Read/Modify/Write RAW data tools?

[Panopeeper]

Kwash, are you working with Windows or Mac? If Mac, do you have Crossover or some other simple Windows simulator?

[ejmartin]

Don't mind Jonathan, incivility and condescending dismissiveness are the norm.

[Jonathan Wienke]

That's pretty cool, are they glass multicoated filters?  Do you special order them with the CC mix you want?  I've looked and searched a number of threads and pages on Uni and I've never seen these.

My apologies, I can't find the reference I had with spectral response plots of various photo filters. The closest thing I can find is something along the lines of a CC30M magenta filter with a CC05B blue filter or the like (depending on your camera characteristics and the lighting you want to shoot in) stacked over it as necessary to fine-tune the red/blue channel relationship. The filters I linked are multicoated glass, and priced accordingly. Not cheap, but a fairly reasonable single-shot solution if you really want to go that route.

[kwalsh]

Kwash, are you working with Windows or Mac? If Mac, do you have Crossover or some other simple Windows simulator?

I typically do my photo work on a Windows box, but I have both Mac and Linux boxes in the house that I could use for experimenting.  At this point I don't really care about a clean and easy work flow - if I actually find some RAW trick is truly worth the effort then I'll invest the time to code it up.  For now a kludge in any OS is fine.

My apologies, I can't find the reference I had with spectral response plots of various photo filters. The closest thing I can find is something along the lines of a CC30M magenta filter with a CC05B blue filter or the like (depending on your camera characteristics and the lighting you want to shoot in) stacked over it as necessary to fine-tune the red/blue channel relationship.

Thanks Jonathan.  I had seen those filters on B&H, and available as circular threaded mounts as well.  For my purposes I'm really interested in optimizing the R channel so I could probably even get away with a single red filter - though given the cost I'd certainly experiment with gels first.  And yeah, finding spectral curves for filters is a bit of a pain, I've managed to dig up transmission curves for all the B+W filters and curves for the glass that goes into many of the Hoya filters.  Haven't found a good fit for my camera's daylight spectral characteristics in glass - of course there are many, many more options in gels.  Thanks for the links.

[Panopeeper]

I typically do my photo work on a Windows box, but I have both Mac and Linux boxes in the house that I could use for experimenting.  At this point I don't really care about a clean and easy work flow - if I actually find some RAW trick is truly worth the effort then I'll invest the time to code it up.  For now a kludge in any OS is fine.

I can give you a program, which cuts an uncompressed DNG file in three parts, one of them being the data. However, I'm afraid there is something you did not think through: the raw channels can not be assigned to any Photoshop channels. You can process them only as a greyscale image, and I doubt that you will be lucky with that.

Regarding the magenty filter: the maximum gain you can achieve is about 0.5 EV; I demonstrated this in http://luminous-landscape.com/forum/index....st&p=230490
Of course this depends on the actual illumination and scenery.

ADDED

My statement regarding the processing as a greyscale images relates only to the way of loading the raw image data as "raw" type in Photoshop; a program created specifically to that purpose can associate the raw chzannels with their proper meanings, for example when merging two images.

[ejmartin]

I can give you a program, which cuts an uncompressed DNG file in three parts, one of them being the data. However, I'm afraid there is something you did not think through: the raw channels can not be assigned to any Photoshop channels. You can process them only as a greyscale image, and I doubt that you will be lucky with that.

ADDED

My statement regarding the processing as a greyscale images relates only to the way of loading the raw image data as "raw" type in Photoshop; a program created specifically to that purpose can associate the raw chzannels with their proper meanings, for example when merging two images.

IRIS can separate the Bayer array into its component color channels, and remerge them again.  It also has a variety of processing tools.

[eronald]

There's another reason writing back to the RAW format is stupid: unless you're shooting with a digital back, the RAW format is limited to 12 or 14 bits. If you're exposure blending to increase captured DR, limiting yourself to 12 bits is going to leave you with quantization noise problems in the shadows 4 stops sooner than a 16-bit format. Why go through the hassle of shooting and blending multiple exposures, and then waste the effort by saving the results in a format that isn't designed to handle it?

Jonathan,

  Stupid is a very strong word; there are zillions of reasons one might want to filter the raw pixels - maybe the reasons don't appeal to you but that doesn't make them dumb.

Edmund

[papa v2.0]

do you have matlab.


it can read and write raw files and any image type files.

its what i use for my image processing along with photoshop etc

[Jonathan Wienke]

Jonathan,

  Stupid is a very strong word; there are zillions of reasons one might want to filter the raw pixels - maybe the reasons don't appeal to you but that doesn't make them dumb.

It's not so much the filtering of the RAW pixels I think is stupid; it's saving them back to RAW format again when there are other options that will serve the purpose much better and cause fewer issues down the road.