Short answer: To maintain control
Sort of shocking that the verge even thought to ask that…
It would be nice if the major camera companies, or at least a few of them, got together and adopted a fully open format(DNG is not open). Sort of like synth companies did the in the 80s with MIDI. It would only need a few for the others to at least feel some pressure and adopt it.
It’s a major disgrace for humanity that we have closed formats. Both for current consumers and for future archival.
If you want to understand the kind of “logic” that seems to be behind such decisions, I recommend reading this:
It’s completely nonsensical, but there you go.
I don’t know about other cameras but the common excuse that only the manufacturer knows how to correctly convert the sensor output to RGB has to be true for Sigma/Foveon.
Here’s the metadata (not Exif) for an SD9:
IMG00169.X3F.txt (16.6 KB)
Have a look at “ImagerSettings” for example and compare those to the DNG spec.
Reverse-engineering those is almost impossible. D. Crow’s command-line raw converter was quite a poor but welcome approximation and the subsequent development LibRaw was better but still not equal to Sigma Photo Pro, even today.
Sigma’s adoption of Adobe’s approximation format DNG in later years brought joy to many but only to those who “just have” to use LR … and BTW who likes 150 MB image files anyway?! That’s 3X bigger than current Sigma/Foveon X3F files which are big enough in the first place!!
In any case, my computer has been Adobe-free for over a decade, grump.
P.S. I have a vague recollection that one well-known manufacturer dropped their DNG option (Hasselblad?).
And curiously my Lumix LX-1 has a TIFF option which I rarely use …
I understand that the camera manufacturers don’t want to shackle themselves to Adobe’s format. But I don’t get why they’re halfheartedly trying to keep their own format a secret. Are they intending people to not be able to edit their files? Why even write those files then?
In any sane world, I’d have thought that camera manufacturers would meticulously document their file formats, not hide them behind NDAs and such. But at least there doesn’t seem to be any restriction to reverse-engineering the formats.
And besides, who said we’re living in a sane world.
I’ve been wondering this as well. Why intentionally restrict support of your cameras in third party software? They sell hardware, not software, so the more widespread the support, the better, I’d think.
Only explanation I could come up with is, that they get paid by companies like Adobe for disclosing the secrets.
I suspect it’s more that they get used to hiding their hardware behind NDAs to avoid industrial espionage and it just ends up being a habit that pervades their entire operation (i.e. nobody ever says “this NDA is a step too far”, or at least there’s never a good commercial reason for changing their practices).
Apart from the RAW formats or DNG, there are already very good replacements for the old 8-bit JPEG format. In my opinion, it would be a great advantage and desirable to replace the cameras normal JPEG with the output of JPEG-XL. From the outset, if necessary, there would be much more picture information to start and play with it. This would also be a great further development for ever newer monitors with 10-bit and HDR.
In any case, you do not need the full range of data of an RAW. But if you wanted and need it, there is a way to still use it.
Please explain which parts of a raw data range are not needed.
Especially in the case of my Sigma SD9 camera which only outputs raw data full size and JPEG thumbnails.
There was a missing “not” or one at the wrong place in my post. Sorry Regarding me, I don’t need RAW data or files in any case or very often. I would have the JPEG from the camera. Sometimes it would be good to have some more image information available for post-processing. I would also find it very practical in further workflow and processing, if a JPEG-XL from the camera were available.
Frankly, if the camera did the lens correction, demosaicing, and transformation to a standard color space, in a linear format preserving bit depth, 99% of the time I would just take it from there and not bother with the raw.
No problem, thanks for the explanation.