"Official" Audyssey thread (FAQ in post #1) - Page 1739

Gary, ... it was not meant to take things off Topic.

Just to give an idea of what XT-32 was processing with my system.
Seeing sometimes helps to give a clearer understanding of what another is trying to describe in text.

Terry

I thought it was an appropriate post IMO since you gave your thoughts on Audyssey in your room and how you left the settings as is. Plus, we always love pictures

+1. Fully agree, I am also one who is a fan of pictures. Seeing is believing. Pictures tell the story, etc....

Mitch, ... just a layman on this end so I cannot answer your question.

I am of the thinking that the Audyssey XT32 did it correctly.

Can I say that I am positive, nope not at all.

But I have to in this case assume the XT32 program works better than me and what I think would be correct.

I guess my thinking on this is the following,
why run XT32 if the person is going to over-ride the program and it's settings.

But I speak only from a "Layman's" point of view and certainly not one of the Experts here on AVS.

I can tell you that the system just simply sounds Awesome !!!

I'm well satisfied with what I am hearing after the Tech ran the XT32 program.

Probably did not answer your question or concern but I did give it a shot anyway.

Terry

I kinda got jealous with Keith getting all this attention for the Audyssy FAQ's, so I decided to update the Audyssey Setup Guide.

Seriously, many thanks to jdsmoothie for giving me the nudge, and some initial ideas.

I have attached the MS Word document so we can discuss the updates. I would like to save it as the more current .docx file extension; are there any issues with that? (Edit: AVS cannot accept .docx file extensions, apparently)

The changes are annoted with red text, and I have utilized comments for areas where I know I need help and/or clarification.

If something was not changed that you think needs to be, lets discuss it.

(Edit: See more recent posts for the current revised revision )

Thanks.

Mark

+1

The problem with 'hitting the stops' on the trims, Terry, is that you have no way of knowing if Audyssey would have trimmed it back even more if it had been able to. The maximum negative trim for the satellites is -12dB. If yours is set to that, there's no way to know that it wouldn't have been -15dB if it was possible. That's why it's never a good thing to see the trims maxed out. The trim setting results from a combination of factors - speaker efficiency, amplification, room size etc. If the sub is trimmed down to the stops (-15dB for subs) it's not so much of a problem to resolve because the sub has a volume control. For the other speakers (unless they are powered 'active' monitors, it's a trickier proposition - the usual advice is to use a pad (an attenuator) in the relevant circuit. A 5dB pad might be appropriate in your case. Raise it with your installer - meanwhile, if it sounds great, don't sweat it.

Keith, I think we need to split this problem of using attenuator pads into two cases.

1. Those who have "separates" (a pre/pro and a separate power amp) can easily use this solution of attenuating the signal in order for Audyssey to set speaker trims within the available trim range, because those pads can go between the pre/pro line out and power amp line in. In other words, a) physically the pads can be placed outside the boxes and b) they are analog.

2. But those with an AVR will not have access to the pre/pro line out vs. power amp line in stage, coz in case of an AVR this stage is inside the "closed box".

Recently it was batpig who came up with the solution, but it will require some additional attention from AVR guys experiencing maxed out trims on the minus side. After Audyssey setup, playing the MV level of the internal test tones (-30 dBFS band limited pink noise outputting 75 dB at the MLP at 0 dB Master Volume setting) at, say, +5 dB and measuring with an SPL meter should allow the balancing of the relative speaker levels to match each other, i.e. in this case all speakers are to be trimmed to 80 dB at the MLP. If this is still not enough, another +5 dB up of the MV may be necessary to be able to measure SPL at the MLP at 85 dB.

This method should not throw off the Audyssey EQ filter settings in any way, yet is essential for DEQ to work properly. Note, the boo-boo in this case will be that the new Ref.Level for film material will now be at -5 dB or -10 dB MV setting instead of the commonly accepted 0 dB reading.

Hi Mark - oh yes... I'm loving it

Will read your revised guide later...

Terry has a prepro and separates. Not sure what your point is about them being analogue - so is the output to the power amp.


You can also get inline pads for speaker lines you know.


You know that the internal test tones bypass all processing, including Audyssey? The accurate way to do it after Audyssey setup would be with a test disc. You’re not going to get an accurate result using the internal tones because of any adjustments Audyssey has made.



I can see how that would work but the pads seem a heck of a lot easier. And as the internal test tones bypass Audyssey I'm not sure either that the point wrt to DEQ applies.

Good for Terry. My point for them being analog was to clarify that attenuator pads can only attenuate an analog signal. I've heard of people who were confused and wanted to to insert a pad into the HDMI flow.


I think speaker level attenuation (in other words: power level attenuation) could be a bit problematic, think of the heat build up on such attenuators. I don't think it would be a practical idea. Powering up the final signal, then attenuating a portion of that power would easily lead to global warming!!!


Can agree with you Keith, but let's see. The internal test tones are a band limited (500 Hz-2 kHz) signal, by standard the limited band means any speaker in the world should perform in this range. Meantime, Audyssey starts with level trimming prior to setting up the EQ filter for the whole frequency range, doesn't it? Please argue with this point!


I think the point of a new ref. level reading on the MV display with pads in the signal flow is a valid one, especially for DEQ to work properly for film materials.

Yes I see now why the black levels are so good.

CRT-ish

You have??! Blimey...


You must have heard of analogue volume controls, Feri They're speaker level attenuators and have been for decades...


Not sure what your point is. The internal test tones bypass Audyssey. So whether you play them before or after running Audyssey isn’t material. If Audyssey isn't in the equation, then you can use the internal tones in any way you like. But if you want to use test tones *after Audyssey* (which is what you said) then you need to use a test disc.


The easiest way for Terry to solve the maxed out trim problem is to use a pad. It has no effect at all on DEQ - it is in circuit before Audyssey goes anywhere near the signal to create a filter to modify it. It's no different to using a volume (gain) control on the power amp, if his power amps have them (like the Emotiva UPA2s that I own for example).

Keith, ... thanks as I knew you or someone really off into this would know what was up.

In my case it sounds Awesome as is so I am not going to mess with a good-thing.

Terry

Keith, volume controls are never placed between power amp out and speaker in (aka speaker level). Speaker level is too high to do proper attenuation, therefore volume control is always done at the preamp stage or at the input stage of the power amp at line level like in your case before the signal gets powered up.



We're talking about setting relative trims in order to balance speaker levels to each other. Nothing to do with Audyssey EQ'ing, IMHO. Even Yamaha YPAO can do the same. (Well, actually it does!) Turning off Audyssey after setup will not default to initial "no-setup" of trims, just the FIR filters are gonna be by-passed, but the trims and distances will remain intact. Check the menu to verify this case.



If Terry would insert the pads without rerunning Audyssey IMHO it would affect both the reference level at 0 dB and DEQ, yet in case a pad is inserted then Audyssey should be run again. Especially for DEQ, coz now its getting lower level signals that will surely throw off the equal loudness curves DEQ is gonna set for this new ref. level.

Hey guys,

Been a while since I have needed to post in here but I just recently came across a once in a lifetime (for me at least) deal on a Paradigm Sub 25 which I couldn't pass up. This means that I will be going from a dual subwoofer setup to a single sub setup.

My receiver is an Integra DTR-70.3 with XT 32. My current subs are Paradigm Monitor SUB 12's. As some of you may be aware, Paradigm subs can be individually calibrated with their PBK software (calibrated microphone, uses computer for calculations, robust). However, with a dual sub setup I was always told to skip PBK since it does not EQ both subs interacting together and just let XT 32 handle the calculations alone.

Now that I am switching to a single sub, would it be beneficial to use PBK on the subwoofer before running XT 32......this way XT 32 will have an already flat response/"perfect sub" to work with?

Thanks in advance!

I'm not sure what you are trying to tell me, Feri. Terry can use pads in between the preamp and the power amp. The only point I was making wrt to speaker level attenuators is that they are nothing unusual and have been around for decades - potentiometers. Analogue volume controls.


Of course but again I am not sure what you are telling me. The point I am trying to make is that the internal test tones BYPASS Audyssey, so there is no point in using them after running Audyssey if you want to take account of the Audyssey filters. Audyssey will have modified the pink noise when it created the filters, so using the internal test tones has no relevance to what comes out of the speakers once Audyssey has been run. You would need to use an external source for the pink noise test tones and then you would hear the tones in the real world, post-Audyssey.


]

Who said anything about not re-running Audyssey?


No - it has nothing at all to do with DEQ. If you use a pad between the preamp and power amp all you are doing is sending a lower voltage to the power amp. Audyssey hasn't even entered the equation at that point. Then when Audyssey is run, it 'hears' a lower level of signal and doesn't need to set the trims so low they max out. It doesn't affect DEQ in the least.


Of course.


Putting a pad in has no impact at all on DEQ unless you didn't bother to re-run Audyssey - but nobody has suggested that one should not re-run Audyssey. Of course Audyssey needs to be re-run - that is the whole point - you re-run Audyssey and the trims don't get maxed out to -12dB because of the pads.

I think this is covered in the FAQ, BCJ.

How do I set the controls on my subwoofer before running MultEQ? See point No 9.

I'm placing dual subwoofers in my room with XT32 as the EQ.

The room is 13'W x 15'L with a 8ft. ceiling. About 1600 cu. ft. sealed.
It is treated with heavy drapes over the windows, carpeted and five 4'x8' rigid firberglass acoustic panels. It's not a dead room or overly live.
MSPaint of the room... ignore the sub location, that is old:


The MLP is almost dead center in the middle of the room.

Using a Denon 4311ci with all 11 speakers and dual subwoofers.

Using REW and a old-style Radio Shack meter, here is a graph with both subs under the left and right speakers:


Same, but with XT32 applied:


I used the receiver's Auto-EQ to set the volume and distance before I made the measurements.

Here are the subs under the speakers, which are mounted on the wall above them:

Left:


Right:


It is rather important that I pick a good location, as in a week, I will be cutting holes in the floor for infinite baffle manifolds and be putting these in them:


And after that, there is no going back!

My next plan is to do some sweeps with the mains turned on, perhaps try a couple more sub positions. I'm actually kind of happy with those graphs.... should I be? Any other ideas?

Lots written on this . Tryputting your side sub 1/3. From front wall. You need to test nd crawl a bit

Yes, there is quite a bit of good stuff along the lines of the JBL white paper. In fact, I tried getting close to the "midpoint on the sidewalls" positioning that worked best for the JBL tests of dual subs:



Obviously, it's not that great for my room....

I'm curious to see what Audyssey does with my sub/main x-over point when I do sweeps with the mains turned on and a 80hz x-over.
I also want to see what it does with the dip around 130hz.

Have you discussed this with someone who has installed the infinite baffle manifolds? I would be concerned that even though you find a good placement for the floor-standing subs, the same position may not be good at all for infinite baffle manifolds. Like you say, once you cut the holes, you are committed. Good luck!

My inclination is to go with jmschnur's recommendation of looking at 1/3 wall locations. You don't want a symmetrical pattern such as 1/3 from the front wall on one side and 1/3 from the back wall on the other side. My inclination would be to try one sub on the front wall 1/3 from the right corner and the other sub on the left wall 1/3 of the way back from the front wall. That might provide complementary patterns from the two subs.

Front right/left corners are, I think, to be avoided since you may have people sitting not far from the rear corners. I.e., stay 1/3 of the wall away from any front corner.

I gather that your current plots are for both subs on the front wall under the mains. My guess is that what you are seeing has heavy participation by modes at the 15' room length (30' round trip) by the direct wave colliding with the reflected wave off the back wall. Those two waves are in phase at about 37 Hz and 74 Hz and out-of-phase in between. That's where your two peaks are, with a valley in between. It's late and I'm fading, but I think that's what you are seeing. If so, moving one sub off the front wall should help.

While what you have is within XT32's ability, the less challenging the "before" is, the better the "with Audyssey" will be ... or should be. Look for sub locations that present with less dominant modes.

The other thing to look for is deeper base. It appears that your test subs are being pegged with an in-room -3db point of about 25 Hz, below which Audyssey does not provide correction. If other locations indicate correction to a lower frequency, that is a good thing. Of course, your new drivers should move you in that direction without help from room modes.

Hopefully others will chime in. I have only enough understanding to be dangerous.

Yes, this method works decently for floor-mounted IB's. It's the people who cut holes high on the wall or in ceiling's who have to get creative with measuring techniques.


I will try this.


I'm kind of happy just to have that.


You are being quite generous to my current subs. I will keep a lookout for a lower -3db point.



Yes, I'm hoping four 18" drivers with 27 liters of displacement and 2000 watts will provide a slight upgrade in that area. Below 10hz is the plan.


I think you are being modest. That was one of the most coherent and knowledgeable interpretations of a REW sweep I've ever read. Thank you so much.

+1. Harrison is a great asset to this thread. I wonder where he's been hiding all these years Do you think there is any chance at all that markus is his evil twin? LOL.

FAQ UPDATED



One new question added, prompted by Harrison's possible issue with maxed out trim on satellites.


e)6. Some of my trim levels are at the limits of their adjustment (‘maxed out’). Do I need to change anything?

I should be receiving an Onkyo TX-NR809 (MultEQ XT) receiver today and I have been studying all the information about Audyssey that I can.

I noticed this in the FAQ about using two subwoofers:

"If you have the MultEQ or MultEQ XT version of Audyssey, things are more complicated. Using either version, you will get best results by connecting both subs to ONE outlet on your AVR with a Y-cord. This applies even if your unit has two sub outlets - simply use Sub Outlet A or 1. This is because MultEQ and MultEQ XT cannot EQ two subs as two subs. When you run MultEQ or MultEQ XT, you will find that Audyssey now pings both your subs as if they are one and sets the level and distance as if they are one and then goes on to EQ them as if they are one."

My receiver has two subwoofer outputs, but according to the owners manual, the same signal is output from each jack. If the same signal is output from each jack, then wouldn't the result of hooking each to the seperate outputs on the receiver be exactly the same as using a Y-cord and hooking both to just one output?

Thanks for your help.

I honestly don't understand that one either. I have an Onkyo TXNR1007 which also has MultiEQ XT as well. I have two Rythmik subs and I use the two separate sub outputs because it does seem to treat them separately. I get separate distances and levels. One is corner loaded and Audyssey trims it appropriately. I've never measured the result with REW or anything like that but my bass is very smooth. IMO, you should set your subs up with the separate outputs. You can also try it both ways and see what you like best. All I know is that with the separate outputs, my system sounds incredible!

The problem with doing it this way is that XT cannot properly EQ two subs.

If you look at how XT32 does it first it becomes more clear. XT32 pings each sub separately and sets the level and distance for each sub. Then it pings both subs together and EQs them as a working pair. This means that both subs are treated as one as far as EQing them in the room is concerned.

XT does not have that capability. So if you set them up as two using both sub outputs on the AVR they each get pinged in turn and each gets its level and distance set as you say. But then each one gets EQd separately too. But because they never play separately, you can see that this won't give the best result - in fact the EQ of one could be working against the EQ of the other. You could have corrected a peak with one sub only for the other sub to reintroduce a boost at that same frequency. Neither sub knows what the other is trying to do, from a EQ POV.

The best result will be obtained by following the FAQ and using a Y splitter, setting the subs equidistant from the MLP if possible and ideally using matched subs. They will then be EQd as one sub - small differences in distance will be way less important than the correct EQ.

Why not try it and see?