XXHighEnd

Ultimate Audio Playback => Your questions about the PC -> DAC route => Topic started by: pedal on May 27, 2010, 04:26:25 pm



Title: Problems between XX and RME DigiCheck?
Post by: pedal on May 27, 2010, 04:26:25 pm
Hi Peter,

I use the RME FF400 as soundcard/spdif out to my DAC.
RME has a freeware called DigiCheck, where I can analyze the signal (music) stream. Also, it has a recording function, recording the stream to harddisk and saving it as a .wav file.

It seems that XX and DigiCheck doesn't comunicate perfectly, because I get audible noise when recording. (I record, save to harddisk and can replay the file in XX or in digiCheck itself).

Also, when I use the audiospectrum analyzer I get (too much) high frequency above 20kHz. It's like all my 96kHz albums has a lot of supersonic content. It is hard to belive. Maybe it's the distortion showing?

Is this a "generic" problem because of XX (I know som dCS DAC's doesnt work with XX). Or should I increase to latency or something?

All the best!


Title: Re: Problems between XX and RME DigiCheck?
Post by: PeterSt on May 27, 2010, 04:35:31 pm
First a question : which XX version are you talking about ?


Title: Re: Problems between XX and RME DigiCheck?
Post by: pedal on May 27, 2010, 05:19:05 pm
First a question : which XX version are you talking about ?

09-y8c.

(Sorry for troubling you. I could of course try to install Foobar and check if it works better. But since you and others here at the Forum also use RME hardware/software, maybe you had some experience in this matter).


Title: Re: Problems between XX and RME DigiCheck?
Post by: PeterSt on May 27, 2010, 05:24:22 pm
It's ok Pedal, only I don't like to connect my RME stuff at this moment.

Allright. What Engine do you use, and if Engine#4 what Mode and Q1 setting ? If Special Mode, what is the latency in samples ?

Btw, I don't recognize your problem, but it is not difficult to create distortion by means of missing samples. That's why the questions.


Title: Re: Problems between XX and RME DigiCheck?
Post by: pedal on May 27, 2010, 06:30:38 pm
It's adaptive mode, 1024 Latency, Engine 4, Q1=1.
(Didn't have time yet to try yout latest version).


Title: Re: Problems between XX and RME DigiCheck?
Post by: PeterSt on May 27, 2010, 06:51:19 pm
Ok, what I can imagine - a long shot though - is that the latency you use (which will be net 1024 samples in your case)

a. is too low to allow for headroom to capture at the same time;
b. implies too short buffers to allow for digital capturing.

Ad b.
I don't know how the (buffering for the) capturing works, but if you have one buffer of 1024 samples in the driver, and one buffer of 1024 samples in the buffer at the program side (hard to explain), then either or both may already have been overwridden before the capturing is done (or even started) with it.

You may try whether Engine#3 gives the same poor result; From my own experience that should work normally (I don't think I ever tried capturing with Kernel Streaming).


But is this about your loopback (and room/speaker correction stuff) ? and if so, how come you notice this only now ?


Title: Re: Problems between XX and RME DigiCheck?
Post by: pedal on May 27, 2010, 07:21:16 pm
Switched to Engine 3, but same problem. I can hear the music but highly distorted.
No, no loop or DSP at the moment.

I will try with other player and revert.


EDIT: I needed a fairly high (1024) latency in the past because I was using the internal DAC in the FF400. Now I have got my new Buffalo II DAC, and I use the FF400 only to provide the spdif out. I had forgotten about this point, but right now I managed to reduce latency to 512 with glitzfree play of 24/192 music files. Isn't lower latency supposed to sound better? :-)


Title: Re: Problems between XX and RME DigiCheck?
Post by: manisandher on May 27, 2010, 07:22:54 pm
I'm not sure if this is relevant, but I can get clicks and pops when I have DigiCheck on, i.e. not when recording, just when it's on and XX is playing.

However, this has everything to do with the buffer size. The following buffer sizes give the following results:
32 - Digicheck crashes
64 - clicks and pops on playing through XX
256 - works fine

As I mentioned, this is NOT when I'm recording, but just using something like the Totalyser.

DigiCheck's a really neat piece of software though...

Mani.


Title: Re: Problems between XX and RME DigiCheck?
Post by: manisandher on May 27, 2010, 07:37:39 pm
Also, when I use the audiospectrum analyzer I get (too much) high frequency above 20kHz. It's like all my 96kHz albums has a lot of supersonic content. It is hard to belive. Maybe it's the distortion showing?

I think it's the distortion.

I've just tried a RR 24/176.4 and there is a -105dB peak in noise at 25KHz, which gradually dies down. The highest peak I'm measuring at 81KHz during playback is around -90dB.

If I reduce the buffer and introduce noise, then indeed, there is a lot of HF energy going all the way up to 81KHz.

Mani.


Title: Re: Problems between XX and RME DigiCheck?
Post by: PeterSt on May 27, 2010, 07:51:55 pm
Mani - I was hoping for you to jump in. Thanks a LOT ... as always.

Guys, I am sorry, when I talked about "recording" this is what I meant to imply by using Digicheck. Because that is just "recording" as such ...

I'd like to add :

Quote
If I reduce the buffer and introduce noise, then indeed, there is a lot of HF energy going all the way up to 81KHz.

This is a dangerous thing. I mean, why wouldn't this happen during normal playback and an actually too low buffer size. You may not hear it as such, but it can be there anyway. This is why the knowledge of "is it still bit perfect" is of major importance. Which brings me to this :

Very nice that XXHighEnd can decode HDCD, but not so nice that it can't be shut off. I mean, now you can't use the HDCD light on your processor etc. anymore to have a brief check (ok, most won't have that anyway, but still).

:yes:
Peter


PS: If I overlook something let me know !


Title: Re: Problems between XX and RME DigiCheck?
Post by: pedal on May 27, 2010, 08:02:32 pm
Thanks Mani!

Mystery solved: I just increased the Q1 value some steps up (Q1 = 4). Then everything works glitz-free, I think. (Remains to be confirmed from long term testing).

Which means that my music-PC still can be a Foobar-free zone. :-)

MANI: You should check out the Global Record function. It's like a digital tape recorder. Just one push and whatever you are playing is recorded in WAV, same resolution as source.


Title: Re: Problems between XX and RME DigiCheck?
Post by: manisandher on May 27, 2010, 08:07:41 pm
Very nice that XXHighEnd can decode HDCD, but not so nice that it can't be shut off. I mean, now you can't use the HDCD light on your processor etc. anymore to have a brief check (ok, most won't have that anyway, but still).

Oh, but I can... as long as it's a hirez HDCD. HDCD indicator on DAC is coming on with buffer=32 and Q1=1, so I guess no noise here...

Mani


Title: Re: Problems between XX and RME DigiCheck?
Post by: PeterSt on May 27, 2010, 08:09:10 pm
One other thing before you start thinking about it and I created additional confusion :

At normal playback, KS Adaptive Mode can not be other than bit perfect, unless you have buffer errors (and a few per track are perfectly allowed although not 100% good of course). With KS Normal Mode and Engine#3 this shouldn't happen, but with KS Special Mode we're kind of asking for troubles. This is where the HDCD light comes handy (but I must make an option to shut off HDCD decoding first).

Thanks,
Peter


Title: Re: Problems between XX and RME DigiCheck?
Post by: PeterSt on May 27, 2010, 08:10:49 pm
Quote
as long as it's a hirez HDCD

Haha, at this moment I can't even think of how it would work (usefully). But I didn't investigate it either.


Title: Re: Problems between XX and RME DigiCheck?
Post by: manisandher on May 27, 2010, 08:22:51 pm
You mean hirez HDCD? I'm not sure what it's doing exactly, but the HDCD flag is there.

Mani.