Demagnetizing CD's !!

[Craig Sawyers]

Originally Posted by kalee20

I've often been surprised that CD players don't have a couple of LED's fitted, one to indicate that error correction has been applied (which I can imagine happens reasonably often, but won't affect the sound), and one that indicates error detection, and interpolation, has taken place (hopefully far less often).

I seem to remember reading that CD-ROMs have an error rate of <1 bit in 10^9...

The early Cambridge Audio CD1 was a four module player, one optional module of which being comprehensive error read out

https://www.hifinews.com/content/cam...io-cd1-vintage

Craig

[jamesperrett]

The issue with these CD tweaks is that sometimes they do make a difference. In an ideal world they shouldn't, but I've found that certain CD players sound different when playing different discs containing identical material. I always surmised that this was down to imperfect isolation between the digital and analogue power supplies or a jitter sensitive DAC.

The biggest difference was between a pressed CD and a CD-R copy of the same disc which I confirmed contained the same data. The CD-R sound had a slight fuzziness to it on the fairly cheap Philips player that I was using. At the same time I was also seeing if burning at different laser powers had much of an effect (using Plextor's Varirec system) but the results from this experiment were less clear cut.

[Radio Wrangler]

That's the usual argument in favour of separate CD player and DAC. But how separate is separate?

It's also an argument for ripping a CD to a lossless file with all errors fixed be re-reads if necessary and only on completion starting it playing.

David

[G.Castle]

Of course the CD fixes work! If you buy into them, then the seller has some of your money.

If, the sound quality is so bad that it's noticeable by human ears then it is a fault or poor design, neither of which can be rectified by mucking about with audiophool ephemera in the home situation by Joe public.

Radio Wrangler's suggestion is an engineering solution to a percieved problem, so, why not just rip all music to a memory in a microprocessor based system, and then just worry about the bit rate and transparency of the A/D circuitry?

Oh, wait...

There's a hole in my bucket

[agardiner]

I tried demagnetising my CDs as it promised to increase the audio separation. I also trimmed off the edge, bought gold audio cables, gold mains cable and purified the electricity supply. The audio became so separated that it fell out of the speakers flooding the house.

[kalee20]

That sounds good and successful!

I'd be worried that after demagnetising CD's, the speakers or mains cables would have to be purged - involving a send-away pay-through-the-nose-for service, or hire of a pay-throgh-the-nose-for purgificator, to flush out all those magnetised electrons!

[paulsherwin]

No, you just call them and they do it over the phone. Just give them your card details.

[Pfraser]

My 2d worth: Many years ago, I watched an Open University programme which explained CD technology. An engineer from Philips demonstrated their error-correction system, using a disc which had been modified so as to produce errors(!) Four thin lengths of black tape were affixed radially at equal distances.

The machine he used featured switchable error correction. Without it, the result was unlistenable, rapidly-chopping noise. With it: perfection! He did concede that the corrected output was not identical to the uncorrupted original, but he stated that only listeners with very close knowledge of the music would notice anything amiss.

So...are there any credible real world scenarios in which an off-centre or otherwise compromised CD might benefit from fettling, if it is to be played in a machine which lacks the error correction power of the best products?

In my experience, CDs either play or they don't. One point which is perhaps missed in these entertaining and intriguing investigations is the clear difference in sound quality one may discover when comparing a first generation Sony machine with a later, lighter, slimmer and ostensibly more modern unit - in my experience, an Aiwa. Granted, the machines were being used in a broadcast studio and not in a mass market hifi system, but still...

[paulsherwin]

An incorrectly pressed or damaged CD will depend on the player design and error correction mechanism to produce acceptable sound. This is a fault condition though.

Some CD players sound better than others, but this is (normally) nothing to do with the way that the CD data is being read. Some DAC designs are better than others, and the analogue stages are as variable as any analogue preamp.

The point about this mad audiophool CD stuff is it's supposed to affect the digital data stream somehow, which it demonstrably doesn't.

[emeritus]

I still have a recording that OU programme on VHS somewhere. I haven't played it for years, but my recollection is that they actally cut four radial slots in the CD.

[paulsherwin]

Philips did make a big show of this once CDs launched, though it's hardly relevant to how CDs are actually played. The first generation CDM0 transports were indeed very good at reaquiring the digital track after an interruption.

[kalee20]

There's a difference between error correction (where the data bitstream is 'made good' thanks to redundancy in the encoding) and 'patching' where an error has been detected, but it's not possible to know what the correct dataset should be - so 'best guess' algorithms are used to make it as good as possible.

The former will be completely inaudible, but not the latter.

[Radio Wrangler]

The true audiophile believes that even the correction of a perfectly correctable error is a problem... theorising that the correction of that bit has to create a surge in power consumption on the supply rails in the CD machine and that this will come through on the outputs.

They somehow fail to understand that all bits whether errored or not go through the same process every time so there is no error corrector which gets swung into action out of its parked state. Every bit creates similar current impulse demands on the supplies (CMOS logic has no static power consumption, but grabs a slug of charge from the positive rail to charge a node from '0' to '1' and dumps the same charge into the ground rail when it goes from 1 to 0) whether it's good or bad.

Silly to the point of laughable. The digital stuff is very very frequently taking current surges and supplies must be adequately separated from the analogue stuff on the tail end.

Back to demagnetising CDs.... If you have a disc which benefits from demagnetisation, doesn't that mean that your CD machine needs demagnetising and any other discs played on it since it was last demagnetised? Have you been playing any discs from untrustworthy sources which could have magnetised your machine?

David

De-magnetised? Ha! I have them exorcised as well!

[jjl]

Originally Posted by kalee20

There's a difference between error correction (where the data bitstream is 'made good' thanks to redundancy in the encoding) and 'patching' where an error has been detected, but it's not possible to know what the correct dataset should be - so 'best guess' algorithms are used to make it as good as possible.

The former will be completely inaudible, but not the latter.

Yes, the latter is an uncorrectable error where interpolation, muting or whatever will be employed. How often this happens depends on the physical condition of the CD, but a paper on the original CD system specifies the interpolation rate for a bit error rate of 1e-4 as 1 sample every 10 hours. There are also uncorrectable and undetectable errors where a spurious sample value will be output; the paper specifies that these occur at a rate of 1 sample every 750 hours at bit error rate of 1e-3.

John

[agardiner]

I just tried demagnetising my MP3 collection...

[Craig Sawyers]

Alternatively, attenuate possible RF fields

https://img.freepik.com/premium-phot...583.jpg?w=2000

Craig

[GMB]

I think people forget that reading CD data by a computer, at its leasure, is quite a different problem to reading it in real time that a CD audio player must do.

I am quite sure sometimes there is a significant error rate and I too am disappointed not to see warning lights on CD players to indicate that error patching is being done...

Many decades ago I was part of a team that developed a mainframe disc filing system that replaced the manufacturer's OS. We had a suspiscion that disc drives made a lot more errors than they admitted to by flaging CRC check errors so we added quite a lot of software error detection information in every sector in spite of the loss of disc space. On day one the error rate flagged was so high that we had to back off to the old system, but analysis revealed that we had been right and that our software detected errors were all true! It took the disc engineers weeks to make the disc drives work well enough for our new system.

I suspect that if we put our old system on a modern PC we may well see the same high error rate and likewise from CDs. The surprise was not just how many errors were not spotted by hardware error checks but also how little obvious damage it seemed to cause to the running system.

[stevehertz]

I recollect reading that when cleaning CDs it is best to use a radial motion. The reason being that radial imperfections ie scratches are easier dealt with by the machine than ones that are 'around' the disc. This suggests that Philips' (or anyone's) demos to illustrate a machine's error correction capabilities using radial scratches or strips of tape are choosing the least onerous type of damage for the test.

[paulsherwin]

It's more that radial scratches confuse the tracking less, so the laser is less likely to lose the data track.

[stevehertz]

Originally Posted by paulsherwin

It's more that radial scratches confuse the tracking less, so the laser is less likely to lose the data track.

Yes, that's what I read.