NOTE: The galeon mailing list is no longer active. The list archives are made available for historical reasons.
On 7/13/05, Rob Raskin <raskin@xxxxxxxxxxxxxxxxxxxx> wrote:
Yes. That is what I am suggesting. This is mentioned in the GeoTIFF specs as a possible use for that parameter pair. Given that there is no concept of raster spacing in the vertical direction, how else might you interpret the vertical scaling parameters? Even if we created 2-D vertical-longitude profiles, these should have x- and y- axes within the GeoTIFF, for broswers and other software to render correctly.
Rob,I had always assumed that the "z" dimension stuff was intended to apply to a possible extension of the TIFF format to hold 3D arrays
of values for a given "sample". The libtiff API has also always hasa "z" parameter that seems to imply this sort of 3D capability in TIFF though I have never understood why. I personally don't think it makes sense for the "z" dimension of geotransmatrix to be used to describe the value of pixels while
the x/y values talk about position of pixels. This is especially confusing usage in the situation where the value being describedis not elevation but some other physical quantity like radiance, pressure, water accumulation, etc. For that matter, the geotransmatrix
gives no opportunity for different scale/offset for different samples (bands)though that might clearly be necessary in some applications. I am also not aware of anyone using the "z" dimension in geotransmatrix in the way that you suggest. But it is certainly feasible.
Best regards, -- ---------------------------------------+-------------------------------------- I set the clouds in motion - turn up | Frank Warmerdam, warmerdam@xxxxxxxxx light and sound - activate the windows | http://pobox.com/~warmerdam and watch the world go round - Rush | Geospatial Programmer for Rent
galeon archives: