Big news coming from Sound United in 2023!

[Karls_speaker_works]

I watched that video up to the point he started showing what he called strange, about the loss of bass output at high volume settings using the offsets. Then I lost interest because the gentleman seem to forget Audyssey's DEQ offsets rely on the volume setting only so without seeing his trim level setting it is hard to assess how "strange" or interesting behavior he highlighted.

I turned the video off also. It seemed he was bewildered that it was acting exactly the same as a similar function in my 12-year-old Pioneer VSX-820 that boosts lows/highs at light volume. There's another control (can't recall name) that boosts surround levels at light volume. I like the way these functions work - they are as expected and desired.

 

[-Matt-]

What I would like to see is comprehensive measurements of the DEQ not only at various volumes and reference level settings, but also with signals of dynamically changing amplitude (i.e. not just set an amplitude and run a frequency sweep). If the DEQ is truly "dynamic," as its name implies, it might change its FR characteristics not just with those static settings, but also by determining the actual level and dynamics of the material (although that could introduce dynamic coloration).

...Another finding is that the DEQ boosts the surround sound not just below 500 Hz, but over the entire frequency range into the highs, which causes an imbalance even with Audyssey corrections enabled. That explains why I heard the side surrounds playing too loud compared to the fronts and center with the multi-channel source material.

I think Audyssey Dynamic EQ only adjusts its boost based on the master volume setting, and if you think about it it makes sense. It would be truly awful if the frequency response constantly changed depending on whether the currently playing music was in a loud or quiet passage.

I posted some measurements on the effect of Dynamic EQ settings here. Don't know if you will find then comprehensive enough (feel free to post your own) but they serve to illustrate what the feature actually does.

 

[lc6]

I think Audyssey Dynamic EQ only adjusts its boost based on the master volume setting, and if you think about it it makes sense. It would be truly awful if the frequency response constantly changed depending on whether the currently playing music was in a loud or quiet passage.

I posted some measurements on the effect of Dynamic EQ settings here. Don't know if you will find then comprehensive enough (feel free to post your own) but they serve to illustrate what the feature actually does.

Thanks. Your measurements seem to more or less match those in the video I cited. With the DEQ on in your system, the fronts and center have a significant bass boost below 150 Hz (and only slightly rising with decreasing frequency), but the highs are boosted much less and only above ~10 kHz (which does not match the ISO 226:2003 curves at the typical 60-80 dB SPL).
It also looks like in your system the bass is significantly peaking around 80 Hz, which might lead to "boominess" depending on the source material.
Your perception of the surrounds running way too hot is consistent with mine, and I agree this poses volume adjustment problems.

I also agree that the DEQ should only use the volume setting and not the signal level or transients as input, but I still would like to have that verified by measurements (I am not sure if the AP, REW or similar have this kind of capability; probably not).

Anyway, while I will keep experimenting and try to re-run the Audyssey across all 8 reference locations (even though I do not have a second row of seats), I believe it will be hard to overcome those DEQ deficiencies at low volumes.

 

[-Matt-]

This is the suggested formation of Audyssey microphone measurement positions...

(I think this is supposed to be a view from above and behind the chair).

I just use this formation around the main listening position. I.e. 1 is where your head goes, 2 is a foot to the left, etc.

Edit: About the peaking at 80Hz, I have since re-done my MSO settings (3 subs) and the response is much smoother. The important thing shown in those measurements is the difference between each line (not the shape of the reference DEQ off curve).

 

[lexelot]

The bad news is that, per the above video, the DEQ is what it is and cannot be adjusted. The good news is that my hearing is not entirely crappy yet.

If you listen at a level significantly lower than reference, use the DynamicEQ offset. All DynamicEQ does is increase bass and high frequencies based off of current volume compared to the set offset.

So for example, if the offset is 0db and you’re listening at 0db, DynamicEQ is nonfunctional.

If you listen at -40 and the offset is 0db, it’s going to apply a huge boost.

If you listen at -40 and the offset is 15db, it will apply a boost but significantly less than when offset is 0db.

If you listen at 0db and the offset is 15db, it will actually apply a trim to bass and high frequencies.

If you are hearing boomy bass with it on, I strongly recommend adjusting the offset.


Also if you aren’t using the MultEQ app, I recommend trying Flat setting instead of Reference. It gets rid of the default midrange dip that Audyssey adds and sounds a lot better on most speakers.

 

[lc6]

If you listen at a level significantly lower than reference, use the DynamicEQ offset. All DynamicEQ does is increase bass and high frequencies based off of current volume compared to the set offset.

So for example, if the offset is 0db and you’re listening at 0db, DynamicEQ is nonfunctional.

If you listen at -40 and the offset is 0db, it’s going to apply a huge boost.

If you listen at -40 and the offset is 15db, it will apply a boost but significantly less than when offset is 0db.

If you listen at 0db and the offset is 15db, it will actually apply a trim to bass and high frequencies.

If you are hearing boomy bass with it on, I strongly recommend adjusting the offset.


Also if you aren’t using the MultEQ app, I recommend trying Flat setting instead of Reference. It gets rid of the default midrange dip that Audyssey adds and sounds a lot better on most speakers.

Thank you. As I wrote in on of my previous posts, I did try the 4 DEQ reference settings, and found only the 5 and 10 dB ones to be usable in my setup. I also tried the Flat setting (a "neighbor-pleaser" for apartment-type locations, as the description implies), but found it to cause an audible distortion.

 

[lc6]

This is the suggested formation of Audyssey microphone measurement positions...

View attachment 234375
(I think this is supposed to be a view from above and behind the chair).

I just use this formation around the main listening position. I.e. 1 is where your head goes, 2 is a foot to the left, etc.

Edit: About the peaking at 80Hz, I have since re-done my MSO settings (3 subs) and the response is much smoother. The important thing shown in those measurements is the difference between each line (not the shape of the reference DEQ off curve).

FWIW, from the AH video I am now watching:

 

[-Matt-]

Exactly, if you use multiple measurement positions closely surrounding the main listening position then it will help to average over the room response (which can vary even as you move your head).

I use MSO with multiple subs to try to get sub response to be smooth over more seating positions.

 

[ryanosaur]

(I think this is supposed to be a view from above and behind the chair).

Correct.

1 is where your head goes, 2 is a foot to the left, etc.

The recommended mic placements are supposed to be 2' apart for Audyssey. (Not greater than 2' to be certain.) Correct, regarding the pattern.

 

[-Matt-]

I think that the above discussion serves as a reminder that modern AVRs are hugely complex devices with many, sometimes confusing, options for configuration. The various options allow the user to completely transform the perceived sound. (It requires a considerable investment in time and effort to get the best out of such a system).

This means that our subjective listening impressions are almost completely useless in determining the quality of the underlying hardware. This is why Amir's measurements are so important. (Although, at the same time, IMHO, we shouldn't fixate on maximum SINAD; beyond a certain level of transparency it makes very little difference).

 

[-Matt-]

The recommended mic placements are supposed to be 2' apart for Audyssey. (Not greater than 2' to be certain.) Correct, regarding the pattern.

Haha! I noted that, on the app, it said they should all be below 2' from each other. If you do position #2, 2' to the left of #1 the then diagonal ones (#5 & #6) will be further than 2' from #1. Hence you need to do a grid with <2' spacing. Since I'm at heart a metric person, I actually aim for 40 ish cm spacing! (I mentally rounded to 1 foot for US convenience)!

 

[peng]

If you listen at a level significantly lower than reference, use the DynamicEQ offset. All DynamicEQ does is increase bass and high frequencies based off of current volume compared to the set offset.

So for example, if the offset is 0db and you’re listening at 0db, DynamicEQ is nonfunctional.

If you listen at -40 and the offset is 0db, it’s going to apply a huge boost.

If you listen at -40 and the offset is 15db, it will apply a boost but significantly less than when offset is 0db.

If you listen at 0db and the offset is 15db, it will actually apply a trim to bass and high frequencies.

If you are hearing boomy bass with it on, I strongly recommend adjusting the offset.


Also if you aren’t using the MultEQ app, I recommend trying Flat setting instead of Reference. It gets rid of the default midrange dip that Audyssey adds and sounds a lot better on most speakers.

I don't think flat disables MRC. You need to app to do it, flat or not.

 

[lexelot]

I don't think flat disables MRC. You need to app to do it, flat or not.

I measured both on REW with my old X4500H and flat removed MRC completely. Not sure if things have changed but I know it wasn’t present a few years ago.

 

[Sancus]

I think Audyssey Dynamic EQ only adjusts its boost based on the master volume setting, and if you think about it it makes sense. It would be truly awful if the frequency response constantly changed depending on whether the currently playing music was in a loud or quiet passage.

Yes, to perform fully automatic, proper equal loudness compensation, your electronics need to be able to analyze the entire track. Which is impossible for an AVR to do, as it's not the playback device. Without that capability, any attempt at equal loudness compensation will always be undone by the varying digital levels in the actual content. So it can only ever be an approximation that requires user input.

FWIW, from the AH video I am now watching:

Not really sure how that quote was used in the video(way too long), but the key part of it is "global EQ". Audyssey is global EQ(that is a flaw, if it was local EQ like MSO then it would be much better at multi-sub).

The reason behind Audyssey's multiple-seat measurements is so that it doesn't global EQ the other listening positions straight to hell, basically. If you really, truly only care about one seat, it's best to take all the measurements close to that seat. It's still a good idea to take 6-8 measurements though to prevent it trying to EQ small variations in the mids and highs. You can do 8 measurements all within 1ft of a single listening position if you want. This applies to any room correction that is working full range. (some people use the moving microphone method to shortcut this but it's sadly not compatible with Audyssey)

You're still going to get better results if you can modify the target curve with the app, most likely. The app also lets you restrict Audyssey to changing things below a certain frequency, like 500hz, which can be useful if you have really flat speakers that don't really need correction high above schroeder.

 

[Chromatischism]

I do not listen at very high volumes, so the author's finding that with some reference settings the DEQ actually reduces bass and highs

DEQ will never do any cuts to bass or treble unless you are over reference level, aka into positive territory on the relative scale, or over 80 on the absolute scale.

But one of the takeaways is very valuable: at moderate levels, the 0 dB reference setting does produce too much of "boomy" bass.

I use 0 and I do not have boomy bass. I have some of the best bass attainable. If it gets boomy, it means it is enhancing problems that already existed in your bass response due to room modes.

Also, in terms of highs, it perceivably (i.e. by more than a dB or two) boosts frequencies only over 15 kHz, which most people cannot hear anyway. It does not boost frequencies in the 4-10 kHz range, which the equal-loudness contours suggest should be amplified as well. This explains why I did not hear any improvement of highs with the DEQ on.

I can hear the changes to the highs with DEQ, and my measurements show it starting at about 7 kHz, but it is more noticeable by 10 kHz:

Another finding is that the DEQ boosts the surround sound not just below 500 Hz, but over the entire frequency range into the highs, which causes an imbalance even with Audyssey corrections enabled. That explains why I heard the side surrounds playing too loud compared to the fronts and center with the multi-channel source material. The author says he therefore does not use the DEQ at all

This is true, but the solution to not use DEQ for that reason is silly, IMO. I simply reduce the surround channel volumes and I'm back to having a great, balanced sound bubble, and I don't lose the benefits of Dynamic EQ.

 

[-Matt-]

...just blasting me, as some other members keep doing, for not following their instructions down to the letter, incl. getting a subwoofer, etc.

Reading back through this thread I can see why @lc6 might feel a bit under attack. At the same time I think most members are just trying to make suggestions to help you to get the best from your system. (This has been a long road for many of us who have gotten used to Audyssey's quirks over the years).

So, I want to thank @lc6 for sharing their impressions of the 3800. Many of us are keen to hear any news of this new range of products. (And I don't disbelieve that you are unsatisfied with your current bass sound). Thanks for taking the time to write down your impressions and for engaging in the subsequent discussion.

The part in the subjective review that has caused controversy (IMHO) is concluding from the listener perceived bass (after first setup) that the power amplifier sections are weak or incapable of driving your speakers. (It isn't impossible that they are). But there are so many ways to mess up the bass in the configuration that this should be checked first before making such a claim. In the past Denon have proven to be pretty competent at amplifier design; so it will likely require proof in the form of AP measurements for many of us to believe that this model has such a flaw.

 

[peng]

I measured both on REW with my old X4500H and flat removed MRC completely. Not sure if things have changed but I know it wasn’t present a few years ago.

I don't know, may be you are right, or not.... I also plotted such curves but it was a few years ago and I don't remember the results anymore. If your measurements for both flat and ref were done under the exact same conditions, then you must be right (see edit below). Answers to such questions on forums are as expected, not consistent, but I thought I read from Ask Audyssey before that flat simply means no high frequency roll off, other than that it says the filters would work the same way as the reference curve, and that MRC can only be disabled with the Pro software and the Editor App (obviously the new MultEQ X too).

Again, can't argue with your curves. Though it doesn't matter either way for those who use the App for other changes because most (if not all) the changes made won't be in effect if the flat curve is selected.

Edit: Just found a thread in which multiple posts reported MRC is on for the flat curve as well. I think I am going to do some measurements and keep the plots on file this time.

This is the linked to that thread:

audyssey flat or reference, subwoofer

The app worked like a charm on my iPad. In fact I felt it easier. I just plugged in my room dimensions to find my Schroeder frequency for my room, let Audyssey donits thing like always then made my alterations. Not EQing the full spectrum has done wonders in my room. The Audyssey Ref and Flat...

forums.audioholics.com

I just finish checking it with REW/Umik-1 and can confirm MRC is active even with the flat curve, as reported by others. The only way to disable it is to use the Mobile App, Pro, or MultEQ X. No idea why your REW plots show flat has it disabled!

 

[Keened]

I think Audyssey Dynamic EQ only adjusts its boost based on the master volume setting, and if you think about it it makes sense. It would be truly awful if the frequency response constantly changed depending on whether the currently playing music was in a loud or quiet passage.

I posted some measurements on the effect of Dynamic EQ settings here. Don't know if you will find then comprehensive enough (feel free to post your own) but they serve to illustrate what the feature actually does.

It's Perceptual Volume Leveling (DEQ) vs Absolute Volume Leveling ('dynamic compression').

A lot of people are looking to solve two problems: maintaining 'good sound' at lower volumes which means adjusting the frequency response proportionally to the master volume and fixing poorly mixed program material or highly variable program materials (explosions/whispers/bad vocals or a multi-album playlist). Unfortunately the only way to seamlessly mix the two is to process the program material in full before running it through the DEQ. You can use geometric mean gain adjustments but without future knowledge it may still give a bad result at times.

 

[GalZohar]

If you find Audyssey boosting highs too much (with Audyssey on vs off, no dynamic EQ), then you should try limiting the EQ range. You should probably try that even if you don't find Audyssey boosting the highs too much. Let Audyssey EQ only the bass, then use Dynamic EQ to get the bass levels back.

Note that even at -15, if your master volume is not at least -15 (you should really use relative levels and not absolute, it's much easier to compare), then you are still boosting bass compared to DEQ being off.

In any case, most of your experience seems to be due to how you use the room correction system and the settings, rather than the "Denon sound" (which really is as flat as we can hear, at least according to measurements of all previous Denon models). You could get the same results even with Dirac if you set it to give those results. Unfortunately, Audyssey isn't the "fire and forget" kind of thing they market it to be, it is at least as complicated as the more expensive competitors (say Arc or Dirac, not talking about super high end), if not more! You should really learn how to use it in order to get good sound.

 

[Jon AA]

Can you explain this part?
I don't understand the pictures either.

Sure. The way Audyssey sets the levels between the speakers and subs, really assumes a target curve that is flat in the bass (the only way it was available for many years). When it's not flat, things need adjustment. In short, when you add a bass boost to a speaker channel, it boosts the bass per your directions above zero. When you add that same bass boost to your sub curve (which you should), it lowers the level of the sub such that your boost is at zero instead of above it.

So, in the picture I showed with a target curve having a 5 dB boost in bass, the speakers at low frequency are at +5, the sub is at 0. So, when you do that you need to increase the sub level 5 dB after calibration so the target curves match and the levels will match at the crossover point (wherever you put it).