Bibble is a respected raw converter, developed by Bibble Labs Inc, with some unique features and many enthusiastic users. It is one of the few significant raw converters (perhaps the only one) that can’t process DNG files if Bibble doesn’t have details for the camera built-in. Given the undoubted expertise of the development team, this omission is caused by ideology, not incompetence. It almost certainly reduces their market.
I’ll use quotes from a recent thread on their support forum to illustrate misconceptions that many people still have with DNG. Although these quotes do not come from the Bibble team, my discussions with the team over a long period give me confidence that they would largely agree with these quotes. I am not popular in their forum, so I’m responding here! People from the forum are welcome to comment here (as long as they are not abusive).
Pretty well everyone in the Bibble forums who asks for full DNG support has a reasonably accurate understanding of what can be done with DNG.
Second of all, Bibble should just get over its ideological objections to the format, and just add support for converted DNG. This should be a top priority. Although I don’t use DNG for photos I took myself, I do occasionally get files from other photographers. Some of those are in DNG format. Not being able to collaborate or share photos with people who use DNG is a complete showstopper for me.
Precisely! Bibble supports the most popular cameras very quickly. (It currently supports a little over 200 cameras). But photographers will typically want their raw converter to handle all the images they have to handle, and that may include some from cameras that are not supported, or some from other people. (Officially, the number of cameras that can be converted to DNG is nearly 300, but unofficially there are up to 30 more).
Being able to convert the new camera’s files to DNG means that you’ve got support from your “favourite” software until it releases a version that supports the camera: and the option to keep your money in your pocket if the new software release that officially supports your camera isn’t a free-of-charge “point” release.
Perhaps Bibble see this as a threat. Adobe have lived with this for a long time; right from the start, they knew that photographers could stick with Photoshop CS & ACR 2.4, yet support cameras released many years after that software was released. Adobe don’t use “new camera support” as a way of forcing photographers to upgrade to new releases of Photoshop or Lightroom, (although many uninformed photographers believe otherwise).
How about old camera, new software, that upgrade you want to pay for or qualify for free, but has dropped support for your camera…. Remember B5 dropped support for a lot of older cameras, which was very frustrating for some – ask the D1X users. The lucky ones got support a over a year later – like the D1X – others are still without, with no prospects of ever being supported (Fuji users). I have no doubt that B6 when it comes will drop a load more old cameras, and people who will be itching to hand over the money to get their hands on all the cool new features will hold off because their camera isn’t supported. Supporting converted DNGs gets round that and keeps the clients in the fold, otherwise you might as well put a link to the LR site in the FAQ page under the question ‘My camera is not longer supported, what do I do?’
This is the classic OpenRAW fear; long term loss of processability of older images. And it is a problem for which DNG is the only known solution, which is why archivists and conservations are moving towards DNG as their way of preserving raw images for future use.
I’ve said this before: DNG support can be as easy or as hard as software developers choose to make it.
Yes! It needs a can-do rather than a can’t-do attitude. Very many raw converters where the development team have a can-do attitude have provided this level of DNG support for years. Bibble is almost unique in having such a persistent can’t-do attitude.
Perhaps the trick is to have a first release that achieves most of the requirements reasonably cheaply, then to refine the capability in later releases.
The arguments against full DNG support are like a fart in a lift. They are wrong on many levels. Let’s try to tease them apart.
But for a raw converter to get the max out of any raw file independent of format it needs to be profiled for that converter. Because of the normalization process of the Adobe DNG converter, profiles generated for the native raw file will not deliver identical results when used on a converted DNG.
How many errors can there be in such a simple statement? Let’s see:
- DNG files contain camera details, corresponding to a “profile”. So the raw converter can extract the profile from the DNG file, and doesn’t need to be profiled for that camera.
- There is no such “normalization process of the Adobe DNG converter” that would disrupt raw conversion. The raw image data is copied across from the original file from the camera to the DNG file. A DNG file is a genuine raw image file! (Even the Makernote from the original raw file is copied across to the DNGPrivateData field in the DNG file in a documented way).
- There is no evidence here that identical results won’t be delivered from the converted DNG file. After all, the raw image data is the same. But anyway, so what? If Bibble won’t handle a DNG because Bibble hasn’t been profiled for a particular camera, that is an unsupported camera! This is the difference between supporting a DNG image, but possibly not in the way that Bibble might hypothetically support it if it ever gets round to it, and not supporting that camera at all! Which makes more sense to buyers?
Proper DNG support would enable extra cameras to be supported, yet there is an argument that Bibble can’t afford to do it. Why could those other raw converters do it? Won’t the extra flexibility save effort rather than cost effort, or attract extra buyers and so pay for the extra work?
So you want a small organization that is already resource constraint to spend time on re-profiling everything? What a waste. There are more important things to be dealt with.
As demonstrated above, there is no need to re-profile everything. Instead, the need is for a one-off development to extract camera details from a DNG file and turn them into the equivalent of a Bibble profile, as closely as possible. Each of the camera details has to treated on its own merits. CFAPattern is DNG’s way of identifying the sequence of Reds, Greens, and Blues, in the sensor’s colour filter array. Obviously Bibble must have such data in its “profile”, and the conversion is likely to be easy. But the ColorMatrix1, which is part of the colour profile, may be significantly different from how Bibble stores the data, and is probably the hardest thing to convert. In other words, Bibble’s colours using data from the DNG file are likely to be somewhat different from what they would be using its own profile. The user may need to do different colour modifications.
But do users always use Bibble’s colours without modification? Often the raw converter’s colours when first opening a file are just the starting point; after all, Adobe products always start with the DNG values, and they survive in the marketplace! If the alternative is not being able to handle the image at all, what would users prefer?
Honestly, i never understood the point of converted DNGs (native DNGs make sense for me) as they’re: … based on reverse engineering (i doubt that canonikon & co gave adobe the specs of their raw formats); potentially lossy (maker notes anyone?) if someone figures out how to decode a up to now unknown tag tomorrow, it can be applied to the original raws, the already converted DNGs will not have that possibility.
Bibble’s profiles are also based on reverse engineering! And the Makernote is copied across to the DNGPrivateData field in the DNG file in a documented way.
I understand the idea that DNG seems like a common format that is like a JPG or TIFF file but it isnt. If this was the case then I’m sure that BibbleLabs would do exactly as has been requested and provide DNG support. However DNG interpretation would require additional work for every camera. Its just that Adobe has built this already into their products that makes it seem like it is a good idea. Adobe wont release this code as they say it is proprietary as it is part of Photoshop. In reality for every camera CCD there is a DNG equivalent so 1 RAW = 1DNG.
Everything in that statement is demonstrably wrong! (And it is unwise to be “sure that BibbleLabs would do exactly as has been requested and provide DNG support”. They are driven here by ideology, not technology).
- Obviously there isn’t extra work for every camera! All those other raw converters that can handle raw images from cameras that were launched after the software was released don’t need extra work for every camera! (That list is years old, and covers just the subset I tested. There are lots more than that).
- Adobe hasn’t built this [presumably extra work for each camera] into their products. ACR 2.4, for Photoshop CS, was released in January 2005, yet supports more than 200 camera models launched since that date via DNG. No one needed to go back and change that software for those cameras.
- The fact that Adobe won’t release its code is irrelevant; does the Bibble team really need code from Adobe? Can’t they write high-quality code themselves? Of course they can!
- The reason why the statement “1 RAW = 1DNG” is wrong is that DNG has lots of metadata that enables a single DNG format to currently handle about 300 different camera models, and lots more in future. It is unique in this respect; no other raw file format does this, nor will they.
This re-profiling nonsense is a recurring theme. It has been present in Bibble forums for years, and perhaps the people concerned haven’t looked outside the forums to update themselves for years?
No, DNG won’t help. The big work is profiling the camera. Decoding the RAW file is easy when it’s already done once (aka the RAW file format is deciphered). If reprofiling is needed, the camera is sometimes needed to take new pictures.
DNG is increasingly being used for long-term archiving because it doesn’t need such re-profiling later. After all, there may not be a working Nikon D70 in 50 years time for the purpose. The DNG file contains what is necessary within it, and the future raw converter then needs to follow the specification and use that data.
The facts, as demonstrated by the DNG specification and by the experiences of lots of other raw converters, are incompatible with the anti-DNG views that have persisted in the Bibble forums for years. It is valid for the development team to say “yes it can be done but it isn’t the best use of our time at the moment”. It is unfair to customers to pretend that it can’t be done, and that re-profiling is necessary.
Here are articles, typically written by professional archivists and conservationists working for respectable organisations, who suggest or recommend DNG for archiving purposes, where future software can’t be relied on to have camera profiles built in, and there may not be working models of those cameras available for profiling purposes:
- Planning for US Library of Congress Collections: Sustainability of Digital Formats – Still Image – Preferences in Summary
- American Society of Media Photographers – dpBestflow: Raw File Formats
- universal photographic digital imaging guidelines (UPDIG): File formats – the raw file issue
- Archaeology Data Service / Digital Antiquity: Guides to Good Practice – Section 3 Archiving Raster Images – File Formats
- University of Connecticut: “Raw as Archival Still Image Format: A Consideration” by Michael J. Bennett and F. Barry Wheeler
- Inter-University Consortium for Political and Social Research: Obsolescence – File Formats and Software
- JISC Digital Media – Still Images: Choosing a File Format for Digital Still Images – File formats for master archive
- International Digital Enterprise Alliance, Digital Image Submission Criteria (DISC): Guidelines & Specifications 2007 (PDF)
- The J. Paul Getty Museum – Department of Photographs: Rapid Capture Backlog Project – Presentation
- American Institute for Conservation – Electronic Media Group: Digital Image File Formats
- Archives Association of British Columbia: Born Digital Photographs: Acquisition and Preservation Strategies (Rosaleen Hill)