Improving the Camera: The Image

[Hrvoje Simic]

Thanks. Looking forward for the MX & Helium versions.

Could you post one image but with all versions of the prime on? So its easier to just se how they affect the image. You can alway mark the one that you think it appropriate to use on the given image.

Sure, coming up today.

 

[Hrvoje Simic]

4 examples, I've modified the naming convention a bit.
All have differing contrast, but keep the full DR, not slamming signal extremes. Also, saturation levels stay put. Less visible difference side by side, more visible in scopes and toggling between them. That is a good thing.

This route allows quick toggling between luminance distributions without the need to compensate saturation. It also allows more flexibility in approach to contrast while grading into the Primer. Each will predetermine the outcome and since they are finely tuned per sensor response the risk of noise enhancement by particular curve modification in post is reduced.

01 Base

Very mild contrast with most room for shaping toe and knee. Best for high DR scenes and "milky" filmic routes.

02 NTR


More defined and medium contrast while keeping mids stable.
Best for judging accurate exposure with minimal strain to signal.

03 Vivid

Stronger toe and knee.
Best for punchier imagery.

04 Push

Additional push upwards while keeping the blacks and highlights.
When you want/need that extra push but want to avoid noise trouble.

 

[Adrian Jebef]

Best thing I've seen in awhile. You're on Steve Yedlin's wavelength. Nice work.

Want any Alexa65 clips to play with?

 

[Chris McKechnie]

Very interesting indeed. Currently using the Alexa LUT for helium, but def'n interested to see what your helium primer looks like once it's ready. Let me know if you want helium content to play with.

 

[Hrvoje Simic]

Best thing I've seen in awhile. You're on Steve Yedlin's wavelength. Nice work.

Want any Alexa65 clips to play with?

Thanks Adrian.
Sure, hit me on exert at omeneo dot com. I'm doing Alexas sometime next week so I can see how what translates.

Very interesting indeed. Currently using the Alexa LUT for helium, but def'n interested to see what your helium primer looks like once it's ready. Let me know if you want helium content to play with.

Sure Chris, same mail.

 

[Chris McKechnie]

Tried PMing...what's your email?

 

[Andy Roberts]

Tried PMing...what's your email?

exert at omeneo dot com

 

[Joel Arvidsson]

Nice done, like you say, its very obvios how the different primers affect the images when toggling. Im sold on this, hoping we will se primers for the Helium soon. When do you think it will be ready. In a month or two or NAB?

 

[Aaron Lochert]

I have some free time this month to shoot some Helium test images for you, Hrvoje.

Is there anything you need in particular when doing this analysis?

 

[Hrvoje Simic]

I have some free time this month to shoot some Helium test images for you, Hrvoje.

Is there anything you need in particular when doing this analysis?

That would be very useful. I have mid grey weather over here plus a few projects in post.
You can get in touch on exert at omeneo dot com. As well as anyone willing to help.


...



Dragon.

ISO 800, 5600K

1:1



RG3 DC2

Log3G10/RWG & 02_NTR Primer

 

[Craig Parkes]

That's amazing work Hrvoje, the separation of tones between the micro hairs on his lip and the underlying skin looks incredible, and the fact that the face skin tone is so boosted but the thumb and neck appear only mildly so would almost make me suspect a selective key. Would love to see the impact of this combo on darker skin tones as a comparison.

 

[Mike Nagel]

Hrvoje,

could u run the primer of your choice on this RED test image available here - 3rd image, the girl

http://www.red.com/sample-r3d-files

Thanks.

 

[Hrvoje Simic]

That's amazing work Hrvoje, the separation of tones between the micro hairs on his lip and the underlying skin looks incredible, and the fact that the face skin tone is so boosted but the thumb and neck appear only mildly so would almost make me suspect a selective key. Would love to see the impact of this combo on darker skin tones as a comparison.

Thanks Chris.

The impression of "boosted" is mind playing tricks after the first image conditions the colour interpretation. That's the effect of chemistry of visual system. Try looking at the second image isolated visually and after brain "forgets" the first.

The selective key association is the projective effect of analytical property of mind, using known references in attempting to rationalize the new experience.

No keying or grading on these. Signal is re-mapped in transformation from base color mapping to viewing in micro particulars and once the hues and their separation are adequately treated, their perceptively natural relations come out and a rich palette emerges. Digital typically doesn't do that, only film. Unless you are Fuji.

Hrvoje,

could u run the primer of your choice on this RED test image available here - 3rd image, the girl

Already did.
Posting it online - not gonna happen. I learn from my mistakes.

 

[Wayne Morellini]

You are way off.

I love your creative mind, desire for improvement and will to share your thoughts and ideas with community. Sometimes it seems that it goes a bit overboard with expectations and few assumptions here and there, but that tends to go with the package of a progressive mind. Some don't do anything and don't have these challenges, and some wonder, dig and explore, hitting bumps on the way. I can just share my experience to help. Nothing I write is meant to get your Spirit down. I understand it might have seemed the way you interpreted but English is not my native language and word interpretation and phrasing differs between languages and cultures in context and tone and it takes time to develop an adequate transform. I'm not working just on optimizing cameras.

Anyway, don't let some of the comments put you down, guys mean well. Think of these as just friendly pokes and this forum as the interactive pool of consciousness balancing itself through feedback of various perspectives. We are all on the same side, it just takes joint work to properly pass and assimilate the feedback.

Best vibes sent to the creative mind in the land down under. Cheers, man.

Thank you for your gracious reply Hrvoje. I try to be gracious in my replies (well reasoned is a form of graciousness) but sometimes some people around just do my head in. No amount of love or reason seems enough for them. A few barely related perception examples of the norm most of the time: A few days ago I rang up about another parcel that turned up 2000 miles from me, the call was dead except echoing got my own voice, which sounded totally alright and 10x better than I expected, and wondered why people had trouble hearing me. Just wait for this. I call back, I get someone and give them the tracking number, they stuffed up the first 3 numbers three times. I say 0009.. They say 009.. I say triple 0.9, they say 029 and it goes on like this, when it starts, I think why does it happen to me everytime knowing my voiced is impeccable at the time. Being gullible, I just keep repeating it simply. I go to see about an overdue amount to pay that got missed , the woman seems to be imperviouse to understanding the obviouse thing I'm wanting to pay, and repeatedly going onto something seperate in misunderstanding 5 times (a result of her jumping to a false conclusion and not listening). After I got the statement, it clicked in her head. But after having went out for a walk finally and gotten sick from it hours before, and standing in line for 10 minutes I just politely gave up and left having trouble walking and making it beck to my car (been trying to get in there for a few weeks). This stuff all of the time, and not even the really bizarre odd and statically side out there stuff, that all happens most of the time, that just rarely happens in real life. In the minor cases, and in general human reaction, people have disregulated emotional reactions to things in analysis, a divergence to good analytical ability. They go off on some tangent to true analysis, rather than keeping it real and seeing such things as an aspect to the analysis (meaning, is it really offensive or wrong, and wherever, and how, the emotional aspect integrates into the analysis. They say the spiritual stuff doesn't exist, but given extreme statistical odds there has to be someone where a lot of things go wrong. But you can't do business like this. However, same in general in this forum.

However, what I am on about isn't as you say, somethings are friendly, or sarcastic, jibes, some are just not listening hearing understanding disruptions or actual deliberate attempts by people that are of limited intelligence . I'm really happy to see real progress, objectively, I put into others progress, I might say if it is wrong minded, but the objective is good progress. I'm happy to see them suceed, but a troll I'm not so happy to see succeed. I don't remember if there is a real troll here (a troll, not in the sense of somebody that disrupts purposely, but something worse than that), some arses, but maybe not a real troll. At least that is something. However, we are not talking about what you say, but purposeful disruptuve intent instead of listening understanding or hearing, a day they tend to keep circling coming back with new areas of disruption when faced with the truth, sometimes they pretend not to get it, both minor trolling.

Perspectives, and sspects. At the moment I'll have to leave this, I have being having some trouble writing, as I am watching stranger things as I'm writing, and it is drainingly tedious. This 80's nostalgia is annoying. Couldn't they shoot it with modern values but a conservative non cliche 80's timeline. On top of that, my assessment yesterday says that the very small amounts of damage are probably causing my slow down in thinking and memory issues. So, I'm going have to rest. But it is about aspects, not just perspective, problems and solutions are made up of aspects. Perspectives are tainted, you acknowledge that aspect right or wrong, as an aspect, a concept, and go on. I've spent tens of thousands of hours on technical things. Thousands of scientific and engineering periodicals, many more, mostly technical research, web pages and forums as is now the technical side of things. You get to know how to design through or around bumps. You have to consider the aspects to break through the limitation of perspective. Sorry for the philosophical design stuff. The higher you look up the tree, often the more things change.

 

[Mike Nagel]

Already did.
Posting it online - not gonna happen. I learn from my mistakes.

well, u do keep the surprises coming. I have no idea how to understand that statement ;-)))

staying away from (mis-)interpretations of ur response: u offer sensor optimization for RED. this well lit RED footage. What seems to be the problem ?

crop out the chart, don't care about that.

 

[Wayne Morellini]

Hrvoje, I like how you have pulled down the top end and emphasised its color. I hate how they render bright skies on 12 stop Sony sensors used on cheap cameras. I imagine Gh5 owners might be appreciative of that. I certainly would be interested in using it.

 

[Hrvoje Simic]

well, u do keep the surprises coming. I have no idea how to understand that statement ;-)))

That it is for the good.

I am irrelevant here. If you leave Egos out of the equation and replace the barrel with a Cave, outcomes should improve.

staying away from (mis-)interpretations of ur response: u offer sensor optimization for RED. this well lit RED footage. What seems to be the problem ?

There is no problem, Mike.
Just ethics, psychology and properties of light and visual perception.

Problem often emerges from desires and projections and so far you have contributed a few times in that department on this thread. If didn't have huge respect for your work, saw a bunch of great posts you made and knew you mean well and have firm basis, after a few comments you made I'd ignore your posts. It is in human nature to assume the negative first so I kindly ask you not to and I'd still enjoy a talk over a drink.

 

[Hrvoje Simic]

Thanks Wayne, I appreciate your input. All cool.

 

[Mike Nagel]

That it is for the good.

I am irrelevant here. If you leave Egos out of the equation and replace the barrel with a Cave, outcomes should improve.

And now we're talking. Because that cave is my home ;-)

There is no problem, Mike.
Just ethics, psychology and properties of light and visual perception.

Problem often emerges from desires and projections and so far you have contributed a few times in that department on this thread. If didn't have huge respect for your work, saw a bunch of great posts you made and knew you mean well and have firm basis, after a few comments you made I'd ignore your posts. It is in human nature to assume the negative first so I kindly ask you not to and I'd still enjoy a talk over a drink.

buddy... I'm gonna be very honest with you because everything else does not help you or anybody else. Your solution can only be very limited. It can never be a silver bullet for all footage. Calling them primers is a very good choice. And as you know best, there is a reason there are various primers.

Because a LUT is static one trick pony. It cannot know the light properties (of a given shot) and it cannot dynamically adjust to that, or anything else that is happening in a given frame. That just comes with the LUT approach.

Getting the hooks in the SDK and starting to address values there offers more control.

Your LUTs are very intriguing. But in most results, the 1D is off (not a problem per se). Sometimes less obvious, sometimes more. Some skin tones in some examples here are very unrealistic - because the 1D is off. Not because of the RG3/4 reference shot, because simple matter of fact the 1D is lifted (or has an unnatural curve in that particular stimuli range for this particular shot). I can see it and others here have commented on that. It will need to be adjusted. Which, naturally, will change colors. Which is all fine and good, everything is a moving target.

But that's why I've been asking for an example of a finished shot and then YOU to explain what u've done in the signal pipeline for this particular shot, ergo stating an example workflow. So that peeps here know what you (the creator of these LUTs and landlord of THE DOMAIN ABOVE THOUGHT) do to get the most out of ur LUTs easily.

Easy non-ACES example: debayer RWG > linear > pre-adjustment > map to log (RLF, etc) > normalize shot & 1st treatment > Primer LUT > adjust 1D and/or signal range > then work on keys/windows/ROI/etc.


None of the statements made here in this thread is to "down your product" (matter of fact we are all excited), because there is nothing you can do about any of this. Your LUT cannot dynamically adjust to what is happening in a frame, ergo: it will work great sometimes, and sometimes not so much. Not a problem, we'll just don't use it (on that shot).

But telling some peeps that their "eyes plays tricks on them" and that they don't have a clue yada yada yada - when there are colorists here that have treated at least as much footage as u have (that is putting it mildly) - is just funny. And they have good eyes and can see what is going on.

Naturally, you're simply defending your baby, and criticism is hard to take. After 18 months in the lonely cave of a cold barrel ;-) (This new script is writing itself !) But it doesn't make much sense to argue about something that your solution can never provide.

Now, IMO, your solution is a no-brainer purchase given the very, very fair price and the nice results that it can provide that we have seen already.

Good, now that I have your attention, let's talk about the girl. U ask which one ? Well Hrvoje, u old Casanova, the one sitting on the chair. On red.com ;-)

I am very aware - as I explained in lengths above - that ur solution is not a silver bullet. So I don't expect anything. It may work on this shot nicely, or it won't. This is publicly available footage. Everybody can use it for tests and evaluation of workflows (for this particular lighting scenario) and it's also always interesting to see what other colorists or other approaches bring to light (pun intended) - from the exact same footage.

Your primers are a new option. So why not give it a go ? I mean it's not that u can prevent the outcome. All owners of your LUTs can see what is happening on that shot and post it anywhere.

Or are u telling me that u - the master who released the beast - are a scared little chicken...... ?

 

[Hrvoje Simic]

Oh...look...what do we have here.

And now we're talking. Because that cave is my home ;-)

Cave has projections on the wall and by staring at them for too long one interprets them as real and the cave as home. While some cool things are awaiting outside...away from the security of the known.

buddy... I'm gonna be very honest with you because everything else does not help you or anybody else. Your solution can only be very limited. It can never be a silver bullet for all footage. Calling them primers is a very good choice. And as you know best, there is a reason there are various primers.

Okay then. Thank you for explaining to me the adequacy of naming I chose for my tools.

Me and a growing bunch of others are pretty happy with being "limited" this way.

No one said anything about a "silver bullet" and for Magic Bullet and similar solutions there are other addresses.
This is something else.

Because a LUT is static one trick pony.

And every camera + its signal mapping have a fixed part of signal deviations.
That part which makes colours "not so good" overall.

It cannot know the light properties (of a given shot) and it cannot dynamically adjust to that, or anything else that is happening in a given frame. That just comes with the LUT approach.

It doesn't have to know the light properties, it fixes to a possible extent what camera and base mapping generally do to the signal, regardless of particular light.

It doesn't have to dynamically do anything, that's what a colorist is for, to shape the signal according to scene and adjust for light related offsets.

Getting the hooks in the SDK and starting to address values there offers more control.

Sounds good.

You go play with SDKs and more power to you. I play with cameras and test their base response mapped in camera maker's choice and I do what I do.

All cool.

Your LUTs are very intriguing. But in most results, the 1D is off (not a problem per se).

It is "off" for your expectations. There is a very good reason (a bunch of those) for these responses being designed this way and they are related to further signal usage and approach in post.

Sometimes less obvious, sometimes more. Some skin tones in some examples here are very unrealistic - because the 1D is off
Not because of the RG3/4 reference shot, because simple matter of fact the 1D is lifted (or has an unnatural curve in that particular stimuli range for this particular shot).

Third and last attempt.

There is a reason these are called Primers. What you see as "lifted" is designed that way to allow you more room for shaping knee and toe and not having slammed signal extremes as a typical generic LUT does, deciding a part response for you, where you have to dig the data out and inherit other color space + image mamipulation tool related signal distorsions when you do. If you do not wish that as a starting point, there are 3 others.

This way you can approach the signal shaping differently, based on scene and approach preference while having color accuracy and separation improvements built in the same tool.

I can see it and others here have commented on that. It will need to be adjusted.

Yes Mike.
So you open the grading app and you adjust it per scene.

Which, naturally, will change colors.

Yes, colorists do that typically.

And these prep the signal.

But that's why I've been asking for an example of a finished shot and then YOU to explain what u've done in the signal pipeline for this particular shot, ergo stating an example workflow.

Example workflow:

1)You put the Primer on the node (not timeline)

2) You adjust WB in Raw settings

3) You do the appropriate signal adjustments in log before the Primer

4) You do the appropriate tweaks in video after the Primer

5) You avoid auto chart matching by dumb algorithms because Primers do a better job fixing the colors

6) You practice and test and get the hang of it pretty soon.

Your LUT cannot dynamically adjust to what is happening in a frame, ergo: it will work great sometimes, and sometimes not so much. Not a problem, we'll just don't use it (on that shot).

You are debating with your strawman.

But telling some peeps that their "eyes plays tricks on them" and that they don't have a clue yada yada yada - when there are colorists here that have treated at least as much footage as u have (that is putting it mildly) - is just funny.

It may be funny. It is also realistic when you know how the visual system works.

And they have good eyes and can see what is going on.

It doesn't matter on good or bad eyes.
It is how visual system works and one image conditions the perception of another.

Naturally, you're simply defending your baby, and criticism is hard to take.

Not really, just pointing to the windmills in case their maker has capacity to perceive them, like others do.

But it doesn't make much sense to argue about something that your solution can never provide.

You are referring to your strawman.

Now, IMO, your solution is a no-brainer purchase given the very, very fair price and the nice results that it can provide that we have seen already.

Thank you for your positive assessment of the purchase of the tool you have invested effort in trolling, and having issues understanding so far, thus creating an impression of your appreciation and not trolling the thread.

Good, now that I have your attention, let's talk about the girl. U ask which one ?

You got my attention with the first overthinking trolling attempt and I was being considerate to you despite that. Which you abused. A few times.
Now I'm doing hygiene.

The one shot with light pollution from three types of lighting (at least) with highly suggestive colour background affecting color impression in the foreground, which should somehow serve as a test for colour performance and attempt to satisfy your curiosity while in troll mode. Which I attempted to avoid discussing due to consideration to others but you just couldn't let it go.

Or are u telling me that u - the master who released the beast - are a scared little chicken...... ?

Shaking completely and loosing feathers due to stress.

But I'm loving your attempts in explaining how LUTs work or how the image should look like with one click of a button.