WS_Lam@adobeforums.com wrote:
> With the new Nikon P6000 announcement, there're lots of angry voices about Nikon's decision to use a new RAW format. The new RAW format can only be converted on a system that supports Windows Imaging Component.
>
> Everyone is looking at Adobe... will Adobe be able to support WIC formats even on a OS X platform?


Well there goes Nikon shooting themselves in the foot again.

I switched to Canon when the 5D came out because Nikon was dragging
their ***** going to full frame and the Kodak SLR/n I bought to try to
continue to use my Nikon lenses was next to useless.

I was toying with the idea of switching back with the advent of the D700
but now that's out the window as I can't trust that Nikon won't switch
to this new less accessible format for their future pro level cameras.

Dave Smith

WIC is Windows only.

Even on Windows, WIC is basically useless for Camera Raw/Lightroom type applications.

How disappointing. Just when I thought I was seeing a newly energized, newly enlightened Nikon again raining brilliant products upon the photographic market...it turned out to be just a brain fart. It's the same ol' Nikon boys planning products around the same ol' laminated conference room tables.

The P6000's implementation is really, really...dumb.

Anybody know who to write to at Nikon to encourage them to get a clue?

Hmm, there is no such thing as a "WIC-only" format. Nobody is prevented to read the file bits directly, on any platform. The only real question is to know if Nikon is going to document the inner details of the format and/or provide a separate SDK to decode the files.

Besides that, I'm a bit perplex about Thomas's comment regarding the uselessness of WIC for ACR/LR-type applications. As far as I know, WIC supports raw development (exposure, contrast, white point, gamma, tint, saturation, sharpness, noise reduction, color profile, tone curves, rotation and rendering mode quality - see IWICDevelopRaw).

That's a lot of ACR/LR-type processing.

Cheers,
Axel

according to this
http://www.dpreview.com/news/0808/08080702nikonp6000.asp
and
http://press.nikonusa.com/2008/08/new_coolpix_p6000_offers_digit.php

it'll also be supported by viewnx. i don't think it'll be limited to wic interface alone. probably both. they don't say WIC-only anywhere.

so 3rd party developers should be fine.
BTW anywone tried to use IWICDevelopRaw with nikon's raw codec? or canon's? does it allow to set/get parameters like sharpness or it's just a viewer?

true. i was thinking that since view nx doesn't have the system requirements of wic interface there are methods other than wic to read the data.

if it's entirely crypted then that would be a real problem.
sony/canon encrypt their own data too. see this interview.

http://www.dpreview.com/news/0504/05042701davecoffininterview.asp

it does i know...
nothing to do but wait for samples to appear on the web

"Thomas Knoll - 8:19pm Aug 7, 08 PST (#4 of 17)

WIC is Windows only.

Even on Windows, WIC is basically useless for Camera Raw/Lightroom type applications."

Interesting, is this Adobe's nightmare scenario, with Microsoft offering "free" raw conversion if you have Vista?

Is this also the real reason for Adobe's push for DNG?

> Is this also the real reason for Adobe's push for DNG?

Let's leave the commercial considerations aside. In what way do you believe DNG would protect Adobe from any competition?

Images in DNG format have yet to be processed, exactly as the native raw images. Every program, which can process a native raw file can process its DNG version very easily. In fact, it is easier to process the DNG files than most native raw files, because the DNGs are cleaned up already, i.e. the garbage what the cameras create under the disguise of TIFF is turned into a straigforward format.

>The WIC interface supports (i.e. forwards) those functions, which the designers of WIC found important enough on the day of the definition. This is a narrow-minded approach; in fact it had become outdated before it has been published. I don't think that the independent raw processor products will have difficulty to remain competitive.

They already have a hard time. See the lag when new camera comes out, and see the default conversions you get from their first release. Why do you think most users prefer CaptureNX over pretty much everything else as far as conversion quality is concerned? The Nikon NEF codec for WIC exhibits the same basic conversion characteristics as CaptureNX, except that any app can take advantage of it. Besides that it's very easy to extend COM interfaces, if the need arises. If something crucial is missing from WIC 1 then it will be in WIC 2.

>If one is confident, that the creator of the decoder can do the best noise reduction, sharpening, etc. then one can use their own raw processor, like Nikon Capture, Canon DPP, etc., there is no need for WIC based raw development.

You accuse WIC designers of being narrow-minded (your opinion, anyway) and you follow on by being even more narrow-minded. What about batch converters, application without a user-interface running in image processing farms etc, how do they use Capture or DPP (or ACR, for that matter)? A WIC codec just enables a *wide* array of apps to use the raw files, without waiting several months for 3rd parties to catch up and several more months for the correct "conversion profiles" to be found.

The codecs works on day one, and if recent history is any indication, are hard to beat regarding conversion quality, e.g. the NEF codec supported the D3 immediately, and already supports the D700 while D3/ACR users had to wait, and are still suffering strange color rendering at high ISO with the latest ACR and profiles (I just read that in the ACR forum) a *year* after the camera was released. And where is the D700 support? What about the P6000? When will 3rd parties actually match CaptureNX and the NEF codec, let alone beat them?

Another example: the Canon RAW codec for WIC comes with 24 different converters, presumably each tuned for different sensors/chipsets/whatever (just install it and check C:\Program Files\Canon\RAWCodec130).

It would be naive to assume that 3rd parties can just re-do all this work and tune everything for every Canon camera and beat them all along on all counts, then do the same for all Nikons, Sony, Pentax, Olympus, you name it, and do it again everytime a new camera comes out. It's just a daunting task, not very likely to succeed because some of the data is kept secret and will likely always remain secret.

Everyone is asking for a standard way to access raw files, and some would like to see their container format become that standard, but manufacturers apparently have other plans. Then why not standardize on a common, higher-level access method and let the manufacturers do the "last mile", i.e. decode the secret data and interact with the files themselves. If the control offered is not sufficient, then why not ask for more control and have the relevant parties extend the interfaces? If a platform does not have an extensible plug-in infrastructure like WIC, why not write one, this way manufacturers would release, say, a WIC codec for Windows and an XYZ codec for the Mac, both bundled with the camera, and all software could immediately take advantage of them without needing to be rewritten/adapted. It's called device-independance and it already exists for different device classes, such as printers or display adapters. Why digital cameras would be any different?

> Except that the "straightforward format" still leaves at least some bits out: any encrypted (or undocumented) part of the original raw file that could not be understood must be either left out or copied as an opaque BLOB, who's content will be just as alien to the DNG-enabled apps as it was to the other 3rd-party raw converters.

> DNG as an universal format is an illusion until the manufacturers truly documents all their proprietary details

I am not happy to be forced in the camp of the DNG apologists, but the critique coming from laypeople is often doing just that.

DNG can not be held responsible for the fact, that manufacturers literally hide, sometimes even encrypt the information relevant for the rendering of the image.

It is a fact (some will be disputing it), that DNG *is not a substitute format for any specific raw format*. I do not convert my native raw files in DNG. I regard DNG as something *in the right direction but not at the right point*.

However, the *real issue* is NOT that manufacturers are *hiding* or even *encoding* their proprietory data.

The *real issue* is, that the development on this area is too fast to be followed in and accomodated by the standard, and Adobe is one of the last places suitable to do the job of unification.

Camera makers are coming up with new features on the daily basis, and the DNG specification is limping years behind. THAT is the real impediment on the way of DNG becoming a useful format.

> Why do you think most users prefer CaptureNX over pretty much everything else as far as conversion quality is concerned?

Why do you think most (NIKON) users prefare CaptureNX? Where did you get this information?

ACR become much better overnight with the new profiles, at least regarding Canon. I guess CNX's advantage on the color reproduction area too has become smaller, and most of ot can be eliminated by more profiles.

> Besides that it's very easy to extend COM interfaces, if the need arises

I suggest you to stay on some area you understand something of.

> What about batch converters, application without a user-interface running in image processing farms etc, how do they use Capture or DPP (or ACR, for that matter)?

What do these subjects have to do with WIC? Have you never saw raw converters with bach conversion interface? Do you really think that this is a subject of substance?

> The codecs works on day one

The codec works when its creator adopts it to the newest camera. How long have people been waiting for a Nikon codecs before it finally came out? What about a 64bit codec?

> And where is the D700 support? What about the P6000? When will 3rd parties actually match CaptureNX and the NEF codec, let alone beat them?

This critique is justified. The architecture of ACR SUCKS, it is responsible for the mentioned problems, which appear everlasting.

> It would be naive to assume that 3rd parties can just re-do all this work and tune everything for every Canon camera and beat them all along on all counts

1. Beating the manufacturers converters is nice, however not essential. What percentage of the buyers of a model do you think are purchasing in the first two-three months after the annnouncement?

2. Supporting a new model is not the question of time for the development. It is the failure of the ACR architecture, that it takes so long. The factual support could be done much earlier.

I have a product, which works with native raw files and, of course, with DNG. The adoption of a new camera may take a few minutes, or it may take much longer, if one discounts the spectral response of the sensor. It is not a straightforward issue.

> It would be naive to assume that 3rd parties can just re-do all this work and tune everything for every Canon camera and beat them all along on all counts, then do the same for all Nikons, Sony, Pentax, Olympus, you name it, and do it again everytime a new camera comes out. It's just a daunting task, not very likely to succeed because some of the data is kept secret and will likely always remain secret.

I hope for you you don't wenture on the area of software manufacturing; you are a born loser on that area.

> Then why not standardize on a common, higher-level access method and let the manufacturers do the "last mile", i.e. decode the secret data and interact with the files themselves

Because "we" do not trust those suckers to do this the best way. Is that not simple?

>That's actually pretty funny...you know what the real secret is? There ain't no secrets
>You honestly think they are keeping secrets in their raw file formats?

Just show me the official and complete documentation of Canon and Nikon's maker notes.

Call it secret, call it undocumented, call it what you want, the bottom line is the same: some details are missing. A DNG file with embedded proprietary maker notes is no more transparent, portable or universal than a NEF or a CR2.

I never said that the fact that Canon/Nikon/Sony and others were secretive about the details of their files was a good thing for photographers, what I said however, is that the universality of DNG is an illusion until manufacturers document everything, and all the data can be converted and interpreted. Until that day I fail to see what DNG really brings to the table.

>Writing into a DNG file is just as safe as writing into a TIFF or JPEG. That makes a DNG file a lot more useful...

Agreed on that one. It's nice to be able to write metadata to the files, but this does not solve the proprietary data issue: manufacturers still write blobs in their files, that only them really know everything about. Until they all write everything as discrete, documented EXIF or XMP fields, full understanding of the data will remain a challenge, stored in a DNG container or not.

On a related topic, WIC supports metadata reading and writing and applications can thus read/write metadata in a uniform manner, including fast in-place rewriting if space/padding allows, independently of the underlying file details. This is really powerful, you don't need to know anything about the container, the framework and codecs abstracts the details for you nicely. I know from 1st hand experience that writing does not always work exactly as it should in practice, even with Adobe's own DNG WIC codec last time I checked, but things will get better, metadata writing into proprietary files is on its way to become a non-issue for WIC applications, e.g. Vista can now write into NEFs without knowing anything about them.

>Yeah, and I remember that the Nikon codec (and I believe the Canon one as well) had some problems with that...

Yes, that was *before*, codecs are not cast in stone, they evolve and gets updated/fixed/improved just as every other piece of software.

>Also note that DNG actually taught Nikon and Canon a thing or two about file formats

Hehe, what makes you think it's DNG? Have you ever heard of the upcoming ISO/IEC 29199-2 image container format?

>It has considerable usefulness as a replacement for JPEG

Most certainly, and it also has an IFD structure just like TIFF 6.0 and TIFF/EP so I'd guess that, as a container, it could be made to accomodate raw data just as well.

>I would expect some new cameras in the future offer JPEG XR

MS actually demoed at least one camera with native JPEG XR support a couple of months ago, so that future might not be too far away.

One interesting new file format to watch is Nikon P6000's NRW. All we know so far is that it comes with a WIC codec and that it's not a NEF as CaptureNX won't read it.

FWIW The post-processing forum on Nikonians is full of discussions on Capture NX, with a few on Lightroom and a handful on CR. You could probably make of this what you want, but my guess is that most Nikon users use CaptureNX for Raw.

I find this is compounded by the common Nikonians' criticism of ACR's default rendition - something which has now been beautifully addressed by Adobe.

> Just lurk around in photographic communities like Flickr or dpreview. The consensus is pretty unanimous: CaptureNX converts NEF files better, with more accurate color renditions

What "better" is depends on the customer's need. The colors are more accurate with CNX; this is (was) true re Canon's DPP and ZB as well. However,

1. the color reproduction got enhanced a lot, and it can be enhanced even more with version 1.2,

2. while the color fidelity may be of utmost importance for some, the scope of adjustments may be even more important for others.

(How long have people been waiting for a Nikon codecs before it finally came out?)

> About zero minutes. IIRC the codec came out around January 26, 2007, *before* or about at the same time Vista was released to the general public

Except that they forgot to test it. Though this is nothing new (and not relevant to the subject anyway).

> What percentage of the early buyers would accept not to be able to read their raw files on day one?

This is the sore point. However, customers do not want to *read* their raw files but *process* them in the first place.

> WIC gives any application programmatic access to raw files, with very little effort

BS. What about reading the WIC interface documentation?

DNG does not (and can not immediately) account for new features offered by the cameras. It would require the extention of the DNG specs each time a maker comes out with a new idea not covered by the current specs; this is a huge drawback. This goes back even to pre-DNG times; for example such simple issues as contrast, sharpness and saturation can not be communicated through standard tags (the designers of the EP specs were very shortsighted).

On the other hand, WIC does not help on this problem the very least. Not only, that the WIC interface would have to be expanded with such features; the main issue is, that WIC does not offer data access but function access. The drawback is much worse than with DNG.

The only thing that really matters is, will Photoshop support the P6000 raw format. Until that support is available, people will keep buying G9s.