"Official" Audyssey thread (FAQ in post #1) - Page 517

Originally Posted by penngray

Its not hard to test if it accepts nulls as crossovers actually.

Originally Posted by penngray

Hehe, You do not know much about the DIY community on here or on diyaudio, htguide, etc. Let just say you can not spend $6K for what I have for under $2K...You are aware that people doing DIY are also builders of fine custom speaker solutions that are commercially available?

This is design is just fine so lets instead thinking them as DIY think of them as built by Jim Salk, Rick from Selah, Mark Seaton and Curt C (who builds speaker kits)


This is a different topic though, lets stick to figuring out if Audyssey will mistake a null for a crossover point..

Originally Posted by augerpro

BTW some interesting lessons here. You are most probably using LR filters in the DCX? These sum flat when drivers are in phase and have a big null when out phase. You flipped the mid polarity and yet the response never went to a null. Isn't that interesting? Why do you think this is? This is why way back I warned you that an off the shelf active crossover is not a substitute for understanding the principles. All they do is give people the impression that what they are doing is correct, when later measurement will show that it is not. Now combine knowledge and active crossovers and you have something powerful. Did you ever read any of the books I suggested? Like Speaker Building 201 or Loudspeaker Design Cookbook?

Originally Posted by BobMN55446

I have a Jamo 6510 subwoofer. I just purchased a Denon AVR-1910. I ran audyssey and didn't like what it came back with for the subwoofer. I think my placement was too close to the corner of my room and it was slightly hiding behind an end table. Audyssey showed that it was 22 feet away instead of the 12 that it really was. I am going to move the sub out of the corner and was going to rerun Audyssey.
I am new to this whole HT thing and would like some help with what I should have the sub settings at before I restart Audyssey.
I have 3 dials on my sub - level (volume?), cut-off frequency, and phase? Any help would be greatly appreciated.

Originally Posted by sivadselim

Edit: Your speakers ARE pretty, though.

Originally Posted by audyssey

The sub gets the same logarithmic sine sweep signal as the other channels. It starts at below 10 Hz and ramps up through all frequencies with a logarithmic spacing as that has been found to improve measurement signal to noise ratio particularly at low frequencies. The reason you hear thumps is most likely because your sub has a lowpass filter engaged that you are unable to turn off. The best way to deal with that is to turn the frequency knob to the highest allowable value and leave it there. It's better for measurements and it's better for bass management in the AVR.

Originally Posted by sivadselim

Maybe you don't realize it. He's telling you as nicely as he can that you don't know what you are doing. Based upon my experiences here, gee, I have a strong suspicion that he is correct. You're no Jim Salk. Or the name of anyone else you want to drop.

Edit: Your speakers ARE pretty, though.

Originally Posted by batpig

you CAN lower the crossover frequency. you will just have a small portion of frequencies that are uncorrected by MultEQ filters.

for example, let's say your receiver crossover options are 80 and 100Hz. After Audyssey the receiver has set it to 100Hz. That means the measured -3dB point was likely somewhere between 80 and 100Hz.

If you decide you think the sub is localizable at 100Hz xover, you can lower it to 80Hz. The tradeoff is that there might now be some region (80Hz up to whatever the -3dB point was, unreported but somewhere below 100Hz) that is unfiltered by Audyssey. This tradeoff might make things sound worse, but it might make things sound better to you if you like the sub/sat blend better at that frequency.

AFAIK the consumer Audyssey setups do not choose the xover based upon the smoothest response in your system; they simply report the measured -3dB point to the receiver and the receiver picks the next highest available xover setting. So it is certainly POSSIBLE that lowering the xover setting one notch might produce a more pleasing sound, and the benefit may outweigh the tradeoff (having some uncorreceted frequencies).

I think what you DON'T want to do is make major changes to the xover frequency, as then you are probably going to have a big hole in your response. For example, if your speakers are tiny sats that roll off around 130Hz, and the receiver sets the xover to 150Hz, if you lower it to 120Hz it probably won't be a big deal.... but if you lower it to 80Hz, you are going to have a pretty sizable gap in the frequency response.

Originally Posted by penngray

Batpig, I believe Audyssey probably keeps those corrections in place but they become inacurate since we are adjusting the FR plot by change the crossover. The controlled peaks may be dips when a new crossover is set and this is a great concern form considering I would be moving the 120Hz crossover down to 70 or 80Hz....That is a pretty big change in the FR plot.

Originally Posted by SoundofMind

Bob, see the first post of this thread for helpful info including this link to the Audyssey setup guide:
http://www.avsforum.com/avs-vb/showt...5#post14456895
Therein you will find how to set those controls prior to running autusetup:
Vol at mid point, set the cut-off freq to max or bypass the filters if the sub allows, and set phase to 0. Do a little reading on this thread and you will see that the acoustical distance readings for subs is often longer than measured distance due to signal delays from the sub's internal processing.

After running autosetup check that the 1910 does not have the sub channel trimmed to +/-12. Reset the sub "crossover" to 120 Hz. The sub's output (=low freqs) are not nearly as affected by things such as end tables as hi freq are. So put the sub where it produces smooth, even bass. I strongly agree with getting the sub out of the corner before you rerun autosetup. But I do let my sub hide behind a plant and a chair, as it is very timid.

Oh yeah, how's it sound?

Originally Posted by BobMN55446

SOM,
Thanks for the help. The sound difference is really amazing with Audyssey.
Bob

Originally Posted by homeav

Whether the speakers are good or bad has little to do with room-related nulls.

Originally Posted by penngray

Does Audyssey mistake deep nulls as crossover points?? I say yes it does, Chris says "Yes", "Maybe", "Absolutely Not".....Im not sure he knows.

Originally Posted by sivadselim

Well, .................I don't know about that. But I think you missed my point.

Originally Posted by pepar

I still have yet to see any suggestions on how Audyssey should, or even could, handle nulls differently.

Originally Posted by bradandbree

Isn't the answer obvious?.... camels.

Originally Posted by pepar

I still have yet to see any suggestions on how Audyssey should, or even could, handle nulls differently.

Originally Posted by Gary J

Originally Posted by batpig

penngray - I think the reason some are interpreting your tone as negative is comments like these. You are an excellent DIY builder and you know a lot about speakers, but you are talking to one of the world authorities on acoustics, sound processing, etc. You seem to be asserting that he doesn't understand the product he INVENTED or that he is somehow intentionally obfuscating.

Following along, Chris has been pretty consistent on the "null" issue -- Audyssey's algorithm looks for a specific pattern when calculating the F3 point. It should not mistake a null for the F3 point UNLESS the shape / width of the null mimics the roll-off of a speaker that you would get at F3. And, as pepar asks, what would you have Audyssey do in the presence of such a huge null? Maybe it is CORRECT in recommending the crossover there and having the subwoofers pick up at that point?

I don't think it's fair to imply that he is being inconsistent or, worse yet, does not understand what he is saying!

Originally Posted by LarryChanin

Hi,

Batpig's advice is right on the money. It was provided to a forum member with tiny speakers. It is unclear whether your response was intended to disagree with his excellent advice. Clearly he was not intending to address your situation.

He is merely stating consumer Audyssey setups do not select crossovers based upon the smoothest response. In fact, as has been discussed many times, it is the receiver/prepro, not Audyssey that is making the initial crossover selection. Since the initial crossover settings might not yield the smoothest subwoofer to satellite transition, batpig is merely suggesting that experimenting with small changes in the crossover settings might result in a smoother sub to satellite splice, even if a portion of the response might not be corrected by Audyssey. He very specifically was careful to caution against making major changes in crossover settings.

It is also clear that Audyssey correction is applied independently of any crossover setting. Audyssey measures each speaker and sets filters that attempt to get each speaker's response to match the same target curve. This is necessary to more closely timbre match the speakers and to permit the receiver/prepro to select the appropriate crossovers for each speaker.

I don't follow your comment, "The controlled peaks may be dips when a new crossover is set...". The response is corrected on a per speaker basis without regard to crossover settings. As batpig stated, for consumer setups small changes in the crossover may (or may not) result in a smoother sub to satellite transition, but I don't follow how that extrapolates to peaks are converted to dips.

Larry

This is necessary to more closely timbre match the speakers and to permit the receiver/prepro to select the appropriate crossovers for each speaker.

Originally Posted by penngray

...it seems that we have full defense going on for Audyssey now and it's not really needed.

Originally Posted by homeav

I think that it would be great to see test results (yes time consuming,etc) of how Audyssey responds to nulls.

If Audyssey even did the research that would be even better. I'm more interested in a product when I know it's benefits *and* limitations. Knowing how a product behaves in problem situations - then lets you work around it. (and maybe even appreciate the advantages more).

If it struggles with nulls - so be it (doesn't make it a worse product) . I understand that it's simply impossible for Audyssey to fix/handle every situation. I think Audyssey has high expectations to meet.

Originally Posted by penngray

example, say you have a +5dB peak @ 100Hz and Audyssey smooths that out with a small BandPass EQ (Q =0.1 dB=-3dB Fs=150Hz)

Originally Posted by Jim Hef

I think it may be. Is there an answer that he can give without seeing all your tests and their related equipment that would be based on his experience? Your question is reasonable, but I doubt he has enough data to venture a true response. I've seen this in other threads where questions become "direct" or "blunt", and the creator goes away, not needing to be bothered to be thrust into a situation that he may not want to be in.

Originally Posted by sivadselim

That's not how Audyssey works.

Originally Posted by penngray

Audyssey has zero idea of what Timbre matching means. Its all about EQing and its simply finding the F3 of each speaker regardless of timbre matching.