Obsessive Compulsive Audiophile "OCA" Corrections - Categorization Tools Assistance

[OCA]

The .ady file is essentially a JSON data file. It stores all the FIR filters for every channel, among other settings. These filters are loaded and processed by the receiver exactly as they appear in the file. The real limitation comes from the total number of samples available in each filter impulse response.

For example, the basic EQ speaker filters only have 128 taps which is easy to verify directly from the .ady file. At a sampling rate of 48 kHz, the lowest frequency a 128-tap filter can meaningfully correct is:

48,000 Hz / 128 = 375 Hz.

That’s just fundamental DSP math.

The surprising part is that the units also include internal mechanisms capable of running FIR filters at lower internal sampling rates, such as 6 kHz, effectively giving the filter 8× more usable resolution. Because of this, my initial assumption was that the basic EQ was processing those 128-tap filters at 6 kH just like the subwoofer filters allowing correction up to 3 kHz, which is honestly fine for most EQ tasks. Evo was even designed that way initially as I was certain that should be the method (I do not have basic or XT to test it myself)

But this is where things got confusing.

Users with XT or basic EQ systems eventually measured the post-calibration responses and noticed that the filters weren’t behaving as expected. After a lot of testing, it turned out the filters were not being downsampled. They were being processed at the full 48 kHz, which cripples their low-frequency capability. That means:
• EQ (128 taps): correction limited to ~375 Hz
• XT (512 taps): correction limited to ~93 Hz

Given what the hardware is clearly capable of, it’s frustrating to see them being left like that. Whether it’s oversight or deliberate is beyond me. Having said that, it's possible (although highly unlikely) that about 10 XT users all measured their systems wrong multiple times or a firmware update might silently fix it at some point especially now that people are hearing about it. So, people with these units should keep testing IMO.

 

[funkera91]

Maybe it’s a good idea to post here in this thread as well. I’ve set up my speakers as follows: FL/FR — Linton 85 in an isosceles triangle, with the center — EVO 4.C in the middle, and surrounds — EVO 4.1. The subwoofer is an XTZ 12.17 (on the right). My receiver is a Denon AVR X3600. I performed a calibration using the A1 Evo Express, and this is the file that opened in REW with the graphs before and after. I’m attaching it here → A1 EVO EXPRESS - REW - will be valid till 05.12.2025. I was looking at the graph in REW, but even without knowledge, I see that something is wrong.

The mic was Audyssey mic from the X3600h and I have used a photocamera stand, placed on my listening position, on the height of my ears, pointing upwards. I did 8 times measurements. In the end I dont know if it looks like it should be..

 

[kawauso]

Maybe it’s a good idea to post here in this thread as well. I’ve set up my speakers as follows: FL/FR — Linton 85 in an isosceles triangle, with the center — EVO 4.C in the middle, and surrounds — EVO 4.1. The subwoofer is an XTZ 12.17 (on the right). My receiver is a Denon AVR X3600. I performed a calibration using the A1 Evo Express, and this is the file that opened in REW with the graphs before and after. I’m attaching it here → A1 EVO EXPRESS - REW - will be valid till 05.12.2025. I was looking at the graph in REW, but even without knowledge, I see that something is wrong.

The mic was Audyssey mic from the X3600h and I have used a photocamera stand, placed on my listening position, on the height of my ears, pointing upwards. I did 8 times measurements. In the end I dont know if it looks like it should be..

In my rather unusual setup, the A1 Evo Express didn’t work well, but since funkera91’s system is much more typical, I really hope it works properly for them.
https://www.audiosciencereview.com/...ds/denon-avr-x3800h-review.38574/post-2456076

Denon AVR-X3800H Review

If your auto-EQ can easily handle environments with this many variables—such as vintage speakers or DIY passive subs—then it truly has the potential to become an excellent system. Based on my systems in two rooms and one desktop, DLBC is better than Audyssey XT32+App overall, but Audyssey...

www.audiosciencereview.com

 

[OCA]

Maybe it’s a good idea to post here in this thread as well. I’ve set up my speakers as follows: FL/FR — Linton 85 in an isosceles triangle, with the center — EVO 4.C in the middle, and surrounds — EVO 4.1. The subwoofer is an XTZ 12.17 (on the right). My receiver is a Denon AVR X3600. I performed a calibration using the A1 Evo Express, and this is the file that opened in REW with the graphs before and after. I’m attaching it here → A1 EVO EXPRESS - REW - will be valid till 05.12.2025. I was looking at the graph in REW, but even without knowledge, I see that something is wrong.

The mic was Audyssey mic from the X3600h and I have used a photocamera stand, placed on my listening position, on the height of my ears, pointing upwards. I did 8 times measurements. In the end I dont know if it looks like it should be..

I had to postpone the new release after pushing hard for it all day so I could look at this...
All of the midbass to midrange from both speaker is being heavily absorbed by some half soft, large surface area stuff (wardrobe(s)?) or did you do some kind of room "treatment"? Whatever it is you did in the front side of the room, is not working. C speaker is not correctly tilted toward MLP. The rest is as it should be. Should sound quite good except for lack of mid-bass.

 

[kawauso]

I had to postpone the new release after pushing hard for it all day so I could look at this...
All of the midbass to midrange from both speaker is being heavily absorbed by some half soft, large surface area stuff (wardrobe(s)?) or did you do some kind of room "treatment"? Whatever it is you did in the front side of the room, is not working. C speaker is not correctly tilted toward MLP. The rest is as it should be. Should sound quite good except for lack of mid-bass.

In funkera91’s data, it looks like the 100–200 Hz range is being left a bit too untouched.
You often say that correcting above 200 Hz makes things worse, but with Dirac, I feel that around 300 Hz is fine in many rooms.
Since the position of the “curtain” that sounds comfortable varies from room to room, I make adjustments around that area.

 

[funkera91]

I had to postpone the new release after pushing hard for it all day so I could look at this...
All of the midbass to midrange from both speaker is being heavily absorbed by some half soft, large surface area stuff (wardrobe(s)?) or did you do some kind of room "treatment"? Whatever it is you did in the front side of the room, is not working. C speaker is not correctly tilted toward MLP. The rest is as it should be. Should sound quite good except for lack of mid-bass.

So , Iit

I had to postpone the new release after pushing hard for it all day so I could look at this...
All of the midbass to midrange from both speaker is being heavily absorbed by some half soft, large surface area stuff (wardrobe(s)?) or did you do some kind of room "treatment"? Whatever it is you did in the front side of the room, is not working. C speaker is not correctly tilted toward MLP. The rest is as it should be. Should sound quite good except for lack of mid-bass.

Thank You!
There is no warderobe, also no room treatment. Is it possible, the speaker to be too close to the back wall? Also the center channel is 100% centralised from my listening position also on the level of the ears. So what coukd be the problem for it?

 

[kawauso]

So , Iit

Thank You!
There is no warderobe, also no room treatment. Is it possible, the speaker to be too close to the back wall? Also the center channel is 100% centralised from my listening position also on the level of the ears. So what coukd be the problem for it?

There doesn’t seem to be any particular problem.
(The subwoofer might have a better placement option, though.)
If possible, I might choose to use a conventional bookshelf speaker as the center to improve horizontal dispersion and consistency.
Using the same Linton would naturally be a good match, and the Diamond 12.1 also has a DI close to Linton, so it should pair quite well.

 

[OCA]

So , Iit

Thank You!
There is no warderobe, also no room treatment. Is it possible, the speaker to be too close to the back wall? Also the center channel is 100% centralised from my listening position also on the level of the ears. So what coukd be the problem for it?

Interesting. You should move them somewhat. Try to avoid equal distance to side and rear walls. 50Hz peak is very manageable with EQ but an 250Hz wide output in a very important frequency band is about 6dB too weak in both speakers at their current location:

And both your FL and FR are INVERTED! You need to change their polarity.

 

[kawauso]

Interesting. You should move them somewhat. Try to avoid equal distance to side and rear walls. 50Hz peak is very manageable with EQ but an 250Hz wide output in a very important frequency band is about 6dB too weak in both speakers at their current location:

View attachment 494369

And both your FL and FR are INVERTED! You need to change their polarity.

The dip around 250 Hz might just be a compromise caused by prioritizing other parameters or by placement limitations, so you can't really judge it in isolation.
In my KEF room, even after carefully considering the speaker placement, there's still a dip around 200 Hz without EQ, and that's just how it is.

NoEQ

DLBC

 

[kawauso]

Interesting. You should move them somewhat. Try to avoid equal distance to side and rear walls. 50Hz peak is very manageable with EQ but an 250Hz wide output in a very important frequency band is about 6dB too weak in both speakers at their current location:

View attachment 494369

And both your FL and FR are INVERTED! You need to change their polarity.

If, instead of leaving that dip as it is, you shape the target curve so it gently bridges that region, I think you end up with a much more cohesive feel from the low bass up into the mid-bass. It also seems like the tonal balance between channels will line up better.

With Dirac, if you just limit the correction range with the curtains somewhere around 300 Hz and then tweak the low end with a shelf, it tends to give you that kind of result automatically.

 

[OCA]

And in REW, you can totally fix it with a single operation:

You should understand that DSP is not about drawing nice graphs. The filter above will cause massive pre-echo and will be literally un-listenable. So, would be your 300Hz shelved Dirac result, at least to me. For people out there trying to learn something from all this, focus on "different distances to adjacent walls" advice and always double/triple check your speaker polarities.

 

[kawauso]

And in REW, you can totally fix it with a single operation:

View attachment 494378

You should understand that DSP is not about drawing nice graphs. The filter above will cause massive pre-echo and will be literally un-listenable. So, would be your 300Hz shelved Dirac result, at least to me. For people out there trying to learn something from all this, focus on "different distances to adjacent walls" advice and always double/triple check your speaker polarities.

If you create a straight-line curve in REW, it ends up boosting the area around 200 Hz far too much, and of course that’s going to sound unnatural. On top of that, applying that kind of correction based on a single-point measurement is extremely risky, because seat-to-seat variation will turn even small errors into major problems.

 

[kawauso]

Moving the speakers to reduce the 250 Hz dip might improve that specific region,
but it can just as easily make other aspects of the system worse.

I’m not saying the current placement is “correct.”
Placement always involves trade-offs, and you can’t judge it from one dip alone.

Depending on the room, moving the speakers may:

• reduce the 250 Hz dip
• but worsen early reflections
• or, conversely, the reflections in the current position may be subtle or even contribute positively to presence
• disturb toe-in or directivity alignment
• break the positional coherence with the surrounds
• reduce usability or even WAF in a real living space

I also have no way of knowing how experienced @funkera91 is or how much time was spent refining the current placement.
I can’t “read a REW graph and know exactly how the system sounds,” and I don’t know their personal tonal preference.

What I can see is that the L/C/R geometry, toe-in, and center-channel height all look like the result of careful work, not random placement.
So I hesitate to draw conclusions from a single parameter like the 250 Hz dip.

My own philosophy is simple:

I put the speakers where they sound good first, and then let DSP finish the balance.

When you adjust placement based on only one metric, all the side effects I listed above become much more likely.
That’s why placement is about overall balance, not fixing one isolated dip.

 

[JoseBcn]

The truth is that the work you are doing at OCA is impressive. Thanks to you, those of us who use Audissey have greatly improved the sound in our home theaters.

You deserve all the best.

And now I am sure that with AcoustiX it will be even better and spectacular. Since it will improve what has already been improved.

Audissey should learn from you, OCA.

Will AcoustiX finally be released today?

 

[Times]

A1 Evo AcoustiX has been released about a week ago --> A1 Evo AcoustiX Automated Sound Optimization Tool

@OCA: Are setups without subwoofer supported in this release? If not, what's the most convenient workaround?

 

[Zed]

All Denon Marantz AV receiver systems with a subwoofer should go with A1 Evo Neuron.

For stereo system DSP filters (i.e you can use your PC as your music source with a convolution engine like Equalizer APO, Camilla DSP, JRiver, Roon etc.) "Mastering Digital Room Correction" is the last posted video. I linked useful ones in the spoiler below.

Spoiler: Links

Hey @OCA . Since this was written, is there an evo express equivalent for someone with a 2.1 system running camilla and roon? If not by you, is there anything else? Thanks for what you’re providing!

 

[OCA]

Hey @OCA . Since this was written, is there an evo express equivalent for someone with a 2.1 system running camilla and roon? If not by you, is there anything else? Thanks for what you’re providing!

 

[Daverz]

Here are the results I got for a pair of Buchardt S400 bookshelf speakers (psychoacoustic smoothing):

Phase with 1/1-octave smoothing:

Mag response again with var smoothing: