Hallöchen! Sebastian Kraft writes:
> Am 23.09.2014 um 07:53 schrieb Torsten Bronger: > >> [...] >> >> I don't see an actual difference between returning 1.6e16, inf, >> or nan. The caller has to special-handle it in either case. > > No, the caller should not have to handle these cases. [...] > > Example: Some operations lead to a varA = NaN which actually is > meant to be varA = MAX_FLT and fixed accordingly. If later on we > calculate something like varB = 1/varA we can get a valid output > of the function. However 1/NaN = NaN can never be recovered or > interpreted correctly. > > Can anyone summarize in which lensfun functions the invalid NaN/inf > values occur? Then I will have a look at this... For example, see the function at https://sourceforge.net/p/lensfun/code/ci/099f390d5d7f0df7ae2242f3bbc5973d29442950/tree/libs/lensfun/mod-coord.cpp#l883 The "asin" returns a NaN if its argument is larger than 1, which means that both iocoord's also become NaN. You cannot recover it: For such pixels, there is no counterpart in the sensor image. Tschö, Torsten. -- Torsten Bronger Jabber ID: torsten.bron...@jabber.rwth-aachen.de or http://bronger-jmp.appspot.com ------------------------------------------------------------------------------ Meet PCI DSS 3.0 Compliance Requirements with EventLog Analyzer Achieve PCI DSS 3.0 Compliant Status with Out-of-the-box PCI DSS Reports Are you Audit-Ready for PCI DSS 3.0 Compliance? Download White paper Comply to PCI DSS 3.0 Requirement 10 and 11.5 with EventLog Analyzer http://pubads.g.doubleclick.net/gampad/clk?id=154622311&iu=/4140/ostg.clktrk _______________________________________________ darktable-devel mailing list darktable-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/darktable-devel
