Filmulator + Darktable - Baby Steps - OK1 - Yellow Flowers

Workflow is :

  1. Image development in Filmulator, then
  2. Minor final adjustments in Darktable, pretty much just sharpening and a bit of noise reduction then export to jpeg. Applied the same darktable settings to all duplicates.

This was my 1st attempt. Went overboard with the dark look, probably cos I was also using the dark background in Filmulator, while refining the image. Its easy to fall in love with the various sliders, cos each of them does something interesting. Excellent range of possibilities. Watch out for the Drama Queen - i.e the drama control, very tempting. Very tempting that one. But if you want that old film look, go ahead, overdose on this slider.

Then went back to the drawing board(image below), changed the image borders in Filmulator to white, and aimed for a brighter result. What I really like about this tool are the colours. The most lifelike, of all the tools I have used, obviously this is in my humble opinion. The defaults will not always be ideal for every photo(but that’s going to be the same with any tool), but with a few tweaks in Filmulator, I was able to lean towards what I had in mind, or rather what my mind recalls from the day the photo was taken, as I was aiming for a natural somewhat realistic look.

The controls in Filmulator, feel less scientific and more artistic - hard to describe. More like painting than drawing. I definitely encourage you to try it out. Do not stick with the default, which is always a decent starting point, but definitely this is a tool worth learning, and it has so few sliders, the learning curve is not steep. It is still in beta, so go easy on any critiques of the tool.

I can’t say enough about the colours - Filmulator has some of the best controls for getting this to where I want them to go., or rather to leave them more like what my mind recalls, from the photo session.

1 Like

Not to be left out of the picture, pun intended, here is a version with filmic. I had to adjust filmic controls in a way I had not done before, but the results speaks for itself.

Filmic + a dash of some sprinklings of Tone Equaliser + same sharpening and denoise settings used in images posted earlier. : I was a bit conservative, in my use of filmic, not wishing to push it to unnatural extremes, as the aim of this comparison was to arrive at a natural look in each case

Oh by the way there is another thread - most interesting which compares various raw processors using statistical analysis.

https://discuss.pixls.us/t/comparing-defaults-across-software

*And here is the in camera jpeg : *

For me this “study” has been quite informative, and the conclusion, not statistical obviously, I can only describe what my eye sees in the pictures. Your mileage may vary.

Each raw processors approach needs patience on the part of the user, to fully appreciate what direction the tool has taken. However it is these comparisons that give one a much better understanding of each tool. I highly recommend that time is devoted occasionally every few months to this kind of review/comparison with a few of your recent images.

From what I can see, filmic takes a purist approach that is closest I have come across, to the real life optics, of a camera capturing the scene. Filmulator and the in-camera algorithms, take more liberty with their creative licence, delivering images which can be more “alive”, but in truth, there are elements of more light bending in their approach.

This reminds me of the music industry, where the purer approach is digital, very accurately captured, but some want vinyl, even though we know it is a distorted version of reality. And so much of what we hear is based on the harmonic aberrations of exciters, distortion effects, compressors, valves, transformer, reverb, yet we find it the more pleasing, and things taste bland without these imperfections.

So I guess it could be decided based on intention. If I wanted to publish something scientific, in a journal - definitely filmic seems like the one to choose. But if I wanted to post to facebook, or instagram, then maybe filmulator is the better choice.

There is a similar parallel in the music instrument world. Especially in the category of instruments targeted at the less accomplished players, these instruments are designed to sound good in the shop, one touch - engaging sound, but these sounds have been fed with steroids. Only time and experience lets you realise how hyped they are, and over time you move on to the more expensive instruments which do not immediately sound as hyped, but this more natural expression, matches more closely with the advanced skills of those for whom it was intended. No instant gratification, but a lifetime of true musical pleasure.

So one size does not fit all. And it is gratifying to have improved my appreciation of the directions, that each of these tools takes.

While I absolutely respect Aurelien for his qualified stance on basing everything on a truly structured proven, extensible, future proof foundation, such as he has attempted to do with filmic, I would hope he appreciates that there must be some inherent non linearity in humanity that makes us gravitate towards the impure, the gritty, the out of this world.

We are the product of steamy sessions between a man and a woman, in some warm or cold places, and probably unideal location, many from unplanned accidents, not the thoroughly well planned in a lab supervised fertilisation.

I would hope that we would see in the future a darktable that accommodates both views. Provides the scientifically accurate transforms, but also leaves room for the creative transforms that have underpinned our experience of films, film photography and in-camera jpegs. These augmented reality transforms, also have a place in digital photography, and darktable should give us the opportunity to choose what kind of transform we wish to use.

Yep, Filmulator is a very nice tool, good at what it does. I think you’ll soon get noise reduction, too. What is the source file for noise reduction?

Yes please, steamy hot darktable sessions in weird positions! Give some of that :smirk:

Any implication that you cannot be creative with darktable is completely unfounded. It might not have that particular thing you adore about Filmulator, but you cannot keep rejecting darktable for it.

2 Likes

I’m not really sure what you’re trying to achieve with all these threads. It looks like a comparison between Raw processors but the problem is that you can get good or bad results from all of the available software depending on how you use it. If you were to upload your XMP files and give some idea of what you’re trying to achieve people would be happy to help you to learn the tools. Is that what you want?

6 Likes

I’ve had quite a challenge with developing images with filmic, which looked ok to me.

Now its absolutely impossible to describe what the ok would look like. Only I would know, so its not really something that lends itself to a specific enquiry.

Yes it has been busy in terms of my recent participation on the forum, lots of posts from this “stranger”.

But there has been a method, and I hope my explanation below clarifies this sufficiently for you.

  1. I was not able to get my images to consistently look like a picture, using filmic. Sometimes, it would work, at other times, it was frustrating me. Filmic has a number of controls, and it takes a while to learn them all. A lot more involved that what it replaced or became an alternative to - the base curve module.

  2. The next step in the search led me to try out other tools, and I was able to obtain varying degrees of more immediate gratification from these other tools, typically raw to image, or what we now know to be scene to display referred, tools outside of the internal universe of darktable.

  3. Thereafter only a few hours ago, I watched fortuitously a video on youtube, and saw a really accomplished photographer, displaying pretty impressive images via workflow developed with filmic, which led me to revisit filmic, and the other tools to better understand the context in which each of them excels. And I also invested even more efforts to really get to the bottom of my understanding of filmic. Guess seeing what someone else was able to do with it, which was quite impressive, is a part of the education process.

  4. This thread, with the yellow flower, as you can see starts off with being able to get in my opinion, subjectively pleasing results with filmulator, with not too much effort.

  5. Thereafter, I have given filmic another go at it, and if you read through, the results are interesting. And the images speak for themselves.

As it has transpired with me, the journey and back and forth has not been wasted. More than ever before, it has helped explain to me, and I hope to others also, the love/hate relationship with filmic. Some like it, some do not, some want to like it - Just read comments and opinions all over the web.

The search was to discover what works for me. But it has provided more than that, I also from looking at the pictures posted on this thread, also understand why each image processor evokes a different impression from the same image.

No need to repeat my earlier expressed conclusions. I publish these thoughts here with the hope that others may also glean from journey, and benefit from my thoughts on the issue.

Looking at these images, reveals the whole story. And I hope others benefit from my thoughts and come to their own conclusions.

It has been quite frustrating, and as I found solutions, and answers and explanations, I’ve put these thoughts out there. That’s all.

Where does all this challenge actually come from. We’ve all grown up with a lifelong view of images, in print and on digital media, and our eyes have been accustomed, like vinyl, to things looking a certain way. Absolutely nothing wrong with that expectation. Then some like me encounter filmic, and the image does not look like what we have spent all our lives seeing in film photography, television, films, facebook, or our in-camera jpgs. Neither does it look like the images we are getting in other photo editors, huge dilemma.

Then its taken for me about a year or more, and a very intensive recent two months, to get to the bottom of this. Eventually it all makes sense. And the images show the direction each processing approach has taken.

In summary - filmic takes a purist approach to the transform, than most, and many of us like me have been so accustomed to the less pure approach taken by everything else we have seen. If you look at the images, the filmic version renders the most distinct separation between foreground and background, between what is in focus and what is out of focus. i.e. in truth - a more realistic view, but one which we are not accustomed to seeing, in print and digital media.

So contrary to what some might have thought, this is not about knocking filmic or darktable, simply about discussing, what we see and know. After all this is the discuss space on pixls.

It has been a search to solve a seeming anomaly, which has ended with a bonus, an understanding of why filmic appears to render a bit differently from what our eyes expect (based on a lifelong history of what we have been used to seeing).

I hope others also get something out of this, and especially anyone who has been in the same quandary as I was, with filmic.

I also hope my observations will assist the developers of imaging tools, to understand their algorithms much better, not just from the technical viewpoint but from the viewpoint of especially the uninitiated user who is trying to succeed with their tools. Hopefully good feedback, which for engineers should feedback into their tool improvements.

This reminds me of one of the features that is enabled in 3G telephony - which is rarely used - video calls, because tools must go beyond their intention, to accommodate human behaviour.

We can’t force people to like something. And if there is a clamour by some including me, that a tool is not well appreciated. we should ask - Why?, rather than just dismiss the assertion, with a response that oh - the tool was not built for people like you.

My threads have really helped me outline what the issue is, and eventually understand why we have this issue, and hopefully for those with the patience to read through rather than be dismissive, to see that there is a valid reason for the marmite opinions on filmic.

But there is also a solution - some education to manage expectations. filmic is a decent capable tool, as I have eventually found out, but it has its own unique look(from its own unique algorithms), which is a bit distinct from most other raw image processors, and it may take a while to get familiar with that look., or learn how to complement that look, with the other tools in darktable.

2 Likes

Links please.

Unfortunately we can’t see how you’re using the tools so I can’t see how any firm conclusions can be drawn from your attempts. The results don’t speak for themselves because there are many ways (especially in darktable) to achieve the same result.

I’m the primary maintainer of the darktable documentation and I’m genuinely interested in seeing how people use the tool and what (if any) misunderstandings they might have. That’s ultimately the purpose of good documentation. But we can’t do this without concrete examples and specific discussions.

4 Likes

If you search the pixls forums, the opinions are there, hundreds of them.

You said “all over the web” which implies more than just here. I’ve literally read every post ever published on this forum.

Please use your eyes. You do not need anything more. Its’ in the pictures I published, and I have attempted to explain what I see.

You do not need any tools, or sample files, except maybe some spectacles and a good monitor to see, what I see. I have given you an example of what I observe as a key differentiator in the final image, I hope you can also see the same things.

Sorry. I can’t.

Tha’s not how it works. In all raw processors a skilled person can achieve virtually identical results using different processes.

XMP files from darktable usually contain history which would reveal your “thought” process. tha’s wy we askn.

An’ since we can’t ‘see’ your process we can’t ‘see’ what you’re up against. If it’s filmic that’s givin’ ye troubles, just come on over to Let's learn Filmic RGB! Your one stop shop to understanding filmic-based approach to edits! and post the stuff accordin to rulz an’ we halp ye.

Less wording more lerning :+1:

7 Likes

Your were politely asked to give more information. Why do you insist that this information is not necassary for others? You have it, just provide it.

6 Likes

I feel you’re replying with generic statements instead of anything concrete.
If you like Filmulator, and prefer it over darktable, fine. If you post some photos in the Showcase category, we’ll be happy to admire them. But coming here, doing comparisons without listening, and not giving information when requested, does not benefit anyone. Not even you.

7 Likes

I will make one more attempt to explain what I see.

One thing that will be impossible is to share what I who took the picture saw, in an environment that I see pretty regularly, and am familiar with. That I cannot convey in any form

When you look at the images developed with other tools, in this example, the background that is out of focus is not as natural, and does not separate as distinctly from the foreground flowers that are in focus. i,e without filmic the transition from foreground to out of focus relationship is fuzzy, very busy out of focus area, the images developed with other tools when you really look closely are the ones which are actually the more “distorted”. When you compare them with similar non filmic tools, they look kind of similar in this regard.

Then as it made clear to me, I did a last ditch effort to do the best image I could in filmic, and when I stood back to look at the result, and especially when I compared with the result of filmulator and the in-camera jpeg all of which are published, here, to me it was obvious.

The image with the least “distortion” was filmic, with the most accurate rendering of the image, and the best separation of foreground and background, amongst other things that can be discerned, this being the most visible one.

In comparison, the other non filmic images look like something from another planet. But our eyes see that distortion as “normal” cos thats what we’ve been fed with.

I really hope people read and read everything, rather than come to an unfortunate opinion thinking I am saying something negative about filmic, on the contrary, its success at remaining authentic to the raw image is the cause of its issue, for some - my hypothesis. For some of us, it initially did not look like what we wanted to see. We are not used to seeing images so clinically well processed, with so little distortion from the original.

How much more can I explain this. That’s about as much as I can do.

Your yellow flower image could easily have been edited I think without filmic. I have bounced back and forth between having filmic applied by default and not . I am almost wondering if it was not applied by default and people had to look first at the image and then decide if they wanted to reach for that tool or not just a you do for any other tool in the DT toolbox if there would actually be less of a blame game on filmic when things go wrong with the edit.

1 Like