DxO Optics Pro / dng output / color

[deejjjaaaa]

And/or have customers bitch like hell when they hear this stuff.

I do not recall any bitching and I was reading DxO forums (not recently though)...

Keep in mind, LuLa gets something like 1 million unique hits a month. This is the place to air out that smelly vendor laundry.

I 'd assume that is 999999 hits from LR users   ... people use DxO for certain niche (less niche nowadays) features... consider in Camera DNG and manufacturers not giving a damn about what was/is written here...

If they want to say "look, we use Adobe RGB (1998) for processing and that's the gamut limit" fine. Just don't lie and post a lot of dog poop about a working space that is been used for 13+ years.

16bit raw from MFDB anybody ?

[deejjjaaaa]

Or who don't know they're missing those colours.

Bob...even Schewe is sometimes not aware how ACR/LR works inside and he writes books about that... colors... for as long as the users and their customers are happy with whatever colors DxO delivers.

I suspect not all DxO users fit into either category though.  

true, we find one person and he is welcome to illustrate his issue w/ a raw file to show the critically important missing colors

As Andrew said digital raw files can capture more than the ProPhoto space.

yes, shall we ask Adobe to change its internals ?

There are printers that can render more than AdobeRGB in some colours.  DxO's advice is wrong and their clipping of colours even when optionally trying to output a file tagged with ProPhoto is bad programming.

true, I 'd again assume that it was/is a non issue for DxO customer base... some people consider how ACR/LR work are an example of bad programming (to sacrifice certain things in order to have a real time/close to real time UI feedback)...

Further, their comment

comment from a tech rep... never shall be trusted.

that nothing more than AdobeRGB is needed because monitors can't reproduce it is ridiculous.  Monitors today may not be able to reproduce more than, or even as much as, AdobeRGB but who knows what the future will bring.  A few years ago we had to be satisfied with monitors that could only reproduce sRGB.  Same with printers and papers.  If nothing else, using ProPhoto is future-proofing.  Why go back and have to reprocess hundreds, thousands or 10s of thousands of RAW files to take advantage of display/printing advancements in the future?

original raw files are the only future proofing... having said that - shame on DxO   !!! I bet they were paid by Canon, Nikon, Sony Adobe to use AdobeRGB !

[RFPhotography]

Bob...even Schewe is sometimes not aware how ACR/LR works inside and he writes books about that... colors... for as long as the users and their customers are happy with whatever colors DxO delivers.

No one is perfect. 

yes, shall we ask Adobe to change its internals ?

Not really sure that's necessary.  Their raw working space is a form of ProPhoto and there should be no clipping of colours if images are exported or saved and tagged with ProPhoto.  Is there a practical colour space that's larger that could be used?

true, I 'd again assume that it was/is a non issue for DxO customer base... some people consider how ACR/LR work are an example of bad programming (to sacrifice certain things in order to have a real time/close to real time UI feedback)...

But at least there are ways to work around the 'limitations' Adobe puts in its programming.  I'd not really want to work with a non-gamma corrected image (even DxO does that) but it is pretty easy to back out the Adobe render curve.

comment from a tech rep... never shall be trusted.

Which is unfortunate, but likely true.

original raw files are the only future proofing...

Only if a viable working space larger than ProPhoto is developed.

having said that - shame on DxO   !!! I bet they were paid by Canon, Nikon, Sony Adobe to use AdobeRGB !

You really don't like Adobe.  Almost as much as I dislike Apple. 
[/quote]

[fdisilvestro]

Hi,

There is a way to preserve all the colors (outside AdobeRGB) in DXO, that is to process the files as linear DNG instead of TIFF. This workflow requires that you continue the process in LR/ACR, using DXO only for optical corrections.

It made more sense in the past because of the optical correction that were not available in the previous versions of LR/ACR (Still, DXO's corrections are not just about distortion. There are other corrections for the lens+sensor combination such as "lens softness" that are unique).

A linear DNG has been demosaiced but not color encoded. In LR/ACR you could apply your favorite DNG profile, color balance and you could even generate DNG profiles out of them
.

[fdisilvestro]

IF what DXO says is true (it isn't), how on earth can they allow their users to convert into ProPhoto and worse, when the data is a smaller gamut, Adobe RGB (1998)?

Actually the option for output from DXO are "sRGB", "AdobeRGB" and "Custom". If your want to use Prophoto RGB you need to go the "Custom" way and find the appropriate icc or icm file (some users might not even know where to find such files).

[digitaldog]

If your want to use Prophoto RGB you need to go the "Custom" way and find the appropriate icc or icm file (some users might not even know where to find such files).

It's pointless IF the processing color space is limited to the Adobe RGB (1998) gamut.

[deejjjaaaa]

Not really sure that's necessary.  Their raw working space is a form of ProPhoto and there should be no clipping of colours if images are exported or saved and tagged with ProPhoto. 

the same statement is true for DxO.... Their raw working space is a form of AdobeRGB and there should be no clipping of colours if images are exported or saved and tagged with ProPhoto....

Is there a practical colour space that's larger that could be used?

there are colorspaces containing prophoto, there are operations that can be done in native camera's RGB before any demosaick and color transforms... and please do not assume that somehow prophoto coordinates are (the only or the most) practical (just because they happened to be used in ACR/LR)... there are converters other than ACR/LR and somehow there is no proof that ACR/LR delivers a better result just because of its particular design in that matter.

Only if a viable working space larger than ProPhoto is developed.

working space where ? for postprocessing in photshop ? BetaRGB for example is better even not larger... and as for a raw converter - ACR/LR is just a properly automated photoshop doing a postprocessing on prophoto coordinates/linear gamma image that's it... raw conversion there ends before you have a chance to move a single slider... then what you do is essentially postprocessing

You really don't like Adobe. 

on the contrary... since what is known as "process 2003" became a history, ACR is a good product... and PS of course... but that does not mean somehow that design decisions made eons ago were/are the ideal (they might be still competetive and practical though)

[RFPhotography]

But that's not what a number of people are saying and that's where there appears to be confusion.  It seems that the thinking is that DxO clips colours to Abobe even if the image is tagged with ProPhoto at the end of the chain when the batch is exported. Are yousaying that's not what is happening?  Have you got the proof to support that?

I understand there are larger colou spaces than ProPhoto. The point is are those practically availabe to be used. As far as other raw converters that may allow for more flexibility, again the issue comes down to how practical those are to use.  If they have to be used in cooamnd line mode, or don't have an efficient GUI, or are produced by independent programmers where support isn't availabe, or .... then those are't overly pracrical. 

Working space and for tagging images. It may all be moot with current technology since we're already clipping colours for display/editing purposes even with so-called wide gamut monitors.  I understand that the very basics of raw conversion end when you open the image in ACR/LR (and many others).  And that may already clip some colours that the camera was able to capture.  And that's my point.  If there were a larger colir space that was practical to use as a working space, but also for tagging images in order to preserve colours,that's fine. If not then it's pointless.

I'm not overly familiar with BetaRGB so don't know what advantages it may offer.

[digitaldog]

If we can believe the tech support person quoted here, DxO uses Adobe RGB (1998) as a processing color space. It has of course a fixed gamut. Colors that can be created and fall outside Adobe RGB (1998) gamut will get clipped.

An example of how ProPhoto RGB can be used within a raw processing pipeline, and the gamut implications can be seen here with REAL images:

Everything you thought you wanted to know about color gamut

A pretty exhaustive 37 minute video examining the color gamut of RGB working spaces, images and output color spaces. All plotted in 2D and 3D to illustrate color gamut.

High resolution:http: //digitaldog.net/files/ColorGamut.mov
Low Res (YouTube): http://www.youtube.com/watch?v=n0bxSD-Xx-Q

This video in no way speaks to what DxO is or isn't doing since we have no idea of the technical correctness of the tech support post concerning color management in general. Certainly their take on ProPhoto RGB and bit depth seems quite nonsensical. But point of the video, (which was generated to squash some miss understandings about using Adobe RGB or worse, sRGB instead of raw and using a larger processing and encoding space) is when the goal is to gather and possibly use as much possible colors a digital camera might produce, nether is adequate in terms of gamut size. Colors get clipped.

As pointed out below, ProPhoto RGB is simply one possible container for our image data. In no way do we need to think we can or must fill it with data. The Adobe RGB (1998) container is too small if the goal is to encode all the color gamut potential of our images.

DxO isn't the only converter to use Adobe RGB (1998) not that this is ideal. I'm pretty sure that Apple's Aperture does as well.

[Ligament]

Phenomenal video Andrew. Thanks!

[kers]

Andrew,

Thank you for the very instructive video. 37 useful minutes!
But because of that interesting lecture i have some questions…

At a ( nikon nx2) workshop the instructor suggested not to use the ProPhoto space but the (nikon) adobe wide gamut space instead.

he told me that since Pro-Photo space is so huge the jumps between colors ( even16 bit)  gets bigger than necessary- resulting in less color information in the file. Indeed 99,9 % of the printable colors (HP-Z3100) are inside the adobe wide gamut.
-what do you think of that idea?

the same reasoning could be applied to a color-portrait:
 is it better to use the sRGB colorspace in 16 bit for that purpose?

- what colorspace is best for BW ? 

- I have no idea of what colors my Nikon d800e can capture - How to find out?

Thanks in advance!
Pieter Kers

[digitaldog]

he told me that since Pro-Photo space is so huge the jumps between colors ( even16 bit)  gets bigger than necessary- resulting in less color information in the file. Indeed 99,9 % of the printable colors (HP-Z3100) are inside the adobe wide gamut.
-what do you think of that idea?

If you believe that the HP will be the last or only printer you'll ever use... The facts are, there's good and printable data we can archive from some raw converters outside of Adobe RGB (1998).

And 99% fits? I'd like to see that profile.

- what colorspace is best for BW ? 

That's an iteration from the raw or ProPhoto master.

[kers]

hello Andrew,

i was talking about the  abobe wide  colorspace not adobe 1998

here some images

[digitaldog]

i was talking about the  abobe wide  colorspace not adobe 1998

I don't know what Adobe wide color space is (not Adobe RGB (1998))

Also, the ColorSync utility isn't really the best way to be looking at gamut maps. Yes it's free. No it's not really accurate <g>. ColorThink would provide a much better and more useful gamut plot FWIW.

[kers]

I don't know what Adobe wide color space is (not Adobe RGB (1998))
Also, the ColorSync utility isn't really the best way to be looking at gamut maps. Yes it's free. No it's not really accurate <g>. ColorThink would provide a much better and more useful gamut plot FWIW.

adobe wide gamut colorspace
 http://en.wikipedia.org/wiki/Wide-gamut_RGB_color_space
and see the space attached ( nikon variant)
I also cannot find it anymore in the adobe site - possibly they abandoned it in favor of the even wider ProPhoto colorspace for good reason...
-
I am not so much into color that buying buy Colorthink would be a good investment ...

anyway - thanks for the replies..

[RFPhotography]

Adobe wide gamut aka WideGamutRGB.

This site is a good one for viewing profile gamuts and doing comparisons between two profiles.

While I'm not Andrew, I'll toss my thoughts in for what those are worth.  To suggest that ProPhoto is much bigger than WideGamut is a bit misleading.  If you compare the two on the site I linked above, you'll actually see that WideGamut is larger than ProPhoto in some areas - some greens, some reds/magentas/violets.  ProPhoto does contain some colours that are only 'theorectical' (i.e., outside of CIE LAB).  That may lead to some additional errors in encoding when doing more extreme editing but I'd think those would be fairly minimal.  When you consider that modern digital cameras are also capable of capturing colours outside of ProPhoto why would you want to intentionally clip colours back to a smaller space? 

As Andrew said, while the reproducible gamut of your printer may not exceed Adobe, there are printers that do now.  Technology improves.  If you force colours into a smaller space now then in order to take advantage of technology advances in the future you're going to have to go back and reprocess those older images.  Why do work twice if you don't have to?

[digitaldog]

I also cannot find it anymore in the adobe site - possibly they abandoned it in favor of the even wider ProPhoto colorspace for good reason...

Yes exactly, Wide Gamut RGB as it's actually named by Adobe was installed in Photoshop 5 along with other RGB working spaces. It wasn't an option after that if memory serves me (certinly by Photoshop 7 it was gone). So that web page should be ignored (even though they do reference my Adobe white paper <g>).

I am not so much into color that buying buy Colorthink would be a good investment ...

No reason. I just wanted to point out that ColorThink is one of the only gamut viewers doing the viewing correctly.** If you wanted me to plot your profile there I can do so. I can also give you the exact percentage differences in two profiles thanks to ColorThink as well.

**http://www.colorwiki.com/wiki/Color_Management_Myths_26-28#Myth_26