Audyssey Manual Calibration “OCA’s REW + Audyssey Awesomeness”

[DeLub]

@OCA , Why don't you limit the frequency range to which the correction is applied? This would eliminate the need for the anti-hf correction, right? And as you rightly state: the correction should be limited to the lower frequencies anyway...

 

[ooheadsoo]

@OCA , Why don't you limit the frequency range to which the correction is applied? This would eliminate the need for the anti-hf correction, right? And as you rightly state: the correction should be limited to the lower frequencies anyway...

Pretty sure that algorithm is baked into the receiver/mobile app.

 

[OCA]

Personally, I'd -

* Absolutely 100% put the MiniDSP back in
* Put a "perfect response file" from the Supreme method on the receiver, calibrate the subs with REW+MSO as per typical MiniDSP process
* Run the Audyssey multi-point with the MiniDSP in, on, and calibrated
* Run that through A1

Very curious to see what OCA says though

With older models with just 2 sub outputs, you definetely need a MiniDSP if you are running 3 or 4 subs. You should align them with MSO using minimal time delays and no EQ (and no processing whatsoever on the subs themselves including PEQ). Then Evo will take care of the rest. If you have one of the newer models, no matter how many subs you have (unless more than 4), you don't really need a MiniDSP. Run an Audyssey calibration in directional bass mode and feed that to Evo and it's going to optimize all subs (up to 4) as good as MSO and stay within Audyssey delay limits. They will be treated as 1 sub at the end (as with a MiniDSP) and system will be converted to standard bass mode.

 

[OCA]

Is this step necessary? I own x6700h. Just ran A1 Evo Maestro and bass freq. response looks much worse than MultEQ-X result. I measured it with Umik-1 and REW. Could it be a reason?

Evo is a sound improvement tool, use Power Point for nice graphs.

 

[Mike Lima]

With older models with just 2 sub outputs, you definetely need a MiniDSP if you are running 3 or 4 subs. You should align them with MSO using minimal time delays and no EQ (and no processing whatsoever on the subs themselves including PEQ). Then Evo will take care of the rest. If you have one of the newer models, no matter how many subs you have (unless more than 4), you don't really need a MiniDSP. Run an Audyssey calibration in directional bass mode and feed that to Evo and it's going to optimize all subs (up to 4) as good as MSO and stay within Audyssey delay limits. They will be treated as 1 sub at the end (as with a MiniDSP) and system will be converted to standard bass mode.

OCA, the particular amp in my bedroom is even older and more basic. It is a Denon AVR-S740H. Only has 1 subwoofer output.

 

[PHD]

Is this step necessary? I own x6700h. Just ran A1 Evo Maestro and bass freq. response looks much worse than MultEQ-X result. I measured it with Umik-1 and REW. Could it be a reason?

What do you mean by worse? Take REW measurements of the front + subs obtained using the app and EVO and post them here for direct comparison. Better use psychoacoustic smoothing, because this represents human perception.

 

[Mike Lima]

With older models with just 2 sub outputs, you definetely need a MiniDSP if you are running 3 or 4 subs. You should align them with MSO using minimal time delays and no EQ (and no processing whatsoever on the subs themselves including PEQ). Then Evo will take care of the rest. If you have one of the newer models, no matter how many subs you have (unless more than 4), you don't really need a MiniDSP. Run an Audyssey calibration in directional bass mode and feed that to Evo and it's going to optimize all subs (up to 4) as good as MSO and stay within Audyssey delay limits. They will be treated as 1 sub at the end (as with a MiniDSP) and system will be converted to standard bass mode.

I have never used MSO. One listening position (MLP) 4 subs. So should I do MSO or REW alignment (with my MiniDSP)? Then run Audyssey calibration, then run EVO.

 

[Dovydas]

What do you mean by worse? Take REW measurements of the front + subs obtained using the app and EVO and post them here for direct comparison. Better use psychoacoustic smoothing, because this represents human perception.

Here's what I have in mind. Blue graph is A1 Evo and yellow is MultEQ-X. Sorry, I did not save measurements and did only 1/48 smoothing. I have some issues with fronts and subs integrations with MultEQ-X as there's a huge dip in XO region. Wanted to see if A1 Evo solves it, but got issues elsewhere. You can see that the difference below 55 Hz is quite big. People are very happy with A1 results, so I'm pretty sure I did something wrong.

 

[PHD]

Here's what I have in mind. Blue graph is A1 Evo and yellow is MultEQ-X. Sorry, I did not save measurements and did only 1/48 smoothing. I have some issues with fronts and subs integrations with MultEQ-X as there's a huge dip in XO region. Wanted to see if A1 Evo solves it, but got issues elsewhere. You can see that the difference below 55 Hz is quite big. People are very happy with A1 results, so I'm pretty sure I did something wrong.

Uninstall REW and reinstall the latest beta:

https://roomeqwizard.com/installers/REW_windows-x64_5_40_beta_51-api.exe

Install the EVO Electrified wizard app:

Releases · pulento/A1Evo_Electrified

Electron port of A1Evo. Contribute to pulento/A1Evo_Electrified development by creating an account on GitHub.

github.com

Uninstall your Google play Multi-EQ app and manually install the modified latest version instead:

https://www.avsforum.com/attachments/com-dmholdings-audysseymulteq-1-11-0-1-apk-txt.3650699/

Now use the modified app to take new measurements. Take 3 at the same mic location (MLP). Then use EVO Electrified. Increase frequency range to 10000Hz and maxboost + overall maxboost to +6dB, run and generate the master .adt and load it to AVR.


Now take REW measurements in with AVR in stereo mode (REW must generate mono test tons for L+R+sub) with DEQ off, and repost the results. They should look better than both of the results you posted here.

 

[OCA]

Here's what I have in mind. Blue graph is A1 Evo and yellow is MultEQ-X. Sorry, I did not save measurements and did only 1/48 smoothing. I have some issues with fronts and subs integrations with MultEQ-X as there's a huge dip in XO region. Wanted to see if A1 Evo solves it, but got issues elsewhere. You can see that the difference below 55 Hz is quite big. People are very happy with A1 results, so I'm pretty sure I did something wrong.

There isn't enough detail in your graphs to tell for sure but those dips don't seem to be audible. Trying to boost them may have quite audible distortion effects though.

 

[OCA]

I have never used MSO. One listening position (MLP) 4 subs. So should I do MSO or REW alignment (with my MiniDSP)? Then run Audyssey calibration, then run EVO.

To properly align 4 subs manually is a bit tough. You can align 2 at a time in REW. MSO is free and quite easy to use.

 

[Omartian]

Hi guys-

i initially used REW and minidsp to time align and set crossovers using REW's alignment tool for my 7.4.4 setup. I've been using that and pretty happy with the results. when i came across the post for a1 evo on reddit, it seemed like a good idea to try out.

I went into my minidsp and turned off all of the eq, i then ran the multieq app to run the Maestro MJ script and have loaded the master .ady file back to my receiver. I like the sound especially from the center channel.

I had a few questions.

1. i've seen in other setup vids for the multieq app that you want to adjust filter frequency range from 0-300 for speakers and then turn it off completely for the sub. right now my sub is set to 0-250hz, and speakers are from 0-20000hz. Leave these alone?

2. my crossovers on my optimized master.ady seem pretty close to the values i obtained when measuring w/REW's alignment tool. The exception was center which was set to 60 hz (REW alignment tool said 110 hz) and my atmos speakers at 80 hz (Alignment tool measured best at 120). I have 4 beefy subs (2 HSU 15 VTF-3 MK5 and 2 SVS pb-2000). Should i bump the crossover for the center and atmos up? the faq says to trust the crossovers, but wanted to make sure that these levels seem too low.

3. I have my home theater in the corner of the basement, the master.ady would be the way to go compared to the dynamiceq.ady, correct?

4. My subwoofer volume (+3 db) and distance (24.8 ft) seem off, but not really sure (before it was 0 db and 17 ft).

5. Do you recommend high frequency roll off 1 (selected by a1 evo) or roll off 2?

Sorry if that's a lot and looking forward to your responses.

 

[OCA]

You're on the right track because you're in search for the best sound from your system. Evo will get you 80% of the way with minimal effort. Use it to teach good sound to your ears. It takes time. My best filters have sounded similar to other good ones intitally and it took me days to understand they were more special than I initially thought. Give your hard earned filters ample listening time before you skip to the next optimal procedure. Remember there's always better.

 

[PHD]

@OCA

With default settings, EVO reported a dip removal efficiency of about 94.5% (max).

I've used the EVO Electrified and checked force large fronts and increased boost and an overall boost to +6dB and now the dip removal efficiency has increased to 97.6%

Is this improvement valid/legit? I mean, when calculating efficiency, does EVO account for a possible dip boost, which would only make things worse?

Also, If the better efficiency was due to setting fronts to large, why didn't EVO do it automatically as part of the trial-and-error phase (it did say fronts are not large enough but still set their XO to 60Hz, if I remember correctly)?

Thanks.

 

[OCA]

@OCA

With default settings, EVO reported a dip removal efficiency of about 94.5% (max).

I've used the EVO Electrified and checked force large fronts and increased boost and an overall boost to +6dB and now the dip removal efficiency has increased to 97.6%

Is this improvement valid/legit? I mean, when calculating efficiency, does EVO account for a possible dip boost, which would only make things worse?

Also, If the better efficiency was due to setting fronts to large, why didn't EVO do it automatically as part of the trial-and-error phase (it did say fronts are not large enough but still set their XO to 60Hz, if I remember correctly)?

Thanks.

The improvement is most possibly just graphical and will sound worse. Evo calculates how the response tracks the target and boosted dips will improve that but it does not boost dips, it's not even a screen option.

For solely speaker pritectuin purposes, Large lfe+main option is automatically considered only if front speakers display a bass roll off of 40hz or lower. You can oorionally force it though.

 

[PHD]

The improvement is most possibly just graphical and will sound worse. Evo calculates how the response tracks the target and boosted dips will improve that but it does not boost dips, it's not even a screen option.

For solely speaker pritectuin purposes, Large lfe+main option is automatically considered only if front speakers display a bass roll off of 40hz or lower. You can oorionally force it though.

Thanks. I tried to run again but this time I only used force large and left boosts at 0dB. The efficiency did improve but not by much (95.5%). So part of the improvement was due to the boost or overall boost. Do you also don't recommend to change only the overall boost? If I understand, allowing the overall boost to be set higher than 0dB could only allow to increase the channel levels, not the filters, isn't that correct? Or does the overall boost refer to the total boost by all filters?

 

[OCA]

Thanks. I tried to run again but this time I only used force large and left boosts at 0dB. The efficiency did improve but not by much (95.5%). So part of the improvement was due to the boost or overall boost. Do you also don't recommend to change only the overall boost? If I understand, allowing the overall boost to be set higher than 0dB could only allow to increase the channel levels, not the filters, isn't that correct? Or does the overall boost refer to the total boost by all filters?

Dips are local wave cancellations, boosting will increase the volume of both waves equally and usually only increase distortion.

 

[PHD]

@OCA


Is there a workaround for this error:

1:09:15 AM [IMPORTANT!] Available bass system volume adjustment range → { -10.5dB : 12dB }

1:09:16 AM [IMPORTANT!] SRA - applied adjustment: -2.5dB

1:19:42 AM [INFORMATION] -2dB Audyssey auto-leveling compensation needed for speaker SRA.
1:19:43 AM [INFORMATION] -8dB Audyssey auto-leveling compensation needed for speaker SRA for 'DEQ' ady.
1:19:43 AM [WARNING!] Speaker SRA could not be fully compensated for DEQ auto-leveling!


My SLA/SRA are high-sensitivity speakers and close to the MLP.

What's with the first info about the adjustment range being limited to -10.5 instead of -12dB ?

Then, I see that -2.5dB is needed for SRA, and later -2dB for the master, and for the DEQ auto leveling it says -8dB are needed (-2.5-8 = -10.5dB)

And then I get a warning about SRA not being fully compensated for DEQ although.

The final level for the SRA was set to -12dB. How do I calculate how many dBs are needed below -12dB ? If I know, then I can simply add this amount manually to all channels except the SRA. This better be done by EVO even if it causes deviation from reference level. Better than having un-leveled setup.

Thanks

OK. I figured this out. It's accumulative (-2.5-2-8=-12) so I'm missing 0.5dB for SRA. So I can just add 0.5dB to all the rest channels excluding SRA? I still think EVO should do it automatically because some may overlook this warning.

 

[PHD]

I think I found a bug with the DEQ settings on my older X4400H. The RHL/RHR channels remain +6dB higher than the rest in the DEQ ady. I've tested using Atmos test signals and SPL meter.

 

[Stinga]

I just wish Prof Peng would run this thing and compare this to his well tweaked rat file.
This would say alot for dumb buggers like me.