,

 

the link you are quoting is much more meaningful that many know about ..

a car is a much more difficult environment that a listening room, it's easy to understand that the so called "soundstage" and imaging in your (I mean everybody's) car is not probably very good.

 

A careful reading of your link can help in understanding what Dirac's technology can do...

the result is a physical sound field that you never thought possible in a car, decoupled from the physical loudspeakers and extending as if speakers were placed outside the car.

 

By the way CES visitors at our booth have had a real glimpse into what is technologically possible when no cost constraints are imposed on our researchers and engineers

 

Flavio

[dallasjustice]

DIRAC's pricing reflects that it's the easiest and most effective software DSP solution available. It costs slightly more than it's competitors. However, it offers the greatest value, IMO.

 

Value isn't determined by the forex or marketing terminology. Value is always determined by individual buyers in the marketplace. I personally think DIRAC offers more value than a $10,000 component upgrade, when well setup.

[Boris75]

DIRAC's pricing reflects that it's the easiest and most effective software DSP solution available. It costs slightly more than it's competitors. However, it offers the greatest value, IMO.

 

Value isn't determined by the forex or marketing terminology. Value is always determined by individual buyers in the marketplace. I personally think DIRAC offers more value than a $10,000 component upgrade, when well setup.

 

Dear Dallasjustice,

 

Which mike did you get for Dirac?

 

Boris.

[dallasjustice]

Dear Dallasjustice,

 

Which mike did you get for Dirac?

 

Boris.

Earthworks m23. I also used the earthworks to measure my speakers with the DEQX HDP-4. I applied a gentle filter between 1khz-14khz in the DEQX. It's a nice mic for sure. But not necessary for DIRAC. There are cheaper ones that work well too.

[Boris75]

Earthworks m23. I also used the earthworks to measure my speakers with the DEQX HDP-4. I applied a gentle filter between 1khz-14khz in the DEQX. It's a nice mic for sure. But not necessary for DIRAC. There are cheaper ones that work well too.

 

Many thanks for your reply. I see on Earthworks' website that this mic ends with a XLR connector which suggests that an A/D converter is needed to get the signal to the computer. If I may ask a follow-up question, what did you use to digitize the signal?

[Kal Rubinson]

I use ARC at 88.2 and 96kHZ regularly, and this is the first I've heard about down sampling... I believe that miscommunication comes from the fact that it takes measurements at 48kHZ through its stand-alone application;

I believe that presumption comes from the observation that it downsampled to 48kHz in many of it hardware implementations in popular AVRs and prepros.

[WINDIANRECORDS]

I believe that presumption comes from the observation that it downsampled to 48kHz in many of it hardware implementations in popular AVRs and prepros.

 

Agreed, and so I set off to find out if that was the case in the standalone hardware ...

 

If you go back to post #24, you can see that I went back and measured the output with ARC in the chain; it does 'downsample', if that's what you'd call it. Since I mix/master in 88 or 96, ARC is really only 'downsampling' while I am monitoring - when I actually bounce/create the track after all my work, ARC is not active, and therefore the file is still 88/96.

 

For discrete listening on higher end systems, people may want to be aware that their hi-res files will not be at their max bit rate if ARC is being used, but I still use ARC for listening to music in my studio even when I am not mixing.

[sdolezalek]

Given all the discussion we have had about the effects of filtering on DACs and the implication that a significant part of why 24/192 may sound better than 16/44 is that it allows more flexibility in filter designs:

 

Doesn't room correcting software imply all kinds of filtering right in the middle of the audible spectrum and all of the pre-ringing and group delay issues that come with it???

 

After playing around with Dirac over the last week, I came to the view that although the frequency changes (which were fairly modest) had certain benefits, they robbed the music of much of its impact and "liveliness." I have heard others say: "That's just because you are used to the original frequency curves", but I'm convinced there is more to it than that.

 

Thoughts?

[Superdad]

After playing around with Dirac over the last week, I came to the view that although the frequency changes (which were fairly modest) had certain benefits, they robbed the music of much of its impact and "liveliness." I have heard others say: "That's just because you are used to the original frequency curves", but I'm convinced there is more to it than that.

 

Sdolezalek:

As a fellow Audirvana+ user who spent a lot of time giving DIRAC Live a good trial, I have to agree with you. But I think the impact and life gets lost not because of Dirac's filters, but because their playback engine has to be inserted in the chain--essentially defeating the benefits of A+. I say this because I was disappointed in the SQ of DIRAC Live (versus A+ straight my DAC) even with the filter I made turned off.

Perhaps if Dirac Research ever designs a version that works as an AU plug-in, then I will be able to try it in a more integrated was with Audirvana.

 

The above is not at all meant to take anything away from what is an excellent and refined piece of software. And DIRAC Live is also one of the easiest to use (no small feat). People whose rooms and speakers have some serious problems will realize a great deal of benefit from its use. But my room/system is already quite smooth (the filter Dirac created for me was very mild), and the benefit of flatter response was outweighed by the loss of realism (again, due I think to the engine and not the correction).

 

I encourage everyone to give DIRAC a try in their own systems! It is educational as well.

[flak]

Hello Sdolezalek,

the issue you are raising is an important one, there are several aspects to take into consideration... (forgive me if I'll use some "cut and paste")

 

The first one is that the DSP gain has to be adjusted to the lowest value which does not generate digital clipping.

You'll find the instructions in the manual... its value depends not only from the amount of correction required by your specific listening room but also from the music material and the way it has been recorded so that fine tuning of the gain is the best option if patience and care are an option.

 

The fact is that actual situations vary widely... not only the adjustment varies depending on the input, but it is also dependent on the different correction required in each individual listening room.

Also it should be noted that Dirac correction is applied in the frequency AND in the time domain... and the time domain gain depends on the input signal, the filter and the phase.

In general you'll find that "old school" and high quality recordings require less DSP headroom than modern recordings that follow the "loudness war" guidelines.

 

The other aspect is that as you know room modes create substantial "room gain" at certain frequencies, and because they are resonances they persist in time... as a result when they are removed the "bass impact" is reduced so that increasing the bass in the target curve can eventually be useful to restore the "impact" but without the resonances.

(you will find that if the bass boost was felt as necessary because of an "addiction" to room modes you will be able, if necessary, to gradually reduce it again in time)

 

In other words the fine tuning of the DSP gain and of the target curve will restore the impact and the "liveliness" of the music while maintaning the smoothening of the frequency response and the impulse response correction.

It may take some time and patience but it should be fun also

 

Good listenings,

Flavio

[wgscott]

I think the point Flavio makes about the bass resonances is a very good one. I had a few very sharp peaks that are greatly smoothed out by Dirac. Overall, it sounds much smoother and better, with much less resonance (which I can really appreciate when in the corner of the room opposite my sub), but my initial impression was that the bass lacked the punchiness I had grown to like. (I have a Rel, which is the sub for people who don't like subs, so this in a way is more of a confession of my own shortcomings than anything else). I finally realized I was being fooled into thinking the resonances sounded good, when really what I was appreciating was an artifactual audiophool sound effect. After a couple of months with Dirac, I really now notice the shortcomings of my system without it. These were shortcomings I had almost 4 years to get used to as "normal." Gone also is the BBC dip; at first I thought the high midrange was too forward and shrill, but now when I turn off Dirac it is as if a veil has been lowered upon ... ok, sorry, but you get the point.

[dallasjustice]

Many thanks for your reply. I see on Earthworks' website that this mic ends with a XLR connector which suggests that an A/D converter is needed to get the signal to the computer. If I may ask a follow-up question, what did you use to digitize the signal?

Boris,

That's right. The EW mic needs 48v power to work. I use an RME HDspe Pcie soundcard to convert from Analog to Digital. Because the RME card is for the pro market, there's no built in mic-pre. If you use an AtoD that doesn't have a built in mic pre, like the RME, you will also need to find a good mic pre with 48v power. I selected the Grace M101. The m101 is well known in the pro world for it's neutrality. I think every unit comes with it's own phase and frequency response measurements in the the manual. It's really nice and I highly recommend it.

 

Obviously, I take things to the extreme and it's really not necessary to spend that much on the test gear to get good filters. I've just found a slight improvement with the better test gear; YMMV. If you ever become a DSP-addict, like me, and want to do speaker correction (like I did with the DEQX) the mic really matters and the earthworks is an excellent choice. I've got remarkable results with it. :-)

 

Michael.

[dallasjustice]

I think the point Flavio makes about the bass resonances is a very good one. I had a few very sharp peaks that are greatly smoothed out by Dirac. Overall, it sounds much smoother and better, with much less resonance (which I can really appreciate when in the corner of the room opposite my sub), but my initial impression was that the bass lacked the punchiness I had grown to like. (I have a Rel, which is the sub for people who don't like subs, so this in a way is more of a confession of my own shortcomings than anything else). I finally realized I was being fooled into thinking the resonances sounded good, when really what I was appreciating was an artifactual audiophool sound effect. After a couple of months with Dirac, I really now notice the shortcomings of my system without it. These were shortcomings I had almost 4 years to get used to as "normal." Gone also is the BBC dip; at first I thought the high midrange was too forward and shrill, but now when I turn off Dirac it is as if a veil has been lowered upon ... ok, sorry, but you get the point.

+1 Well said! I think DSP is different for everyone. For some it's an immediate revelation, for others it takes some getting used to and some others never really get it.

 

For several years, I fell into the last category. I was very critical of DSP. I used to think DSP was only for folks with a crappy room or mediocre gear. Now, I am of the opinion that those going without DSP are most likely the ones with a mediocre listening experience, whether they know it or not. Just my opinion, though. I know some disagree with me and all I can say is: I've been there.

 

Michael.

[REShaman]

As an Amarra Symphony with iRC enthusiast, for critical listening, I have a simple routine when sorting out what Filter Gain setting works best when iRC is active. I prefer to be more precise than use a pink noise track to settle what sounds best (as an average). I agree with the comments about liveliness in wanting the music to be as dynamic as possible by regulating the gain to just before clipping and listen. Then turn iRC "off" and listen. Then back "on". Compare and decide. There are times when I turn iRC "off" or "on" in favor of what sounds preferable. The equipment may seduce. But the music's the thing. It is good to be able to choose.

 

Enjoy the music,

Richard

[Boris75]

Boris,

That's right. The EW mic needs 48v power to work. I use an RME HDspe Pcie soundcard to convert from Analog to Digital. Because the RME card is for the pro market, there's no built in mic-pre. If you use an AtoD that doesn't have a built in mic pre, like the RME, you will also need to find a good mic pre with 48v power. I selected the Grace M101. The m101 is well known in the pro world for it's neutrality. I think every unit comes with it's own phase and frequency response measurements in the the manual. It's really nice and I highly recommend it.

 

Obviously, I take things to the extreme and it's really not necessary to spend that much on the test gear to get good filters. I've just found a slight improvement with the better test gear; YMMV. If you ever become a DSP-addict, like me, and want to do speaker correction (like I did with the DEQX) the mic really matters and the earthworks is an excellent choice. I've got remarkable results with it. :-)

 

Michael.

 

Many thanks for taking the time to write such a detailed reply. For now I have ordered a calibrated UMIK-1 from Cross Spectrum labs, but if I am pleased with my first experiments with Dirac, I may be tempted to continue with the more sophisticated equipment along the lines that you mention.

[skipperdude]

Dirac Live Room Correction Suite

 

Dirac RCS | Dirac Research

 

[ATTACH=CONFIG]9601[/ATTACH]

 

Advantages:

 

Very user-friendly and easy to implement.

 

You only need a calibrated USB mic.

 

Apparently uses a high-quality algorithm. Corrects frequency and impulse response. Very low demand placed on CPUs.

 

Separate filters for various sampling frequencies (44.1, 48, 88.2, 96 kHz and possible plans to go higher).

 

Allows easy on/off toggling for A/B testing.

 

Allows user to define a target curve and the frequency range to be corrected.

 

Free two-week trial of uncrippled software.

 

Disadvantages:

 

Forces user to use Dirac's proprietary deconvolver rather than a standard OS X AU plug-in. (Dirac sees this as an advantage).

 

Does not work with Audrivana in iTunes integrated mode (but does with Bitperfect).

 

Cost: Stereo version goes for about US$ 550.

 

One more for the disadvantage column, potentially a big one:

 

Dirac does not really "uninstall." The uninstall utility Dirac provides does NOT uninstall completely. I've confirmed this by uninstalling, reinstalling with a new username, and finding the settings from the previous user still intact.

 

As I'm unfamiliar with Mac OSX, I can't say for certain what this implies. In Windows, I'd be going into the registry to clean things out.

 

But this is, potentially, a concern: There's no question Dirac has embedded something in the OS, something that is not uninstalled when one "uninstalls" through its "uninstall utility." So the questions are...

 

1) Just is it that Dirac embeds in the operating system?

2) Why?

3) How does one get rid of it in Mac OSX?

[wgscott]

Dirac "embeds" exactly one thing, a kernel extension, in the operating system:

 

/System/Library/Extensions/DiracAudioProcessor.kext

 

The uninstaller removes it.

 

It also places user-related files in the user's Library directory, i.e.,

 

~/Library/Application\ Support/Dirac

 

~/Library/Preferences/se.dirac.appl.DiracLiveCalibrationTool.plist

 

Please don't use my thread to slander Dirac, especially when you acknowledge you don't know what you are talking about. Thanks.

 

You can remove these with the following unix command:

 

rm -rf  ~/Library/Application\ Support/Dirac   ~/Library/Preferences/se.dirac.appl.DiracLiveCalibrationTool.plist

[skipperdude]

Dirac "embeds" exactly one thing, a kernel extension, in the operating system:

 

/System/Library/Extensions/DiracAudioProcessor.kext

 

The uninstaller removes it.

 

It also places user-related files in the user's Library directory, i.e.,

 

~/Library/Application\ Support/Dirac

 

~/Library/Preferences/se.dirac.appl.DiracLiveCalibrationTool.plist

 

Please don't use my thread to slander Dirac, especially when you acknowledge you don't know what you are talking about. Thanks.

 

You can remove these with the following unix command:

 

rm -rf  ~/Library/Application\ Support/Dirac   ~/Library/Preferences/se.dirac.appl.DiracLiveCalibrationTool.plist

 

Geez, "slander" is a pretty strong term. I said I didn't know my way around the innards of the Mac operating system, so that it was not self-evident to me how one could fully remove Dirac's software. I don't recall acknowledging that I don't know what I'm talking about. (Although I will acknowledge having no idea what to do with that UNIX command you've suggested.)

 

I do think software that is not completely extracted by a functionality that is, ostensibly, an uninstaller is a big disadvantage...and, frankly, a suspiciously big one. Why is it that their "uninstaller" requires a sidebar knowledge of UNIX so as to actually uninstall it?

 

At the very least, add as a "disadvantage" that one should be familiar with UNIX, as you are, to fully uninstall the software.

 

Now, kindly refrain from slandering me? Thanks.

[wgscott]

Another logic fail.

[Yucca06]

Tried both Dirac Live and ARC2 on Mac OS, and found IK Multimedia ARC2 to be much better (for me at least), and totally integrated through AU plugin in Audirvana.

 

Don't why all this Dirac ravings here, especially at this price...

[flak]

 

Don't why all this Dirac ravings here, especially at this price...

 

What all the fuzz is about?

 

That's a good question... IK ARC2 is a quality product as it is based on Audyssey XT32, a traditional and well known minimum-phase solution.

Dirac Live is a mixed-phase solution instead which can successfully correct both the impulse and the frequency response in an arbitrarily defined listening area... normal listening rooms have a mixed-phase behaviour so that a minimum-phase solution can properly correct only at those regions of frequencies that have a minimum-phase behaviour also.

 

Too lengthy to discuss in a single post, this is a short explanation: http://diracdocs.com/Understanding%20more.pdf

this is a more detailed one: http://www.dirac.se/media/12044/on_room_correction.pdf

 

Flavio