Scene referred, display referred, and the Colour Science of Canon/Fujitsu/Nikon/etc. in Darktable (filmic-rgb module in particular)

[Matt_Maguire]

Wow, there are quite some misconceptions in this post, where to even start… Much of this has already been discussed at length in this forum. Let me make some general remarks:

• base curve, filmic etc. are primarily tone mapping modules, and are generally not concerned with colour grading.
• filmic is designed to work on scene-referred linear data, and it emulates the sort of response that traditional film has to luminosity of light. It provides a parameterised curve to allow quickly and easy scaling of this transfer curve. It assumes that the mid grey point has previosuly been set appropriately by an instance of the exposure module. If you set the default exposure level to something appropriate for your camera, are using your camera’s auto exposure features, and the scene you are taking has a reasonable spead of tones, then generally you’ll get a reasonable looking neutral image out of filmic suitable for further processing. If you need to scale the transfer curve, then there are sliders available to make this fairly simple.
• base curve also generally operates on linear scene-referred data. However, here you have to manually adjust control points on the curve to fit it to the characteristics of your image. Again, if you are using auto-exposure on your camera, and your image has a common distribution of tones, then you will probably get something reasonable. If you have to tweak the curve however, then this is something that can be fiddly and may not be reliably repeatable.
• If you want to make local tone adjustments, then there is the tone equaliser module that can help you do this, and it provides a guided filter to help preserve local contrast which is not something you will get with basecurve alone.
• It’s not entirely clear what you mean by “colour science”, but if you are talking about colour grading, then any “presets” would generally not be applies at the tone mapping level, but rather in modules such as color calibration or color balance.
• to do effective colour grading, you generally want to start with a fairly neutral image with some basic level of white balance, so you are starting at a known baseline. Then, on top of that, you can implement secondary colour grading to implement some sort of creative “look”. If you start with an image already has some sort of look applied (eg. to simulate an OOC JPEG), then you are potentially fighting against this implicit grading to achieve the final look you want to achieve.

This response is already much longer than I had intended; I’m really not saying anything new that hasn’t already been discussed many times before in thisw forum.

[priort (Todd Prior)]

@OK1 I suspect you will get better direct feedback if you avoid writing a several hundred word prelude to your actual comments.

Having said that I do have a few presets but they are not for what you are suggesting. I have one for a very neutral setting of filmic as described in the manual and then I have a couple that turn off a couple of the color corrections and basically that is it for me. In the beginning I fiddled with filmic but once I understood it as many others here have suggested its really mostly about setting and or correcting your exposure and then just adjusting (mainly for me anyways) the white relative exposure and the black relative exposure.

Images can tend to be a bit flat or dark but I always pair filmic with the tone eq to bump the shadows and local contrast and color balance for further color and contrast. There are things you can tweak in filmic but for most of them its rarely necessary and it leads to people trying to get too much out of filmic often with frustration. If you use the EV display its pretty clear what you are doing when you move the sliders so if indeed you do understand the lighting then you should be all set.

Any preset that attempts to provide a look in filmic will likely require just as much tweaking as if you had not crafted any preset IMO.

You could look at the experimental sigmoid module if you are looking for something simpler.

New Sigmoid Scene to Display mapping - #196 by priort

[johnny-bit (Hubert Kowalski)]

Everything takes time to learn and even more time to master.

If you want to learn, go there and post concrete questions and your problems:

Let's learn Filmic RGB! Your one stop shop to understanding filmic-based approach to edits!

Hello Dear Students and Teachers! After recent influx of posts and posters saying in unison that they have problems with filmic I thought we need a place for learning. A place when people “having problems” learn how to solve them and a place where we can learn that those problems are, so we can improve our knowledge about real world usage. So this is this place! If you sometimes think “filmic is too hard” or “I can’t achieve the look i want” or “it takes forever to get close to what i think i …

[OK1]

Long story short. Sadly, going forward, I can only use Darktable for special cases, where I’d like to deviate pretty creatively, from the out of the camera jpg.

For about 10 months, all my edits have been in darktable, and done my best to learn it and produce great results with it, but in some cases, like a portrait photo I took recently, it was such a herculean effort with darktable.

My big mistake was after several weeks of editing this set of photos, in darktable, some which turned out pretty nice, I made four inadvertent comparisons.

1. Took a look at the out of camera jpg, which is embedded in the raw file (you can extract this using a variety of tools downloadable from the net)

2. Tried editing a photo using Sony’s accompanying editing tool. I have owned a Sony for a few months but never bothered to use their editor, as I had stuck to darktable. One thing was pretty clear, the Sony editor was as identical in results (if I did not change any parameters) as the out of camera jpg, as I can imagine.

3. Installed Adobe Photoshop Express, which is where I started a bit over a year ago, and moved from this to darktable, and the results I was getting with Adobe Photoshop Express were quite close, especially in the level of “detail” that was retained in teh sharpening of the image, when compared with all the sharpening options I have exhaustively tried in darktable - highpass, contrast eq, sharpen modules. Pretty similar and retaining same level of detail as with the Sony Editor - Imaging Edge.

4. Installed the Capture One version for Sony, and while being a more difficult tool to use, and learn, than the aforementioned, the tool seemed more “photographic” in its end results.

Without directly comparing with other tools, which I never really bothered to do until I had exhausted all my options in darktable, darktable is a commendable effort.

But IMHO, when I have compared the end results, losing much sleep with darktable over the last month or two, unfortunately results which I cannot share publicly as these are photographs of people, I have to admit, like it or not, there is a huge amount of “inteligence” that has gone into the tools that are not open source or freeware like darktable, i,e the free tools from Adobe(Photoshop Express), Capture One, and Sony, which make a compelling case, for most people to use these, instead of darktable, for a lot of their images.

I still think that there is tremendous value in shooting raw(and moreso raw + explicit jpeg generation - as distinct exportable files - in camera), especially with the knowledge that if ever there were improvements in color science, or image processing in the future, much like we have remastered music, these raw files are our own negatives from which we could develop even better or more fashionable or pleasing results, whenever technology catches up with the abilities of human vision.

We could imagine future where display technologies and maybe printing technologies have advanced to show us more colours and support higher bit depth output, which is already taking place with HDR, 10 bit and who knows one day, we’ll have much more capability on the output end of things. In audio 16 bit was de-facto for distribution and end-user consumption, but things have moved on and now we can have 24 bit high res remasters and remixes (in theory and in practice). Same should apply to video and photos, in the future.

I am aware that somethin akin occurs in the Blackmagic camera world where improvements in their colour science allows you to reprocess pseudo raw video files(cos Blackmagic capture is not really 100% raw) , using an improved proprietary version of their process.

Learning darktable has not been a waste of time, not so. Still a tool I would use on special occasions. The ecosystem and contributors to darkable have opened up knowledge that I had no clue about, such as the current limitations in digital photography, needing to conform images to colour spaces from input to output, needing to reduce the bit depth of images, becoming aware of out of gamut challenges that othr tools never show you, in their effort to keep complexity away from you. So much has been learned by trying to get darktable to deliver the results I wanted.

Over time as my picture taking has improved, and my vision, of what I want to achieve has also improved, I cannot ignore the pretty significant amount of “intelligence” that has been embedded in the proprietary image editors. Straight out of opening up the RAW image, you get a more believable starting point, and while I accept that this starting point is a bit of a straight jacket, roping you into a samey samey image, but its also a safety net. Darktable is like high wire tight rope walking, with very little to stop you from making a right mess of things, which is also good - you learn from the falls.

I do not think the “intelligence” baked into proprietary tools and cameras will ever be available in open source tools, unless some unforseen failure or government supported initiative made this neccesary, so for photo editing, there will be the two camps. Open source, and proprietary, and they from what I have extensively researched, will never look similar. Never. For certain kinds of images, I’ll go with darktable, for others, darktable will have to take a back seat, cos the initial image presented by darktable, for very good reasons, deviates quite a bit from the image presented to you as a starting point in any of the proprietary closed source tools. Thats a fact, not debatable.

But the images presented to you by the proprietary tools, looks much more similar to each other, across various tools from Adobe, Capture One, and directly from Sony.

Let’s be open about this, and not defensive. Darktable will struggle to incorporate this comfortable starting off point, when you open up your raw image. It cannot, unless something unforseen happens, invent this “photo” intelligence, that is locked up in Sony, Canon, Nikon, and Adobe. For quite a bit of the other image editing, sure Darktable can come in handy, once the initial image is developed elsewhere. I could imagine importing a 32 bit tiff into darktable, after inital “development” of photos in some other tool, in some use cases.

I know that these opinions will not be popular, after all this is a forum for open source tools. But I hope I can contribute something meaningful, by pointing out the shortcoming that I cannot see any open source tool overcoming, in the near future, unless something unforseen occurs, such as a kind donation by one of the big players, of their photo intelligence, to the open source community.

I can imagine that many in the free and open source community will challenge my opinions, but please do your own research - take the same images, and edit them in darktable, and in the proprietary tools, and take a good look at the results, especially the starting point of photo development in the tool. Its a lot harder in darktable cos there is something missing in the various start off points for photo development in darktable. I have tried all four approaches and they produce inconsistent results across various kinds of photos.

The four approaches, being to conform the raw file using as a starting point.

1. Using rgb curves or tone curves to add the required contrast, such as a 2.2 gamma.

2. Using LUTS.

3. Using base curves

4. Using filmic.

A lot of the time you get a decent enough result, and “it depends” but you never really know what image is better processed with which initial approach until you try them all, which is a pretty huge overhead to the workflow, only if you have lots of time. Creative - yes, fun yes, efficient - no.

No knowledge is lost or wasted.

My final take is - as my pictures have improved, in the taking of them, there has been less and less of a need to edit. Getting the lighting right, framing, avoiding poor exposure choices, and with all these advances, less and less of a need for extensive edits in darktable, but more of a great need to just have the image pop up properly on first open, as it does on other tools, not having to start this hurdle or massaging the image, to look great from day 1, and cycle through a few of the initial photo development paths open to you in darktable (listed above).

Today, an image I had struggled with in darktable, creating several versions, which is darktable’s strength - great image processing approach, in theory, of which I liked only one or two versions, after many revisions. Then I opened up the same raw file in Adobe Photoshop Express and in less than 20 seconds, just a little bit of extra sharpening, that was it - done. The raw file, straight out of camera as “developed” in Adobe Photoshop Express was 95% done, and I cannot fight with the obvious. Compared the results with all I had achieved in darktable, and no contest. It will take an unforseen event for open source tools to have that kind of “photo intelligence”. Its not just about darktable, its the same with Raw Therapee - the research investment in “photo intelligence”, in comparison to proprietary tools is going to be pretty hard to match. Any non proprietary tool for developing images from the raw file, will face a similar challenge.

[betazoid (Anna)]

Have you tried RawTherapee or ART?
I have used all of the proprietary tools for years, I think I know them quite well, but I think there is no photo editor/raw developer that produces “automatically” better results than RT - most of the time it’s the auto-matched curve that does the “magic”. There are photos where you don’t need to do anything in RT.
And one more thing: the most important thing in raw editing is knowing what you want. If you know where you want to be it’s easier to get there. I suggest, when you open a raw file in dt, you look very carefully at it and think very sharply before you start pushing the sliders. Actually a good exercise for this is just looking at pictures without doing anything with them. Go to the museum.

Eidt: Btw, presets for different cameras is not a bad idea, but we would need to include other modules such as local contrast as well. Filmic alone is not enough.

[Peter]

If I want in camera JPEG look-alike, then I use base curve. There is a profile in the base curve for Canon and Canon EOS 6D.
For art reproduction I use base curve and color calibration together with SpyderCheckr. The results when printing are very close to the original paintings.

I took some raw files and JPEG earlier this day. Check them out. Some with base curve, some with filmic. The most time consuming was to add NR, contrast equilizer x2 and sharpening. CC0 https://drive.google.com/file/d/1RcKcdH9PjGucOcOWmUG7kuS5LWdD8Qm-/view?usp=sharing

But I prefer my own preset.

The raw files from my DJI Mavic Pro is another story. In that case I recognize what you are saying.

[paperdigits (Mica)]

@OK1

I can imagine that many in the free and open source community will challenge my opinions, but please do your own research - take the same images, and edit them in darktable, and in the proprietary tools, and take a good look at the results, especially the starting point of photo development in the tool. Its a lot harder in darktable cos there is something missing in the various start off points for photo development in darktable

If you like the look of the camera jpeg, that is good for you. I like the look of some of the Fuji simulation modes.

However, those looks are someone else’s look and feel, not mine. For my (artistic) photos, I don’t care about the embedded jpeg look at all, I want to put my own spin on the image.

My own personal opinion is that if you want to match the look of the camera jpeg, darktable may not be the tool for you. Sure with some effort you can get there, but there are faster and easier ways to get that, if that is what you desire.

What I want is a tool that will give me a lot of power to edit my images, even if that comes with some learning.

I’ve never used Lightroom, but if I did I would be willing to bet that I would find it quite limiting and discard it rather quickly.

[priort (Todd Prior)]

Until you present some examples its pretty hard to refute or support your claims. Just use a playraw if the issue is all your images are of people that you can’t share. It seems you are really straying in to this “I want something that looks like my jpg file” to which most people would say then just use the jpg. There may be something to what you say but really I have also used a few commercial programs and what really is interesting if you pixel peep they are actually not rendered that well. Images in DT can often seem not to have the punch that is added by commercial product and when you first look at them you can be fooled by the over jacked color and contrast in many case. Again without specific examples we can’t really have a conversation. A picture is worth a thousand words and here it would help…

[anon41087856]

@OK1

Long story short. Sadly, going forward, I can only use Darktable for special cases, where I’d like to deviate pretty creatively, from the out of the camera jpg.

Do you realize that hundreds of people are using it right now with no issues ?

@OK1

there is a huge amount of “inteligence” that has gone into the tools that are not open source or freeware like darktable,

No, there is not. There is no big secret in image processing, manufacturers do their color profiles, add custom curves, some do some sort of local contrast adjustment (breaking news : dt local contrast uses the same local Laplacian operator as Lightroom “clarity”, the algo is documented and published here : Local Laplacian Filters: Edge-aware Image Processing with a Laplacian Pyramid).

@OK1

who knows one day, we’ll have much more capability on the output end of things. In audio 16 bit was de-facto for distribution and end-user consumption, but things have moved on and now we can have 24 bit high res remasters and remixes (in theory and in practice). Same should apply to video and photos, in the future.

Bit-depth is irrelevant for what we do, 8 bits is plenty for display, it’s still enough for lower-tier HDR.

@OK1

Darktable is like high wire tight rope walking, with very little to stop you from making a right mess of things, which is also good - you learn from the falls.

[EN] darktable 3.0 live - fast workflow for pros/time-constrained photographers

@OK1

pointing out the shortcoming that I cannot see any open source tool overcoming, in the near future, unless something unforseen occurs, such as a kind donation by one of the big players, of their photo intelligence, to the open source community.

Shortcom… what ? https://www.reddit.com/r/DarkTable/comments/nt2i65/darktable_is_fantastic/

@OK1

the research investment in “photo intelligence”, in comparison to proprietary tools is going to be pretty hard to match.

You mislead. darktable aims at tooling the user’s intelligence. That’s where we failed : assuming users knew what they were doing, or would learn at some point. Proprietary is aimed at making magic in-place of the unskilled so you can get the joy of feeling like a photographer while actually having no idea what you just did.

You either put on a CD or start to play on some instrument, but don’t blame the instrument for not having pre-recorded stuff or blame the juke-box for not conveying your vision.

[snibgo (Alan Gibson)]

I’ll make the obvious point that Sony know how their cameras make JPEGs, so their software knows how to make a JPEG from the raw that looks exactly like the OOC. My Nikon D800 came with software that could do the same trick, which was neat, but I really hated the software and never bothered with it.

Moreover, although the Nikon OOC JPEGs were (and still are) quite good, I could easily make an image that I preferred to the JPEG, manually with any software. This is partly because I know what my image is aiming for, which elements I want to emphasise and so on. And I can see different ways to process the image, sometimes in ways I hadn’t considered when taking the photo.

Personally, I suspect that OOC JPEGS should be considered harmful, and ought to be banned. They encourage the view that this is the way the photo should be processed, against which we should judge the paltry efforts we make from other software. We might call the OOC a starting point, but it is really an end-point, the result of many assumptions and decisions, and changing any of those at that late stage generally makes the result look worse. It is easier to start from scratch.

We can be blinded from seeing that the image might be better in a lower or higher key, or with background colours muted, or extra contrast in this part, or with that colour tweaked.

Rather than “press this button to look like the OOC”, I would prefer a button that randomised all the settings, to process the image in a way I hadn’t considered.

[OK1]

Glad you appreciate some of the challenges that I may not have articulated perfectly.

You truly understand my challenge when you state so correctly, and I paraphrase here, “that the filmic module alone may not be ample by itself”, for “developing” the raw image to a point that is good enough for viewing, and any further editing. You mention and I agree that some other modules such as local contrast would be needed.

Without getting too deep into the technicalities, that was my experience that to get to a basic comfortable starting point where I could look at an image, and be satisfied that I was seeing a decent representation of what the camera captured, I needed to use a couple of modules, such as exposure, (Filmic or base curve or 3D LUT) and sometimes more, to get an image to a place where I was comfortable with what I saw, to then begin to get to step 2, which is fixing any issues, and then step 3 which is adding my own spin or artistry.

If one spends too long at stage 1, creatively one is exhausted before one gets to step 2 and 3. And my experience has been that our eyes and brain are not accurate precision tools but operate relatively, i.e after too much time spent on one image, one is no longer objective cos any further adjustments are relative, and ist easy to go too far, with editing cos one has spent so much time on an edit, that you need to take a break or come back the next day to objectively see what appeared to you to be a minor edit, which had taken your image beyond the intention you had.

Kinda similar to needing to take a break when you are mixing audio or video.

I hope my comments will be seen not as in any way knocking darktable or open source tools, but as a realistic if somewhat frustrated feedback, from a real user. I once wrote software for a living for at least 10 years, and supported various software products and projects for at least another 10 years, so I value the effort that has gone into the darktable product. Phenomenal achievement. It may be left of field opinions like mine, that may stir those who have the ability to create the needed change in further darktable development.

But I am extremely glad I can close this comment on a most positive note. I earlier spoke about the possibility of seeking some other tool which may permit me to develop the photos for further editing in darktable (or GIMP - now that is a lot more complex and frightening than darktable). I do love the organisation and workflow of darktable which has come in leaps and bounds over the most recent 18 months.

I was initially thinking I’d use one of the other editors I mentioned to “develop” the raw image, i,e transform it into something that looks acceptable, then convert this to a 32 bit Tiff file, i,e with as much of the information retained as possible, then bring this into darktable for any further tweaks. One finds similar workflow where image development is started in Adobe Camera Raw, and or Lightroom and final editing is done in Photoshop, or imag development is done in darktable, for further editing in GIMP.

Its very early days yet, and I have only spent about the most recent 30 minutes downloading, installing, quick read through the available documentation, and opening up a few images, but I must say I am impressed with the immediate quality of the default appearance of the images in the tool, as it were - straight out of raw development using default settings, absolutely no changes by me and I feel like I’m using one of the commercial tools I mentioned earlier, but this one is free and also open source (albeit no binaries for Mac - only for Windows and Linux). I speak of Filmulator - its not an advanced photo editor, but it, within just a few minutes of using it, seems to have solved all of my challenges with the various methods of basic photo development stage 1, which I had in darktable.

Straight away, you get a very impressive - possibly too impressive, like I said I have only used it for less than 30 minutes, starting point of a very well developed image, at par or in advance of what I experience with the non open source image editors. It is early days, very early days, but looks promising. The number of controls are few, well laid out, and I like what I am seeing in the images, and thats before I touch any controls. I forsee my having to do very little in Filmulator - maybe only white balance, then export as tiff and continue photo edit in darktable, which I can then use for its great workflow.

But as I said, its early days, but out of the blue I remembered I had heard of this tool, and thought to try it out. Previously with all my work done in darktable, it was not needed.

But if we are all open here, and truly open, what I can see is a win win for everyone.

Can we imagine a world in which the photo development features of filmulator were reused in darktable.? That would be a wonderful world, where you could use what you wanted, as your photo development tool or tools in darktable directly. My point is - yes I have made a fuss that I think darktable photo development foundations are in need of enhancement in certain ways, but its not as bad as it seems, the solution - is already there somewhat. And I’m happy to be a go between and connect the development teams of darktable and filmulator to see what they could do together, to really raise darktables game, and bring in literally a far greater number of users, who would not face the challenges I have faced, so far.

I may be speaking too early, but with what I have seen in filmulator, there may be no need for presets, cos the default “developed” image is certainly at par or better than anything I have seen in any tool, including all the great non open source ones I have mentioned.

So here I am, having to eat my own words, within just a few hours - there is great open source software out there, with available code, that appears to mimic or exceed the “out of the box” photo intelligence that I have referred to as “color science”. Somehow the developers of filmulator have got this challenge of mine sorted - and it might just be the perfect next addition to darktable.

May I appeal to all those involved. After over a year of trying to make darktable work for me as my only photo editing tool(I do not need anything as extensive as photoshop and darktable is a cool intermediate between the kind of features in both lightroom and photoshop - a bit of both), I had to eat my own humble pie, and admit, it was not working for me, and I have aleady eplicitly stated the specific area which I found challenging.

Its not about the thousands who are using it and not complaining. If we are all humble it could be just one, just one person who has the guts to speak out and say what many others are respectfully coping with, or not even aware that there could be a better way of doing things, and say it with all honesty, that may make darktable better for current and future users.

Lets see where this leads, I only ask for significant humility on the part of all involved, including the developers of darktable. I do not belittle the work involved, not at all, but I am certain, that if this area of darktable were resolved, in the way I think has already been achieved in filmulator (and who knows, maybe there are other open source tools with similar capabilities), it would be one very big leap for mankind.

Please note its not about having identicum with a Sony or Canon or whatever camera or model’s JPEG image intelligence, no. I was not looking for such, I simply wanted to open an image and see it well presented, without my having to labour over it again - after all the work it took to capture the image in the 1st place - Filmulator does this - instantly very decent, very well presented images, straight from raw. I am not an optics expert and have no clue how it is achieved, but I have tried it out, and to any who are skeptic, it will only take 30 minutes of your time to come to the same conclusion. Great images - straight out of camera raw.

I’ll test out Fiilmulator and how this weaves into my workflow, but this is something everyone else can also try out for themselves, its free and I assure you, one look at a complex image viewed in Filmulator will set you thinking, there must be an optimal way forward for enhancing darktable to incorporate a similar instant - straight out of the box excellence of the initially developed image.

Someone in the open source world has figured out the photo science(at least to an extent that I am initially pleased with) and done a pretty good job of it, let’s not be defensive and reinvent the wheel, its already there, and we can reuse it in darktable.

I had to abandon what I thought was the best - i.e darktable as my only tool and admit that I needed better workflow, and yes it was difficult for me to admit cos I had invested so much effort and time in darktable, but the images starred me in the face, it was time to change.

If all in darktable community can see the benefit of better, darktable can have an amazing future.

One look at the default image you import into Filmulator and you will come to the same conclusion, darktable needs a similar capability. Not my opinion only - anyone can try it out for themselves.

[ggbutcher (Glenn Butcher)]

The big bollox to auto-tone is the handling of sensor saturation and the single-exposure across the sensor. For all the smarts behind it, the metering mode can only dial in one shutter speed and/or aperture.

Most metering modes are about exposing for a middle gray somewhere in the scene. Depending on the lighting of the scene, that may push the parts with more inflicted light energy into sensor saturation, blown highlights. Take that into whatever software, either in-camera or on your computer, and a default tone curve and highlight reconstruction handle that.

However, some prefer to record meaningful data across the entire scene, ETTR or whatever metering. This in some cases pushes the desired middle region into darkness, requiring lift with the tone curve. That lift requires some judgement regarding sufficiency; an algorithm attempt will leave one unsatisfied in most conditions, so you’re back to messing with the sliders.

I shot a chorale concert this weekend conducted outdoors, where the group moved from venue to venue. Some were in bright sunlight, others with various mixes of shade. I expose to preserve data, so I spent a bit of time coming up with curves for each situation. As I do this for each shoot, I study my parameterization, but I cannot come up with any useful heuristic for automation.

If you want the software to do the work for you, IMHO you’re missing out on really understanding your medium…

[OK1]

Dear Aurelien,

Your amazing work with darktable has opened up both the challenges of all digital photography in general, and made me seek to find answers. Over the next few days, please re-read my comments.

With the same verve that I have brought up the challenges I consider need resolving in darktable, I have with the same enthusiasm, spoken up again, (see my prior comment to this one), about discovering that someone in the open source community appears to have developed a solution that fulfils the missing link in darktable.

Its an open source tool called Filmulator(at filmulator.org). have no affiliation with the tool, but having tried it out, it looks promising, as something that could be incorporated into darktable as a module, with the support/permission and hopefully assistance of the core developers ideally, as one more option to filmic and basetable or 3DLUT modules, as one more approach to getting an initial highly polished end result which can then be edited further.

Of course one would have to look at dissecting what Filmulator has done and like turbocharging a car, see what parts can be reused and where to fit them most appropriately into the darktable workflow.

I may be simple minded, but there seems to be a lot of aspects that can be copied across and reused.

Reviewing a small set of images which had proved pretty difficult for me to get to look right in darktable, in Filmulator, and all my concerns were gone. Sharp images, great colour - straight out of the box - straight from raw, with absolutely no adjustment by me - none. I could save them as jpeg straightaway and be happy with the result, but then I could export as tiff, and bring them them into darktable for further editing…

Please take a look at filmulator - we need something like this in darktable. It seems to completely address all my concerns, and I could imagine a world where we could incorporate the aspects of filmulator that I have found a bit challenging in darktable, liek the immediacy of a well developed image, with enough controls to tweak (if you so choose), and for those who need this filmic or any other modules in darktable can be part of the workflow.

In my use of darktable, I sometimes combine using LUTS with filmic(using the opacity to just add a dash of filmic), for creative effect to create images that neither of them on their own can create. Rules including all the scene referred and display referred can be broken, fortunately darktable gives us the opportunity to break rules, which is what is great about it.

I think we have an opportunity to take all the amazing work that you and others have done in darktable, and close some of the challenges that may prevent darktable from having a much wider audience.

I am reminded of one of the supposed tenets of object oriented programming, and open source - reuse.

You have been and can be an even greater catalyst for change in darktable’s fortunes, and bring changes like this, such as reusing filmulator image processing code, in darktable, to the wider public, not just nerds like me, who have had the patience to stick with it thus far.

I’m not too far away, in currently sunny United Kingdom, and I’d love to meet you sometime not too far off in time, in person, when proper travel is permitted for non urgent visits.

I am confident that you’ll see the value in my challenges with darktable, and the opportunity to add even more photo intelligence into darktable for those who want a bit more instant results, as a starting point in their photo editing journey.

It is such a relief when you open up a photo editor and love what you see - straight out of the box, without having to first fight with a few controls, to get it to an initial satisfactory view point.

Sure filmic is a huge improvement upon the base curve - definitely so, but we can do more, and we need you as the champion, cos you have the abilities to deconstruct filmulator code and port over its capabilities in the photo development workflow, to darktable, which anyone who uses it will find pretty excellent.

Beyond basic photo development, there are thinks like the sharpening used in the out of the box image (default) in filmulator which stood out to me, when I looked at a landscape image with some grass, that was pretty impressive. I have enjoyed the sharpening tools in darktable, but the default sharpening I saw in filmulator is pretty special, do take a look at this tool and its code, for yourself, and do let us have your thoughts on the suggestion to incorporate some of its benefits into darktable.

I am going only by results. I loved the immediate image - default, which filmulator gave me, without any tweaking, and only wish I could also have that in darktable, which is the kind of wish that made me start this thread on the forum.

Best Wishes.

My French is terrible, not sure how to sign off… in French…

[OK1]

Please read all my comments on the thread.

It gets better. There is an open source tool which is also free which I discovered - Filmulator, which does what I was hoping to achieve in darktable, and it does this straight out of the box, default.

Take any image, import int into Filmulator and once you open up the edit tab, straightaway you have a polished beautifully presented image, which you can then tweak a bit further, if you want to.

If we could have something like filmulator in darktable, it would be a bonus for people like me. And those who love using the modules like filmic or base curve, would still have that choice, if the features in filmulator were incorporated as an additional module or additional alternative modules, and can also work in sync with any other legacy modules, so end uses can as they say in English - pick their own poison (Choice of tools).

I opened up less than 15 images straight from raw in filmulator and I was absolutely impressed.

You do not have to take my word on this. 30 minutes of your time to download install and import an image, and you will come to the same conclusion, everyone, including those who love tweaking, could do with something like this in darktable. Results speak for themselves, great images, which is after all the whole point, not the tools.

Filmulator makes all th hard work you took to capture the photos, look great straight from raw. Whatver photo intelligence they are using in this tool, we could do with all of it working inside darktable.

There will always be those who will never drive an automatic, so the most modern cars, especially the more expensive ones, which I do not drive but only see on Youtube, let you have it both ways - manual or automatic - you choose. You can have it both ways.

If it works out well, I’ll do the initial photo development for things like white balance adjustment in filmulator and bring in the image via tiff format into darktable for final finishing touches. Open source all the way, if only they could be perfectly integrated darktable and the features of filmulator.

[anon41087856]

I know filmulator, I had beers with its developer 2 years ago. It personnaly doesn’t give me the results I’m looking for, and also doesn’t look anything like actual film I have shot. Also, I think Johannes tried to re-implement Filmulator in vkdt and noticed the actual code doesn’t match the reference papers it implements.

The thing is, when I take a picture, I know exactly the post-processed result I want before pressing the shutter : whether I will process it B&W or color, if I want something dreamy/soft or dramatic/sharp, what kind of colors and feeling I want in post, meaning how I will push or pull the actual colors of the scene. I also manage the in-camera exposure taking into account how filmic will handle it (meaning usually grossly “under-exposing” by at least 1 EV or shootings in “highlights priority” mode).

When I’m starting an edit, I know where I want to go. dt gives me the least obtrusive way to get it, and that’s what I got fairly easily : https://www.instagram.com/photo.aurelienpierre/.

I suspect many users have no clue what they want upfront, so they fall-back to emulating OOC JPEG as a starter and randomly push and pull sliders before accepting what seems to look best. Working with no aim and with no workflow is going to get you nowhere in darktable because there are simply too many options to roam the software aimlessly while still wanting “some” result.

The inverse is true too. Knowing exactly what result you want gets you nowhere in Lightroom or Filmulator. You have to abide by the creative decisions forced upon you by the editor. It’s very frustrating to know what you are doing and have to use Lightroom. But I bet what I call lack of freedom and forced decisions are what you call “intelligence” and “guidance”.

[OK1]

Please read my other comments - use filmulator - import any image, see for yourself, and compare with the workflow as we currently have it in darktable.

We have an opportunity to go one better and make darktable even better, without alienating anyone who loves the tinkering with the existing modules. I tried out filmulator for a few minutes, and was very impressed.

Please do the same, I am 100% assured you’ll come to the same conclusion.

Its open source, and those involved in darktable can take advantage of it.

[OK1]

Dear Aurelien,

I have made my point. I can do no more. If I had the ability, I would code the changes in darktable, and create my own fork, and wait for the rest of the darktable community to bring the code into mainstream darktable.

I do not yet have the immediate ability to do so. But it’s not an issue, until someone ses the benefit of having someting similar in darktable, I’ll “develop” my images in filmulator, and continue any further edits in darktable. My problems are solved.

I am certain, someday, the ideas expressed in this thread will be revisited and someone will do the needful. It does not have to be filmulator code reused, I’m sue you understand what I mean, the functionality.

But its a democracy, if those in charge of darktable, consider it not to be a good idea, to have similar functionality in darktable, not a problem. We will wait. It’s only a matter of time. I am extremely confident that this idea will sit in the minds of all those involved, and its only a matter of time.

Good ideas are good ideas, and it does not matter where they come from. I saw the results of filmulator and its a great idea, to have in a photo development tool. Any photo development tool should have something like this.

I rest my case, until further notice.

[priort (Todd Prior)]

I have tried it i the past and it does produce some nice results for a simple tool but I have to tweak just as many sliders and then I am going to have to export the image and continue on in another program to handle noise color and a myriad of other issues…so its not going to save me time to have to manage two programs to achieve the same result…

[priort (Todd Prior)]

On thing that many people fail to remember in these discussions esp when they go comparing software to something like an Adobe product is that this is open source software written by developers on their own time to suit first and foremost their needs, desires and vision not ours. We have the privilege to tag along for free or participate but we need to remember they are not marketing anything to others. Of course good ideas wherever they come from should always be welcome and considered and there is in fact a lot of code sharing at times where it makes sense.

Darktable has a philosophy to present the raw image with as much control and transparency of process as it can. In doing so out of the gate the image is “really” raw and often unimpressive when compared to commercial product. Then you go to work and craft your edit. The user should know and understand that and move on if that is not "their " expectation.

The current tool set works for you for some things and now you have found something to offset what doesn’t so that’s great. For me the benefit is not there and I will stick one program for the most part as that works for me the majority of the time. I also have Photoflow, Filmulator, RT, ART, Affinity Photo, GIMP ON1 photoraw Photoscape X and Picture Window Pro 8. From time to time I revisit all of these to see what changes have been introduced or if I am curious I try an image here and there but up to this point DT consistently seems to give me the best results in my hands…

[spidermonkey]

I think there is mental barrier for many people that if the starting point doesn’t look quite as nice as their OOC jpg, then they are dealing with a picture that is worse to start off with, so they don’t want to engage with it. There is also the belief that this is what came out of their camera so that is the ‘correct’ image. This is perhaps why commercial software developers invest so much time in making the defaults look like the OOC jpg, even if there are already lots of adjustments behind the scene. darktable developers, understandably, don’t want to spend their time doing this.

Would it be worth having a splash screen for new user installs that says that the starting point won’t look like the OOC jpg, and suggest that they need to adjust contrast, saturation, exposure and local contrast. Having those few controls on the quick adjustment panel too, would mean that they can easily adjust their images to taste.

It might mean there are less comments “darktable is broken, the images look fine in the lighttable but then they look all dull when I open them in darkroom, what is wrong?” It must get tedious for people having to repeatedly explain this.

I must admit, I fail to see what the problem is, after adjusting those few things, my image will always look better than the OOC jpg, after around 30 seconds worth of adjustments. Local contrast plays a big part, I think, to give images that punchy look.

There is also the fact that if somebody has paid for something, they will probably make sure they put more effort into learning how to use it. Can’t do much about that one.