"Official" Audyssey thread (FAQ in post #1) - Page 2028

 

Good point. But at least, if you really wanted it, there is evidence showing exactly how the wings do keep the plane in the air. ;)  And i bet the people who claim to have that evidence wouldn’t withhold it from you saying "for those who don't know how the wings keep the plane in the air, the evidence won’t tell anything, and for those who do know, the evidence isn't needed".  One thing is for sure, it isn't hot air that keeps the plane in the air :) 

Sorry guys to beat a dead horse but the more I think about it the more confused I am. When I set my IB sub level to 75db during the initial Audyssey setup, I am at about 50% volume on each channel of my Behringer ep4000. After calibration Audyssey then trims the subs an additional 10db. So basically I would need to set the gain controls to about 25% to achieve 0 trim after Audyssey calibration. The Behringer manual states the knobs on the front of the ep4000 are really not volume controls. It basically states that at 50% on these knobs I am limiting the ep4000's output to 50%. If I set them to 25% to achieve 0 trim in audyssey I am down to 25% available power from my sub amp. I hear people who worry about bottoming their subs so I cant understand how my sub could keep up with 3 Lascalas and 4 Heresys with only 25% of the ep4000's available output.

And that is what makes me leery of it.  If the quieter notes are more EQ-ed to sound as if they are less quiet while the louder notes are less EQ-ed, that is a musical distortion, imho.  Linking the correction to overall level setting makes more sense even if there is the capability to do "dynamic" EQ.

Hi Kal, you are raising an interestion point here. Care to expand a bit on why you think it's a "musical distortion"? AFAIU, the second-tier makes sense because when DEQ chimes in we are no longer at 0 dB reference level, thus without it the spectral/tonal balance of the contents would change from soft to loud parts as the music plays, while the ultimate goal is to hear the program material with the same tonal balance as we hear it at ref. level. This is taken care by DEQ's 2nd level compensation.

On a second note, in order to be able hear the difference it would be interesting if the second-tier could be turned off, but we know this can not be done by the user.

I would listen and enjoy at what you think sounds the best. If your sub isn't exhibiting any signs of stress somewhere in between the two settings, or even with the +8.5dB boost, then have fun! I know it takes that big of a boost to get my sub back up to a flat response (measured with calibrated system) so it really is up to you to either trust Audyssey, or get some measuring gear and see for yourself what Audyssey is doing I think if the majority of the people on this thread actually bought a measurement rig, they might be very surprised at what they have settled in on as a "good" sounding system. The whole preference vs. reference thing, right?

How big is your IB? 4 subs? 15's? IB's are very efficient, so being at a 25% spot on your attenuation knob is not a bad thing. The attenuation does not cut the "power" of the amp also btw, it simply "attenuates" the signal to get it into the proper range. I have with 8 sealed 18's my amp's gains are a little below 25%.

I have 4 18's.

I pulled up the manual on line and didn't see anything like that. Looks like these are typical input sensitivity controls. In a "pro" (garage band) setup, they would mostly be used to make sure that the devices 9like a mixer) in front of the amp don't overdrive the input section causing distortion before the ampe even begins to really do its thing. Also to establish an appropriate gain structure (usually thought of as setting up the entire system for lowest noise and appropriate output).

The actual power amps always have the exact same gain, so how loud (ie how much power output) the amp gets is controlled by the level that reaches the power section from the input stage. If at 25% of the volume control, you get the subs properly calibrated with the mains, then whenever the signal to the subs is maxed (essentially 40 dB louder at 115 dB for the LFE channel) the amp will indeed get 40 dB louder because that's all it can do. It simply responds to the input. So if, theoreticlly, your setup prevents you from achieving full power from the Behringer with the input signal from your receiver, that's because the added power is never ever needed to reporoduce the sound at full correct volume. That's called headroom and most folks think it's a good thing.

Also, play a real bass-heavy soundtrack with a lot of ULF (Ultra-low freq) material and I am willing to bet the 18's will still start knockin pretty good listening to music or normal program material with a system that capable won't make them move like many are talking about. You also probably have 8-10x the displacement of the people talking about bottoming their subs, haha.

Dynamic adjustments based on signal level are necessary if the equal loudness curves behave non-monotonically. Audyssey obviously believes this is the case although the ISO curves do behave monotonically. Here are Dynamic EQ equal loudness curves measured from an AVR preamp output.



The following graph shows the dynamic behavior of compensation curves for three different input levels with MV at -30dB.

That's quite a first post urwi!

Now Feri doesn't have to make the graphs that he feels are unnecessary, so thanks

Yes, welcome, urwi!!

Hey batpig, I wrote: ...the graphs unnecessary for those who have the grasp! LOL What else did you expect, eh?!

Here's a great welcome to urwi!!

Now hopefully Keith will also u'stand DEQ two-tier operation!! Right Keith?

Just because something isn't *necessary* doesn't mean it's not cool to see it in action.

Since we're talking about it, I would also dispute the idea that it's not *necessary* per se. People understand the principle of equal loudness contours, so by your logic it's not "necessary" to see graphs of DEQ's corrections for those who understand the principle, but it's still valuable to see actual graphs of the pre-outs at different MV settings to see the actual correction being applied by DEQ.

Trust but verify, right?

bp, i don't think we were discussing equal loudness contours in general this time, but the unique two-tier implementations of Audyssey which if not understood by someone there will be no graphs in the world that would explain it to him. But OK, let's move on. I'll be much more interested in Kal's reply.

OK.  First, the spectral balance differential between loud and soft instruments played simultaneously exists at 0dB reference level and is inherent in the composition.  The question is whether, at lower levels, the compensation can be accurately scaled so that the same relationship is maintained.

 

However, there are no real level standards for music recordings, so reference is arbitrary.  Have you ever played a string quartet recording at reference level at home?  It is too loud.

Also, one must rely that Audyssey can distinguish between simultaneous loud and soft instruments and treat them differently.   The only argument for this, off the top of my head, is that the louder instrument will mask the quieter one so that treating them the same makes no difference.  Perhaps so.

 

My (admittedly) limited experience with DEQ is that it muddies the sound of music at any non-reference level.  I use it only for broadcasts of sporting events.

 

I can’t help with the Sub trim problem unfortunately, but I am not sure you are totally clear on the purpose of the controls on the ep4000. The things that look like volume controls are in fact gain controls. Turning them down doesn't in any way limit the output power of the amps. What is happening when you turn down the gain is that the amps can deliver their rated power from a lower level of input.  So when you turn down the ep4000's gain controls to the 50% you mention you are not "limiting the output" to 50%. You are getting the same output but from 50% less input. Does that make sense?

 

The controls on a typical subwoofer work in the same way - they are gain controls and having them set very low doesn't mean that the sub's output will be limited. It just means that they are seeing a higher level of input, so their gain control is turned down accordingly. 

 

+1.

 

What I don't fully understand, Feri, is that about a dozen posts back you said you were "finished" with this topic. Seems not...

 

 

I do not agree that you understand the way DEQ works either. I think you think you do though.

Edited by kbarnes701 - 3/19/13 at 10:38am

I had previously looked into this a little bit. The Radio Shack sound meters have always been known to need corrections applied when using their meter readouts, especially in the frequencies below about 31Hz. You will get a lot of hits when you Google "Radio Shack sound meter corrections". Here is one such table which seems to be the same as found everywhere http://www.subwoofer-builder.com/SPL-corrections.htm. These have been widely accepted for probably at least 10-15 years.

However the discrepancy from this arises when you use the Radio Shack sound meter as a mic with any sort of frequency response software/hardware. According to Acoustisoft there is a difference using the RCA outs versus the built in meter readouts. They provide measuring software with calibration files and make this note about the Radio Shack meter:

"Other calibration files exist on the internet for this unit, these calibration files apply when reading the meter movement, not when taking data from its RCA output jack. These units are very constant gain below 500 Hz (flat frequency response) and do not require further low frequency calibration." http://www.acoustisoft.com/micpreamp.html

This seems to be supported by your links showing measured comparisons between the Radio Shack meter used as a mic where they were obviously using the RCA outs versus the built in meter. Again however the correction charts needing to be used when using the Radio Shack meter readouts have been pretty well accepted as necessary for a number of years. So basically we have two different situations: one where the Radios Shack meter is accurate and one where it is not, both depending on how the meter is used.

 

I have always thought the same thing. In real life, the spectral balance changes if the 'volume' is turned down - eg if the musicians play less loudly, or you sit further away from them. As there is no reference standard for the 'loudness' of recorded music, then it seems impossible to me to 'correct' for deviations. Of course, with movies, it is an entirely different thing. But to be fair to Audyssey, I think they too would say that DEQ has limited usefulness for music, hence their provision of RLO which is a kludge aimed at 'improving' the way DEQ works with music sources.

 

Of course, there is also the issue of DEQ raising the level of some parts of the source more than others, depending on its loudness and I have never fully understood how DEQ 'knows' that a softer passage is not supposed to be softer. If it raises the level of an intentionally softer passage, then it is introducing distortion. It is all moot to me because I only ever listen to music (seriously) in stereo (call me old fashioned), other than music in movie scores of course, which often I admit does sound wonderful. 

that seems to be a pretty arbitrary and self-serving distinction, and a pretty pessimistic view of humanity and our ability to learn. Why would a visual representation not be helpful in demonstrating a specific concept that someone doesn't understand? How is that different than someone not understanding the concept of equal loudness contours in general, and then showing them a graph that demonstrates what is happening? I simply can't understand why you are trying to force this weird distinction on this narrow topic.

Hi, everybody. I'm researching the purchase of a new sub, and I've run across a few comments about Dynamic EQ having a very detrimental impact on subwoofer performance. I haven't experienced this to be the case myself, which has me wondering if:

a) Dynamic EQ is an issue at all?

or

b) if it is an issue, is it an issue in all implementations, or only as implemented in some specific AVRs or processors?

Also, I ran across this post in the Rythmik thread.

http://www.avsforum.com/t/1214550/official-rythmik-audio-subwoofer-thread/8670#post_23083505

Most of it is over my head, but they're talking about reference levels, and my understanding is that Dynamic EQ isn't doing anything at reference level, and should have no impact. Is my understanding correct?

Is there any truth to that post?

With all due respect bp (+ a million more), may we move on?

Good to hear another person's take on this. I also prefer leaving DEQ off for music, and due to the lack of music recording level standards, I simply set the MV at what sounds to me like a realistic level for a 'live' experience, i.e. how a string quartet, solo piano or orchestra sounds in a concert hall from the 5th row, or an acoustic guitar in a bar or smaller than concert hall venue, from 25' away etc. I take cues from live music to set the MV and leave DEQ off for music.

When a live pianist plays pianissimo in real life, I'm going to hear the lows and highs a little less than the middle notes/frequencies. I don't particularly want DEQ to dynamically boost those frequencies on the fly to 'restore their prominence'as this would be a distortion of what it would sound like when a live pianist plays softly.



Max

Excellent find sir! Thanks for the additional info. Alas, with the digital readout RS meters, this situation has gotten better than the older metered version. I will do some additional testing of my own to verify though

I've been (as usual) on the fence about DEQ, but I'm leaning toward leaving it off for all music except maybe some two-channel stereo sources as well. I'm finding that DEQ boosts the surrounds more than I really like on MC recordings (majority being BluRay concert), especially if I'm going for a "live" feel. OTOH, I tend to listen at -10 to -20 db in our room, and our subs are near field (either side of MLP sofa). But it depends on what role the surrounds play in the mix: if they're primarily used to create "ambiance", I find it less annoying than on mixes where instruments have been relegated to the back channels.

Maybe it's just trying to keep those sounds from becoming inaudible?

Remember Roger's post from last year:

Dynamic EQ doesn't "distinguish between simultaneous loud and soft instruments". It simply looks at the overall level and varies equalization of lower frequencies accordingly.

Modal decay might be too low in that particular room. Raising low frequencies (like Dynamic EQ does) will increase the audibility of such room problems.
Edited by urwi - 3/19/13 at 11:44am

Nope. It's a bunch of technical gobbledygook intended to make folks think the poster is much more well informed than he is, and to make the average reade go, "Wow... that totally went over my head. This guy really SOUNDS like he knows his stuff though, so I guess he must be right".

He doesn't really understand acoustic processing OR how bit depth is used. Oh, 16-3 bits leaves 13 bits. Huh? And the whole, "it sounds good with cheaper/crappier subs because they have lower resolution". Really? Subs are digital devices? Who's he trying to fool? Many users here have subs that conparatively speaking, make Rhythmik subs the cheaper/crappier subs, but DEQ sounds just fine.

Yes, it IS possible in some rare cases, for DEQ to cause less capable subs to bottom out where they didn't/wouldn't without DEQ. The normalization boost can occasionally make that happen with subs approaching their limits in the lowest frequencies, but knowing this, proper setup can eliminate that problem. Ridiculously capable subs might not even flinch under similar conditions though, simply due to having ample headroom, but starting off with an optimized setup (ideal sub, speaker and MLP placement, optimally some acoustic treatment etc.) tends to ameliorate problems. For those who cannot or will not optimize placement AND have bottoming problems due to DEQ, there's always the option to turn it off.

Now this might all sound odd in light of my previous post before this one, but I DO use DEQ when watching movies below Reference because movies actually HAVE a Reference level and I don't/can't always watch at Reference. Properly recorded concert and or musical BDs though, are viewed at Reference without DEQ. And I don't use it for music.



Max