DNG or RAW

DNG

22 (26.8%)

RAW

60 (73.2%)

Total Members Voted: 81

[Schewe]

I understand your method eventually accomplishes the same thing when you save as DNG after completing the raw editing.  Is there a reason for this approach, in particular for keeping the raw copy as well as the DNG?

Well, I shoot with cameras and back that produce large raw files...I have a lot of 1Ds MIII, Phase One P65+ and IQ 180 files. For example, an IIQ raw file from the IQ 180 is about 79MBs (the back is an 80 MP capture). If I convert that to a non-lossey DNG the file size grows to about 96MBs on disk. But here's the kicker...if I modify the original raw file and save the edits to a .xmp file, that XMP file is about 8-9 KBs on disk. If I change the modifications to the DNG, the edits are stored in the DNG (which can be a good thing).

But, if you do your backups based on file modification dates (which is standard), if I modify a raw file, 8-9 KBs must be backed up. To back up the DNG, the whole 96 MBs file must be backed up. Multiply that file size difference by how many captures to shoot and you'll see that editing the raw files–not DNG files saves a lot of disk space.

Yes, when I'm "done" editing a raw file, I'll often convert that file to DNG (usually inside of Lightroom) so I can bake the edits & snapshots and metadata into the image. Prior to that, I edit the original raws...

That fact that I've modified my workflow for my own personal use does not in the least mitigate that current situation where camera makers are spawning off new, undocumented raw files for every new camera released (expect for the several camera companies that use DNG).

Note: one of the reasons I segregate my raw editing from edited files is that I often work in both Camera Raw and Lightroom. In the case of Lightroom, I'm not so worried about writting the .xmp metadata out to the file because the LR catalog is backed up separately. With ACR, that's not the case and any/all changes must be written to .xmp either in the side car (smaller) or the DNG (larger).

[Bryan Conner]

John, I have read your responses as well as the other posts in this thread.  I have seen opinions on why dng could be a bad idea.  It has been said that offering dng file format as an option could add to the price of a camera or to the cost of producing the camera.  It has been said that offering dng file format as an option could....    I have not read anything in this thread or in any other source that factually states anything negative that offering dng format as an option will do.  I have read lots of reasons that offering dng format as an option will make a positive change in the digital imaging world to the end user.

I am looking for facts.  Does anyone have any facts to present that clarify why it is a bad idea to offer dng file format as an option for the user to choose, or not to choose at his own discretion?

[jrsforums]

John, I have read your responses as well as the other posts in this thread.  I have seen opinions on why dng could be a bad idea.  It has been said that offering dng file format as an option could add to the price of a camera or to the cost of producing the camera.  It has been said that offering dng file format as an option could....    I have not read anything in this thread or in any other source that factually states anything negative that offering dng format as an option will do.  I have read lots of reasons that offering dng format as an option will make a positive change in the digital imaging world to the end user.

I am looking for facts.  Does anyone have any facts to present that clarify why it is a bad idea to offer dng file format as an option for the user to choose, or not to choose at his own discretion?

Bryan, all those coulds are based on experience of working with complex computer system's hardware/software/firmware...which is what today's cameras are. Translate could as would.  

Canon and Nikon are not going to open their books to give you facts....and I am not going to relate facts about the company I worked for.  Anyone who has worked in this or related industries, particularly major firms, would understand.

[jrsforums]

In terms of DNG, Thomas was getting pretty pissed off having to decode each and every camera's raw files. In developing Camera Raw, he came up with a process of reading the camera metadata and converting the metadata in order to apply normalized default settings. It was the process of taking the raw image data and metadata and normalizing the adjustments that led to developing DNG. Thomas reasoned that the basic raw image data and the image metadata should not be hidden from easy access.

.....

Give Adobe shit for all the mistakes they have made, but don't give them shit for the things they have done on our behalf.

Interesting, but a bit irreverent stories.  Pretty much common to any view of inter-industry company interaction.

Adobe has been good for us.  I like an use there software....even with the "rental" (as long as they don't make LR rental only).  I, for one, do not relate the rental action to any benefits or DNG activty.

However, as you relate, one of the primary reasons for Thomas creating DNG was to avoid decoding all the multiple camera raw files.  I understand the benefit to Adobe....to date, other benefits to users are minimal.

[Tony Jay]

John, I have read your responses as well as the other posts in this thread.  I have seen opinions on why dng could be a bad idea.  It has been said that offering dng file format as an option could add to the price of a camera or to the cost of producing the camera.  It has been said that offering dng file format as an option could....    I have not read anything in this thread or in any other source that factually states anything negative that offering dng format as an option will do.  I have read lots of reasons that offering dng format as an option will make a positive change in the digital imaging world to the end user.

I am looking for facts.  Does anyone have any facts to present that clarify why it is a bad idea to offer dng file format as an option for the user to choose, or not to choose at his own discretion?

Bryan, the real issue has nothing to do with DNG per se.
It is all about the volatility of your images.
Perhaps you don't care whether your RAW images are still accessible in ten, twenty, or fifty years time, or a century from now.
(Rhetorical observation only.)
But on an industry level this is a massive problem.
There are already plenty of orphaned RAW formats out there.
The software business itself is very volatile and one should not make any assumptions that one's favourite RAW converter will still be available in years to come to make sense of your legacy RAW formats.
The fact that, with some effort, it will be possible to reverse engineer the format, as is done now by Adobe and others, in the future does not mean that it will happen.
Also, software that might still convert early RAW formats may not run with current hardware/software configurations.
That situation will continue to get worse as computer hardware and OS software evolve.
This forum is replete with whizz kids and geeks (affectionate term) who can continue to run old software/hardware combinations and maintain them.
This might work on a hobby level but it isn't sustainable at an industry level.

Currently we wait up to six months for someone to reverse engineer the avalanche of new formats that are released every year so that one can do RAW conversions with one's converter of choice. This is not an issue if one chooses to work with the manufacturer's RAW converter, but for most in the industry this option appears not be the one of choice.

What is really required is a universal RAW format that is generated by all cameras that will make the process of RAW conversion trivial. Trivial in the sense that any RAW converter can process it without any reverse-engineering being required.
Why DNG is being punted currently is that it is the only candidate on the horizon.
DNG is in the public domain with its entire specification available as a SDK and has been offered to the ISO for standardisation.
The standardisation thing does not mean that the format cannot evolve just that the ISO will control how it changes.
All the hot air about how this will stifle camera sensor development in the future is just that - hot air.

The current DNG standard is heavily influenced by the TIFF-EP standard. Most (nearly (all)) current RAW formats are also essentially the TIFF-EP with proprietary metadata influencing how the manufacturer's RAW converter will handle the RAW file.
So, strangely a lot of proprietary RAW formats look a lot like DNG, just with proprietary metadata thrown in.
(This is the basis for Schewe's regular comments about secret RAW formats not really been secret anymore - they are just a pain in the ass to reverse-engineer.)

From an industry perspective photographers are only just coming to terms with digital asset management with most of the current emphasis being entirely focused (pun intended) on the survival of 0's and 1's. The problem is that it will be all for nothing if in a hundred years time there are half a million legacy RAW formats that need something to convert them with.
What do you think the chances are that your format(s) will still be supported?

From an individual perspective, as I began this post, it may not matter to you personally. After all in a hundred years YOU will be dead. So too will your images be dead.
But maybe, just maybe your images are worth something to somebody, perhaps only your family, but, maybe your images are worthy of being hung in the Smithsonian or the British Museum, and, maybe, that is not currently recognised right now.
Even if the 0's and 1's survive (still unlikely at this stage) is any software going to be available to convert your assiduously conserved digital images.
From an industry perspective the sad thing is that slides from twenty years ago stand a much better chance of surviving a hundred years than current digital image files.

Are the current issues solvable? Yes, I believe so.
Does DNG absolutely have to be a part of the solution? No!
But currently there is no universal RAW format and the bottom line is that we do need a universal RAW for archival purposes.
A universal RAW format, by itself is not enough but it is an essential part of the solution.

If someone can come up with a better universal RAW format I am all for it - bring it on.
Until then, if we have any sense, lets not throw the baby out with the bathwater.

Tony Jay

[Schewe]

If someone can come up with a better universal RAW format I am all for it - bring it on.
Until then, if we have any sense, lets not throw the baby out with the bathwater.

Well, on the Planet Earth, the number of people who "could" possibly come up with an alternative to DNG could be counted on one hand and the initials TK head the lists...Other than Dave Coffin (dcraw) and maybe a couple of others, I can't think of another individual or group that could beat out the guys above...DNG has been offered to the ISO, let's see where it ends up.

[Bryan Conner]

Tony, thanks for your well thought out response. You and I are on the same side of the fence.  I understand everything that you said and I have agreed with everything for the past few years. All of my images for 2012 and most of 2013 are all dng files. The fact that all raw processors do not read dng files is the only reason why some of 2013's images are not dng. 

I see the benefits of dng format in digital photography.  I agree that it is not so much about dng, or it should not be about dng and/or Adobe.  I can not think of any logical (to me) reasons that offering dng would be a bad idea. 

I have not heard anyone give any reasons that show that making dng an option is definitely a bad idea.  No one has even shown evidence that making dng an option probably would be a bad idea.  I am not interested in hearing or reading about someones predictions that they apparently got from their magic 8 ball.  I want facts.

[Tony Jay]

...You and I are on the same side of the fence...

I did actually figure Bryan, nonetheless I thought it important to lay out a few key facts and put the whole issue in context for all.
Sadly this road was travelled earlier last year and the result was polarisation of positions.
But you never know...

Tony Jay

[ErikKaffehr]

Hi,

It is still nothing that hinders proprietary tags to be include in DNG files. The vendor can add any data in a DNG as well as in a proprietary file format. Any program that can read a file can ignore tags it doesn't recognize.

Regarding spectral response data, it is normally handled by colour transformation matrices as far as I understand. DNG offers two one for 2800K and one for 6500K and a recommended way to interpolate between the two. As far as I believe to know, Adobe actually measures the spectral response of the sensors they develop DNG profiles for. (It is not terribly hard, you need a monochromator for that).

As a side note, I made a limited test of colour accuracy on my P45+ and Alpha 99 using both Capture One and Adobe Lightroom. Lightroom consequently outperformed C1, even if I would say that the winner was the Alpha 99 by a wide margin. That is of course not the whole truth. We can see millions of colours and the ColorChecker has just 16 colour patches, that happen to be industry standard. But correctly reproducing the 16 ColorChecker colours is not a bad start. 

In honesty, I started looking at colour calibration as I hated colours coming out of the P45+ with Adobe standard profile, so just matching ColorChecker decently is not enough.

My results are here: http://echophoto.dnsalias.net/ekr/index.php/photoarticles/79-p45-colour-rendition

As a side comment, as a long time Minolta/Sony user I am accustomed to "Sony color", I also feel that Sony has a AWB that is biased towards a bit cold (bluish rendition), when I shoot a grey card I almost always prefer the auto white balance by Sony compared to grey card exposure.


Best regards
Erik

Again, it's really no different in that respect from all Raws that follow the TIFF specifications. Common info is known and shared, e.g. where the top left sensel within the image data block is located, and what pattern the Bayer CFA has and thus what color that top left pixel represents. However, it does not say exactly what spectral transmission that filter color had. So it may be a very pure Red or one with a trace of Green transmission as well to increase its transparency and make the camera more sensitive to light.

That info, required for very good color conversions may be tied to a patent for a very special Red dye that that camera maker can produce for a fraction of the cost of its competitors, but it bleaches faster than the Red from competitors. They do not want to tip of their competitors, because they might want to also change to a cheaper Red dye or switch source, which would reduce the competitive advantage. Only the original manufacturer will know that they will have to modify the Raw conversion for Red a bit faster as time goes by, to compensate for the rate of bleaching, by e.g. using the number of exposures that's being recorded in the maker notes, combined with the average exposure level that's encoded in common tags in the EXIF section.

Now when Madmanchan does a very good job, he may be able to quickly find an assumption for the Red characteristics to use during conversion for new cameras that's pretty close to what is optimal for the Adobe conversion engine, and within a month or two the Camera will be supported, with the DNG converter than can be used during the first ingestation/conversion into Lightroom. But he won't know of the fading characteristics of that dye because Adobe wouldn't pay royalties for that invention. So that same file will convert the same over time, but later Raw Captures will deviate in color.

Now this is just a silly example, but it might also be something like unique image verification data that's tagged to the file for use in law enforcement situations, or a piece of code that uniquely identifies the retina of the person who looked at the EVF to make that kiddy porn, or a feature that unlocks 3D image data that was encoded by splitting sensels into a left image and right image, which also allows to adjust depth of field after taking the image based on depth clues.

Intellectual Property is big business, and companies will protect that. They will even make life a bit less convenient for people that insist on using a certain brand of converter because those people feel that it suits their workflow better. Those people may perhaps not be able to change DOF after the fact, but just get the left or right image, just like they took it.

Cheers,
Bart

[Bart_van_der_Wolf]

What is really required is a universal RAW format that is generated by all cameras that will make the process of RAW conversion trivial.

Hi Tony,

I'm wondering how that could be possible when the hardware that generates the image data differs? Those hardware differences will need to be converted into human viewable images, and the Raw images will always be different because the input/hardware/firmware is different.

Again, we're not talking about the image data that the ADC produces and gets recorded in the image data section of Raw files. That data is tagged in all formats, including DNG, and all converters read out the same numbers (one hopes). It's what happens with that Raw data that differs based on how intimately one knows what is additionally recorded in the proprietary Raw Maker notes, and how to interpret that image data.

That proprietary data will not be surrendered by the IP holders, because it represents incurred cost and future value. It can also make a competitive difference, even if the current differences between final results may seem small from the casual observer's position (and clever software can reduce those differences even after the fact). That's why there cannot be some universally agreed encoding of future features, because that would hurt competitive advantages. Competition is what fuels innovation and hopefully reduces end-user cost.

There are still many innovations ahead in sensor technology, so timing the introduction to market of those innovations is also critical for companies. Not spilling the beans too soon can also mean a lot of additional sales of more mature technology, but early warnings will halt most sales with the current sales models and alert competition.

Cheers,
Bart

[ErikKaffehr]

Hi,

I have no issue with proprietary information. What I want is:

- A well publicly documented standard container format, universally supported
- A set of well defined essential tags, including white balance,  colour conversion matrices  and other essential data
- Bad pixel mapiing, bad pixel columns, dark frame data
- Proprietary data needed for proprietary processing is fine with me, I can use a proprietary processor if I need proprietary functions.

I am a bit concerned about having dozens to hundreds of different data formats. I don't know how many of those will be readable in say 10 or twenty years.

I would also say that standards are good things, and I don't think standards limit evolution. Many standards are extensible.

Best regards
Erik Kaffehr

Hi Tony,

That proprietary data will not be surrendered by the IP holders, because it represents incurred cost and future value. It can also make a competitive difference, even if the current differences between final results may seem small from the casual observer's position (and clever software can reduce those differences even after the fact). That's why there cannot be some universally agreed encoding of future features, because that would hurt competitive advantages. Competition is what fuels innovation and hopefully reduces end-user cost.

There are still many innovations ahead in sensor technology, so timing the introduction to market of those innovations is also critical for companies. Not spilling the beans too soon can also mean a lot of additional sales of more mature technology, but early warnings will halt most sales with the current sales models and alert competition.

Cheers,
Bart

[Bart_van_der_Wolf]

Hi,

It is still nothing that hinders proprietary tags to be include in DNG files. The vendor can add any data in a DNG as well as in a proprietary file format. Any program that can read a file can ignore tags it doesn't recognize.

Correct, and that's why there is no benefit for Camera makers to drop their proprietary Raws. It would add the burden to make sure that all tags that are not used by their cameras (even their own legacy ones) are set to values that might not inadvertently screw up the conversion of their Raw data in whatever application is used. They have no control over the negative side effect that others may create in their Raw converter, based on some interpretation that should have no use for the camera.

When a Raw converter produces substandard conversions, then that's up to the converter manufacturer to explain or fix. When a Camera maker screws up a parameter that has no use for them, the blame will fall on them. Therefore it will add effort/cost to avoid such mistakes.

Cheers,
Bart

[digitaldog]

I understand the benefit to Adobe....to date, other benefits to users are minimal.

Minimal FOR YOU jrsforums (whoever that may be). For you. There are others here who find the benefits far, far more than miniumal. But you obviously don't care about anyone but yourself based on your writings. You prefer to speak for everyone too (kind of dumb). You can't muster up any reason why those of us that desire a DNG should be kept from using that format. And you're the inventor of very deep rabbit holes based on pretty shallow thinking. Please don't use DNG. Please let the rest of us decide what's a useful or non useful workflow (based on facts).
Please make an attempt to separate your loathing of Adobe and think about your fellow photographers who intend to use Adobe solutions even if you have scrubbed your computers of all their products and file formats out of (what protest?).

[digitaldog]

Quote from: BartvanderWolf on January 04, 2014, 06:50:59 AM
Hi Jeff,

You make it sound as if they are doing this pro-bono, in their spare time. Fortunately that's not the case, they get paid by Adobe, because it helps the company's bottom line ... (they would not likely get paid to waste company resources). And since Adobe is not in this game for charity, they must believe it will benefit their investors to do the effort.

In point of fact, both Camera Raw and DNG started as unpaid sandbox projects that Thomas wanted to do

Well that effectively kills that silly anti DNG argument! The format started pro-bono (which isn't salient in terms of issues with DNG), but now Thomas and every other raw converter engineer outside the camera makers DO get paid to hack these formats so that's a waste of money and engineering resource. Either way you slice it, bringing up the economies here shows the silly nature of a very time limited proprietary raw format.

Cost or no cost to build a DNG from a new proprietary raw, the need is simply unnecessary and has a negative affect on a group of people. All this shows the need for an open raw format, not the opposite. Doesn't matter of there are two highly qualified people who could hack any proprietary raw format, I can't and few of you can either. Some of the so called arguments in favor of proprietary raw formats are based on some of the silliest ideas I've heard outside the US political arena in a very long time!

[Bart_van_der_Wolf]

Well that effectively kills that silly anti DNG argument!

Hi Andrew,

Apple also started from a garage as a hobby project to build some circuit boards for fellow hobbyists. That's as irrelevant to this discussion as the suggestion that Thomas is more charitable than mother Theresa (which she wasn't really) is phony. Thomas is an enhousiast with great skills that were used for the benefit of Adobe. Nothing wrong with that. But it has nothing to do with the discussion, it's totally irrelevant.

Besides, for that DNG conversion the proprietary Raw still needs to be reverse engineered, because hardware changes. DNG changes nothing about that, except that the process is a bit more efficient for Thomas/Adobe. Nothing wrong with that either.

Cost or no cost to build a DNG from a new proprietary raw, the need is simply unnecessary and has a negative affect on a group of people.

No, YOU find it unnecessary (despite having access to DNGs), and would prefer others to pick up the cost (time/expense/competitive disadvantage). Adobe, understandably from their perspective, also wouldn't mind if others took the burden on their necks. You and they are entitled to that opinion, but you at the same time find it difficult to accept that others (those camera makers) at the receiving end disagree. Seems a bit shortsighted.

I can also think of a few things that would fit my personal preferences better than the current situation, but I tend to dream with my eyes closed. During the day I try to be realistic, which is not the same as being in favor of proprietary formats.

Cheers,
Bart

[digitaldog]

Apple also started from a garage as a hobby project to build some circuit boards for fellow hobbyists. That's as irrelevant to this discussion as the suggestion that Thomas is more charitable than mother Theresa (which she wasn't really) is phony. Thomas is an enhousiast with great skills that were used for the benefit of Adobe. Nothing wrong with that. But it has nothing to do with the discussion, it's totally irrelevant.

Come on, really? First, no one stated Thomas was more or less charitable than anyone else. Thomas did go about trying to fix an issue. Whether he gets paid or not, there's a cost to this problem we're discussing, someone, somewhere pays, it's a problem that has zero reason to exist! That's the bottom line.

Besides, for that DNG conversion the proprietary Raw still needs to be reverse engineered, because hardware changes. DNG changes nothing about that, except that the process is a bit more efficient for Thomas/Adobe. Nothing wrong with that either.

Which is exactly the justification for the camera makers providing this for us like they do the JPEG. Or are you suggesting that when a new camera is produced, there is zero cost to the manufacture’s to provide that JPEG but there always is with raw? If so, evidence please (along with the cost which I'd be happy to pay for and likely am somewhere).

No, YOU find it unnecessary (despite having access to DNGs), and would prefer others to pick up the cost (time/expense/competitive disadvantage).

I'm happy to pay and already do. So do you or any other user who's NOT using the camera manufacturer's raw converter.

[jrsforums]

Minimal FOR YOU jrsforums (whoever that may be). For you. There are others here who find the benefits far, far more than miniumal. But you obviously don't care about anyone but yourself based on your writings. You prefer to speak for everyone too (kind of dumb). You can't muster up any reason why those of us that desire a DNG should be kept from using that format. And you're the inventor of very deep rabbit holes based on pretty shallow thinking. Please don't use DNG. Please let the rest of us decide what's a useful or non useful workflow (based on facts).
Please make an attempt to separate your loathing of Adobe and think about your fellow photographers who intend to use Adobe solutions even if you have scrubbed your computers of all their products and file formats out of (what protest?).

As usual you and your friends turn to vicious personal attacks....

You can use DNG today...or not...your choice.  You just can't get it directly from most camera manufacturers, but you do not want to listen to any reasons why it may not be of advantage to them.

"loathing of Adobe"....what "hole" did you pull that from...  Just another way you dismiss logical discussion into personal attacks.

[Damon Lynch]

But you obviously don't care about anyone but yourself based on your writings.

Hi Andrew, I admire your passion in favor of DNG. Good for you that you feel strongly about it and that you want to encourage others too. But to be honest I find the moral side of your advocacy wholly unconvincing. In saying that I don't mean to undermine your argument -- on the contrary, I encourage you to strengthen the moral aspects of your advocacy. Consider Dave Coffin's stance, for example: he states he does his work not for financial gain, but to win the respect of his peers. He undoubtedly has the track record to back up his stance. He makes his valuable work and genuine expertise on decoding RAW formats available freely, in every sense of the word free. He clearly does care about others. I assume you admire his track record for more than just his technical expertise.

However based on what you wrote earlier, at the same time you're apparently happy to lock up in a proprietary format the metadata on how we process our RAW files. From my perspective this is a contradiction. I'm glad that this metadata is part of the XMP spec (and therefore DNG spec) and that an increasing number of tools we rely on to do our work utilize it. We need much more of this, not less, for both pragmatic and moral reasons.

And finally absolutely nothing I read in jrsforums' arguments suggested to me that he or she is any less moral than any of the rest of us! I feel characterizing him or her otherwise is unfair.

[digitaldog]

He clearly does care about others. I assume you admire his track record for more than just his technical expertise.

Absolutely!

However based on what you wrote earlier, at the same time you're apparently happy to lock up in a proprietary format the metadata on how we process our RAW files. From my perspective this is a contradiction. I'm glad that this metadata is part of the XMP spec (and therefore DNG spec) and that an increasing number of tools we rely on to do our work utilize it. We need much more of this, not less, for both pragmatic and moral reasons.

I'm not happy locking up anything that I could use. However I understand there's all kinds of functionality that is and should be proprietary. I gave several examples including Photoshop Layers (timely when some Adobe customers are thinking of using other image processing soltuions and some assuming all that data will transfer).

Is there some proprietary data, data that can only be used by the camera manufacture's converters which becomes non proprietary months after the new format ships due to the labors of other's (Dave included)? Does that data change by virtue of it being understood? It is still not used, but deciphered such every other converter can now access the meat and potatoes (the raw sensor data) such they can render the image? I believe not but I'm open to learning more. If this temporary proprietary data, data that I'll never use is only proprietary for a few months, causing others an inability to access all the data as they please, what good is it being proprietary in the first place?

[Damon Lynch]

I'm not happy locking up anything that I could use. However I understand there's all kinds of functionality that is and should be proprietary.

So you're against proprietary file formats except when you're in favor of them? More specifically, you're against them when you see that they detract from your workflow, but in favor of them when you perceive they are not. It seems you don't pay much attention to the workflow of others when it comes to these aspects of post-processing. Some people do want that data to be non-proprietary, for all kinds of reasons. It may not affect you directly but it surely affects them. I don't mention this to "score points" or something similarly juvenile, but only to say that reality is not as black and white as seems.

Is there some proprietary data, data that can only be used by the camera manufacture's converters which becomes non proprietary months after the new format ships due to the labors of other's (Dave included)?

We are using the same terminology to describe different phenomena here. When Dave Coffin figures out how camera manufacturers have encoded data in a file format he has decoded it. The file format remains proprietary.  For example, MS Word .doc files are in proprietary format, even though the format has been reverse engineered by programmers who never worked for Microsoft. By contrast, MS Word .docx have never been proprietary, are fully documented, and are an ISO standard.