Roger Howard twisted the bytes to say:
Roger> On Mon, November 27, 2006 9:22 am, Daniel M. German wrote:
>> I think it is a good idea, but it will probably be problematic
>> Should the output be a set of DNGs of the map image? Will that be
Roger> I'm not sure I understand this. If you were asking me what I'd expect as
Roger> output, it would be just like if I used normal TIFF on input... I'd have
Roger> the option of individual remapped DNGs ("layers) or a single, blended
Roger> remapped DNG.
The problem is the single blended output. Computing masks and
feathering will have to be performed on this. I suggest multiple DNG
would be the first approch and then we can take it from there.
In order for this to even start we will need to build a DNG library to
read and write the files. Any volunteers? Is there any other
GPL-compatible project doing it?
>> I think DNGs will need to be supported at all stages of the workflow,
>> including enblend.
Roger> Indeed, I'm certain it will; keep in mind I only jumped into this
Roger> conversation when there was some concern that a raw file as souorce would
Roger> require demosaicing in the processing pipeline; I just jumped in to point
Roger> out that we have a compromise available in the form of demosaiced "linear"
Roger> DNG. I know this will take some effort, and as always appreciate there are
Roger> priorities and that you guys are doing a great job on your own dime.
Daniel M. German "Work. Finish. Publish. "
dmg (at) uvic (dot) ca
replace (at) with @ and (dot) with .