DVD As Media For Digital Image Distribution?

Buran asks: "I work in a research lab at a major university. We currently record a lot of visual data from a light microscope on large Panasonic optical disks. They're so old that replenishing the supply is iffy, and we can't get replacement recorders or players anymore. We want to move to DVD for image storage because they're much smaller, players are widespread, new computers can play them, and because there's a lot of storage space on each disc. However, I'm having trouble finding information on DVD recorders (certainly we have the technology to make them!) and players with high-quality digital output (for making digital movies for sending to collaborators over the Internet.) I'm hoping the Slashdot community can provide suggestions. I'm all set up to import FireWire video from digital sources, now all we need are the recorders and players." While a neat idea, I think that it would be easier to use something more open. How about tying a video microscope to a PC, saving the images in a format like JPG for distribution on a CDR? Thoughts?

Re:Loss of data in JPG

[setecastronomy]

AFAIK, a JPEG saved at the highest quality setting loses no data, at least not any that can be seen. It does tend to have problems with text and line drawings, though, so if you need *perfect* replication of the original, just go with PNGs. Otherwise, JPGs provide excellent compression with virtually undetectable data loss, up to level 20 or so (in my experience).

In the satellite lab...

[chowda]

We had a dish grabbing images of much of the east coast every hour or so from 3 NOAA sats with a resolution of ~1.2km/pixel (iirc) so the raw data after being captured was anywhere between 20M and 50M. It was highly dependant on the angle of approach and the time of day.

At any rate, each image was stored directly on to an 8G DAT drive (which was changed every 7-10 days) and then a rectangle around the state of Maine and the Gulf of Maine was cut out and placed in a directory on the hard drive. I would manually line up the physical coast on the image with where the computer believed the coast was with some fancy (read expensive) software. Then each images would be processed by a script I wrote (and some of that fancy stuff too). To make a long story longer, the channels of the data would be combined to get the temp of the land and sea (and clouds) into a data format called MCSST where every pixel directly translates into a ~1.2km square's temp, called MCSST or Multi Channel Sea Surface Temperature. MCSST is the stuff the researchers and students used for doing science. We also made gif's and thumbnails of the entire state/gulf as well as focus in on a few areas like penobscott bay. I then wrote a simple CGI to display a gallery of the thumnails and allow you to view any image you wanted as they became processed each day. We allowed anyone who wanted the raw data to access that as well.

The point of this long winded story is that we let people who were interested in our data see the light weight gif's before they wasted badwidth or disks (writable dvd is WAY more expensive per gig than DAT) on the huge data. It wouldnt have been hard to set up a burner to burn from the DAT to make custom data dvd's from a shopping cart type deal on the the web site.

Why Videomicroscope?

[Workpad z50 User]

Why limit the "open" solution to a video microscope connection? A straight-to-digital-media (plural for multiple masters) of ditigal data could be useful for digital video microscopes as well as any digital datastream (DV, audio, etc.)?

I understand why you suggest a CDR versus a DVD-RW or DVD-RAM, though. Until the format winner is manifested, I wouldn't invest a lot in either technology (not that I think DVD-RAM can overcome the to-market lead DVD-RW enjoys).

Anyway. Long format, digital video recorded real-time (hi-res) onto permanent media. Let's get it goin'!

Why not just use PhotoCD?

[spectecjr]

PhotoCD is out there, already available, and good for storing quite a few high-quality images. They're also readable on most good DVD players (IIRC).

Alternatively, Panasonic make DVD recorders now, for DVD5 format disks.

Optical Media

[Spacecase]

I run a large imaging operation, and we are also looking into DVD as a storage medium, the problems is the equipment just isn't available yet. My suggestion is to look into getting a HP Surestore Optical Jukebox system. These systems use SCSI interfaces to a jukebox that has a maximum capacity of 238 discs. Currently each disk can hold 5.2 GB making a total storage capacity of 1.2 TB. I have also heard rumors that the 10.4 GB disks will be available. The benefit of this system is it uses standard SCSI connections, has a standard disk access system (with API) and is not bleeding edge. Take a look at HP's website and see if they fit your requirements. Spacecase

do DVD-R or DVD-RAM work on Linux at all?

[Jamie Zawinski]

I'd like to use a DVD-sized device for backups, but I'm having a hard time tracking down information about using DVD-RAM and/or DVD-R under Linux at all...

Which DVD-RAM and DVD-R drives work with Linux?

I've heard a rumor that discs written by DVD-RAM drives tend to be wedded to the drive that wrote them, and can't be read on other drives. Is that true?

If I don't need to re-write the discs, but just want static backups at around 4G capacity, what should I use, DVD-RAM or DVD-R? I would have guessed DVD-R, but it looks like DVD-RAM drives are around $600 while DVD-R drives are around $5000. What is there to recommend DVD-R over DVD-RAM?

Are these discs as reliable as CD-R discs? If I verify them immediately after writing them, then put them in a box, can I count on them still being readable in ten years?

Can DVD-R discs be read in normal DVD-ROM drives? My laptop's CDROM is really a DVD-ROM -- if I fed it a DVD-R, would it be able to read my backups, or is there a difference between video and data DVDs?

Loss of data in JPG

[Louis_Wu]

Well, JPGs might be a bad idea, as a recent article [kuro5hin.org] on kuro5hin [kuro5hin.org] pointed out. The compression in JPGs reduces the resolution of detail, and smooths out the picture so that loss of data is less apparent. I don't think that a research lab wants to purposefuly destroy data. I don't know enough about the details of each format to know which would be better, but JPGs have a big drawback.

Louis Wu

Thinking is one of hardest types of work.