Welcome to our community

Be a part of something great, join today!

  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Profile: SC 4.1 250ISO

I have been asked a few times "if Red is rated at 320, why do you have to set your light meter to 160 for an accurate reading?"

Because the lightmeter puts 18% grey in the wrong place.

EDIT: More completely, the lightmeter puts 18% grey in the correct place for film negative, but the wrong place for the RED ONE.

EDIT: Even more completely:

If you imagine the histogram as being backwards, then 18% grey is in exactly the right place with the meter at 320 and the camera at 320. However, as the RED ONE is not film negative and instead is recording a positive image, you need to place your midtones higher. The range is still filled appropriately with an EI of 320.
 
Thank you for offering this profile and the attendant discussion, Macgregor. I'm looking forward to testing it...these are the types of generous acts that keep me coming back to this website.
 
The profile is very interesting, especially after I've discovered during a shooting some expert guys did a similar thing setting values in the camera instead loading the profile.
The values were quite similar although the ones of Mcgregor are more complete.
I've tried to load a clip in apple color, it keeps all the values of the profile except the curve but the look is very similar to the one in redalert. I think in some way apple color maintains the curve values although they are not accessible as in redalert. Is it correct?

I believe you are correct. Programs reading through the SDK should honor curves set in camera/redalert and display them, but will not load them as editable. Otherwise the footage would not look the same.

Specifically with Color, Apple uses bezier style curves while Red Alert does not, so the curves from cam/RA cannot be brought into Color as settings.
 
This is really generous of you Macgregor - I've been doing some testing with this profile and I'm REALLY impressed. I've been getting some very "filmic" results. I hate using that word, but I can think of nothing more appropriate. Thanks for saving my highlights!

Andrew Wilding
 
I believe you are correct. Programs reading through the SDK should honor curves set in camera/redalert and display them, but will not load them as editable. Otherwise the footage would not look the same.

Specifically with Color, Apple uses bezier style curves while Red Alert does not, so the curves from cam/RA cannot be brought into Color as settings.
Yes, for this reason I said the look is very similar in Color, not the same but really close. I think if you shoot with this profile and stay in rec 709 also the raw and the proxies are quite close (the raw is better of course) and I think this sort of consistency is useful.
 
I can't seem to get the .RLK file to be seen when I try to import the look. I've tried resetting looks, user profile, etc. Nothing seems to work. I'll keep trying. Anyone else experience this?

EDIT: I figured it out. I didn't have a Profiles folder created on the SD card (Don't worry, I slapped myself, no need to pile on). Once I dropped Mac's look file in there, it saw it and I was able to import it.
 
Just got onto this thread. Looks like I was missing on some quality education through some quality entertainment. Too funny. :rofl:

Good work Macgregor. This is something I always wanted to set up myself but never got around doing it and I could not have done it better anyway. Thanks for sharing; really appreciated.

If only Sanjin took the time to read/understand what you wrote instead of... :head_explode: ... but then, we would not have had the entertainment part.
 
Hi steve, having the same problem.

What do you mean create a profiles folder on the sd card?

I would love to try it out.

You can create a folder on the SD card called Profiles or just export a look profile in camera to your SD card which will create the folder for you. Then drop SC look profile into that folder. It will now appear when you go to import looks. I believe the folder name is in all capitals (I'll have to double check).
 
Hi all, this last weekend we had a red one workshop in Madrid where we did some tests of how the ISO settings affect the image. We also compared the SC 41 against different ISOs to explain who it works.

Take a look at this:

320.jpg

Shot exposed using false color and the grey card. 4K 5000ºK +80C. 320ISO. Master Prime 40mm. B17. Overexposed areas are completely blown out. No way to recover in post.


640.jpg

Same situation, incrasing ISO setting to 640. Exposed using false color and the grey card. Image looks quite similar: highliths are shown clipped in LCD screens as well as Redcine (but in reality, there´s some hidden info in the highlights that you could recover in post by lowering the ISO back again to 250, pain in the ass). Focus is on the fruits not on the model, sorry.


sc.jpg

Again, same lighting conditions, but this time we loaded the SC 4.1 look profile. Exposed using false color and the grey card. No overexposed areas anymore. The image looks the same in LCDs and Redcine, and QT proxies.
 
The TRICK:

Now let´s see those same screen caps in linear mode:

320linear.jpg


640linear.jpg


sc41linear.jpg



You see? The shot taken with the sc41 and the 640iso setting look about the same. The reason for this is that the grey is mapped in the same zone in both cases. In the 320ISO setting image, the mid gray is mapped in a zone too high to achieve enough dynamic range to resolve the contrast situation of the lighting. And that´s the reason why using look profiles with curves, such as the sc41 help creating better looking images and more control during shooting.

We concluding during the workshop that the ISO setting on the RD ONE should be called differently to avoid confusion. Zone system setting maybe? "IV" mode vs "V" mode?...
 
Really enjoying the work you're doing here, Macgregor...not just the testing, but the artistry of your other work. If you're ever in New York, let me buy you a drink...
 
Or you could leave the ASA at 320 and just expose more for the highlights...

This shows the problem I have though with how the metadata is read and converted -- you'd think at 640 ASA rating, you'd see an improvement in highlight latitude, but since there is just a simple overall signal boost to convert it to 640 ASA, you get the same amount of clipping, though there is a stop more highlight information in the highlights (and less in the shadows) if you look at the RAW files at 320 ASA.

So for dailies converted from RAW using the metadata, you would get an inaccurate idea of your dynamic range recorded when you change ASA ratings. The Red Alert conversion should be doing a bit of a gamma curve adjustment or something to take advantage of extra highlight information when using higher ASA ratings.

It's also why when beginning a D.I. you have to remind the people doing the conversion to ignore all the metadata, or just convert everything at 320 ASA at 5000K.

What I don't understand Miguel is why your special profile has to start with a 250 ASA base instead of 320 ASA, if the point is to hold more highlight information. Couldn't you get the same results by playing around with the profile but at a 320 ASA base?
 
I did some tests previous friday on this, haven´t had the time to process and analyze the images just yet.

The idea of using the 250ISO base (as I have understood it) is that the RED cameras LCD and False Color actually corresponds with a normally set up light meter - when using the combination of the SC 4.1 profile and the REC 709 viewing space.

I guess the reason Macgregor did this, is that when having your RED setup without any gamma profile (LOOK file), in 320 ISO, and viewing in REC 709 mode, you will tend to stop down about 1&1/3 of a stop, compared to what your lightmeter tells you to do, set up at the same 320 (@24fps@172.8 f.ex).

So the SC 4.1 profile is supposed to do three things (again, as far as I have understood):
1- Put middle grey in the middle of your tonal scale available in the RED
2- Makes RED ISO more like lightmeter ISOs compared to the cameras actual response
3- Allows you to see this picture more accurately on your monitors on set

My experience from the set of the test was that it seems to do this - but when it comes to how accurate it actually is, and if I´m happy at exposing my midtones at tha cameras effective 640 ISO - I will have to see when we have done our analysis.

PS and thanks to Macgregor for sharing his experiments in this field.
 
I still don't understand why you have to select a slower-than-320 rating in order to create a profile that exposes your midtones more in the 640 ASA rating area... if the point is to decrease clipping, why start out with a slower ASA rating like 250? Why not, let's say, 400 ASA instead?
 
I guess because you´re basically fooling the system - you are putting a different interpretation of your visual information than what was designed by RED.

REDs original design is married to the ISO of the camera, and as this original interpretation is putting your midtones very close to the highlights in your recorded range, you cannot just increase the ISO overall - you must curve the gammacurve from its original linear appearence (linear here only in the meaning of how it appears in REDAlert), and thus increase midtone ISO while keeping your highlights hanging in there.
 
This of course not only means that your camera, dailies and editing material becomes more WYSIWYG, but that you will expose - and then probably also light - differently from the original setup of RED1.

This part is what I am most anxious to look into in our test material.
 
I have been using the SC 4.1 profile for a few jobs now, and it's very good. Better protection for whits and better looking qt proxies.

I guess you can set the iso to 320 and have same effect, but if set to 250 it matches a real lightmeter better.
 
Back
Top