Basics in RT: Used dynamic range compression among others.
Then to Gimp, dodging and color toning.
3 Likes
20190809-RickettsGlen-0006.CR2.xmp (6.8 KB)
My take in Darktable 2.6.2. Nice photo but very underexposed. I went for triple fusion with the base curve + an exposure correction (thatās how underexposed this was). That got the curve ballpark to where it should be. After that, a bit of shadows and highlights, and filmic - I rarely use base curve and filmic together but seems to work here.
You could go more crazy on these shadows; thereās a lot of detail there. But it detracts from the main feature, which is the water fall.
Other than that I did a bit of noise reduction, sharpening using the equalizer and some tweaking of the a and b curves to get some color back.
This combined with an alignment phase is a part of Googleās processing pipeline that they use as part of their āHDR+ā mode.
However their pipelineās alignment functions might behave badly around the waterfalls. Not sure if you had a tripod - if you did, then alignment would not have been an issue and you could have done a simple averaging stack. That reminds me, Iāve got a tripod burst series of shots from Buttermilk Falls in Ithaca I still havenāt stackedā¦
Lighting can be brutal in Ricketts Glen - you frequently have direct sun visible on some leaves in the frame, but most of the rest of the shot is forest shadow, so no single white balance setting will make everything look nice. It doesnāt help that the first time I went there I forgot my white balance reference, and the second time I went there I had family start being a PITA any time I took more than 10 seconds for a shotā¦
@Entropy512 Yea Ricketts Glen is very tough to shoot. It was my first time there and the day was far from ideal. Thankfully I only live 1 hr away so it is quite easy for me to go back whenever I wish. I do feel blessed to live so close to such a beautiful place.
I do find it quite amusing how at Ricketts the other hikers stop and stare at you while you are setting up a shot and waiting for decent light for like 20 mins. Then they just snap a shot with their cell phone and just move alongā¦ I canāt help but wonder what they are thinking while watching me setup a shot.
Yes I do have a tripod a pain to have to keep setting it up but so worth having and using.
Tuesday seems like decent weather I might head back up there and check out some different spots. Trying to scout it the best I can for the Fall.
You want hard, try doing a multishot panorama at Waters Meet on Memorial Day weekend.
(One of my favorite shots, which is often my desktop wallpaper, is a pano made with 3 exposure bracketing and stiched in Hugin then fed to enfuse. Sadly I set the center exposure of the bracket too long, and even the shortest exposure had highlight clipping. That shot involved a LOT of waiting for the current aim point of the camera to be clear.)
For me its 2-2.5 hours depending on traffic. The first time I went there, on my way back I got stuck behind a box truck doing 40 in a 55 the whole way - except for the ONE passing zone where he sped up to 60!!!
I have never been there in the fall - Iāll likely be out of town for most of the peak colors, and even if iām in town Iāll likely hit one of the gorges in Ithaca. (I live an hour from Ithaca).
Edit: If you havenāt already, Iād strongly recommend getting a trekking pole (Leki or similar), itās a wonderful thing when navigating those wet stairs. Well, itās a wonderful thing when hiking in general. Theyāre frequently sold in pairs, Iāve found that having two can get in the way when youāre also trying to use a camera, but one is a huge improvement. Leki sells a version (the FS) with an integrated tripod bolt, but Iām not a huge fan of it. The design tradeoffs that make a good shock-absorbing trekking pole make it a really poor monopod (extremely bouncy). You will get a lot of āgoing skiing?ā comments on the trail thoughā¦
I could only imagine holiday hellā¦ I tend to avoid going out on holidays 
. I have seen people with those trekking poles will for sure consider getting one will be very useful at Ricketts for sure some really steep slippery paths and stairs.
My main priority is finding a better lens for my T7i that is affordable as well as figuring out a way to go pure open source for my photo editing. LR/PS are so good and run so well on my Mac it is hard to find other options I am content with. Really tough to move away from something that works well
Hah, yeah! I know the pain of trying to migrate to a new workflow after youāve āgotten usedā to another one.
I primarily started moving to darktable because compressing dynamic range with exposure fusion | darktable was a significantly less labor-intensive version of a technique I used extremely frequently. Unfortunately that particular implementation has some serious limitations/issues, and a decision has been made by the darktable development team to simply let that feature wither and die as ādeprecated, only kept for compatibility with old editsā rather than allow it to be fixed, because in extremely rare situations it might fail and cause a minor halo (a behavior Iāve not seen in 2-3 weeks of working with a āfixedā version of the module for my own personal needs).
So I am finding myself where my workflow needs cannot be met without either an extremely labor-intensive process or using out-of-tree unmerged patches against dt, so I need to find an alternative workflow that has the potential of meeting my typical workflow needs in the future. For the time being Iām clearing my massive backlog of photos that need to be worked on using out-of-tree patches before I switch workflows and develop another huge backlog.
You got lucky with clouds to handle potential scene dynamic range challenges, so Iām not going to try and do my own ātakeā on handling your particular scenario other than give tips from someone else who has hiked the Glen specific to the capture part of workflow including making it less likely youāll slip and fall off of wet stairs. 
Handling color variation due to differing illumination within a scene has never been my strong point - which is not a good combination with my tendency to hike through the woods!
Where is the decision handed down? I donāt see anything like that in the PR.
My original PR, or Edgardoās PR (which I fully supported as splitting the functionality into a separate module and moving it later in the pipeline WAS a good idea) which is referenced in discussion near the end of mine?
Because it was most definitely handed down in Edgardoās PR. Iām fine with objecting to the remaining implementation issues (Edgardo jumped the gun on requesting code review, youāll notice I had a number of review items against his PR that were unaddressed), but the module was Havoc Penningtoned from orbit based on mere concept. It was also clear that a third party who came in to object had not bothered to read the code or understand how it ACTUALLY operated, since all of his objections appeared to be based on the typical āoriginalā use case of Mertens algorithm, which generated a tonemapped output from the original camera inputs (preventing further processing from occurring in scene-linear space), as opposed to the intended use case of exposure fusion as described in the link above, which is to tonemap the output near the end of the pipeline. (As to the constant ābasecurve is too early in the pipelineā argument - it no longer has to be since iops can be reordered. In every since case where Iāve used fusion and/or basecurve, they are at the end of the pipeline with nothing after them. Iām A-OK with ensuring that they default to ānear the endā because thatās how they should be usedā¦)
It looks like you have another PR open that seems to be going OK.
Just because people disagree with you doesnāt mean itāll get rejected, and Edgardo closed his own PR, so I think that means you all should try again.
Thereās absolutely no metric by which my approach is better than Edgardoās was. The only reason it appears to be āgoing OKā is because all further discussion on the future of fusion in darktable was on Edgardoās PR (because it was better in almost all ways, and absolutely better in terms of concept of splitting the functionality out to default later in the pipeline. The only issue with his approach are some code structure approaches that may have had negative performance impacts, but these were NOT cited as the reason for nuking the effort. After all, those issues could have been fixed.). Thereās absolutely no metric by which my PR would get merged when Edgardoās was refused, given the rationale given for why his got nuked.
When I say that Iāve been using a fixed form of fusion for 2-3 weeks without halos - Iām referring to Edgardoās implementation with a couple of minor tweaks, not my original PR. In theory I could take his work back into the basecurve module, but that is not the right way to go under any metric I can think of.
His approach was (other than some minor low-level details), simply better. Splitting the functionality out into a different module is THE way to do it.
Edgardo closed his PR because it was made very clear that there was no point to expending any further effort given that it was nuked without any actual testing or discussion to verify the claims being made. The only reason mine is still open is because I havenāt bothered to close it.
This has nothing to do with the Play Raw. Please start a new thread or split this.
1 Like
But none of the people that commented on the thread have veto power in the project.
Waiting until the time is right for a code review is a good thing, and I think you all should try again once something usable has been committed.
Per Afreās comment, Iām replying to this with a PM.
I love seeing the different takes people had on this image. It is a great example of why youād want to shoot in raw instead of letting the camera automatically choose the one ābestā interpretation of the data.
I particularly enjoyed hearing what software and techniques people are using! Itās great for people like me who are only just beginning to learn.
While to my eye the LightRoom version has a lot of appeal, I have to admit that after a decade of seeing the hyper-real results LR can achieve, my brain is a bit fatigued. I think my eyes are ready for the 2020s, when hopefully photography will swing back to appearing a little more ānaturalā.
blj, if you donāt mind, could you please list the exact steps you did in LightRoom so the experts here can tell us how the same process could be done in Free Software? Thanks!
@hackerb9 Welcome to the forum btw. I can do my best to explain what I did in Lightroom. First and foremost I am not a fan of the crazy fancy edits people tend to do with Lightroom/Photoshop I try to keep my edits minimal and more natural looking.
That particular image I did in Lightroom does not have much done to it. I did a new version uncroped that I did more work too that looks much nicer but did not post here. As this forum is dedicated towards open source. But either way that first image I did a few slight things to it.
-
Cropped the image to a square crop. Thought it would look good at the time but I much prefer the uncroped image after doing a second pass at it.
-
The image is underexposed due to weather circumstances and having to predict my exposure. I did not have much time to tweak it so I did end up under exposed but modern cameras are pretty good at keeping the detail anyway in a raw. I wish I slowed the shutter a bit more. So to compensate I increased the exposure by 2 stops in Lightroom.
-
Cooled off the image a bit. The auto white balance was a temperature of 6600k which I felt was too warm so I pulled that down to about 6000k
-
Pulled down the Highlights and pulled up the Shadows with the highlights and shadows sliders to recover detail lost in the bright and dark areas of the image. Doing this makes the image appear flat because it is pulling the histogram extremities closer to the mid tones. This will be corrected in the next steps.
-
Set the white and black point of the image with the whites and blacks sliders. (I donāt think these sliders correspond to what it means in the open source softwares. In Lightroom I believe they just operate similar to a tone curve just a quicker way to do it because in Lightroom you canāt stack curves like in DT). For me I tend to set the whites to just under where they clip. The blacks I always let clip to personal taste as I like deep shadows spattered through the image as it adds depth in my opinion.
-
Added some vibrance with the vibrance slider I prefer this over saturation as it makes it much harder to over saturate the image as in Lightroom vibrance only pushes colors to saturation not beyond and in the case of people that slider protects skin tones where saturation slider does not. Mainly a personal preference. I wanted to add some more color pop.
-
Added a touch of contrast as well as a smidge of clarity (similar to a local contrast) and some texture with the new texture slider. I love the new texture slider works great on landscapes makes the details crisp without having to add sharpening.
-
Local adjustments were done with the brush tool as well. The right Rock Wall I bumped the exposure by another 0.10 and pulled up the shadows some more to get some more detail back.
Also on the right the part of the wall where the waterfall is cutting through the sun hit nicely so I brought that up by 0.20 on the exposure to keep it nature but get it to pop more. -
Lastly I added a vignette to the image to help pull the eye in more to the falls.
Again not my favorite edit I prefer my second go uncroped. I also did a bit more work to my uncroped version which is not posted here. But those are the steps I did on the LR image in my OP.
If anyone wants me to post my uncroped Lightroom edit to this picture let me know in a ping or pm or something and I will be more then happy to put in this thread. It is up on my Flickr page as well I think I have that link in my profile.
1 Like
