Why do digital sources sound different to each other?

[Leif]

No.
The important thing is that CD drives in computers read the disc multiple times, if they miss some data they go back and read it again, and again, and again. The '30 X speed' allows the drive to make multiple sweeps in quick time, it is not necessarily 30 X linear speed.

Hi-fi CD transports get one go in real time, data on CDs has a degree of redundancy, this enables some data to be missed but the data stream remains 100% accurate. If sufficient data is lost that the data stream can not be 100% accurate, then the transport's error correction circuit is called on.

It is my opinion that this final factor, error correction, is responsible for the differences heard though to be fair there is no objective proof of this.

I take your point that the drive is operating in a different mode. Nevertheless, my decent (but extremely cheap by hi-fi standards) PC drive is ripping the disc with bit perfect accuracy in typically 1/20 of the time it would take to play it (dbPowerAmp with AccurateRip enabled). If it can do that, surely it is not unreasonable to expect that a drive could get the data off the disc in 'real time' with extremely good accuracy.

I wondered about that. However, a PC has far more processing power than a CD drive, and can control the hardware in a more sophisticated way. So it may well be the case that the standalone CD player can only do a one pass read strategy. That said, you'd think the higher end ones would do multiple reads. I'm sure an embedded Linux system could handle that. Maybe the makers prefer to throw in sophisticated error correction for the one pass strategy instead. It probably avoid bringing in new expertise.

 

[davedotco]

No.
The important thing is that CD drives in computers read the disc multiple times, if they miss some data they go back and read it again, and again, and again. The '30 X speed' allows the drive to make multiple sweeps in quick time, it is not necessarily 30 X linear speed.

Hi-fi CD transports get one go in real time, data on CDs has a degree of redundancy, this enables some data to be missed but the data stream remains 100% accurate. If sufficient data is lost that the data stream can not be 100% accurate, then the transport's error correction circuit is called on.

It is my opinion that this final factor, error correction, is responsible for the differences heard though to be fair there is no objective proof of this.

I take your point that the drive is operating in a different mode. Nevertheless, my decent (but extremely cheap by hi-fi standards) PC drive is ripping the disc with bit perfect accuracy in typically 1/20 of the time it would take to play it (dbPowerAmp with AccurateRip enabled). If it can do that, surely it is not unreasonable to expect that a drive could get the data off the disc in 'real time' with extremely good accuracy.

I wondered about that. However, a PC has far more processing power than a CD drive, and can control the hardware in a more sophisticated way. So it may well be the case that the standalone CD player can only do a one pass read strategy. That said, you'd think the higher end ones would do multiple reads. I'm sure an embedded Linux system could handle that. Maybe the makers prefer to throw in sophisticated error correction for the one pass strategy instead. It probably avoid bringing in new expertise.

I think you underestimate just how complex it is to get all the information in one pass. The data on a CD is recorded with a degree of redundency so that minor reading errors are corrected 100%, less minor errors need error correction, so it is not that simple.

I have a vague recollection of a multi scan player being developed some years back but I can not remember any names or details. Though if you think about it, process wise, this is the same as ripping a CD and streaming it after the fact.

 

[BobWH]

... Though if you think about it, process wise, this is the same as ripping a CD and streaming it after the fact.

Quite. Which leads to the question: if it so difficult to get an accurate real time data stream from a CD why bother with it, and why not just rip and stream? After all, it is relatively straightforward to determine that a bit perfect rip has been obtained, so one can say with certainty that the data from a real time play of the same CD is at best going to be exactly the same or, possible more likely, a bit worse.

 

[davedotco]

... Though if you think about it, process wise, this is the same as ripping a CD and streaming it after the fact.

Quite. Which leads to the question: if it so difficult to get an accurate real time data stream from a CD why bother with it, and why not just rip and stream? After all, it is relatively straightforward to determine that a bit perfect rip has been obtained, so one can say with certainty that the data from a real time play of the same CD is at best going to be exactly the same or, possible more likely, a bit worse.

An argument that many people have been putting forward for some time.

Linn used just this argument when they dropped CD players in favour of streamers.

 

[Andrewjvt]

... Though if you think about it, process wise, this is the same as ripping a CD and streaming it after the fact.

Quite. Which leads to the question: if it so difficult to get an accurate real time data stream from a CD why bother with it, and why not just rip and stream? After all, it is relatively straightforward to determine that a bit perfect rip has been obtained, so one can say with certainty that the data from a real time play of the same CD is at best going to be exactly the same or, possible more likely, a bit worse.

Because (and I'm not judging here) some people still prefer to stick in a physical disc.

I don't understand why but its a personal preference for some.

 

[Al ears]

... Though if you think about it, process wise, this is the same as ripping a CD and streaming it after the fact.

Quite. Which leads to the question: if it so difficult to get an accurate real time data stream from a CD why bother with it, and why not just rip and stream? After all, it is relatively straightforward to determine that a bit perfect rip has been obtained, so one can say with certainty that the data from a real time play of the same CD is at best going to be exactly the same or, possible more likely, a bit worse.

Because (and I'm not judging here) some people still prefer to stick in a physical disc.

I don't understand why but its a personal preference for some.

True. It's not actually that difficult to get accurate real-time data from a CD either.... that's what they were designed to do.

Play a CD in a CD player, now play a perfect RIP of that CD through whatever equipment you like. Hear any problems???

Thought not. ;-)

 

[Leif]

I think you underestimate just how complex it is to get all the information in one pass. The data on a CD is recorded with a degree of redundency so that minor reading errors are corrected 100%, less minor errors need error correction, so it is not that simple.

I have a vague recollection of a multi scan player being developed some years back but I can not remember any names or details. Though if you think about it, process wise, this is the same as ripping a CD and streaming it after the fact.

Yes, that was exactly my reasoning. But you are not ripping the whole CD, you are ripping it fast enough to buffer the stream for playback.

 

[paulkebab]

rip it. I'll take the lot lol.. I was with it until that last comment rip it fast enough to stream?? A rip is a rip, stream it in real time coz it won't stream any faster or it'll sound really fast.

 

[Leif]

rip it. I'll take the lot lol.. I was with it until that last comment rip it fast enough to stream?? A rip is a rip, stream it in real time coz it won't stream any faster or it'll sound really fast.

I can rephrase it if you like. I meant read it faster than usual, so there is time to read each bit multiple times to allow for read errors, and then buffer the read data in memory until it has been played (streamed) to the digital output or th internal DAC.

 

[chebby]

Because (and I'm not judging here) some people still prefer to stick in a physical disc.

I don't understand why but its a personal preference for some.

I keep all of my CDs after rippng them. (In lots of black Rymans CD storage boxes in the cupboard under the stairs.)

I like to retain the possibility of playing them again even if it seldom actually happens.

Otherwise they are all ripped in 320kbps AAC VBR (with correction) for speech/drama/comedy/documentary, or ALAC (music) and played via AirPlay from my Mac Mini, or from synced libraries/playlists on my iPhones (6 and 7) & iPad Mini 3 (all 128GB).

I also pay £9.99 every month for streaming Apple Music and should use it more than I do

 

[chebby]

Some history: About 9 years ago I bought my first DAC (Firestone Audio Fubar USB and 'Supplier' PSU) to play from a laptop. It partnered an Arcam Solo-Mini. (I sold the DAC and PSU to 'LDC' an ex member here.)

After a year or so I sold the lot to get a Naim Nait 5i-2, CD5i-2, NAT05 system and bought a Beresford TC-7520 DAC (and fitted LM4562NA op-amp replacements).

This DAC proved so enjoyable that I had 'retired' the CD5i after a few months! (It stayed in it's box until I sold the whole system two years later.)

Since then it's been integrated DACs. (Marantz M-CR6xx and Quad Vena.)