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

[ncpilot]

First, thanks OCA for all your efforts...

I've watched most of your calibration vids multiple times trying to reinforce and understand workflow.

Question regarding using REW for measurements based on techniques in "audyssey a.r.t."., because I want to use DEQ (and get my money's worth out of my umik!)

Are these steps correct? #3 and #5 are my big questions...

1) create an ".ady" file with my receiver first with all bass and volume management off
2) get .ady into MultEq app
3) **do I turn on DEQ in app editor??** **do I get rid of the mid-frequency dip in editor??**
4) (after sending .ady file to computer) run html script to create "perfect speaker".ady
5) ** load the perfect speaker file back to receiver??** (pull into app, then send to receiver)
6) run rew to get all measurements
7) continue with steps in video

Thanks!

 

[OCA]

First, thanks OCA for all your efforts...

I've watched most of your calibration vids multiple times trying to reinforce and understand workflow.

Question regarding using REW for measurements based on techniques in "audyssey a.r.t."., because I want to use DEQ (and get my money's worth out of my umik!)

Are these steps correct? #3 and #5 are my big questions...

1) create an ".ady" file with my receiver first with all bass and volume management off
2) get .ady into MultEq app
3) **do I turn on DEQ in app editor??** **do I get rid of the mid-frequency dip in editor??**
4) (after sending .ady file to computer) run html script to create "perfect speaker".ady
5) ** load the perfect speaker file back to receiver??** (pull into app, then send to receiver)
6) run rew to get all measurements
7) continue with steps in video

Thanks!

Any ady file with responseData replaced with perfect speaker - {1,0,0,0,....} settings on all detectedChannels will do. If you will make a new ady file from scratch, ending the calibration at the end of the minimum required 3 measurements will suffice. Extract&Convert_ady script will do the perfect speaker conversion. Other settings in the ady file don't really matter at this stage. But after you transfer the converted ady file to the AVR and before making measurements with REW, make sure:

• All speakers are set to large and speaker distances are set to 0
• All speaker volumes are set to 0
• Dynamic EQ is ON and at your preferred setting
• Audyssey is on and at Reference setting
• Dynamic volume is OFF

After you complete your measurements, determine your speaker level/distance/crossover settings, enter them in the XL file and upload them to the ady file with the script. Finally, generate your filters and upload them to the ady with the UploadREWFilters2ady script. In the final ady, you can turn Dynamic EQ on if you want but you can do that in the AVR after transferring the ady as well. In the ady, there are no settings for Dynamic EQ variations anyway.

 

[ncpilot]

Perfect!!

Thanks so much!

 

[ncpilot]

Another quick question, since you seem to be monitoring this thread... 2 questions...

1 ) Can you explain your speaker abbreviations in your sweeps file? Some of them don't correspond to anything listed in the Denon manual (like SLA, SRA).

**EDIT** I figured it out by brute force, I renamed your sweeps to match the speaker abbreviations that Audyssey uses (as found in the .ady file) ***

I use Top front+rear speaker settings in my 4700H Denon (for Auro and Atmos compatibility). I seem to have used the correct files, but now not so sure.

Anyway, I'm going to evaluate my workflow again since I get a Subwoofer distance greater than +7 meters from closest speaker, and I can't get the cross-correlate align to get the sub in phase with all the other speakers (I have 2 SVS subs connected to 1 output of Denon, use the subs 3 band EQ).

2) When going through sub alignment, since they are using only 1 output of receiver, I can't dial in any distance delay needed between subs. I do have phase control with the SVS app, and the only way to get rid of a horrible dip in response curve is to invert polarity on one sub (180 degrees). Any implications of doing it that way? Alternatives?

Thanks again. I'm getting closer, but my blood pressure did rise quite a bit last night... LOL

 

[ncpilot]

**EDIT** reversed Center wiring, re-ran sweeps, and all the impulses go in the same direction now... ****

Ok, I lied, more questions.

Yesterday's Audyssey run said my center speaker was out of phase (but last time I ran it a long time ago, it said the same thing, so I reversed the wiring).

Looking at all the impulse curves, (having run REW sweeps with the Center reverse wired), it looks like the Front Left has inverted phase peaks. If I invert phase in REW, then all the peaks match in direction magnitude.

So on the basis of that, should I reverse polarity of my Front Left? Does it matter?

Just remembered, I've got 3-way F/C/L speakers with an open back behind the mid-range driver (like a tunnel, not really a port)....that's probably going to create measurement anomalies...

 

[ncpilot]

second attempt going faster and more smoothly...

After cross-correlate and vector avg of my 3 speaker measurements distilling it all down, and all the other stuff... doing the final cross-correlate still won't align the woofer...

Following thru with the IR shift in the spreadsheet, the subwoofer is more than 7 meters from the closest distance...

 

[OCA]

Can you explain your speaker abbreviations in your sweeps file?

All these speaker abbreviations are directly from the actual ady file Json and need to be kept as is. SBL and SBR are Surround Back Left&Right speakers. SLA and SRA stand for Surround Left/Right A. There used to be two sets of surround speakers A&B in the days of 5.1 surround for Music and Movies in high-end Denon/Marantz receivers because each 5.1 required different positioning of surround speakers. So, there was also SLB & SRB which is nowadays extinct (I think).

the subwoofer is more than 7 meters from the closest distance

This is quite typical with subs especially if they have internal DSP circuitry adding extra delays to the line. MiniDSP introduces even longer delays if its filters are engaged. You can manually adjust SW impulse peak to other speakers by measuring the distance between impulse peak with CTRL+mouse right click method. Aim for an earlier impulse peak of the sub if the main peak requires more than 6m adjustment. With two subs, you need to also set up relative distance (delay) and volume (trim) settings between subs in the ady file. Audyssey applies only 1 correction for the summation response of both subs and keeps it in SW1 custom curve points. You can find more info in this thread:

How to manually time align multiple subs to acoustic timing reference speaker?

Hi, I'm following OCA's tutorial on manual Audyssey calibration, and I've stumbled upon an issue regarding how to time align multiple subs to the acoustic timing reference speaker. I have two subwoofers (SW1 & SW2), which I have time aligned to each other using the alignment tool. SW2 needs a...

www.avnirvana.com

 

[ncpilot]

Hah... I already saw that thread when searching earlier today...

I have both subs hooked up to a single sub-out on my receiver with a Y adapter, originally played with room placement to get a half decent looking raw REW curve. I was under the impression based on research that full system alignment would be easier with one combined sub compared to two separate ones. Then both the receiver and REW would see "only" 1 subwoofer.

My Audyssey file only has SW1...

Should I be using both sub-outs?

So, when I do the final cross-correlate align, I should leave out the subwoofer? Then manually align to earlier peak with the other speakers?

Not sure how to do that, is their a function in REW to do that, or just do manual measurement and put IR shift in the spreadsheet column?

 

[OCA]

Should I be using both sub-outs?

Yes, you can improve the combined bass response a lot by just adding a delay to one of the subs.

So, when I do the final cross-correlate align, I should leave out the subwoofer? Then manually align to earlier peak with the other speakers?

Correct.

is their a function in REW to do that

Yes, CTRL+mouse click method I mentioned above. You can see it being done in many of my video tutorials.

 

[ncpilot]

Thanks for your replies...

Revisiting data I think I've got a bigger problem. In my final align SPL, then cross-correlate align impulse, I noticed this (before and after for just front 3 speakers):

I think because of the large initial spike for Center channel, REW is getting "confused", even though maximum impulse is around negative 100 for all three lines, it's not aligning the way I would expect.

So if I invert a speaker (I think I inverted the Left), then align, it seems to be better, even though REW is not aligning the maximum impulse peaks.

I looked at another set of measurements I took earlier, and the Center channel has that large initial peak before the max impulse.

My speakers have an open "duct" behind the mid-range (in a 3-way system), and I suspect I'm going to have greater issues with reflected sound?

Does the impulse of Center channel imply that the second stronger impulse is reflected? (and the initial one is the direct?)

 

[ncpilot]

As a followup, I'm finding the cross-correlate align will randomly align to the non-dominant peak for either the Right or Left channel. Haven't figured that out yet.

However, in a flash of consciousness in my 69 y.o. brain....... I relized I had originally tried to optimize subwoofer placement based on the best looking combined SPL curve, but now understand (I think more correctly), I should be placing them to get better phase alignment.

After doing that, and running some more traces, it seems I can pretty easily align all other speakers to the sub, now I'm getting about 3 meters... but I've got to run another set of more careful measurements with the new sub placement...

 

[Mattknz]

I started with the json editor today. When I get to the subwoofer there is no response data. Is there any way to fix that or do I need to do another set of 3 measurements?

 

[OCA]

I started with the json editor today. When I get to the subwoofer there is no response data. Is there any way to fix that or do I need to do another set of 3 measurements?

That's the second sub. Audyssey measures it and aligns it to the first one in relative distance (delay) and relative volume (trim) then keeps the total response of SW1+SW2 in SW1 and also uses only one set of filters for the total response in SW1. So, what you see is not an error. If you switch to "directional bass" then it will EQ SW2 separately but is only useful in stereo only systems. In some models, SW1 and SW2 will both be named SWMIX in the ady.

 

[Mattknz]

That's the second sub. Audyssey measures it and aligns it to the first one in relative distance (delay) and relative volume (trim) then keeps the total response of SW1+SW2 in SW1 and also uses only one set of filters for the total response in SW1. So, what you see is not an error. If you switch to "directional bass" then it will EQ SW2 separately but is only useful in stereo only systems. In some models, SW1 and SW2 will both be named SWMIX in the ady.

I don't have the option of directional bass. What do I do now, just leave the subs response data blank ?

 

[OCA]

I don't have the option of directional bass. What do I do now, just leave the subs response data blank ?

What exactly do you intend to do in the Json editor?

 

[Mattknz]

What exactly do you intend to do in the Json editor?

I am following the video and am at the point of replacing audysseys response data with the 0's from the excel spreadsheet you provided. I have done FL, FR, C, SR, SL with no problems. Now I am trying to do the SW but there is no response data to replace.
In the video you state that you have to replace the response data for every channel including the subwoofer.
There is no data to replace for the subwoofer.
I want to enter the 0's in the response data for the subwoofer but can't.
Do I need to run Audyssey again in order to obtain response data for the subwoofer so I can replace it with 0's or is there another way I can do it?

 

[OCA]

I am following the video and am at the point of replacing audysseys response data with the 0's from the excel spreadsheet you provided. I have done FL, FR, C, SR, SL with no problems. Now I am trying to do the SW but there is no response data to replace.
In the video you state that you have to replace the response data for every channel including the subwoofer.
There is no data to replace for the subwoofer.
I want to enter the 0's in the response data for the subwoofer but can't.
Do I need to run Audyssey again in order to obtain response data for the subwoofer so I can replace it with 0's or is there another way I can do it?

I see, if you have two subs, second sub will not have response data. If it's a single sub and there's no response data, that's weird. You're following an older video btw, Audyssey ART video has links to scripts that automate the whole process. If you run into problems, share your ady file with me and I will have a look.

 

[Mattknz]

I see, if you have two subs, second sub will not have response data. If it's a single sub and there's no response data, that's weird. You're following an older video btw, Audyssey ART video has links to scripts that automate the whole process. If you run into problems, share your ady file with me and I will have a look.

I'll have a look at the ART video. Thanks

 

[Mattknz]

Well, I'm starting to feel like a little silly. I am stuck again on what appears to be a simple step.
I am now trying to follow the 'ART video'.
I am using the Multieq Editor app.
I have put the 3 position .ady file from the app into the Extract and Convert thing (for the editor app) and it has created an .ady _PerfectSpeakers zip file.
When I try to open this I get 'cant open file.....as archive'
I've messed around with the properties settings for this file it didn't make any difference.
I can view all the data in notepad.
Is there a specific app I need to be opening it with?

 

[OCA]

Well, I'm starting to feel like a little silly. I am stuck again on what appears to be a simple step.
I am now trying to follow the 'ART video'.
I am using the Multieq Editor app.
I have put the 3 position .ady file from the app into the Extract and Convert thing (for the editor app) and it has created an .ady _PerfectSpeakers zip file.
When I try to open this I get 'cant open file.....as archive'
I've messed around with the properties settings for this file it didn't make any difference.
I can view all the data in notepad.
Is there a specific app I need to be opening it with?

Open REW, double click on the zip file, select all the .txt files that show up and drop them over to REW with your mouse. REW will open them all as measurements.