Pages: [1]   Go Down

Author Topic: RLF = Cineon  (Read 23440 times)

fredjeang2

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1376
RLF = Cineon
« on: July 24, 2015, 07:09:02 pm »

It is the same curve indeed.



Ps: should I delete this thread? It seems that it does not interest anyone no?
« Last Edit: July 27, 2015, 03:38:05 pm by fredjeang2 »
Logged

fredjeang2

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1376
Re: RLF = Cineon (edited with images better than words)
« Reply #1 on: July 25, 2015, 09:04:37 pm »

Lol...watch the scopes! So true. I will put this
As my mantra in the forum signature.

Yep, doin the editorial luckyly,
So as you point lots of sins can be fixed in editorial
Style.

In fact this topic came by accident because I could
do it by the traditional way in fusion and resolve, actually didn't even boder to use the flat log sauce but kept redgamma4 as it.
Just that I thought about something alternative and
It became a research and an obesession to understand their color science.
But you know me,
I'm a little bit like the pitbull: when got something
In my teeth, I bite and don't let it go.

Yeah, those 320 isos are strange. I was expecting also 800.
The car shot is also 320.

Why people are shooting Red at iso 320?

Ps: and I agree with you, I'm not obsessed with the flat files. Flatten is not going to cure all crap. And when things get really nasty, it requires the talent of specialists.
But at the same time, the more I try, the more I learn, more tools I got in my arsenal and less threads like those I'll have to post.
So I'm not for flatten all files just because it's going to magicaly compensate sins. The thing with Red, is that their redgamma is a curve on top of the redlogfilm.
So it's good for general purposes because it makes life easier, but the purest untouched material (not talking about the linear) remains their redlogfilm. So in the end, the flat Red curve is the one to use for complete CC control because it doesn't have any fancy stuff on top of it.

Now, should cameramen shoot redlog? Ummm...in an ideal world many say yes. I know, clients pull their hair-off when they see a flat image on the monitors, but can't we Lut the client's monitors?


Cheers.
« Last Edit: July 27, 2015, 03:41:33 pm by fredjeang2 »
Logged

D Fuller

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 608
    • AirStream Pictures
Re: RLF = Cineon
« Reply #2 on: July 28, 2015, 12:11:31 am »

It is the same curve indeed.


Ps: should I delete this thread? It seems that it does not interest anyone no?

Sorry, I've been busy. It was interesting, but at a quick read, I couldn't figure out what you were trying to do. I've never used Fusion, so it would have taken more time to understand it than I had.
Logged
business website: www.airstream.pictures
blog: thirtynineframes.com/blog

D Fuller

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 608
    • AirStream Pictures
Re: RLF = Cineon (edited with images better than words)
« Reply #3 on: July 28, 2015, 12:58:40 am »


Yeah, those 320 isos are strange. I was expecting also 800.
The car shot is also 320.

Why people are shooting Red at iso 320?
Who knows? People who know what they're doing sometimes shoot Red MX at 320 to get less noise in the shadows. BUT they pay close attention to the hi lights to be sure they don't clip. That almost never works outside if there's sun, but can work really nicely if you are in studio and controlling the light, or if it's cloudy so the light is flat. But like BC, I'd say you get better results more often at 800. YMMV

Red Dragon is better shot at 320. And it has the DR and ability to handle the highlight rolloff that makes it a good wy to go in a lot of circumstances.

The thing with Red, is that their redgamma is a curve on top of the redlogfilm.
So it's good for general purposes because it makes life easier, but the purest untouched material (not talking about the linear) remains their redlogfilm. So in the end, the flat Red curve is the one to use for complete CC control because it doesn't have any fancy stuff on top of it.

Nope. This is not true. And I think it may be causing you some confusion.
RedLogFilm is a gamma curve, RedGamma is a gamma curve, RedGamma2, 3 & 4 are gamma curves.

RedLogFilm is not what the picture looks like as "Raw" data, it's what the picture looks like with a curve that matches what you'd get with  a Cineon LOG gamma in a film scan. It is applied to the image data just as RedGamma2 is, it's just a different curve.

If you want to see what the picture looks like without any curve, you have to choose "Linear" for gamma.

But you're right that RLF is good for Color-Correction (the best choice, IMO) but not because it's untouched. Rather, it's because the data is interpreted on a curve that makes it easier to work with using tools like those available in Resolve (which happened to be developped in a world that delt sith Cineon gamma film scans).

Now, should cameramen shoot redlog? Ummm...in an ideal world many say yes. I know, clients pull their hair-off when they see a flat image on the monitors, but can't we Lut the client's monitors?

Cheers.

No, I don't think so. For two reasons. One is that if there are clients on set, they won't be able to make meaningful judgements about the pictures they are looking at. (But yes, you could run RLF at the camera and a LUT on the client monitors.) The second reason is more important. The DP ought to see the effect of their lighting in something that resembles what the final look will be. In the old days, when we shot film, the DP looked through the viewfinder. He was looking through the lens and filters that would affect the pictures, and he was looking at the real-world imagein its full tonal range. Today, the closest you can come to that is not RLF or some other weird, flat gamma, it's something like what the final output gamma will be. That's how the light needs to be judged.

There is good reason to check exposure in RLF occasionally to see where the highlights and shadows are falling, and if anything is really clipping (I have a button on my Epic programmed to allow me to do that), but it's not a very good way to know what your light is doing to your actors.

Just my opinion,

DAF
Logged
business website: www.airstream.pictures
blog: thirtynineframes.com/blog

fredjeang2

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1376
Re: RLF = Cineon (edited with images better than words)
« Reply #4 on: July 28, 2015, 06:48:07 am »


Nope. This is not true. And I think it may be causing you some confusion.
RedLogFilm is a gamma curve, RedGamma is a gamma curve, RedGamma2, 3 & 4 are gamma curves.

RedLogFilm is not what the picture looks like as "Raw" data, it's what the picture looks like with a curve that matches what you'd get with  a Cineon LOG gamma in a film scan. It is applied to the image data just as RedGamma2 is, it's just a different curve.

If you want to see what the picture looks like without any curve, you have to choose "Linear" for gamma.

But you're right that RLF is good for Color-Correction (the best choice, IMO) but not because it's untouched. Rather, it's because the data is interpreted on a curve that makes it easier to work with using tools like those available in Resolve (which happened to be developped in a world that delt sith Cineon gamma film scans).


DAF


Yes. That's what I was trying to say and maybe it's my
Text that was badly written but this is exactly the message
I wanted to express. That's why in my post
I wrote in parentesis "not talking about linear" but yes it was
Not very well written.

The untouched Raw is of course linear. In Fusion if I disable
The redgamma (remove curve) or use directly the metadatas
And switch to linear, this reaveals the raw untouched.

What I meant by "the purest untouched" talking about RLF, and here I agree that it was not well written, I was refering to
A gradable material to work with. Rlf is less alterated than redgamma wich is another curve added to it (a contrast curve and therefore not as "pure" or untouched. But you're
Right, let's use more precise lenguage and my frenglish parentesis
Saying "not talking about linear" was a bit confusing.
Rlf is simply one less curve than their redgamma whatever flavour. But yes, RLF is actually a curve and not the linear raw. And it's based on the Cineon standart. But redgamma is 2 curves and therefore
As a starting point for grading, it's too alterated. It works
For quick grades but not for more sophisticated looks
Where rlf is really the way to go.

In Fusion, I could see that Rlf is Cineon standart, exactly
The very same curve.
If I take the linear data and apply a standart cineon node, I obtain
Their rlf. Tested and verified.



So, sorry for my english, I sometimes write fast, and I think
My explainations are good and maybe there're not.
So for the readers a recap:

- linear is the raw datas untouched. (obtained by or canceling the gamma on the read node or switching to linear at a metadata
Level. It is the same)

- rlf is simply cineon standart log curve applied to the linear
That red called rlf. (this is what I called purest or untouched thinking of grading but Daf made me realise that my post
Could be confusing because indeed. The real purest untouched
Material is of course linear. But we're not going to use this
In an app like Resolve to color correct)

- redgamma 2, 3, 4 is another curve applied to the chain wich
Is a sort of S curve or contrast curve. It is linear+rlf (or cineon)+contrast curve.
It's function is to give a sort of fast viewable material
To work with when you do not need sophisticated grade.
It's an easy way to go somewhere quick. But it does not
Give the same latitude for grading seriously because
One more curve is one more curve, and not yours.

Now, if you're back into linear, you could decide to
Not use the cineon standart (or the rlf) and use another log
Curve. It is possible. Nobody is stucked with the rlf.
We can just cancel every curve and apply our own ones.
But it's risky. That means one has to be a colorist expert
Not to screw it. (this is when I started to screw it because
My experiment consisted in canceling all gamma and
Start to build my own curve and I'm not yet at the level
To do that so I started to put myself into a mess I didn't
Control. In other words, I was tricking the Cineon values
According to "my eyes" and that's guessing). Therefore, staying with rlf gives the
Confidence of consistency within a standart and assure
Predictible results. But a really good colorist can work
With the curves he-she wants.
« Last Edit: July 28, 2015, 02:18:16 pm by fredjeang2 »
Logged

fredjeang2

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1376
Re: RLF = Cineon
« Reply #5 on: July 28, 2015, 07:57:37 am »

Maybe cause I learned with film, but to me the world is 7 stops.

Even if you start with flat no curve digital it still breaks down to 7 stops if you work it to look like film and digital doesn't roll to the highlights and flare like film, it just hard clips.

Some better than others.



Now with raw it's not a fix all and the best thing raw does is allow you to easily match grades in post.

I have two R1's MX and one Scarlet MX.  The R1's are virtually identical, the Scarlet smoother and a little less noise.  I can match the grades from the r1's perfectly the Scarlet takes more work.

Maybe that's why I cancelled the Epic.  

Anyway, as D Fuller says, your talking about gamma and not real curves, so think of RED gamma as a slider that just flattens, but doesn't curve like a film stock.

That your gonna have to make either in camera, or in post or both.

If a DP uses the camera right, then it's gonna look on the studio or field monitor a lot like the film look.  The beauty of raw is you can move things around a little to match, but their is no magic in it.

If we're working hurried and one R1 is 100 degrees off on kelvin, or the tint isn't exact or the knee, then it's ok, because once you set a grade they will match as long as you keep the asa the same and you don't miss exposure.  

If you go too far from each camera they won't

Now in regards to the dragon, I've just seen tests, not hands on.   I don't care about 6k, but I do care about highlight roll off without that hard digital clipping.

Arri seems to have that down a little better than RED and I've thought about the Arri because of that, though I'm kind of stuck in the RED workflow and don't want to change.

I guess I should try a dragon and see, but once again I live in a 7 stop world.

IMO

BC

P.S.  I've heard the film guys forever talk about stock with a gazillion stops, though few if any of them worked in post.  They just shot to the numbers and handed it off, letting telecine and grading do what they do.

The good thing about film was they know their limits, crafted the scene to work and didn't usually exceed the values.

In reagards to asa, I can shoot the scarlet from 600 something to 1200 ok, the R1's I leave at 800 and at low key night scenes can move them to 1200 without noise as RCX pretty much handles 1200 well, though I don't make a habit of anything other than 800.

PS 2.  Even with stills I continue to work in 7 stops.  Stills are easier cause making in a sky or window is easy compared to digital cinema, but I still see in 7 stops and guess that's why I've never concerned myself when I hear a sensor will do 14 stops.

ps3.  One trick on short takes or if you have a director or production that will allow it in studio is to run a mist machine.   Fill the room, let it get almost gone and then the highlight roll off is beautiful with holding the hard details.  (don't try this on long days, long productions because everybody will lose their mind) but for short form advertising it works well.

ps4.  As DF says, always give the clients on set the best look you can on the monitors.  Don't make em guess, make em go ohh, ahh.  If you don't you spend more time talking than working and you lay it on the colorists to try to figure out what it should be.

Grading shouldn't be a fix, just part of the process.


IMO

BC


Coot, I've read your post several times and don't understand it.
I don't know what"s happening with my english but I'm not centered ultimatly. It costs me to read a writte. Maybe this endless heat wave we have here in
Europe because I see everybody alterated, tired, moody and slow.
I'm working in slow motion. Life seems unreal. I'll
Never forget the 2015 summer.

What I don't get are the 7 stops world.

Why not working with RLF as a starting point, and Aces,
Not putting any idt in Resolve so you get a huge amount of lattitude
To work with in both gamma and color?

At first it looks like time consuming but on the other hand
Who wants to drive a Ferrari at state speed limit?

I don't see Coot driving a car at the cop's speed limit anyway,
maybe this scene is closer to the reality



It's a bit the same. I don't mean by that that all crap can be
Done in capture because post is going to save the sins.
No. But why not entering post with the widest possible range of
Values?

My point is that if we have a starting point for grading that actually is built "for" Resove as Daf pointed: "a curve that makes it easier to work with using tools like those available in Resolve (which happened to be developped in a world that delt sith Cineon gamma film scans)".
And if the Academy took the time to bake Aces (for one time they are working...),
Let's use it.
« Last Edit: July 28, 2015, 04:17:22 pm by fredjeang2 »
Logged

John Brawley

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 77
Re: RLF = Cineon (edited with images better than words)
« Reply #6 on: July 29, 2015, 09:12:11 am »


Why people are shooting Red at iso 320?



Because that is the RED, RED MX, EPIC native ISO.  Or it's certainly the view of many that use it (myself included).

RED often say it's good practise to expose as if it's ISO 800 because that protects the highlights a little more.  In fact they say their sensor doesn't have a "native ISO" at all , because ISO doesn't apply in the way they work with RAW files.

I've found 320 works well and keeps the noise away.  It tends to have less headroom but more shadow discernment. (more stops of shadow less stops of highlight)

I haven't shot RED for a couple of years, I kind of gave up on them.  I'm still not seeing a lot that excites me enough to get back into using them again.

JB



Logged

fredjeang2

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1376
Re: RLF = Cineon
« Reply #7 on: July 30, 2015, 01:58:16 pm »

Logged
Pages: [1]   Go Up