XXHighEnd - The Ultra HighEnd Audio Player
March 19, 2024, 05:56:40 am *
Welcome, Guest. Please login or register.
Did you miss your activation email?

Login with username, password and session length
News: August 6, 2017 : Phasure Webshop open ! Go to the Shop
Search current board structure only !!  
   Home   Help Search Login Register  
Pages: [1]
  Print  
Author Topic: z10 | Phase Alignment (advanced usage)  (Read 6826 times)
0 Members and 0 Guests are viewing this topic.
PeterSt
Administrator
High Grade Audiophile
*****
Offline Offline

Posts: 16827



View Profile Email
« on: June 25, 2012, 11:30:55 am »

USE AT YOUR OWN RISC 1

THE PHASE ALIGNMENT FEATURE IS ONLY ALLOWED WHEN YOUR AMPLIFIERS BLOCK DC.

Do you hear ? when this is not the case, you will fry your speakers. Never, NEVER forget this, and you might at switching to another amplifier. SO DON'T forget.

Do not THINK your amplifiers will block DC - KNOW it. Do as follows :

Connect the DAC to power amplifier(s) or to the pre-amplifier and that to the power amplifier(s).

Detach the speaker cables at the output of the amplifiers. Notice that what you are going to do has to be done for *all* of the outputs. So, left and right and if you use more (bi wireing, cross-overs etc.) do this for all :

Switch on everything.

Take a Voltmeter, set it to DC and set it to a range that it will see in the 200mV range. You can start higher (like 5V) but in the end you must check the mV range as well;

Connect the legs of the Voltmeter to the +/- of the output terminals. Polarity is not important. Now :


This button (Red is active) is your extra protection for when it is clear that you can not use Phase Alignment. This way you can not click this by accident :


This really activates it as long as the digital output volume is less than -9dBFS (so, towards the -144).

Set your volume in XXHighEnd to -30dBFS (see last picture). Better not activate Volume Normalization so the output will always be the same as set (at the left side).

Now play something and look at the Voltmeter. It is not allowed to show a significant DC Voltage. 100mV is already in the danger zone, knowing that the output at -9dBFS will show something around 800mV in that case. So what you actually should see is something like zero.

If you see something which indeed is in the tens of milli Volts, stop XXHighEnd playback and look again. Is the same figure still there ? then the "DC Offset" you see is inherent to your amplifier(s). It is not the best, but it is so and you were used to that. No big deal within itself. But don't let XXHighEnd add DC to this inherent Offset, because that is not allowed.
Notice : It can well be that an inherent Offset of, say, 80mV decreases when XXHighEnd plays. In that case this is because XXHighEnd implies DC of the other polarity. Watch possible polarity changes on the Voltmeter, because from +80mV to -20mV really is a 100mV change. Thus, any significant change is not good.



USE AT YOUR OWN RISC 2

This is what this little guide ends with as well, and it should be 100% clear to you that it is not guaranteed to work for your situation.

When you passed the above test, never forgetting to do it again with a new or other amplifier, it is not said at all that you will be in the clear; Despite you will be out of the danger zone, possibly plops are your part during playback or the end of it.

Watch this and bear in mind :
From the beta testing of 0.9z-7 it has come foreward that actually nobody is able to put up a consistent audio stream throughout. It sure can be done because we at Phasure can do it too, but let's say it takes too much knowledgde in general and especially from the XXHighEnd program to get your stuff right regarding this. What does this mean ?

In order to *prevent* the plops because of unexpected stops, the unexpected stops are anticipated upon and the program, say, gently ends playback. But, it *does* end playback. Generally you may notice this when going from one album to the other (both being in the playlist and subsequently played) but it can also happen in the middle of a track. So, even if one sample is missing, playback will stop. Yes, a good test within itself ...

The above will happen when the stream is interrupted and Phase Alignment is engaged. It will not happen without Phase Alignment being engaged. Then, like you were used to, the Audio Engine waits for what's to come, and one or two samples "stall" will not be noticed. However, when it needs longer like seconds, you will notice a gap, obviously. Maybe not in between tracks or albums because you think it belongs, and anyway, generally it is no big deal. With Phase Alignment engaged playback will just stop and you will blame XXHighEnd.
This is your good right, but in the mean time that plop was avoided which otherwise may happen each x seconds; remember, the stalling of the stream of 0.001ms is really enough for that.

So what to do when you notice that playback stops regularly : improve your system. It might be good for SQ as well. yes


When the above is arranged for adequately, you *still* can be in trouble. This time you implied buffer settings which can not deal with the proper treatment of Phase Alignment. No real rules for it exist, but let's summarize it (very) roughly -so you can understand- that if the gently dealing with it by the program needs 1 second, but buffers are 1ms of length, things can not be dealt with. There's no time given for it. Remember, this is the rough idea because it is far more complicated.


Before you think the text below is too long to read through, better do it after all. Why ? there are settings advised, and when these advised settings bring you bad sounds (like plops) something is not working for you. Various causes can be the matter, and as said, a not consistent audio stream is one of them (read about it below as well). So, while it would be completely inaudible to miss a sample here or there, now this will create less or more loud ticks and things may sound like getting damaged. So before you *are* going to read the text below, operate like this to your own benefit (and this is not a question but rather an order) :

Set your DIGITAL attenuation in XXHighEnd to 30dBFS or more (so, in between -30 and -144). Apply the settings you will see below (with emphasis on the two Addendums you will run into) and try some playback. ESPECIALLY when playback is normally going on, no ticks or anything should be audible. It it does, press Stop, sit back and think what may stall your stream. But also interpret whether it ever can be a stalling stream. So, if the ticks are at irregular intervals it is hard to imagine that it is an impaired stream because you would have noticed before (without Phase Alignment). In this case we may even think in the direction of electronics not coping. So here too, be careful.
But if the ticks are at regular intervals, especially at SFS boundaries, chances are 100% that something in your PC is not on par.

And on the other hand, when you hear ticks especially when pressing Stop, this will be about wrong settings for the feature. But, knowing we advised settings which should work (below) it is still dangerous.



This is for Engine#4 / Kernel Streaming only.

The advanced usage consists of it being very easy to imply anomalies during playback, which or are unavoidable, or come from other settings not being applied in consistency with the Phase Alignment feature. IOW, expect to encounter things like "what a sh*t player this is" when you are new to XXHighEnd and use this feature too soon. Otherwise ?

Otherwise the Phase Alignment can well be the best invention for Music Reproduction through loudspeakers of the last century together with the current one so far.

Its working will remain secret, but what can be told in general is that this comprises of a very mild form of digital signal processing (DSP), but completely lossless like the Digital Volume of XXHighEnd.

What happens (for sure without saying it all) is that a better alignment of the phase of individual frequencies take place. For the result you should be perceiving enormously more pure and undistorted sound.

Once it would be known what actually happens regarding the processing, you'd see that what happens is completely legal (up to : shouldn't it have been done right from the start like this ?). But, what it causes is completely "different" at the same time, and this works througout everything (in the chain). This is so different, that you might be scratching your head during the first days of listening to it. However, A/B comparing left aside, you really must go back to the normal setting (not activating Phase Alignment) after a couple of weeks uncousciously listening to it. You will be totally shocked how much distortion the normal playback means consists of. Really distorted sound, and you can only learn what that is after listening for a longer while to this super-pure sound.

A warning

It is expected that the effect of the Phase Alignment feature works out totally different for everybody. Why ? because it eliminates inherent distortion in each part of your chain, but, which may be out of distortion to begin with (and that to a more or lesser degree). But it also works the other way around : when parts of your chain operate distortion-free already to this regard, the feature may just imply extra distortion. Now watch out at judging this, because at first you (nobody) will know what this is about, because the kind of distortion will only get known to you once you heard it for a fairly long time without, and then go back (like it was layed out above). So, of course you will start to A/B right away, but this won't work. What merely will be the case is that the sound comes to you as "strange" (much depending on the type of music you play). This "strange" is never to be judged as wrong right away, unless it really is. Clear extra distortion is part of that, so if you perceive that right away, just don't use it anymore ! But otherwise ? try to let it go for several days, no matter how strange it comes across. In the mean time watch for this most pure sound; could be women voices to watch for.

One of the phenomena you will recognize is a perceivedbly less bass output. This for sure is to be ignored and instead try to get the real merits of it. So, maybe now you can suddenly recognize the bass instrument, instead of "bass(sy)" alone. Even for Phasure NOS1 users this may still be the case, them already being used to a similar thing, and this makes it "worse" (thus better) again.

It is especially to be noted that amplifiers may trip their shortcut circuitry. Not that fuses will blow, but protections in there may not dig what is happening here. This is only theory and it may not happen to anyone, but *if* it happens, switch on that circuitry and don't use Phase Alignment anymore.  Or remove the circuitry, because this really isn't about anything catching fire or shortcutting or anything of a bad kind. Phase Alignment just causes an abnormal state of (schoolbook) things, and electronic engineers may have created protections for this unexpected behaviour.

Activating it

Sadly, it is the most complicated part of XXHighEnd now, but for the best cause, so let's bite ourselves through some "constraints" we must live with.

First off, it won't work with (XXHighEnd's !) Digital Attenuation of less than 9.5dBFS. So, the Digital Volume of XXHighEnd *has* to be used. Otherwise it won't engage.
Note : When it is engaged, this can be seen on the Coverart Wallpaper by the denotation of ** around the Absolute Phase indicator. So, this will show *~* for Normal Phase and *I* for Inverted Absolute Phase. When _~_ or _I_ shows, it did *not* engage. And notice that the -9dBFS we talk about, incorporates the Volume Normalization. So, when you set the Digital Volume to -12dBFS but Volume Normalization made -6dBFS of that, Phase Alignment will not engage (without further notice, but watch the Coverart WallPaper).

Then there is the result of the Latency as the product of Device Buffer Size, Q1 value and Q1 Factor value;
(also see here for understanding this : Q1 Factor (huge latency setting - more advanced usage) )
It is to be assumed that at the combination Device Buffer Size = 4096, Q1 = 30 and Q1Factor = 1, it all works well. Watch out now, because when your sound device is incapable of this already very high buffer size (which is 4096 x 30 x 1 in this case), your chances get less in being able to use it.

Now want some complexities ?

All right. Even at lower buffer sizes it may still work well, but like described in the link just given, now even more factors play a role. But let's first describe better what "does not work" means in this respect :

When it "does not work", it only does not work as intended, and this means without any clicks or plops. They may be fairly loud, but not dangerous. They may happen when playback stops or when the Digital Volume is changed. Notice that small clicks may be audible at doing the latter anyway (this is avoidable but complex). So, concentrate on stopping Playback. If you hear no ticks/plops there, all is fine.

Athough it is just a description approaching reality, envision you are in a pressured room, in there audio playing. You will understand that a few things happened to you sound. Now, the "keeping under pressure" fails instantly when audio playback itself stops for whatever (unwanted) reason. What happens ? suddenly the air expands. Thing about an air yet breaking the sound barrier;
When music playback stopped for unexpected reason, it is fairly common that it also can start for unexpected reason. Now things go the other way around, and suddenly air retracts. Not sure what to compare it with, but that too makes sound.
Point is : when audio stops/starts outside the control of the program, there is no "time" to slowly unpressure (at stop) or pressure (at start). It goes with that little bang. Therefore it is mandatory that you can play without hickups of any sort.

Now, what influences this ?

- The shorter the buffer, the earlier it happens.
- The higher the volume (less attenuation) the earlier it happens.

... with the notice thar "the earlier" at the very same time means "the more loud".

Small summarization in between :
So, Phase Alignment itself will always work, but whether you can stand the ticks/plops when unavoidable for your situation is something else. Notice that a small tick will be no problem (you will get used to that), but that a loud plop may put your hairs upright because of the startle.
To keep in mind : when you run into the unavoidability of this, it will be foremost because your buffer size can't be long enough. You still have a chance by making it less severe, by means of fairly much digital attenuation you (can) apply. Think of -22.5dB being fairly much as such, giving you way better opportunities than e.g. -12dB.

The technical workout of the solution, meaning : once Phase Alignment worked by itself, implies that any "much too long" buffer size will imply repeats of the last buffer when stopping Playback. So, this now is about the other way around : we have a sufficiently long buffer in order to cover for the plops, but now it is so long that the mechamisms used when playback is stopped will cause at least one buffer to replay. And, at the highest "huge" settings, this is around 30 seconds (also see earlier given link again).
Thus mind you, for completely other reasons you may be working with "Huge Latency" (as how we call it), but now engaging Phase Alignment makes that Latency really occurring (otherwise you won't notice it at all).

Of course it backfires to you from that other end as well : when you are into Ultra Low Latency Special Mode Kernel Streaming, you just can't use Phase Alignment, no matter your sound device / driver is capable of the long buffers. But of course in that case you should try it, and of course you will. After that you will decide what is better net for your situation.

Addendum :
A month or so after this was written originally, it was found that a very low Split File Size (SFS) solves about all of the theoretical problems we can run into. So, an SFS of 2 allows for a Device Buffer Size of 4096, a Q1 of 14 and a Q1Factor of 1, which gives the net result of a fairly short buffer size while Phase Alignment works beautifully; in this setting the repeat of the last buffer (by sort of mechanical organization) is very short but also very long in frequence. For net result it will even be difficult to detect it (think like 50ms worth of audio repeating 10 times now).
This is important, because it will allow everybody to use it, while at first it was thought that only the "huge buffer sizes" would allow for it, that sure not being able to dial in for all DACs.

Activating it, again


See mouse pointer.
The IPhase button is accompanied by a small + and - (plus and minus) button. Clicking either of these activates Phase Alignment; Remember, whether it really engages depends on the Digital Volume (see above).
The ToolTips on these buttons will tell you when to use which as well as it will tell about all the combinations which theoretically exist. Quite many, and all will sound different. Which is right depends on your system, so take notice of that. But generally when you receive the best sound from Normal Phase, the minus button is yours. And the other way around, when your normal setting is Inverted Absolute Phase, the plus button is the one you'd want.

Not quite done

For highly experimental usage, the Settings contain a "Phase Strength" field. This is filled with 0 by default, and it should stay like that. However, it is there for a reason, and it can be set to 1, 2 or 3. It is not even advised to try it, but since the field is in there you will at some stage (but seriously, better not try it).
What will happen is that indeed the strength of this application is going to be higher with no single technical reason of how it can sound better. And really, it should be worse. However, this now acts as a smoothener for something we could call "over stressed systems"; Hard to explain, but look at it as amplifiers which are way out of control of the woofers, and this now will soften that. So, a really bassy system may get better of this after all. May, because it really will be a strange thing, and it virtually anticipates on things being fairly wrong in the first place, so the correction is (to be) inappropriate.
The down side of this ? you could guess it : all gets stronger. On that matter, here is that little list again with a third additional topic now :

- The shorter the buffer, the earlier it (the plop) happens.
- The higher the volume (less attenuation) the earlier it happens.
- The stronger, the earlier it happens.

So better avoid this Strength parameter.

Addendum :
Referring to beforementioned SFS setting of 2 which works "the best", it has been found also that applying a Strength of 1 sure can improve. Again it is difficult, because it first required the general nature of sound (or quality of it) "depicted" by the SFS=2 setting before it started to work out. So, despite the warnings above, an SFS of 2 with further settings as described under the earlier Addendum, seems quite safe in order to try a Strength of 1. Sadly, again later it was found that with this Strength of 1 and an SFS of 60, sound seems to get as analogue as can be, hence totally different and seemlingly way better again. Sadly, because with the further settings the same, you won't be free of small ticks at stopping or Volume Change etc. Still harmless, but not free of anomalies (which SFS = 2 seems to provide).


We at Phasure can not take responsibility of things or your ears getting damaged by the usage of it; even for advanced XXHighEnd users things can go wrong just by implying unthougtful settings. And otherwise whatever can happen is unforseen.
USE AT YOUR OWN RISC.
« Last Edit: September 10, 2012, 07:17:30 pm by PeterSt » Logged

For the Stealth III LPS PC :
W10-14393.0 - July 17, 2021 (2.11)
XXHighEnd Mach III Stealth LPS PC -> Xeon Scalable 14/28 core with Hyperthreading On (set to 14/28 cores in BIOS and set to 10/20 cores via Boot Menu) @~660MHz, 48GB, Windows 10 Pro 64 bit build 14393.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 14/-/0/0/*1*/ Q1Factor = *4* / Dev.Buffer = 4096 / ClockRes = *10ms* / Memory = Straight Contiguous / Include Garbage Collect / SFS = *10.13*  (max 10.13) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / Stop Desktop, Remaining, WASAPI and W10 services / Use Remote Desktop / Keep LAN - Not Persist / WallPaper On / OSD Off (!) / Running Time Off / Minimize OS / XTweaks : Balanced Load = *62* / Nervous Rate = *1* / Cool when Idle = n.a / Provide Stable Power = 1 / Utilize Cores always = 1 / Time Performance Index = Optimal / Time Stability = Stable / Custom Filtering *Low* (16x) / Always Clear Proxy before Playback = On -> USB3 from MoBo -> Lush^3
A: W-Y-R-G, B: *W-G* USB 1m00 -> Phisolator 24/768 Phasure NOS1a/G3 75B (BNC Out) async USB DAC, Driver v1.0.4b (16ms) -> B'ASS Current Amplifier -> Blaxius*^2.5* A:B-G, B:B-G Interlink -> Orelo MKII Active Open Baffle Horn Speakers. ET^2 Ethernet from Mach III to Music Server PC (RDC Control).
Removed Switching Supplies from everywhere (also from the PC).

For a general PC :
W10-10586.0 - May 2016 (2.05+)
*XXHighEnd PC -> I7 3930k with Hyperthreading On (12 cores)* @~500MHz, 16GB, Windows 10 Pro 64 bit build 10586.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 14/-/1/1/1 / Q1Factor = 1 / Dev.Buffer = 4096 / ClockRes = 1ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = 0.10  (max 60) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / All Services Off / Keep LAN - Not Persist / WallPaper On / OSD On / Running Time Off / Minimize OS / XTweaks : Balanced Load = *43* / Nervous Rate = 1 / Cool when Idle = 1 / Provide Stable Power = 1 / Utilize Cores always = 1 / Time Performance Index = *Optimal* / Time Stability = *Stable* / Custom Filter *Low* 705600 / -> USB3 *from MoBo* -> Clairixa USB 15cm -> Intona Isolator -> Clairixa USB 1m80 -> 24/768 Phasure NOS1a 75B (BNC Out) async USB DAC, Driver v1.0.4b (4ms) -> Blaxius BNC interlink *-> B'ASS Current Amplifier /w Level4 -> Blaxius Interlink* -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere.

Global Moderator
Pages: [1]
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1 RC2 | SMF © 2001-2005, Lewis Media Valid XHTML 1.0! Valid CSS!
Page created in 0.154 seconds with 20 queries.