XXHighEnd

Hi all,

It seems useful to share an experience, already because it may sound intrigueing. It is about something we all actually knew, but aht I now could emphasize.

At working on one of the Firewire interfaces for the NOS Phasure 1 DAC, I ran into a, say, very fragile one; It allows "ticks" very easily, and it is also "capable" of keeping on playing while it actually should not, or with hiccups otherwise. In this case this merely looks like a nice and neat "even" skipping of samples, like 1 out of 10, 1 out of 9, 1 out of 3, 1 out of 2 etc. This by itself can be incurred for by buffer settings, and a larger buffer creates this "effect" (the buffer will be overflown). From another angle this could be called "fragile" also : when 1 sample out of 20 is skipped, this may go unnoticed, but will degrade sound for sure. So, all very tricky.

Back to the "ticks" I mentioned, many of you may recognize that making the PC more busy can incur for fragile ticking, that ticking very much sounding like vinyl ticks. Not everybody may notice this, and when the DAC is connected via e.g. my Fireface, this never happens. Via a Juli@ (PCI) though, it does, and to me it is unclear what actually causes it, and so far I blamed in on "PCI" or "in the PC".
This particular Firewire interface - not being much different than a Fireface, and (also) SPDIF connected - does the same.

Does the same ? not completely;

If I move the mouse, some tick-tick-tick is audible.
If I hover the mouse over a button, trrrrrrrr is audible;
If I let a Tooltip popup, TRRRRRRRRRRRR is audible.

Other things don't seem to incur for this, nor does disk I/O. It is just the mouse interactibg with the objects on the screen.
Note that the TRRRRRRRRRRR from the Tooltip already makes the music unrecognizeable.

Now, what should worry you is that before the tick-tick-tick (very tiny ticks) are audible, we might be long in the stage of samples being skipped "1 out of 20 or 40 or 50" etc.; This will be inaudible by itself, but how does it sound ?

I didn't take the time on finding out what exactly was skipping (if at all) or actually happening, but figure a thing like this may be happening to us all the time, that making just the difference between "the Services" being on vs. them being off. Note (and I sure checked that) that when "the Services" are off, not any mouse movement could incur for any ticking.
But as I told earlier, at leaving the Services On, any changing pixel on the screen causes the whole screen to be redrawn, and I calculated that any random screen I/O was about 10000 times more slow with the Services On. So, the least this does is using heavy cpu cycles, and might you ever have seen a "glowing" button (its border colours slowly getting more and less) which is for showing it as the default button (I think) - at each gradation of the color, the whole screen is repainted. This would be a typicle situation my interface would keep on ticking ever and ever, until that button stops glowing.

If I change the buffer size, all ticking might go away (audibly !!) and the popup of a tooltip may cause one or two ticks only. Or none, if the buffer size is changed further. Audibly ... but what will really be happening ? 1 out of 30 ? 40 ? 50 ? 10 perhaps ?

So, besides any GUI present influencing sound in general, I now ran into this phenomenon which I didn't know before. It will be not much different (wrong, it will be much worse) from the screen copy below, that outburst showing as the difference between Attended and Unattended playback, the Attended creating such an outburst at the interval the time cursor moves. I never heard that cursor moving, but you can bet that what you see below must - or should be audible. It will be in the middle of many other anomalies incurred for by actually any program running.

Keep this in mind when you think "the h*ll with Unattended", and instead think Unattended will avoid these influences to a large extend. It really does, and that is what Unattended is for.

Peter

If I move the mouse, nothing happens, even if it brings on front the taskbar.
If I make active xxhe window or Console window, nothing changes.
If I browse through the gallery, there can be an occasional tick trr or something like that.
If I transfer files through usb or the lan, the trrr and general dirtyness are almost sure.

I have large buffers (2s) but they are not enough. Most is culprit of the poor asio drivers of the Quartet.
Waiting for new ones or your DAC, whichever comes first :)


I'll give unattended another chance, but i hate to not be in control of everything my pc is doing.

Telstar,

Try Unattended, YOU ARE IN CONTROL, REALLY!

I like it now, and before I didn't.

ALT-X brings xx back in a flash, press stop and xx brings back wallpaper, press "I" button and brings back desktop icons.

And when loading files to C: you actually see xx do that. (instead of leaving us in the dark)

I think its like this: Unattended has become better, and Attended has become more sensitive to cr*p going on in your PC. (and thats audible)
 :)

So The difference has become bigger

OOOOK, looks like i have to :)

BTW, i'll be sending you an email later about an interesting speaker that i have found (before the vacations).