Subscribe to the Podcast:
On this episode of The Wandering DP Podcast I talk about understanding how RED Cameras change their performance capabilities based on ISO settings.
Hope it helps!
Thanks for Listening
Thanks again for listening to the show. If you find the content valuable please share the podcast using the social media buttons on this page.
Also, reviews for the podcast on itunes are extremely helpful for the getting the word out about the show. If you can take a second to write an honest review that would be greatly appreciated. The more reviews the more people the show can reach and I do take on board all the comments to try and make the show as good as possible.
And lastly, don’t forget to subscribe on Itunes or Stitcher to make sure you are always up to date with the latest episodes from the podcast or even better, sign up to The Wandering DP Newsletter to get automatic updates on new podcasts as well as blog posts and exclusive downloads. To join click the link below.
How bright do you tend to go to get a fat neg but still have the ability to make the image look dark in post? Is your go to technique using a lut or do you stop it down for a sec to show the director?
Do you preview the raw/log view on camera and use a separate lut on the flanders?
On the last few jobs I have used REDGamma4 which is much more contrasty than the other REDGammas. I don’t output a different LUT to the flanders.
I don’t use a stopped down LUT I just know some DPs who do.
I try and get the fattest negative making the image look how I want it. If it is a dark scene a lower ISO will get you the fattest negative. Basically counterintuitive to what you would think. The only thing that matters is how much light is filling up the bucket.
The higher the ISO the more space you are leaving in the bucket but of course you have to balance the good with the bad. You need to decide where you want your info mapped. Shadows vs highlights.
I light to my eye, check the ratios in the FSI, then check the stoplights on the epic or the raw clipping on the alexa.
If things are clipping and I will adjust the extremes if necessary.
I appreciate you trying to explain using the different ISOs on the RED. I’m wondering if you can clarify a few things for me? I apologize for all of the questions.
The ISO-DR shifting principles do not apply to the Alexa, correct?
You said that it was best to use the low light filter and shoot with low ISOs in low light, but you mentioned that you were using the filter and shooting at 1280? Can you clarify why you were shooting at 1280 and using the low light filter? Were you trying to get more DR in the highlights? Then why use the low light filter?
Also, I believe I understand the “when” but I’m still trying to understand the “why” nuts and bolts between the ISO and the DR. For example I’ll use two scenarios. In scenario 1, you are wide open at f2 shooting at twilight. In scenario 2 (like Ep 34) your background is 6 stops over and you go from 320 to 640 to gain a stop of latitude in the highlights while then closing down a stop.
Let me try to explain how I am understanding it and please correct me if I’m wrong. Changing the ISO up and down makes it more sensitive to higher and lower foot candle values respectively while keeping a consistent amount of latitude and also changing the IRE values of the image by a relative stop. The camera sensor sensitivity’s middle point (aka middle grey) would move along the latitude of sensitive foot candles… So by going up with ISO, it maps the DR up and makes it sensitive to the higher foot candle values while also making everything a stop “brighter” in IRE values. Please tell me if I have this wrong.
So my question is in Scenario 1, would it make sense to shoot at the lowest possible ISO to keep the sensor’s sensitivity in the lowest “foot candle values”. If you were to do this the image would seem “dark”, but would I be right in saying that it would seem dark because of the contrasty gamma profile? If you were to see it in REDLogFilm, you would in fact notice that you could see alot more into the shadows?
So in scenario 2, by changing your ISO from 320 to 640, you are making your IRE values a stop brighter and adding an ND (or stopping down) to negate that change in IRE value all while then shifting the dynamic range to accommodate the extra stop in the highlights?
Lastly, considering ISOs for grading/noise. Changing the ISO during production sets the shifts in DR in stone and you cannot shift it in post, correct? In a typical grading situation, if you were working with twilight footage that was shot in low ISOs, would you first start working with the footage in say, REDlogFilm so that you could have access to all of the info and use curves/levels to get to a brightness/IRE values that you wanted? Or would you stay in REDgamma4/3 and up the ISO in the metadata to get to IRE levels you were comfortable you were comfortable with?
Sorry for the endless questions. Really appreciate the help and time you’ve put into the podcast. Thanks Patrick.
No worries Matt. I will answer the best I can.
1. Yes the DR/ISO thing happens with the Alexa. That is why they say 800 ISO gives you 7 stops over and 7 stops under or parity so most people use it at 800.
Drop to 400 and you will lose a stop in the highlights. Go to 1600 and the inverse occurs.
2. You’ve got the idea more or less but I think you may be overthinking things.
Think of it like this:
The sensor is a bucket.
It gets filled with light.
The bucket is real and exists whether you can see it or not and it has a certain maximum capacity that you cannot alter.
Not enough light and there will be noise and shitty color/quality.
Too much light and you will never get back the light spilling over the sides.
You can’t see the actual bucket.
You can only see the bucket with these special glasses. These glasses are ISOs.
ISO is a tool to make the light level inside the bucket appear to more full or less full.
ISO doesn’t effect the light though.
To get a fat negative that will look tonaly interesting you want to get as much light in the bucket without spilling.
Say you are shooting a scene like the cafe one.
You look at the bucket @ ISO 320. You see you have lots of light in the bucket. Actually a little too much light.
If you boost to ISO 640 you can trick yourself into thinking that the bucket is WAY too full. So you ND or stop down to eliminate actual light (not perceived light).
When you stop down the overall amount of light now fits in the bucket. You moved everything else in the image down a stop so you could read the brightest spot.
The tricky part is finding the amount of texture (which is induced from lack of light hitting the sensor) that you are happy with in relation to the stops captured.
Max Goldman talked about shooting the Alexa at ISO 1600. He liked the texture of starving the sensor of light. It changes the way the image looks. It also changes the way he lights on set.
So while you can change the ISO in post you can’t change the way he lit it. That is why it is important to understand ISOs.
I shot at 1280 with the LLO for a few reasons.
I shot with the LLO because that is what was in the camera as most of the commercial was going to be in dark environments I didn’t have time to light.
I shot at 1280 as there was too much light to fit in the bucket at 320 and the texture at 1280 is what I preferred for the project.
I wanted to keep the highs in the high contrast areas and through testing found I could spare shadow stops when necessary.
Finally I would never grade anything in RedGamma whatever number. You are losing tons of info if you do that. Tons.
Grade in Redlog Film if you want to see the true capabilities of what the sensor has captured. Sometimes I use ISO changes to bring up a shot. Sometimes though I will just use the Resolve tools. All depends.
In the end the RED is much like the Arri or any other pro cam now. If you shoot something at ISO 800, light it well, craft it well, it will look amazing. You only really have to toy with these ideas if you are interested in squeezing the last drop from the fruit.
If I did all the litte ISO tricks on a job and Max Goldman shot the same job all at ISO 1600 on the Alexa you would notice our personal style differences (and the giant gap in our skills) long before you would notice the ISO tricks of saving a stop so it doesn’t blow out.
Thanks for the reply man. A few things you said I found really helpful.
“So while you can change the ISO in post you can’t change the way he lit it”. This really makes a lot of sense alongside choosing ISO for texture while knowing the limitations of the way it’s changing the DR.
I appreciate the help because I feel that its time to squeeze the last drops. Knowing the details makes you more confident as it becomes second nature.
Thanks for the help! Looking forward to more.
Check this out for a visual if you haven’t already seen it: