wtfplay project - it's official

[Sligolad]

@sima66: One more thing came back to me when BIOS was mentioned. That was during very first tests with Jonathan (the first release I think). He had HPET disabled in BIOS because it was recommended for JPlay or some other player. Enabling HPET improved situation dramatically.

Actually now that you mention it Fryderyk that is the one item I have not disabled in some time, and you are right it was early days of JPlay and AO but later I did not find it made much difference for JPlay so I have always left it on.

[goon-heaven]

I have an RME HDSPe AES32 soundcard - is it possible to add suitable driver please?

Hi Steve,

have you seen my post before about Class Compliant Mode with FF UCX? If your RME card allows CC mode it will be recognized. you might check RME's manual.

Good luck
Thomas

Hi Thomas,

Yes, I checked for this CC Mode availability on the HDSPe - but found nothing in manual or gooogle. The card is internal PCIe - no switches to play with. You are fortunate!

[frd1996]

In the next build I will add the drivers for RME PCI devices.

[joelha]

In the next build I will add the drivers for RME PCI devices.

I'd like to put in a request for aif files.

Really enjoying the player.

Thanks a lot for that.

Joel

[goon-heaven]

In the next build I will add the drivers for RME PCI devices.

Excellent. Thank you!

[jrling]

might be a little more resolution in this maybe in kernal 11

Code: Select all

uwtfplay -p 60 -d hw:1 -f 256 -n 1 somefile.flac

Hi nige
Looking at your command line I saw -n 1 in it.

The lowest value for n is 2. So putting -n 1 you will get the default value in practice.

Here is Para 3.7 of the Manual -

The number of periods can be specified by adding -n M or --periods=M command line parameter, where the M is the number of periods. The minimum number of periods is two.

If you do not specify either of the two parameters above wtfplay/uwtfplay will use default values that are:

period size of 2048 frames (-f 2048)

period count of 3 (-n 3)

So you are actually running -n 3 (the default)!
Just thought you (and others) might like to know.

Jonathan

[nige2000]

i never RTFM :-()
Although im not sure 2 was always the min
Doesnt make huge difference anyhow
Initially i thought mqn had more detail air reverb etc and wtf had better balance fluidity and lower noise
Lowering the buffer to 256 512 or 1024 for me improved reverb resolution decay etc

[jrling]

i never RTFM :-()
Although im not sure 2 was always the min
Doesnt make huge difference anyhow
Initially i thought mqn had more detail air reverb etc and wtf had better balance fluidity and lower noise
Lowering the buffer to 256 512 or 1024 for me improved reverb resolution decay etc

Glad to be of service!

N 2 was always the min according to the FM! And, as I work in a software house and did help F with proof-reading the M, I always do.

Jonathan

[satshanti]

It's been quiet for some days, so I thought I'd do an intermediate report on my tweaking progress.

Because I didn't want to blindly follow the recommendation without trying it for myself, I tested the sound on default settings both with HPET enabled and disabled in BIOS. Until now for any music listening (through MQn) I've always had it disabled, and I think also Audiophile Optimizer recommends it. Anyway, I ran a series of test tracks and found that Fryd is right: it does sound better with HPET enabled, more control, tighter and more detail. What threw me off at first, was that the default settings make the sound in my system ever so slightly harsh or over-detailed and it sounded smoother and more pleasant with HPET disabled. In the end I realized that enabling it was the more accurate option.

After some tinkering with the -f -n and -p settings, I noticed that for me the -f setting has the most clear impact on the sound, so I started with that. I've not come to any definite conclusions yet, but every single setting makes a clear difference. In my system (AMD CPU with SA9023 USB receiver chip) the lower settings of 512, 256 and 128 were too messy and loose. The high setting of 8192 was too controlled and tight. Funny enough the default 2048 didn't sound as good as the ones directly above and below it. These are my two favourites: 4096 and 1024. The former is the most detailed, textured and accurate. The latter is really musical and has plenty of PRAT. I'll have to do some more testing with other tracks or the -p and -n settings to come to a definite conclusion. And who knows, maybe I'll find that the default 2084 setting is in fact the best compromise between detail and musicality. :-)

Anyway, I'm having a great time, and am really enjoying the excellent sound. Played some final songs with -f 4096 and -n 2 yesterday and was blown away, definitely the best sound ever on my system. To be continued...

[jrling]

It's been quiet for some days, so I thought I'd do an intermediate report on my tweaking progress.

Because I didn't want to blindly follow the recommendation without trying it for myself, I tested the sound on default settings both with HPET enabled and disabled in BIOS. Until now for any music listening (through MQn) I've always had it disabled, and I think also Audiophile Optimizer recommends it. Anyway, I ran a series of test tracks and found that Fryd is right: it does sound better with HPET enabled, more control, tighter and more detail. What threw me off at first, was that the default settings make the sound in my system ever so slightly harsh or over-detailed and it sounded smoother and more pleasant with HPET disabled. In the end I realized that enabling it was the more accurate option.

After some tinkering with the -f -n and -p settings, I noticed that for me the -f setting has the most clear impact on the sound, so I started with that. I've not come to any definite conclusions yet, but every single setting makes a clear difference. In my system (AMD CPU with SA9023 USB receiver chip) the lower settings of 512, 256 and 128 were too messy and loose. The high setting of 8192 was too controlled and tight. Funny enough the default 2048 didn't sound as good as the ones directly above and below it. These are my two favourites: 4096 and 1024. The former is the most detailed, textured and accurate. The latter is really musical and has plenty of PRAT. I'll have to do some more testing with other tracks or the -p and -n settings to come to a definite conclusion. And who knows, maybe I'll find that the default 2084 setting is in fact the best compromise between detail and musicality. :-)

Anyway, I'm having a great time, and am really enjoying the excellent sound. Played some final songs with -f 4096 and -n 2 yesterday and was blown away, definitely the best sound ever on my system. To be continued...

Glad to hear that your experience is exactly the same as mine and my 'sweet spot' ended up at -f4096 -n2 -p60 (uWTFPlay Kernel 12 which is clearly the best one).

I am amazed that several people say that they cannot hear any clear differences between different settings. On my modestly powered Intel J1900 board with Hiface EVO into dual 1704 NOS DAC, it did not take me more than a few seconds of a test track to decide if a setting was better or worse. Getting all three settings to come together as a 'Holy Trinity' took a little longer.

In fact, I still think that if they cannot hear differences, then perhaps they are getting the syntax 'wrong' and are actually listening to the default settings each time! Or their replay chain is not sufficiently resolving (unlikely on this forum) or a few too many glasses of red have been imbibed - or a combo of two from three!

We are fortunate that is clear.

Jonathan