Feedback / Development: Filmic, Baby Step to a V2?

This has nothing to do with my point that the communication around this subject is too hostile and juvenile to allow for any meaningful progress. As time goes on, the number of people weighing in on these threads has dropped from dozens to around four, and it continues to drop. It’s not hard to see why- any opinion or thought is criticized, mocked, and degraded with a barrage of insults and profanity. Pretty soon these threads will just be one person spewing endless rants full of vitriol, to themselves, ignored by everyone else. When that happens- I did warn y’all, many times :man_shrugging:t3:

Carry on man. I really don’t mind. I get a ping as this was a thread I originated.

I’ve tried to point out the shortcomings of shovelling around stimuli, and I’m reasonably confident that no advancements can be made along such a pursuit.

If folks want to talk about clipping, and relative wattages, and other things that are dead ends, go nuts. Fill your boots.

In the meantime, I’d love to have all of my posts deleted, if that is possible. I don’t mean anonymized, I mean deleted, to inhibit web scraping by statistical nonsense software.

Is this possible?

PS: Every time you post this “I’m done”, you continue to show up over and over and over and make little chirps. Do me a favour, and please fulfil my request to delete all of my posts from this forum.

Thanks.

It is not, sorry. You can ask @bartv for a special exception, but we don’t delete posts as to not break existing conversations

Even in light of the nightmare of statistical scraping?

I’d really appreciate it.

I believe it is mainly a language barrier. Troy’s language is extremely scientific and technical, but the issue is that this is being said in a form full of artists who just care about the results constantly improving from a color and a realism standpoint (since try as we might, we will not be able to fully understand what is being said).

We want to promote understanding in this case no matter how long it takes, not chase people off the forum. Keep in mind that losing Troy_S from the Blender scene may end up being a long-term detriment to advances in color workflow because he will just take his talent and know-how elsewhere. I hope you are ready to pay for software that contains the latest he is cooking up.

And people wonder why many of the devs. see this forum as just noise.

2 Likes

Has less to do with a language barrier, and more to do with applying any of this thread to actual production.

When thread participants are continually told things like “curves are meaningless, math is silly” - then the thread loses most any value as it applies to actual production… as the controls we deal with, to change color - are curves, and adjusting math values.

Being continually berated, because we have actively decided that we dont’ want to discuss color correction techniques in terms of “how we feel about it” and “our brain doesn’t even know what color is” and “let’s have a cognition” - yeah, not helpful nor particularly of much pragmatic value. We’re making digital images, not … sitting in a drum circle and smoking peyote with Stephen Hawking.

We just had one participant drop out of this convo due to the above, who apparently had SOME idea of the techincals of the topic, and how to apply them. So which is the bigger issue - that, or your personal dislike of Joseph?

4 Likes

AgX got into Blender did it not? That would count as an application of what was talked about in this thread.

Blender still allows people to have notorious six, color ringing, and burnt-out highlights in their work so nothing was taken away here. The new whitepoint feature also allows for some that Hollywood teal and orange grading to add on top of it as well.

My heart honestly breaks when I see top-tier developers like Troy get bashed like this, because I know that a lot of applications tend to rise and fall with the quality of its community. Blender is the only thing FOSS has in terms of a full DCC suite, let us not ruin it.

6 Likes

I’m sorry, you’ve said on multiple occasions that Blender devs completely ignore all this because noise… so which is it?

I honestly doubt that’s going to happen; standard forum practice has been that accounts can be anonymized, the posts remain. Everything people converse on this - or any website - shouldn’t be broken apart simply because a participant decides they no longer wish to participate.

No one is attempting to ruin Blender. Saying “this conversation, or this or that PART of a conversation, is not constructive” is not ruining Blender.

clipping is easy to detect from a gradients perspective: It’s the same as banding: The color suddenly stops varying smoothly in basically the same way.
Only real difference is that the clipping really “is happening” data-wise, and presumably can’t be resolved by such simple tricks as, say, dithering.

The thing is, you beat on and on about this on your drum, giving the same five example demonstrations over and over, but there isn’t really anything to show for your current preferred approach.
Like, yes, those visual demonstrations are very interesting and all, but what would that actually mean for a color management approach?

It’s not just, that you keep being negative and hostile towards people who, perhaps, know less than you, or simply are coming from a different angle than you, although that is a big part of it. It’s also that what you give instead has nothing, like, obviously actionable.

So all that negativity ends up doing is make people feel bad, with not a single glint of how things could be fixed. Not that people should need to fix a situation they didn’t cause in the first place. But not having any path forward whatsoever makes this even worse.

10 Likes

Hi everyone,

I’ve been following this forum for two years now, reading every thread with great interest and admiration for your discussions and progress. Although I don’t yet have the knowledge to actively contribute, I’m trying to learn as much as I can, hoping that one day I might be able to add something valuable to your conversations and collective efforts.

Lately, however, I’ve noticed that the atmosphere on the forum is becoming increasingly tense. I understand that the topic you’re working on – managing color in an innovative way – is complex and requires both creativity and precision. In such a challenging project, it’s easy to feel frustrated, especially when progress doesn’t come as quickly as we would like.

It seems to me that everyone here wants to contribute to the development of this concept in the best way possible, but concerns about different approaches might be leading to unnecessary tension. It’s important to remember that each of you brings something valuable to the discussion, whether it’s deep mathematical knowledge or a fresh perspective on the problem.

Perhaps it would be helpful to try to understand and appreciate the different perspectives. Creating separate threads for various approaches might help develop them without conflict, while also providing space for collaboration on the shared goal.

I’m confident that by working together and respecting each other’s contributions, you can achieve something truly exceptional. From my vantage point, I just wanted to express my support and hope that you can continue this project in a spirit of mutual understanding.

4 Likes

troy_s in my eyes it’s like Michel Jordan, who wanted to motivate his colleagues by often being rough with them. But it was because he wanted to achieve high goals. I think it’s worth considering whether these are personal insults or a desire to stimulate activity that should be understood in a more human way. Would it be helpful to designate a map/route for such a complex project?

4 Likes

I too hope that Troy decides to stay around. His communication is abrasive and brash, but I find the conversation interesting, even though I too can not add anything to it. To my mind this step is/was inevitable, technologically we are now at a point where this makes sense to get a grip on color management. Before we didn’t have HDR displays and the GPUs to handle them. Now that the technology has matured some, this is the next logical step.

One of the problems I see with this is that it’s like ray tracing from the light source instead of the camera. We have a lot of variables on the human side that we can’t yet account for. This makes it difficult to find a general solution (if there is one?). In the future maybe everyone will have a profile of their sight that can be an input into the color management of an image?

Troy can be harsh, personally I have never have never been treated badly by him. I have read everything he has posted , and even when I don’t understand at all, I have kept my mind open to it, and over time I have realized that most of the time when he gets defensive its because its something has already been explored before without any good results from it. I have been able to get some really good results by following his ideas and exploring different ways to deal with them.

Many people have applied my own tools and image formation in many productions, short films, documentaries, commercials and even I have recently used it in a CG productions.

My take on AGX ideas its called 2499, the main version is a DCTL for Davinci Resolve as I believe OCIO can be extremely limiting , but I have made and OCIO version too.

My Version:

Genco made an even better version of my OCIO, along with some great alternatives:

3 Likes

I’ve also been following this and similar forums and trying to understand what was being communicated. A lot of Troy’s recent messages do read as frustration, and as of now he really is considering account deletion
I kinda hesitated to pitch in since I’m no expert on digital color (not like I would have changed much), so thank you for taking the words off my fingers

We- myself included, I recognize and admit my failing- have gone far off-topic here. This thread is about color science, not discussing users. Please stay on topic

3 Likes

Thing I realized when having my first chat w/ troy_s is that ‘Colour’ is science and ‘Looks’ are alchemy. That suddenly crystalized my perception and made all confusion go away.

2 Likes

Inversion of the CIE XYZ Color Space Transformation

Today, I was looking at bright light filtering through a curtain in my room. My eyes gradually adapted to the bright light without causing the other parts of the room to appear darker. However, when I focused on the darker areas, the curtain seemed to appear brighter.

CIE XYZ and Color Perception

In the CIE XYZ color space, the brightness of a color (luminance) is represented by the Y component. Our eyes adapt to different levels of brightness (Y) and compensate for differences in luminance by altering our perception of color.

Adaptation to Brightness: As our eyes adapt to high brightness levels, the Y component might stay constant, allowing us to perceive bright objects, like light behind a curtain, in a stable way. Changes in the perception of other objects in the room are due to visual adaptation rather than actual changes in the chrominance values (X and Z).

Perception of a Brighter Curtain: When you focus on darker areas, the Y values for those areas are low, which increases the contrast between them and the curtain, making the curtain seem brighter.

Inversion of the CIE XYZ Transformation

In the context of the CIE XYZ color space, it could be considered that when inverting the transformation:

Adapting to Brightness (Y) Alters Our Perception of Chrominance (X, Z): When our eyes adjust to bright light, the luminance (Y) can influence how we perceive chrominance. If Y is high (bright light), the chrominance might appear less saturated, causing the curtain to appear more neutral in color.

Adapting to Darkness Increases the Perception of Chrominance: When Y is low (dark areas), the contrast in chrominance (X, Z) in brighter areas, like the curtain, may become more pronounced, making the color appear more vivid.

Has the Approach of Obtaining Y First and Then Subtracting X and Z Already Been Tested?


I took a quick screenshot to illustrate this.

1 Like

It may be a bit late to post this given the circumstances, but I’d like to know how robust this is as a formation path:

I haven’t tried it yet, but if someone is willing to purchase this I’d like to hear from them any strengths and flaws that it may have. The developer set up shop just last month, and I haven’t seen his products recognized anywhere outside of Reddit so far