A common use case for optical discs, at least when they were common, is for backup and archival. Recordable discs are (sort of) WORM media that could provide the ability to safely store vast amounts of data inexpensively. While these are mostly out-of-fashion nowadays, I’ll be spending quite a bit of time examining this technology at the trailing edge, in part thanks to the valuable donations by pepst.

While many that are using optical discs for archival are generally focused on Milleniata/M-Disc DVD or perhaps the slightly less proven M-Disc BluRay/BDXL solutions, or the formerly available DataTresorDisc, there was a market for archival media before the arrival of such inorganic media that mostly focused on having a gold reflective layer and resilient, albeit still organic, dyes.

Today’s post mainly focuses on documenting the design features, operation, burning and read-back of samples of such discs – the actual archival life is not considered (sorry to those that will find this disappointing – but I don’t have an environmental accelerated life testing chamber at my disposal).

Mitsui Gold 74 Minute 650MB CD-R (GD74)

When I started my CD recording journey in 1997, we had relatively limited choices of media here in Australia. There were relatively standard CMC Magnetics discs under a variety of brand names, but if you were looking for quality and archival, most of the retailers here stocked Kodak’s Gold Ultima at a reasonable price. The disc itself was quite golden on both sides, having phthalocyanine dye rather than the aqua-coloured Formazan hybrid that were previously found on Kodak discs.

But there was a slightly more premium option that only a few local computer shops stocked – the Mitsui Gold discs from Japan. These had a legendary reliability reputation, at least compared to the cyanine dyes of the time and had I known at the time, I might have made some deliberate effort to purchase some. It all made sense, as Mitsui Toatsu Chemical were the inventors of the dye and process.

After years of regret in not experiencing such discs, thanks to a donation from pepst, I’ve managed to get two samples of just the bare disc. The outer packaging would likely have looked like this (unrelated completed auction listings).

The disc logo top is translucent gold background with black text consisting of a labelling area. Production codes are in three groups of four digits, printed in ink dots in the clear clamping area. The discs were often printed with “Made in Japan” on the top, so I wonder if this was a later production by European plants instead. Nevertheless, I was surprised to see that the underside was less golden than I had expected. The Kodak Gold Ultima discs are really gold – the reflective layer only looks barely more golden than an ordinary one.

These photos from my DSLR are perhaps a bit more true-to-life in terms of colour – this sample is already burned, but looking in the stamping code area, the reflective layer looks mostly silver to me …

A closer look at the code in the stamping area has the text “MG07C0-026186146”. I presume MG is Mitsui Gold. The ink dot numbers are “1057 2531 3358”.

From ImgBurn -
Free Sectors: 333,233
Free Space: 682,461,184 bytes
Free Time: 74:05:08 (MM:SS:FF)
MID: 97m27s55f (Mitsui Chemicals Inc.)
Last Possible Start Time of LeadOut: 74m05s10f

The ATIP data from different tools is shown above – not much generosity on that 74-minute rating, providing just 5 seconds and 10 frames more.

The Overburning Test dialog in Nero CD-DVD Speed gives a Dye type of “5”.

Running multiple tests gave some inconsistent results – 76m31s60f is the shortest from three runs, but I’d say that it’s perhaps not a guarantee.

As I didn’t want to write both of my samples – I selected one and let my BenQ DW1640 write a useful image file that was not overburned but close to the full capacity. This was done at 12x CLV, the slowest supported speed (the only other being 16x). This would be relatively contemporary with the disc as most drives of the time were likely in that 4x-16x era. Going too slowly is not necessarily good, but neither is going too quickly.

For a disc that’s probably about 28 years old, it’s quite amazing that the burn reads back smoothly and error rates are relatively well controlled, well below the 220 blocks/second error rate limit. Jitter is quite acceptable for a CD burn with no C2 errors on the DW1640.

The DH16A6L also gives a smooth readback and shows a lower error rate (it usually does) but slightly increasing with diameter. I suppose this is a data point that tells us that well-stored old optical media may be perfectly usable … assuming you have a writer that can do a good job with it – the DW1640 is one of my outstanding exceptions.

“Designed for Photo Applications”

If you had asked me at the beginning of last year what my favourite CD-R was, I would have answered “those black coloured substrate discs” which were very PS1-esque. But thanks to pepst, I now have a new favourite.

I don’t know the origins of this disc in terms of how they were sold – presumably they may have been bulk packaged for sale to photo processing shops to use to burn scanned film JPEG files to. Regardless, it’s very visually striking – the central portion being a “regular” CD-R but the outer portion text and imagery on a mirror smooth finish. Of course, that area cannot be used to carry data, but it is still covered in a phthalocyanine dye that’s only pale yellow.

In real life, it looks a bit like this to my DSLR.

When written, the dye darkens, but only just a bit. *So perhaps it’s worth asking, what makes this disc “designed for photo applications”? *The choice of phthalocyanine dye is certainly a good one for longevity, but the reflective layer appears to be a normal “silver-looking” layer, so that’s not ideal.

---

Putting a picture of a DSLR camera on it? Perhaps not …. at least, functionally.

Using my scanned image for measurement, I can find that the data area of the disc occupies 7.81cm diameter across the disc. This means that it’s basically the equivalent of an 8cm CD-R but with a “halo” of 2cm either side of “dead” space.

**But I think this is where the genius is. **Initially, you might think that using an 8cm CD-R would just be equivalent, but this dead edge is actually a big help. From a protection standpoint, mishaps with drive loading, edge impacts and scratches usually affect the outer edge more than the inner edge. Light exposure from the side of jewel cases in storage also tends to fade the dye from the outer to inner. By ensuring no data can be burned in this vulnerable position, it’s basically setting up the user for success – but you could do the same with just an ordinary CD-R just by not filling it up.

While it might seem trivial, I still think this is both unique and despite appearing to be a gimmick at first, might actually work!

The stamping area has “SCSI – 23 – PHOTO01” and “F4495” written in it.

There is an inked code of “V2220004” in the clamping area and we can see the sharp edge between the data area and the smooth silvered area where there is no groove.

From ImgBurn -
Free Sectors: 103,348
Free Space: 211,656,704 bytes
Free Time: 22:59:73 (MM:SS:FF)
MID: 97m24s26f (Unknown)
Last Possible Start Time of LeadOut: 23m00s00f

Retrieving the ATIP using various tools reveals that the disc has a 23-minute capacity (or about 201MiB) capacity. The manufacturer is identified as Computer Support Italcard (or Italy). This is significant, as this manufacturer is who eventually took over Mitsui’s assets after they exited the optical media business in 2003, forming MAM-E (Mitsui Advanced Media – Europe), operating out of France and its existing operations continued as an American spin-out called MAM-A. Nevertheless, this gives this particular disc some pedigree …

However, the disc dye type is identified as “6” rather than the “5” of the GD74 above, so perhaps that reflects a later version of the dye.

I ran a bevy of overburning tests on the disc with the lowest result returned being 25m09s51f. However, it would seem this was not to be as I did try overburning two different discs for testing – the first to 25m05s00f and the other to 24m40s00f and while both “claimed” to complete their burns, netiher was readable to the end, so quality scans were only done up to the 23m00s00f range.

The first sample went to my Lite-On iHAS624 A, at 16x CLV. The write claimed to be successful but the overburn was not readable without error.

Testing on the DW1640 showed smooth readback but again, there is a read error in the overburned area. The actual block error rate is low and very acceptable, but the jitter is surprisingly high, averaging almost 20% which I consider unacceptable. This can be the result of burning “unknown” media or using more modern format burners on older media.

The DH16A6L seemingly agrees, but cannot judge the jitter. I suspect the disc itself is not at fault – just the iHAS624 A being not a great CD burner for this type of disc.

Using the BenQ DW1640 at 12x CLV, I tried a less aggressive overburn which claimed to have completed.

Unrecovered read errors still occurred in the overburned area, but the block error rate is still quite low and jitter is much better averaging about 8.6%.

The DH16A6L seems to agree with this. So while the media may not appear to be more than a “decorated” silver-gold type phthalocyanine CD-R, it is made by the vestiges of Mitsui and perhaps with the mindset that the annular buffer of 2cm around the outer edge would limit the effect of many forms of damage and dye fading that might occur due to sideways light exposure (e.g. via the edge of a jewel case). Aside from that, it looks quite cool, making it my favourite CD-R.

Verbatim PhotoSave DVD-R (8x)

This third disc comes to me also thanks to pepst – the Verbatim PhotoSave DVD. I must have been sleeping under a rock since I’ve never heard of these, even though they are seemingly still available in some corners online as “new-old-stock”.

While I don’t have the outer packaging, it would look something like this, with the promotional material targeting it as a no-fuss, easy-to-use, three-click automatic photo backup solution. It hails from 2009 and the included software is compatible with Windows 2000 through Vista. The bare disc itself is branded clearly on the top, with an area for labelling on the bottom and with the translucent area broken up by a pattern. The underside of the disc looks like a regular MCC disc, but with some data already burned on the disc.

Images as taken by my DSLR, which provide a bit more of a true-colour representation.

The stamping code on this disc is ZE0077-DVR-147C. Looking closely with the microscope, we can see a thin hairline power calibration area/program management area write, the lead-in, DVD-R pre-embossed area, but also a hairline to the border-in of the next session where data can actually be appended to the disc.

----------------------------------------------------------------------------
Unique Disc Identifier : [DVD-R:MCC 02RG20  ]
----------------------------------------------------------------------------
Disc & Book Type :       [DVD-R] - [DVD-R]
Manufacturer Name :      [Mitsubishi Chemical Corp.]
Manufacturer ID :        [MCC 02RG20  ]
Blank Disc Capacity :    [2,298,496 Sectors = 4.71 GB (4.38 GiB)]
----------------------------------------------------------------------------
** INFO : Hex Dump Of 'Media Code'-Block Listed Below
** INFO : 4-Byte Header Preceding 'Media Code'-Block Discarded
** INFO : Format 0Eh - Pre-Recorded Information In Lead-In
0000 : 01 40 c1 fd 9e d8 52 00  02 85 0d 10 88 9a 80 00   [email protected].........
0010 : 03 4d 43 43 20 30 32 00  04 52 47 32 30 20 20 00   .MCC 02..RG20  .
0020 : 05 88 80 00 00 00 02 00  06 09 0d 13 87 78 80 00   .............x..
0030 : 07 88 80 00 00 00 00 00  08 05 15 0d 10 0b 0a 00   ................
0040 : 09 98 08 0f 0b 78 88 00  0a 80 00 00 00 00 10 00   .....x..........
0050 : 0b 07 19 15 a7 88 75 00  0c 99 bc aa a2 20 22 00   ......u...... ".
0060 : 0d 00 00 d0 00 00 00 00  00 00                     ..........

** INFO : Hex Dump Of 'Control Data Zone'-Block Listed Below
** INFO : 4-Byte Header Preceding 'CDZ'-Block Discarded
** INFO : Format 10h - Physical Format Information Of Control Data Zone
0000 : 25 0f 02 00 00 03 00 00  00 26 12 7f 00 00 00 00   %........&......
----------------------------------------------------------------------------
[ DVD Identifier V5.2.0 - http://DVD.Identifier.CDfreaks.com ]
----------------------------------------------------------------------------

The disc itself is a standard MCC02RG20, but I’ve not encountered this 8x DVD-R variant before simply because I was more of a DVD+R person at the time – MCC003 was more my thing.

The disc itself was pre-burned with a Pioneer DVR-112 based on the recorder mark. The included software is version 3.5.0.395 from Soft-R Research LLC, originally named SoftR3.exe but named PhotoSave.exe on the disc. This particular file has a size of 1,884,264 bytes with MD5 of 41c47e7dbf6c117b53f1dca250034936.

As I only had one sample of this interesting media, I didn’t want to burn it up.

Without burning the disc, I could get this far on a Windows 11 machine – the initial dialogue that allows one to select photo sources and commit the burn. But I reasoned that it should be possible to copy this disc, albeit, it was more difficult than expected.

Recreating a PhotoSave DVD

The PhotoSave DVD has an interesting structure. Using ImgBurn, we can see that the disc has the following properties:

Status: Incomplete
State of Last Session: Empty
Erasable: No
Sessions: 2
Sectors: 1,248
Size: 2,555,904 bytes
Time: 00:18:48 (MM:SS:FF)
Free Sectors: 2,203,936
Free Space: 4,513,660,928 bytes
Free Time: 489:47:61 (MM:SS:FF)
Next Writable Address: 93952

As a multi-session, open disc, it’s not easily replicated. You can certainly image the disc to an ISO, which will give you a file of 2,555,904 bytes with an MD5 of e6ff47838d328ccb92e99f60c90a48fe, but to write the disc is something else.

If you try with ImgBurn, you’ll find it cannot be replicated properly as ImgBurn cannot produce “open” discs.

Status: Complete ***
State of Last Session: Complete ***
Erasable: No
Sessions: 1 ***
Sectors: 65,264 ***
Size: 133,660,672 bytes ***
Time: 14:32:14 (MM:SS:FF) ***

The lines marked in asterisks differ from the original disc. This is frustrating, so I decided to give Infrarecorder a try – it’s supposed to be able to do tricky things …

I thought leaving the “Close disc after writing” and “Pad data tracks” unticked would be enough … but the disc it created was the same as Imgburn with the same properties. Both of them seemed to enlarge the session despite the filesystem being much smaller.

In the past, I might have tried CloneCD/CloneDVD but since I don’t have them handy and I’m not sure they can handle open discs, I decided to give Nero 6 a try. It’s usually pretty good about things – I imaged the actual disc and burned the image using Nero 6.

Status: Incomplete
State of Last Session: Empty
Erasable: No
Sessions: 2
Sectors: 65,264 ***
Size: 133,660,672 bytes ***
Time: 14:32:14 (MM:SS:FF) ***
Free Sectors: 2,203,936
Free Space: 4,513,660,928 bytes
Free Time: 489:47:61 (MM:SS:FF)
Next Writable Address: 93952

The resulting disc had an enlarged session size too, despite the filesystem being the same (much smaller) size. At least the next writable address is the same … so will it work?

Nope. It didn’t like it. Perhaps it’s simply not happy that the disc is not the same, so let’s see what happens when I try burning the ImgBurn image using Nero instead.

Status: Incomplete
State of Last Session: Empty
Erasable: No
Sessions: 2
Sectors: 1,408 ***
Size: 2,883,584 bytes ***
Time: 00:20:58 (MM:SS:FF) ***
Free Sectors: 2,203,936
Free Space: 4,513,660,928 bytes
Free Time: 489:47:61 (MM:SS:FF)
Next Writable Address: 93952

The size discrepancy is smaller, but it seems Nero had padded the session in some way, so the numbers aren’t the same. This copy didn’t work either.

I had to get the big guns … out comes Linux and cdrskin.

With these options, I wrote the ImgBurn image to a blank disc and … this time the numbers all matched the original. But still, I get the error that the disc is “not a valid Verbatim PhotoSave disc”.

Digging Deeper

By now, I’ve ruined a few blanks, so I’m annoyed but at least not much out of pocket. I decided to take a look into the executable to see if I can learn the secret of how it determines which disc is valid. I didn’t even bother to disassemble it – it looked like a Visual C++ 7 executable.

After searching for strings, I came across “IDS_MB_NOTAVERBATIMDISC” and the following string is the hard-coded media code. I wonder what would happen if I just patched the bytes that follow to match the RITEKF1 I had on hand …

So I applied this patch to the image and burned it to a disc.

Success! No tricks here – you’re seeing PhotoSave running on a RITEKF1 16x DVD-R under Windows 11. Through looking at strings in the executable, it seems this program traces its heritage back to the CDSoft-R Cryptex CD-R from 2007 as there are still references to encryption in the program. Perhaps it’s just been reskinned.

Verbatim PhotoSave Software

The software can automatically scan your pictures folder, your whole drive, that of a digital camera card to find photos to back-up, which makes it a near-automatic back-up proposition. But in the manual mode …

… you have a sort of “composer” window where you can add files or folders to your disc, with a small preview on the side. Some basic photo operations, like rotating, are supported.

The window automatically traverses folders if selected, with the files being added to a specific folder if you wish. A small preview is given to the side and the capacity consumed is indicated.

Once all photos are placed into the composition window, you can decide whether you want to close the disc. The disc’s capacity is indicated by the graphical bar and in text below.

Once you hit record, the progress bar appears and the write is performed. You don’t get a choice on write speed, you don’t get any estimated time. It’s bare bones as can be, which might be bad news if your burner doesn’t do well at the default write speed. At least the original MCC media has a good chance of working properly with most drives. Once completed, you reach the final screen where you can click “Finish” and the program closes.

If a disc is not closed, then on the next start-up, you get the option to explore or add more photos. If you choose to add more photos, the current session’s consumption is highlighted in a different colour. Nevertheless, the process is the same – but if I tick the box to close the disc, then no further appending will be possible.

In that case, the burn proceeds as usual, except that the disc is closed after the burning completes.

When the disc is closed, the program is not “unlinked” from the filesystem. Instead, it is still set to Autorun, but it will only allow you to explore the disc, with an automatic countdown before launching Explorer (or you can eject or cancel). All image files are just burned to the disc in their respective folders – there’s no special organisation beyond that.

While this whole set-up is amazingly simple and works under Windows 11 still, it’s perhaps a bit too simple and I can’t see it having sold too well. For one, if you’re under Mac OS or Linux, there’s nothing in it for you. For another, free burning software certainly existed, so the premium these discs would have commanded may have been difficult to justify. They also don’t do anything specific to better safe-guard your data either – no redundancy, no way to automatically produce multiple copies. It’s also very specific about needing their media – if you are backing up too much for one disc, it can “span” the backup across multiple discs, but they all have to be PhotoSave discs. So perhaps it’s a bit of a way to make a self-burning disc, but perhaps not all that appealing to consumers.

Memorex Pro Gold Archival

Now this was a lucky find on a thrift run that I did towards the end of 2025. Visiting a Salvos Store in St Marys, I managed to come across a pack of five Memorex Pro Gold DVD-R Archival going for a very reasonable price. Even if it wasn’t, I would have bought it anyway to satisfy my curiosity.

While Memorex products are often Ritek-aligned, this is one area where I don’t think Ritek are involved. The main giveaway is that the disc is Made in France, which highly suggests this is a MAM-E product (i.e. of Mitsui heritage), but it is packaged in the USA (where MAM-A would be the expectation). These discs advertise a 24k gold reflective layer, 8x recording in the DVD-R format with a limited lifetime warranty. The date on the packaging suggests these discs would have been around 2006, making them just around 20 years old this year.

Rather unusually, they come in a “crunchy” plastic case (the sort that splinters, like CD jewel cases) of a similar size to a DVD library case, but not as thick. It’s hinged to open but the inlay card sits very loosely in the front, taking every opportunity it can to become misaligned.

The rear of the inlay card contains some tips for content preservation – better quality scans of the inlay are shown below.

The disc itself has a translucent gold logo top, but the disc does have a gold reflective layer.

This is somewhat evident as the underside which is usually purple has a more gold-purple colour.

This is more clearly indicated in the DSLR photos of the disc.

The stamping area code has a TDK format – “DVD-R 5301A2” with a laser marked code “E0317”. There is an inked code in the clamping area – “5353D1110539 3 G”.

----------------------------------------------------------------------------
Unique Disc Identifier : [DVD-R:TTH01]
----------------------------------------------------------------------------
Disc & Book Type :       [DVD-R] - [DVD-R]
Manufacturer Name :      [TDK Corp.]
Manufacturer ID :        [TTH01]
Blank Disc Capacity :    [2,298,496 Sectors = 4.71 GB (4.38 GiB)]
----------------------------------------------------------------------------
** INFO : Hex Dump Of 'Media Code'-Block Listed Below
** INFO : 4-Byte Header Preceding 'Media Code'-Block Discarded
** INFO : Format 0Eh - Pre-Recorded Information In Lead-In
0000 : 01 40 c1 fd 9e d8 52 00  02 85 0d 0c 87 89 90 00   [email protected].........
0010 : 03 54 54 48 30 31 00 00  04 00 00 00 00 00 00 00   .TTH01..........
0020 : 05 ba a3 22 32 20 02 00  06 09 0a 15 98 89 90 00   ..."2 ..........
0030 : 07 88 80 00 00 00 00 00  08 05 15 0c 0e 09 04 00   ................
0040 : 09 96 05 0b 0b 78 88 00  0a 80 00 00 00 00 10 00   .....x..........
0050 : 0b 09 18 14 97 88 85 00  0c b8 9a b8 83 00 30 00   ..............0.
0060 : 0d 00 00 d0 00 00 00 00  00 00                     ..........

** INFO : Hex Dump Of 'Control Data Zone'-Block Listed Below
** INFO : 4-Byte Header Preceding 'CDZ'-Block Discarded
** INFO : Format 10h - Physical Format Information Of Control Data Zone
0000 : 25 0f 02 00 00 03 00 00  00 26 12 7f 00 00 00 00   %........&......
----------------------------------------------------------------------------
[ DVD Identifier V5.2.0 - http://DVD.Identifier.CDfreaks.com ]
----------------------------------------------------------------------------

Indeed, the media code block indicates TTH01 which is a TDK MID, suggesting the use of TDK dye.

Looking closer at the edge, there seems to be a gold “spot” too – I wonder how consistent the manufacturing is, as the reflective layer is a bit uneven towards the extreme outer edges.

Tracking Error and Focus Error (TE/FE) tests on the DW1640 show exceedances for 16x and 12x recording.

At 8x, there is an exceedance for TE at the extreme outer edge but 4x and 2.4x are both fine.

My BenQ DW1640 was given the honours with this disc, burning at a recommended 4x.

The readback had some fall-backs towards the end, likely due to TE/FE issues. The scan shows excellent jitter and a stable, but averagely-high PIE rate. PIFs were also somewhat elevated towards the inner diameter, but nothing that would cause readability issues.

The DH16A6L gave a smooth readback curve and a very similar appraisal of the disc’s error rate. Where results coincide, this usually means good news for the reliability of the test.

Even the grumpy SHM-165P6S seems to agree mostly that the PIE rate is elevated but almost stable across the disc, with the PIF declining slightly towards the end of the disc.

The jitter is stable, but somewhat higher than the limit line – but I consider this normal as this drive maybe somewhat optically compromised. Unfortunately, this seems to be because of a notable amount of peak shift, especially for pits, so the strategy isn’t quite perfect. The beta level is within limits, but a little on the low side.

So in the end, not a perfect burn – perhaps my Pioneer DVR-111L could have done better, but I’d rather not expend another disc to find out as this burn is still stellar in terms of its jitter level even if its PIE/PIF rates are average.

Conclusion

Optical discs and archival seem to go hand-in-hand. Even before inorganic media became available, discs with gold reflective layers and durable organic dyes were marketed for archival use. Other times, the discs appeared more ordinary, but the case of the “Designed for Photo Applications” disc, with a 2cm ring of “dead zone” which likely avoided many sorts of damage affecting the data area. In the case of the Verbatim PhotoSave DVD-R, it was a case of providing pre-recorded software to make photo back-up easier for the uninitiated.

It was fun to see that even old optical media, stored well, can still perform well today given the right drive. But it was even more interesting to have worked my way around the PhotoSave DVD-R, allowing me to replicate a copy for documentation on modern media without spoiling the original disc. Despite its age, it even works under Windows 11, with its roots hailing back to the CDSoft-R Cryptex CD-R with references still remaining within the executable file.

While organic dyes are known to be light sensitive, under proper storage conditions, they can still last sufficiently long and so inorganic discs are not a “must” for long term storage. However, inorganic discs are generally more stable and less sensitive to environmental conditions – so perhaps you might have to consider cost and availability as well. But as the Memorex documentation indicates – it’s not just the media, you have to be able to write, handle and store it well.