Archived posting to the Leica Users Group, 2009/05/21
[Author Prev] [Author Next] [Thread Prev] [Thread Next] [Author Index] [Topic Index] [Home] [Search]The Epson software for the RD1 was available as a stand alone programme and a plugin for Photoshop, which I liked a lot. I thought this was a good idea. Frank On 21 May, 2009, at 00:36, Clive Moss wrote: > I have run the experiment (with both Nikon lenses and the Canon G9 > using DxO) and the difference is eye-popping. > > I use NEFs because I use DxO software for troublesome conversions and > to correct pincushion and barrel distortion. DxO does not work with > DNGs - it wants to use the proprietary data in the NEF file. > In the future I expect that manufacturers will start taking a total > systems (and therefore proprietary) approach to lens/sensor/raw > conversion to optimize the final image. > > There have been some small steps in this direction: > Leicasonic with the LX3/D-lux 4 approaches to correcting distortion > Nikon correcting some color aberrations in Capture NX2, and rectifying > the 10.5mm fisheye. > Third parties like DxO and PTlens correcting distortion. > Even Leica used that approach to solve the IR problem, even if it was > unintentional! > Fujitsu is using non-standard sensors to improve dynamic range > > In the long run I expect the total systems approach to dominate -- it > will let smaller lighter lenses (especially zooms) and possibly lower > cost sensors produce superior images. At the low end, the emphasis > will be on cheaper components. At the high end, the emphasis will be > on better images. I expect that as processor power improves even depth > of field could be enhanced using algorithms that based on the > properties of specific lenses. > > What I am leading up to is that an open standard like DNG used by > camera manufacturers will inhibit a total systems approach. It is too > early in the life of digital cameras to freeze and standardize such a > critical component of the total system. > > Adobe's expertise does not lie in lens/sensor design, and even their > approach to raw conversion led to howls (murmurs?) of protest that the > camera's output was superior to Adobe's. Adobe (rightly) gave in to > its users and produced profiles that are closer to the camera's output > than to Adobe's idea of what the output should look like. > > It be better for the industry if Adobe opened up their architecture > and allowed third parties to plug their raw converters in to Lightroom > or Adobe Camera Raw, save their parameters in the database. Then Leica > (or anyone) could use their magic sauce to cook the raw conversion as > they pleased. Third parties like DxO could provide alternative "looks" > as they do with DxO Filmpack, so that the user could say "make this a > Kodachrome" and bingo, it is. > > (and no, the CS3 to CS4 upgrade is not worth it - I upgraded out of > curiosity, but I cannot say that it was worth it to me. YMMV) > -- > Clive > Blog: http://clive.moss.net/blog > Photographs: http://clive.smugmug.com > > > > On Wed, May 20, 2009 at 2:44 AM, Geoff Hopkinson > <hopsternew at gmail.com> wrote: >> Frank perhaps you could run an experiment there if you wished and >> let us >> know the outcome? You'll need a lens with known significant >> correction >> problems to conduct the experiment. > > _______________________________________________ > Leica Users Group. > See http://leica-users.org/mailman/listinfo/lug for more information