Dirac room correction software, compatible with Audirvana?

[TommyCat]

After i switched from Audirvana/Amarra to PureMusic I haven't had any issues with Dirac and Yosemite. Let me check this the next few days, but sounds promising

 

kk

 

Prior to switching, any chance you were experiencing static or, worse, no audio issues?

[livelistenlearn]

Thanks for the updates, TommyCat + KnockKnock. sounds like it may be better to wait this one out a bit longer.

 

cheers.

[Artifex]

Unfortunately, Dirac is still unusable for me, due to the crackling and popping noises with Audirvana and JRMC etc. I really wish I had held off from buying it until they fixed the problem.

[flak]

Hello Artifex,

 

did you try the fixes for the Yosemite Audio Bug suggested in this thread at posts 201 and 203?

http://www.computeraudiophile.com/f11-software/official-os-x-yosemite-thread-22030/index9.html#post368663

 

Flavio

[Artifex]

Hi Flávio,

 

Thank you for your message and suggestion! No, I have not tried this yet. Unfortunately I'm now traveling and won't be able to try this for a couple of weeks, but will let you know once I do so! Thanks again for highlighting this potential workaround!

[LBob]

FWIW I have gotten the same problems with Pure Music. It's a strange problem though. It can disappear for a while only to return.

[LBob]

Hello Artifex,

 

did you try the fixes for the Yosemite Audio Bug suggested in this thread at posts 201 and 203?

http://www.computeraudiophile.com/f11-software/official-os-x-yosemite-thread-22030/index9.html#post368663

 

Flavio

 

I tried it today and it did not work. When I started up the computer things seemed to load sluggishly/slowly. Dirac took several seconds to appear. When I tried to play a 44k file, the dac showed it as 96. Tried restarting the program but that didn't help. Used the command to undo the previous change and things went back to normal. Perhaps the 10.10.02 is a bit different than the original?

[occamsrazor]

Hi, Well I finally got around to trying the Dirac Live trial today, after receiving a UMIK-1 microphone a few weeks ago. My system: Mac Mini > Audiolab MDAC > Hypex Ncore monoblocks > ATC SCM-11 speakers. The first time I tried it the effect was awful. Even during the testing sweeps I could hear what sounded like a blown woofer on my left channel. So I tried a second set of measurements again carefully placing the mic each time. The effect was the same. Also the curve seemed like it was doing a huge amount of correction...

 

 

With now over an hour spent on it I was rather disappointed. And then I realised that my amps were in fact still plugged into my old broken speakers which I use only for parties and had connected over the weekend, speakers that I don't mind getting trashed and which do indeed have a blown driver. Oops!!!

 

OK, time to start from scratch again with the ATCs... This time the curve looked much better... I didn't try anything except the default target.

 

 

Impressions: Well first off it worked with iTunes but when used with Audirvana (in iTunes integrated mode) I got a horrible crackling whenever I played a song. Then I tried Flavio's Yosemite fix mentioned in post #104 which did indeed fix the problem:

 

http://www.computeraudiophile.com/f23-dsp-room-correction-and-multi-channel-audio/dirac-room-correction-software-compatible-audirvana-15480/index5.html#post398915

 

I've tried various Audirvana configurations. Integer mode seems useable but I had to turn off "Direct mode" and uncheck "deactivate completely iTunes own playback".

 

Regarding sound quality... it's somewhat hard to judge and my feeling of it varies between different songs. Overall it does seem to increase the soundstage and add a clarity to vocals, I guess that's the increased treble the optimized curve is applying. It's reducing the bass a bit too. On some songs the effect is noticeable and definitely an improvement - vocals are clearer, bass is reduced but less "boom" on some parts of songs - on other songs less so. There's one song where I actually prefer it without but I'd say the majority do sound better with Dirac on.

 

Is it worth it? I'm reserving my judgement until I've had more time to play with it. I do find the price rather steep for what it is. If it was 50-100 dollars I'd definitely get it, but at around $450 for non-EU buyers that's a not-unsignificant amount. I'm always hesitant to pay that much for software when you never know if it'll get broken by some system upgrade etc. Also I'm keen to see the upcoming AU plugin version of Dirac, as it seems a simpler solution to using Dirac with Audirvana, and will likely hold off on buying Dirac Live until I see what kind of licensing/price and quality the AU plugin is.

 

PS - I'd be very appreciative if any Dirac experts can comment on the second curve posted above, the one with the ATC speakers. I'm just interested to know how to read it, and whether it reveals anything interesting, or not. Thanks

[dtb300]

Regarding sound quality... it's somewhat hard to judge and my feeling of it varies between different songs. Overall it does seem to increase the soundstage and add a clarity to vocals, I guess that's the increased treble the optimized curve is applying. It's reducing the bass a bit too. On some songs the effect is noticeable and definitely an improvement - vocals are clearer, bass is reduced but less "boom" on some parts of songs - on other songs less so. There's one song where I actually prefer it without but I'd say the majority do sound better with Dirac on.

 

Play around with the curve. Double-Click on the curve to create a new "point" and move/drag it up or down .5dB-1dB. If you want more low end, play around in the area 20-500. If you want less vocals, play around 1k-6k range. Do it in small increments, save the curve to a new name, then activate it in DAP.

 

It is like a equalizer but does it across a range of frequencies. And with my testing I have found it sounds better than what JRiver offers using their Parametric EQ or regular EQ (I am a Win2012, JRiver user).

 

I am still in testing mode myself and not sure if I will stay with Dirac or not.

[LBob]

Regarding sound quality... it's somewhat hard to judge and my feeling of it varies between different songs. Overall it does seem to increase the soundstage and add a clarity to vocals, I guess that's the increased treble the optimized curve is applying. It's reducing the bass a bit too. On some songs the effect is noticeable and definitely an improvement - vocals are clearer, bass is reduced but less "boom" on some parts of songs - on other songs less so. There's one song where I actually prefer it without but I'd say the majority do sound better with Dirac on.

 

Is it worth it? I'm reserving my judgement until I've had more time to play with it. I do find the price rather steep for what it is. If it was 50-100 dollars I'd definitely get it, but at around $450 for non-EU buyers that's a not-unsignificant amount. I'm always hesitant to pay that much for software when you never know if it'll get broken by some system upgrade etc. Also I'm keen to see the upcoming AU plugin version of Dirac, as it seems a simpler solution to using Dirac with Audirvana, and will likely hold off on buying Dirac Live until I see what kind of licensing/price and quality the AU plugin is.

 

 

I thought it was reducing bass too, but what is actually happening is that a rather large bump is being reduced. I had the same initial impression, but after a month or so I realized that I wasn't losing bass. In fact, I found that the bass was much more detailed. I had been losing a lot of information because of my bass bumps and the resultant bloating. You might also notice that you are suddenly hearing more midrange detail because of reducing the bass bump. Try a few more projects. In my case I tried the "couch" option first and was not really wild about it, but when I tried "chair" it knocked my socks off. Playing with the target curves can also help. You might also find that Dirac grows on you.

 

One question: I tried the fix in the Yosemite thread and it didn't work at all. What version of Yosemite are you using?

[occamsrazor]

In my case I tried the "couch" option first and was not really wild about it, but when I tried "chair" it knocked my socks off. Playing with the target curves can also help. You might also find that Dirac grows on you.

 

Thanks, I'll definitely give the "chair" option a try. I'm still keeping an open mind, and as I say I do feel there is a positive effect, I'm just not sure that - for me - it's significant enough for the price.

 

One question: I tried the fix in the Yosemite thread and it didn't work at all. What version of Yosemite are you using?

 

I've heard some people talk about "dropouts". I wouldn't say I ever had that. What i had was a very nasty and continuous ragged-sounding crackling. Hard to describe but it sounded like the speaker was totally blown and I couldn't listen for more than a second or two, was even scared it might damage my speaker. I should add this was on my "good" speakers. The terminal command fix removed that problem entirely. My system is running OSX 10.10.2 and Audirvana 2.0.8 and is connected to my Audiolab MDAC via asychronous USB. The MDAC's maximum rate over USB is 24bit/96khz. I only ever use A+ in iTunes-integrated mode.

 

When you enter the command (best to copy/paste it) a warning came up about issuing sudo commands and required me to enter my mac login password, I'm assuming you did that too?

 

PS - I used this command:

 

sudo nvram boot-args="dart=0x0"

 

...not the one that supposedly coexists with Trim Enabler:

 

sudo nvram boot-args="dart=0x0 kext-dev-mode=1"

 

...as I'm not running TRIM on my SSD anyway. I don't know if that matters but thought I'd mention it.

 

What I do notice is that using Dirac with Audirvana sometimes leads to some weird artifacts or delays whenever I change the volume in iTunes now, whereas it didn't before when using only Audirvana. However this only occurs for a slit-second during the change, so is quite liveable with.

[SteveS1]

I'm lucky that the version within Amarra seems to work fine and always has done. So what follows assumes you can get it working with your set up.

 

Looking at your second graph, that is a nasty bass hump. As you get more familiar and smarter with the use of Dirac it is possibly to take a single measurement - go sraight to filter design and just screen shot that measurement. Then come out. The reason I did that a few times was just to check whether I could minimise the more obvious issues by adjusting speaker or seating placement. You need plenty of patience but I considered it a worthwhile afternoon and although my physical adjustments were small - they did improve things. Then I did my measurements very carefully and not getting carried away with the height differences - a few cm above or below the initial one. The room is reasonable size and I made two projects one chair and one sofa. The results for me has been a clear preference for the sofa filter which would now have to be torn from my cold dead fingers.

 

In my case I don't have any large humps but definitely a significant difference between the speaker sitting on a heavily furnished side and the other sitting in open space. This has an impact of the image, pushing it a little to one side. I also have a bit of mid bass suck out, although not a lot and a gentle rise or two in the upper mid that could become a bit distracting and caused a bit of smearing.

 

The iRC has resolved all of this and the sound is really excellent. Warmer, smoother but with better clarity. Bass is now extended and even, the image is central and deeper with appropriate recordings. I am very happy and it is sobering to switch the filter out. Amazing what your ears get used to - so some adjustment does take place as not using iRC sounds worse than I remember it.

 

With your rather dramatic bass hump, perhaps create some more points on your curve and (very gently) follow the shape of your original trace using much lower values than the trace of course and save the new target curve, then optimise it and create an alternative filter allowing a little more of that hump into the sound. That would be where I would look if I couldn't get a default curve how I wanted it having got used to it. I haven't been able to improve on my default and am very happy with it.

 

Hope some of this helps, I think it is worth you persevering. I've helped a couple of friends put this into their systems and they are delighted. I have absolutely no connection with the product or any Hi Fi or audio trade btw, other than as a customer.

 

Steve

[occamsrazor]

Hi Steve,

 

Looking at your second graph, that is a nasty bass hump.

 

I'm very new to understanding these graphs... Are you talking about the bit around 40-200hz? Or the spike from around 25-30hz?

 

With your rather dramatic bass hump, perhaps create some more points on your curve and (very gently) follow the shape of your original trace using much lower values than the trace of course and save the new target curve, then optimise it and create an alternative filter allowing a little more of that hump into the sound.

 

I'm curious what is the theory behind following the original curve somewhat. Isn't the point to negate the original curve? I saw Liam posted a similar suggestion in post #38 of this other thread:

 

http://www.computeraudiophile.com/f23-dsp-room-correction-and-multi-channel-audio/drc-digital-room-correction-it-poor-relation-room-treatments-21239/index2.html#post401170

 

Is that what you're suggesting?

 

Thanks again, your advice is much appreciated....

 

PS - Can you save your measurements so you don't have to do them again each time, but can just load them, apply a different target curve, optimise, and save to a new filter?

[SteveS1]

@ the hump.

 

Yes the 40-200hz, the mid to higher bass is likely to be much more of an issue untreated I would have thought.

 

@ following your trace.

 

Yes that's right, but I would limit it to a very gentle 'nod' toward that shape at a much lower level.

 

@ saving your measurements.

 

Assuming it works just like the IRC in Amarra - you save your measurements as a 'project'. You can then load the project at anytime. I never bother saving target curves as I have problems loading them. But if you adjust the target curve, then optimise it, you can then save the new optimised curve as a new filter.

 

Take a screen shot to remind you of what you did. To be honest, if you get your speakers and listening position optimised then create your project(s) you can play to your hearts content. But beware of creating spikes or moving the target too much - that's unlikely to help.

 

A filter may take a little getting used to (just as moving your gear to a different room would). If you are used to some significant room colour it can take a little adjustment. For example, if you have a bit of overblown bass - a flatter response can seem a little bass light until you realise that it's just more even and accurate bass instead of boosted at some frequecies.

 

Likewise a rising upper mid or top end could have been interpreted as 'lively' and a more natural balance appear a little dull at first. All good fun.

[LBob]

Occamsrazor,

 

I used the exact procedure (password and scary warning etc) and the sudo nvram boot-args="dart=0x0" command. Strange. I wonder if it could be disk related. You apparently have a newer mac mini than I do if you are using a SSD. I have a 2011 with a traditional spinning hard drive.

[occamsrazor]

I used the exact procedure (password and scary warning etc) and the sudo nvram boot-args="dart=0x0" command. Strange. I wonder if it could be disk related. You apparently have a newer mac mini than I do if you are using a SSD. I have a 2011 with a traditional spinning hard drive.

 

Mine is an "Early 2009" model, so older than yours. I replaced the hard disk with an SSD myself. Maybe the issue will reappear, I've only been playing on it a few hours since I did the command. But it had an immediate effect when I did it....

[occamsrazor]

Some more observations....

 

The crackling is back! Sorry to have given some false hopes, it seems the crackling is back. What is weird is I can't seem to work out the cause, or rather make it repeatable. This is what happened...

 

1. Load Dirac processor, load Audirvana and iTunes.

2. On the first song get immediate crackling, same with subsequent songs.

3. I thought maybe the nvram fix had got lost on a previous reboot, so I did it again and rebooted.

4. Repeat (1) - no more crackling.

 

At this point I thought that the solution was good, but maybe just wasn't surviving a reboot, but I now think that's wrong because....

 

5. Listen to about two dozen songs all perfect, and then the last one suddenly there is crackling again and for subsequent songs.

6. Changed nothing at all except rebooted, loaded the apps, and started playing the same song - no crackling.

 

What does this mean? Absolutely no idea to be frank, but the non-repeatability of both the crackling and the "solution" implies to me that the problem is a complex one.

 

Anyway, back to the sound....

 

Taking the advice of others who replied, I re-did the measurements and then tried to create a curve that "followed" the bass hump a bit:

 

 

Immediately I felt this was better than the default target I used yesterday. The vocal clarity and soundstage was still there, but without the loss of bass that had annoyed me. The one song that yesterday I felt was better without Dirac, is now better with.

 

Then I was looking at that frequency response graph. My ATC SCM-11 speakers are supposed to have a very flat response curve, as are my Ncore monoblocks, and the speakers are specced by ATC as "Frequency response (-6db): 56hz-22khz". So what I don't understand is, if the lowest bass is 56hz, what is all that stuff going on below 50hz on the graph, where is it coming from? I should add there are a bunch of soft furnishings in my room, which is 7m x 4.5m, as well as a hard wood sofa very near the speakers, and also wooden floors. I don't know if that is relevant. As an experiment I tried to sort of diminish it via a new target curve:

 

 

I can't say I can hear much difference between the two to be honest though....

 

What I notice the most doing A>B comparisons between Dirac and no Dirac is the change of soundstage, especially for vocals. It's hard to describe but without Dirac the vocals feel like they are coming from two speakers simultaneously, and with Dirac they feel feel like they are coming from a point in the middle of the speakers, if that makes sense. It's an odd effect, on some songs I'd say it's just "different" not better not worse, on others it's quite dramatically better.

 

Finally, to answer my question "Can you save your measurements so you don't have to do them again each time, but can just load them, apply a different target curve, optimise, and save to a new filter?" the answer is of course "yes". The "Save" button at the bottom-left of the screen saves the entire set of measurements to a file you can re-load later.

 

At the moment if you use Audirvana and want to continue doing so I couldn't recommend Dirac right now, the combination of the two together just seems unstable. At least until the rumored AU plugin is released, which sounds to be a much better solution. But I'm keeping an open mind about Dirac generally and want to do some more testing with different curves...

[LBob]

Some more observations....

 

The crackling is back! Sorry to have given some false hopes, it seems the crackling is back. What is weird is I can't seem to work out the cause, or rather make it repeatable. This is what happened...

 

1. Load Dirac processor, load Audirvana and iTunes.

2. On the first song get immediate crackling, same with subsequent songs.

3. I thought maybe the nvram fix had got lost on a previous reboot, so I did it again and rebooted.

4. Repeat (1) - no more crackling.

 

At this point I thought that the solution was good, but maybe just wasn't surviving a reboot, but I now think that's wrong because....

 

5. Listen to about two dozen songs all perfect, and then the last one suddenly there is crackling again and for subsequent songs.

6. Changed nothing at all except rebooted, loaded the apps, and started playing the same song - no crackling.

 

What does this mean? Absolutely no idea to be frank, but the non-repeatability of both the crackling and the "solution" implies to me that the problem is a complex one.

 

 

I've had the same experience that you have had with the crackling, but it's not just with Audirvana +, it's with all the players I use. It's a Yosemite/Dirac thing I believe, as my problems only started after switching to Yosemite. For me all I need to do is hit pause for a few seconds and it straightens itself out.

[occamsrazor]

I've had the same experience that you have had with the crackling, but it's not just with Audirvana +, it's with all the players I use. It's a Yosemite/Dirac thing I believe, as my problems only started after switching to Yosemite. For me all I need to do is hit pause for a few seconds and it straightens itself out.

 

For me even quitting A+, iTunes, and Dirac and restarting them all didn't always resolve the issue, but a reboot usually did. Pausing or changing songs, once the problem occurred, didn't help. Although that was all with A+ and Dirac, not tried other players.

[il Carletto]

I had the noise issue since I moved from an iMac 2011 to a Mac Mini late 2014 and Yosemite.

Then a friend of mine gave me his Empirical Off-Ramp 5 to try. Bingo, a big SQ improvement with its I2s connection to my PS Audio DirectStream and the noise issue's disappeared.

After one month I bought an Audiobyte Hydra X+ to use as usb>I2s bridge, hopeful it works as the OR5 did.

The noise is back... :-(

[Hailey]

Gosh I just tried Dirac with UMIK-1, etc., and I'm afraid I may have destroyed my system. The sound was just gone. Now I've uninstalled Dirac and still nothing. Agh, why did I ever monkey around with something that was working just fine?

 

Does anyone have any experience in Dirac-reversal? I've uninstalled, but the damage appears to have been done...somewhere. Operating system reinstall, then? (I'm running Yosemite). Thanks...

[Liam]

Gosh I just tried Dirac with UMIK-1, etc., and I'm afraid I may have destroyed my system. The sound was just gone. Now I've uninstalled Dirac and still nothing. Agh, why did I ever monkey around with something that was working just fine?

 

Does anyone have any experience in Dirac-reversal? I've uninstalled, but the damage appears to have been done...somewhere. Operating system reinstall, then? (I'm running Yosemite). Thanks...

 

Have you checked Audio MidI to make sure the correct inputs and outputs are selected? Did you make the correct selections when you installed Dirac?

[Artifex]

Hello Artifex,

 

did you try the fixes for the Yosemite Audio Bug suggested in this thread at posts 201 and 203?

http://www.computeraudiophile.com/f11-software/official-os-x-yosemite-thread-22030/index9.html#post368663

 

Flavio

 

Hi Flávio,

 

I'm finally back home and tried this fix. Sadly it doesn't work for me. Still lots of popping and crackling sounds with Audirvana if I use Dirac.

 

To be honest I'm pretty unhappy that I paid over $600 for a piece of software that doesn't work, but thanks for your suggestions.

[Hailey]

Have you checked Audio MidI to make sure the correct inputs and outputs are selected? Did you make the correct selections when you installed Dirac?

 

Yes did check that. As for the correct selections during installation...well, I endeavored to. Perhaps I erred.

 

In any event, ended up doing a clean reinstall of the Mac OS. That took care of the problem.

[KnockKnock]

My crackling noise is just gone. I made a playlist with the songs that did result in static noise, now playing this playlist for more then an hour without ONE click or anything.

 

I have tried with optical out, USB to Dynaudio Hub, USB (M2Tech HiFace2), bothwired and wireless to the Dynaudio Focus 200 , and no problem anymore. Don't know why, but I am happy now