Big news coming from Sound United in 2023!

[GXAlan]

Tangentially: I've been curious why not go back to AKM? Perhaps AKM isn't ready to meet demand for large production runs, yet? Last recent news I saw is that they were operational and had released some new product as well...

Any insiders have an AKM update?

Not an insider but the reports were that AKM were resuming production earlier this year with volume shipments in “third quarter” of this year. So the products under design now would probably have AKM chips. So it has to be the X3900 or X4900 before we might see AKM involved again.

There is a danger than the PCM5102A is cheaper, doesn’t affect sales, improves profits… and they end up permanently downgrading the DAC.

 

[ryanosaur]

with volume shipments in “third quarter” of this year.

That would be the piece I didn't see (or catch) from the reports I read. Thank you.

It will be interesting to see what companies choose to do as AKM product becomes readily available again. I think for some companies, they will choose less expensive parts as long as it is not detrimental to performance, but others will want the best measuring no matter what and AKM has that reputation.

Eager to see measurements for this new generation of gear!

 

[DerRoland]

There is a danger than the PCM5102A is cheaper, doesn’t affect sales, improves profits… and they end up permanently downgrading the DAC.

This can be true, the „accident“ in the 600/700 generation, that the low price range had the same sinad as their sota 8500 can‘t been their intention.

 

[CyrusTheGreat_600BC]

There is a danger than the PCM5102A is cheaper, doesn’t affect sales, improves profits… and they end up permanently downgrading the DAC.

It won’t affect sales of x3800+ series, since supplies and stocks were low for the past two years and price was high, and many people did not upgrade because of that. Even with the increased prices, with the current inflation people are getting used to the new norm, and x3800 will have a record sale compared to 3700. It will be a shame if Denon won’t see this fact and would keep putting the lesser chips in for x900. By then, the competition will catch up to them in this market segment unless another stronger than Covid virus comes out.

 

[WHQL]

The A6A has a bad error where the center channel has low SINAD. The front channels are competent.

I am really seriously considering the A8A and measuring it. YPAO RSC has the equivalent of Trinnov for sound field mapping even though it’s actual EQ and control is rudimentary.

I had the Sherwood Newcastle R972. I am sure SINAD is low but the 3D sound field made up for it

The A8A checks a lot of boxes for me too. Would be great to have some real measurements. If the price drops down below 3K, I would seriously consider it. I cancelled my X3800 preorder right before it shipped and preordered the X4800 instead. I am content to wait and see how the X3800 shakes out after the software matures and there is an experienced user base.

 

[Newman]

I don't have either, but we don't know already that the AV10 has at least the following advantages:
- 4 independent subwoofer outputs vs the AV8805's 2
- Dirac Live support (supposedly "full support" whatever that means, via FW upgrade/update
- Reference class near flagship ESS DAC IC, ES9018k2M vs the original AV8805's ex flagship AKM AK4490 and the newer after factory fire AV8805's average class ES9010K2M
- ALL inputs supporting 8K / 60Hz and 4K / 120Hz
- improved HDAM version (same version number, but improved, same as those in the SR8015), note that the SR8015 is the only Marantz that measured almost as good as the Denon AVR-X8500H even in non preamp mode.
Others will hopefully add to the list.

Thanks peng. Nobody has added to the list yet. I can add 15ch instead of 13ch.

So, nothing new in terms of streaming support, Dolby/Auro/DTS surround or 3D audio tech, no new Dolby Vision/HLG/HDR support? These areas are not my forte, so maybe there is nothing new to include, lol.

cheers

 

[Chromatischism]

So, nothing new in terms of streaming support, Dolby/Auro/DTS surround or 3D audio tech, no new Dolby Vision/HLG/HDR support? These areas are not my forte, so maybe there is nothing new to include, lol.

Nothing to add with video, but a lot of people seem to be looking for DTS:X.

 

[Jon AA]

Can you show the results of a specific speaker, once with flat curve and once with a 400Hz shelve filter with 5-10db boost, when the EQ is limited to 200-300Hz? Or specify which 2 graphs show that.

Sure. The first two were close but this will simplify. Flat target:

2nd order shelf at 400 Hz, 8 dB:

You can see EQ'd full range, there is a slight drop in overall level, but not enough to bring the average of the entire curve to zero. Limit the correction to 200 Hz and:

You definitely do loose some boost as you describe. But in this case, I'm not sure what you'd want it to do. You don't want a brick-wall dropping off a cliff back down to zero where you limit the EQ. You can see it's maintaining the same 2nd order shape and slope, but it's missing out on all the boost that should have been building from 1000 Hz down to 400 because the EQ is turned off in that area.

I'm not sure why one would specify a large boost centered at 400 Hz and then cut off the EQ at 200? What should it do? My suggestion would be if one wants to run a target curve with such a large boost in the lower midrange, you need to actually let the EQ work through the lower midrange.

Conversely, if the bass boost is added down where most people would want to add it--centered at 120 Hz for example:

Here you see EQ'd full range, since the boost is gone by 400 Hz, the top end stays at 0, and the entire curve is clearly at zero or above:

Limit the EQ to 200 Hz and it makes very little difference:

Just a tiny difference in boost at the bottom (likely due to the large dips at 600 and 700 of the uncorrected signal bringing down the average in the 400-2K range), but small enough to hardly be measurable.

But again, if you look at all those curves, it looks like Audyssey put the average level of the 400-2K range at zero.

And like I said in the beginning, for the main channels you can do whatever you want with the target curve below 400 Hz without worrying about changing the overall level. This is not true for the sub channel, which is why it will require a level change after calibration if a target curve like this is run. The sub channel for comparison with the 120 Hz shelf:

 

[peng]

Nothing to add with video, but a lot of people seem to be looking for DTS:X.

AV8805 has DTS:X, and iirc, DTS:X Pro upgradable via FW.

 

[Dj7675]

This is not true for the sub channel, which is why it will require a level change after calibration if a target curve like this is run.

Have you heard if Audyssey is planning on ever applying sub level adjustment automatically?

 

[GalZohar]

Sure. The first two were close but this will simplify. Flat target:

View attachment 236433

View attachment 236434

2nd order shelf at 400 Hz, 8 dB:

View attachment 236435

View attachment 236436

You can see EQ'd full range, there is a slight drop in overall level, but not enough to bring the average of the entire curve to zero. Limit the correction to 200 Hz and:

View attachment 236437

You definitely do loose some boost as you describe. But in this case, I'm not sure what you'd want it to do. You don't want a brick-wall dropping off a cliff back down to zero where you limit the EQ. You can see it's maintaining the same 2nd order shape and slope, but it's missing out on all the boost that should have been building from 1000 Hz down to 400 because the EQ is turned off in that area.

I'm not sure why one would specify a large boost centered at 400 Hz and then cut off the EQ at 200? What should it do? My suggestion would be if one wants to run a target curve with such a large boost in the lower midrange, you need to actually let the EQ work through the lower midrange.

Conversely, if the bass boost is added down where most people would want to add it--centered at 120 Hz for example:

View attachment 236438

Here you see EQ'd full range, since the boost is gone by 400 Hz, the top end stays at 0, and the entire curve is clearly at zero or above:

View attachment 236439

Limit the EQ to 200 Hz and it makes very little difference:

View attachment 236440

Just a tiny difference in boost at the bottom (likely due to the large dips at 600 and 700 of the uncorrected signal bringing down the average in the 400-2K range), but small enough to hardly be measurable.

But again, if you look at all those curves, it looks like Audyssey put the average level of the 400-2K range at zero.

And like I said in the beginning, for the main channels you can do whatever you want with the target curve below 400 Hz without worrying about changing the overall level. This is not true for the sub channel, which is why it will require a level change after calibration if a target curve like this is run. The sub channel for comparison with the 120 Hz shelf:

View attachment 236441

View attachment 236442

I was getting different results, I will have to try again when I get the chance. I checked with my android app but it was messy, as it is only Audyssey XT. With XT32 (from my friend's file) I see similar results to what you get. Maybe he messed up some setting in the MultEQ-X?

 

[lc6]

I would like to share additional findings about the 3800H.

I conducted additional power measurements. First, I let the AVR warm up a bit because my previous measurements showed that its power consumption changes over time as its temperature rises. The mains was 120 V / 60 Hz, as measured by the power meter. When idle, with the ECO On the AVR consumed about 46 W (fluctuating a fraction of a Watt) and with Off about 76 W. Then I went to the Speaker Connection in the setup and changed all settings from the default Speaker + Pre-out to Pre-out Only. This had no effect on the measured power consumption, which implies that power amps are simply disconnected from the built-in preprocessor and continue to be fully biased and consume power for no good reason, even in the ECO On mode (while this finding was already speculated about in other posts, the actual measurements prove it).
Intuitively, these amps should be switched off (by means of disconnecting power via a relay or, better yet, by a signal to disable at least their output stages so there is no bias current flowing through them). That would be especially beneficial with the external Class D multichannel amp connected to the AVR.
So it looks like the only option to reduce power consumption in pre-pro mode is to turn ECO On. Incremental savings could be realized by switching off the front display, but that only works in Pure Direct mode. Now, while many owners do not care about AVR power consumption but only great sound, I'd suggest they consider the current geopolitical situation. Winter is coming.

After I restored the Speaker Connection settings, without entirely exiting Setup I went to the Audio > Audyssey menu to experiment with settings. When I entered the dialog screen, the MultEQ XT32 setting was set to Reference, the Dynamic EQ was set to On with the Reference Level Offset of 0 dB, and both the Dynamic Volume and Audyssey LFC were set to Off. I selected the MultEQ XT32 setting and with the right arrow on the remote selected the L/R Bypass. To my surprise, this also automatically turned on the LFC!
I noticed once before that when I entered that setup screen from scratch, the LFC was set to On even though I was quite confident I did not turn it on or mess with it at all. But I have never observed this happen afterwards even though I was playing with the MultEQ XT32 settings quite a lot to compare their effect on bass. So the event today (which I could not reproduce despite several attempts) proves that this AVR may randomly enable the LFC, perhaps by misreading the saved configuration. This could explain why I keep having trouble with the deep bass. Now, this could be a result of a software glitch or a problem with my particular early-production unit (I suspect the former). Time will tell as more people will get their hands on the 3800H. But I wanted to report this here in case other owners notice the same problem.

Then I manually adjusted channel levels using the built-in noise generator (aka Tone Generator). I used a calibrated SPL meter with C-weighting and slow averaging in the 60 dB range, placed at the exact same primary listening position as the Audyssey's #1 mic position. The result was that while L/C/R levels were unchanged, the side L/R surrounds changed from -2.0/-2.5 dB to -3.0/-5.0 dB, and the L/R front heights changed from -2.0/-3.0 dB to -3.0/-3.5 dB (there is another software issue: in the automatic measurements and in the manual setup, the order of surround and height L/Rs is, strangely, reversed!). That explains why I perceived the surrounds playing too loud after the Audyssey correction. Also, I had to separately bump up the center channel by +3 dB on the remote's Option pop-up menu to make the dialog intelligible.

 

[Chromatischism]

I'm glad you found the cause of your bass problem. But let me caution you about those internal test tones.

The pink noise there does not use the Audyssey filters. That means you're getting completely uncorrected levels, which are not representative of your room correction in use. Further, the tones used by Audyssey are a little different than standard pink noise -- they claim it is better for calibration.

This means the Denon levels menu pink noise is really only for an uncorrected system. Alas, useless.

Finally, the spectrum and weighting you are using on your SPL meter is probably different than what they are using.

 

[LaserGuruGuy]

Chris Kyriakakis (CTO of Audyssey Labs): The best approach is to use a system that is based on impulse response measurements. Swept-sine or log-chirp or MLS test signals are sent to each speaker and the response is measured. After deconvolution of the measured response with the original signal you obtain the impulse response of each speaker in the room. There are ways to average these in the time domain or in the frequency domain.

That’s from 2014 but still pretty much what they do. The receiver does the chirp and the deconvolution. The apps or receiver then compute the taps for the fir filter.

 

[lc6]

I'm glad you found the cause of your bass problem. But let me caution you about those internal test tones.

The pink noise there does not use the Audyssey filters. That means you're getting completely uncorrected levels, which are not representative of your room correction in use. Further, the tones used by Audyssey are a little different than standard pink noise -- they claim it is better for calibration.

This means the Denon levels menu pink noise is really only for an uncorrected system. Alas, useless.

Finally, the spectrum and weighting you are using on your SPL meter is probably different than what they are using.

I disagree that the "Tone Generator" function is "useless" because the noise is unprocessed. The whole point is to compare the channel levels Audyssey selected after measuring the system with its uncalibrated mic to those determined by a calibrated SPL meter. Sure, the source noise and C-weighting spectra may differ, but I can say that after the level adjustment based on measurements with the SPL meter, the system sounds a more balanced with a sample Atmos source material, although the surrounds still seem to run too hot.

BTW, last night I observed for the second time a "glitch" I described previously: while streaming a movie source material, and without touching any remote controls, the TV suddenly declared that the AVR was off, so it switched the source to itself, started streaming a news channel and playing sound through its internal speakers. But when I manually switched the source back to the AVR, I was able to continue to watch the movie. The eARC connection between the AVR and TV is via a HDMI 2.1-capable 3-ft cable, so this is not an issue. And I do not believe this is a problem with the TV because I have not observed this issue when evaluating the RX-A6A. This "glitch," again, could be caused by a firmware bug or a problem with early-production hardware. I just wanted to report it here in case some other owners of the 3800H experience it.

 

[Chromatischism]

I disagree that the "Tone Generator" function is "useless" because the noise is unprocessed. The whole point is to compare the channel levels Audyssey selected after measuring the system with its uncalibrated mic to those determined by a calibrated SPL meter.

It is useless because what you hear isn't what you're getting.

 

[bo_knows]

It is useless because what you hear isn't what you're getting.

So Gene at Audioholics and other YouTube reviewers are wrong? I swear I thought he mentioned in one of his videos that he doublechecks the levels manually with the SPL meter after Audyssey calibration. Don't mean to "stir up the pot" or anything like that but just for my own understanding.

 

[Chromatischism]

So Gene at Audioholics and other YouTube reviewers are wrong? I swear I thought he mentioned in one of his videos that he doublechecks the levels manually with the SPL meter after Audyssey calibration. Don't mean to "stir up the pot" or anything like that but just for my own understanding.

He's not wrong. In fact after I posted measurements proving the Levels menu tones bypass EQ, Gene mentioned such to Jeff from Audyssey in one of their videos, who concurred.

If he's using an SPL meter he must be using pink noise from a source external to the AVR. It's still not perfect, but it's a lot better.

 

[lc6]

Why is it external pink instead of white noise -- it is imposing a heavily-shaped spectrum on the C-weighting curve of the RadioShack SPL meter (shown in one of his videos)?

Anyway, back to the basics: if, after Audyssey measurements, the 3800H displays certain speaker levels, I would assume that they represent averages from which relatively small FR corrections are made. So over a wide spectrum of internally generated noise, things should still average out and be comparable to the "unprocessed" levels measured by the external SPL meter in the Tone Generator mode. Not so?

 

[Chromatischism]

Audyssey uses a modified pink noise, so if you want to "check" its accuracy, you would need the same, and they haven't released it. They say they plan to.