There’re 2 kind of image editors: pixel based and vector based.
A pixel editor works on a predefined image of x by y pixels by changing the pixel values. This can be done on individual pixels level or group level. The essence is that individual pixels are changed.
So I think a raw converter exists out of the conversion part, the creating of an RGB raster image and after that just the pixel editing part. Some tools can be greyed out in the converter or handled different when using a jpg like white balance or exposure correction.
I myself shoot pictures for their content. I try to archive a nice picture with the camera. A sky replacement is replacing a part of the content. In journalistic en documentary photography strict forbidden.
Ok, lett me refrase myself.
There are raw applications which can stack images and blend then together like This one
You can change composition details by pasting two to six “workspace previews” on top of each other and create a new scene and there for i call this a hybrid. It’s equiped with pixel editing software but in the base jt’s a rawdeveloper.
Photoshop is the fullblown magicbox you can create a image by pulling stuff out other images and never go out to make the image yourself. You can edit pixels as in fill in a blank canvas. I call this a Pixel editor.
DxOpl is a rawdeveloper with some need tricks , which uses pixel manipulation software, but i my eye’s it’s not a pixel editor as in replacing and altering a scene out of reality. Granted repair and cloning and colorreplacement is also altering reality but hence every adjustment is altering reality.
Replacing a sky is photoshop area, enhancing sky structure, color and exposure masking objects out is developing skills.
Maybe DxOPL follows the rawdeveloper above and create some tools to stack and select in layers maybe not. Personally i find it difficult to judge if i need it or not.
If it’s there i want a good working tool not a “blunt shizzle” which ruining my work.
When above’s is brand selective and reduced in price i think i am tempted to find out. As in it never hurts to look over the fence to compare.
But my main developer wil stay dxopl, the improvements it does on optical level and noisereduction, which is key with m43 sensors, is outstanding. There toolbox is in the last two years expanded in a good direction so i am confident it will serve me well.
@OXiDant
Basic we say the same.
Every pixel editor is manipulating pixels, but when we talk about the difference between a rawdeveloper and a pixeleditor we refer to the fact that that pixeleditor can change the content of the image, is more developed for that. Working on the sky in PL doesn’t replace that sky and changes the content. Replacing does change the content. I would say leave that to other programs.
Because using stock skies is how most people will likely use it. So far, the only Luminar sky replacement images I’ve seen anywhere have used the included stock skies rather than images captured by those using that feature.
I thought you could not use your own sky in luminar 4 but only those stock you get with the software or buy on their store.
Correction:
Now you can instantly change the sky in your pictures. Smart algorithm even selects the best orientation for the sky. You can use many pre-installed skies, or easily add your own sky photos.
I totally second your opinion. You can get acceptable results with the tools we already have and in a reasonable time.
A luminosity mask would have made the process much easier.
As a personal opinion: this sky color only exists for a minute, a bit less saturated, early in the cloudy morning, like this it looks a bit dodgy to me - also because of my pain trying to edit so that I show the proper color of that sky in the morning.
your right, i was pushing it to the max to see if it’s holds credibility.
saturated too much. too edgy trees. looked a bit if i got a UV polaroidglasses on of a dodgy salesstore LOL, here the rawfile _1060609.RW2 (18,9 MB) have fun with it
What this all comes down to is whether it would be prudent for DXO to expend resources developing a feature similar to the one already in Luminar. My opinion is that it would not be a good use of their limited resources. There are many more important feature upgrades and new feature requests in the backlog that should take precedence over special effects like sky replacement.
Just bought Luminar 4 (with sale + discounts, £45, silly not to) and find the sky replacement is pretty good.
ON1 and Affinity can also cope with it, albeit a little more tricky and more work.
Quite agree that I also would like DxO to focus on bringing in luminosity masks, colour pickers and such like and leave the fancy image editing stuff to other programs, of which there are plenty.
Not so, and not what Luminar recommends: “While the filter includes sample skies for you to use, many will want to use sky photos of their own creation”. Good advice how to capture your own sky pictures (and how to properly use them): https://manual.skylum.com/4/en/topic/ai-sky-replacement-guidelines-2 No idea (and of little relevance here) how “most people” will use it.
Would be banned from contests if I didn’t use my own skies!
The ones supplied by Skylum are fine, but for many people they will simply be OTT and look ridiculous.
I have stock sky images I create of various sizes and both orientations and as long as they are the same resolution and approximate size (to avoid stretching pixels to fit) as the area you want to replace they will be fine as I have already found. Certainly far quicker than other software, and you can blend the sky just as you would in any other program.
Below good article “Fast Sky Replacement” from Amateur Photographer. It also shows many examples using non-stock sky images for people who did not see any. This is clearly rather powerful tool which, when used properly, can produce excellent results (this statement applies to just about any photo software!).
Now, should DxO use their somewhat limited resources to go this path? I am not sure, it may or may not be where they want to go with their software. It really depends on how limited their resources are.
True as in powerful tool but those examples showes also that for me as a amateur photographer it is a rather brut change of captured moment and over steps my feel of artistic freedom of enhancement.
getting rid of disturbing objects? yes.
changing WB/colors to my memory/liking? yes
altering tonality/Dynamic Range of a image to see the details in shadow and highlight? yes
blurring surroundings to point out a subject? yes
changing a hole image by adding a sky which wasn’t there?
hmm, only when the image is disturbed by a sky which wasn’t captured well but in reality was there recovery so to speak…
So no cloudy stormy sky’s wile it was a sunny quit day. (the flipover showes the problems, shadows and lightfall is too difficult to match on the ground.)
So if they make a tool to replace blown sky for a “blueisch realistic sky” we would see if we where out there in a second. YES PLEASE, much faster then what i did.
But pixelphotoshopping library of sky’s to make your own reality. No not for me, i like my memory alike the image and i am not selling the image so for that i need no improvement.
that’s my honest opinion others may think otherwise. That’s fine.
The difficulty for DxO, Mark, is that there are many conflicting opinions as to which outstanding feature is more important … such that it’s impossible to satisfy all expectations.
That is correct, which is why I qualified my statement with “My opinion”. However, while I have no idea what DXO’s thinking is on this subject, I suspect that a feature to replace skies would not be a high priority for them. But, I’ve been wrong before.
mine looks processed, colors are “enhanced” yours looks like a “blue light moment” more natural kind of light. But my mind is telling me too blue in bushes and/or grass.
mine and your mid one.
mine starts to scream…
Hmm, i think i need a v3 of my processings version. less vibrance or less contrast/clearview.
And i think i need to read in on splittoning use.as i said i tried it , didn’t catch the use and left it.
split toning is the instagram filter basically. I think it’s necessary to create the warm/cold contrast in a good picture, if you were there at the wrong time of the day. The art in using it is the balance slider, I believe.
And actually you can do everything with using layers or local adjustments but it’s often a question of speed. Which one is faster, so that you have time for other things. And that’s why I think PL should finally offer a luminosity mask…
I am personally not in favour of a sky replacement tool. What I like about DxO, Capture One and LR Classic is that they provide me with the tools that I need so that I - and only I - decide how to use them to correct and enhance my photographs. All this AI that you can find in Luminar is the very reason why I personally don’t like Luminar.
There’s a market for each of us and for different kind of softwares, DxO is just not that software.
- also because of my pain trying to edit so that I show the proper color of that sky in the morning.
