> I'm writing a spec that would require submissions from at best uncompressed > raw files and at worst least jpg compression files, interpolated by the > photographer and submitted as RGB tiff.
Delivery of Raw data is accepted by some libraries, but then they run all the adjustments to your work. If you have some "signature" in some aesthetic aspect of the image, then you are not delivering it. > Having recently played with a Canon D60 and seen very (very) slight edge > artefacts I'm considering which cameras may qualify as serious film > replacement platforms. > The time has come. Digital cameras and photolibraries. I would say that no one can judge any camera in the first place before finding out the group of specs and running the tests you are talking about in the first paragraph, as the D60, even the D30 and all of those 6Mp cameras in general can provide what you ( and libraries ) are loking for, if you provide the adequate treatment to the files. > the spec is for 55mb (A3 @300dpi). This would involve > interpolation from the camera original files. > I think the criterion need to be carefully thought out both in terms of jpeg > / tiff / raw and also sharpening and interpolation methods. If you consider interpolation as a necesary step to achieve 55Mb files then jpeg is not an option and you are doomed if your original files are in jpeg. Being very optimistic , the possibilities of sucess are extremely limited. If you think of capturing in Raw, processing to obtain 16 bit files, work with them, including the interpolation and come back to 8 bits just prior to delivery you should handle well. >my experience in recent weeks have shown that my > tried and tested method of upsizing with a CCD camera has not yielded the > best result when comparing different approaches with CMOS files. THe "battle" betwen CCD's and CMOS chips is important, but out of our control, and only if Nikon is working in CCD powered camera, then some comparisons could be made, since the current offer from Canon and Kodak ( for 35mm DSLR) is only CMOS. COntax has not recovered from its drama, unless they offer a new breakthrough discovery in the coming PMA, which should be opening in a few days. There are indeed LOTS of expectations regarding this PMA. > Similarly, some camera files have to have a minimal amount of sharpening or > end users will never buy on the basis that they are too soft (D30 / 60). I am all for testing, and trying to define specs, since this comment sounds pretty much like an urban myth and could be understood and accepted if formal evidence against it would not be available, but it is. All those cameras bring either an IR filter or what they call a "Low Pass " filter infront of the chip and this is the reason all those cameras provides original, native files with some degree of softness ( assuming you are using quality lenses, not plastic cheapos) and this includes the 1Ds. Sharpenning methods to counter this softening effect ( First Pass Sharpnening in Luminosity, High Pass Filter in soft light) have been devised , proven and confirmed as effective to overcome this aspect of file capture in 35mm DSLR camerasand this step must be included in the workflow. THere is no reason whatsoever( well, maybe ignorance of the facts) to condemn any camera because of this. Check the links in the archives. I have gone ( with some files) from the D60 up to 75 Mb@300 dpi with fantastic sucess. Anything smaller than that is just a piece of cake. You can even choose from different methods to upsize , depending on what is the content of the file( if you already have soft focus, or blurrs, etc) and also choose your best method to upsize a perfect file to a bigger size. Same with sharpenning. If you want to really run a complete study on this step ( interpolation) you will have to take a look at both Photo Retocuh Pro which is a rather expensive( but amazing) solution and my big favorite, less expensive alternative, which I would rate it the BEST of them all is the app called S-Spline2 , which you can get for trial from http://www.s-spline.com. Forget about stair interpolation( even though this one works well too), forget about Genuine Fractals, run the S-spline and see for yourself. Perhaps it is just wise to wait until the PMA and see what is coming to the market before preparing such specs and tests? All the best. Jorge Parra APA/ASMP www.jorgeparra.com =============================================================== GO TO http://www.prodig.org for ~ GUIDELINES ~ un/SUBSCRIBING ~ ITEMS for SALE
