Welcome to our community

Be a part of something great, join today!

Log in
  • 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 :)

Now its confusing me....

Yes, the RAW recorded R3D images would be the same.

What I'm trying to get at is that although what you set the ISO to matters not, it is vitally important to think about what it means in terms of where the recorded code values go. That means understanding that there are two situations (probably more) when you'd go to a higher ISO:

1) it's dark and you need to see what you're doing
2) it's bright and you want to capture more highlights

Graeme
 
Ah. okay. Thanks Graeme. I think, hope, that swept the clouds of confusion away for me.

Setting the "correct" ISO increases my ability and accuracy to make better judgments: need more fill light, need an nd to prevent the clipping, etc. And B16 will just make it easier.
 
Jason Ing said:
This is my current understanding: If ISO is just metadata, and it does not matter what it's set to, then it's not affecting the actual data. It is what it is. By changing the ISO, I'm affecting how it displays on the lcd/monitor. That's it.
Click to expand...

But if you set your lenses at F22 maybe you'll only be able to use ISO3200.

Even if ISO is just "metadata" there's still the exposure decision in the game.

As Graeme pointed:

Graeme Nattress said:
If you set the camera to that value or not, it does not effect the recorded RAW data, but never-the-less, you must make that exposure decision.
Click to expand...

Edit: I didn't read the last page and I was redundant. But hey, sometimes redundancy is good.
 
Graeme Nattress said:
I think of ISO as a decision on where you want to place your exposure. Now, as ISO is metadata, it does not matter what ISO you set the camera to, but, you must think about ISO, where you want to place mid grey in the RAW data. If you set the camera to that value or not, it does not effect the recorded RAW data, but never-the-less, you must make that exposure decision.

Graeme
Click to expand...

Can that decision be made in post, as its meta data, if the raw R3D data :detective2: is unaltered?

edit: I think that was covered already, so question 2, Can`t equally good decisions about how to spread your image code values be made using histograms or data analysis in post?
 
I imagine not if you didn't expose properly. My interpretation of what Graeme is saying is to expose for your ISO (lighting, iris, shutter) and dial it in when you "expose" in Redcine/ALERT.

If you shot for ISO500 on set, you're probably going to end up there in post, meta data or not!
 
Sorry to add another voice to the rabble, but i think i can help.

So for instance you light a subject with one 300w light(a lilli for example) and you set your f-stop at like f22 or something. This is going to be way underexposed no matter which way you look at it(if you can see anything at all).

The ISO metadata isnt going to help you at all here, if there is nothing there, there is nothing there. You can bump it to like ISO3200 and all you'll get is a screen full of noise.

However if you shoot it wide open and for arguments sake is "exposed correctly"(you be the judge of that) what the ISO can do for you is either bring more detail in the shadows out(if its there) with possibly the cost of a bit of noise at say ISO500, or conversely clean up the image by dropping the ISO to say 100, so you get a real clean nice image, but at the cost of not being able to see shadow detail. in terms of the 18% gray, what your saying in the first instance at ISO500 your saying i want 18% grey to be like (picking a number out of thin air) 3000 out of the 4096. So everything gets boosted up and you cna see whats in the shadows, at possibly the cost of clipping highlights. Or you can say i want 18% grey to be like 1024 out of 4096, so you get like a 3/4 of the range for anything above 18% grey. and only 1/3 of the range for things below. And the metadata just tells the application to make that change. This results in clean shadows (because its all black) and nice gradations of highlights (coz they have 3/4 of the range to spread out across)

Now on set what the iso can do is allow you too see what your going to see, and play around with noise/cleaness shadow/highlights.

Obviously usual disclaimer about me not really knowing much and if im wrong please correct.

Cheers
Mark
 
Jeas the more I read the more questions I have, I'm inclined to ask a few basic questions to try and clarify a few matters:
It is clear that while shooting, the monitoring is done using the chosen ISO via what ever LUT is specified in the camera's menu (default REC-709). This information is stored in the metadata separately to the raw. This is so we can reference this information later in post when cc or encoding for a variety of purposes. I.e : on shoot we will be able to better predict the resulting look of the footage when dealt with using an HD LUT such as REC-709 in post. There for it is important that the redcam histogram references the post LUT footage in camera. Am I still on track?
1) Why do I keep reading the importance of placing the for instance 18% grays at a certain code value between 0-4095? In what way does this effect the recorded raw footage if at all?
2) Why are people recommending to shoot at a higher ISO on a sunny day more sensitivity means less light needed no? (This is coming from a bit of still film experience I've got).

Sorry for the amateur question been working in the industry for years and have never needed to know any of this. Always believed in being as savvy as possible and seeing as we're all posting in regards to these topics. As well as this directly effecting the nature of the work on our productions I would really appreciate a clear answer.
Thanks.
 
Where you place mid grey defines how many stops above, and how many below you have. Say there's someone standing in a field on a bright sunny day. You set the camera to a low ISO and get it so that the person is nicely exposed. Low ISO means high mid grey code value, which means not much highlight lattitude, but a very low noise image. You've blown out the sky. Set the camera to a high ISO, get the person exposed right. Now you've kept the sky, but the person is a bit noisier.

That's the thought process. You don't have to set the ISO in camera, but you must go through the thought process. Once you've thought out your exposure, why not set the camera ISO to match your decision?

Graeme
 
Hi Graeme, I like the sound of Redspace.

One of the problems I had before, was in some situations I'd underexpose to avoid clipping, but the rest of the image was falling into darkness, but I never knew how far to push it. Sounds like this will help me.

But I have a Question : I did find it sometimes hard with a high contrast scene to make everything fit nicely into a correctly exposed image. In particular Im thinking of one shot where : if i drop the curves in redcine the sky looks deep and blue, but my subject is dark and muddy. On the flip side, if I use curves to try and bring up the mid-levels of the subject and make it punchy, the sky goes washed out in colour. Perhaps this is just an example where there's too much contrast in the scene and I needed some fill, but does redspace help to push the lower-mid tones up into the upper range?
 
Yes, I think it will help you, and yes, that is one of the goals.

If you can't get midrange and highlights together, that might be a case where you leave the midrange darker, and do secondaries on the mids in colour correction?

Graeme
 
Thanks Graeme. I'll look forward to using it.
 
Don't forget that these decisions in ISO can further mitigate the use of ND's, shutter and the aperture setting all of which are not meta data. So yes, you should be making these decisions on set. The beauty of the meta data portion of this is that once it gets to post you have "room to play."
 
This has definitely been a source of confusion for many including the DPs I have to explain it to. The key take away is that when choosing an ISO you make other decisions like exposure, NDs, etc that are NOT metadata.
 
Graeme Nattress said:
Setting the right ISO is important.

Say you've got controlled studio lighting. Say there's a small dynamic range to capture. Setting a low ISO here will help, as then you'll put mid grey at a high number, and you'll record a totally noise free image. You'll see the histogram such that you're filling it, without clipping, and the image in the LCD will look good too.

But if you're in low light... Setting a higher ISO will allow you to see what you're shooting. If you can't see it, you can't shoot it. Or set a low ISO and add more light.

And if you're outdoors, with wild highlights / sky, setting a high ISO will allow you to have a nice looking image, but with mid grey recorded lower, giving ample highlight protection for a great post grade.

Graeme
Click to expand...

Why not just keep it at the native 320 and open up as much as possible without clipping?

That way you will always have a direct look at what the sensor actually sees rather than what the metadata tells you. For example, if I set the camera to 1,000 ASA today on build 15, the meters on the camera will warn me of clipping long before the sensor/RAW data actually clip. Accordingly, I can never judge the true sensor exposure.

If I leave it at 320 and expose accordingly, I can always tell exactly what's happening.

No longer valid with build 16?
 
Brook Willard said:
Why not just keep it at the native 320 and open up as much as possible without clipping?

That way you will always have a direct look at what the sensor actually sees rather than what the metadata tells you. For example, if I set the camera to 1,000 ASA today on build 15, the meters on the camera will warn me of clipping long before the sensor/RAW data actually clip. Accordingly, I can never judge the true sensor exposure.

If I leave it at 320 and expose accordingly, I can always tell exactly what's happening.

No longer valid with build 16?
Click to expand...

If you do not have enough light in your scene to see anything on the monitor at 320, then you will have to go to a higher ISO. When you open up your clip in RC, you are not going to jump back down to 320 because it would be underexposed again. So if you are going to end up at 500, might as well view 500 while shooting.
 
Hmmm. I generally shoot everything at 320-500 depending, then process at 200-250 with a curve to compensate.

In build 16 will we shoot at 500 and still process at 500, or will we still be "reeling" in the highlights by lowering the processing iso?

Also, In build 15 if I lower my processing iso to 100 will I be in effect, softening the image, even if I bring it back to proper exposure with curves?
 
You should always try to "ETTR" or expose to the right...

Graeme
 
You must log in or register to reply here.
Back
Top