but as you admit DNG and indeed Quicktime CAN just be a wrapper for an underlining format... that has always been my understanding though how that manifests in current practice is beyond my knowledge and understanding.
That's not precisely what I wrote. And their capabilities in this regard are quite different.
DNG format is designed to be an extensible container, derivative of a Tagged Image File Format base design, but the DNG file format specification defines a precise structure for the original sensor data + metadata which raw native files need to be reorganized to. Because the DNG file is a container, the original raw file can
also be encoded in a byte stream and added as a tagged element of the file (what I referred to as 'wrapper'ing it into the file), which the DNG Converter application knows how to extract if a user wishes to recreate the original native format raw file. This encoded native raw file is not what is read by any application parsing DNG files.
This is completely different from the QuickTime file format, which has nothing to do with the notion of a TIFF container and has far more defined (and far far more complex) structures in it. A QuickTime file can contain a huge diversity of data types and formats. How you would wrapper, say, an MPEG-4 video into a QuickTime file is a completely different order of business from how you 'wrapper' a native original raw file into a DNG file. (Also remember that the ISO specification for MPEG-4 format was derived from the QuickTime movie format, not the other way around.)
.. You are probably better placed to answer the question to what difference it makes to have a file that is recorded directly to DNG - which i am guessing what you say happens with the Leica M8/9 and a DNG converted from a E-5 ORF ?
When a camera creates a DNG file as its original file, the only thing that separates that DNG file from a DNG file converted from a native raw image file is that the camera organized the sensor data and camera metadata rather than an external application which re-organizes the elements. All the data elements are the same, the structure and identification of those data elements are defined by the publicly accessible specification.
... My point is that if the act of creating a DNG rewrites the original RAW data then maybe thats why Aperture gets its knickers in a twist
Maybe Apples lack of support may be politically more than engineering driven? who knows....
Creating a DNG file doesn't re-write the original raw data, it copies it into a well-disclosed structure. The short of this is that Aperture isn't interpreting the camera calibration table correctly with the DNG file output from the Pentax K-5 and the Olympus E-5 DNG files created by DNG Converter processing the .ORF file. Exactly why is for the Aperture engineers to determine, if they feel it's important enough, etc.
Looked at historically ...at first, Aperture would only work with DNG files for cameras that it interpreted the native raw files for. This meant that it was using its own calibrations and decode strategy rather than that offered by the DNG specification. Then, because several important cameras were released which produced DNG original files, Aperture's capability was expanded to process any mosaic data DNG file up to DNG spec rev 1.2. This missed linear RGB encoded DNG files and missed the DNG files which relied upon DNG spec rev 1.3 to include parametric lens correction metadata, a new form of metadata that required changes to the DNG specification in order to be used properly. Now it's been updated to handle them, but some errors continue and it still can't read linear RGB encoded sensor data at all. This is not an outstanding implementation of DNG file compatibility, IMO.
I don't doubt that there might be some political background to these issues. But that's irrelevant to me as a user.
These sorts of inconsistencies in the way Aperture handles the processing of raw files ... never mind that we disagree on UI design ... as well as the always slow and "when we get around to it and if we think it's important enough" attitude towards supporting new camera models, are at the heart of why I have not yet considered Aperture as a serious competitor to Lightroom with regard to raw image processing.