2007.01.17 02:25 "[Tiff] Elevation Data", by Craig Bruce

2007.01.17 15:55 "Re: [Tiff] Elevation Data", by Joris Van Damme


It seems difficult to make a completely interoperable image using floating-point samples. I have heard that some software which supports floating-point TIFF checks all the sample values in order to best understand how it should be displayed (assuming it can be reasonably displayed at all).

In practice, you are correct. There's just not much floating point image data written, and the few that do write it aren't always concerned with interchange.

But I do believe there is a consensus emerging, in that the description in my previous message does apply to the biggest part of what is 'out there', and the others have all kinds of ranges and as such don't qualify to derive a consensus from.

What, exactly, is the problem with using SMinSampleValue and SMaxSampleValue for this?

Just convention, historical use and especially lack of use of that tag, arising from the fact that the specification clearly says it's worth nothing really except in some statistical sense (and who knows what that might mean - iow, it means nothing).

If the spec were clear and binding on this, maybe things would have been different. But it isn't.

In any case, please don't take any opinion I've expressed on this issue as written in stone. I'm merely trying to see what makes sense on the one hand, and on the other is de facto supported by at least some bigger part of what I have gathered as test images from different sources over the years.

It makes sense for floating point grayscale data, or any data under that Photometric tag, to assume the range of 0.0 to 1.0, simply because there is no other intrinsic range on this. For example, 0.0 to 255.0, makes no sense, unless there is something intrinsic in the concept of grayscale that binds it to 8 bit ranges in particular. Amongst the few testfiles I have, some part applies this range 0.0 to 1.0, and others do all kinds of ranges. I do that 0.0 to 1.0 range myself in my proprietary code. And I've seen little or virtually no use of (S)MinSampleValue and (S)MaxSampleValue sofar, and don't use it myself either.

In need to stress again, I mean nothing here as written in stone, I'm just trying to see a common denominator in de facto usage that seems to make some sense.

Best regards,

Joris Van Damme
Download your free TIFF tag viewer for windows here: