[BrianR]

The long term survival of dng, or tiff, or any other format doesn't seem terribly important. It's not like overnight all programs will lose the ability to read dng. If it is replaced, expect it to be slowly over time and with batch converters to change over your existing dngs to something readable by whatever the new software at the time is. Count on your user error causing any future loss or if it happens long after you've passed on, that no one alive at the time cared to archive your stuff, probably because weren't interested in even looking at it in the first place. Consider the people who have 5.25" floppy disks and no hardware to read them... they screwed up during the transition periods (which isn't even over yet).

[newmikey]

The long term survival of dng, or tiff, or any other format doesn't seem terribly important. It's not like overnight all programs will lose the ability to read dng. If it is replaced, expect it to be slowly over time and with batch converters to change over your existing dngs to something readable by whatever the new software at the time is.

The long term survival of PEF, or tiff, or any other format doesn't seem terribly important. It's not like overnight all programs will lose the ability to read PEF. If it is replaced, expect it to be slowly over time and with batch converters to change over your existing PEFs to something readable (DNG) by whatever the new software at the time is.

[c.a.m]

5.25" floppy disks and no hardware to read them...

Ha, I transcribed all of my 5.25-inch floppies to 3.5-inch technology. I'm future-proof!

Seriously, though, you're right. A transition would take time, and we should be well prepared to move our files to other formats.

On the question regarding DNGs or PEFs, I prefer PEF format, because Photo Ninja - my RAW developer of choice - sometimes has trouble indentifying the lens types. Something about an extra tag or two in the DNG files from Pentax cameras. Anyways, the PEF files work fine.

- Craig

[BrianR]

The long term survival of PEF, or tiff, or any other format doesn't seem terribly important. It's not like overnight all programs will lose the ability to read PEF. If it is replaced, expect it to be slowly over time and with batch converters to change over your existing PEFs to something readable (DNG) by whatever the new software at the time is.

The long term survival of *insert format here* doesn't seem terribly important. It's not like overnight all programs will lose the ability to read *insert format here*. If it is replaced, expect it to be slowly over time and with batch converters to change over your existing *insert format here*s to something readable (*insert shiny new format here*) by whatever the new software at the time is.


(I think this is a pretty safe statement for anything so widespread as the standard imaging formats)

---------- Post added 08-31-15 at 03:58 PM ----------

Ha, I transcribed all of my 5.25-inch floppies to 3.5-inch technology. I'm future-proof!

Absolutely, so did I. I skipped moving the 3.5" to CD or DVD, but years before I ditched my 3.5" drive I had gone through all my floppies and transferred any of the ones I cared about onto a hard drive. At that time, a small HD was capable of storing the data of roughly a bajillion 3.5" floppies.

[stevebrot]

I would think that if Adobe had no dog in that hunt they might have no problem letting the standard be in the public domain as it pretty much is now anyway.

The DNG specification is proprietary (not open), but published. All versions of the spec are available in full on the Adobe Web site. Although standardization has been suggested, the DNG specification has not been formally taken up by any standards body. As such, DNG's features and direction reflect Adobe's vision and needs in regards to product integration* and such with no real community involvement. That would change should it ever become an independent standard administered by an entity such as ISO.

A DNG standard would be very cool, though broader industry buy-in and participation would likely be required before ISO or another standards body would be interested.


Steve

* The "import as" and "export as" DNG features of Lightroom are a good indication of Adobe's intent is to promote DNG as THE medium for image data exchange between Adobe products. Cataloged DNG files may then be handled by multiple apps through common parsing components with full non-destructive and transparent input from all apps.

[stevebrot]

AFAIK we're talking about container formats (like mkv for video) where the content is essentially the same, but packed in different wrapping. So shouldn't the answer be 'It doesn't matter', because it's trivial to shift files from one to the other? Also regarding convenience of reading them,

I believe you are correct. Both DNG and PEF files contain TIFF/EP*-compliant documents and any valid TIFF/EP parser can handle both formats**. The process of conversion from PEF to DNG involves mapping the content of compatible "tags" in the PEF to corresponding "tags" in the new DNG file. It is my understanding that compatibility between the two is currently very high meaning that an in-camera DNG is essentially equivalent to an in-camera PEF. The same may or may not be true for PEF converted to DNG by an Adobe product.


Steve

* TIFF/EP is not the TIFF image format. They are related, but not the same.

** TIFF/EP forms the basis for many RAW formats, the most prominent of which (after PEF ) is probably Nikon NEF.

[stevebrot]

Or what is the Native Digital Format????

Bits as content in a TIFF/EP tagged field.


Steve

[stevebrot]

DNG is Adobe's open standard

DNG is a specification, not a standard. Adobe has published the specification and encouraged camera makers and software vendors to support it, but that is only the first step towards becoming a standard. Standardization is driven by broad adoption and active community involvement by all stakeholders.

The lack of such could not have been more clear than the matter of the pixel-shift DNG files from the K-3II. Ricoh bent the DNG rules with the result that those files were (and continue to be) non-compliant to the DNG specification. They failed to import as a multi-image files into Adobe products as a result. Adobe has since issued a patch to remedy and with any luck the next version of the DNG spec will support the Ricoh-style data structure. Thumbs up to Adobe for responding. Thumbs down to Ricoh for pressing the matter.


Steve

[stevebrot]

Something about an extra tag or two in the DNG files from Pentax cameras.

Yes, the Pentax metadata has ambiguities...


Steve

[wildman]

That would change should it ever become an independent standard administered by an entity such as ISO.

though broader industry buy-in and participation would likely be required before ISO or another standards body would be interested.

Yep. There really should be a default RAW standard on all serious cameras which wouldn't necessarily rule out the possibility of a proprietary standard if a company felt they had something particularly unique to offer rather like Pentax and the DNG/PEF thing. I don't mind proprietary so long as as it's use is not obligatory.

[MadMathMind]

Yep. There really should be a default RAW standard on all serious cameras which wouldn't necessarily rule out the possibility of a proprietary standard if a company felt they had something particularly unique to offer rather like Pentax and the DNG/PEF thing. I don't mind proprietary so long as as it's use is not obligatory.

Honestly, it really doesn't matter because RAW is useless without the camera manufacturer to provide enough info for reconstruction. If the camera maker goes out of business and support for RAW conversion drops, you won't be able to much with a DNG, although it will at least technically open. But I don't think that's worth much.

In that respect, RAW is always proprietary regardless of what container it's in.

[wildman]

RAW is useless without the camera manufacturer to provide enough info for reconstruction.

Of course that's precisely what a independent third party RAW standard would avoid.

Standard - Conforming to or constituting a standard of measurement or value; or of the usual or regularized or accepted kind
A RAW standard is independent of the hardware by definition - the manufacturer simply makes camera bodes that write files that conform to this independent standard.

Unless I'm totally misunderstanding your post.

[newmikey]

Of course that's precisely what a independent third party RAW standard would avoid.

Standard - Conforming to or constituting a standard of measurement or value; or of the usual or regularized or accepted kind
A RAW standard is independent of the hardware by definition - the manufacturer simply makes camera bodes that write files that conform to this independent standard.

Unless I'm totally misunderstanding your post.

A "RAW standard" is quite a meaningless terminology seeing that future cameras will continue to have added functionality not even contemplated today. One cannot devise a "standard" based on a moving target. This is what the K-3 II case has now proven to us. Best you can expect is for DNG to cover a few generations of model updates which are similar except for MP count f.i.

A future camera might have a need to store data not yet even imaginable to us now such as multi-angle images, Lytro-type light-field information, some newer type of GPS stabilization, newer EXIF fields etc. etc. (multiple images with pixel-shift would be a nice example of course which we could not have thought about 10 years ago). Who knows, potentially even the TIFF/EP standard of the raw sensor data storage inside the DNG and many other formats such as PEF, NEF etc. may be challenged as insufficient. In that, DNG will be in exactly the same position as proprietary formats.

DNG is a great solution but in its lack of future-proofing, it is no better or worse than PEF (or NEF, CR2, ORF etc. for that matter). It works out very well if you have different bodies which can all produce DNG raw files, streamlining your raw workflow if your software cannot handle a mixture of formats in one go, it is a great temporary solution when the camera proprietary format is not yet interpreted by one's favorite bit of software (until the next update, that is). It's also great for exchanging raw files with others who lack the software required (if you would ever want to do so).

First and foremost, to me at least, it seems like a solution without a problem. If Adobe would drop support for proprietary raw formats in ACR tomorrow, would that be acceptable to anyone? If it is not, why worry? If your PEF is supported by today's software, chances are it will be by all future updates to that software as well.

DNG can be handy once in a while and you can convert your PEF to DNG at any given time when needed. Adobe guarantees the conversion does not cause any data- or quality-loss, don't they?

[acoufap]

DNG is a great solution but in its lack of future-proofing, ...

In software development there are a lot of standards and the targets are always moving. No problem because the basic standards are permanently developed accordingly. This takes time but it works. Just look at the HTML standard how it has changed the last years. So what is needed is a process that guarantees that DNG will be accepted as a standard and the specification should be developed further by an open committee according to the moving target "camera technology".

The situation with DNG to me seems a bit like where Pentax once was with the K-Mount bayonet specification. There were some manufacturers including Agfa Selectronic, Alpa, Chinon, Cosina, Petri, Ricoh and Vivitar (most if not all manufactured cameras) that Pentax could win for the K-mount. The most important ones Nikon, Canon, Minolta, Contax, ... didn't join. When introducing the automatic mode Ricoh included the famous ricoh pin. That was out of scope of the specification. Like pixel-shift for DNG today. Maybe a standardization process hasn't been established for the K-mount. So this happened. But I don't know enough of the history so I better close up this statement. It was just an idea of parallelism.

The lack of such could not have been more clear than the matter of the pixel-shift DNG files from the K-3II. Ricoh bent the DNG rules with the result that those files were (and continue to be) non-compliant to the DNG specification.

The question I would pose is why did Ricoh/Pentax not support generating one computed raw representation by pixel shift as they do with multi exposure mode. If they would, nobody would have a problem to process this pixel-shift file with other software than PDCU. Additionally Ricoh/Pentax could have offered file options for the multiple raw representations within one DNG file (the existing one) and storing each representation in its own DNG file.

[stevebrot]

The question I would pose is why did Ricoh/Pentax not support generating one computed raw representation by pixel shift as they do with multi exposure mode.

I guess because they chose to support it in the same manner as in-camera HDR.


Steve