Set Dynamic Volume and Dynamic EQ to ON or OFF ?

m. zillch said:

Two Three questions about that:
A. Does Audyssey and/or Sound United concede that they indeed do this and if so do you have a link?
B. Assuming they do this, do they say why exactly, or is it just a major blunder they have never corrected year after year? [I'd assume it is true of all years of production, not just recent.]
C. Why is their error described as "3 to 6 dB" in the video, or "as much as -6dB above" ? Sure I can see how due to some glitch in the software it might theoretically always de-emphasize some output channel by some fixed dB value, but I'm scratching my head as to why the error it makes is a variable amount.

Thanks!

Click to expand...

I made a big point about Audyssey’s post calibration bass volume problem at a press event. I said flat out this issue needed to be fixed and you can’t assume people will use Dynamic EQ.

The other issue I made a point of was the need to load multiple calibration profiles in the AVR to switch them. They listened to my second comment with the new 2021 models. Here’s hoping that the Audyssey fix is next, but that’s an Audyssey not Sound United issue 😁

Audyssey assumes users will use dynamic volume—that’s seemingly the root cause. I think it’s fair to say this is their philosophical angle but not one that everyone would agree with. The bass issue is consistently measurable post calibration on an SPL and with one’s ears.

I’ve seen this across all Audyssey versions on Denon and Marantz review units I’ve reviewed and so has Gene. I presume it’s prevalent on other brands too over the years.

Gene and I can speak to what we’ve measured in our respective rooms across multiple review units and what others have related. There are lots of threads here on AVS forum about it as well.

The variability is consistently 3 to 6 db lower. While one would naturally presume it would be a consistent volume difference, it’s not. Remember also that people have different subwoofer configurations such as one, two, or four. This is strictly a post calibration issue and since everyone’s rooms are different, sub placement is different, and the curves/filters able to be applied are different that is the most likely source for the variability. When looked at from that angle, a 3-6dB range then makes total sense and it’s probably related to the amount of gain that Dynamic EQ would be applying to compensate for bass in that users particular room.

TheoN said:

I made a big point about Audyssey’s post calibration bass volume problem at a press event. I said flat out this issue needed to be fixed and you can’t assume people will use Dynamic EQ.

The other issue I made a point of was the need to load multiple calibration profiles in the AVR to switch them. They listened to my second comment with the new 2021 models. Here’s hoping that the Audyssey fix is next, but that’s an Audyssey not Sound United issue 😁

Audyssey assumes users will use dynamic volume—that’s seemingly the root cause. I think it’s fair to say this is their philosophical angle but not one that everyone would agree with. The bass issue is consistently measurable post calibration on an SPL and with one’s ears.

Click to expand...

I presume you meant that Audyssey assumes users will use Dynamic EQ, (DEQ), not Dynamic Volume, (DV). DV, is an entirely different technology than DEQ. DV is basically variable dynamic range compression tied to the Master Volume Control. It doesn't change frequency response.

TheoN said:

I’ve seen this across all Audyssey versions on Denon and Marantz review units I’ve reviewed and so has Gene. I presume it’s prevalent on other brands too over the years.

Gene and I can speak to what we’ve measured in our respective rooms across multiple review units and what others have related. There are lots of threads here on AVS forum about it as well.

The variability is consistently 3 to 6 db lower. While one would naturally presume it would be a consistent volume difference, it’s not. Remember also that people have different subwoofer configurations such as one, two, or four. This is strictly a post calibration issue and since everyone’s rooms are different, sub placement is different, and the curves/filters able to be applied are different that is the most likely source for the variability. When looked at from that angle, a 3-6dB range then makes total sense and it’s probably related to the amount of gain that Dynamic EQ would be applying to compensate for bass in that users particular room.

Click to expand...

Audyssey calibrates the system for flat in-room response at Reference Level. The following "After "graph shows a flat response, (other than a remaining null at 80 Hz that is uncorrectable by any EQ, Audyssey included.) The point is that the "target curve" for Audyssey is flat in-room response. This graph is completely representative of every Audyssey result I've ever achieved in my own experience.



At any level below Reference Level, the bass will be perceived by the listener to be "too low" due to human hearing deficiencies at lower frequencies.. This is were Dynamic EQ compensation comes into play.



DEQ is Audyssey's interpretation of the Fletcher-Munson curves, (aka "Equal Loudness Compensation", or ELC). Here are the currently agreed upon set of ELC curves, (red), which, as you can see, are somewhat different than the Audyssey DEQ compensation:



Audyssey calculates trim levels based on the 1st measurment position, but it also performs "Volume Normalization" to compensate for the sum total of the boosts or cuts the EQ has implemented. The final subwoofer level setting represents "unity gain" of the subwoofer channel output.

Whenever I have run Audyssey, I always measure the SPL post-Audyssey at the 1st measurment position at full Reference Level, and it always results in the subwoofer level being equivalent, (+/- 1 or 2 dB), to the other speakers. It is mandatory to measure this with an external test tone source. The receiver's internal test tones bypass the Audyssey EQ's, and are not representative of the post-Audyssey result. If the internal test tones are measured post-Audyssey, it would not be at all unusual to find the levels 3 or more dB too low. However, using external test tones, and exposing them to the Audyssey EQ filters, they are much closer to the speaker levels, which is the ONLY thing that matters in terms of relative level calibration. The only time absolute level measurements are important is if there is a need for absolute calibration to Reference Level. In that case, an external measurement system with a finely calibrated microphone would be far better than the variability of the Audyssey mics, (which are "batch calibrated" to +/- 3 dB, the full range of which is equivalent to the 6 dB range you found in your measurements.)

Also, Audyssey integrates the levels of the full bandwidth of frequencies in the subwoofer range, measured using a "frequency sweep test", whereas most subwoofer level test tones, (internal and external), use bandwidth limited pink noise, limited to 40 - 80 Hz. This alone can have a significant impact on the level trim results vs. "measured SPL". To accurately measure the SPL of the full bandwidth of the subwoofer channel, a wide bandwidth pink noise test tone should be used.

I don't follow Audioholics, so I'm not aware of your measurements or how you perform them, but mine certainly don't follow the pattern you stated. I also use DEQ all the time, and never engage DV.



Craig

So I've learned from TheoN's most recent post that Audyssey does not concede this "underestimated by 3 to 6dB subwoofer output level error if not engaging DEQ" issue is real, as I suspected. Craig has pointed out largely where I was going with this.

I've always assumed the reason people not using DEQ often like to manually dial in 3 to 6 dB more sub level is because most people don't listen at Reference level, 0 dB, (its too loud in my home at least) so due to Fletcher-Munson equal loudness curves we perceptually hear it as being "bass shy" otherwise, even though it technically isn't.

Solutions?

• turn up the volume to 0 dB
• turn up the sub bass level to taste
• engage DEQ

If, as I understand it, Audyssey intelligently tells the AVR/AVP it suggests certain cuts, prioritized over boosts [in order to preserve amp power and cone excursion, minimizing distortion in both], so the new average range output decreases by say 3 to 6dB, yet this reduction is not compensated for in matching to other channel levels, I guess that could be an issue but I'd find it rather strange that people like us in forums figured this out but Audyssey never has.