Why ASA? Why Meters?

[Nathan Troutman]

I'm trying to understand better why were using ASA ratings on RED - besides giving the Film segment something to equate with? It just seems to be adding a layer of technical superfluousness to what is natively a digital camera. The sensor's native sensitivity is 0 DB of Gain. From there you can either add or subtract Gain, right? Maybe you can shoot at 1000 ASA with RED. I know for myself I'd rather know that I'm shooting at +12 DB of Gain than an ASA rating.

Second question. Why use a light meter with a digital camera that is essentially a large light meter? Am I missing something here? Can't you just look at the histogram or waveform (any chance of a waveform view besides the histogram in Red?) and see exactly what the exposure of your scene looks like?

[M Most]

Why use a light meter with a digital camera that is essentially a large light meter? Am I missing something here? Can't you just look at the histogram or waveform (any chance of a waveform view besides the histogram in Red?) and see exactly what the exposure of your scene looks like?

When you're getting ready to shoot, yes. When you're lighting, or when you're designing a lighting plan for a location, you really don't want to lug around the camera to figure out what units to use and/or what your lighting ratios are. A light meter and an intended shooting stop give you a very good starting point whether you happen to have a camera up and available or not. Not to mention that unless you have variacs on every lighting unit, or run everything through a lighting board, it takes time to adjust light levels. If you've never been a DP, and/or you've never shot in a professional situation (film or video, doesn't matter) it's likely you wouldn't understand the usefulness of these tools. But if you have, they are tools that you consider indispensable.

[Greg M]

Nathan,
For the most part the Red camera uses cine lenses that do not have an auto iris function. The camera is not a "large" light meter, it is a capturing device. As with any camera the measurement of light is critical to capturing great photography. A light meter is not required, but if you want be able to capture nice images a basic understanding of ratios and metering will be very helpful to your craft.
The last thing you want to do is use your gain to adjust the light levels in your camera. I would suggest you pick up a few books on photography to better understand the basics of light and metering, it will have a big impact on the quality of your work.

[Jeremy Hughes]

Well if you just listen to the Gain, you're going to need to know the native ASA at least in order to understand what +12dB is, right? You must keep in mind that gain is relative to the sensor and ASA is not. Also, if you're shooting 2K or 4K, you're shooting RAW. And you will have the ability to change the gain in RED ALERT! or REDCINE.

Also, it's dB for decibel (0.1) and DB for a decabel (10) if there was such a thing. I'm not being cruel with this correction am I? I hope not. :)

[Graeme Nattress]

Nathan's question is a very valid one. MMost answers it well, that a light meter is more portable than a camera. The other reasons are:

Some people like to work with light meters and are happy doing that,
Some like to work with scopes, histograms and the like and are happy with that,

And some will use all the tools available to them to make a nice picture, and be happy with that ecelectic approach.

Gain in camera, BTW, is metadata only. It doesn't alter the captured data at all.

Graeme

[Anders Holck]

Just to add to the answers here:

Gain is not absolute. It's relative to the base sensitivity. Asa/iso is an absolute rating and therefore very useful especially when lighting.

When working in a team and your gaffer knows the Asa/iso rating, you can ask him to light a scene to the F-stop you want to shoot at. It's pretty neat when you are tied up at the camera doing another image.

Edit: Ok, I'm now officially getting too slow to post. Nothing new I guess....

[Nathan Troutman]

Also, it's dB for decibel (0.1) and DB for a decabel (10) if there was such a thing. I'm not being cruel with this correction am I? I hope not. :)

No you're not - I just have an over-capitalization problem:help:

Great responses from everyone. What I'm not understanding is how changing RED (there I go again) from 320 ASA it's native sensitivity to say 1000 ASA isn't just a change in gain? But Anders made a great point I completely forgot about in terms of being able to have a rated number and f-stop that your gaffers can then use to light the scene (using a meter) until a camera is up and ready. So we know the native sensitivity is 320 ASA - isn't it still easier to just have in camera a simple 0 through +18 dB number scale to show what the camera is doing rather than using ASA numbers? It's like having an HVX and using 200 ASA as it's native sensitivity and then instead of +3, +6, +9, +12 of gain you'd have 300 ASA, 400 ASA, etc. This to me doesn't communicate effectively what's actually happening.

[Chris Kenny]

What I'm not understanding is how changing RED (there I go again) from 320 ASA it's native sensitivity to say 1000 ASA isn't just a change in gain?

It is just a change in gain. Digital gain, specifically, since Red doesn't (currently?) have adjustable analogue gain.

[Chris Gearhart]

Nathan, you are exactly right. I think it is precisely because the camera is using RAW that perhaps we are getting loose with terminology. It is what amounts to changing the LUTs to make the image whatever we want (non-destructively, of course), and so we call it what we want.

There is some analogical sensibility to doing so, however, since giving an ASA rating for our meta-image tells more to a DP than, say, "+12db gain applied." It's like saying "the sun rose today". Technically, the earth rotated today, but that's not what has meaning in my world.

[Craig Schober]

It's like having an HVX and using 200 ASA as it's native sensitivity and then instead of +3, +6, +9, +12 of gain you'd have 300 ASA, 400 ASA, etc. This to me doesn't communicate effectively what's actually happening.

exactly. that terminology would turn off a lot of video guys to the hvx. red is a digital cinema camera and no matter how new shooting raw and reviewing footage on a laptop minutes after a take is, you're still selling the camera to a lot of old school film guys. they would get scared off by all this gain and raw talk so red marketing equates the sensor's sensitivity to an iso rating they can relate to and lenses with t-stops that make them more comfortable. you don't need any of that stuff to make great looking images with red. you only need a basic knowledge of light ratios and a histogram in the viewfinder.

[Nathan Troutman]

you don't need any of that stuff to make great looking images with red. you only need a basic knowledge of light ratios and a histogram in the viewfinder.

Am I the only one that would also really like to have a waveform view? I like the histogram but to me a waveform is far more descriptive as to what is doing what exactly where - when you look at it.

[dchenevert]

It is just a change in gain. Digital gain, specifically, since Red doesn't (currently?) have adjustable analogue gain.

Is the RED analog gain adjustable in firmware?

If so, then: with what granularity could this gain be adjusted? i.e.,
per-pixel, per-row, per-frame?

I've wondered for a while about the possibility of trading resolution for
DR. Imposing an alternate-row or alternate-frame variation in gain
(i.e. +1 stop on the even rows, -1 stop on the odd rows) would provide
more info in the shadows and highlights, at the expense of some spatial
resolution.

[Joe Taylor]

Hope this is not off topic. IF Red's native ASA is 320, would that be its default ASA? If so, can the user work below 320 ASA to theoretically have an even cleaner (less noise) picture? For example, many digital cameras have a 0 gain for their default setting. The operator will also have a -3 gain setting as an alternative. Will RED do the same, ASA wise?

[spawnofbill@comcast.net]

Hope this is not off topic. IF Red's native ASA is 320, would that be its default ASA? If so, can the user work below 320 ASA to theoretically have an even cleaner (less noise) picture? For example, many digital cameras have a 0 gain for their default setting. The operator will also have a -3 gain setting as an alternative. Will RED do the same, ASA wise?

From what I understand, which is mainly based off of a knowledge of DSLR sensors, shooting below the native ASA is feasible, but somtimes can actually show a degradation of quality. I say sometimes because this is kind of a hot topic, some people claim to see it, others think they're crazy. But basically with it involves a lot of sensor magic which I don't fully understand.

[chuck colburn]

Plus it's awkward to tote the camera around with a big duffision dome on the front to use it as an incident meter. lol

[jbeale]

Hope this is not off topic. IF Red's native ASA is 320, would that be its default ASA? If so, can the user work below 320 ASA to theoretically have an even cleaner (less noise) picture? For example, many digital cameras have a 0 gain for their default setting. The operator will also have a -3 gain setting as an alternative. Will RED do the same, ASA wise?

This has come up a number of times. The answer, at least for the present, is no. The analog gain (setting native ASA/ISO equivalent sensitivity) is fixed and it can be neither increased nor decreased.

http://www.reduser.net/forum/showthread.php?t=4981
http://www.reduser.net/forum/showthread.php?t=4904

[Harva Raj]

Second question. Why use a light meter with a digital camera that is essentially a large light meter?


When you're lighting, or when you're designing a lighting plan for a location, you really don't want to lug around the camera to figure out what units to use and/or what your lighting ratios are.

i heard that the DP of Transformers was seen using a DSLR as light meter in the sets.is it true?

[Bob Melville]

exactly. that terminology would turn off a lot of video guys to the hvx. red is a digital cinema camera and no matter how new shooting raw and reviewing footage on a laptop minutes after a take is, you're still selling the camera to a lot of old school film guys. they would get scared off by all this gain and raw talk so red marketing equates the sensor's sensitivity to an iso rating they can relate to and lenses with t-stops that make them more comfortable. you don't need any of that stuff to make great looking images with red. you only need a basic knowledge of light ratios and a histogram in the viewfinder.

I'm a little bit confused by the argument that somehow T stops are no longer relevant and that all you need to make great looking images with Red is a basic knowledge of light ratios and a histogram. Lighting ratios are only part of the overall image. Depth of field is crucial to creating the look. Red lenses are cine style lenses for a reason. They allow precise control over depth of field by allowing the user to determine parameters such as focal length and aperture. An additional factor is how far the camera is from the subject. ISO is part of the discussion because it is another variable in choosing the aperture and thereby ultimately deciding what is in focus and what is not in focus.

Bottom line is this. If you tell someone who is experienced the focal length of the lens and show them a waveform and a histogram they can't tell you whether both George, standing 20 feet from the camera and Will standing 30 feet from the camera are in focus. That person will be scrambling to catch a look at the production monitor. If you want to spend a lot of money do all your lighting by looking at monitors.

If you meter the scene and tell someone who is experienced the focal length and the T stop they will be able to tell you whether or not you can have both George and Will in focus without looking at monitors. Also, because they understand the variables involved they know what to do to change depth of field if the director wants it. Maybe they change lenses, move Will 3 feet forward, add or decrease light so they can change the aperture? Let's get even more precise, which eye is in focus in the closeup? It is also pretty easy to pop a spot meter into the shadows and have a precise understanding of how light and shadow is going to work in the shot.

The whole point of the RED camera and the lenses which go with it is to allow complete control over the image. There is a reason Red is not selling broadcast video style lenses. In fact, they are selling a set of film style primes. There is a reason Red lenses are not auto focus and exposure is set manually. It is to provide the digital cinematographer with control. That control, no matter how it is augmented digitally by further information provided by waveforms or histograms, results from the knowledgeable application of basic principles of photography.

Jim Jannard and the whole Red team have done something remarkable and come the spring I will have a Red camera in my hands. I will marry that camera with the absolute best lenses the wonderful Red team can produce, basically a new twist on old style cine lenses. When it comes down to it you can't beat great glass. And one of the coolest things I will experience is manually setting that aperture.

[Gopher77]

Am I the only one that would also really like to have a waveform view? I like the histogram but to me a waveform is far more descriptive as to what is doing what exactly where - when you look at it.

I'm with you, in fact I think it'd be cool if the little lcd in the back of the camera could display those. Then the AC or camera op could simply push a button and look at either one and make any adjustments necessary.

[Craig Schober]

I'm a little bit confused by the argument that somehow T stops are no longer relevant and that all you need to make great looking images with Red is a basic knowledge of light ratios and a histogram. Lighting ratios are only part of the overall image. Depth of field is crucial to creating the look. Red lenses are cine style lenses for a reason. They allow precise control over depth of field by allowing the user to determine parameters such as focal length and aperture. An additional factor is how far the camera is from the subject. ISO is part of the discussion because it is another variable in choosing the aperture and thereby ultimately deciding what is in focus and what is not in focus.

Bottom line is this. If you tell someone who is experienced the focal length of the lens and show them a waveform and a histogram they can't tell you whether both George, standing 20 feet from the camera and Will standing 30 feet from the camera are in focus. That person will be scrambling to catch a look at the production monitor. If you want to spend a lot of money do all your lighting by looking at monitors.

If you meter the scene and tell someone who is experienced the focal length and the T stop they will be able to tell you whether or not you can have both George and Will in focus without looking at monitors. Also, because they understand the variables involved they know what to do to change depth of field if the director wants it. Maybe they change lenses, move Will 3 feet forward, add or decrease light so they can change the aperture? Let's get even more precise, which eye is in focus in the closeup? It is also pretty easy to pop a spot meter into the shadows and have a precise understanding of how light and shadow is going to work in the shot.

The whole point of the RED camera and the lenses which go with it is to allow complete control over the image. There is a reason Red is not selling broadcast video style lenses. In fact, they are selling a set of film style primes. There is a reason Red lenses are not auto focus and exposure is set manually. It is to provide the digital cinematographer with control. That control, no matter how it is augmented digitally by further information provided by waveforms or histograms, results from the knowledgeable application of basic principles of photography.

Jim Jannard and the whole Red team have done something remarkable and come the spring I will have a Red camera in my hands. I will marry that camera with the absolute best lenses the wonderful Red team can produce, basically a new twist on old style cine lenses. When it comes down to it you can't beat great glass. And one of the coolest things I will experience is manually setting that aperture.

i was only referring to luminance because of the nature of the thread and light meters. obviously, there are dozens of factors that go into creating a pretty picture. some are made through use of technical charts and some are made by just "eyeballing" the frame. i mentioned f-stops because they don't mean much to most videographers but they do to filmmakers. focus, however, means a lot to both groups. red is pulling customers from both of these pools so people will use whatever works for them. but in the end, i only need to see what's in viewfinder or lcd to make a great looking image. it's worked for 100 years of filmmaking and that won't change anytime soon.

[Stephen Williams]

[QUOTE=learmonth;93958]
If you meter the scene and tell someone who is experienced the focal length and the T stop they will be able to tell you whether or not you can have both George and Will in focus without looking at monitors. /QUOTE]

Hi,

Thats not correct, DOF is calculated using 'F' Stops, T stops do not have a direct relationship with 'F' stops so don't help when trying to calculate DOF.

Stephen

[Bob Melville]

[QUOTE=learmonth;93958]
If you meter the scene and tell someone who is experienced the focal length and the T stop they will be able to tell you whether or not you can have both George and Will in focus without looking at monitors. /QUOTE]

Hi,

Thats not correct, DOF is calculated using 'F' Stops, T stops do not have a direct relationship with 'F' stops so don't help when trying to calculate DOF.

Stephen

Hi Stephen,

Here is the quote from the ASC Manual, page 59. "The "T" stop number is defined as being the true "f" stop number of a lens if it is completely free from all reflection and absorption losses." Continuing "The T-stop can be considered the "effective" f-stop."

You are right, there can be some variation between the f stop and the T stop. For example, zoom lenses with a lot of internal elements can affect the transmission of light a great deal which increases the disparity between the f stop (a mathematical calculation) and the T stop (a measurement of the actual transmission of light through the lens).

This is all an elaborate way of saying that I should have just used the term "f" stop in my previous argument and not brought in T stop. So consider this a "find" and "replace" comment.

Thanks

[Stephen Williams]

[QUOTE=Stephen Williams;93968]

Hi Stephen,

Here is the quote from the ASC Manual, page 59. "The "T" stop number is defined as being the true "f" stop number of a lens if it is completely free from all reflection and absorption losses." Continuing "The T-stop can be considered the "effective" f-stop."



Hi,

A T stop is about light transmission, and is an effective F stop for LIGHT TRANSMISSION.

DOF is calculated from the F stop, the calculations use a measurement from the front entrance pupil of the lens & not the film plane. DOF tables provided by lens manufactures for a specific lens will be corrected to measure from the film plane & in T stops. This is one reason why tables from different manufactures vary & Zoom lenses appear at first glance to have different DOF characteristics. However with a Zoom the entrance pupil may well be 12 inches or more in front of the film plane, back off the camera those 12 inches and the DOF & Image size on the film plane will then be identical.

Stephen

Stephen

[Anders Holck]

Currently the ISO adjustment in camera is metadata only. I believe that when the .R3d file is loaded inside RedAlert the algorithm applies a "digital gain" adjustment to the image. So shooting ISO 640, is really ISO 320 +6db post gain.
If you shoot ISO 160 it would apply -6db of gain.

This will of course also attenuate your clipped specular highlights as well (If there are any), but highlight recovery might help this as there is at least 1 stop (6 db) in the RAW file.

Most Canon DSLR's have a base sensitivity of ISO 100 (actually ISO 125), and the ISO 50 mode (ISO 64) has one stop less lattitude in the highlights, which implies that a negative post gain function could do the same job. So that feature is probably there only for making low noise JPG's in camera.

http://a.img-dpreview.com/reviews/CanonEOS5D/Samples/DynRange/iso.jpg

Finding the base sensitivity would need you to find unity gain by setting the AD gain of the chip to hit maximum digital intensity at full well/photosite saturation. setting it lower would basically just limit your dynamic range, as the sensor would clip before you get full digital intensity. after that you rate the camera based where you want to utilize the dynamic range.

(Graeme? :bye2: )

[Rocco Schult]

Am I the only one that would also really like to have a waveform view?..

No, you're not. Its a digital camera and it makes sense. But the histogram is more widely used, so maybe we remain the few who really want.


I'm with you, in fact I think it'd be cool if the little lcd in the back of the camera could display those...

Though the idea is lovely, the LCD is lacking resolution at least and contrast for sure. Alternatively switchable to the LCD would way cool.
'Picture/Histogram/Waveform' or the like. Don't know about all the other options the LCD has and it therefeore might total BS to integrate another position to the menu, but I like the idea of having a waveform handy (and built-in).