"Official" Audyssey thread (FAQ in post #1) - Page 1648

Dynamic range is the ratio of the softest part of the sound/signal to the loudest part.

What is the softest part of music/movies - silence (0 dB FS)?

I thought the discussion was about how DEQ gets the surround levels wrong because the varying loudness compensation was not applied to the surrounds when they had the mixers tweak the levels when MV was reduced? I didn't read on Ask Audyssey where anyone ... ask audyssey if that was the case, and if so, wouldn't that have changed what the mixers did with the surround levels.

This RLO chatter is a bunch of hooey. The darn thing - DEQ - is designed for movies. Music is so all over the place that, if one cares about accuracy, it is impossible to adjust DEQ correctly, and album to album is different. For music, turn it off and play it loud!

Jeff

How is 90+ db average listening level LOWER than 85?

The calibrated SPL of a monitoring system is completely unrelated from what is coming out of the speakers if you aren't concerned with an industry standard...

If the goal is to mix louder (without hearing damage) you must lower the monitor level to achieve more SPL if you are to also intent on using as much of the available headroom available on you recording medium.

You are assuming that if the engineers calibrate the monitors, that they stay at the calibrated "0", which they do not in my experience...

Once again, without a standard, "reference" is meaningless...

"Compressed" seems to be shorthand for compressed dynamic range. I understand Dolby DRC, so I understand your use of terms!

Maybe even play it soft, medium, or loud.

Markus.. your argument only has relevance if you are constantly changing the monitoring level over the course of mixing a piece of content, which of course no one does when mixing on a given system (that's not to say they don't check the mixes at various monitor settings, but as a rule most music mixers set it and forget it... when they want to check for "changes" it usually requires going from mains to mid or near fields, or out to the car or iPod for a check..)

If you lower that monitors, you push up the gain into the mix buss to achieve the same SPL..

You are starting to run around in circles the concepts of a calibrated SPL monitoring level vs. how hot the recorded signal is vs. the SPL's coming out of the speakers... three different things, related yes, but none of the three wholly determinant of the others (i.e. if you calibrate to a certain SPL, and want to listen at a certain level, you can either turn up the monitor pot or push into the recorder or a complimentary combination of both.... however, if the result is 90db average SPL out of the monitors, equal loudness contours isn't relevant.)

As you stated, Dynamic EQ works optimally, and only as designed, when you know the reference level the material was mixed at..


It works as intended because the monitoring levels were higher... you are confusing the monitor calibrated level with SPL again..

I was... and as I just stated in another post, the calibrated SPL has no bearing on how loud the mixer listens at... that is determined by how loud they turn up the monitoring pot, and, if it's not painfully obvious by now, is how most music mixers work in order to print hot and listen loud.

Sure, whatever. But as there is no standard, it's all a guess .. and a constantly changing one at that. Better to turn it off or use it however you like, but to blather on for page after page after page is silly.

Thanx, they also mention it on the THX site:


This makes as little sense as their previous recommendation of placing the rear speakers together (only to use processing to make them sound like they're apart). If the above is to be believed, the reason to spread the rear speakers apart is if you switch from lossy (DD, DTS) or hi-rez lossy (DD+, DTS-HD HRA) to lossless (TrueHD, DTS-HD MA). Why would a small difference in fidelity between DD+ and TrueHD (assuming the difference is even audible) need a change in rear speaker configuration?

I think ASA was born out of 5.1 post-processing and still had a place for the 6.1 matrix/discrete formats. It was also addressing the issue of having one speaker configuration for all content. With 7.1 and, ESPECIALLY, hi-res MCH music, the efficacy has greatly diminished. But I'd guess that there is "inertia" and they cannot overhaul their stance overnight. Hence the walkback that has stopped at an awkward place.

Some time ago I moved to having two completely different speaker configurations - the 7.1 THX and an ITU one - that I physically swap in and out. And then swap Audyssey Pro filters as well. Actually, it was Pro that made this feasible. A kludge, but a feasible one.

Anywho, that's my $.02.

Jeff

The author in your sig points to research from Robinson & Whittle in 1960 and Sivonen & Ellermeier in 2006 which demonstrates that "sounds arriving from the side-front, about 60° will appear to be louder than those from the front and louder than those arriving from the symmetrical directions in the rear hemisphere". Same sound, different perceived loudness (depending on direction).

True, since THX never matrixed stereo rears (only dual-mono rears using Surround EX). But even dual-mono rears would still have benefited from spreading the rears physically apart (less hotspotting in the middle of the back wall, better rear envelopment, fewer imaging reversal artifacts).
Would've been a little more understandable if they had recommended different configurations for 5.1EX/6.1 versus discrete 7.1, but the lossy vs lossless reason makes less sense. But, hey, I'm not their marketing person.

It is exactly the ideal solution that I described, having metadata to reflect each song's loudness.

Your solution:

I have a redo of my theater rear planned and part of that is to dis-collocate the rear surrounds and wire them for tripole (M&K) pattern from the monopole they are now.

Jeff

You won't listen to the music without lowering the playback level, either by reducing the monitor level if your playback system or lowering the level of the source... It will be unbearably loud if you do not. .

Either way you compensate , without knowing what the reference level was during mixing, accurate reproduction of the frequency response is impossible, but because we will tend to listen to music within an acceptable range around where it was intended, I don't think the issue is a great one of concern.

Yes, I mentioned that this method is also a good solution, but not as surgically precise as Mogorf's metadata as his works track by track, and Intelli works on the overall source.

I've just read the last 4 or 5 pages. My Conclusion: DEQ is garbage. I'd rather let my brain do the compensating than let DEQ compensate it incorrectly.

It's wrong because you are still ignoring the fact that music people don't leave the monitor at a calibrated level.... They will push the mix buss into the recorder, and turn the monitors up from where they calibrated at to get the desired spl.

This maintains a more balanced signal to spl ratio than your drawing reflects.

Calibration has nothing to do with where they mix at.

Does this topic need any more debate

Music, as a general rule, is mixed at a louder and higher level than film content.

What more is there to discuss?

Correct


Probably. But it does not matter.


Sorry to disagree with the right honorable gentleman, but this is not related. Monitor level has no bearing on the average level of the mix. Pop music is EQ'd and further peak compressed after the recording session is finished. That's what defines the final properties in the release medium.

My new Onkyo 709 is yielding some very weird results with the mic calibrations (seems like some other people, not all, having similar issues since the last firmware update).

It yields a VERY tinny hollow sound after the 8-point XT measurement (almost like it's rolling off everything BUT the very high end). It defaults to "Movie" Audyssey curve, but i noticed that switching it to "Music" made everything sound fantastic (and much better than pre-calibration).

So regardless of the weird issue with the "movie" measurement, is there necessarily anything wrong with just leaving it in "Music" as far as Audyssey is concerned?

thanks

All that matters is the SPL in the consumer's home.



Because DEQ does not understand the difference between reducing the MV to compensate for loud recordings vs a choice to play them back at reduced volume. That's why it needs RLO, or alternatively, some means to normalize program loudness (metadata, Intellisound) before the MV sees it.