XXHighEnd

Hi,
I finally bit the bullet and purchased XX.  :sign0144qp7: With some advise from Peter and using the explanations of the tooltips, I now got XX working "errors/distortion/pop free" with following settings:

- KS Adaptive (Buffer 512, found using method described in tooltip)
- Scheme-3
- PlayerPrio / ThreadPrio = Nothing / RealTime
- DAC settings: 24bit / 192kHz, 32 bits, 96 and 192 only (since soundcard driver doesn't support other)
- Split file size 60MB

I tried Special KS Mode but fund that I had to set the Q1 to 270 or higher (buffer at 1024) and even then after a while the sound would start to distort. So, I went back to Adaptive. Anyone know, if it is still worth trying even higher Q1 settings?

However, I wanted to try out Anti Imaging with Double or Quad. This works but only for a few minutes(between 5 and 15), after which I will get a "Too many buffer errors" message and the sound stops.
Does anyone know how to get around that?


Any help is appreciated!

PietPara

PietPara,

You are going to have to increase your buffer size for Double and Quad. Repeat the method described in the tooltip for each of these separately.

As an example, with my RME, with NO Double or Quad, I can set the buffer to 32 in Adaptive mode and I get no buffer errors or pops 'n' clicks. Switching to Quad (with Arc Prediction), I have to set the buffer to 512 to get the same result.

Special mode with Quad Arc Prediction remains difficult for my setup - I have to set the buffer to ~200 to get a clean signal. This is substantially higher than Peter's 22 samples... or Russ's 1 sample! (or is it 2 samples nowadays?).

Mani.

-16
Set Device Buffer Size : 48  Samples of latency (Q1) : 1  (see more below for the applied KS Mode)
-17
-18
-19
-28
-29 (Start Playback)
12:42:53.9668654 First ChangeWP call
Special Mode !

I've been flipping between 2 and 1, but most of the time it is 1. By processing all the tracks before play starts , and disabling the OSD, I found the occurrence of the occassional click/pop is only every 5 to 10 minutes or so. It depends on the album as to whether I actually notice these or not. Fair to say that quiet passages are more susceptible to hearing a click.

Russ ( with a saturated nucleus accumbens   :grin: )

P.S. welcome PietPara

Thanks guys, I'll try that. I could have know but since the tooltoip says that for Adaptive it is important to set the buffer to the actual buffer size of the device, otherwise adaptive is useless, I thought I shouldn't change that buffer size anymore.

So I interpreted the buffer size incorrectly(?).

PietPara

Hey there PP - Allow me ...


It depends what you did and how you interpreted things in the first place;


So, if you found that, it just should be allright. But remember, you can only have found that when no (or not too many) Buffer Errors are reported !


This is where you might have gone wrong ...
So, if you want to play with Double or Quad, you should have tried to find the buffer size with that (Double or Quad) active !
But ... I don't think this was ever made clear by me anywhere. It is true though.


Keep in mind : For Special Mode the Device Buffer Size does nothing. It only presents you Q1 ranges which are different per setting.
Also to keep in mind : Special Mode always has a lower latency than Adaptive Mode, so if you can perceive that lower latency is better sound, it still is worth it. But merely : it is just another kind of engine, and it will sound differently because of that.
Lastly, the setting found by Adaptive Mode is actually unrelated to what you can do with Special Mode. So :
1. Adaptive Mode will let you find a low latency setting easily;
2. Special Mode means a lot of hassle at trying, but when you found it at last it will be lower than Adaptive, and thus "better". But also different.

Peter


PS: After using Special Mode for two weeks or so, yesterday I went back to Adaptive Mode, and as happened to me earlier I am under the impression that this has a more analogue sound. Less stressed too maybe. Just my 2c.

Thanks Peter,
When running in Adaptive without Double or Quad with a buffer at 512 I get no Buffer errors at all and no opos either. At buffer of 384 I get buffer errors, so according to tooltip explanation, I should use 512. And tooltip explains that this is the actual devices buffer size.
Therefore, I do not understand why I would still need to change the buffer again when trying Double or Quad. Doesn't that contradict the tooltip explanation of the buffer having to be set to the actual device buffer size?

Actually, I must correct myself after more experimenting, the buffer errors occur only under Quad, not with Double.


And regarding Special Mode, how is it explained that the distortion only starts after a few minutes of playing, before that being fine? Should I still increase the Q1 or does this point to a another issue with e.g. the soundcard?

PietPara

The general answer to your main question is : you are not setting a buffer size or anything, but a latency. And the possible latency is subjective to the number of samples per time unit.


Possible reason :

Because after those few minutes something happens in your PC which requires more resources than the low latency stuff can bear (could be the loading of a track (part).


Although that could help, it is better to take away the resource eating thing. But, you must know what it is. For example, if it is about track loading, you could decrease the Split File size parameter.

But it is rather complicated, and I'm afraid there is no single answer to it. :no:
Peter

Hi Peter,

Work with two PCI-cards on W7 64bits.

1:Asus Xonar Essence st
2:Audiotrak HD2

Both cards work fine via analog out and Arc Prediction double and quad in apdative or any other mode.
When feeding an outboard dac using dig. out (SPDIF) of both cards, just double works, NOT SO QUAD! No sound with quad, but pointer moves.
On the Asus card I do have incidental clicks in the playback (at digital througput SPDIF). Maybe this because of not beeing able to do 88.2 and 176.4 and samples up in a wrong way? Can choose 96/192 and anti image, but like to stay in arc pred. though.
Do you maybe know a reason or have a solution?
Next to these remarks am very happy with the sound, as ever.
 :) Ed

Difficult Ed ... I don't think the ticks come from what you think (but I don't know where otherwise from either -> should be a resource problem). If you can proove that this is with all modes (especially KS Adaptive compared to Engine#3) I will start to think of other things, but I don't think you will find this is the same all over.

You merely bring me to think about something else, which already slipped my mind with PietParra's problem;
He will be using the Xonar analogue out, and I (now) actually wonder what that does to the "bit perfect" thing. I mean, chances are fair (as how I think now) that this is subjective to the OS resampling anyway. It may even depend on how the card is organized, but notice that the drivers for SPDIF pass through and analogue out will be different for sure. So :
For the AudioTrak (I know) you can proove this by looking at the control panel. Does it output the rate (at analogue out !) you depicted ? or does it output the rate as set in the OS at the driver properties - Advanced (for shared mode) ?
If the latter, forget about analogue out from such a card. If the former I currently wouldn't know what the difference can be, BUT there's always additional processing in the card then (we call that D/A conversion, haha).

Maybe with this story you can come a bit further with it ?
Peter