Dirac > Audyssey XT32 ... Sure, Always?

[GalZohar]

Seems like the Multeq-X PEQ is just an additional way to make editing the target curve easier. Basically, it adds a bump with Q and amplitude similar to a PEQ filter, but in the end it's just a target curve modification. With enough (lots of) dedication you could probably achieve the same with the app. It's still not actually going to apply a PEQ filter. If you had a bump that Audyssey couldn't fix, adding a PEQ at that frequency to compensate will most likely not do anything useful.

Seems like MultEQ-X is basically making the measurement process much nicer as well as the target curve editing much easier. If you don't need either of those, you probably won't get any real benefit from this very pricy application, and should just get the 20$ phone app. Using Dynamic EQ and limiting the correction frequency while keeping the curve flat - I don't see what I could further improve with the app.

Although with MultEQ-X it would be faster to re-do the calibration as I could easily delete/re-measure bad measurements due to outside noise etc. With current app you have to re-do all speakers for the last measurement point, and can't change anything once done. But if you don't re-calibrate very often this might not be worth the price premium either.

 

[peng]

Seems like the Multeq-X PEQ is just an additional way to make editing the target curve easier. Basically, it adds a bump with Q and amplitude similar to a PEQ filter, but in the end it's just a target curve modification. With enough (lots of) dedication you could probably achieve the same with the app. It's still not actually going to apply a PEQ filter. If you had a bump that Audyssey couldn't fix, adding a PEQ at that frequency to compensate will most likely not do anything useful.

Seems like MultEQ-X is basically making the measurement process much nicer as well as the target curve editing much easier. If you don't need either of those, you probably won't get any real benefit from this very pricy application, and should just get the 20$ phone app. Using Dynamic EQ and limiting the correction frequency while keeping the curve flat - I don't see what I could further improve with the app.

Although with MultEQ-X it would be faster to re-do the calibration as I could easily delete/re-measure bad measurements due to outside noise etc. With current app you have to re-do all speakers for the last measurement point, and can't change anything once done. But if you don't re-calibrate very often this might not be worth the price premium either.

In my opinion, those fake PEQ that MultEQ-X has is most likely better than the real PEQ because the modified filter set will still be FIR, that's better than IIR inherently. I have been doing it with just the App and to me it is much more effective than the minidsp+REW for the 20-300 Hz range.

 

[GalZohar]

In my opinion, those fake PEQ that MultEQ-X has is most likely better than the real PEQ because the modified filter set will still be FIR, that's better than IIR inherently. I have been doing it with just the App and to me it is much more effective than the minidsp+REW for the 20-300 Hz range.

But if you had a dip that Audyssey couldn't flatten out to your flat target curve, it probably still won't flatten it if you add a bump in the target curve. You have to first understand why there is still a bump/dip in the response post-audyssey before thinking you can correct it with a PEQ adjustment to the target curve.

 

[hemiutut]

In my opinion, those fake PEQ that MultEQ-X has is most likely better than the real PEQ because the modified filter set will still be FIR, that's better than IIR inherently. I have been doing it with just the App and to me it is much more effective than the minidsp+REW for the 20-300 Hz range.

Doing the equalization for minidsp as you do the filters with REW ?
Automatically or manually?
If you do them in automatic I tell you that you have much room for improvement than if you do it manually and even more if you then confirm it using the RTA of REW.

Written with translator

Greetings

 

[raistlin65]

But if you had a dip that Audyssey couldn't flatten out to your flat target curve, it probably still won't flatten it if you add a bump in the target curve. You have to first understand why there is still a bump/dip in the response post-audyssey before thinking you can correct it with a PEQ adjustment to the target curve.

Is it that it can't? Are they choose not to have Audyssey do that?

I had always thought it was by design. That Audyssey was configured not to boost dips to avoid pushing the speakers and/or the receiver's amp to distortion during dynamic peaks at the Audyssey calibrated reference volume.

Is there something about FIR filters that prevents correcting dips?

 

[GalZohar]

Is it that it can't? Are they choose not to have Audyssey do that?

I had always thought it was by design. That Audyssey was configured not to boost dips to avoid pushing the speakers and/or the receiver's amp to distortion during dynamic peaks at the Audyssey calibrated reference volume.

Is there something about FIR filters that prevents correcting dips?

If it limits the boost, then no mater how how you set the target curve at that frequency, that limit will still apply. That's what I'm trying to say. Can't/Won't doesn't matter for this matter. That goes for anything that Audyssey didn't correct the first time around - It probably won't do any better after you add PEQ trying to correct something you didn't like in the first run.

If you get any correction via modifying target curve that flattens the end result which wasn't flattened when you used a flat curve, I would love to understand what prevented it from being flat in the first place that was bypassed with the modified target curve.

 

[raistlin65]

If it limits the boost, then no mater how how you set the target curve at that frequency, that limit will still apply. That's what I'm trying to say. Can't/Won't doesn't matter for this matter. That goes for anything that Audyssey didn't correct the first time around - It probably won't do any better after you add PEQ trying to correct something you didn't like in the first run.

If you get any correction via modifying target curve that flattens the end result which wasn't flattened when you used a flat curve, I would love to understand what prevented it from being flat in the first place that was bypassed with the modified target curve.

If I'm reading this correctly, it sounds like to me this is all speculation on your part.

Why don't we let some of the people who are actually using the PEQ tell us if it's not possible to correct dips?

 

[Newman]

better than the real PEQ because the modified filter set will still be FIR, that's better than IIR inherently

FIR>IIR is not a given, it seems, for audio correction. Both seem to have their proponents, among the experts.

 

[peng]

FIR>IIR is not a given, it seems, for audio correction. Both seem to have their proponents, among the experts.

Agreed, and thank you for pointing this out. I am talking about in general and in theory as it applies to audio REQ applications, overall I believe if one must choose between the two, FIR seems better for audio room eq assuming processing power is not the limit and implementation is not the issue. Of course this is a lot of assumptions lol.., and you are right about both have their proponents among experts. As an EE I may be naturally biased so I am more interested in related articles by engineers such as Siemens and not those by other engineers or scientists such as those from Dirac, Audyssey, or Anthem. From what I read in that article, if processing power is not the limitation, I would pick FIR, and I don't think the Siemens engineers would be biased in favor of one over the other as Siemens don't really have a horse/audio in the race, they are just engineers who knows about filter design for various applications, not hifi audio/req specific.., but ultimately I rely mostly on the results I was able to achieve with Audyssey.

In practice, FIR and IIR filters have important performance differences as summarized in Figure 8.

Figure 8: Summary of differences between IIR and FIR filters.

An IIR filter has an advantage that for a similar filter roll off as a FIR, a lower order or number of terms can be used. This means that less computations are needed to achieve the same result, making the IIR faster computationally. However, an IIR has nonlinear phase and stability issues. It is a bit like the fable of the tortoise and the hare. The FIR filter is like the tortoise in the race – slow and steady, and always finishes. The hare is like the IIR filter – very fast, but sometimes crashes and does not complete the race.

Siemens DISW

There is also the very short comparison on the minidsp website and they don't really have a product (as far as I am aware of) that uses FIR filters.

FIR vs IIR filtering

In this application note, we will explain the difference between FIR ("finite impulse response") and IIR ("infinite impulse response") filtering.

www.minidsp.com

With the continuously improvement processors become available, it seems that FIR based REQ software will pull ahead anywhere, if not yet. That's just based on my own understanding of the two, having read as many articles on both, including some comparisons.

Dirac claims to have been using both and implemented their so called "mix phase" approach, vs Audyssey's minimum phase.

https://www.diymobileaudio.com/threads/how-dirac-live-works-a-mix-phase-filtering-iir-fir-approach-to-room-correction.421271/

 

[peng]

Doing the equalization for minidsp as you do the filters with REW ?
Automatically or manually?
If you do them in automatic I tell you that you have much room for improvement than if you do it manually and even more if you then confirm it using the RTA of REW.

Written with translator

Greetings

It was only after spending many hours trying every trick I knew at the time, auto, manual (using the REW's own equalizers), before I removed all the interconnects and put the USD225 2XHD back in the box. Using it by itself was not as good as using it with Audyssey but in the end the 0.5 or may be 1 dB better at some frequency points was not worth all the extra heat, cabling and potentially clipping due to the 2 V limit.

Put it this way, I am getting FR flat from 20-200 Hz within +/- 1 to 1.5 dB, 1/12 smoothing, without messing up impulse response, phase etc., why would I bother adding another device in the loop? Again, how much can you improve to from +/- 1 to 1.5 dB, 20-200 Hz, even just 20-130 Hz using the minidsp, would you mind sharing your REW plots? That's just my opinion/experience obviously. I may take my 2XHD out of box when I am ready to spend another USD 200 to try Dirac in my two channel system. Until then, if someone offers me $100 I would sell it.

Last edited: 5 minutes ago

 

[Flak]

Dirac claims to have been using both and implemented their so called "mix phase" approach, vs Audyssey's minimum phase.

That's correct and we use both... you may find this document interesting even if not recent:

https://www.dirac.com/wp-content/uploads/2021/09/On-equalization-filters.pdf

 

[hemiutut]

It was only after spending many hours trying every trick I knew at the time, auto, manual (using the REW's own equalizers), before I removed all the interconnects and put the USD225 2XHD back in the box. Using it by itself was not as good as using it with Audyssey but in the end the 0.5 or may be 1 dB better at some frequency points was not worth all the extra heat, cabling and potentially clipping due to the 2 V limit.

Put it this way, I am getting FR flat from 20-200 Hz within +/- 1 to 1.5 dB, 1/12 smoothing, without messing up impulse response, phase etc., why would I bother adding another device in the loop? Again, how much can you improve to from +/- 1 to 1.5 dB, 20-200 Hz, even just 20-130 Hz using the minidsp, would you mind sharing your REW plots? That's just my opinion/experience obviously. I may take my 2XHD out of box when I am ready to spend another USD 200 to try Dirac in my two channel system. Until then, if someone offers me $100 I would sell it.

Last edited: 5 minutes ago

I am not going to put measurements with REW because I do not want polemics.
You realize that among the engineers themselves they do not agree on how to
to open the pandora's box and that jump for example the defenders of equalization only up to 300-500 Hz, those who defend the entire frequency range or the same creators of room correction programs.

But yes, where there is a manual, powerful and customizable equalization, I'm not going to put it, I don't want any controversy.
all automatic calibration.
I do recommend that anyone who does automatic or manual EQ, then look at what happens with the RTA of REW for example.
I have no doubt about that, but you have to know how to apply such equalization and it takes much more time to optimize a system.
Let alone if we want to do it manually in multichannel vs a 2.0 or 2.1.

Regards

 

[ad_fletch]

Makes me wonder why Yamaha still offers "Manual YPAO" PEQ then. Surely after using YPAO, particularly the multipoint RSC variant, manual EQ tweaks would suffer from the same measurement mismatch issue described above...?

Having said that, my inner tweaker would like to have the option without spending $1.2k on an X3700H + $200 for MultEQ-X. Nb no Denon discounts in my country.

 

[peng]

Makes me wonder why Yamaha still offers "Manual YPAO" PEQ then. Surely after using YPAO, particularly the multipoint RSC variant, manual EQ tweaks would suffer from the same measurement mismatch issue described above...?

Having said that, my inner tweaker would like to have the option without spending $1.2k on an X3700H + $200 for MultEQ-X. Nb no Denon discounts in my country.

You don't have to spend $200. The App that runs on smart phones and tablets cost $20.

 

[Steve Dallas]

You don't have to spend $200. The App that runs on smart phones and tablets cost $20.

Right. And Ratbuddyssey allows you to see all your measurements and edit your target curve accordingly for free.

 

[ad_fletch]

Ah ok, somewhere earlier in thread I thought I read that PEQ input/conversion was new feature of windows app.

 

[Chromatischism]

Ah ok, somewhere earlier in thread I thought I read that PEQ input/conversion was new feature of windows app.

It is. The mobile app works based on a simple curve editor. It is the PC app that's the more beefed up of the two.

 

[GalZohar]

Ah ok, somewhere earlier in thread I thought I read that PEQ input/conversion was new feature of windows app.

The way of modifying the target curve with PEQ-style (and other styles too) parameters is a new feature of the windows app. However, you could always just modify the curve manually or with the Ratbuddyssey to achieve similar or maybe even identical results.

 

[peng]

It is. The mobile app works based on a simple curve editor. It is the PC app that's the more beefed up of the two.

Ah ok, somewhere earlier in thread I thought I read that PEQ input/conversion was new feature of windows app.

That is not quite the case especially if one uses it with Ratbuddyssey. Keep in mind that simply editor also allows you to diable/enable MRC, set EQ frequency range, save as many curves as you want, just to name a few of the very useful feature.

The window app uses an emulated PEQ format for data entry only. The resulting filter set will still be FIR based, not PEQ at all and in my opinion that is "better" anyway. With the $20 IOS/Android app, if you use Ratbuddysey in certain ways (I tried so I know it works), such as using some extra anchor points (frequencies), I am quite sure you can get similar results, that is without the ability to enter the "Q"s that the windows app has.

If you don't use Ratbuddyssey or other 3rd party software then I would say the window app is far superior, but with Rat and if you are good at it, I think the IOS/Anroid app is practically (in terms of just the EQ part, without the fancy interface that the windows app offers) just as good and it even let you tweak the left and right channel individually. Overall, I would still prefer the windows app because of the very nice interface and the useful information provided by the graphics, but only if they drop the price to $100, or $200 if the license is transferable to allow for at least one AVR/AVP upgrade, if not two, or say for within 8 years.

 

[Miker 1102]

My only real complaint with Audyssey XT32 is the default reference and flat curves. If you simply run Audyssey in a Next/Next "wizard" like function, it doesn't sound good. Audisssey's belief appears to be flat frequency response is the desired goal. This belief is flawed as people prefer a downward sloping curve. It basically sounds both bright and no base at the same time. If they would add a better downward sloping default curve (like a harmon type curve) it would make things much easier. My favorite instructions for getting Audyssey to produce good results is via this" https://docs.google.com/file/d/1YH_eNHRCxKFCwXMddi28kzXqnUwYHfrD/edit?filetype=msword

Al least I have had good results with it.

Best example I have read. I changed my curve, Limited the frequency. Audessy at base line sounds awful..to the point where I question Denon enthusiasts. They also have poor instructions.