DoP "PoP" and how JRMC gets around this
The problem is that when DoP stops there is no defined "level". After DSD playback via DSD finishes, the DAC output may be at any given of 64 Levels and as no more data is send it will stay there.
There is a 1 in 2 chance that the level the DSD DAC outputs is either more negative or more positive than nominal "maximum music level". There is a 3 in 4 chance that the level that remains at the output of the DAC is greater than half the peak level of the Music.
When then PCM Playback starts it starts with silence (empty samples as no data is yet sent).
This empty sample forces the DAC to output 0V. This happens so fast, there is no way to mute this. The DoP Playback has already stopped and PCM is being send while we are still trying to get the mute on.
This is the "big pop", it is simply an instant return from X (where X will have a very high chance of being at a very high level respective to the music, positive or negative) to zero.
The way to avoid (or at least substantially reduce) this is click is to play some extra DOP Samples (we are unable to determine exactly how many J-River actually adds as they are able to virtually eliminate this issue) representing silence after the nominal playback is finished. By playing silence the DAC output will return to Zero, so when PCM kicks in it will be at zero or very close, so we get no or a very small click.
The ASIO implementation of DSD incidentally takes care of that and hence has no click. In the case of other DAC Chips that convert DSD internally to PCM before filtering and playback this whole issue obviously does not happen!
As so far they represent 95% or more of all DSD DAC's out there, this problem is hence quite rare.
We hope this sheds more light on this matter.