Are CD drives really this inaccurate? How could our software even work?

[maxhodges]

Hi, I'm new to dbpoweramp (just downloaded it a few minutes ago). I'm very surprised to hear that 2% of all ripped CDs have errors. How is that possible? We use CD drives to install software. If the error rate were that high, our software wouldn't run correctly. Surely computer CD and DVD drives are able to read billions of bits without a single mistake. How could CD players reliably play music? If a single bit were off it could be inaudible (if it was the least significant bit), or we'd be jolted by loud clicks if it were the most significant bit. I find it really hard to believe or understand why the ability to read a disc would be so unreliable? Surely this isn't true right?

[Dat Ei]

Re: Are CD drives really this inaccurate? How could our software even work?

Hey maxhodges,

How could CD players reliably play music? If a single bit were off it could be inaudible (if it was the least significant bit), or we'd be jolted by loud clicks if it were the most significant bit. I find it really hard to believe or understand why the ability to read a disc would be so unreliable? Surely this isn't true right?

CD player have a special handling for errors of Audio CDs. In case of read errors they can interpolate values. As long as there ain't too much errors, you will not notice anything.

My own experience is that I have 27 defective CDs with permanent read errors in my collection of about 2.400 CDs. That's a bit more than 1%, but I handle my CDs very, very carefully. My new CDs of the last 1 1/2 year have been only 2 to 3 times in a drive, because I've ripped them right after the unboxing. Since then they remained in the shelf.


Dat Ei

[maxhodges]

Re: Are CD drives really this inaccurate? How could our software even work?

So despite handling them very carefully, if 1% don't read correctly that could mean catastrophic failure in the case of installing software. Seems the computer industry couldn't tolerate that. These days we install most software online but there was a period of a decade or more when the majority of all software (operating systems included) was installed from CD and DVD. Seems very unlikely that a 1% fallout rate would have been tolerated by the computer industry.

what do you make of this article? Complete bullshit or is he just using $2000 CD transports which happen to be a hell of a lot more accurate (perfect in fact)?

Do you hear random loud clicks when playing CDs?

The reason you don&*8217;t hear such bit errors is because the CD format has a sophisticated error detection and correction system to prevent such errors from ever reaching the CD player&*8217;s DACs (or a transport&*8217;s digital output jack). In fact, CD&*8217;s error correction system can completely correct an error of up to 4000 missing consecutive bits. We&*8217;re not talking about interpolation (filling in missing information with a best-guess approximation) or error concealment, but full, 100% correction of those 4000 bits.

Errant or missing data longer than 4000 consecutive bits are interpolated so that the error is less audible. If the error is too great to be interpolated, the CD player simply mutes the output. The proponents of the &*8220;bit error&*8221; theory explaining sonic differences between transports suggest that this interpolation is nearly continuous and affects such sonic parameters as soundstage depth and reproduction of timbre.

My experience suggests otherwise. When I worked in a CD mastering lab, we converted Philips CD players into CD error analyzers for the factory&*8217;s QC department. We tapped into the flags in the error correction chips for presentation to a PC running custom software to plot the errors. This allowed us to see exactly the frequency and severity of data errors on CDs. CD data errors are categorized by a letter and two numbers that tell you the error&*8217;s severity with great precision. I&*8217;ll spare you the details here, but suffice to say that we could determine with tremendous accuracy exactly what was going on in the error correction circuits, and if any uncorrected errors occurred.

Uncorrected errors (interpolation) not only never occured, but never even came close to occurring unless the disc was damaged. In fact, the CD&*8217;s error correction system is far more robust than it needs to be.

Moreover, I sometimes performed a bit-for-bit comparison between a CD master tape and the CD replicated from that tape. I did this on about ten different CD/master tape pairs, representing perhaps 56 billion bits of data. Not once did the system detect even a single bit error.

http://www.theabsolutesound.com/articles/transport-bit-accuracy-two-theories-explaining-the-sonic-differences-between-cd-transports-1/

[Spoon]

Re: Are CD drives really this inaccurate? How could our software even work?

Software (DATA CDs) use a different format with way more error correction than audio cds.

[mjt5282]

Re: Are CD drives really this inaccurate? How could our software even work?

Our ears do not have API integration into the accurate rip database so we are not aware when the scratches, smudges, poor pressings, etc impair the perfect transfer of digital audio going down our ear canals into our brains... Error correction and interpolation also occur "in real time" in CD players. Anyway, the world is moving to checksummed TCP/IP downloads and away from physical media. I also rip after taking the CD out of the case. Saddest thing happened a couple of weeks ago - taking a CD out of it's case it cracked .... ARRRRRGGGHHHH !!!

[garym]

Re: Are CD drives really this inaccurate? How could our software even work?

CD player have a special handling for errors of Audio CDs. In case of read errors they can interpolate values. As long as there ain't too much errors, you will not notice anything.

This!

[maxhodges]

Re: Are CD drives really this inaccurate? How could our software even work?

OK but that does't explain how, for over a decade, we've installed software and operating systems from CD and DVD media. If the error rate was 2% as AccurateRip reports, our software would fail catastrophically. You can't just interpolate the data. Why doesn't our "File" menu sometimes say "Pile" instead? These days we mostly just download software, but there was a long period of time where hundreds of millions of us were installing everything from CDs, and if the error rate was this high it's just something I can't see the computer industry tolerating. Software wouldn't work so well and big companies like Microsoft would have invested countless millions in creating more reliable media and drives to solve that problem. As a software engineer it just doesn't make sense that you could randomly scramble some bits of code and it wouldn't cause some problems.

[maxhodges]

Re: Are CD drives really this inaccurate? How could our software even work?

found this article which explains error correction and concealment circuitry

http://www.digital-recordings.com/cdcheck/cdmore.html

[maxhodges]

Re: Are CD drives really this inaccurate? How could our software even work?

what you you guys think of this article. His claims about errors never happening seems to contradict AccurateRip. They can't both be right. Maybe he just never has had any scratches or dirt on the CDs he tested?

When I worked in a CD mastering lab, we converted Philips CD players into CD error analyzers for the factory&*8217;s QC department. We tapped into the flags in the error correction chips for presentation to a PC running custom software to plot the errors. This allowed us to see exactly the frequency and severity of data errors on CDs. CD data errors are categorized by a letter and two numbers that tell you the error&*8217;s severity with great precision. I&*8217;ll spare you the details here, but suffice to say that we could determine with tremendous accuracy exactly what was going on in the error correction circuits, and if any uncorrected errors occurred.

Uncorrected errors (interpolation) not only never occured, but never even came close to occurring unless the disc was damaged. In fact, the CD&*8217;s error correction system is far more robust than it needs to be.

Moreover, I sometimes performed a bit-for-bit comparison between a CD master tape and the CD replicated from that tape. I did this on about ten different CD/master tape pairs, representing perhaps 56 billion bits of data. Not once did the system detect even a single bit error.

http://www.theabsolutesound.com/articles/transport-bit-accuracy-two-theories-explaining-the-sonic-differences-between-cd-transports-1/

[garym]

Re: Are CD drives really this inaccurate? How could our software even work?

OK but that does't explain how, for over a decade, we've installed software and operating systems from CD and DVD media. If the error rate was 2% as AccurateRip reports, our software would fail catastrophically. You can't just interpolate the data. Why doesn't our "File" menu sometimes say "Pile" instead? These days we mostly just download software, but there was a long period of time where hundreds of millions of us were installing everything from CDs, and if the error rate was this high it's just something I can't see the computer industry tolerating. Software wouldn't work so well and big companies like Microsoft would have invested countless millions in creating more reliable media and drives to solve that problem. As a software engineer it just doesn't make sense that you could randomly scramble some bits of code and it wouldn't cause some problems.

Not sure, but as a user and not a technical person in this field, I've been installing software programs for almost 40 years using tape, 5.25 floppy drives, 3.5 drives, CD-ROM, DVD-ROM, USB drives, downloads, etc. (as well as making copies of data to floppies, CD-ROM, etc.). Thinking back, I probably have had 1 or 2% failure with all these mediums. With purchased software, I can recall having to get replacement disks, etc.

And I've purchased over 5,000 CDs over the years. And a small percentage of these have been "bad" from the get go and required replacing.

[mville]

Re: Are CD drives really this inaccurate? How could our software even work?

I've been installing software programs for almost 40 years using tape, 5.25 floppy drives, 3.5 drives, CD-ROM, DVD-ROM, USB drives, downloads, etc. (as well as making copies of data to floppies, CD-ROM, etc.). Thinking back, I probably have had 1 or 2% failure with all these mediums. With purchased software, I can recall having to get replacement disks, etc.

I agree. Removable media, prior to flash drives always had a (low) failure rate and I had to get replacements in such cases.

[Dat Ei]

Re: Are CD drives really this inaccurate? How could our software even work?

Hey maxhodges,

As a software engineer it just doesn't make sense that you could randomly scramble some bits of code and it wouldn't cause some problems.

and as a software engineer you will know that optical media don't play a role in the IT industry. Optical media have been very critical especially in the backup sector. In the beginnings there have been big problems with different CD types (CD, CD-R, CD-RW, CD+RW) and different surfaces (silver, golden, bronze). I have in mind that data CDs have checksums to recover and restore from errors of more than one bit.

Aside of that: the use of Audio CDs is different to the one of Data CDs. How often do you use a Data CD and how often do you use a Audio CD? While we have usually CD/DVD drives with trays in the PC, we see a lot of slot-in drives in the audio enviroment of cars.


Dat Ei

[maxhodges]

Re: Are CD drives really this inaccurate? How could our software even work?

and as a software engineer you will know that optical media don't play a role in the IT industry

Maybe you're a lot younger than me (i'm 42), but it most certainly did play a huge role in the IT industry. Starting from around 1988 when the Yellow Book standard was published until the mid-2000s, CD-ROMs were popularly used to distribute software for computers and video game consoles. But seems I found the answer!

Unlike an audio CD, a CD-ROM cannot rely on error concealment by interpolation; a higher reliability of the retrieved data is required. To achieve improved error correction and detection, Mode 1, used mostly for digital data, adds a 32-bit cyclic redundancy check (CRC) code for error detection, and a third layer of Reed&*8211;Solomon error correction[5] using a Reed-Solomon Product-like Code (RSPC). Mode 1 therefore contains 288 bytes per sector for error detection and correction, leaving 2,048 bytes per sector available for data. Mode 2, which is more appropriate for image or video data (where perfect reliability may be a little bit less important), contains no additional error detection or correction bytes, having therefore 2,336 available data bytes per sector. Note that both modes, like audio CDs, still benefit from the lower layers of error correction at the frame level.[6]

[garym]

Re: Are CD drives really this inaccurate? How could our software even work?

Aha. Thanks for digging that up. That makes a lot of sense to me regarding differences between data and audio CDs.

[mville]

Re: Are CD drives really this inaccurate? How could our software even work?

All optical media data has in-built error correction. Reed-Solomon codes are also used when encoding Audio CDs. If you have the time and can understand it, try reading The Art Of Digital Audio by John Watkinson.

However, error correction does not fix all manufacturing defects.