Lightroom-DxO Round-Trip

[fdisilvestro]

Hi,

A few observations,

1) Linear DNG output from DXO Optics Pro is not new of 9.5. It has been there a long time ago, at least from version 6.

2) The new thing is the LR plugin that allows to export a raw file from LR to DXO (LR usually exports rendered files to external editors). In the past, DXO OP had the ability to read from LR catalogues, so you could use LR as DAM, select the images, and then open the catalogue from DXO and select a specific collection. This worked with DXO 6 and probably DXO 7 with LR up to version 3. This new plugin is even better than the old way to read LR catalogues.

3)If you convert your raw files to DNG without keeping the original raw, then you cannot use this plugin, since DXO does not accept DNG as input (except those created directly in camera, as few models do)

4)The output DNG from DXO is a linear DNG meaning that it has been demosaiced and interpolated, because this is a necessary step before applying optical corrections.

5) The output DNG is less processed than a TIFF. It has not been encoded to a color space (so you avoid the limitation of Adobe RGB if you output TIFF). Additionally you can apply your preferred DNG profile and you can even create a DNG profile with the Xrite tool from this linear DNG.

6)When to use this route: If you want the Prime noise reduction and/or DXO optical corrections, which for some camera / lenses combinations are superior to Adobe's in my own experience, especially the "lens softness".

7)Note if you go this route: Apply white balance in DXO if you have blown out highlight. The explanation for this is long and has been posted in this site a while ago. If you do not do this and wait until returning to LR to apply white balance, you might end with color casts in those highlights.

You may try DXO for free for a month

[Charlene McKinnon]

you could loose a lot of valuable data

Theoretically, I suppose, although in practice - I can't tell there is any color loss / degradation.

[Charlene McKinnon]

there is also DxO Viewpoint.

Thanks for the tip. I just tried it - notes:

• it reads tif/jpg only but you have to have the raw too for the exif info it needs for module/profile-based corrections.
• it supports (module based) distortion, volumetric anamorphosis (rarely an improvement in my opinion - better to warp using Photoshop) and perspective corrections (good), but not lens vignette, nor sharpness falloff, nor chromatic aberration (too bad).

Anyway, I agree - if those are the corrections you need (and you want raw conversion by Lightroom), then it may not be a bad way to go...

[digitaldog]

When I import a NEF camera raw file into Lightroom with LR converting it to DNG on import, is the resulting file a "linear DNG?" What other types of DNG are there?

No, probably not linear but the actual raw data. Linear is an option, off by default.

[JimAscher]

Hi,

A few observations...

3)If you convert your raw files to DNG without keeping the original raw, then you cannot use this plugin, since DXO does not accept DNG as input (except those created directly in camera, as few models do)....

To my disappointment, and following a long email correspondence with the very cordial DxO customer service people, I learned that DxO does not support all DNG's created directly in camera.  It apparently does for the later model Leica's but not for my Ricoh GXR M-module, reportedly for the reason that that specific camera has been discontinued by Ricoh.  So, I can use DxO for converting raw from other of my cameras (with the exception also of my Foveon-sensor Sigma), but not DNG's from my Ricoh GXR. 

[Damon Lynch]

Hi Charlene,

I find the volumetric anamorphosis correction the most useful feature in Viewpoint, especially for faces near or at the edge of the frame. When you open a TIFF in Viewpoint via Photoshop from LR, and it's asking for the original RAW, that's a bug. I find the CA removal in LR to be sufficient for my needs.

Damon

[Charlene McKinnon]

Hi Damon,

I find the volumetric anamorphosis correction the most useful feature in Viewpoint, especially for faces near or at the edge of the frame.

I too find it useful in some cases. It's just that a lot of cases I try it, and it make the photo look better in some ways but worse in others - no net improvement, or net negative...

When you open a TIFF in Viewpoint via Photoshop from LR, and it's asking for the original RAW, that's a bug.

Could be, but it needs access to the raw file for exif metadata, in order to do module-based corrections to jpegs etc.

I find the CA removal in LR to be sufficient for my needs.

Same here - Lr's CA removal seems to work quite well so far...

[john beardsworth]

Because DxO's image corrections need to alter pixels with their algorithms for lens corrections.

What I was actually suggesting, Jeff, was that DXO might have chosen to output a regular DNG rather than a linear one, storing their adjustment parameters as metadata in the same way as LR/ACR but to DXO's own namespace. LR would certainly preserve that metadata in the file, even if it doesn't do anything with it (it already preserves other unknown metadata). I think the DNG spec also allows DXO to write their own embedded previews and speed up rendering each time the DNG is reopened in that program.

[eliedinur]

What I was actually suggesting, Jeff, was that DXO might have chosen to output a regular DNG rather than a linear one, storing their adjustment parameters as metadata in the same way as LR/ACR but to DXO's own namespace. LR would certainly preserve that metadata in the file, even if it doesn't do anything with it (it already preserves other unknown metadata). I think the DNG spec also allows DXO to write their own embedded previews and speed up rendering each time the DNG is reopened in that program.

John, I don't understand what would be the point of their doing that. You would have a fully editable DNG with part of its metadata useless to LR and part of it useless to DxO. How would you go from there to a single viable RGB image file that incorporates contributions from both converters?

[john beardsworth]

John, I don't understand what would be the point of their doing that. You would have a fully editable DNG with part of its metadata useless to LR and part of it useless to DxO. How would you go from there to a single viable RGB image file that incorporates contributions from both converters?

The value would be the fully-editable DNG would contain all its metadata, as it is designed to do (metadata not understood by LR isn't "useless"). I guess I am assuming that one would use one or other converter, not make adjustments in both. Do people really want to do that?

[fdisilvestro]

The value would be the fully-editable DNG would contain all its metadata, as it is designed to do (metadata not understood by LR isn't "useless"). I guess I am assuming that one would use one or other converter, not make adjustments in both. Do people really want to do that?

Yes, that's the whole purpose, you might perform optical corrections and noise (prime) corrections in DXO and the rest of the edits in LR (color, tone, etc).
One specific example: for the nikon 70-200 VRI, the optical correction from LR does not compensate uniformly for vigneting, leaving one or two corners darker than the rest. The corresponding module of DXO works just fine.

Regards

[Fine_Art]

I've been playing around with DxO Optics Pro 9.5's new trick: the ability to import a raw file from Lightroom and return a linear DNG.  If I understand this properly (as usual, the explanatory material on the DxO website is less than pellucid), by emitting a DNG, DxO Optics Pro allows a Lightroom user to take advantage of DxO's sophisticated optical testing and correction technology while still retaining all the functionality of LR—except, of course, Adobe's demosaicing algorithms.  For the moment, at least, I'm only interested in DxO's automated sensor-and-lens corrections, with all additional processing performed in LR.

I'm encouraged by the initial results.  I've attached a small center crop from a much larger image made with a Nikon D800E and Nikon's 24-120mm f/4 zoom.  The first attachment is a 1:1 JPEG of the raw file as demosaiced by LR after applying LR's automatic lens corrections.  The second attachment is the same raw file demosaiced by DxO after applying DxO's automatic corrections for the sensor and lens.  I didn't do any manual sharpening or make any other corrections with either tool since my objective is to compare how the two perform the raw conversion and correct for known optical imperfections in the capture.  I've only worked on a few images so far, and only those shot with this particular sensor-lens combination, so it's too soon to come to any definitive judgment.  But in every test I've made, the DxO image looked better than the LR equivalent—sometimes dramatically so.

My only concern is that I may be giving up some of Lightroom's functionality by having DxO perform the conversion and turn over the linear DNG to LR.  It appears LR can still properly modify the white balance of the DxO-emitted DNG.  Is there anything else I might be giving up, or should otherwise be concerned about?

Why does the LR version look so mushy compared to the DxO version?

[digitaldog]

The value would be the fully-editable DNG would contain all its metadata, as it is designed to do (metadata not understood by LR isn't "useless"). I guess I am assuming that one would use one or other converter, not make adjustments in both. Do people really want to do that?

How is metadata that LR can't understand anything but 'useless' to it and the end user unless they render? How can one move from raw converter to raw converter with raw data and two differing sets of instructions only one converter can understand?

[john beardsworth]

Imperfectly, but possibly of value, as C1 has shown by reading LR adjustments.

Storing more than one raw converter's adjustments in the DNG metadata is what the X in xmp is all about.

[Chris Kern]

Why does the LR version look so mushy compared to the DxO version?

I'm not sure I understand your question. One of the points I made in the post that started this thread was that the rendering with default DxO optical corrections seemed to produce a better initial image than the Lightroom rendering with Lightroom's lens corrections.

In both cases, the sample images were produced without any manual capture sharpening.  Even though the D800E effectively has no AA filter, my experience has been that some capture sharpening is still necessary.  The reason I want to emphasize that I performed no manual capture sharpening is that one of the mainstays of the DxO optical correction suite is a correction for "lens softness."  I don't know whether Lightroom's lens correction module for the lens I was using includes a similar property.  The samples I posted represented a small center crop of a much larger capture file (see attachment) so small differences in edge acuteness are noticeable.

Again, my experience is that DxO often seems to provide a better starting point than Lightroom for the camera sensors and lenses I use.  That especially seems to be true with the D800E, perhaps because the sensor's high resolution reveals optical defects in a lens that wouldn't otherwise be as objectionable and DxO's sensor-lens correction model does a superior job of neutralizing those defects.  I haven't attempted to perform any rigorous testing to determine whether additional processing with Lightroom would compensate for DxO's initial benefits—and I'm not sure any generalization is possible because there are so many variables that might need to be considered: output format (i.e., print or screen), degree of cropping, size of final image, etc.  I'm working on the assumption that this is something that needs to be determined on an image-by-image basis.

What I've tentatively concluded after reading all the responses in this thread is that I'm not giving up much by having DxO perform the initial raw conversion and then continuing to process the DxO-emitted linear DNG in Lightroom.  And of course the original raw file is always available, so I can still have the option of Lightroom-only processing if I'm not satisfied with what I get from DxO.

[digitaldog]

Imperfectly, but possibly of value, as C1 has shown by reading LR adjustments.
Storing more than one raw converter's adjustments in the DNG metadata is what the X in xmp is all about.

I'm all for saving all the data, but what do you do with it? I have a raw file inside a DNG with two sets of instructions unique to each product which renders the data. I've got C1 and LR xmp. Now what and in what order to utilize both on the raw data?

[Chris Kern]

Storing more than one raw converter's adjustments in the DNG metadata is what the X in xmp is all about.

I'm not certain if this accomplishes what you're proposing, but DxO Optics Pro writes a sidecar file containing its proprietary metadata into the directory where the raw image is stored.  The file doesn't have a dot-x-m-p extension, presumably to avoid overwriting any sidecar for the same image created by Lightroom.  (It's saved as <raw file prefix>.dop.)  The purpose of the sidecar apparently is to permit you to resume processing in DxO wherever you suspended it in a previous session.  In other words, DxO doesn't apply its default adjustments to the raw image, but rather the adjustments you were using when the previous session ended as documented in the sidecar.  The only purpose of DxO's linear DNG, as far as I can tell, is to give Lightroom a partially-baked file to use when you continue processing in LR.  DxO doesn't use the linear DNG at all; it always operates on the raw image.

[john beardsworth]

DNG files don't need sidecars - it's an advantage of the format. Instead DXO could write to their own section of xmp inside the file itself, and read it when resuming processing. But this capability is different from the desire to process Dxo results in Lightroom.

[Chris Kern]

DNG files don't need sidecars - it's an advantage of the format. Instead DXO could write to their own section of xmp inside the file itself, and read it when resuming processing. But this capability is different from the desire to process Dxo results in Lightroom.

DxO doesn't use the linear DNG.  It operates on the original raw file.  I guess it could read metadata from the linear DNG instead of the sidecar, but I'm not sure what the advantage of that would be, since if and when processing resumed, DxO would again be applying its corrections to the original raw.

[john beardsworth]

I think we're going round in circles here, Chris.