Do we know definitively yet whether the 6 other channels were silent?
And also if bitstreamed audio has this issue? For instance, a Dolby TrueHD or DTS Master Audio 7.1 track downmixed into 5.1.
As far as I can tell, the issue was only present when sending 8 channel PCM from a Windows machine to the Denon to output as 2.0.
I'm also not clear if the Denon was set up as 7.1 but 'Pure Direct' forced the down mapping to 2.0. Or if the Denon only had 2.0 speakers configured.
Excellent. This is asking the right questions. The quick answer is the limited (as practical) testing that ASR does exposed a symptom of the problem but isn't sufficient to determine the entire scope of the problem to say where it would or would not affect.
We really don't know the root cause of the problem to confidently predict the scope of where the problems might manifest. It could be more than the scenario that was in the test or very limited. Denon would need to do a far more comprehensive set of tests to isolate the problem that is beyond the scope of ASR.
Bottom Line:
It would be impractical for ASR to test all of the possible combinations to narrow the problem down and establish the scope of the problem. Denon would need to do so and provide a fix. But if they are
not going to be transparent about it, then consumers (outside of the fanbois club) would have a dilemma.
1. Is there a firmware fix to this? Not all firmware can be updated. For example, if the root cause is an issue with the upstream HDMI receiver in processing the channels as received as part of the 2.1 protocol (or there is a hardware issue), that may not be fixable by firmware update but might require a card swap (I believe NAD did this for one of their cards).
2. Is Denon going to fix this for future units but let it slide through for current units sold if they think it is inaudible to most if not all even with the problem? Would one have a philosophical objection to this approach?
Personally, if I was looking for an AVR and this was in contention, I would skip or hold off until there was an explicit firmware update that even hinted at a correction related to this issue and not purchase (at this price level) with that cloud hanging over it. But that is just me, not trying to encourage or discourage others.
Technical discussion for the so-inclined:
A very high-level logical flow of the processing would be
Receive HDMI protocol data and split into up to 8 streams of data -> Process data and meta headers to check for encoding and decode into appropriate number of PCM streams (1-16) -> DSP Processing for up/downmixing -> Muting of channels (speakers/pre-out) set as not in use. I assume that what channels comes out of pre-outs is exactly what would come out of the speakers and so they are both muted the same way for unused channels.
I have left out additional steps that may happen such as artificial synthesis of surround data to create additional channels when not part of the original content (the different DSP synthesized surround modes).
The problem could be in any one of those steps consistent with what has been observed. But the scope of the problem where it could affect not just the L and R but also other channels has not been established and not practical to do so for ASR.
All we know is that if the data is received as 2+ channel PCM (not any encoded streams) and we force a down-mix to 2 channels, there is noise introduced in those channels. We also know that if it receives just two streams (of PCM), it does not show the symptoms.
We don't know if there is noise introduced in the other channels in that scenario but muted downstream (it could be muted to the amp but not the pre-outs or to both etc).
If the bug is in HDMI receiver near the ports where it introduces spurious data into the other channels as received, it could have implications in many more cases (or not). Whether those would be audible or just show up in measurements is a different question.
If the data is detected as two channel bitstream encoded (say EAC3 or DTS 5.1), then the other channels could be ignored even if corrupted and the streams processed correctly with downmixing. But show up in all PCM processing where there could be noise in all channels (but potentially muted).
If the corrupted data in other channels affects the determination of the encoded stream (say 2+ channel encoded 5.1.4), it could impact the "cleanliness" of what comes out of any of those channels with or without down-mixing (whether audible or not).
If the problem is in the determination of whether the channels are PCM or a bit stream encoding, then there could be issues even when encoded bit stream is sent (although this is likely to have far more audible impact than some additional noise).
If the problem is in the down-mixing algorithm, then the problem would manifest even in encoded streams sent to the unit and beyond HDMI (for example DTS 5.1 sent through the optical to a 2.1 or less speaker configuration).