Luminous Landscape Forum
Raw & Post Processing, Printing => Digital Image Processing => Topic started by: eronald on February 23, 2007, 01:14:53 pm
-
My article "Adobe Lightroom, a lucrative surrender" is up on my blog.
http://monitor-calibration.net (http://monitor-calibration.net)
Enjoy.
Edmund
-
I'd say more of a tactical retreat than surrender. I think DNG still has some legs.
-
I'd say more of a tactical retreat than surrender. I think DNG still has some legs.
[a href=\"index.php?act=findpost&pid=102639\"][{POST_SNAPBACK}][/a]
A few legs - others like Hasselblad have been amputated.
At the rate digital SLRs are multiplying DNG should be a centipede by now
Edmund
-
I think wide use of DNG or some version of a standard RAW file structure is inevitable. But not tomorrow.
I don’t know enough about RAW file structure to know if there’s a real performance advantage (smaller file, bit depth.. ?) to keep RAW file format proprietary. But I think sooner or later whatever advantage there is will be lost in the noise. Literally.
-
At the moment the camera manufacturers have a lot of freedom to innovate, and improve imagery inside their own proprietary Raw processing programs. When a single processing method becomes standard they will lose this freedom, which they see as a competitive advantage. A unique format is inevitable as you say, but only when the industry matures.
DNG was an attempt to take over the world by creating the new JPEG. I guess Adobe owed it to their shareholders to try, but they failed.
Lightroom is a very nice program and users will ove it, but its nicest feature is precisely the fact that it works with zillions of different camera manufacturers' files.
Edmund
I think wide use of DNG or some version of a standard RAW file structure is inevitable. But not tomorrow.
I don’t know enough about RAW file structure to know if there’s a real performance advantage (smaller file, bit depth.. ?) to keep RAW file format proprietary. But I think sooner or later whatever advantage there is will be lost in the noise. Literally.
[a href=\"index.php?act=findpost&pid=102688\"][{POST_SNAPBACK}][/a]
-
A new Jpeg? That comparison really isn't valid on any level. They are two incredibly different formats and both are very different from a licensing standpoint.
That aside, you say DNG is a failure in your blog but you fail to qualify that statement in any way. The statement is made in the first couple paragraphs but you don't support the argument further. How is DNG a failure specifically? If anything fact shows the opposite is true as cameras from smaller companies (other than Canon/Nikon) are slowly starting to support it as does all 3rd-party raw developing programs available (including lightroom).
Can you qualify your argument in any way or is this simply a move to bring some more traffic to your blog?
[span style=\'font-size:8pt;line-height:100%\']*Edited: re-worded slightly for less-cafinated clarity :) *[/span]
-
I like the way you complain about Adobe PR's unwillingness to supply some random blogger who gets absolutely no traffic and cannot even figure out how to configure his site template correctly with free copies of their software when a 30-day trial version is freely downloadable. You're really stickin' it to the man every time you spam links to your informationless non-site on every forum and mailing list you can find there, champ.
-
Edmund,
With all due respect, your blog piece about Lightroom and DNG is unsubstantiated nonsense, and does you no justice.
If you have any familiarity with the program at all you would note that DNG is the lingua franca of Lightroom. Want to edit a file and transfer it to another editing program as a raw, but retaining all metadata edits? Just export it from Lightroom as a DNG.
Yes you can also do this using proprietary raw formats and XMP sidecars, but the ability of DNG to retain both the complete raw data as well as editing metatdata betweern applications is unique.
Michael
-
Edmund,
With all due respect, your blog piece about Lightroom and DNG is unsubstantiated nonsense, and does you no justice.
If you have any familiarity with the program at all you would note that DNG is the lingua franca of Lightroom. Want to edit a file and transfer it to another editing program as a raw, but retaining all metadata edits? Just export it from Lightroom as a DNG.
Yes you can also do this using proprietary raw formats and XMP sidecars, but the ability of DNG to retain both the complete raw data as well as editing metatdata betweern applications is unique.
Michael
[a href=\"index.php?act=findpost&pid=102945\"][{POST_SNAPBACK}][/a]
Michael,
Let me thank you again for hosting this incredible forum where people bring differing opinions, but debate with astonishing courtesy and a total lack of personal attacks.
In the past, I have received some very angry hate mail about my position concerning DNG, and I am thankful to see someone disagreeing with me so politely.
As you are aware, my opinions are the result of a certain maturing. Before writing my pieces, I did some technical research, and took the pulse of the industry:
I have looked through the DNG specs when they were published, and I have played with dcraw, which is source code incorporated in many programs that input Raw files.
I then briefly discussed some of the technical issues involved in the Adobe Raw conversion model with Thomas Knoll, of Photoshop, and also talked at considerable length about some technical stuff with Christian Poulsen, the hands-on CEO of Hassleblad, who opted out of DNG.
Thank you for letting me take what you say under careful consideration, and think about it before I publish my forthcoming analytical piece about Lightroom. As you know I like to market my writing as that of a controversialist, and like to defend positions which appear at first glance quite absurd.
Today's image (Lion's Mane Medusa) is remarkable and certainly does you credit as a photographer. Unfortunately I do not have your skills with a camera.
Edmund
PS. At no point do I say that DNG is bad. I'm just saying that when it came out it was presented to the public, by Adobe, as a "universal" Raw format, but it hasn't caught on with the camera makers. I'm sure it does meet Adobe's own needs for Lightroom perfectly, as if it had been designed for them And I do say that Adobe has listened to its customers when building Lightroom, and that I believe it will be a bestseller !
-
Edmond,
Wow, that sounds impressive. However you still don't substantiate your argument nor do you provide any actual information.
My advice to you would be to correct your article and provide something that reflects the research and thought you claim to have put into this. And do it quick. Otherwise you'll soon be brushed off as a sensationalist simply looking to stir up the pot for the sake of gaining traffic to your blog and no one will take you seriously.
-
Edmond,
Wow, that sounds impressive. However you still don't substantiate your argument nor do you provide any actual information.
[{POST_SNAPBACK}][/a] (http://index.php?act=findpost&pid=103014\")
Let me be clearer: Adobe launched DNG several years ago as a universal camera format, shortly after Thomas Knoll wrote the original ACR, and they were feeling the pain of supporting a new Raw format every time a new camera came out.
In fact Adobe was not even certain that they had a legal right to read all Raw files. There was a big controversy because Nikon would not play ball and encrypted their white balance data so ACR could not employ it legally without violating the DMCA.
[a href=\"http://news.com.com/Nikons+photo+encryption+reported+broken/2100-1030_3-5679848.html]http://news.com.com/Nikons+photo+encryptio..._3-5679848.html[/url]
This was amicably settled. To date, more than half the cameras on the market that write Raw files are made by Canon and Nikon together -care to dispute that ?- and do not use DNG as a native format. Thus DNG has failed to become a success as a camera format.
Adobe now markets an excellent piece of software, Lightroom, that needs to read about 130 formats rather than just one. They will make good money on that but their attempt at unification of the Raw format has failed decisively and they will have to spend engineering hours on supporting every new camera as it comes out. And they remain hostage to a future Raw file encryption by some manufacturer.
Edmund
PS. I'm not against an open file format, in fact I'm desperately for it. But I don't think that DNG is what the camera world needs, and I think that's the real reason they didn't adopt it. If it had genuine technical advantages beyond what people are already doing, it would have made more headway.
-
Edmund,
How about putting some of that research into the article? As it stands, it really is an assertion rather than anything more weighty. You don't even mention evidence that might be in favour of this assertion, such as LR being unable to make full size preview DNGs or update previews.
LR signals nothing like what you pretend - its handling of multiple raw file formats merely acknowledges dangerous competition in that market space and has nothing to do with DNG.
John
-
Edmund,
How about putting some of that research into the article? As it stands, it really is an assertion rather than anything more weighty. You don't even mention evidence that might be in favour of this assertion, such as LR being unable to make full size preview DNGs or update previews.
LR signals nothing like what you pretend - its handling of multiple raw file formats merely acknowledges dangerous competition in that market space and has nothing to do with DNG.
John
[a href=\"index.php?act=findpost&pid=103022\"][{POST_SNAPBACK}][/a]
John,
Industry insiders will have got the point now: Format wars happens because the winner gets a lion's share of the loot. Sony made a lot of money selling VCRs but they would have made much more if their own design, Betamax, had won. They still weep over that loss.
In the same way, Adobe will be very succesful with Lightroom and its successors, but they would have owned the consumer Raw market if DNG had been adopted by the major players in the SLR market .
Although the DNG format itself is open, it's not obvious to all how to write a Raw decoder for all the variants. As the originator and keeper of the DNG format, Adobe could have licensed its own and very good DNG conversion technology everywhere Jpeg is now found - high-street minilabs, DVD recorders , web browsers, portable phones , even iPods. And maybe in the end even cameras.
Adobe failed. The major SLR vendors refused any form of standardisation. Microsoft may in the future be in the position to enforce a common format, but then it will be Microsoft's format and not Adobe's format.
Edmund
-
John,
Industry insiders will have got the point now: Format wars happens because the winner gets a lion's share of the loot. Sony made a lot of money selling VCRs but they would have made much more if their own design, Betamax, had won. They still weep over that loss.
In the same way, Adobe will be very succesful with Lightroom and its successors, but they would have owned the consumer Raw market if DNG had been adopted by the major players in the SLR market - Adobe could have licensed DNG conversion technology everywhere Jpeg is now found - high-street minilabs, DVD recorders , web browsers, and maybe even iPods.
[a href=\"index.php?act=findpost&pid=103030\"][{POST_SNAPBACK}][/a]
[span style=\'font-size:8pt;line-height:100%\']emphasis added[/span]
What?? DNG in web browsers and DVD recorders? iPods? (yes, because a linear raw format would be so beneficial for my AC/DC album art. Ha, ha!) Do you even know what a DNG is?
Implying DNG could be licenced as a replacement for Jpeg shows how absurd and ill-informed your stance is. DNG is a open raw format and does not require licensing.
-
[span style=\'font-size:8pt;line-height:100%\']emphasis added[/span]
What?? DNG in web browsers and DVD recorders? iPods? (yes, because a linear raw format would be so beneficial for my AC/DC album art. Ha, ha!) Do you even know what a DNG is?
Implying DNG could be licenced as a replacement for Jpeg shows how absurd and ill-informed your stance is. DNG is a open raw format and does not require licensing.
[a href=\"index.php?act=findpost&pid=103038\"][{POST_SNAPBACK}][/a]
Daniel -
I can understand you disagree with me, but at least quote me exactly. I said that Adobe could have licensed not the DNG format, but *DNG conversion technology*. They have invented DNG, they have beautiful converters that allow viewing of many variants of DNG files, which many of the best engineers in the industry spent years in writing. This is very valuable intellectual property, even if the format itself is open.
If i may be allowed an analogy, as you are obviously a technical person, the C++ language is standardised, and anybody can write a compiler. However a good optimising C++ compiler is still a valuable piece of intellectual property, even though totally open and free C compilers do exist - like gcc.
Edmund
-
For those requesting research to back up his claims, the following highly scientific process can be seen at work behind every Edmund Ronald blog posting:
Step 1. Read a discussion or public press release from a large corporation posted to the Apple ColorSync mailing list.
Step 2. Summarize the handful of points you vaguely understood in three sentences or less with plenty of name-dropping in order to make yourself sound authoritative by association.
Step 3. Post a message to the Apple ColorSync mailing list in order to inform everyone there that you've just regurgitated all of the information that had been posted previously on a website with your name on it.
Step 4. Hold your breath and wait for attention.
Step 5. This is a paragraph of text that could go in the sidecar.
Step 6. This is a paragraph of text that could go in the sidecar.
Step 7. This is a paragraph of text that could go in the sidecar.
Step 8. This is a paragraph of text that could go in the sidecar.
Comparing the storage of bitmaps with some metadata attached to the compilation of object code is just plain ridiculous. You have the sensor data and you write it out. If there were any optimization beyond lossless compression taking place in the pipeline then it wouldn't be raw sensor data anymore. At what point in the interpretation of image data is there even any opportunity to do things like loop unrolling or register coloring that might possibly make this a valid analogy? The CMM is the only thing I can think of, and again, that is not an issue at the point of file creation.
-
Comparing the storage of bitmaps with some metadata attached to the compilation of object code is just plain ridiculous. You have the sensor data and you write it out. If there were any optimization beyond lossless compression taking place in the pipeline then it wouldn't be raw sensor data anymore. At what point in the interpretation of image data is there even any opportunity to do things like loop unrolling or register coloring that might possibly make this a valid analogy? The CMM is the only thing I can think of, and again, that is not an issue at the point of file creation.
[{POST_SNAPBACK}][/a] (http://index.php?act=findpost&pid=103076\")
As people are asking me for links to information I used in researching my little piece of total nonsense, here is one to Dave Coffin's dcraw source code (text). This is a very nice free (GPL) converter which does *very well indeed* on Raws if you can feed it a profile for the camera. The highlight recovery feature is extraordinary and well worth having in your toolkit.
Dcraw builds on considerable published work. The comments in the code allow one to see what problems are addressed in a Raw converter, thereby satisfying this poster's curiosity on the degree of difficulty involved.
[a href=\"http://cybercom.net/~dcoffin/dcraw/dcraw.c]http://cybercom.net/~dcoffin/dcraw/dcraw.c[/url]
And of course here is a link to Dave's main page
http://cybercom.net/~dcoffin/dcraw (http://cybercom.net/~dcoffin/dcraw)
Of course, dcraw is also an excellent testbed for testing improvements to Raw processing algorithms as it supplies all the I/O. Readers of this thread may find the "Other Photo Decoders" list instructive.
Edmund
-
I taken a look at dcraw, and the dng specification. Dcraw apparently can decode raw images in a generic way, but dng has an important advantage: It can describe the original raw data into it's own file format, in a very detailed way, including the sensor response, and all the data needed to decode the image as it's supposed to be seen, now and into the future.
Adobe's approach is the most correct, the raw format must be standarized in order to assure the correct opening of the file, now and forever. In this case, Adobe's experience managing several raw formats from all the different manufacturers has found (in my opinion), the only secure way to assure the future opening of the files. DNG is build on top of the TIFF standard, so the open community can make their own dng converter if they wish.
Doing a comparison between raw and a device driver is incorrect, because nobody cares if a printer can be user in 10 years, but everybody wants to be able to open their pictures forever.
I have seen OpenRaw's page, even filled the survey, but the reality is that i don't see any effort to make their own format. Are they only asking the manufacturers to publish their raw formats?
Manufacturers had always formed consortiums to define standards, i don't know what the camera manufacturers are waiting to do that. The dangerous proliferation of many incompatible raw formats are the manufacturer's responsibility, and they are showing almost unbelievable irresponsibility.
As people are asking me for links to information I used in researching my little piece of total nonsense, here is one to Dave Coffin's dcraw source code (text). This is a very nice free (GPL) converter which does *very well indeed* on Raws if you can feed it a profile for the camera. The highlight recovery feature is extraordinary and well worth having in your toolkit.
Dcraw builds on considerable published work. The comments in the code allow one to see what problems are addressed in a Raw converter, thereby satisfying this poster's curiosity on the degree of difficulty involved.
http://cybercom.net/~dcoffin/dcraw/dcraw.c (http://cybercom.net/~dcoffin/dcraw/dcraw.c)
And of course here is a link to Dave's main page
http://cybercom.net/~dcoffin/dcraw (http://cybercom.net/~dcoffin/dcraw)
Of course, dcraw is also an excellent testbed for testing improvements to Raw processing algorithms as it supplies all the I/O. Readers of this thread may find the "Other Photo Decoders" list instructive.
Edmund
[a href=\"index.php?act=findpost&pid=103106\"][{POST_SNAPBACK}][/a]
-
Norfindel,
Thank you for having looked at the references and participating in this debate in an informed fashion.
Rather than put all my arguments on this site, I have placed them on my blog. I assume you have read my further article entitled "Raw: The Adobe Way, the Microsoft Way and the Open Way (http://photofeedback.blogspot.com/2007/02/raw-adobe-way-microsoft-way-and-open.html)".
I totally agree with you that DNG is an intereresting proposal - we desperately need standard formats for Raw, and we need more such proposals for standardization.
But then you say that because of Adobe's experience DNG is "the only secure way to assure the future opening of the files". Here I must disagree - there are numerous alternative formats possible, and a standardization process involving the camera makers would have much more political chances of success than a proposal that is quite visibly being resisted by them.
In fact, I very much agree when you say that the camera makers should create a format standardization consortium. In which of course Adobe should participate. A combined "designed by committe" format might be as cumbersome as most products of collective design, but like with XML the ultimate benefits in terms of interoperability might exceed the kludge factor of the format.
And fully support your statement that " The dangerous proliferation of many incompatible raw formats are the manufacturer's responsibility, and they are showing almost unbelievable irresponsibility." This is very true: They are playing with our images !
As far as dcraw goes, it is just code rather than a spec. But in the absence of a common spec it is a viable interim solution. Dcraw is in fact being used by many software writers, commercial and otherwise to solve the Raw Maze problem now. I "prefer" dcraw to DNG because DNG is just paper, while dcraw is is actually usable by anyone.
Integrating dcraw in Windows and Photoshop as a plugin to each would however provide a simple way to be sure that there is a "standard" way to decode Raw formats available in the near future (10-15) years, and might also offer a reference implementation for a future draft standard for Raw formats. One of the main weaknesses of DNG is that it does not offer such a publicly available reference implementation.
Edmund
I taken a look at dcraw, and the dng specification. Dcraw apparently can decode raw images in a generic way, but dng has an important advantage: It can describe the original raw data into it's own file format, in a very detailed way, including the sensor response, and all the data needed to decode the image as it's supposed to be seen, now and into the future.
Adobe's approach is the most correct, the raw format must be standarized in order to assure the correct opening of the file, now and forever. In this case, Adobe's experience managing several raw formats from all the different manufacturers has found (in my opinion), the only secure way to assure the future opening of the files. DNG is build on top of the TIFF standard, so the open community can make their own dng converter if they wish.
Doing a comparison between raw and a device driver is incorrect, because nobody cares if a printer can be user in 10 years, but everybody wants to be able to open their pictures forever.
I have seen OpenRaw's page, even filled the survey, but the reality is that i don't see any effort to make their own format. Are they only asking the manufacturers to publish their raw formats?
Manufacturers had always formed consortiums to define standards, i don't know what the camera manufacturers are waiting to do that. The dangerous proliferation of many incompatible raw formats are the manufacturer's responsibility, and they are showing almost unbelievable irresponsibility.
[a href=\"index.php?act=findpost&pid=103976\"][{POST_SNAPBACK}][/a]
-
As people are asking me for links to information I used in researching my little piece of total nonsense, here is one to Dave Coffin's dcraw source code (text). This is a very nice free (GPL) converter which does *very well indeed* on Raws if you can feed it a profile for the camera. The highlight recovery feature is extraordinary and well worth having in your toolkit.
Dcraw builds on considerable published work. The comments in the code allow one to see what problems are addressed in a Raw converter, thereby satisfying this poster's curiosity on the degree of difficulty involved.
http://cybercom.net/~dcoffin/dcraw/dcraw.c (http://cybercom.net/~dcoffin/dcraw/dcraw.c)
And of course here is a link to Dave's main page
http://cybercom.net/~dcoffin/dcraw (http://cybercom.net/~dcoffin/dcraw)
Of course, dcraw is also an excellent testbed for testing improvements to Raw processing algorithms as it supplies all the I/O. Readers of this thread may find the "Other Photo Decoders" list instructive.
Edmund
[a href=\"index.php?act=findpost&pid=103106\"][{POST_SNAPBACK}][/a]
That's great, but once again, has absolutely nothing to do with anything, and I suspect you're simply posting links to C files in the hopes that you will dazzle us with your ability to seem more technically inclined than you actually are. I'm sure this sounds pompous, but I was writing C code to pull data off of camera sensors back in the days when VGA resolution grayscale models were in the tens of thousands of dollars, so I'm afraid you're going to have to do more than paste links to some code you had nothing to do with if you want to impress me with this "research." I can post links to complicated tax laws on the internet all day long, but it wouldn't make me a tax attorney.
The reason why there are so many variables and opportunities for innovation involved in raw file decoding is because there are so many variables in the raw data itself. One company might put one byte here, and another company might put it somewhere else entirely, leaving it up to a skilled engineer to do a lot of guesswork based on careful disassembly--which is what dcraw is: guesswork. Often these guesses end up being quite wrong, and that's when you start seeing problems like the little maze patterns Aperture used to draw in some files.
The entire point of DNG is to establish a common format with zero guesswork. You don't need to write 20 page long switch statements to deal with manufacturer-specific de-mosaicing algorithms or byte-ordering because the file describes its contents and their locations for you. Zero ambiguity means zero opportunity for innovation and zero opportunity for intellectual property suits. Nobody is going to patent a DNG converter because there's nothing for them to patent. Have you actually read the DNG specification? Did you understand anything in it?
Now if you're going to talk about post-processing features of the converter such as highlight recovery, then yes, those are all very nice and fancy and up to each vendor to perfect on their own, but they have absolutely nothing to do with the file format itself. They are creative adjustments applied to the data by the user and the underlying format of that data is essentially irrelevant as far as they're concerned. By the time you get to the point of making visual adjustments you already have a fully decoded bitmap.
The camera may also massage its own data (noise reduction etc) in firmware prior to writing out the file, but again, that has nothing to do with the file format, and the raw converter has no influence over this process. It must accept whatever the camera spits out as gospel regardless of what may or may not have been done to it before it got there.
A little information has always been a dangerous thing, but thanks to the internet, every sad attention whore with too much time on their hands on the planet can pass themselves off as an expert. The unfortunate side-effect to all of this begging and brown-nosing is that some people might actually see your work and remember what they saw. Having seen both your photos and your writing, I'd probably go back to applied mathematics and give up on this whole internet celebrity dream if I were in your shoes.
-
I apologize, publicly and basely to the nice people at Adobe - apparently there is a reference DNG decoder.
http://www.adobe.com/support/downloads/detail.jsp?ftpID=3350 (http://www.adobe.com/support/downloads/detail.jsp?ftpID=3350)
Edmund
-
Let's have a bit more civility and less name calling please.
Discussions should be about ideas. Not personalities.
Michael
-
A little information has always been a dangerous thing, but thanks to the internet, every sad attention whore with too much time on their hands on the planet can pass themselves off as an expert. The unfortunate side-effect to all of this begging and brown-nosing is that some people might actually see your work and remember what they saw. Having seen both your photos and your writing, I'd probably go back to applied mathematics and give up on this whole internet celebrity dream if I were in your shoes.
[a href=\"index.php?act=findpost&pid=104083\"][{POST_SNAPBACK}][/a]
Orange, I would take your comments more seriously if you couched them more politely. You are probably right about some issues, you are wrong, in my opinion about others, but if you had talked like this at any scientist's conference you would have got thrown out. If you aplogize for the personal insults above, I will address your comments. However direct personal insults do not encourage me to do so.
Edmund
-
HOT DAMN, orangekay! All I can say is OUCH! And to think I thought my Blahnik heels were a great source of pain. Talk about molar-chomping on a Ph.D. and spittin' it right out and then grinding your Doc Marten steel-toed-heel into it. Suddenly, being pushed into an on-coming NYC subway train looks like fun. Maybe it's time for the treadmill, dear.
Btw, I'm looking for a new Rep, can I interest you in joining my team?
-
To date, more than half the cameras on the market that write Raw files are made by Canon and Nikon together -care to dispute that ?- and do not use DNG as a native format. Thus DNG has failed to become a success as a camera format.
[a href=\"index.php?act=findpost&pid=103019\"][{POST_SNAPBACK}][/a]
Well, the figure I have heard would be closer to 75% combined market share for Canon and Nikon.
Considering the initial hopes of Adobe, it seems pretty obvious to me that DNG is a failure as we speak.
This has, by the way, nothing to do with the value of DNG. The idea is good, but it was not proposed by the right company. My opinion has always been that a standart raw format would only be doable if it were created by Canon and Nikon themselves. Anyone even remotely trying to put herself in their shoes would find this pretty obvious actually.
Aren't you gonna laugh at me if I come ring at your door and try to force you to start speaking Esperando instead of English?
Regards,
Bernard
-
Re my apology to Adobe, I have just tracked down the license of the Adobe DNG SDK, and it's very nice. It looks like Adobe are going the extra mile.
http://www.adobe.com/support/downloads/dng...k_eula_win.html (http://www.adobe.com/support/downloads/dng/dng_sdk_eula_win.html)
---snip
1. LICENSE GRANT
Software License. Subject to the restrictions below and other terms of this Agreement, Adobe hereby grants you a non-exclusive, worldwide, royalty free license to use, reproduce, prepare derivative works from, publicly display, publicly perform, distribute and sublicense the Software for any purpose.
Document License. Subject to the terms of this Agreement, Adobe hereby grants you a non-exclusive, worldwide, royalty free license to make a limited number of copies of the Documentation for your development purposes and to publicly display, publicly perform and distribute such copies. You may not modify the Documentation.
---snip
Edmund
-
Bernard,
Not only are you right politically, but there are also some truly technical reasons involved.
Simply put, when film reigned, cameras had completed their contribution to the image once the the shutter had clicked behind the lens. At that point the processing lab took over. Now cameras do some processing work themselves *after* the actual exposure has taken place.
Intensively post-processing and massaging the sensor data is now a way to gain competitve advantage for the camera corps - and they want to be able to do it outside the camera -in a sealed black box converter - without their competitors seeing what they're up to. This massaging needs to be done on ALL the data the camera has, which may include a lot of details about the specific lens mounted on the camera, and its settings when the shot was taken. Even ambient light sensing can count - color appearance models take ambient light into account.
Think of it as every company developing its own proprietary, finely tuned and thus more capable, version of Dxo.
Christian Poulsen of Hassleblad told me during Photokina that his company sees postprocessing Raws rather than "just" converting them as a competitive advantage that is being asserted with the H3D, and he wants to do so even more.
We can see Leica going there too in the M8 - lens detection by coding and then vignetting correction. Although the M8 outputs a DNG file, this is in fact is the product of much in camera processing , I believe. I think it's a sort of intermediary-format DNG rather than a real Raw.
From information that has come my way, I would think that Nikon and Canon are going down that road too.
As long as picture quality improves, photographers won't complain, but such "secret sauce" processing obviously goes against the idea of a one size-fits-all Raw format or processor.
It has been observed before that the audio/video industry goes through this repetitive 2-phase cycle of innovation and standardization. I guess we'll have to wait some time for the second phase of the current cycle.
Edmund
This has, by the way, nothing to do with the value of DNG. The idea is good, but it was not proposed by the right company. My opinion has always been that a standart raw format would only be doable if it were created by Canon and Nikon themselves. Anyone even remotely trying to put herself in their shoes would find this pretty obvious actually.
Bernard
[a href=\"index.php?act=findpost&pid=104215\"][{POST_SNAPBACK}][/a]
-
This has, by the way, nothing to do with the value of DNG. The idea is good, but it was not proposed by the right company.[a href=\"index.php?act=findpost&pid=104215\"][{POST_SNAPBACK}][/a]
There is no "right company" unless there's a not-for-profit organization who mandates these decisions of behalf of all users and players.
I thought Digital Negative (DNG) was brilliant the moment I first heard the term. It makes perfect sense! The real problem is cock fighting: the corporate suits who believe their company shouldn't make it easier for another company to dominate, even if there's no monetary gain and the single idea makes universal sense.
I'm sure the question has been raised more than once: "Why should we (Canon or Nikon) make it easier for Adobe?" One would think the answer to that question lies in the number of Canon/Nikon photographers who earn a living using Photoshop.
The way I see it, Canon and Nikon aren't going anywhere and I can't imagine my workflow without Photoshop. The sooner some corporate leader with balls steps forward and says, "This is good for the future of all photographers!," the better. Given my experience with men, I won't be holding my breath.
-
I don't think the execs at C and N are dumb.
Maybe they're really getting extra quality out of their postprocessing - in this age of software, having your own postprocessor is like having your own brand of glass that allows you to make special aberration-free lenses which your competitor cannot imitate.
Maybe the execs are
the corporate suits who believe their company shouldn't make it easier for another company to dominate, even if there's no monetary gain and the single idea makes universal sense.
[a href=\"index.php?act=findpost&pid=104273\"][{POST_SNAPBACK}][/a]
-
I don't think the execs at C and N are dumb.[a href=\"index.php?act=findpost&pid=104277\"][{POST_SNAPBACK}][/a]
For the record, I never claimed anyone was dumb. I just happen to think cockfights are for the birds.
-
DNG may not yet be the format of choice for camera manufacturers.
But, DNG does have one huge plus, which I make full use of:
It is open source.
Therefore, if at some stage in the future, both Canon (my camera of choice) and Adobe no longer existed, I could still access my images using a third party.
As a result, even though I process my RAW files as they emerge from my Canon, I save *everything* I shoot as a DNG and keep them in more than one location. DNG turns my files into "negatives" that I will be able to use in any "enlarger".
DNG gives me long term security. Nothing else does.
D.
-
DNG is not open source. For something to be open source it needs to be released under a OSI approved license (http://en.wikipedia.org/wiki/Open-source_license#OSI_approved_licenses). The most common is the GPL.
That is not to say DNG is not freely available. It is under a royalty free license (http://www.adobe.com/products/dng/license.html) authored by Adobe.
-
Strictly speaking, being open source is not enough, by itself, to guarantee the long-term viability of DNG files because you still need a program and an operating system to run that program on to get at and do something with your DNG negative. But that's probably semantics. At the moment, that seems like a pretty safe bet. Nothing in computers ever seems to go completely away. It would add a minor complication if/when an upgrade to DNG is released. There will probably be software to process DNG for a long time to come, so long as there is a need.
But C, N and others could go a long way toward negating the effect of DNG (if they wanted to) by adopting a strategy of releasing their proprietary raw formats into the public domain after, say, 5 years of the release of a camera model. That gives them time to make money from unique proprietary technologies and (maybe) software sales but still addresses the long-term worries of creators of the images. So far, there's no sign of them doing that although since their raw formats have been deciphered anyway, it may be an exercise that they see no point engaging in. This is idle speculation, of course, I have no idea what the manufacturers' strategies are.
I detect some vehemence against the manufacturers for acting in a proprietary way. I understand this completely. Our images do belong to us and we don't want to be held to ransom to get at them. But it's not 100% black and white, I don't think. In the film days, everyone was better off by standardizing on common colour development protocols. But nowadays, when each one of us can build our own proprietary darkroom, the manufacturers could take the attitude that preserving our images for posterity is our problem, not theirs.
If you write a novel using Microsoft Word, it's good practice to save an ASCII text version somewhere. There's a better chance of being able to read that text file in 50 years that today's Word version, I would guess.
If a manufacturer gains a marketplace advantage over competitors by bucking that trend and fully supporting an open format like DNG, that might change their collective minds. So long as it doesn't affect sales, though, they may decide that they don't need to do anything.