wtfplay project - it's official

[jrling]

Satshanti

Thanks for this in-depth review, which is very insightful.

I continue to believe though that YMMV from others. Certainly it does to me, as any setting above p 60 is def worse.
BUT I think there are two possible reasons for that:

1. With Frd's help I have now found out that HPET on my Bay Trail J1900 Mitac PD10BI board is not being enabled by WTF. This is because Frd's tests at boot time assess HPET timing accuracy and decide if they are good enough to turn it on. In my case, it is not and the BIOS is awful and has no HPET switch or any other performance tuning settings available to the user

2. The J1900 is a low-powered board. CPU hard soldered. I think therefore that as the Priority setting is increased, it starts affecting other Linux kernel priorities as the CPU is not powerful enough to adjust to the higher demands and then it all goes pear-shaped.

Well that's my theory.

Jonathan

PS In UK I think your expression is "Neither Fish nor Fowl"!

[frd1996]

1. With Frd's help I have now found out that HPET on my Bay Trail J1900 Mitac PD10BI board is not being enabled by WTF. This is because Frd's tests at boot time assess HPET timing accuracy and decide if they are good enough to turn it on. In my case, it is not and the BIOS is awful and has no HPET switch or any other performance tuning settings available to the user

Before we jump into big conclusions on this point, here is what I found out.

Linux kernel tests HPET accuracy at boot. Initially I thought that this is the problem, but this is even simpler. In the case of Bay Trail platform HPET is not even tested. There is a specific workaround in Linux kernel to disable it and there is a reference to the Intel documentation. I took a quick look at it and it appears that in certain power management states of Bay Trail, HPET will be disabled, while CPU will be still running. The allowance of such situation ruins completely HPET usefulness. That is why Linux kernel developers decided to disable it. Hopefully I understood it correctly and this is not too technical.

I think the primary target for Intel Bay Trail are the Windows based tablet devices.

F

[cvrle59]

I am bad, I have to admit it, I live on someone's else hard work. I'm listening music right now with -p80, -f4096, -n2, so my system sounds the best, ever.

[satshanti]

I am bad, I have to admit it, I live on someone's else hard work. I'm listening music right now with -p80, -f4096, -n2, so my system sounds the best, ever.

:-) I'm glad! Don't feel guilty, just enjoy it! :-)

[satshanti]

Well, the verdict is in now, it seems. I chased down the cause for the slight congestion and sharpness in my system. It was the stereo to binaural conversion process I use to prepare the files in my music library for listening on my headphones. I'll update my original thread about that in one of the coming days. It's definitely worth revisiting that.

Anyway, the sound opened up and the hint of harshness disappeared, which meant that it's now clear that an f-setting of 4096 is in fact the optimal one.

So a certain consensus seems to be emerging, that the sweet spot lies at:

-p 80 -n 2 -f 4096

If some more of us have the same finding, Fryd, it might make sense to make that the default setting in any new versions. Just a thought.

Now, having found and solved the conversion problem, the sound quality is nothing short of absolutely amazing. No matter what I throw at it (well, within limits :-)) it's such a joy to be so fully immersed into the music, sitting there in front of the musicians, be it a grand piano, a string quartet, an orchestra or a bunch of folks in a jazz club.

Once more my sincere thanks for such a great music player.

[Claus]

Just got this little gem up and running. I am hearing a 3d sound stage that I have never heard before. Not practical for me as I have the TV connected and it won't show the resolution, so I need to have separate screen set up... but it might be worth it... congrats! Great job on this player!

[frd1996]

Here is a BIOS tip that I received from few users: disabling Intel Hyperthreading (HT) helps. Please try it. I am wondering if this is a generic case or it is limited to specific hardware.

[rickmcinnis]

I remember being told to reactivate it.

But I didn't.

[Sligolad]

I went and bought a Programmable Keypad recently to help me with Beta testing new versions of Wtfplay and it certainly makes life so much easier for testing.
Bought the Accuratus 24 keypad from here but on further research the X-Keys pad may have been a better option as it seems you can copy text between keys which makes life easier rather than typing all the text for each key when a lot of the text would be similar. http://www.keyboardspecialists.co.uk/sh ... keyboards/

With the keypad I could easily have a track or album already programmed to play on a single key with all personal preference priority, buffer, etc. modifiers available with one keystroke once wtfplay has booted when changing between different kernels.
Even having the bulk of the commands ready on keys and pointing at you music folders is an easy setup so all you gotta do is type in your music destination.
For the CUI version this programmable keyboard could do it all.

The good news is for me recent beta testing has shown an improvement in new wtfplay kernel tweaks which i heard before i got the keypad so looks like some really nice audio improvements coming the way of those who have discovered this great player.

Hopefully there will be some good news soon!
Cheers, Pearse.

[joelha]

Pearce,

Thanks for helping to test this new version of wtfplay.

I absolutely can't wait for this update.

Joel