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 :)

Basic rules of exposure

OMG when i look at some of my footage according to these linear light levels, a lot of it is at least 2 stops under...what tools do we have in camera to monitor the true linear light levels?
 
The problem with monitoring in linear light is it's bloody hard. I think original Frankie was like that, and quite frankly, I don't know Jarred managed to get the images he did off it. For starters, it's nigh-on impossible to see what's going on in the shadows, and a properly exposed image looks under-exposed, and hence you tend to over-expose if monitoring in linear light.

I find the best guide to be the RAW levels bar-graph. That's what works for me.

Graeme
 
OMG when i look at some of my footage according to these linear light levels, a lot of it is at least 2 stops under...what tools do we have in camera to monitor the true linear light levels?

I get the same impression from looking at a lot of my stuff in linear light... most shots I could have at least opened up one stop if not two...

I think what´s really important is to make sure you reduce contrast as much as possible so you can move the histogram further to the right without any unnecessary clipping... you either use more fill light or you try to apply shading to parts of the image with ND grad filters...

it gets tricky with extremely bright highlights like reflections in a chrome car bumper for example... you have to check the RAW file, look at it in false color mode to see where exactly in the frame you´re highlights are and decide if they are really worth protecting... sometimes it´s better to let a highlight clip in exchange for some more tonal values in the shadows and midtones in order to get less noise...
 
Sometimes it´s better to let a highlight clip in exchange for some more tonal values in the shadows and midtones in order to get less noise...

Good advice. More tonal values can make a big difference for cameras that have very heavy compression as it prevents quantization noise and posterization.

I think the most significant benefit of increasing exposure, though, especially in cameras like the RED ONE where the compression is relatively light, is the increase in the number of photons (signal) and reduction of photon shot noise (in absolute numbers) and read noise.
 
Good advice. More tonal values can make a big difference for cameras that have very heavy compression as it prevents quantization noise and posterization.

I think the most significant benefit of increasing exposure, though, especially in cameras like the RED ONE where the compression is relatively light, is the increase in the number of photons (signal) and reduction of photon shot noise (in absolute numbers) and read noise.

Now Daniel, all this photon stuff, it confuses me, remember my request at the beginning of this thread was keep it simple, if you can put that stuff in layman's terms, well i would love that :w00t:
 
The problem with monitoring in linear light is it's bloody hard. I think original Frankie was like that, and quite frankly, I don't know Jarred managed to get the images he did off it. For starters, it's nigh-on impossible to see what's going on in the shadows, and a properly exposed image looks under-exposed, and hence you tend to over-expose if monitoring in linear light.

I find the best guide to be the RAW levels bar-graph. That's what works for me.

Graeme

So what are the advantages of linear light monitoring? What is it really useful for. How does that pertain to the RAW image?
 
Now Daniel, all this photon stuff, it confuses me, remember my request at the beginning of this thread was keep it simple, if you can put that stuff in layman's terms, well i would love that :w00t:

Sure! The simple version is: the more light, the better. Until, of course, you clip the highlights. That's too much light. The balance of these two is the essence of ETTR (Expose To The Right).

In situations with extreme compression (e.g. 8-bit MPEG-2), the number of tonal values are far too few to describe the light, which can problems such as posterization (the opposite of "smooth tonal gradations"), especially if the exposure is too small to utilize all the available tonal values. In this situation, ETTR helps two things: it adds more light and it increases the use of tonal values.

However, in cameras that already have enough precision to fully describe the light, such as the RED ONE, increasing exposure doesn't really help the tonal values, because there's already plenty of them to go around, even in extreme underexposure. (If there wasn't, it would indicate a weakness in the REDCODE compression, and I haven't heard anyone mention it.) Instead, increasing exposure increases the amount of light, or signal, relative to noise (i.e. signal to noise ratio), which results in an improved image.
 
So what are the advantages of linear light monitoring? What is it really useful for. How does that pertain to the RAW image?

I guess judging from our discussion with Graeme it makes no sense to monitor in linear light at all. It´s only good for checking your files in RedAlert in order to see how much of the 4096 tonals values you actually used.

As we were able to see in linear light mode, one or two stops of underexposure can cost us a lot of tonal values and is likely to ad more noise to our image if we try to bring the exposure back up in post. We should try to "expose to the right" (pushing the histogram to the right by opening up the aperture of the lens) as much as possible and always keep in mind, that in a 12 bit linear file, the brightest stop actually uses half of the 4096 tonal values...

if you have that extremely bright reflection on a chrome bumber in frame, just let it clip... don´t give away 2048 tonal values just to save that little highlight from clipping... it´s likely to be not more than a tiny speck in your image... and to protect it you´re likely to lose a huge amount of tonal values that you need for the rest of your image, the mids and the shadows...

Always check the RAW view and make an informed decision about what hightlights are worth being protected and how much noise your willing to accept in order to do so.

That´s my suggestion. Comments are welcome...

Cheers,

Kay
 
Take a look at these images... the contrast is very high. I had no lights available. The only thing I was able to do in order to open up the aperture a little and lift the shadows of the interior without loosing too much of the sky was to use a ND.6 grad filter and shade the sky on the right side of the frame. You can see the gradation if you look at the dashboard. Because of that I lost two stops exposure in the mirror which was kind of sad because you can´t see much of the reflection of the driver anymore... and some of the highlights around her head are still clipping... it clearly was a compromise, but one I was willing to make because I wanted to protect some of the highlights in order to not have a completely blown out sky and still get as much tonal values in the shadows as possible.

If you look at the somewhat darker linear light representation of the image to the right, all 4096 tonal values were used as good as possible...

If anyone has a better idea how to treat a situation like this please let me know...
 
Yep, thanks everybody who is contributing to this it really is very helpful.
 
The same strategy worked even better in this shot. I used an ND grad .6 filter to shade the left part of the frame (red line apx. representing the soft edge of the filter), helping me to reduce the brightness of the sky by two stops. Although there still is some clipping in the highlights, I was able to open up the aperture and gain a lot of extra tonal values in the mid tones and shadows without loosing all the details in the sky. Quite a good compromise.

The focus puller I worked with on this project has shot a feature with dop Slawomir Idziak (who shot black hawk down). He told me that Slawomir has a huge collection of grad filters, some of them hand painted by a polish artist. For example a grad filter for a two shot of two people sitting at a table, with only the negative of their silhouettes shaded in a certain color and densitiy. Building a collection like that could be really useful in certain situations to help control contrast.
 
That's the beauty of the FALSE COLOR option in camera - you can see exactly where the highlights are clipping and can base your final exposure on that in real time. In combination with RAW view and the RAW exposure bar graph you have an extremely powerful and accurate way of getting the maximum image tonality from your 12bit sensor.

It's also why I recommend to all our clients that they put the light meter away and trust the camera, because it is giving you the best possible information as it is directly affecting the actual sensor.

Cheers for an excellent thread!

Paul
 
That's the beauty of the FALSE COLOR option in camera - you can see exactly where the highlights are clipping and can base your final exposure on that in real time. In combination with RAW view and the RAW exposure bar graph you have an extremely powerful and accurate way of getting the maximum image tonality from your 12bit sensor.

It's also why I recommend to all our clients that they put the light meter away and trust the camera, because it is giving you the best possible information as it is directly affecting the actual sensor.

Cheers for an excellent thread!

Paul

I still think the light meter is of great value when lighting a set, helping you to make quick judgment on contrast ratios. I also use my color meter again, that I didn´t use in a long time, to get my color temperatures closer to 5000K, but for the final exposure in the camera I agree with you, my best friends are RAW view, false color mode and histogram (I guess with RAW exposure bar graph you mean the histogram in RAW view mode?)
 
... Set one of your user buttons to RAW view so you can quickly check how much more leeway you still have in the RAW file before you clip. This can be very useful in shots with extreme contrast.

Kay

Hya Kay,

thank you for the great hints to how you are approaching the exposure thing.
The raw view thing works great for me! But whilst testing for myself i did set one of the user buttons to raw view and now after switching around sometimes i do get confused of which view is the raw view.....maybe i am missing something. Is there a chance to set the overlays to depict the colorspace you are in?

And if I set the view generally on RAWview in the menue and than do push my configured user button - the ouput changes anyway.....but to what view?

Can anybody explain to me how I can figure out what view i am in?

Thank you,

All the best
 
Hya Kay,

thank you for the great hints to how you are approaching the exposure thing.
The raw view thing works great for me! But whilst testing for myself i did set one of the user buttons to raw view and now after switching around sometimes i do get confused of which view is the raw view.....maybe i am missing something. Is there a chance to set the overlays to depict the colorspace you are in?

And if I set the view generally on RAWview in the menue and than do push my configured user button - the ouput changes anyway.....but to what view?

Can anybody explain to me how I can figure out what view i am in?

Thank you,

All the best

Hi Markus,

I had the same problem in the beginning and I wish there could be a foolproof way like an overlay that tells you what viewspace you are using.

During production it can happen that you accidently press the button and make wrong judgements... but with a bit of experience you can tell pretty well if you are in redspace or RAW view... redspace applies a gamma correction that remaps the tonal values further to the right in the histogram... if you look at your image the bright parts will get a boost when you switch from RAW to redspace and the histogram will move to the right aswell...

If I get confused I look at a highlight that is close to clipping and toggle the view space and when the clipping becomes bigger I know I´m in redspace...
that´s my work around... does somebody have a better method? Please let us know...
 
Hi Markus,

I had the same problem in the beginning and I wish there could be a foolproof way like an overlay that tells you what viewspace you are using.

During production it can happen that you accidently press the button and make wrong judgements... but with a bit of experience you can tell pretty well if you are in redspace or RAW view... redspace applies a gamma correction that remaps the tonal values further to the right in the histogram... if you look at your image the bright parts will get a boost when you switch from RAW to redspace and the histogram will move to the right aswell...

If I get confused I look at a highlight that is close to clipping and toggle the view space and when the clipping becomes bigger I know I´m in redspace...
that´s my work around... does somebody have a better method? Please let us know...

Hya Kay,

Thank you for the Quick reply. I will praktice an Bit and hope i do not get confused on set. I tought already that the histogram shifting to the right must be redspace. Thank you for Clearing that point. And yes - question to everyone - somebody any better ideas?

All the Best
 
Take a look at these images... the contrast is very high. I had no lights available. The only thing I was able to do in order to open up the aperture a little and lift the shadows of the interior without loosing too much of the sky was to use a ND.6 grad filter and shade the sky on the right side of the frame. You can see the gradation if you look at the dashboard. Because of that I lost two stops exposure in the mirror which was kind of sad because you can´t see much of the reflection of the driver anymore... and some of the highlights around her head are still clipping... it clearly was a compromise, but one I was willing to make because I wanted to protect some of the highlights in order to not have a completely blown out sky and still get as much tonal values in the shadows as possible.

If you look at the somewhat darker linear light representation of the image to the right, all 4096 tonal values were used as good as possible...

If anyone has a better idea how to treat a situation like this please let me know...

Great thread guys, I have learned a lot from this.

Kay, I carry sheets of foamcore for situations like this. Cut it up and us it for a reflector to fill shadows a little bit.
 
Great thread guys, I have learned a lot from this.

Kay, I carry sheets of foamcore for situations like this. Cut it up and us it for a reflector to fill shadows a little bit.

I wonder if these new small battery powered LED panels would help in a situation like this .... Flolight ... it seems like Coollight is also making on of those...
 
Back
Top