Now you're not going to like this. No, not at all.
CDs go off with time (please refer to my previous post CD - seedy and Completely Dreadful)
I brought a Revox A77 back from Blighty (£120 on EBay), cleaned the heads and listened to a tape copy I'd made of a Nimbus CD (Vaughan Wiliams, NIM5019) recorded in June 1984.
I bought and copied the CD the same year. The CD exhibits all the appalling colouration I've being binding on about while the tape (apart from a bit of print through, not surprisingly) sounds fine. A bit tizzy but acceptable after a glass or two.
I suggest you all start looking for Revox's and record your precious CDs before melt down.
Don't buy shares in the 'transfer your old tapes/LPs to CD.com'
and please don't try to tell me that a tape recording acts as a low pass filter-it doesn't.
CDs go off with time (please refer to my previous post CD - seedy and Completely Dreadful)
I brought a Revox A77 back from Blighty (£120 on EBay), cleaned the heads and listened to a tape copy I'd made of a Nimbus CD (Vaughan Wiliams, NIM5019) recorded in June 1984.
I bought and copied the CD the same year. The CD exhibits all the appalling colouration I've being binding on about while the tape (apart from a bit of print through, not surprisingly) sounds fine. A bit tizzy but acceptable after a glass or two.
I suggest you all start looking for Revox's and record your precious CDs before melt down.
Don't buy shares in the 'transfer your old tapes/LPs to CD.com'
and please don't try to tell me that a tape recording acts as a low pass filter-it doesn't.
Must be either the cheese or the wine....
I have directly compared several cd's (some remastered) with their lp counterpart.
Some of the original cd's are dreadful - cut off below 100Hz and 10000Hz. Most re mastered cd's can hold their own sound wise against a lp, but have the benefit of 0 surface noise.
I'm not aware of any cd colouration beside what I mentioned above.
audio-kraut said:
I'm really sorry that audio kraut (you'd get prosecuted for having a name like that in the UK BTW) can't hear any CD colouration. What sort of programme material do you listen to ?
How old were the CDs you were listening to?
As I understand it (and it is, rarely, possible for me to be wrong) there is no error correction/checksum on CDs so why should we assume the medium will maintain 100% integrity with time.
Bear in mind that, certainly in the early days, the technique for pressing and coating CDs was at the limit.
Anyone who has tried using CDROM as a computer backup material will have (or is going to have) a disasterous data loss one day - OK a writeable CD is technically different from a pressed audio CD but software on commercially pressed CDROMs also goes unreadable.
Elsewhere on this forum the dire effects of 'digital clipping' have been discussed and would seem to be more aurally unpleasant than their analogue equivalent. I can't see why random scrambling of the bits on a CD shouldn't have a similar deleterious effect.
Incidentally the tape deck I used for the comparison shows no detectable difference switching between input and recorded sound. The tape was originally recorded on an A77, not the same one of course.
There has been report of so called CD bronzing, a physical deterioration of the media. So far, I think this has only been reported to happen with CD pressed at one, or perhaps a few, particular plants, one of them in England, during certain years when these plants used a different method and/or material composition than the others. There has been a lot of discussion about the lifespan of CDs. While CDRs have been claimed and estimated to last at least 50 years for some silverlayered models and up to 200 years for Kodad goldlayered, the correpsonding estimate for pressed CDs, which use an aluminium layer, was somewhere between 10 and 20 years. I still haven't heard of anyone getting problems with old CDs apart from scratching etc. or the aforementioned bronzin problem.
Once in a while I bring out the first two CDs I bought and which are now over 19 years old, to see if the are getting aged. Prompted by this thread, I just did so again. There is absolutely no visible deterioration, apart from very minor hints of scratching. To check the data integrity, I ripped the first and last track on one of them and the last on the other with EAC. I used the highest quality setting and EAC worked very fast, encountering no problems at all. No errors were reported and the track quality was reported as 99.9 % in one case and 100.0 % for the others. That doesn't necessarily mean an ordinary CD player will not have problems with these disks, but at least the data is there, fully readable.
Contrary to previous claim in this thread, there are two levels of error correction on audio CDs. In addition to this, the data is cleverly distributed to minimize the risk of losing consecutive samples.
However, we still don't know how long CDs last, and their life is easily cut short by accidental mishandling. It is thus not a bad idea to rip at least the most precious ones in the collection and store on data CD/DVD as backup. Store several backup copies if your paranoid, and try storing them under ideal conditions (dark, right temperature and humidity). Regularly make new backups of the backups if you are worried.
I can see no point at all to use analog tape as backup for a medium which already was digital from the start.
Once in a while I bring out the first two CDs I bought and which are now over 19 years old, to see if the are getting aged. Prompted by this thread, I just did so again. There is absolutely no visible deterioration, apart from very minor hints of scratching. To check the data integrity, I ripped the first and last track on one of them and the last on the other with EAC. I used the highest quality setting and EAC worked very fast, encountering no problems at all. No errors were reported and the track quality was reported as 99.9 % in one case and 100.0 % for the others. That doesn't necessarily mean an ordinary CD player will not have problems with these disks, but at least the data is there, fully readable.
Contrary to previous claim in this thread, there are two levels of error correction on audio CDs. In addition to this, the data is cleverly distributed to minimize the risk of losing consecutive samples.
However, we still don't know how long CDs last, and their life is easily cut short by accidental mishandling. It is thus not a bad idea to rip at least the most precious ones in the collection and store on data CD/DVD as backup. Store several backup copies if your paranoid, and try storing them under ideal conditions (dark, right temperature and humidity). Regularly make new backups of the backups if you are worried.
I can see no point at all to use analog tape as backup for a medium which already was digital from the start.
I think that's quite significant.
I had a couple of (newspaper giveaway) CDs loose in the car a while ago, duing a cold spell. Within a couple of weeks the humidity had got to them - the silvered layer had deteriorated badly, and they were completely unplayable. It seems likely that similar deterioration could easily take place on a smaller scale over time under ordinary domestic conditions.
At least there's not much which will render vinyl completely unusable
Once ripped and stored on harddisk it is very easy to store and keep your material save.
A 400GB hardisk is completely copied in an hour!
And with a failure rate of 1bit every hundreds of billion Bytes.
It must be the absolute safest way! Cause those faillures are easily repeared due to the operating systems file structure.
And if you still don't trust it. You just have to buy a mirror disk.
With the current pricing of HDspace at aprox. 0,4$ / 740MB
A save copy will still cost you less than a dollar!
You don't even have to try that with tapes!
Regards Simon
A 400GB hardisk is completely copied in an hour!
And with a failure rate of 1bit every hundreds of billion Bytes.
It must be the absolute safest way! Cause those faillures are easily repeared due to the operating systems file structure.
And if you still don't trust it. You just have to buy a mirror disk.
With the current pricing of HDspace at aprox. 0,4$ / 740MB
A save copy will still cost you less than a dollar!
You don't even have to try that with tapes!
Regards Simon
Christer said:
Since no CD medium has been around for more than 30 years we have no means of estimating their longevity other than guesswork. CDRs are notoriously unreliable. We do have (post below this) documented evidence of a 2 week lifetime.
I'm not advocating the use of tape as a CD backup medium, I originally copied the CD to tape for a friend who went to his grave without knowingly handling digital media.
Is Christer telling me there's redundant data on audio CDs for error correction? I thought the read rate was about 40KHz so how can this error check data be accomodated?
If you dump audio data to HDD how do you get it out again ?
I have a few 10 year old pressed CDs suffering from corrosion from the edge, some have eaten in as far as 1cm and also have pinholes all over the place. These are normal commercial pressings, but I do live in a hot and humid climate.
In my office I have had CDR fail after one week, but this has to be a burner/media mismatch. From magazine reviews of writers, recordable DVD is definitely going to be a reliability lottery.
I plan to move everthing to a hard drive raid array with FLAC lossless compression asap.
In my office I have had CDR fail after one week, but this has to be a burner/media mismatch. From magazine reviews of writers, recordable DVD is definitely going to be a reliability lottery.
I plan to move everthing to a hard drive raid array with FLAC lossless compression asap.
stevers said:
Of course we can never know for sure the life time of anything until we have examples that actually are as old as the predicted life time. Any forced ageing test is an estimate that may give the wrong answer, whether we test CDs, paint, paper, photographic film or whatever. Forced ageing tests have, however, shown significant differences between different brands and types of CDR, for instance. For pressed CDs, at least the more pessimistic estimates of 10 years seem not to be the case, since, as I pointed out earlier, my oldest CDs are over 19 years and still seem in perfect condition. On the other hand, that is no proof that all CDs will last 19 years, even if handled with great care.
Indeed I am. The 44.1 kHz is the sample rate per channel, which is something entirely different than the read rate from the disk. There is a lot of extra data stored just for error correction. I don't know what's the best reading on this, but here are some links for a start
home.btconnect.com/geffers/cd.html
www.cdmasteringservices.com/cderrors.htm
sirdavidguy.coolfreepages.com/CIRC.html
www.cd-info.com/CDIC/Technology/CD-R/Media/testing.html
Either though your soundcard or by burning a CDR which you play in your ordinary CD player. The latter was the option I intended when I suggested storing backup copies as data CD/DVD. The idea is to store the ripped content as either wave files, or if you want to save space, use some lossless compression like FLAC.
Some people rip their CDs and burn on CDR because they think the CDR version sounds better than the original, depending on the choice of media, of course. Feel free to experiment.
davidsrsb said:
Yes, the climate most probably is an important factor. Besides that, is there some correlation between those problem CDs? Same record company? Pressed att the same factory? AFAIK this so called bronzing problem has been heavily or even entirely correlated with certain manufacturers during certain years.
This idea of CD deterioration comes up from time to time.
First and foremost - CDs are all encoded with reasonably agressive error correction codes. The references above are an OK intro - although the second one (cdmasteringservices) is so full of errors that it is an embarassment to the company, and essentially mostly wrong.
The design of the error correction on CDs was quite carefully crafted. It is particularly good at allowing error free reproduction in the face of scratches on the surface, and holes in the substrate. Many CDs are playable with no non-correctable errors at all - and are thus totally error free. As some CD's either age, degrade, or are just plain bad, the rate of non-correctable error's may increase. But even here you need to be very careful. 100 non-correctable error's on a CD means an average of a little more than one a minute. So one sample may end up being interpolated per minute. You would not be able to hear this.
Eventually a CD may become so damaged that the error rate is impossible. Most CD players will mute the audio at this point.
So, can a CD really go bad in a manner that it sounds worse than it did. Well yes. But not due to errors in the data. Remember that poor quality CD players have very poor clock recovery, and the jitter in the clock can be affected by a great many second order effects in the player. A CD that has become physically harder to play may cause all sorts of issues. One that may be of particular importance is a poorly designed data de-interleave and error correction unit. As it works harder, or at least differently in the face of a mix of errors and error free data, it may introduce much more jitter into the output data than with a new CD. Rotten CD player designs seem to assume that the encoded clock from here is OK to run into the DAC. No surprises it could sound bad, worse, whatever.
But the culprit is really poor CD player design. It is very very unlikey that the CD is actually producing a degraded data stream - indeed if it did, and you heard it - it would not sound like a reduction in fidelity. It would sound like it was utterly broken.
First and foremost - CDs are all encoded with reasonably agressive error correction codes. The references above are an OK intro - although the second one (cdmasteringservices) is so full of errors that it is an embarassment to the company, and essentially mostly wrong.
The design of the error correction on CDs was quite carefully crafted. It is particularly good at allowing error free reproduction in the face of scratches on the surface, and holes in the substrate. Many CDs are playable with no non-correctable errors at all - and are thus totally error free. As some CD's either age, degrade, or are just plain bad, the rate of non-correctable error's may increase. But even here you need to be very careful. 100 non-correctable error's on a CD means an average of a little more than one a minute. So one sample may end up being interpolated per minute. You would not be able to hear this.
Eventually a CD may become so damaged that the error rate is impossible. Most CD players will mute the audio at this point.
So, can a CD really go bad in a manner that it sounds worse than it did. Well yes. But not due to errors in the data. Remember that poor quality CD players have very poor clock recovery, and the jitter in the clock can be affected by a great many second order effects in the player. A CD that has become physically harder to play may cause all sorts of issues. One that may be of particular importance is a poorly designed data de-interleave and error correction unit. As it works harder, or at least differently in the face of a mix of errors and error free data, it may introduce much more jitter into the output data than with a new CD. Rotten CD player designs seem to assume that the encoded clock from here is OK to run into the DAC. No surprises it could sound bad, worse, whatever.
But the culprit is really poor CD player design. It is very very unlikey that the CD is actually producing a degraded data stream - indeed if it did, and you heard it - it would not sound like a reduction in fidelity. It would sound like it was utterly broken.
Christer,
Thanks for the info, like a CD my errors (maybe) are corrected.
Francis,
I'm using a RAKK DAC so I'd expect to be getting good audio output.
There's something going on I don't pretend to understand.
Most of my old CDs have slowly developed the same mid range harshness. Sadly once heard it is easy to detect on a lot of modern CDs as well. I listen to a fair amount of live music so it isn't some non linearity developing in my ancient lug 'oles and the effect is independent of the reproduction equipment.
I know you can't all get BBC R3 by satellite but if you can (and no doubt there are equivalent other broadcast services) try listening to a mix of live and recorded music on one session. Obviously the whole tx is digitised but the difference in sound quality when changing from live to CD is remarkable - and I think we can assume they are using reasonable CD players. This isn't even a particularly good digital transmission medium but the live stuff is generally comfortable to listen to while recordings are, again generally, fatiguing.
Thanks for the info, like a CD my errors (maybe) are corrected.
Francis,
I'm using a RAKK DAC so I'd expect to be getting good audio output.
There's something going on I don't pretend to understand.
Most of my old CDs have slowly developed the same mid range harshness. Sadly once heard it is easy to detect on a lot of modern CDs as well. I listen to a fair amount of live music so it isn't some non linearity developing in my ancient lug 'oles and the effect is independent of the reproduction equipment.
I know you can't all get BBC R3 by satellite but if you can (and no doubt there are equivalent other broadcast services) try listening to a mix of live and recorded music on one session. Obviously the whole tx is digitised but the difference in sound quality when changing from live to CD is remarkable - and I think we can assume they are using reasonable CD players. This isn't even a particularly good digital transmission medium but the live stuff is generally comfortable to listen to while recordings are, again generally, fatiguing.
I listen to anything from rock to classic, including modern classical music - i.e. stockhausen, xenakis, paert etc.
So my taste ranges and is reflected in about 1500 cd's and about the same amount of lp's.
My oldest cd is "blind faith" bought in 1986, so twenty years old to the date. I experience no deterioration of sound or of the physical state of the cd.
Luckily however we live in a fairly dry climate.
I have to mention something however - when doing a close listening test with dire straits "bia" cd, I heard a noise on the cd as if someone had introduced a triangle wave in the background.
That "noise" is missing from the lp. Do not hear that in other cd's so, and only can be heard when listening at a 1m distance from the speakers.
Perhaps, far from your ears deteriorating, your ability to hear has become more refined.
As an analogy, remember the wines that you found quite acceptable in your youth (assumung that you drink wine
). How do you find them now that your palate is more experienced? Can you detect similar faults in some more expensive examples?
Most hard disks currently sold are intentionally designed for a useful life of 5 years, look at the datasheets if you don't belive that. Furthermore, their average life when used frequently is 4 years or less, particularly for low-cost low-quality stuff (like Maxtor).
Hard drives also suffer from surface defects and modern ones include sophisticated built-in error recover techniques. The latest diagnostic and error handling system is called S.M.A.R.T. It was introduced due to the constant requirement for cheaper drives and due to the fact that manufacturers were no longer capable of producing defect-free disks, nor disks capable of operating for years without thousands of bad sectors appearing, for such a low cost.
Due to the extremely low magnetic substrate quality (again low cost!!), all modern S.M.A.R.T. hard drives have already dozens or hundreds of bad sectors when they are sold brand new, but these sectors are listed in internal tables in the firmware of each drive and they are not used, they are just skipped and sector numbers are automatically adjusted as if they didn't exist.
As the original surface defects of the drive grow (quicker and quicker!!), more and more sectors become bad. S.M.A.R.T. drives overcome this by reserving a pool of free sectors (not accessible to external software, only used by drive firmware) and transparently relocating the data from unrecoverable bad sectors to fresh ones of this pool. There is a table stored in non-volatile memory (and recorded in the own drive) that tells what sectors have been relocated to where, so the process is fully transparent to the operating system and to the user. Everytime a sector becomes bad, it's just transparently replaced by a new fresh one in other place.
As time goes by, more and more sectors become bad, but you are not told about that because the drive is automatically relocating them and hiding the errors to the operating system. This happens happily until the relocation pool and the relocation table (that may have room for several thousand bad sectors) run out of space. What a smart system, eh? Then the drive can no longer "repair" itself and suddenly becomes flooded of read and write errors after a few weeks (note that the rate of appearance of new bad sectors appears to be quadratic or exponential, so as the drive ages, the amount of new bad sectors becomes impossible to handle!!).
This is the best case, since due to firmware design errors, most Maxtor drives, and some from other manufacturers like Seagate, become no longer detected by the BIOS (and thus you can't even recover part of your data!!) the next time they are powered on when the sector relocation table is full.
I know because I happen to own two of these buggy Maxtor drives: 20Gb and 40Gb. The 20Gb drive has already failed, it is no longer recognised by BIOS and unrepairable after less three and half years of usage (I've even tried specific third-party software for firmware repair). The 40Gb drive has already more than half of its relocation entries used (87 of 153) even though it has been manufactured only two years ago and has less than a year and half of usage (just as an emule drive), so it has probalby 6 to 12 months of useful life left before total failure.
So don't trust modern hard drives, nowadays CDs are far more reliable!!
P.S. In comparison, I also own two old non-S.M.A.R.T. hard drives, sized 426Mb (Conner) and 1Gb (Quantum). Both are more than 10 years old and have been used 24/7 for several years (still powered 24/7 now, but in sleep mode most of the time). While they don't have any built-in sector relocation capabilities, recent surface tests revealed that there is not a single bad sector in them!! Note that both Conner and Quantum are no longer in business, obviously they just coudln't keep up manufacturing such high quality drives at the lower and lower prices required by the market.
Hard drives also suffer from surface defects and modern ones include sophisticated built-in error recover techniques. The latest diagnostic and error handling system is called S.M.A.R.T. It was introduced due to the constant requirement for cheaper drives and due to the fact that manufacturers were no longer capable of producing defect-free disks, nor disks capable of operating for years without thousands of bad sectors appearing, for such a low cost.
Due to the extremely low magnetic substrate quality (again low cost!!), all modern S.M.A.R.T. hard drives have already dozens or hundreds of bad sectors when they are sold brand new, but these sectors are listed in internal tables in the firmware of each drive and they are not used, they are just skipped and sector numbers are automatically adjusted as if they didn't exist.
As the original surface defects of the drive grow (quicker and quicker!!), more and more sectors become bad. S.M.A.R.T. drives overcome this by reserving a pool of free sectors (not accessible to external software, only used by drive firmware) and transparently relocating the data from unrecoverable bad sectors to fresh ones of this pool. There is a table stored in non-volatile memory (and recorded in the own drive) that tells what sectors have been relocated to where, so the process is fully transparent to the operating system and to the user. Everytime a sector becomes bad, it's just transparently replaced by a new fresh one in other place.
As time goes by, more and more sectors become bad, but you are not told about that because the drive is automatically relocating them and hiding the errors to the operating system. This happens happily until the relocation pool and the relocation table (that may have room for several thousand bad sectors) run out of space. What a smart system, eh? Then the drive can no longer "repair" itself and suddenly becomes flooded of read and write errors after a few weeks (note that the rate of appearance of new bad sectors appears to be quadratic or exponential, so as the drive ages, the amount of new bad sectors becomes impossible to handle!!).
This is the best case, since due to firmware design errors, most Maxtor drives, and some from other manufacturers like Seagate, become no longer detected by the BIOS (and thus you can't even recover part of your data!!) the next time they are powered on when the sector relocation table is full.
I know because I happen to own two of these buggy Maxtor drives: 20Gb and 40Gb. The 20Gb drive has already failed, it is no longer recognised by BIOS and unrepairable after less three and half years of usage (I've even tried specific third-party software for firmware repair). The 40Gb drive has already more than half of its relocation entries used (87 of 153) even though it has been manufactured only two years ago and has less than a year and half of usage (just as an emule drive), so it has probalby 6 to 12 months of useful life left before total failure.
So don't trust modern hard drives, nowadays CDs are far more reliable!!
P.S. In comparison, I also own two old non-S.M.A.R.T. hard drives, sized 426Mb (Conner) and 1Gb (Quantum). Both are more than 10 years old and have been used 24/7 for several years (still powered 24/7 now, but in sleep mode most of the time). While they don't have any built-in sector relocation capabilities, recent surface tests revealed that there is not a single bad sector in them!! Note that both Conner and Quantum are no longer in business, obviously they just coudln't keep up manufacturing such high quality drives at the lower and lower prices required by the market.
Ha! CICO (Crap in, Crap out)audio-kraut said:
dnsey said:
Perhaps, far from your ears deteriorating, your ability to hear has become more refined.
As an analogy, remember the wines that you found quite acceptable in your youth (assumung that you drink wine
Possibly, although I'm now 62 and (having had a fairly interesting life) have quite severe tinnitus.
I record our village band (Oh God, an old deaf fart) and listen to the 7 1/2 tape and the real thing and.... there ain't much difference.
When I was 16 Arthur Radford took me to a pub in the Mendips and we recorded Acker Bilk + the Paramounts on an old 15" ps recorder and replayed it on Arf's STA 25.
Then we went to Bristol Cathedral and recorded the choiristers.
I've never heard anything better.
You guys have lost your way. Take a CD with all its complexity, stuff it onto HDD via God only knows what software, try to restore it (again via GOK what) then do a DA conversion (rules, levels, original AD conversion parameters) then listen to it.
The result is rubbish. Complete, absolute, sad rubbish.
I'll gladly host a RubbishFest here (Limousin, Central France) and you anoraks can bring you're HDs, Cds, DVDA's, and my old Revox will **** all over them.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.