Rank: Member
Groups: Member
Joined: 9/28/2011(UTC) Posts: 31 Location: Germany
Thanks: 2 times
|
Hallo, I have a question about the Buffallo II´s (or III´s) behavior when you input a digital signal of more than 100%. For example, I use Jriver media playback software, and under normal conditions the output level shown in the program goes often above 100% (normal: 90-105%). Of course this increases when one plays around with digital volume controls ore equalizers, which sometimes let the volume to be increased up to 200% (VLC-media player) or more. In jriver there is an explicit setting to prevent this extra 1-6% level increase over the 100% during normal playback. So I´m not sure if I should use it, I guess it compresses the dynamic range, or something... Of course this is all in the digital domain, so I´m not sure if "clipping" is the right term. But does this digital "level enhancement" influence the analog stability of my hifi-system. For example, if I increase the digital volume too much, does the buffalo´s I/V stage start to clip? Is it all right to let the level go over 100% in a reasonable range? A little bit of internet research showed that the real life clipping behavior depends on the DAC construction. 0DBFS+I own the nuforce uDAC also, and this one apparently DOES clip in the analog output when the digital volume is at 100% (because of the extra 1-6% in digital level). So, how does the buffalo DAC behave in this area? Thank you, Florian
|
|
|
|
Rank: Administration
Groups: Administration, Customer Joined: 10/24/2006(UTC) Posts: 3,979 Location: Nashville, TN
Thanks: 25 times Was thanked: 89 time(s) in 83 post(s)
|
The best solution is never to feed such a signal to your DAC, Boosting beyond 0dbfs is not a great idea to begin with. What I do know is that digital attenuation in the ES9018 does not ever boost the signal it only ever attenuates or leaves the signal alone. So 0dbfs is exactly 0dbfs. For even the possibility for the DAC to pass a clipped signal it would have to come in clipped to begin with. This is a question best posed to ESS as really the datasheet does not describe the actual behavior when a signal comes in clipped to start with. Edited by user Saturday, October 15, 2011 2:51:12 PM(UTC)
| Reason: Not specified
|
|
|
|
Rank: Member
Groups: Member
Joined: 9/28/2011(UTC) Posts: 31 Location: Germany
Thanks: 2 times
|
OK, thanks.
I guess that the "normal" jriver output with just a little over 100% level on some recordings should be fine, and just increase distortion on the output a little bit. But to be safe, I´ll check the "do not exeed maximum"-button in the dsp settings, so output is limited to 100%. I guess I set up an email to ESS as well, if I have the time.
Thanks Florian
|
|
|
|
Rank: Administration
Groups: Administration, Customer Joined: 10/24/2006(UTC) Posts: 2,868 Location: Massachusetts, USA
Thanks: 2 times Was thanked: 141 time(s) in 134 post(s)
|
I think it has the very real possibility of clipping the output stage, and providing a nice dose of DC (which is what clipping is) to your amp and speakers. Edited by user Saturday, October 15, 2011 5:09:22 PM(UTC)
| Reason: Not specified
|
|
|
|
Rank: Member
Groups: Member
Joined: 12/24/2008(UTC) Posts: 100 Location: Seattle, WA
|
Unless you ask Jriver to perform any manipulation of the data it will output exactly what the source bits are and there should be no need to activate additional clipping protection. The default is bit transparent to the sound driver.
Then it depends on the sound driver whether you have any manipulation. ASIO, Kernel streaming, and both WASAPI modes are bit transparent. Jriver's volume control only attenuates.
Cheers
Thomas
|
|
|
|
Rank: Member
Groups: Member
Joined: 9/28/2011(UTC) Posts: 31 Location: Germany
Thanks: 2 times
|
Hallo Thomas, ok, that is clear. What I find irritating is that jriver shows me over 100% level on very "loud" recordings, although I don´t use data manipulation and have ASIO output to a dedicated rme sound card. Edited by user Monday, October 17, 2011 11:05:43 AM(UTC)
| Reason: Not specified
|
|
|
|
Rank: Member
Groups: Member
Joined: 12/24/2008(UTC) Posts: 100 Location: Seattle, WA
|
Hallo!
I don't know how they compute the percentages. It could be that they indicate clipping in the source stream (multiple consecutive samples at full scale) with a level over 100%. However, no attenuation or soft limiter can restore the signal at that point.
In fact once the limiter kicks in it affect all the samples, so I would still suggest to leave the bits alone. I seem to remember that the RME card ship with a bit inspection tool. What does that show?
Cheers
Thomas
P.S.: I am using Jriver with a VST based convolver for room correction so I intentionally change the bits but the curve has been normalized with a constant attenuation as to avoid any accidental clipping.I came to realize that the discussions about linearity and distortion are pretty academic if your room response has 10db fluctuations and the RT60 times are not uniform across the frequency band.
|
|
|
|
Forum Jump
You cannot post new topics in this forum.
You cannot reply to topics in this forum.
You cannot delete your posts in this forum.
You cannot edit your posts in this forum.
You cannot create polls in this forum.
You cannot vote in polls in this forum.