Topic: Spotify - high quality, legal, free, streaming music.

[DavidW]

Yeah, most people can't tell the difference I'll wager.

+2  I find the extreme setting to be pretty darned cool.

[flyingdutchman]

Yeah, most people can't tell the difference I'll wager.

I can and so can many discerning listeners.   I still buy CDs, but focus a lot on SACD quality sound.  Compression and the increase in noise via MP3 does not make a recording better nor equal to higher resolution, friends. 

[DavidW]

I can and so can many discerning listeners.   I still buy CDs, but focus a lot on SACD quality sound.  Compression and the increase in noise via MP3 does not make a recording better nor equal to higher resolution, friends.

Compression artifacts don't sound like noise.

[flyingdutchman]

Compression artifacts don't sound like noise.

No, they don't, but what's your point?  I used the word and for a reason.

[DavidW]

No, they don't, but what's your point?  I used the word and for a reason.

My point is that mp3s are not noisy.

[flyingdutchman]

And my point is that increased loudness is noise.  Increased loudness amounts to noise and lack of clarity and resolution.  Loudness, noise and fatigue is a constant complaint among listeners today.

[DavidW]

And my point is that increased loudness is noise.  Increased loudness amounts to noise and lack of clarity and resolution.

You're talking about dynamic compression, which is NOT what mp3s do.  They compress in the frequency range, but not the volume range.  Also a compressed dynamic range only results in noise if it hits the maximum peak which only occurs in poorly mastered pop music, and has nothing to do with mp3 vs other forms of compression.

[flyingdutchman]

You're talking about dynamic compression, which is NOT what mp3s do.  They compress in the frequency range, but not the volume range.  Also a compressed dynamic range only results in noise if it hits the maximum peak which only occurs in poorly mastered pop music, and has nothing to do with mp3 vs other forms of compression.

This engineer says it best:
http://www.youtube.com/watch?v=Do1FJ5BcqSY

And classical music is not immune to poorly mastered recordings, but the compression found with MP3s increase loudness (noise) to the point that you are losing resolution, clarity, and increasing listener fatigue.  It happens to me whenever I listen to an MP3 of any music.  Higher resolution SACDs improve in all aspects of clarity, musical beauty, and eliminate listener fatigue. 

[DavidW]

The loudness war has nothing to do with mp3 compression, you either misunderstood what you've seen or the presenter in the link misspoke.  I'm not going to watch the link because I already understand both the loudness war and how mp3 compression works.  You are simply confusing two types of compression.

[71 dB]

My point is that mp3s are not noisy.

As a matter of fact they are noisy because of information compression. The noise is just hopefully completely masked by the music itself. That's the whole idea of lossy audio formats.

[DavidW]

As a matter of fact they are noisy because of information compression. The noise is just hopefully completely masked by the music itself. That's the whole idea of lossy audio formats.

I'm more alert now so I'm going to try this again... and hopefully add enough information that others will find it useful.

Mp3 compression is based off of removing parts of the music that is masked by other parts to be inaudible.  The scheme does not mask noise, it is removing parts of the waveform that are not audible to the human ear.  The result of overly compressed music is not noise... it's lack of detail.

Another way that mp3 compression works is by filtering out high frequency (treble) and low frequency (deep bass) because our ears are much less sensitive to those ranges as compared to mid range, and we are less likely to miss it.  And again the result is lack of detail, lack of sparkle and oomph.  It sounds less detailed and lively... but it doesn't sound noisy.

So what is a compression artifact?  When trimming the waveform at too low a bitrate too much is removed and the sound is distorted.  Not noisy but distorted.  A concrete example that most are familiar with is cell phone conversation compressed at 32 k.  Good enough to make out what is being said and who you are talking to, but the voice doesn't sound natural, it doesn't sound noisy just a bit distorted.

Alot of streaming music is at 64k and you sometimes hear this distortion, but mostly it just sounds lacking in detail and dull.

At 128k many people start to hear the sound as transparent-- sufficient detail and lacking in these kinds of transients that you can hear at 64k.  Not as dull either but it could still sound better.  You need better gear to need better bitrate (as opposed to earbuds on mobile devices).

Many blind tests put 160-192 range as reaching transparency even on quality headphones, amps, speakers. 

Now if you want to develop a discerning ear, stop listening to 320k mp3 and pretending that it sounds bad (it sounds great and nobody has ears good enough to distinguish between 320k and flac).  Start listening to really low bitrate mp3s and work your way up.  I really think that when people talk about mp3s sounding noisy they are just picking up the noise floor on their crappy phone.

[Leon]

I'm more alert now so I'm going to try this again... and hopefully add enough information that others will find it useful.

Mp3 compression is based off of removing parts of the music that is masked by other parts to be inaudible.  The scheme does not mask noise, it is removing parts of the waveform that are not audible to the human ear.  The result of overly compressed music is not noise... it's lack of detail.

Another way that mp3 compression works is by filtering out high frequency (treble) and low frequency (deep bass) because our ears are much less sensitive to those ranges as compared to mid range, and we are less likely to miss it.  And again the result is lack of detail, lack of sparkle and oomph.  It sounds less detailed and lively... but it doesn't sound noisy.

So what is a compression artifact?  When trimming the waveform at too low a bitrate too much is removed and the sound is distorted.  Not noisy but distorted.  A concrete example that most are familiar with is cell phone conversation compressed at 32 k.  Good enough to make out what is being said and who you are talking to, but the voice doesn't sound natural, it doesn't sound noisy just a bit distorted.

Alot of streaming music is at 64k and you sometimes hear this distortion, but mostly it just sounds lacking in detail and dull.

At 128k many people start to hear the sound as transparent-- sufficient detail and lacking in these kinds of transients that you can hear at 64k.  Not as dull either but it could still sound better.  You need better gear to need better bitrate (as opposed to earbuds on mobile devices).

Many blind tests put 160-192 range as reaching transparency even on quality headphones, amps, speakers. 

Now if you want to develop a discerning ear, stop listening to 320k mp3 and pretending that it sounds bad (it sounds great and nobody has ears good enough to distinguish between 320k and flac).  Start listening to really low bitrate mp3s and work your way up.  I really think that when people talk about mp3s sounding noisy they are just picking up the noise floor on their crappy phone.

This reflects my own understanding of compression.  The mp3 files downloaded from Amazon are ~256k and for me sound as good as a CD.  I am aware there is a difference in the size of the file and the amount of digital material that makes up the sound, but the end result (which is a product of the equipment I use and age of my ears) is that a 256k mp3 is perfectly satisfactory to me.

I rarely buy CDs anymore for this reason.

[71 dB]

Noise is not just white or pink noise. It's any kind of signal that doesn't belong to the original recording. So, if you have distortion, you have noise.

This is very easy. Take the original uncompressed sound file (rip it from a CD). Then make a compressed lossy sound file from it (e.g. a mp3 file). Then load the original and lossy
versions to a wave editor and subract the other from the other. Everything that is similar with these files is now gone and the difference is revealed. That's the noise I am talking about.
It's not hopefully audible to you but it is there.

Another problem with high amplitude signals is that they may work perfectly uncompressed but when you start removing information the waveform chances it's shape and it might
become "impossible" within the dynamic range available, hence introducing harmonic distortion. People don't know this an so they make distorted mp3s when the amplitude of the
original signal is close to maximum.

[DavidW]

This reflects my own understanding of compression.  The mp3 files downloaded from Amazon are ~256k and for me sound as good as a CD.  I am aware there is a difference in the size of the file and the amount of digital material that makes up the sound, but the end result (which is a product of the equipment I use and age of my ears) is that a 256k mp3 is perfectly satisfactory to me.

I rarely buy CDs anymore for this reason.

Same here, everything from amazon, itunes and eclassical sound fantastic, so only time I'm going to buy cds will be if I can't buy the download (like the Auryn Q).

[nesf]

Many blind tests put 160-192 range as reaching transparency even on quality headphones, amps, speakers. 

Now if you want to develop a discerning ear, stop listening to 320k mp3 and pretending that it sounds bad (it sounds great and nobody has ears good enough to distinguish between 320k and flac).  Start listening to really low bitrate mp3s and work your way up.  I really think that when people talk about mp3s sounding noisy they are just picking up the noise floor on their crappy phone.

Pretty much my thinking. Now if downloads weren't on average more expensive than the CDs for me to buy...

[Leon]

Pretty much my thinking. Now if downloads weren't on average more expensive than the CDs for me to buy...

I agree.  Used to be you could realize a significant savings by choosing mp3s but today often there is no difference, and I am beginning to see mp3s costing more than the CD.  Pure greed, IMO.

 

[DavidW]

I agree.  Used to be you could realize a significant savings by choosing mp3s but today often there is no difference, and I am beginning to see mp3s costing more than the CD.  Pure greed, IMO.

 

I skipped the Krips Mozart symphony set for that reason, it really annoyed me!!

[Ataraxia]

I haven't noticed that; I probably haven't been paying attention.

Are you comparing, say, a CD from Amazon marketplace to an Amazon download?

[nesf]

I haven't noticed that; I probably haven't been paying attention.

Are you comparing, say, a CD from Amazon marketplace to an Amazon download?

Well Amazon download isn't available to me being Irish. iTunes and eClassical etc are often more expensive than Amazon and JPG and similar for the equivalent CDs. Especially with boxsets.

[DavidW]

Well Amazon download isn't available to me being Irish. iTunes and eClassical etc are often more expensive than Amazon and JPG and similar for the equivalent CDs. Especially with boxsets.

The itunes $10 per album pricing is really inflexible. 

You should look into google music, it's insanely cheap like wow!  They have albums for a quarter and prices that would make amazon blush for being so cheap.

https://play.google.com/music/listen#start_pl