[SciPy-User] How to create multi-page tiff files with python tools?

[Stéfan van der Walt]

Apologies, this message was meant for the scikits-image list.  Please
continue discussions there.

2009/10/3 Stéfan van der Walt <stefan at sun.ac.za>:
> 2009/9/30 Ralf Gommers <ralf.gommers at googlemail.com>:
>>> I think the problem is that Frederick Lundh is the only one who has
>>> permission to add/change code base.
>>> I still find it very suspicious that somewhere on the PIL website it
>>> states that you can pay (a lot of money) for a "special license" to
>>> get early  access to the development version - so even you are
>>> providing (free) patches via the mailing list, you would have to pay
>>> to get access to the patched version !?
>>> A couple months ago I asked for an explanation but didn't get a reply.
>>>
>> Yeah that is very odd. An attempt to put the I/O part of PIL in a scikit may
>> be enough of a push to improve that situation. The only other important
>> Python library I can think of that was this inert is setuptools, and look
>> what happened there.
>
> I wonder if we shouldn't take the plunge and add OpenImageIO as a dependency?
>
> Here's the list of features (from their website):
>
> - Extremely simple but powerful ImageInput and ImageOutput APIs for
> reading and writing 2D images that is format agnostic -- that is, a
> "client app" doesn't need to know the details about any particular
> image file formats. Specific formats are implemented by DLL/DSO
> plugins.
>
> - Format plugins for TIFF, JPEG/JFIF, OpenEXR, PNG, HDR/RGBE, Targa,
> JPEG-2000, BMP, and ICO formats. More coming! The plugins are really
> good at understanding all the strange corners of the image formats,
> and are very careful about preserving image metadata (including Exif,
> GPS, and IPTC data).
>
> - An ImageCache class that transparently manages a cache so that it
> can access truly vast amounts of image data (thousands of image files
> totaling hundreds of GB) very efficiently using only a tiny amount
> (tens of megabytes at most) of runtime memory. Additionally, a
> TextureSystem class provides filtered MIP-map texture lookups, atop
> the nice caching behavior of ImageCache.
>
> - Supported on Linux, OS X, and Windows.  All available under the BSD
> license, so you may modify it and use it in both open source or
> proprietary apps.
>
>
> I really don't have much hope for PIL.  The development process is
> closed and slow.  Once you ignore your community, you are pretty much
> done for.  The only reason PIL still exists is because it is useful,
> but let's face it: we can easily rewrite 80% of its capabilities at a
> multi-day sprint.  Perhaps we should.
>
> Regards
> Stéfan
>