New Filter request(s)

[PeterSt]

I wish I knew myself what is going on really. This is a text from someone who doesn't frequent the forum :


  I've managed to play without hearing any more ticks. I've got the SFS down to 1 now, each time hitting the Blue LED as I adjust the SFS lower.

  Now the sound quality is fantastic. More pin point focus, recording ambience is clearly captured, and the highs are simply gorgeous.  It is really hard to imagine what will it be when this new filter is combined with NOS1a.

Forget about this. This was a mistake (I was told).
Sorry.

[christoffe]

Joachim,

If you prefer 8x in general you can also select the 176400 filter (and use 8x).

Regards,
Peter

Hi Peter,

got it. This setting is more pleasing for the speakers and has an impact of the location of the instruments in the recording studio too.

With 16x the piano was in the back of the studio, and moved now (8x) forward, which is more realistic. (Stanley Clarke Trio)

Joachim

[charliemb]

One other problem with AI:  hitting alt-N during unattended playback causes "DAC can't play... check settings.." (I think it's that one).  Hitting alt-P does not resolve the problem.  I need to hit alt-X and hit play button for it to play again.

The shortest path to getting to the next track is to hit, while a track is playing in unattended mode, hit alt-X, alt-S, select next track and click Play.  Well, at least that works without ever getting error.

So hitting alt-X appears to be the key.

More data on this problem:

The same error happens when I do:
- play unattended using AI set to 4x
- hit alt-E to pause.
- hit alt-P to play   (error happens here).

When I come back to alt-x, engine 3 is already unloaded.

Curiously,  the following doesn't result in the same error:
- play unattended using AI set to 4x
- hit alt-X and wait for xx to come up,
- click on the pause button
- click on the play button.

Plays just fine and xxHE goes into unattended.

[charliemb]

One problem I must report about AI:  AI is not multitasking as I expected from reading your previous posts.  I only see one yellow CPU-core light and albums load slowly as before.  Example,  one Spiro Gyra album has 8 tracks with all tracks nearly 5 minutes long, and the longest track is 5:49.  Well AI prodded through these 8 tracks one by one and I timed it with a stop watch.  This album took 2:09  to start playing. 

You can see my PC specs below.  This is a fast xxHEPC and I'm running it at 2.5GHz; I don't use playback drive. 

New Data Point on this 2:09 situation with AI.

I have "found the problem."  Recall that I set my upsample rate to 4x for my 192 DAC...

-AI multitasks every time and just fine when the source is 88/24 or 96/24 HD material.  Conversely, 
-AI fails to multitask when the source material is 44.1/16.

I'm still not sure if the problem ties to 88 vs. 44 --OR-- 24 vs. 16, could be either. 

That might be enough info? 

BTW, I see/hear no problems with AI and multitasking with 88 or 96.  So probably no reason to disable it.

[Edit: I found one album 48/24.  It multitasked.]

[PeterSt]

Thank you Charlie.
Peter

[christoffe]

Hi,

After some tests with the different filters (rev. “e” & “g” ) the revision “e” (sent with the NOS1a) is with 8xAP the preferred one on my system.
Rev. “e” with 8x AP brings a seamless acoustic pattern, nothing protrudes. The soundstage is in a perfect balance.

Best SQ I ever had.

Joachim

[charliemb]

New Data Point on this 2:09 situation with AI.

I'm still not sure if the problem ties to 88 vs. 44 --OR-- 24 vs. 16, could be either. 
...
[Edit: I found one album 48/24.  It multitasked.]

I found some 44/24.  It also multitasks.

This confirms that the lack of multitasking on AI is a 16 bit vs. 24 bit issue.

[PeterSt]

Can't this be the difference between WAV (not) and FLAC (does) ?

[PeterSt]

I found the ticking issue (and solved it). Now for the DAC Needs = 24 noise ...

A new 1.186-h will emerge hopefully soon (1.186-g download has been removed).

Peter

[PeterSt]

New Data Point on this 2:09 situation with AI.

I'm still not sure if the problem ties to 88 vs. 44 --OR-- 24 vs. 16, could be either. 
...
[Edit: I found one album 48/24.  It multitasked.]

I found some 44/24.  It also multitasks.

This confirms that the lack of multitasking on AI is a 16 bit vs. 24 bit issue.

But not here. BUT :
While writing this, I suddenly recall that it works at the album level. So, load a Playlist with tracks from several albums and then it won't work (and never will - just saying).
So that will be it.

Peter

[charliemb]

Can't this be the difference between WAV (not) and FLAC (does) ?

Uh oh...  I think it could be.  I will confirm tonight. 

But I hope not because most of my CD collection is ripped to AIF using dbPowerAmp, AIF being like WAV.

[PeterSt]

No no, WAV and FLAC both work. AIF(F) will too (not that I tested that).
On another note I used that AI version myself for a couple of weeks and if it would have not used multithreading I would have known. This with the notice that I used it with upsampling ti 705600 and a 430MHz set processor ...

[PeterSt]

See New Filter request(s) - XXHighEnd 1.186-g for the original text coming along with the Filter Beta.

Unzip the 4 files in the download (at the bottom) over your current 1.186 or higher folder after removing XXEngine3, XXHighEnd and AutoHotkey from memory.

This solves the ticks.
With thanks to those who noticed it was in the right channel only (now it took me minutes only to find the culprit).

24 bit "DAC Needs" noise takes more time to solve (so later) but it won't play now with the Custom Filter engaged.
Once more apologies that this could happen (but I guess that's beta-beta).

Peter

Edit : Download once again removed.

[charliemb]

New Data Point on this 2:09 situation with AI.

I'm still not sure if the problem ties to 88 vs. 44 --OR-- 24 vs. 16, could be either. 
...
[Edit: I found one album 48/24.  It multitasked.]

I found some 44/24.  It also multitasks.

This confirms that the lack of multitasking on AI is a 16 bit vs. 24 bit issue.

But not here. BUT :
While writing this, I suddenly recall that it works at the album level. So, load a Playlist with tracks from several albums and then it won't work (and never will - just saying).
So that will be it.

Peter

Actually, here is the data.  See a pattern?  This took much of the night to find these.

Album	Rate/Bits	Format	Multitasks?
==========	========	=======	=======
Hothouse Flowers	44/16	Flac	Yes
Mumford and Sons	44/24	Wav	No
Other Aiff (many)	44/16	Aiff	No
Stravinski (DG label)	96/24	Aiff	No
Tedeschi Trucks Live	48/24	Flac	Yes
Many other HD	96/24 or 88/24	Flac	Yes

Certainly not 96 vs. 44.  Certainly not 16 vs 24.

[PeterSt]

Re-chance (see my previous post in this topic);

What I wanted to say but completely forgot :
The 1.186-h version (but the 1.186-i still) contains "restored" code from what has been tested to work a few weeks ago but which code was eliminated because whatever it does, it does in the NOS1 driver now. However, nobody can receive this driver yet. But because for me "it" is already arranged for in the driver I use, it can't go wrong in the restored code so I can not test it. But for you it can go wrong, and so it did for 1.186-h which has been available for a short period of time.

Thanks to one user who reported this to me, and tested the next iteration of the code, not all that many suffered.

Still it can be so that some "GPIO" error message occurs. I think this can happen :

- Right after the NOS1 had been switched off and on again (killing the sound engine should solve this (blue led) and retry playback);
- When the NOS1 Control Panel has been removed from memory (just cross it away when you had it up and do not chose "File - Remove from memory").

In any event meant error messages can only occur for NOS1(a) users.

Remember, paste the files from the zip over your current XX 1.186 (or higher) folder; nothing else. First kill XXEngine3.exe, AutoHotkey.exe and quit XXHighEnd.

Peter