These discs were of the Ritek brand. I agree with your comment regarding Taiyo Yuden, I have been using them exclusively since 2008.

Chad

Ritek is a good brand just like Taiyo Yuden. Statistically, sometimes you just wind up with a "bad egg".

Peter

"Do bad sectors happen at the time of burning"

I've read that there are correctable errors on disks when they are burned. If the blocks are readable with correction, it's considered a "good disk". I've also run the Plextor utilities that counts the errors on burned disks, but I have so little life left I'm looking for better things to do with my time. Note the term "looking".

"...or is it possible as the disc ages that they can be created?"

I've experienced a lot of disk failures with disks burned in that timeframe (2005 and earlier). At the time I was using Maxell media. Never again. I had good luck with Maxell tape and CD blanks so I mistakenly thought that would translate to DVD-R media.

In deference to Steve Mann, none of this discussion is valid without listing the actual data about the company that actually manufactured the disk. Brands are meaningless, anymore.

Two weeks ago, I created two Blu-ray disks with the contents of 11 DVD-Rs that I burned in 2005 and all except for one copied without incident . Luckily, I had two copies of the failing disk. Now, if I really want to keep the material, I save it to two hard drives.

All, yes, everyone of them, disks are burned with bad sectors. This is the nature of the process, it does not create a perfect disk, but rather a passable disk. Sectors do not age and create bad sectors.

What is the issue, is that IF you have a lot of marginal sectors, then the playback is going to vary considerably. It may play well enough in the original burner, but not play well at all elsewhere.

That is why, when burning discs, your quality control checks are always done outside of the original burner.

In the past I have used KProbe to check the "soft" errors of discs I have burnt, but it is not compatible with SATA drives and it doesn't test BDs.

Fortunately Nero DiscSpeed is also available for free and seems to do everything. Note that in order to do the "disc quality" check, you need a compatible optical burning drive as made by LiteON. (Pioneer and LG are not suitable). I am currently using a LiteON iHAP322 DVD burner on one machine and a LiteON iHBS312 BD burner on the other.

After burning a disc I check the soft errors and if they seem a little high, I discard that disc and burn another. Soft errors are corrected using error correcting code, and I have never seen a disc with zero soft errors. It is only when the error correcting code is unable to correct the errors that you get a hard error that will cause problems.

I have noticed that the soft errors can increase with time, but I have not made any systematic study.

One of the worst brands I have tried is Ritek, although recent CDs I burnt for a friend seemed no worse than the CMC CDs that I usually use. I use Verbatim DVDs because they seem better than Ritek, Sony or CMC. I have only tried Verbatim BDs. Note that I am talking about the brand embedded in the disc, not what is shown on the box. (Some years ago I had some Verbatim DVD-Rs made by CMC.)

If you are seriously concerned about the quality of burns and want some way of montitoring degradation before it is too late, I strongly recommend that you start using DiscSpeed.

For myself I don't trust DVD or CD discs anymore. I have some music CD's that are yeeeears old and some of them skip and jump about. I've now ripped all my music to high bitrate MP3 files on a hard drive which I have backed up.

Most people go with DVD because of tradition not realizing that technology has advanced to high definition and DVD's just don't cut it anymore. I see the trend is going more and more to media players capable of playing high definition media. Do you still use your VHS to back up your movies?

Sure putting all your eggs in 1 basket is risky but at the same time you can backup your data easily. Since hard drives don't age with time but rather in hours of flying time. You can keep a healthy backup hard drive for many years even after the guarantee has expired provided you remove the backup dive from you PC to keep flying hours low. When IDE drives were being replaced with SATA ones it was so easy to transfer the media.

Acronis (free) is a wonderful tool that checks the health of hard drives by reading the SMART chip then displays health in % and notifying you of imminent failure.

I still give my clients DVD's but also copy the movies in HD as data to DVD's but warn them it'll only last 10 years.

For what its worth.
I have a CD given to me in 1995 which had a lot of freeware and (other) utilities for Win 3.1 and 95. XTree Gold is there, remember it?
It is a MaxellCD R74H disk and has always been in its jewel case with really opaque labels. It wasnt archived as for years it was in my box of bits and pieces and often used to install on clients systems.
I just slipped it into the DVD burner on my win 7 64 bit machine and while it recognizes that something is in the drive it wont read a thing. Reaching for an old Cd burner from the growing pile of suspect drives i cant bring myself to throw away and powering it via a USB adapter I can however read all that is on it with I suspect 100% performance. Win 7 wasnt very happy about some of the files and wanted to go looking for converters etc. Most graphic files were quite readable and some photos were quite good quality for the era.
PDFs were as per original in Adobe reader.
This means that if you have old disks which mean something it will pay to keep an older player handy.
Incidentally I have just about got all my analogue video 8 on to digital but have had to buy four second hand cameras so far.
Brian

Whatever happened to the new "forever" media that was supposed to be out in 2012? I think it was called Millenia or something like that.

M-Disc is available. I have a drive from LG that burns it, but I've had such poor results with so many other media types I haven't felt the urge to try one.

Why errors in the first place? What is the cause?
JJK

Errors occur during burn for a number of physical and sometimes logical reasons. Much more technical jargon understanding would be needed to go about explaining it, suffice it to say, it is a fact of burning.

What several recent posts above affirm, is that the disc burning process is part and parcel of the burner it is created on along with the media choice. Over the years, the burner development has somewhat laxed into less than severely tight standards. Hence the reason for always making a quality control step part of the process, that elaborated on above posts, and frequently checking and maybe even doing a "re-burn" to a second disk on later media and equipment in following years.

Just bit off subject, but what is the information about the storage life of SD or other memory cards?

Dave T2

People here have already hit on the key point: use quality media, as there IS a big difference.

I used Ritek DVD-R blanks for years until I saw data somewhere that showed degradation, then I switched to TY and Verbatim. These two brands are almost universally considered to be the top of the food chain..

And, yes, disks can age and gradually get worse. I just recently threw away the data I had collected over the years for a Ritek DVD-R I had burned back in 2003. Right after burning, I did a Nero scan and printed out the graph. Pretty clean, some errors towards the outer (4GB+) reaches, but still easily correctable and with a pretty high score of 96.

Then I did an identical scan and print of that disk on the same drive (NEC 3550) six months later. Significantly higher correctable error rate, again on the outer portion of the disk, gradually rising from about 2GB outwards. Still a decent score in the 80's.

Then, last year, another scan just for the heck of it, again on the same drive. Utter crap beyond about 2GB, uncorrectable errors, score of zero.

So, yes, something in the disk changed over those eight years. I should note that my earliest TY disks show minimal change after about 6 years.

DVDs ALWAYS show some read errors, that's the nature of the beast. But with several layers of robust error correction, the disks can be trusted on a bit-for-bit level if quality media is used. I have C drive backups using Acronis that were copied to DVD-R years ago; they are bit-for-bit identical today with the hard drive copy they came from (they produce the same MD5 code).

EDIT: After writing the above I looked on my system to find the oldest large file in one of my hard drives. It was an Acronis-generated 4.33GB file, part of a larger C image dated August 2008 (4 1/2 years ago).

I went out to the building housing my data-DVD backups and located the disk that had been burned with that file at that time, using Nero. Sure enough, using "MD5check" on both files, same MD5 hash: 49C6E35F9AFDC2A2178E1C9F85F06627. The disk was a Taiyo Yuden, burned at 12x speed.

If you scan a disc just recently burnt, chances are that it will be pristine, but a disc from some years ago can contain dust and fingerprints (not mine of course!). Make sure that you clean old discs thoroughly before scanning them. I blow on them while brushing outwards with a lens brush.

>> People here have already hit on the key point: use quality media, as there IS a big difference

This is very true, also, use newer technologies. CD-Rs are the worst (on average, there are good ones) and it is not hard to see why. Get one, use something sharp to scrape at the top of the disk. You will see something similar to thin tin-foil coming off and the disk it self is clear plastic. On CD-Rs the data layer is actually on the outside of the disk (on most).

DVD is an improvement. Both better error correction and better material.

Blu-Ray is a step up again with significantly improved both error correction and material used to create the disk. If you want to put anything on optical disk for storage, get good quality Blu-Ray disks, they will most likely be better than DVD and both are better than CD-Rs.

Can't vouch for the robustness of BluRay, since I've never worked with them. But for CDs and DVDs, you are correct--the data layer on a CD is on the very top of the disk, easily damaged by a ball-point pen or just plain wear and tear. In a DVD, the data layer(s) are in the very middle of the disk sandwich. Also, because of the much smaller track pitch and pit sizes, DVDs use a much-more-robust error-correction protocol.

Terje's comments in the post just before this one, suggests some misunderstandings and mis-representations of technologies involved.

1) CDRom discs were and are the best build for data streams integrity - suffice it to say there are many reasons why this is true.
What is true is about CD-ROM media is that the older, and particularly gold surface, are really the best of all, burned at 2x speeds. Later versions of media have been cheapened and today's discs can very quite a bit. And CDrom media do not include the protective surface quality of a Blu-ray disc.

2) DVD media of all types is the worst of all optical media for data integrity, again for a number of not so uncomplicated reasons. However it has included one standard that CDrom was not successful with - that being a video stream that could be played with reasonable quality on less expensive set-top players.

3) Blu-ray disc tech does include a different form of error balancing so that burned errors play less role in data integrity - but in real world is still not as good as the integrity of CDROM on archival media. Blu-ray did introduce protective surfacing of the exposed disc so that dust and fingerprints can better be eliminated. Even so that protection layer is very thin, and is not a good practice to push it too far with repeated cleaning. Blu-ray does have the best media balance model with being able to record BD-ROM data, many different kinds of video-streaming for Blu-ray set-top players, and can in fact hold a DVD image as a container for "perhaps" archive storage use.

Dredged this thread up again because I stumbled upon a study by NIST, the heavyweight standards body in the USA. Back about seven years ago they began a serious investigation into CD and DVD data integrity over time. The results are here.

Pretty dry reading, but here is a relevant portion of that report. NIST says they used 15 different media (5 different DVD-R brands) and found quality varied significantly among brands, with some unusable between 0-15 years and others good for >45 years.

Since I have to assume NIST bought a cross-section of brands, I would place Verbatim and/or TY at the top, since it's been known for years that these were "premium" brands. Also, I think they found the ±RW disks to be more tender, so they probably made up part of the lower range. Unfortunately, though NIST kept meticulous records, they have refused to release the brands performing the best!

Still, this reinforces my impression that a HIGH QUALITY DVD-R should last a very long time. If there is a different interpretation, I'd be interested in hearing it.

Comparing a broken data file with a broken movie file, you could recover a broken movie with lost frames whereas a broken data file is totally useless. DVD's with a 95% confidence level and Murphy on the loose...........?

Think of the old cine movies in a box in the garage. Films gone brittle. If I remember you had to put them in a area with vinegar to soften the film. Anyone ever try this?

Craig

I was under the impression that the NIST study was examining the use of CDs and DVDs to store data, not video. In other words, they were looking for unrecoverable bit errors, not blocks on the screen or the playability of movies. I can't do it right now but I will try to review the study some time to see if my assumption is wrong.

If a data file is damaged, it doesn't mean the whole file is junk. There are ways of recovering the good portions, and depending on the nature of the file the loss might be minor.

The 95% criterion was the reference. I guess they could have made it 99% or any other figure. The point is to have a reference that one can then compare to other storage options. Hard drives? Out of maybe 15 purchases, I've had 3 failures over the past 15 years or so, though two of those were taken out by a rogue power supply (but so what? Hard drives need power supplies, after all). Solid-state storage can apparently show errors over time, though I have no experience with it. Tape comes with its own issues, as does film. Never heard about the vinegar trick.

The laser-read optical mediium really is remarkable--I mean, think back to when CDs or DVDs first were mentioned. You probably smiled to yourself and thought, "What a great idea! Just a bunch of dots inside a rugged plastic layer! Nothing has to touch it! It doesn't care about minor scratches or fingerprints! Amazing!

This is an extremely important topic: if you create something, but then it disappears after a few years, then you might as well become a professional sand castle builder.

While most of the posts above are accurate, there are several common misconceptions which are perpetuated which, if acted upon, could cause someone to do something that would result in eventual loss of their work or, at the very least, cause them to do all sorts of unnecessary and counterproductive activities.

Digital Data Integrity

All digital storage devices lose bits. This is true of RAM, flash memory, CDs, DVDs, hard drives, SSDs, etc. One of the first things you learn in computer science is the concept of error detection, and its companion, error correction.

Media doesn't just have errors when it is starting to go bad: it has errors when it is brand new. This is normal and is unavoidable. In order to make the media usable, the designers build in a huge amount of redundancy that allows the underlying firmware in the device to not only detect when there is an error, but then to automatically correct that error. To do this requires a large amount of additional data. Your brand new 3TB disk drive could probably store 4TB or more if all error correction data redundancy were removed.

Good vs. Bad

With CDs and DVDs, it is actually quite easy to test whether you have a large number of errors, or a small number of errors. Why would you do this? Because while the redundant data and error correction can completely and totally correct for a small number of errors, at some point, if the number of errors per megabyte becomes too large, the error correction fails, and then you DO lose data. This is true whether the bits represent data files, or whether they represent video.

You don't want this to happen.

There are several things that affect the number of errors on a freshly-burned DVD: the media; the device used to burn the disc; the software used for burning; and the settings used in that software. However, here's the key thing to know:

Unless you really do something wildly unconventional, the only thing that matters is the media.

I could scream every time I see a post where someone claims that burning at a slower speed will make better discs. This is not true (unless you burn at a speed faster than the rated speed of the discs or the burner). Some people claim to get better burns with Nero, or ImgBurn, or DVDA, or Encore, etc. I can't make quite a sweeping a statement about this, but from everything I have read over the past twenty years (going back to my first CD burning experiences), the issue with these programs is usually black and white: either they work with your burner or they don't.

Quality Tests

The main point of this thread is that DVD and CD discs appear to be going bad as they age.

This is true: they do degrade.

The excellent accelerated aging tests that were linked to in the previous threads clearly show that some media ages more rapidly than others. Unfortunately, that test did not reveal what media was being tested, and because the media is not revealed, it is quite possible that the really good media (see below) were not included in that test. However, there ARE sites where extensive tests on almost all media have been done, and that information is freely and readily available:

MyCE Blank Media Tests

DigitalFAQ

At the first site linked to above, thousands of users have measured and then shared their experience with various brands of media, with this effort spanning more than a dozen years. These tests accurately show the number of errors on new media, freshly burned. What clearly emerges from these tests is that most media is really bad, and only a few brands are good. Even within those brands, there are good products and bad products. You need to purchase the exact brand specified, and then be very careful you do not get a fake. You can usually do this by purchasing from a reliable supplier.

Aging

There are many poorly-research anecdotal stories about discs failing after short periods of times. This thread contains many examples of this. What is usually missing are the two most important things that contribute to this outcome: how good was the disc when it was first burned, and how was it stored?

Using DVD Speed, I have done these tests many times over the past decade. Here is my finding: I have found my good media is showing virtually no decline over 6-10 years. The three good brands I have used over time are: Maxell MXL-RG01 (2X media), Ritek G04, and Taiyo-Yuden TY02. Premium The Taiyo-Yuden is now branded by TDK [edit: TDK is wrong; they are branded as JVC] and is the brand that pretty much everyone recommends for single-layer DVDs.

JVC (Taiyo-Yuden) DVD-R

I use the DVD-R version because that's what I started back when DVD-R had slightly better compatibility with 1st generation DVD players. I think most people now use the DVD+R. I too use DVD+R when I burn dual layer, and I use the only media that is ever recommended for that, the Verbatim dual layer discs:

Verbatim (95123) DVD+R Dual Layer (DL) 2.4X

Prior to making this post, I just did a test on half a dozen discs that I have burned over the last decade. None of the burns done on any of the discs mentioned above show any measurable decline. By contrast, my experiments with other discs, including some real garbage, show massive error rates. However, these may have been there from the day they were burned (I don't test every single disc, because I have burned many thousands of them). I cannot say for certain that the error rates are caused by aging.

As for storage, the discs must be kept in total darkness, and should be kept in a stable-temperature environment. Lower humidity is better than higher humidity. If you don't do both these things, you will see problems.

As an acrhivist and as an engineer, I am totally confident that if you use high-quality DVD and CD blanks, that they should last at least as long as well-stored film. I have transferred still-photo negatives from the late 1890s and have transferred B&W 16mm movie film from the late 1920s, and both were in almost as good condition as the day they were developed. I have also transferred similar still and movie film that was only a few decades old, and it was in terrible shape. Some of the non-Kodak color emulsions have completely faded. Some badly-stored film has developed the "vinegar syndrome" and can no longer be viewed or transferred.

Storage conditions matter.

Based on my tests, and based on everything I've read, I expect that properly-stored DVD and CD discs will survive at least as long as film.

Finally, despite my assurances above, I think it would still be wise for everyone to get into the habit every few years of taking a few sample discs from their inventory of burned discs and doing the simple DVD Speed test which measures precisely the number of correctable (and uncorrectable) errors on your disc. This program used to come free with Nero. If you find any disc that appears to have a higher-than-normal error rate, make a copy onto good media, and you should then be good to go.

Oh yes, the MAM-A media. I have burned a few of these, and they have much higher initial error rates because their reflectivity is lower. I was put off by this, and therefore haven't done much with them since my initial tests.

An excellent summary from johnmeyer and compulsory reading.
(But I thought that after the purchase of JVC media division Taiyo Yuden uses the JVC brand?)