11-28-2012, 09:09 AM
(This post was last modified: 11-28-2012, 09:12 AM by Brightcolours.)
[quote name='Klaus' timestamp='1354091363' post='21101']
Just to mention ... RPP uses DCRaw.
[/quote]
RPP uses 3 interpolation methods (which is what demosaicing is), which have been implemented by the developer himself. That RPP uses DCRAW to read different RAW file formats is not very interesting, in regards to what we were discussing in the other thread... Just about all RAW converters use DCRAW to read the files, not all use its interpolation methods. That is the beauty of the DCRAW source, one is free to use it how one wants.
To me it seems not very advantageous to let a DNG converter handle demosaicing (the interpolation to fill in the 2 missing channels) for the Fuji RAW, since that is the step where loss of detail can/will take place. Not sure why the Fuji results look like they have been quite heavily AA-filtered, though.
For people who can not run RPP because it is written for Mac OS X, Raw Therapee (available for Mac OS X, Windows, Linux) also is quite good in detail recovery during demosaicing, and offers 4 different interpolation methods.
Back the the 35mm f2 Canon samples I posted, what you are writing is incorrect. The images 3 and 4 I posted cover the whole range from in focus to totally OOF, including very smooth transitions. It makes no sense to only then look at the very OOF background and say that all lenses will perform likewise, ESPECIALLY when the Sigma 35mm f1.4 samples show that the sigma does not perform likewise.
http://blog.sigmaphoto.com/wp-content/up...4-1200.jpg
http://blog.sigmaphoto.com/wp-content/up...266529.jpg
The Sigma shows double line tendencies in the transition from focus to very OOF. Most clearly you can see it with the thumb, where it has a double "edge".
My 3/4 samples do not show that in any way, look at the leaf edge with the butterfly, look at the OOF flowers, they do not show that.
Anyway, in the article from lensrentals I linked to the author also has noticed the same: "...the backround bokeh is not nearly as smooth and buttery as the Nikon or Canon 35 lenses.". He of course is referring to the f1.4 lenses he compared it with.
Again, the 35mm f2 was just one of the lenses I listed from lenses with a similar focal length that have a more benign bokeh characteristic. Does not mean the 35mm f2 is faultless, nor top lens. It just got lifted out by someone who knew I use that lens.
Just to mention ... RPP uses DCRaw.
[/quote]
RPP uses 3 interpolation methods (which is what demosaicing is), which have been implemented by the developer himself. That RPP uses DCRAW to read different RAW file formats is not very interesting, in regards to what we were discussing in the other thread... Just about all RAW converters use DCRAW to read the files, not all use its interpolation methods. That is the beauty of the DCRAW source, one is free to use it how one wants.
To me it seems not very advantageous to let a DNG converter handle demosaicing (the interpolation to fill in the 2 missing channels) for the Fuji RAW, since that is the step where loss of detail can/will take place. Not sure why the Fuji results look like they have been quite heavily AA-filtered, though.
For people who can not run RPP because it is written for Mac OS X, Raw Therapee (available for Mac OS X, Windows, Linux) also is quite good in detail recovery during demosaicing, and offers 4 different interpolation methods.
Back the the 35mm f2 Canon samples I posted, what you are writing is incorrect. The images 3 and 4 I posted cover the whole range from in focus to totally OOF, including very smooth transitions. It makes no sense to only then look at the very OOF background and say that all lenses will perform likewise, ESPECIALLY when the Sigma 35mm f1.4 samples show that the sigma does not perform likewise.
http://blog.sigmaphoto.com/wp-content/up...4-1200.jpg
http://blog.sigmaphoto.com/wp-content/up...266529.jpg
The Sigma shows double line tendencies in the transition from focus to very OOF. Most clearly you can see it with the thumb, where it has a double "edge".
My 3/4 samples do not show that in any way, look at the leaf edge with the butterfly, look at the OOF flowers, they do not show that.
Anyway, in the article from lensrentals I linked to the author also has noticed the same: "...the backround bokeh is not nearly as smooth and buttery as the Nikon or Canon 35 lenses.". He of course is referring to the f1.4 lenses he compared it with.
Again, the 35mm f2 was just one of the lenses I listed from lenses with a similar focal length that have a more benign bokeh characteristic. Does not mean the 35mm f2 is faultless, nor top lens. It just got lifted out by someone who knew I use that lens.