"Official" Audyssey thread (FAQ in post #1) - Page 1105

Well to a certain extent, I can't really change the size of the room, and until it is built I can't even make changes to the acoustical treatment, so this portion of the 'discussion' is completely theoretical.



Hmmm, so based on this, is there any point in worrying about calculating or measuring the RT?

What is perceived by a listener?

A room with an echo is easy to hear, as is an overly dead one... or are you suggesting something else?

Hi Al,

How'bout this:

1. Love your room "as is".
2. Place your speakers in a reasonable way. (Never forget about WAF though!)
3. Run Audyssey as per the setup guide.
4. Grab a cold beer!
5. Enjoy!

...and a turkey leg...


Mike

Now that's a plan!

Hi Markus,


VERY interesting reading, thks.

Hugo

Wide > Heights Audyssey recommendation.

Have any articles been released on the data used for this statement? The logic on Audyssey's site seems reasonable but I could imagine an equally persuasive argument saying that all channels in the same x-y plane doesn't make sense, so some info on the z-axis is critical.

Were tests done based on a targeted movie? the a sound swirling around a perfect listening position? how immerse rain coming from all channels sounds?

Sure I can try this for myself, but I am interested in data an justifications for saying one is better than the other.

The Denon 4810 forum had quite a discussion on heights vs wides from early adopters. No studies done, just listening. Wides were seen as having a bigger impact. I added wides because it was so easy to do and have great results. Try it, you'll love it.

Although we have not published our research, the "Wides first" recommendation comes from studying both acoustical and perceptual benefits. Side reflections with the proper delay, level, and spectrum combine with the direct sound from the front to increase our perception of auditory source width. Height information is second in importance. But, this is not about swirling sound effects and helicopters. It's about extracting the information needed to reproduce a more natural soundstage.

I heard it demo'd at CEDIA a few months ago at the Integra exhibit. The room was tall and a decent width, and had some "volume." I use the word "room" loosely as it comes in a box truck and is set up on the trade show floor. No acoustical treatments, per se, but the walls were carpeted. The first demo we sat in the first row ... way too close. The Wides were discrete and the Heights not noticeable. We "went around" and came in for another demo. This time we sat in the back (the second row) and were the only attendees. That being the case, we asked if we could skip right to the movie and A/B'ing DSX with basic 5.1.

At this point I forget what the content was, but I remember the Wides integrated and expanded the front soundstage. And so did the Heights. I watched and listened carefully for "audyties" (© 2010 pepar All rights reserved ), but there were none. Something panned across the screen and continued offscreen, but not in such a way that made me think that the image should have been wider. It sounded exactly as I would have expected it to sound. Birds took off onscreen and flew upward ... and they also sounded exactly like I would have expected.

Jeff

If one had 7.1 bipoles (except center) all around would you still recommend "Wides first"?

You aren't kidding; lowest price I could find is $1300 ea.

Do you think it's fair to say it's a combination of a standard absorber and a diaphragmatic absorber tuned to the low freq?

Yes. Side information is more important than Height even if it's slightly more diffuse due to bipoles.

Yes, that's the basic principle.

Chris,

I've found that my height speakers (non-bipoles) aren't adding much. I think its because they can't be mounted quite high enough in my room, as well as, because my mains (and wides) are bipoles. I have 8' ceilings BTW.

Or maybe because I haven't listened to the right demo material. Other than Ratatouille, what do you recommend. I remember you mentioning other material a long time ago, but that was before I had a DSX setup.

If you need a specific scene in a specific movie to justify it it doesn't seem like it is a good long term thing.

My mains are already high 5'5" tweeter level (on bookshelves). I just put my spare speakers right on top of them and they almost hit the ceiling. hah. It actually creates more problems than it adds in features, but I am NOT doing it right. I look forward to wall mounting them and lowering my mains to ear level to see what it sounds like.

Real life is pretty 2 dimensional in terms of real life audio content... you know, people talking to you, dog barking at you, etc. My main interest in the processor added channels is for video games. With games I'm fascinated by abstract an unnatural sound-stages if it makes the experience more engrossing. Typically it doesn't at all. But the helicopter and spinning sounds scenario is more common since your digital voice actor come from all angles.

Heights vs. Wides continued...

Does the effect have anything to do with processing difficulty? I could imagine its easy to extract an infinite number of channels in the x-y plane from a 5.1 mix, but adding anything in the z-axis is more artificial/difficult to do properly. Is any of the wides > heights based on the inherent limitations of processing, or even if you create a proprietary format with discrete z-axis information (just for demo/testing purposes) is it not useful?

Yes! Very cool read. thanks.

Hi Doug,

As with properly calibrated surround speakers, you should only miss the content when it's gone. If you're looking for effects speakers that obviously add (i.e. change) to the intent of the content then DSX is not what you are looking for.

Heights in DSX are not about helicopters. Actually, they provide information that adds to our perception of depth of stage. We learned about this from dozens of recordings in concert halls where we hang mics 50-60' above the orchestra in the proscenium.

Since your Heights are not high enough (due to low ceiling), you could experiment by adding more delay to them. Try listening to well recorded live music and focus on the depth of stage as you turn DSX on and off.

.. I didn't hear the birds fly off the top of the screen ... ?

Thanks Chris. You actually confirmed what I am hearing... more depth... but not extra content. I thought I was supposed to hear sounds coming directly from above (extracted from the fronts) that would made me notice those speakers more. Instead I am experiencing a larger front sound stage. I thought maybe because I had bipoles for fronts and wides that they were "covering up" the correct height speaker experience.

I like what I am hearing, but just because I like what I am hearing doesn't mean its optimal. OIOW, it could sound better if I put forth a bit more effort.

Do you have some literature that talks more in depth about DSX? I have read bits and pieces from this forum, the Audyssey website, and other A/V websites, but I'd like to read something a bit more in depth... like a white paper.

Its not that I need specific scenes to justify DSX heights. I want specific scenes to listen to as well as expected results so I can validate what I am hearing compared to what Audyssey is intending for me to hear. Make sense?

Maybe perhaps like cooking shows? They give you the recipe, even walk you through most of it, then show you what to expect in the end.

Yes you did! We also used ch. 2 from WallE where he is climbing the stairs as he moves to the right. That R. Height channel sure makes him climb all the way up. But that was the intent for WallE and the birds, right?

Yes, perfect sense. I was being over critical -- I apologize.

It should do something with any good soundtrack. But it is nice to hear from an expert what the effect should be with a given sequence so you know you are getting the most out of the system.

So, the height channels are based, in part, off of data recorded by microphones high up in recording halls (not exactly, but you understand...)?

So, the leading development question might have been: "in an open, acoustically well designed room, what is the nature of the sound that comes from above?" Following that, you can extrapolate that signal a processor from the original sound which would come from the left and right channels. Or based on a microphone at the listener position in the concert hall and at the instrument location.

That would have the effect of modifying the sound stage and opening it up, rather than making cool helicopter effects. Not that anyone should have anything against helicopters.

That is entirely different from my initial assumption that it used processing of front and rear data to 'guess' just how high up that helicopter was.

I much appreciate responses directly from the company and people who know that is going on. It is incredibly awesome. Thank you.

Hi Chris,

By "adding more delay to (the heights)" do you mean going into the Speaker Distance setting in my MultEQ XT menu and decreasing (or is it increasing?) the distance from what Audyssey sets so that, in effect, I'd be adding more delay to them?

Along a similar line, please explain what effect the Soundstage adjustment (on my Integra 80.1) has when using Audyssey DSX. The owners manual isn't very clear on that, at least it isn't to me

Thank you for what you and the rest of the Audyssey team is doing. I love it!

Cheers,
SB

You're pulling my leg, aren't you, Professor?

No! I was trying to imply that expectation is part of the experience. I actually do a demo like this in class: everyone points to the flying helicopter above their heads and nobody believes that the grand piano is flying even though they are both rendered using the exact same method.

If you are entering distance, then you need to decrease it so that more delay is applied to "align" it with the other speakers.

That's just a level trim that controls the speaker pair (Wide or Height) together. You can do the same thing with the individual trims, but this is more convenient.

Yes. We set up microphones in many different halls and then compared the signals in the various mics to those used in more traditional locations (e.g, a particular seat row, over the conductor's head, etc.). Then we developed an adaptive filter method that uses the standard microphone as the target and continuously adapts the modifying filter applied to, say, the height microphone until the two are indistinguishable in a listening comparison. Then, that "virtual mic" filter is fixed. It contains in it the acoustical information of that space and the unique way it modifies the mic signal. If you do this over enough spaces you will start to find some commonalities and patterns and that was the basis of this research that eventually led to DSX.