the darktable 3.0 video series

Hi’ @aurelienpierre
Thank you for an excellent and interesting video. I like your paper and pen approach, it makes you “digest” the information gradually instead of displaying a full and complicated slide from the outset and then start explaining. I’m looking forward to the next video(s)…:blush:!

In the meantime two questions: You mention that the global tone map tool is broken. Is the tone mapping tool also broken?

From your video, I understand that you always need the base curve (or filmic). However, in many cases or maybe in all cases the base curve is not enough. You need further adjustments. Why is that the case?

I’m sure you recognize the following picture. The contrast between the shadows and the highlights is much bigger than my eyes recorded it at the time?


The tone mapping (local) gives the expected results, compared to the litterature, but produces halos as a side-effect (which is consistent with the algorithm). I would say the result is bad, but consistent with what we expect. The global tone mapping doesn’t reproduce the results shown in the litterature.

It’s a matter of (acquired) taste. Digital photograhy has made us use to very high contrast because the first sensors had very poor dynamic range. When the first Sony alpha 7 went out, people complained they got washed images and had to push the contrast a lot in post. It really depends what you expect to see, therefore what you are used to.

Your eyes should have recorded more contrast than the sensor if you are a standard human being using a consumer camera.

Another Youtuber gave me the idea to look into the stats of my channel…

Over the past 90 days, my videos have been watched by 100% of men between 35 and 65 years. 14027 views, 2451 cumulated hours, not a single woman there.

Ladies, what am I doing wrong ? I don’t mind having a sausage party there, but I suspect the subject will be of some interest for you too (you take pictures too, right ?), so the content is what it is (you can’t escape technics forever), but I can still try to make it understandable… if I knew what doesn’t work for you.

1 Like

I don’t think 100% of men 35 - 65 watched your video. You’d have a lot more views :wink: Rather 100% of your viewers were men in that age range.

Also… Why does it matter?

I have come across women who enjoy and excel in STEM. Technical information shouldn’t be a hindrance to those who desire it. How does YT arrive at these stats by the way?

Well, language… 100% of the viewers are 35-65 men.

When you have a certain percentage of a certain social group represented in a community, you expect to find a similar percentage ± sampling errors in every sub-group of said community. That’s how probabilities work. When you don’t, something non-random is screwing those probabilities, and you would better figure out what and why, when that certain social group is already a minority in said community. Otherwise, you just perpetuate inequalities and such.

Women represent 3% of OpenSource contribution on Github in 2017 and 45% of photographers in the US, so 3% × 45% = 1.35% of women would be a fair expected minimum in my audience. 0% is off-charts, considering the sample has 14000 members. Something is not normal here (in the statistical sense).

1 Like

That I don’t know (and will probably never will).

My point is that a stat of 100% isn’t a good one since there has to be outliers or at least ones indicated by confidence. I am questioning how YT determines the sex and age. Probably using other stats. There is also the case where no one is watching and the video is auto-playing. :stuck_out_tongue_closed_eyes:

Google knows mostly everything on you, especially when you have an account (you have at least to type your birthdate). Errors should be fairly distributed among the sample for such big samples so I wouldn’t try to look for funny discrepancies as an explanation.

I am an outlier :grin:
I really enjoyed your video, it is well constructed and very informative. Thank you Aurélien :slight_smile:

1 Like

@betazoid Just to check the numbers, have you watch the video?

I began to watch it but decided that I could not watch more than 10 minutes at a time so I am not yet finished :stuck_out_tongue: but I intend to watch it until the end
Anyway I clicked on the video atleast once and I guess it downloaded completely, I don’t think I closed the tab/window.
The internet sometimes thinks that I am a man because my last name is a male first name (my email address begins with my last name) and sometimes I even get emails form ladies from Ukraine or so who want to marry a European man.
Ok. I will watch it completely today and @aurelienpierre please check the stats again in the evening.

1 Like

Please help to understand this. I can’t get your answer to fit with the information in your video.

According to your excellent video my eyes perceive less light than the sensor register when the light source emits a lot of light = highlights? Therefore, the difference between dark areas and light areas (the contrast) is perceived less than the sensor register?

When I shoot a picture the shadows always seems darker and the highlights lighter in the photo than I perceived the situation. Is this due to the base curve?

Ok. I will try to give you feedback. The first 20-30 minutes were like this for me: 2-3 minutes talking and then come 1-2 important sentences. Maybe I can give more/better feedback if I watched the full video.
Edit: after having watched the whole video I can say this: first, as far as I am concerned, I think I did not understand about 50% of, but I think I understood the important things. As far as other, most women are concerend: well women are just not interested in videos such as this. They do not want to know the how and why, they are interested in the end result, the bottom line and the practical consequences. But do not underestimate the practical skills of women and their creativity. I think women understand science differently, on a more intuitive level than men. We cannot explain it with words but if we did not understand physics at all we could not even learn how to walk.
Maybe one more edit: e.g. I do not know what a matrix is. I probably learned about matrices at school but that was very long time ago and I don’t remember any more. All I know is that the word matrix is a combination of the latin word mater (mother) and radix (root).

I don’t know. My (female) photographer friends know what kind if software I am using since a long time but I somehow cannot make them even have a look at Rawtherapee or darktable. I don’t know why. They like Lightroom and Photoshop and the Nik plugins. They like (automatic) effects that are easy to use. They are not interested in trying something new, they like what they are used to. I think they would only try open source software if most people were using it. If I talk about open source photo editing software with men they seem more open, but I don’t know whether any of them have actually tried them after talking with me. As far as men are concerned, many of them simply don’t know that there are free alternatives. But apparently most photographers want Adobe. Also, some people, especially women, seem to be afraid of open source software. They think open source software will destroy their computer or something like that.

… and a lot of the photographers use Apple products simply because it is like it is. It’s pressure of the environment, mindset and ease of use (every goes it, right?)

I’m also struggling to get the new concept right. My workflow was simple: disable base curve, correct white balance, lens distortion, jump to tone curve and do the basic adjustment then use equalizer and/or local contrast and then if necessary add more instances of tone curves. Now I’m trying to figure out what I should do :wink:

Ahhh I see what you don’t understand. Sensor and eyes clip at a certain light intensity (glare limit for the eyes, or saturation limit for the sensor). Since your eyes + brain compress highlights, they push that clipping limit further than the sensor, so you see way beyond the limits of your sensor. Meaning when your sensor sees “white”, that white is actually light grey to your eyes.

I don’t think that particular to women. It’s a side-effect of our time: “I don’t need to understand, some software will do magic for me, I just need to know which button to push”. That backfires real quick, and this kind of users is subscribed for life to gurus who will think for them and tell them what to do (it’s a big issue with continuous education too: such workers constantly need seminars and training, because they can’t adapt and learn by themselves if they don’t understand what their tools are actually doing).

The point of science is to add a rationnal layer on top of intuition, because intuition has proven to fail when problems become complex and mix a large number of variables. So the scientific framework provides a way to split complex problems into small chunks, and then solve small problems locally. Understanding “intuitively” is not understanding at all. We all know intuitively that a bike becomes more stable as its speed increases, and is very dangerous at very low speeds. Yet no physicist has ever come with a proper model to describe how speed and equilibrium are linked for a bike. We know how to ship space rockets on the Moon, but we have no idea how a simple bike works (and that doesn’t mean we can’t make it work). Science is less intimidating once you figure it’s just about finding what relationships exist between things.


I think many studies have proved that women have more or a better intuition than men. It is about the subconscious mind where very many calculations are executed, e.g. when we walk or ride a bike. If we could access the knowledge of our subconscious minds we would know much more about science than we do today. Intuition is exactly about complexity and large numbers of variables.