Disappointing Sound Quality from Ripped CDs

[thewinelake.]

This is similar to the other thread wondering how the Rega SaturnR can sound so great. Seems to me that jitter may well be more audible than one would believe.

 

[andyjm]

Although Sonos is strictly 16-bit the output is actually 24-bit, giving it 8-bits of volume control before the signal is interfered with in any way, so you can turn it down pretty low before anything should change.

True for the analogue output, but the OP is using the digital output. Clearly the digital information has to be 'interfered with' or the volume would remain the same. Using the digital volume control in the Connect in this manner will degrade the sound quality of the digital output.

Your advice about setting the output to fixed will ensure there is no loss of quality.

https://en.community.sonos.com/troubleshooting-228999/zp80-output-specif...

This post explains, 8 years ago, how the zp80 (the Connect's predecessor) gives volume control down to -48db without losing any audio data.

So, just to confirm, the data is 'interfered with' by using the digital volume control. The data coming out the Connect is not the same as the data going into the Connect unless the volume control is 100%. This is exactly what you would expect - as I pointed out earlier, if the data was the same, the volume would be the same.

The article you linked to points out that with 16bit data, the Connect can use the extra resolution of a downstream 24 bit DAC to reduce the volume without reducing the resolution as it has 8 bits to play with that otherwise would have been wasted. There is some merit to this argument, but it does assume that the downstream DAC is 24bit and that the Connect is getting its maths right.

The safest solution (and if you are doing an A/B comparison like the OP the only solution) is to keep the Connect output at 100%, then you know the output data will be the same as the CD transport and a valid comparison can be made.

 

[andyjm]

This is similar to the other thread wondering how the Rega SaturnR can sound so great. Seems to me that jitter may well be more audible than one would believe.

There are a number of studies published that try to establish the point at which jitter becomes audible. It is a complex subject as it is not just the amount of jitter, but the type of jitter and its distribution that seems to matter.

When I last looked at this, the Connect was one of the poorer performing transports when it came to jitter performance. For those who care, the Squeezebox Transporter (at 15pS RMS) is I believe still the best performing transport in this respect.

None of this matters if the downstream DAC employs jitter mitigation strategies. According to NAD, the resampled PWM approach used in their DAC (in the case of the OP) makes it completely immune to input jitter. If this is so, then it shouldn't matter how badly jittered the input S/PDIF signal is, unless it causes bit errors - and I have never seen a transport that bad.

Assuming NAD know what they are talking about, there will be no difference in sound quality between the Connect and his transport (with the connect at 100% volume). As he appears to be able to tell the difference, there is either something amiss somewhere or our old friend expectation bias has shown up to the party.

 

[The_Lhc]

but it does assume that the downstream DAC is 24bit

I doubt there have been many dacs from the last ten years that weren't 24-bit capable, the chipsets are dirt cheap.

 

[Hugh50]

Thanks for all the suggestions; the problem was the Sonos volume control. I switched this to fixed and it's hard to hear much difference. I only played one track, but that's all it took to confirm the problem. I still think the Transport was marginally better, but I would have to play a lot more tracks to be sure; if it is, based on the huge difference the volume control made, I think it's likely to be the Sonos being in-line that's the weak link.

Later today I'll try outputting directly from the Synology to the DAC, cutting out the Sonos. This won't be practical for daily use, but will be interesting to know if it's worth upgrading the Sonos in future.

One thing is clear - don't use a Sonos volume control unless it's for practical purposes; the difference is dramatic; at least to my ears.

 

[BigH]

Thanks for all the suggestions; the problem was the Sonos volume control. I switched this to fixed and it's hard to hear much difference. I only played one track, but that's all it took to confirm the problem. I still think the Transport was marginally better, but I would have to play a lot more tracks to be sure; if it is, based on the huge difference the volume control made, I think it's likely to be the Sonos being in-line that's the weak link.

Later today I'll try outputting directly from the Synology to the DAC, cutting out the Sonos. This won't be practical for daily use, but will be interesting to know if it's worth upgrading the Sonos in future.

One thing is clear - don't use a Sonos volume control unless it's for practical purposes; the difference is dramatic; at least to my ears.

Thats good thought it may have been something like that.

How do you control the volume now?

 

[The_Lhc]

Thanks for all the suggestions; the problem was the Sonos volume control.  I switched this to fixed and it's hard to hear much difference.  I only played one track, but that's all it took to confirm the problem.  I still think the Transport was marginally better, but I would have to play a lot more tracks to be sure; if it is, based on the huge difference the volume control made, I think it's likely to be the Sonos being in-line that's the weak link.

Later today I'll try outputting directly from the Synology to the DAC, cutting out the Sonos.  This won't be practical for daily use, but will be interesting to know if it's worth upgrading the Sonos in future.

One thing is clear - don't use a Sonos volume control unless it's for practical purposes; the difference is dramatic; at least to my ears.

 

What sort of level did you have the Sonos volume control at previously?

 

[Hugh50]

Thanks for all the suggestions; the problem was the Sonos volume control. I switched this to fixed and it's hard to hear much difference. I only played one track, but that's all it took to confirm the problem. I still think the Transport was marginally better, but I would have to play a lot more tracks to be sure; if it is, based on the huge difference the volume control made, I think it's likely to be the Sonos being in-line that's the weak link.

Later today I'll try outputting directly from the Synology to the DAC, cutting out the Sonos. This won't be practical for daily use, but will be interesting to know if it's worth upgrading the Sonos in future.

One thing is clear - don't use a Sonos volume control unless it's for practical purposes; the difference is dramatic; at least to my ears.

Thats good thought it may have been something like that.

How do you control the volume now?

The M51 Dac has digital volume control that works without affecting sound quality. I set the Sonos to use it's own volume control for convenience to use with apps, so it was set to whatever level the listener required; it's easy to switch on and off, so when convenience more important than quality I'll revert to the Sonos control temporarily.

 

[MajorFubar]

i have found that flac is inferior to wav when listening to hi-res music. Try a couple of differnet codecs like wma lossless and even wav on one song and see. but would highly recommend starting with wav and working backwards, as this will eliminate the loss factor on the files and point you towards any other potential problems with your set up.

With the OP's problem solved, I feel less quilty taking this thread slightly off track and pointing out for the benefit of future clarification that the differences you think you hear cannot be there and they are nothing more than a perfect example of expectation bias. Not one of us is immune to it, but unless you truly 'get' that FLAC and ALAC are basicaly ZIP files for audio and they unpack on-the-fly into the original data bit for bit without corruption or loss, you probably won't ever be able to unhear the differences you imagine are there, or truly accept they are imaginary. It's even possible to prove FLAC files are identical t the original WAVS with a null test, and this has been done, many times. This is coming from someone who for a few stupid minutes was once convinced that rips of the same CD from two DVD drives actually sounded different. So like I said, none of us are immune, even those of us who know better.

 

[BigH]

Thanks for all the suggestions; the problem was the Sonos volume control. I switched this to fixed and it's hard to hear much difference. I only played one track, but that's all it took to confirm the problem. I still think the Transport was marginally better, but I would have to play a lot more tracks to be sure; if it is, based on the huge difference the volume control made, I think it's likely to be the Sonos being in-line that's the weak link.

Later today I'll try outputting directly from the Synology to the DAC, cutting out the Sonos. This won't be practical for daily use, but will be interesting to know if it's worth upgrading the Sonos in future.

One thing is clear - don't use a Sonos volume control unless it's for practical purposes; the difference is dramatic; at least to my ears.

Thats good thought it may have been something like that.

How do you control the volume now?

The M51 Dac has digital volume control that works without affecting sound quality. I set the Sonos to use it's own volume control for convenience to use with apps, so it was set to whatever level the listener required; it's easy to switch on and off, so when convenience more important than quality I'll revert to the Sonos control temporarily.

I see, how does the Sonus sound if you set volume to 100%.

 

[Benedict_Arnold]

Try ripping a song in WAV or AIFF.

Or FLAC.

 

[busb]

Thanks for all the suggestions; the problem was the Sonos volume control. I switched this to fixed and it's hard to hear much difference. I only played one track, but that's all it took to confirm the problem. I still think the Transport was marginally better, but I would have to play a lot more tracks to be sure; if it is, based on the huge difference the volume control made, I think it's likely to be the Sonos being in-line that's the weak link.

Later today I'll try outputting directly from the Synology to the DAC, cutting out the Sonos. This won't be practical for daily use, but will be interesting to know if it's worth upgrading the Sonos in future.

One thing is clear - don't use a Sonos volume control unless it's for practical purposes; the difference is dramatic; at least to my ears.

Thats good thought it may have been something like that.

How do you control the volume now?

The M51 Dac has digital volume control that works without affecting sound quality. I set the Sonos to use it's own volume control for convenience to use with apps, so it was set to whatever level the listener required; it's easy to switch on and off, so when convenience more important than quality I'll revert to the Sonos control temporarily.

A friend brought over his M51 some time ago to compare with my M-DAC - it (M51) was definitely better. Both do digital volume pretty effectively but not perfectly (both convert to 32bit internally so have quite some headroom for volume manipulation).

I no longer feed my M-DAC's output directly into my power amp but now use the analogue volume control on my fairly new preamp due to wanting to add a record player later this year. The difference between the two methods is minimal to the point where I can't tell which is better

 

[manicm]

This is coming from someone who for a few stupid minutes was once convinced that rips of the same CD from two DVD drives actually sounded different. So like I said, none of us are immune, even those of us who know better.

I'm actually not so sure about this. EAC has behaved totally differently on a laptop I had for 6 years and my new one. On my new one the drive does not like EAC optimised for error correction at all - it's extremely slow. I've thus optimised EAC for speed, and yet I think rips on my newer laptop sound better.

 

[MajorFubar]

This is coming from someone who for a few stupid minutes was once convinced that rips of the same CD from two DVD drives actually sounded different. So like I said, none of us are immune, even those of us who know better.

I'm actually not so sure about this. EAC has behaved totally differently on a laptop I had for 6 years and my new one. On my new one the drive does not like EAC optimised for error correction at all - it's extremely slow. I've thus optimised EAC for speed, and yet I think rips on my newer laptop sound better.

Unlike analogue where there are uncertainties and vagaries there's no such thing with digital files. Two rips are either the same or they are not. You can easily prove it one way or the other by loading them both into an audio editor like Audacity, on separate tracks. Align them up sample-perfect (that's absolutely critical, you can't just align the start of the files together and assume that's it), and if you get complete silence when you invert the phase of one file, they are identical, because they've mathematically cancelled each other out to null. This is what's called the null test, and its outcome (either way) is incontrovertable. That's what I did, and silence is what I got. I had been fooled.

 

[manicm]

This is coming from someone who for a few stupid minutes was once convinced that rips of the same CD from two DVD drives actually sounded different. So like I said, none of us are immune, even those of us who know better.

I'm actually not so sure about this. EAC has behaved totally differently on a laptop I had for 6 years and my new one. On my new one the drive does not like EAC optimised for error correction at all - it's extremely slow. I've thus optimised EAC for speed, and yet I think rips on my newer laptop sound better.

Unlike analogue where there are uncertainties and vagaries there's no such thing with digital files. Two rips are either the same or they are not. You can easily prove it one way or the other by loading them both into an audio editor like Audacity, on separate tracks. Align them up sample-perfect (that's absolutely critical, you can't just align the start of the files together and assume that's it), and if you get complete silence when you invert the phase of one file, they are identical, because they've mathematically cancelled each other out to null. This is what's called the null test, and its outcome (either way) is incontrovertable. That's what I did, and silence is what I got. I had been fooled.

Yes this is regurgitated often, and I don't necessarily disagree with you, but my current laptop drive renders EAC unusable if I optimise the settings for accuracy over speed, which leads to me to believe that drives can have a bearing on the rip. Is it not possible that the ripping software can't deal with drive quirks?

 

[BigH]

This is coming from someone who for a few stupid minutes was once convinced that rips of the same CD from two DVD drives actually sounded different. So like I said, none of us are immune, even those of us who know better.

I'm actually not so sure about this. EAC has behaved totally differently on a laptop I had for 6 years and my new one. On my new one the drive does not like EAC optimised for error correction at all - it's extremely slow. I've thus optimised EAC for speed, and yet I think rips on my newer laptop sound better.

Unlike analogue where there are uncertainties and vagaries there's no such thing with digital files. Two rips are either the same or they are not. You can easily prove it one way or the other by loading them both into an audio editor like Audacity, on separate tracks. Align them up sample-perfect (that's absolutely critical, you can't just align the start of the files together and assume that's it), and if you get complete silence when you invert the phase of one file, they are identical, because they've mathematically cancelled each other out to null. This is what's called the null test, and its outcome (either way) is incontrovertable. That's what I did, and silence is what I got. I had been fooled.

Yes this is regurgitated often, and I don't necessarily disagree with you, but my current laptop drive renders EAC unusable if I optimise the settings for accuracy over speed, which leads to me to believe that drives can have a bearing on the rip. Is it not possible that the ripping software can't deal with drive quirks?

Some software compares your rip with all the other rips it has done for that cd then it fixes any issues. On one of mine most tracks were good, 99.9% accurate, 1 track was out a bit and it corrected it but I could not tell a difference. So yes it can deal with quirks as long as it has enough other rips to compare with.

 

[MajorFubar]

Yes this is regurgitated often, and I don't necessarily disagree with you, but my current laptop drive renders EAC unusable if I optimise the settings for accuracy over speed, which leads to me to believe that drives can have a bearing on the rip. Is it not possible that the ripping software can't deal with drive quirks?

Not sure tbh. There have only been about 3 CDs out of my 600+ CD collection which my USB DVD ROM struggled to rip. All three were long CDs and in all instances the problem occurred with just the last couple of tracks. The rips of these tracks had noticeable skips and crackles. I forget what two of the CDs were, but I think the third was one of the Stones' "40 Licks" CDs. On all occasions the CDs ripped fine on my iMac's internal drive. But of the tracks which ripped without error, the rips from the USB DVD drive and the iMac's internal drive were absolutely identical once you aligned the waveforms precisely.