[HH] Burn custom art on a DVD

[Tom Metro]

Jerry Feldman relayed:
> We are using special DVD burners, called light scribe. These types of
> burners, have the 3mW laser source which has ability to engrave the
> patterns on the other side of DVD.
> http://en.wikipedia.org/wiki/LightScribe

Sure, somebody mentioned LightScribe, but I failed to look up how it works.

While it doesn't alter a substance under polycarbonate, it still seems
like the changes made by the laser are superficial (quoting Wikipedia):

  The surface of a LightScribe disc is coated with a reactive dye that
  changes color when it absorbs 780 nm infrared laser light.

Is this really going to result in the grooves you require for your
application?


> We are planning to use these types of burners to make the patterns on
> our coated DVD.

Hmmm...perhaps "our coated DVD" is a clue. Are you using custom made DVD
blanks that use something different from the reactive dye used on the
normal blanks? That might explain how you'll accomplish producing groves.


> There are several softwares on the internet that can control this type
> of DVD burners to engrave the patterns on the back of DVDs, but most of
> them are not accurate enough for our purpose.

Usually making software accurate is the easy part. Making the hardware
accurate is hard. So the fist thing I'd look at is whether your
requirements exceed the precision capabilities of the hardware. Chances
are the existing commercial software already takes maximum advantage of
the precision the hardware offers.

Have you looked up any specifications from a LightScribe recorder? Have
you talked to an application engineer at a LightScribe drive
manufacturer? (This may be challenging, if they have been discontinued.)

The Wikipedia article also notes:

  Companies such as HP, Samsung, LaCie and LiteOn have discontinued or
  are phasing out LightScribe drives as of June 2013[5] with only LG
  still manufacturing drives.

  As of the beginning of January 2014, LightScribe.com HP's official
  LightScribe website has been removed.

which raises a bunch of other questions:

Are you developing a one-off project, or a product that is going to be
deployed into the field?

If the latter, where are you going to source your LightScribe drives if
they have been discontinued? (Is LG still making them?)

Is the code going to need to work with whatever brand LightScribe drive
the user happens to have? If you are exceeding the typical precision
capabilities of a LightScribe drive, you may find that only a specific
drive model is capable of hitting your requirements.


> So I was wondering if we could have a interface code or software to
> control the the laser beam and its position.

If the hardware is in fact capable of the precision you require, the
drives adequately available, and the LightScribe API documented, and the
limiting factor only the commercial LightScribe tools, then it should be
fairly straight forward to find a software developer to implement this.


It appears there are some Linux packages available to create LightScribe
labels:
http://www.thepowerbase.com/2011/10/lightscribe-in-linux-make-gorgeous-discs-in-ubuntu/
http://www.lacie.com/us/support/drivers/driver.htm?id=10061
http://www.pawtec.com/lightscribe/

Some of these link to files on http://download.lightscribe.com/, which
appears to still be working.

I don't see any source files for any of these packages.

The last link does link to an SDK package, which I'm guessing contains a
binary library and header files.

I also see there are two Qt GUI apps:
http://sourceforge.net/projects/qlscribe/
http://sourceforge.net/projects/qscribe/

which have source, and are currently maintained, but are built on top of
the same binary library provided by lightscribe.com.

If the binary driver supports the precision you require, modifying that
GUI may get you there. (And if so, contacting the lead developer on that
project might lead to finding a developer already well familiar with
these APIs.) (But I'm pretty doubtful that the GUI is the limiting factor.)

If the binary driver doesn't do what you need (but the underlying
hardware does), and the manufacturer doesn't provide documentation for
the hardware interface (treating it similar to a "winmodem" situation
where big chunks of proprietary, low-level code resides in the driver),
then you could be facing a large project to reverse engineer the hardware.

 -Tom