XXHighEnd

Here is an explanation -as far as possible- of the various Q parameter settings.

Q1 - Good old DAC influencer

Is 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.

Q3 - Dictates the interval Q2 should to its job

Btw, the exact job of Q2 remains a ... :secret:.
This "interval" is not about some rational thing, but the higher the slider, the more Q2 becomes active.

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 !

Q5 - Resonating the influence

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.
:wacko: :wacko: :wacko:


Lastly : At the moment of writing this, not one single second listening to the results has happened. In other words, the working has been tested to be without errors (this can - and did happen without sound), so it is pure stupid theory as said above;
From here on "we" will start to learn whether theories may become practice. When nobody can perceive any difference (and watch out for the placebo pills !) theories are proved wrong, and at least this means of influencing will be removed out of XXHighEnd again. But :
This then just proves theories are wrong, knowing that unintended SQ changes happen, and it will be a matter of better analyzing what happened in between such unintended SQ changes. But that is for later. Maybe ... :)

Peter

You truly want to send me to the nut house don't you.

You bet. Better leave this alone. I just started with it myself (honestly), and set Q2 to 8 and Q3 to 24. The others at 0.
Why ? one must begin somewhere. Hahaha

Nope nope, I'll be jumping right in.
But you'll be hearing from my lawyers about my upcoming psychiatric costs associated with the Qs.

Maybe somebody can start some fund raising for this ?
The All Placebo Never Give It Up Fund might be a good name for it.

Btw, I'd swear I perceive a kind of aggressive sound from my Q2=8/Q3=24 setting. But that is way to soon for real judgement of course. Never too soon for a good Placebo though. :teasing:

Let me say that I did not understand a single word.  ;) 

But since I am jovial. Here are my Q settings

Q1=22 Q2=29 Q3=25 Q4=29 Q5=26

:)

Hi Peter,

I am a newbie with XXHIGHEND, but I think I will have a lot of good listenings with it :-)

Thanks for your dedication and time. You usually write long answers ;-) but I can grab my part of it from time to time. Thanks also for your patience against the "quite agressive" answers that some can throw at you in a few forums. Despite this, you keep on going with your beliefs, dedication and persistence !

I am a minimalist that like to keep things simple. I used JRIVER for the last year and to tell the truth, if there are things I can hear, there are others I can't.

I was not using DSPs except for the "fade out" thing (JRiver) that I like when I switch from the middle of a song to another. The cut is less harsh on the speakers...

All the components with switching power supplies are on one AC circuit and the rest on another. I change the process priority to "real time" for the program when running on the laptop. 

I have put a USB to SPDIF (Audiophilleo 2) between my laptop and my DAC (after trying the M2Tech HiFace first). I am happy with my sound system (Bryston Dac, Meitner preamp and amp, Tannoy System 15 DMT speakers). Part of what I have is not new (Meitner are not built anymore and the big Tannoy are from the nineties).

About XXHIGHEND. I have put all the Qs at 0 to get used to the software. I am not sure either about buffers and clock resolution. The Audiphilleo seems to do a very good job at controlling jitter with a very precise clock, but I will experiment with time (and read from the experience of other XXHIGHEND users).

The memory thing itself has made its way to me and I am also interested with the "ringing" thing (I still have to read about this to better understand it)

I have a question about Qs... Are they alterating the jitter (plus or minus) ? This is my first question, but a lot more should be on their way ;-)

Thanks and Regards,

Alain Gregoire
St-colomban, Qc
Canada

Thank you for all the kind words Alain.

About the Q's - yes they do. But in such an indirect way that I'd better say they don't, but then they would do "nothing".

Mind you (and everybody), this is the very first time that I openly admit that this is about jitter. I will still not say how it works, but I guess it has been long enough now to maintain it voodoo like, which it really isn't.
You can't say + and - because it's about the influence on the whole signature of jitter, which btw will work out differently per soundcard/DAC etc.
I really like to stick to this.

That I now at last "admit" this, is because of another anouncement I will be making soon. That one can't go without telling this anyway ... :teasing:

Best regards,
Peter

Hi Peter,

After a few weeks of listening to XXHE, I am stunned by the musicality and the "organic" feeling of it ! I have never heard anything that is near what I am hearing now !

I am not very good in adjustments and the few times I tried to play with the "Q"s did not bring much differences, but maybe this is because of the USB to SPDIF reclocker that I use (Audiophilleo 2) ? I admit I did not try for long though.

But the subtleties, the presence, the warmth... It really surprised me. I was a JRiver user before, I still use it sometimes when I intend to switch a lot between songs (I like the "fade out" "fade in" thing, because it is softer on the speakers). Sometimes also, for reasons that elude me, a song will not start (it has something to do with the headers of the song all in WAV, I should post the log ?). As for the interface, I don't use anything related to the cover art, so an important part of the screen is not used.

The most user-friendly screen (IMHO of course) would be foobar. With a "plain" screen I am happy because I don't need the "arts" other than the music. Like I said previously, I am a minimalist :-)

Still, the sound is awesome with XXHE and I will keep on following all you will put in it !

I am very surprised that there are not more publicity about your product... In my opinion, it is the best I have ever heard !

Thanks for the fantastic work Peter. It reminds me of a sentence I heard very often: "Everything comes from the source". Since it comes from a computer, its weaknesses have to be taken care of.

Regards,

Alain

Thank you very much Alain. I'm glad you like it so much. But the next version (0.9z-6) should really make a difference again.

When you have some spare time, maybe it's a good thing if you put your settings in your signature. Avoids unnecessary questions at problems, may they arise.

Regards,
Peter

Hi Peter,

Thanks for the suggestion about the signature. I added a "few" notes - maybe too much, but...

I am not sure about the differences in sound between Engine #3 and engine #4. The 2 of them are appealing, but the #4 seems to allow more fine tuning. I experienced some problems at first with it looks stable now with the indications in my signature. As for engine #3, all is simpler but I insist on engine #4 (I am curious). Engine #4 use is recent. I still have to check if I will get that error message that prevent some songs to play. I will keep you posted.

I am aware that my CPU is not powerful enough... It will eventually be a silent desktop or another more powerful laptop.

Question: When I use engine #4 I absolutely have to increase the Q1 at a certain level... I understand that it has to do with jitter management, but why do I have to increase it instead of leaving it at zero (as with engine #3) ? And what about the fact that I have a USB to SPDIF reclocker between the laptop and the DAC... Does it affect its work ?

And you are saying that the next update will be different ? I already feel impatient :-)

Thanks again !

Alain

Hey Alain,


In all situations there is no "zero base" or something like that. Thus, also not with Engine#3 (and look at the -4 which can be set).
It is the most easy for you if you look at it all as the setting of internal buffer sizes (which for Engine#4 really is so). Now, "0" won't say that a buffer is zero. It is just a setting and it implies a number of samples in that buffer. Btw, by itself this is not jitter related at all (better forget thinking in terms of jitter).
In any event, there is no base (or 0) setting in all situations. The only thing you could say (and see) is that for Kernel Streaming Adaptive Mode Q1 will be set to 1 by default, which means the buffer size is equal to the Device Buffer Size (see ToolTips on Device Buffer Size etc.).


Normally, yes. :)

Peter

Hi Peter,

Thanks for your response. It makes sense that the Q1 is buffer relevant. I realize that I let the numbers fool me (I knew that it could go under "0", but at this point I still am in learning mode).

I followed your recommendation as for the buffer (1024) with engine #4. Alas, I do not have any information as for the DAC buffer. I will try to get that info from the manufacturer...

As for the USB to SPDIF, when I asked if it was affected, I was in fact asking if it was antagonist...

By the way, sorry if sometimes my writing can be kind of cryptic... I do not always read myself before sending my message.

Regards,

Alain