The DNG from Photolab (and pure raw) is not a ‘complete’ picture. It still needs to be processed to some extent . So whatever application you use to read the DNG , it applies processing to it like a raw developer.
The tif on the other hand has no room for interpretation, it pixels are ‘final’.
If you open the DNG in Lightroom, affinity photo, Capture 1, Rawtherapee, Darktable … It will look different in all of them , just as a raw photo would look different depending on the software you open it in.
A raw file has no 'true, correct rendering ’ (although you could argue that the rendering your camera did to write the jpg file could be the one ). It is always sensor data up to interpretation.
A DNG file written by Photolab or pureraw is no different. Yes, some steps of the raw rendering is already done , and some of the data is altered by Photolab, but a lot of the interpretation still needs to be done . Camera profile / color matrix / basic tone curve just to name a few.
Everything changed with PL4. The DNG produced using the “Optical corrections and NR only” is slightly different from the DNG produced using the “All corrections” option. The WB is left completely in the camera’s sensor space just like in the RAW file. The DNG produced using the “All corrections” option will set the WB to the value set by you and it will set it to the generic Camera Default setting even if you turn off the WB tool similar to what happens when you create a TIFF or JPG. The DNG produced by PureRAW is similar to that produced by PL5 using “Optical corrections and NR only”.
DNG files written by DPL and DPR have been demosaiced (roughly 3x the size), denoised (depending on settings, e.g. DeepPrime) and have had DxO’s colour matrix applied afaik. DNG output looks fairly close in Lightroom to what I see in DPL, except for saturated colours which tend to be shifted by Lightroom’s innermost algorithms.
Anyway, Looking at customized raw files (or DPL’s DNGs) in other RAW developers will show slight differences based on what the manufactures think is the “best” rendering, which means that I can either like what I see - or change it, which does not really matter, unless one works to produce close to life reproductions.
I have found this discussion very interesting but also mystifying. This is because I never go to DNG, using tiff or jpg, and print straight out of DPL or DPR. I just don’t have all the patience so many of you obviously have to work on images to any great extent. I try to get them the way I want them in the camera.
But I must say that I am truly impressed by the one-stop processing of DPR.
On a recent cruise up the Norwegian coast, Northern Lights, moon-lit coastal villages, mountains lit by the famed “Norwegian blue” light or minimal daytime shots came out startlingly good from DPR. I love it. Does anybody really need DPL?
I have become convinced that it is easy enough to use DPL with some limited settings to almost exactly duplicate the DPR output, and to do so creating jpg, tiff or dng rather than just the jpg-dng choice DPR gives you. In addition you get the ability to do much more if you want to.
As I remember the price for the DPL upgrade was less than the price of DPR (new, so without an upgrade price), so it became a choice of paying more for less functionality or paying less for more functionality. An easy choice.
As for me, at the moment I am using DPL as a pseudo-DPR and doing any additional processing in a pixel editor. I am not sure I will continue doing that, so time will tell.
No reason to be embarrassed about asking questions about things you are not familiar with. That is how we all learn, and I was asking the same question when I first starting editing photos.
Photo editors come in 2 general “flavors”, general editing tools like Dxo’s PhotoLab, Dxo’s PureRAW, CaptureOne, Adobe’s Lightroom and others that basically provide enough functionality so that editing photos becomes a process of moving sliders or other controls to change specified parameters to change how a photo looks and a set of other tools that are more “down to the basics” in how you edit photos.
The first type allows editing with simple tools, many of them automatic or semi-automated, and the user does not need to know what is really going on under the covers. PhotoLab is like that. If you want to change the exposure of an image, move the slider. If you want to adjust the color, move the slider, if you want to create a mask to control which part of your image gets adjusted, click on a point and adjust the circle to include what you want, and then move a slider. And just because they are simple and easy to use does not mean that they are not incredibly powerful.
The second type requires that the user get more involved in what he or she is doing. An image is loaded and each time a user wants to make an adjustment he or she adds a specific adjustment overlay, called a layer, to the image. A user may end up with 2 or 3 or 10 or 20 or 100 such “layers”, each controlling one adjustment and each able to specify which part of the image is to be adjusted. It is more complicated, but also generally more powerful. Examples of this type of editor are Adobe’s PhotoShop, Affinity Photo and others.
There are probably far more of the first type of photo editor than the second as the first is easier to use, but the second allows the user more flexibility in what is done as well as offering controls that are generally not available in the first type of editor. For example a pixel editor may allow the user to drag parts of the image to control how the image looks, allows the user to control which points of the brightness curve are to be considered to be at the ends and a lot more.
This is just an overview and if you want to know more you can download trial versions of some of the pixel editors, but often the general photo editors can do what 90% of users want, and in a much easier fashion.
One more thing that I might add is that the differences between the two types of editors is shrinking in that each has started adding features from the other as the competition between them increase. Some examples.
Adobe Photoshop is a well known pixel editor, but it includes a front-end app called ACR (Adobe Camera Raw) that provides much of the same general editing functionality as other general editors, and ACR is also available in Photoshop as a plugin so it can be used for all photo types, not just raw images.
Skyrim offers a series of general editors called Luminar, some of which allow the use of layers, which used to be a pixel editor only functionality.
Photoshop now offers the functionality to create . panorama and merged photo combinations which used to be a specialty type of functionality.
Dxo’s PhotoLab offers “Local Adjustments” which gives users much of the power of pixel editor layers.
General editors now often allow the user of external apps called plugins which provide specific functionality (sharpening, target color adjustments, masking, “artificial intelligence” adjustments) which used to be limited to pixel editors.
In short users can do most things now in general editors that used to be limited to pixel editors and pixel editor users now have simple ways to do much that general editors used to provide. The one thing that pixel editors provide is that they are often far less expensive than general editors. I use PhotoLine, and the upgrade is $29. Photoshop is available for lease at $10/month. CaptureOne is a very good general editor, but is very expensive and Dxo’s PhotoLab, while also very complete and very good, is also pretty expensive.
Thank you. But I rest my case ……I get the image I want mostly in camera. I haven’t the patience or the interest in going further than with the DxO products. Their output produces the prints I, and others, want.
I think before they created a “rawDNG” for plv4 i think, your right.
Most DNG’s are infact lineair pixelbased file like a tiff (denoised and demosiaced and every correction you made in dxo) and have adobeRGB or sRGB colorspace and there fore a set WB. Which causes colorshifts when you open the exported DNG in an other rawdeveloper.
The RawDNG does keep the camera’s colorspace and has every possible WBpoint in the filedata. No cut off or compression of the capured spectrum.
Because of that it’s only optical module and denoising both are working before demosiacing is done.
Demosiacing, which you know cramps camera colorspace in the adobeRGB which is the working colorspace of dxopl, in order to preview it on your screen.
Because wb is set you can adjust lumination and saturation as edit possibility’s. Same as repair and clone.
So what i think dxopl rawdng is doing, it applies optical module features and deepprime, and then just frates a fullblown tiff like file. It is demosiaced but not compressed in the AdobeRGBcolorspace so WB is still a floating point with all available colors that the camera captured.
There is a list of pre-demosiacing corrections dxo does.
One of that is chromatic Abberation correction which is effected by WB.
So IF you had a major colorcast by streetlights or such it’s advised to change WB before export in order to let CA corrected accordenly. (if those settings get used by rawdng export that is.)
Me, i don’t use DNG often, i prefere 16bit tiff, i use dxopl as main developer and only sent out a file for things dxopl can’t deliver. All colors and such are applied and i get less surprizes in colorshifts.
Best working type is let dxopl create rawdng’s so you have less waiting time when you edit on them. (no spinningtime due working optical module).
I haven’t test it but i assume if you copy paste all corrections from dng to raw when you done and want to create end result, you get the best of both.
Working speed in dng and maximal optical correction and denoising on rawfile.