My first attempt at Dirac this morning (on the 3800)

[Dougey_Jones]

The information is accurate. Dirac applies phase correction and Audyssey does not. Frequency and phase response are both in the frequency domain. It's the impulse which is in time domain. You need to learn the basics before making such assumptions.

How big of a deal in the lack of phase correction in real world use?

 

[patoulol]

As mentioned already it doesn't. They leave it up to the brand that has implemented Dirac on their hardware to do that. Some devices, such as the Monoprice HTP1 do but I'm not aware of any others that do.

It's a shame because the DynEQ is a real asset for me. This will be a real obstacle to the DIRAC passage

 

[Dj7675]

It's a shame because the DynEQ is a real asset for me. This will be a real obstacle to the DIRAC passage

It is but there are ways around it or to mitigate it ...
1- calibrate your system to sound good at what your volume you normally listen at. When you are listening at this level no changes are needed
2-presets- all dirac based processors allow a fixed number of presets. You can create one specific to another volume level
3-tone controls-Needing to listen at a lower volume, bump up the bass on a tone control. A movie sounding a little bright adjust the treble. I use this method on my processor (Stormaudio) and quite like it. I have 4 adjustments... brightness, treble, bass, sub level as well as center channel level via web based remote. It does everything I need when I need to.
When I used DEQ I had a love hate relationship with it. I liked how it adjusted bass based on your listening level, but did not like the boosted surround levels so usually left it off. But loudness implementation is something (as well as tone controls) that many Dirac based processors lack. Stormaudio and Trinnov do not have them either. Trinnov doesn't even have the option of tone controls at all.

 

[tjcinnamon]

It's a shame because the DynEQ is a real asset for me. This will be a real obstacle to the DIRAC passage

Same! I love it. I’m gonna buy both because ART will make it a very competitive match.

 

[Dougey_Jones]

It is but there are ways around it or to mitigate it ...
1- calibrate your system to sound good at what your volume you normally listen at. When you are listening at this level no changes are needed
2-presets- all dirac based processors allow a fixed number of presets. You can create one specific to another volume level
3-tone controls-Needing to listen at a lower volume, bump up the bass on a tone control. A movie sounding a little bright adjust the treble. I use this method on my processor (Stormaudio) and quite like it. I have 4 adjustments... brightness, treble, bass, sub level as well as center channel level via web based remote. It does everything I need when I need to.
When I used DEQ I had a love hate relationship with it. I liked how it adjusted bass based on your listening level, but did not like the boosted surround levels so usually left it off. But loudness implementation is something (as well as tone controls) that many Dirac based processors lack. Stormaudio and Trinnov do not have them either. Trinnov doesn't even have the option of tone controls at all.

Agree 100%, it’s a nice feature but the boosted surround levels are intolerable. I feel like they cater heavily to people new to home theater who want as much whiz bang surround channel presence as they can get.

 

[patoulol]

It is but there are ways around it or to mitigate it ...
1- calibrate your system to sound good at what your volume you normally listen at. When you are listening at this level no changes are needed
2-presets- all dirac based processors allow a fixed number of presets. You can create one specific to another volume level
3-tone controls-Needing to listen at a lower volume, bump up the bass on a tone control. A movie sounding a little bright adjust the treble. I use this method on my processor (Stormaudio) and quite like it. I have 4 adjustments... brightness, treble, bass, sub level as well as center channel level via web based remote. It does everything I need when I need to.
When I used DEQ I had a love hate relationship with it. I liked how it adjusted bass based on your listening level, but did not like the boosted surround levels so usually left it off. But loudness implementation is something (as well as tone controls) that many Dirac based processors lack. Stormaudio and Trinnov do not have them either. Trinnov doesn't even have the option of tone controls at all.

for the surround level with DEQ I lowered the levels (-2 db SR SL ) and it's going well.

Your proposal with the stormaudio is attractive but seems complicated to me if I have to adapt it.

my buddy has a Trinnov and did 6 different curves to make it look like a Denon DEQ but it didn't convince him

 

[Dj7675]

for the surround level with DEQ I lowered the levels (-2 db SR SL ) and it's going well.

Your proposal with the stormaudio is attractive but seems complicated to me if I have to adapt it.

my buddy has a Trinnov and did 6 different curves to make it look like a Denon DEQ but it didn't convince him

I probably made it sound more complicated using the storm tone controls... as simple as bass seems lacking, turn up the bass.. Sometimes I will get to an old movie and it will be very harsh/bright so turn the treble down.

 

[OCA]

I didn't have to learn it, I asked Audyssey about it and they gave me the answer. If you are going to insist on your claim, you need to prove it somehow, otherwise you can just say anything you want. Or may be you know more about how they did their filters than them? In that case I would have to wonder why, no sarcasm, am series. I never said anything about impulse response, and you assume I have to learn from you to know that is in the time domain? How good are you with Fournier, Lapace, Z transforms etc.? I do know you are a good programmer, and I appreciate your work. I also never said Audyssey does phase optimization that Dirac does.

I always thought someone seemingly very knowlegeable like you are, would be more open minded and am really sory to see you reacted this way. Telling others to "learn the basics before....", how assumptious! And again, it's not relevant anywhere as I did not make things up myself, I made it clear I was told by Audyssey. May be you can ask them too.

If your rude response is in reaction to the tone of my post, then be clear, I did not anticipate it would offend you. If so, let me apologize right now, and we can move on.

I said Audyssey doesn't apply phase correction while Dirac does and you blamed me for "perpetuating misinformation". What Audyssey told you is completely irrelevant to my statement! I never claimed Dirac does phase correction because Dirac filters change phase response. None of these filters change phase in any significant way. Dirac applies phase correction to fix problems "between the speakers" stemming from asymmetric room reflections, this is actually why phase correction is done if ever and Audyssey simply cannot do that. Their narrative that they don't do phase correction because their filters don't change phase is bit of an abuse of lack of knowledge at best and you didn't help others learn the facts which was my only intention.

Fournier, Lapace

They're spelled FOURIER and LAPLACE and as per your accusation that I don't know about them, attached platform independent (works in any web browser) code I wrote a while back applies Fast Fourier Transformation to impulse responses (in .wav format) and graphs bode diagrams. It doesn't use any existing JS libraries except for graphing tools - there aren't any that can read wave files anyway - and was written from scratch. For what it's worth, I am not a programmer (far from it) but I have an aeronautical engineering degree and we used Laplace transforms quite heavily for solving differential equations in computational fluid dynamics some decades back in school.

I am not posting any of these here to counter argue you or anyone else and please don't take it personal but I usually only post to share information with others who have a desire to learn. The algorithm (I used a simplified version of Radix 2) is much easier to grasp than one would imagine.

 

[OCA]

Should we expect a low-end DIRAC ART?

I guess so. It's well known that even regular Dirac Live is quite different in quality between different versions of MiniDSPs.

 

[Vacceo]

Agree 100%, it’s a nice feature but the boosted surround levels are intolerable. I feel like they cater heavily to people new to home theater who want as much whiz bang surround channel presence as they can get.

For that scenario, play music on surround mode. Esentially that duplicates or triplicates stereo and bam, you have all your channels working without any further equalization.

 

[Dougey_Jones]

I guess so. It's well known that even regular Dirac Live is quite different in quality even between different versions of MiniDSPs.

Can you elaborate on this statement?

Why would Dirac’s quality vary between different hardware implementations? I assumed (maybe stupidly) that Dirac would have been reviewing and certifying that whatever hardware a manufacturer is shipping meets their minimum requirements to run whatever version of their software is being licensed.

 

[Steve Dallas]

Can you elaborate on this statement?

Why would Dirac’s quality vary between different hardware implementations? I assumed (maybe stupidly) that Dirac would have been reviewing and certifying that whatever hardware a manufacturer is shipping meets their minimum requirements to run whatever version of their software is being licensed.

Perhaps your assumption is correct, but Dirac's minimum requirements are lower than imagined?

I can tell you I get much better results running Dirac Standalone on my silent PC than I do running Dirac on my Denon 4800. I assume it is due to the limited number of taps available on the 4800 vs. a virtually unlimited number on the PC.

 

[OCA]

Can you elaborate on this statement?

Why would Dirac’s quality vary between different hardware implementations? I assumed (maybe stupidly) that Dirac would have been reviewing and certifying that whatever hardware a manufacturer is shipping meets their minimum requirements to run whatever version of their software is being licensed.

It's the same algorithm, difference is in filter resolution and is based on the device in question's technical capability (mainly the DSP chip used in it). AFAIK, they change sampling rate (48/96kHz) and even change filter types from FIR to IIR or a hybrid of both depending on the DSP unit.

 

[Dougey_Jones]

Perhaps your assumption is correct, but Dirac's minimum requirements are lower than imagined?

I can tell you I get much better results running Dirac Standalone on my silent PC than I do running Dirac on my Denon 4800. I assume it is due to the limited number of taps available on the 4800 vs. a virtually unlimited number on the PC.

If this were the case, and the taps were only needed on the device receiving the data and making the calculations, wouldn’t that mean that anybody could buy the Audyssey MultEQ-X software, run it to establish the optimal curve and then upload it to a lower tier receiver that only had MultEQ or MultEQ XT but not XT32?

I thought the taps needed to be present to physically implement the corrections but you’re making it sound like that isn’t true.

 

[Steve Dallas]

If this were the case, and the taps were only needed on the device receiving the data and making the calculations, wouldn’t that mean that anybody could buy the Audyssey MultEQ-X software, run it to establish the optimal curve and then upload it to a lower tier receiver that only had MultEQ or MultEQ XT but not XT32?

I thought the taps needed to be present to physically implement the corrections but you’re making it sound like that isn’t true.

I think we have a miscommunication.

When I run Dirac on my PC, there is a standalone processor that provides virtual DSP as a virtual Windows audio device. The result of the calibration in Dirac Live is pushed to that processor as if it is a hardware DSP device. As it uses the PC's CPU instead of a hardware DSP, it can utilize a virtually unlimited number of taps, as it is only limited by CPU capacity and whatever limits the developers put on it for consuming CPU.

On my little silent PC, the Dirac Processor uses 13% of my CPU when sound is playing.

On an AVR or similar, it uses a portion of the onboard DSP chip, and its taps are limited by the capacity of that chip.

Does that help?

 

[Dougey_Jones]

I think we have a miscommunication.

When I run Dirac on my PC, there is a standalone processor that provides virtual DSP as a virtual Windows audio device. The result of the calibration in Dirac Live is pushed to that processor as if it is a hardware DSP device. As it uses the PC's CPU instead of a hardware DSP, it can utilize a virtually unlimited number of taps, as it is only limited by CPU capacity and whatever limits the developers put on it for consuming CPU.

On my little silent PC, the Dirac Processor uses 13% of my CPU when sound is playing.

View attachment 342453

View attachment 342454

On an AVR or similar, it uses a portion of the onboard DSP chip, and its taps are limited by the capacity of that chip.

Does that help?

We’re not having a miscommunication, reread my post. I also run Dirac on an i5 PC and then export to a MiniDSP Flex Balanced, but my comments still apply. If the corrections are calculated on a separate device, then you’re saying that it requires less hardware to implement or run those corrections and I’m just not sure that I agree with that.

I haven’t read that anywhere from either Dirac or Audyssey but it does make intuitive sense. I know somebody asked this same question about doing calcs in Audyssey MultEQ-X and then exporting to a MultEQ or MultEQ XT device and the response was “no, that won’t work”.

 

[Steve Dallas]

We’re not having a miscommunication, reread my post. I also run Dirac on an i5 PC and then export to a MiniDSP Flex Balanced, but my comments still apply. If the corrections are calculated on a separate device, then you’re saying that it requires less hardware to implement or run those corrections and I’m just not sure that I agree with that.

I haven’t read that anywhere from either Dirac or Audyssey but it does make intuitive sense. I know somebody asked this same question about doing calcs in Audyssey MultEQ-X and then exporting to a MultEQ or MultEQ XT device and the response was “no, that won’t work”.

The Dirac Live software requires you to choose a device before you take any measurements. It knows the DSP capabilities of that device before it calculates its filters and creates them accordingly.

I have not used MultiEQ-X and do not know how it works, but it stands to reason it also knows the DSP capabilities of the target device.

Am I closer to the target now?

 

[Dougey_Jones]

The Dirac Live software requires you to choose a device before you take any measurements. It knows the DSP capabilities of that device before it calculates its filters and creates them accordingly.

I have not used MultiEQ-X and do not know how it works, but it stands to reason it also knows the DSP capabilities of the target device.

Am I closer to the target now?

I think I know what you mean now.. Like how on my RZ50, if I used Dirac just on the receiver with Onkyo mic it did 3 positions, but in the Dirac Live 3 software the minimum is 9 positions for tight focus.

 

[peng]

I said Audyssey doesn't apply phase correction while Dirac does and you blamed me for "perpetuating misinformation". What Audyssey told you is completely irrelevant to my statement! I never claimed Dirac does phase correction because Dirac filters change phase response. None of these filters change phase in any significant way. Dirac applies phase correction to fix problems "between the speakers" stemming from asymmetric room reflections, this is actually why phase correction is done if ever and Audyssey simply cannot do that. Their narrative that they don't do phase correction because their filters don't change phase is bit of an abuse of lack of knowledge at best and you didn't help others learn the facts which was my only intention.


They're spelled FOURIER and LAPLACE and as per your accusation that I don't know about them, attached platform independent (works in any web browser) code I wrote a while back applies Fast Fourier Transformation to impulse responses (in .wav format) and graphs bode diagrams. It doesn't use any existing JS libraries except for graphing tools - there aren't any that can read wave files anyway - and was written from scratch. For what it's worth, I am not a programmer (far from it) but I have an aeronautical engineering degree and we used Laplace transforms quite heavily for solving differential equations in computational fluid dynamics some decades back in school.

I am not posting any of these here to counter argue you or anyone else and please don't take it personal but I usually only post to share information with others who have a desire to learn. The algorithm (I used a simplified version of Radix 2) is much easier to grasp than one would imagine.

Come on, so I made a typo, so sorry! If you don't know about Fourier and Laplace transforms, you likely don't know enough about IIR and FIR filters either, and that's why I asked, and it is good to know you actually do know... It was just a question, not accusation so please don't be so sensitive.

Please don't twist what I said, or not said in this case, I have not posted what Audyssey told me so far, so you really shouldn't comment on something that you don't know. What I posted was just from their AskAudyssey website, that's not directed to you anyway, but for others who might be interested in reading. By the way, it's okay if you made some false statements inadvertently, just realize it and moved on instead of doubling down. I would certainly do the same, even just for making silly typo, still my mistake whether I like it or not. Lastly, whether it was hard work for you or not, I really do appreciate your recent contribution in helping those who want to do some tweaking post Audyssey. I wish I had the information a couple years ago when I spent many hours tweaking with Ratbuddyssey (you probably are familiar with that).

Typing on my little phone is tough for me, so excuse me for more typos...

 

[patoulol]

Perhaps your assumption is correct, but Dirac's minimum requirements are lower than imagined?

I can tell you I get much better results running Dirac Standalone on my silent PC than I do running Dirac on my Denon 4800. I assume it is due to the limited number of taps available on the 4800 vs. a virtually unlimited number on the PC.

you need to make sure it's not a psychological preference on your part, we're here on ASR.


Have you done an ABX comparison?