2004.11.18 20:16 "Re: [Tiff] Using TIFFGetField to get a "Rational"", by Eric Vergnaud

When I remember right, you get to values, and must interpret them as divident/divisor

That is certainly true on the TIFF tag value level. However, if I remember correctly, LibTiff does that already, and returns the value as a Float. That seems backed up by the link I mentioned in my other mail in this thread.

Perhaps someone with a normal LibTiff compilation could doublecheck. I'm getting good results in LibTiffDelphi passing a pointer to a pascal Single, and I do think that the C equivalent of that is Float, but it would nice if someone doublechecked.

Yes, here it is:

    case TIFFTAG_XRESOLUTION:
            *va_arg(ap, float*) = td->td_xresolution;
            break;
    case TIFFTAG_YRESOLUTION:
            *va_arg(ap, float*) = td->td_yresolution;
            break;

So you should pass a float* to TIFFGetField.

-------------------------------
Eric VERGNAUD - JLynx Software
Cutting-edge technologies and
services for software companies
web: http://www.jlynx.com
-------------------------------