Well ... where 0.9pd "has something", 0.9q has something wrong. It does not have it, it has a flair in the highs IMO. More digital again.


If you don't agree, just say it. 

Peter

Gerner, beautifully said. But allow me to address this to Gerard just the same. I mean, you both spent a hell of a time to find out what's going on, and although in both your cases this is not exacly clear, you are the winners of ... well, your systems.
Whether you may benefit from an improved version (the 0.9q I just put up) I don't know, but the fact is : those who persist will winn. I tried too (0.9q).

Either of your cases are IMHO rare, or better : as strange as could be;

Gerard, yes, you could say that you should not have so much "overhead" on your PC, but how that is related to one harddrive ... it beats me so far. I mean, XXHighEnd is not accessing that harddrive at all during playback (hence in the middle of a track) and still "you" (your PC) accessing that drive stresses XX. "Stresses" or de(s)organizes, in a mater I cannot comprehend.

Gerner, with you the same, but in an even more parculiar way;
You did not tell it out of politeness, but I will :

As soon as you start Foobar, NOT (and never) playing a single track at all, after a few minutes you have ... well, the same as Gerard; completely distorted sound which won't go away, until stop / restart playback.
Added to that : with *NO* Processor Core Appointment only. Well, if one thing beats me, it's this.
What even beats me more is that 0.9pd does not give you this distortion ... and ... you really received one program only (the 0.9pd) which is unrelated to the playback of sound ...

BUT :
As we know by now, 0.9pd not only solves these kind of problems (not for Gerard I think) but it even expresses a (far) better SQ. Those things must be related.


If not already ...
But you can bet I do not only listen to all of you, but I also will explicitly take this experience into account in order to somewhere, sometime determine what is really going on in playback where it is SQ related. Ok, "we" started off nicely, but although I'd say we're near the top already, I'm sure it still is just the beginning of it.

I know this took 10ths of hours of each of you.

Peter

(the Zip for download is completely at the bottom)


The following features have been added :

• Bigger Coverart for 640x480 resolution (see below 09q01.png).
  
  
• Coverart has been removed from the Playlist Area, hence the anomalies at showing not the full width of track data are eliminated.
  Coverart now shows in the left lower corner by standard.
  
  
• MP3 Icons (see below 09q01).
  
  
• Enhanced Gapless playback for MP3.
  As it turned out this could be better afterall.
  
  
• The text field at the bottom of the Library Area ("Galery Root") has become the main source for Search (see below 09q02);
  When nothing is entered there, the Music Root (path) from the Settings tab is taken, but in the end anything which is entered in that text field overrules.
  
  
• Checkboxes for selecting conveniently instead of normal selecting by means of clicking (with the risk of losing your selections at a wrong click).
  Not for deleting covertart at this moment !!
  See 09q02 and the "Chk" checkbox at the lower right corner; When that is ticked (as in the example) a checkbox appears next to each item in the Libraray Aera, and from that point on selection goes by means of ticking those checkboxes in there.
  
  
• Select All and UnSelect All for above.
  See XX09q03.
  Unlike Ctrl-Shift-Click etc., when all the checkboxes at once are to be ticked / unticked, use the respective menu options.
  
  
• Physical copy.
  Careful : thus far this has been applied for copying with Galery data as a source only !! (and applying it to normal data was not tested and may give anomalies);
  Once a Galery has been created (that consisting of meta data only) a "Physical Copy" of real music data can be made by means of that (.WAV only at this monent !!);
  The reasons for doing things like this can be various, and all is to your own imagination. One realistic example :
  Suppose you have 500 albums (CDs) physically present, and you have the idea (?!?) that those are a backup for your ripped data; suppose also that you have another 1000 CDs you lent and payed money for from a library or elsewhere, and you do not have a physical backup of those;
  Now you could put the lent CDs in a special library (the Metadata of them), and you could regularly backup only those ...
  Just enter the (before created) Galery path in the text field at the bottom of the Library Area (see 09q02 again), select all the visible items (e.g. per checkbox as shown before, or by means of clicking the first item and (Ctrl-)Shift-Click the last item, and press rightclick in the Library Area;
  Go as XX09q04 shows.
  Important : When a "Physical Copy" is made, the coverart is taken from the original folders, and *not* from the folder.jpg files in the Galery data. See more below as well.
  
  
• Now a message will follow after saving a to a Galery (or after a Physical Copy) that saving/copying has been finished. A progressbar shows the progress. When the mouse is put over the progressbar, the current number of selections saved / copied will be accurately shown until the process is finished. For the latter, see XX09q05 (never mind the d:\backup\ example from before in this screen copy, which was just virtual).
  Note : currently the process of physically copying (as well as saving to a Galery, but that's not important) can not be cancelled other than aborting XXHighend.exe by means of TaskManager. This by itself is not a problem, and just allowed (though not much official of course).
  
  
• Folder.jpg files in Galeries are allowed to be set explicitly.
  You may have noticed that when you create a Galery, the folder.jpg files are physically copied to the Galeries in the first place. This is to show coverart from your Galeries as well, which is not derived via the meta data from the original source. IOW, you can have different coverart (like a "Blues" picture) per Galery.
  Note that you have to arrange for this yourself (outside of XXHighEnd).
  At a Physical Copy this the original coverart is taken though, as per the meta data (also see more above).
  
  
• Rename of folders and files;
  Just like with Explorer, folders as well as files can be renamed from within the Playlist Area. Just click on the folder / file name once, and once again, and you are allowed to change the data.
  CAREFUL though : In here -and for your own convenience- more is allowed than you know from Explorer; you can change things anywhere in the path.
  Although everything is to your own imagination, see XX09q06a where a filename is selected to change (note the "Blues" in the LibraryRoot field at the bottom) and XX09q06b where the *path* (!!) has been changed to AAWav\ (also look at the top left of the XXq096b screen !) and thus factually the track is moved into the AAWav folder (never mind the meaning of that).
  This may look awkwardish, but when you get aquainted with select - copy the target path (e.g. "AAwav\") and select - replace that in the source (in the 09q06b upper right example), it can be a relatively fast way to change the structure of your (Galery or real (!!)) data.
  
  
• Support for up to 4 Processor Cores.
  Up to this (0.9q) version, Processor Core Appointment anticipated on 2 cores only, and the cores 3 and 4 (when available) were rendered useless;
  From off this version, cores 3 and 4 will be as useful as they are.
  The Audio Engine (see more below) will run in core number 2 always (unless only 1 core is available).
  
  
• Real time change of Processor Core Appointment.
  The checkbox accordingly in the Settings tab can be changed during playback, and it will take effect immediately.
  
  
• Improvement of Processor Core Appointment in general, or IOW : it was not as efficient as could be.
  
  
• Reset of Processor Core Appointment without reboot.
  Whereas before a reboot could reset the Processor Core Appointment, from now on any regular stop of playback resets all to "normal".
  Careful : "Normal" means evenly spread, and any before (playback commenced) appointed processes will not be explicitly appointed anymore.
  On the other hand, in a normal "home" situation explicit Core Appointment will not take place (uhhmm, but XXHighEnd does just hat).
  
  
• Processor Core Appointement for all Engines (also for XP).
  As said.
  The SQ effect of this has not been judged, but it can be assumed it matters one way or the other. Of course this is addressed to XP users.
  
  
• Better support for problematic Activation situations.
  Hard to explain, and maybe not even useful, but situations appear to exist that an Activation Code can not be transferred from the originally Activated PC to a LAN-connected other. For this "debug data" has been built in, which data users can report back in order to find culprits.
  Sidenote : By now it is near 100% clear that dual boot systems (Vista being involved) report wrong data for the sake of being LAN connected.
  All is not worked out as off now, and this debug data may contribute to just this.
  
  
• ToolTips were actualized (the latest were lacking since 0.9o).
  
  
• Selections (per clicking) in the Library Area remain visible when the Playlist Area does not have the focus.
  Just for your convenience.
  
  
• Offline possibility for getting an Activation Code.
  As it appeared, some people just do not have any means for a LAN and/or Internet connection;
  Per this version a provision has been taken to allow for obtaining an Activation Code without any Internet connection at all.
  Payment is then via normal (worldwide) bank transfer. The details will be given as per PM request.
  
  
• Backslashes (\) are auto added to entered path names whereever applicable.
  Before this was not the case always, while they are mandatory in any case.
  
  

Solved issues :

• The Xp/Vista text could stay visible in the Playlist Area after loading of tracks; An attempt was made to solve this.
  
  

General :

• SQ : (far) Better. Relative to the 0.9pd intermediate version no changes are to be expected, although few listening tests were applied compared to 0.9pd.  However ... if anything ... even more better.
  Edit : 0.9q ... no ..... Not better. In fact, worse.
  
  
• Appoint Processor Cores is available only for the non-demo (hence Activated) version.
  Let's say this is a for even better SQ after Activating, because it really does matter.
  
  

Additional issues :

• The Physical Copy can not be cancelled, unless per TaskManager (which can't harm, but check tha last folder copied).
  In a future version a facility for this will be added.
  
  

Which, as it appears now, is not so easy. I tried to solve it in the upcoming 0.9q version, but ...
In a later version I will address it, not now ...

Can you verify that in the "bad" situation(s), the crackles start right in the middle of a track (say 10 secs from start and 20 seconds before the end at most ?

Also (and sorry I don't know anymore if you told it already), how would you describe the "crackling" you have ? (heavy distortion, small clicks, etc. ?)

Could you retry with the XX folder on C: again (music stays at X:) ?

Are both C: and X: connected by the same type ? I mean, both SATA, or might C: be ATA/SATA and X: be USB ?
(etc.)

I don't think that is necessary Gerard. I just wanted to be sure that I understood well.
Where your problem resorts from I don't know. But IMHO not from any late version of XX.
Not that this helps you, but it helps me a bit.

Gerard, two questions please :

a. Do you say that you actually had the crackles, say, always ?

b. Do you perhaps say that you don't have them in 0.9pd ?



Additionally : Can you verify that you have the crackles with EAC ripped WAVs from your own origine ?
This is not about bad rips, but about the treatment for non-EAC (better : not official red book format) rips as per 0.9o. This treatment is special, and your memory listing from your earlier post may indicate that it is happening ...

Thanks.
Peter

Camera not necessary, I believe you.

Is this repeatable ? I mean, supposed you just rebooted and all, and it took, say, 20 minutes for it to occur; could you let it do exactly the same (same tracks !) again ?

If yes, please then try it WITHOUT reboot, but restart of XX only. Then exactly the same sequence again.

And ... you still think that this is since a certain XX version, do you ? maybe re-check that ?

Gerard,

Didn't you fall alseep by now ?
Must it be the conclusion, almost 24 hours later, you don't have problems when Processor Core Appointment is shut off ?

This is not an official version but it may sound better than the others ...

It contains the controlling section for Vista/Engine#3 only and was meant to be a kind of debug version for some people encountering problems.
Also, it is a version which is near to the coming up 0.9q, although 0.9q is two days of development ahead (hence is different).

The reason why this 0.9pd version is provided, is I myself had a lot of joy listening to it, or IOW, "it has something". And, since changes in the controlling section influencing sound are out of my control (so far), I thought to share it with you. . I mean, before the SQ is destroyed again.

Important :
The withgoing XXHighEnd.exe is meant to replace the XXHighEnd.exe from the 0.9p version. So, to be sure nothing else is destroyed, the below zip only contains that XXHighEnd.exe and nothing else. It would be the exact same situation which was incurred for at my place, where all was kept the same, except for this program.
Rename your current XXHighEnd.exe into something like XXHighEnd.09p before overwriting it with the contents of below zip.

I may see ehhh hear ghosts, but actually there was one other audiophool who pointed it out, while I myself unintentionally ran the 0.9pd version and was immediately happy (compared to 0.9m-1) before I ever found out my "mistake".

If I'm crazy, please say so and just throw away this 0.9pd XXHighEnd.exe.



PS: May it matter : Q1=14, Invert.

What is eating your cpu when you have the problems ...

I'm afraid though it's not so easy to find out; I expect it to be a one-time spike (like you showed before) but which possibly keeps on having the DAC de-synchronized until a stop-start is issued.

Let's not forget : no matter I don't have the problem myself, it should be the program that incurs for it now.

The best test might be : unchecking Processor Core Appointment, quit XX, reboot, and let it run so long that you're sure it won't happen anymore (or of course notice it still happens).

Be careful not to use other types of music data than before (WAV/FLAC/MP3).

Hi Gerard,

Indeed, before it showed 99 I guess ... ? (which it should !!).

A few remarks :

dwm.exe : your glass look in Vista. Better turn that off (into classic I think).

Explorer.exe : no idea why that is using cpu in your case.

General : 12 % useage is way too much (10 % follows from the list). If you sort the cpu column, at the top the cpu users will appear.

Just for background info :

For those who recognize that their soundcards can respond to a 48 sample buffer, do you realize that this means a synchronized and consistent (meaning : throughout hours) refill of that buffer once per 1 milli second  (1/1000 second) ?
And that this factually means a latency of 3-12 times better than the better digital workstations ?
I only want to say : don't mess with your systems, or increase the latency (if you can), or crackles (skipping of samples) will be the result.

XX is quite stressing your system already for response times. Not for cpu useage (look at my graph a few posts back), but if another process gets to the cpu, even the small number of cycles XX needs won't get there (in time !!).

Okay, you knew that ...