Dirac validation

[dallasjustice]

Great before and after posts on Dirac! I've seen similar impressive before and after results in my system under Trinnov, as measured by both the frequency and impulse response plots generated by REW.

Do you get time/distance measures and also phase measures, like in REW, even without a loopback? I thought the loopback was required to get a consistent timing reference.

Thanks Bruce!

Loopback would only be needed to correct the device's FR or to get absolute time values. There's no need to for a timing reference with any of these measurements I have posted. Yes, fuzzmeasure does give absolute time value measurements. Here's a video explaining how it works in fuzzmeasure:

 

[Roger Dressler]

Thanks for the great work on the validation measurements. Really useful.

First of all AES member JJ Johnston ("In the rain" on this forum) opened my mind to the limitations related to condenser mics (measurement mics).

Is there a link to the particular info, or can you summarize? Does it have to do with aiming of the mic?

 

[dallasjustice]

Thanks for the great work on the validation measurements. Really useful.
Is there a link to the particular info, or can you summarize? Does it have to do with aiming of the mic?

JJ Johnston goes by J_J on this forum. His background speaks for itself. Look at post #21; cites to his AES paper:
http://www.whatsbestforum.com/showt...nd-advice-on-in-room-bass&p=230448#post230448

Here's a great dialectic battle between the single measurement dudes and the multi point folks:
http://yabb.jriver.com/interact/index.php?action=printpage;topic=85631.0

I think even Mitchco would be forced to agree that JJ is an authority on this subject. Especially since Mitchco cited to him as such in his comments to his paper here:
http://www.computeraudiophile.com/c...nd-quality-comparison/comments3.html#comments

 

[Brucemck2]

Do either/both FabFilter Pro-Q or Dirac work in J River for files with resolutions greater than 26/96, or, does J River have to be set to down-sample to 88k or 96k"

Thanks

 

[rhbblb1]

Does Dirac work if you use J River as a server? (I would be feeding an MSB Transport Plus.)
Does FabFilter Pro-Q?

Do either/both work for files with resolutions greater than 26/96, or, does J River have to be set to down-sample to 88k or 96k

Yes, I use Dirac with JRiver. Also yes, for now you must down-sample to 88k or 96k.

 

[dallasjustice]

DIRAC is limited to 96k. They have been saying they will update to include faster sampling rates for a while now. I think they are close to doing it.

You could also tryout Acourate. I've been playing with it recently and got some really nice filters. The filters are at least as good as DIRAC. The problem with Acourate is that it's not nearly as simple to use as DIRAC. For me, I've been using DSP long enough now. So, I feel comfortable trying other approaches out. Btw, Acourate does up to 192k. It also can convolve offline. That way you could convert your PCM to whatever sample rates or formats you want to using Jriver as long as your server can do it. Convolving offline helps to reduce computer overhead. If you want to get really fancy with digital filters and oversampling, you could use Acourate with HQplayer. HQplayer can do some pretty advanced digital filtering and also claims to do DSD along with convolution at the same time. This way you could play any format you want, at any sample rate and still use DSP (convolution filter). I am not sure how that's possible, but Jussi is pretty accomplished and seems to know his stuff. I've never used convolution in HQplayer, but I don't think it would be very difficult to do.

DIRAC works with everything. It even works with websites, which is nice if you like to watch netflix or amazon prime movies, like I do. Acourate is very difficult to use with anything that does't have a convolution engine. I think there are ways to use it but it would take some pretty fancy audio routing skills to make all work.

Right now, I just use DIRAC. In the future, I could see a scenario where I would prefer to use both systems.

Michael.

 

[Brucemck2]

Thanks Michael.

Convolving offline may end up being my near term solution. J River says they may add the ability to use VST plugins when using MC20 as a DLNA device, which would solve a lot of my issues.

I'm trying to get as "pure" or "simple" a digital chain as I can get for my hard core two channel setup. Given my very heavily treated room and baffle wall setup, I don't need impulse etc. correction, but, I do need very high quality PEQ.

Having all the PEQ done on the server prior to sending it to the DAC, for any sample rate from 44k up to SACD ISOs, would be about as clean as I can get.

 

[mojave]

DIRAC works with everything. It even works with websites, which is nice if you like to watch netflix or amazon prime movies, like I do. Acourate is very difficult to use with anything that does't have a convolution engine. I think there are ways to use it but it would take some pretty fancy audio routing skills to make all work.

I've routed all PC audio through JRiver for about a year and a half now and use my Audiolense filters on all audio from the PC. This has been done using the JRiver WASAPI loopback feature. More recently I've used the VB-Audio ASIO Bridge which routes all audio through JRiver. JRiver has announced that it is releasing its own audio driver in version 20 that will work the same as Dirac's driver with support for any sample rates, bit depths, and channel counts.

 

[dallasjustice]

I already got my early price for 20 so that news makes me happy. I like the virtual cable but I would prefer to use something like a Jriver WDM. This will make it easier on me to use convolution for everything.

I've routed all PC audio through JRiver for about a year and a half now and use my Audiolense filters on all audio from the PC. This has been done using the JRiver WASAPI loopback feature. More recently I've used the VB-Audio ASIO Bridge which routes all audio through JRiver. JRiver has announced that it is releasing its own audio driver in version 20 that will work the same as Dirac's driver with support for any sample rates, bit depths, and channel counts.

 

[dallasjustice]

I've seen photos of your room so I don't doubt it.

However, why wouldn't want to have a much more advanced method of DSP? Acourate, Audiolense and DIRAC are all light years beyond a simple PEQ. They are probably more transparent too.

Thanks Michael.

Convolving offline may end up being my near term solution. J River says they may add the ability to use VST plugins when using MC20 as a DLNA device, which would solve a lot of my issues.

I'm trying to get as "pure" or "simple" a digital chain as I can get for my hard core two channel setup. Given my very heavily treated room and baffle wall setup, I don't need impulse etc. correction, but, I do need very high quality PEQ.

Having all the PEQ done on the server prior to sending it to the DAC, for any sample rate from 44k up to SACD ISOs, would be about as clean as I can get.

 

[dallasjustice]

Another way of say it: Impulse response and frequency response are two sides of the same coin.

 

[Brucemck2]

I know.

I'm enjoying two very different chains. One is "hard core two channel" where I go MC19 to audiophile DAC to tube preamp (and, I may even add a 300B tube buffer in between the DAC and the preamp). Here, I'm trying to do minimal digital processing, but do need to knock down two room modes and add a touch of very broad band response shaping.

The other chain is "hard core DRC" where I go MC19 (as DLNA device) to reclocker to Trinnov, all in the digital domain, and allow Trinnov to do 3D remapping, aggressive DRC, etc. before going out through the Trinnov DACs.

I like both approaches, "differently".

 

[Brucemck2]

I've routed all PC audio through JRiver for about a year and a half now and use my Audiolense filters on all audio from the PC. This has been done using the JRiver WASAPI loopback feature. More recently I've used the VB-Audio ASIO Bridge which routes all audio through JRiver. JRiver has announced that it is releasing its own audio driver in version 20 that will work the same as Dirac's driver with support for any sample rates, bit depths, and channel counts.

Does this work if using J River as a DLNA device to feed a DLNA renderer? In my case, that would be using J River to feed a modified Oppo 103, which then feeds a DAC.

 

[dallasjustice]

Does this work if using J River as a DLNA device to feed a DLNA renderer? In my case, that would be using J River to feed a modified Oppo 103, which then feeds a DAC.

There's no way to use DSP on DLNA zone in Jriver 19. I've tried with the SMS-100.

 

[edorr]

There's no way to use DSP on DLNA zone in Jriver 19. I've tried with the SMS-100.

When used as a DLNA server, all Jriver does is managing the library and sending data straight from storage location to the rendering device. There is no actual processing in Jriver taking place.

 

[dallasjustice]

When used as a DLNA server, all Jriver does is managing the library and sending data straight from storage location to the rendering device. There is no actual processing in Jriver taking place.

Exactly. I hope 20 will be different. Fingers crossed.

 

[Brucemck2]

Exactly. I hope 20 will be different. Fingers crossed.

Agreed. This would be a huge advance, and would make J River a lot more useful in a networked system.

 

[Ron Kuper]

Some nice updates:

I first tried REW on my windows server. I couldn't get a valid measurement through the DAP filter. I think the problem has something to do with the REW interface with DAP. I really have no idea why it won't work. But trust me: it won't work. I tried everything and it's impossible.

JRiver WDM driver is out and I was able to easily take REW measurements through DAP with it.
To clarify it is:
REW --> MC20 WDM Driver --> DAP --> Pre/Pro (UMC-200)

In MC20 you can easily mute / solo channels so it is pretty convenient and also one can measure with and without Bass Management in MC.

Maybe now someone could finally shed some light into "what's best" - BM before or after Dirac

DIRAC is limited to 96k. They have been saying they will update to include faster sampling rates for a while now. I think they are close to doing it.

For some time now an update from Dirac is available to allow up to 192k.


Ron

 

[dallasjustice]

That's great if it works. I am not using DIRAC anymore. I am using Acourate instead. It's been really tough to get a good measurement using REW and Acourate. Acourate has a pretty robust measurement too so it's not as important as it was with DIRAC. Maybe the WDM will help. I love using it with Spotify though. It's the biggest Jriver using grade I can remember.

Some nice updates:



JRiver WDM driver is out and I was able to easily take REW measurements through DAP with it.
To clarify it is:
REW --> MC20 WDM Driver --> DAP --> Pre/Pro (UMC-200)

In MC20 you can easily mute / solo channels so it is pretty convenient and also one can measure with and without Bass Management in MC.

Maybe now someone could finally shed some light into "what's best" - BM before or after Dirac



For some time now an update from Dirac is available to allow up to 192k.


Ron

 

[Ron Kuper]

That's great if it works. I am not using DIRAC anymore. I am using Acourate instead.

I would love to hear why have you made the switch, would you share your considerations?

Thanks,
Ron