I only posted one example. Believe me that there are many many more. Do the test yourself. And I agree about Qobuz being the best for classical.
Hmmm, just played quoted piece and no clipping occurs hereā¦!?
Note, that Iāve done that without any DSP, like volume leveling or headroom adjustments etc, just set it to indicate clipping - Roon states itās showing single sample overs and I did verify that quite some time ago for myself.
Sublime subscriber and I concur that this is the biggest issue for me as well.
I believe this issue pertains to digital downloads?
Another example here:
Seems very odd. I have seen other users of Qobuz in MQA threads here, complaining of clipping also.
I believe a number stated they moved back to Tidal because of this observation.
I personally only stream, and have never had any issues with clipping, period.
Surely if an issue existed, it would be wide spread and easy replicatable by more users.
Iāve just downloaded one of the examples from this post and while itās close to the wind, sox doesnāt think it clips:
⯠sox 01-01-Lukasz_Gottwald-Kiss_Me_More-SMR.flac -n --norm -R gain 0.0 stat
Samples read: 18422040
Length (seconds): 208.866667
Scaled by: 2147483647.0
Maximum amplitude: 0.988553
Minimum amplitude: -0.988553
Midline amplitude: -0.000000
Mean norm: 0.266512
Mean amplitude: -0.000098
RMS amplitude: 0.357025
Maximum delta: 1.635074
Minimum delta: 0.000000
Mean delta: 0.106302
RMS delta: 0.150322
Rough frequency: 2955
Volume adjustment: 1.012
⯠sox 01-01-Lukasz_Gottwald-Kiss_Me_More-SMR.flac -n --norm -R gain 0.09999 stat
Samples read: 18422040
Length (seconds): 208.866667
Scaled by: 2147483647.0
Maximum amplitude: 0.999999
Minimum amplitude: -0.999999
Midline amplitude: -0.000000
Mean norm: 0.269598
Mean amplitude: -0.000099
RMS amplitude: 0.361159
Maximum delta: 1.654005
Minimum delta: 0.000000
Mean delta: 0.107533
RMS delta: 0.152062
Rough frequency: 2955
Volume adjustment: 1.000
⯠sox 01-01-Lukasz_Gottwald-Kiss_Me_More-SMR.flac -n --norm -R gain 0.099999 stat
sox WARN gain: gain clipped 15479 samples; decrease volume?
Samples read: 18422040
Length (seconds): 208.866667
Scaled by: 2147483647.0
Maximum amplitude: 1.000000
Minimum amplitude: -1.000000
Midline amplitude: -0.000000
Mean norm: 0.269598
Mean amplitude: -0.000099
RMS amplitude: 0.361159
Maximum delta: 1.654007
Minimum delta: 0.000000
Mean delta: 0.107533
RMS delta: 0.152062
Rough frequency: 2955
Volume adjustment: 1.000
So very little headroom but no clipping according to sox.
The track will not clip because it is sent clipped, so it stays below 0dB. Please look at it in audacity.
So thatās not āclippingā but more the results of compression/limiting. This isnāt a Qobuz issue, more a mastering one it seems? Is that fair?
Alrighty, understood, but if it had been clipped digitally it would still clearly be audible, even after shifting the level afterwards, Iād say.
Now, Iād buy that - good one, @killdozer!
Weāre obviously in a new aera of the loudness wars, but with classical music this time around!
I have complained to Qobuz, and they say they send them as they get them. But they could check them and complain. I wonder if this is a copy protection thing from the distributors. Buy it to hear the full quality. Has anyone compared a Qobuz clipped track with a bought version of the same track?
What are they going to complain about? Thereās no contractual obligation to produce a particular standard of mastering. Iāve listened to these, analysed them and bought a couple, and thereās no audible clipping. Yeah, theyāll have crappy DR quite likely, but thereās a lot of that about. If Qobuz were to reject them I doubt theyād get another copy, itās just as likely itāll be another gap in the catalogue.
An example of mine that me (Qobuz purchase 44/24):
Havenāt done any technical analysis yet, but for me distortion is clearly audible during louder passages.
However, I have not yet made a comparison with other download versions.
Is there distortion on both the downloaded and streamed version?
Iāll try myself also.
I just listened to it (local download, Tidal and Qobuz). It seems to be on all versions. But on my download version it stands out the most.
Iāve just given track one a listen (Qobuz streamed) and itās not a great recording for sure. Thereās no digital clipping to my ears, but agree it sounds a bit āsaturatedā here and there, struggling when the bass comes in. I see itās on Bandcamp as well, wonder if that sounds the same? Am again thinking itās a recording/mastering issue, rather than a Qobuz one.
Edit Iām liking this enough to buy the Bandcamp version, @florib we could compare sox stats Iām feeling this is going off topic some.
Wait a minute. Iāll purchase it on Bandcamp (donāt know what resolution).
I think this album deserves it.
I will report later.
Please look at the flac using Audacity. You can easily see the clipping there. The Roon display is too small. In Audacity you should see many near-by peaks at the same level.
Both songs (Qobuz and Bandcamp) seem to be identical, at least visually.
Today I am no longer in a position to carry out such analyses. More info maybe tomorrow.
It really is a lot easier to hear clipping than to see it. In this case the waveform looks like it sounds, a bit bloated in the left channel. I still think itās a poor recording of a nice piece of music. Thereās also the possibility that it sounds as intended.
What we hear is actually what matters.
I would never have criticized this album/song if there wasnāt an audible deficit for me.
And I have to admit that my hearing is clearly damaged.