Here is an explanation -as far as possible- of the various Q parameter settings.
Q1 - Good old DAC influencerIs there since August 13 2007.
Influences incoming jitter at the DAC and the analogue side of the DAC.
Generally a lower setting makes the sound more -as perceived- "accurate" while a too low setting may let music sound too thinny.
The outgoing data stays 100% the same as the source ("bit perfect"). Of course this is outside bit mangling stuff like upsampling.
Q2 - Is believed (!) to influence outgoing jitter.
Is, together with Q3, Q4 and Q5 there since Januari 25 2009 in 0.9w-9, and the main controlling phenomenon for Q3, Q4, Q5. Again all "bit perfect".
With Q2 at 0, nothing changes opposed to Q1 alone.
Q2 (and the others) emerged from pure theory and the experience of a few years people reporting about unintended SQ changes. Thus, from the experience and learning what happened at these unintended SQ changes, now an explicit parameter is there for influencing SQ in "that" area.
The higher the setting of Q2, the higher the influence will be, although certain settings may cancel out like the cancelling waves (opposed to standing waves). Similarly, a certain setting may add up exponentially (like standing waves).
When the Q3, Q4 and Q5 parameters are left at 0, Q2 still influences, though the most mildly.
Watch out though : Even with Q2 at 1 (the lowest zetting when active) this is exactly what could happen at an unexpected SQ change.Q2 has been deprecated since version 1.186.
Q3 - Dictates the interval Q2 should to its job
Btw, the exact job of Q2 remains a ... 
.
This "interval" is not about some rational thing, but the higher the slider, the more Q2 becomes active.Q3 works completely different per version 1.186. See the ToolTip on it for extensive decription and outlay.
Q4 - The strongness of the influence
The higher the number, the stronger the influence of Q2. So note that in combination with Q3, the stronger influence happens at smaller intervals.
Keep in mind : this is just influence, and nothing like increasing sound quality. Whether SQ increases or decreases depends largely on your computer system. Also, what aspect of SQ changes is also dependent on your system.
Note : The 0 setting still influences !Q4 works completely different per version 1.186. See the ToolTip on it for extensive decription and outlay.
Q5 - Resonating the influence (obsolete per 0.9z-9)
Where Q2 acts upon request of Q3 with a strength of Q4, Q5 varies Q3. In other words : the way this is setup allows that the influence can "resonate" upon itself. You can see it so that Q3 is varied by the amount of Q5.
Careful : This red text does not indicate danger, but the importance of this Q5 parameter. Example :
When Q2 is set very low this means that the influence is low. However, think in higher level frequency here. The lower the amount of influence (and this works together with Q3) and the higher the force (which is Q4) the slower the resonance will occur, and the longer it will last. THIS MAY IMPLY AUDIBLE SQ CHANGES THROUGHOUT TIME (meaning each 5 or 10 (or who knows) seconds a change in SQ. Of course this is the most wrong thing to happen. Therefore generally you could say that the higher Q2 and the higher Q5 the more frequent the resonances will occur and the shorter they will last. The main factor here is Q2 itself, and e.g. a Q2 setting of 3 with an Q5 setting of 30 won't help much at avoiding "bad" resonances.
Also, a Q5 of 0 won't help, because the resonances are inherently there.
This too is just theory so far, and might you find an "uneven" SQ at certain intervals which you can't solve, the best solution would be not to use Q2 at all (nothing operates then) or set Q2 at 1 and work only with higher settings of Q3 plus Q4 to your likings.
Q5 - OS Timing (New per 0.9z-9)
Q5 now is a SQ influencing parameter on itself with special purpose.
Hard to explain (or a parameter in secrecy otherwise) but Q5 now works on the 'timing' of the Operating System and it will do a few things on 'granularity' on one side and on noise influence on the other. It works in conjunction with the Clock Resolution setting and implies different effects for when that is set to 'Nothing' versus any of its values.
Clock Resolution set to Nothing :This implies a random treatment of the physical transport of the audio data to the DAC; It may smear otherwise audible microsecond level pulses to a more random nature. SQ can get less 'nasty' of this.
Watch out : The effect of this is infinitely higher for Windows 8 than for Windows 7, and actually the parameter was originally made for W8. For W7 it has its effect too though.
The lower Q5, the more the smearing effect will occur.
Clock Resolution set to any of the discrete values Including 'Minimum for this system') :Underlaying to the above workout of the parameter is a general CPU core 'Task Switching' behavior; Without the 'DAC physical transport' behavior implied by the 'Nothing' setting as described above, the 'Task Switching behavior' will be sustained with the Clock Resolution set to a value other than 'Nothing'. This is regarded as more important for SQ than the 'DAC physical transport' behavior, which is only driver related. And, although the 'Task Switching' behavior is present in both settings, having it active on its own can be important because now the DAC/Interface driver is not influenced. In this case, the lower Q5, the less Task Switching will occur and the higher settings (like 20 and above) will workout to the parameter not being active at all.
In either case (Clock Resolution set to Nothing vs a discrete value) all will not engage when Q5 is 0. With Phase Alignment set active, a Q5 value of 1 is not allowed, which you will be warned about when pressing Play. This is because the value of 1 is special (for either of the both settings) and sub ms intervention will occur, net resulting in a kind of 100% usage of the cpu core which serves the playback. This by itself implies (at least theoretically) another SQ, but continuous playback of the stream is not guaranteed. And, when the sound briefly stalls, the famous 'crack' of Phase Alignment will be your part; With Phase Alignment not set active you can happily try Q5=1, but notice this setting remains special and works out very differently than Q5=2 and beyond.
Lastly, notice that the set Clock Resolution will remain respected as alwas (as before); What you could try *IF* you before could not get satisfaction out of the Clock resolution of 0.5, is setting it to this 0.5 and set Q5 to 2. This most likely now implies an acceptable and better SQ (while the Clock Resolution thus still is 0.5ms !)
