Hi Nick,
When I had a 4010 & AVP I found no difference when bitstreaming with DL4 on or off.
LPCM was clearly better with it on.
I only had a couple of blurays with LPCM soundtracks though so DL4 only really got used for CD replay.
That's a fascinating observation, and gives a big clue about how the AVP works. It suggests that the AVP (like Onkyo and other processors) uses a different audio clock recovery / regeneration process when the processor is doing the audio decoding (though this obviously only works with compressed soundtracks like DTS MA, which are now the norm).
My hands-on experience is that player decoding sounds different to processor decoding, and good players sound better than poor players when they output LPCM audio (which is when the soundtrack is LPCM, or the compressed soundtrack is decoded by the player). This is where I found the A1UD and TUC Oppos to sound good, and the CM 3313 was clearly the best of the lot.
However, with bit-streaming players (processor decoding) the SQ was always very good, and I couldn't tell the difference between players. Bitstream from my cheap LG sounded better than LPCM from any big-bucks player. This suggests either that players aren't decoding properly (once true, but probably not now) or that processor decoding is less susceptible to jitter. That's the empirical conclusion, but explaining and justifying it is more difficult. See my blogs for all the reams I've written about this over the years. Better still read this thread, which tries to explain how sources can sound different, and in particular how sources can sound different for reasons other than jitter.
Advanced topics in HD audio
(My take on this is that ultimately it is about jitter, but only at the point where it really matters – at the DAC itself. When people talk about jitter, they usually talk about the jitter on the digital audio signal – on the wanted signal. My view is that there are all sorts of other unwanted signals from equipment that can affect the final output. (For example, play a CD on an AV system, then completely disconnect the PVR/TV/VP/PJ, and see if that sounds different).)
The clear consensus from discussions with several audio experts is that bitstreaming should make no difference, as amplifier processing is the same. These experts included Bob Stuart, Charles Hansen, John Dawson and Amir Majidimehr – in other words, the people who best understand this, bar none. So that sowed a few seeds of doubt…..
Where I went from there, is that they can only speak for their products, and all receivers and processors don't necessarily work the same. The HDMI Forum is very specific about how the player's HDMI interface works, but leaves much more flexibility for the receiver. I suppose that some receivers use the same audio clock recovery process for both bitstream & LPCM (which is player-controlled dynamic rate decimation of the TMDS (video) clock from the player – a ghastly way to do it). And I also suppose that some receivers have a different clock recovery process for bitstream (like using a phase-locked loop, and ignore the rate decimation) to maintain synchronisation. That would help the receiver to isolate itself from the effects of the HDMI video clock jitter. This is how I explain to myself how bitstream can make the best of all players; its presumably how Onkyo do it, and now I guess that Denon and Marantz do, as well.
That's where I've come from; now here's the bit that's relevant to this discussion. What are the implications for the AVP?
Scenario 1: If the AVP uses video clock decimation for both LPCM & bitstream, both inputs will sound the same, and different players will sound different over HDMI. Denon Link 4, however, will sound better – better even than CineMike players. This isn't consistent with what Jason said, though.
Scenario 2: If the AVP uses a different process for LPCM & bitstream: In this case bitstream will sound better than LPCM, and it may even sound as good as DL4 (though I can't say for sure). Logically, a CineMike player that is bitstreaming to an AVP may also sound as good as DL4, but then, so will every player.
So which scenario is it?
Nick