If a light meter is good enough for Deakins...

Dennis
No need to convince me the 5/55 is native 640 - it says 640 on the tin if you know where to look - the lowest cineEi setting - just like Red is 320 - also written available in the settings. I just rate 1/2 or one stop faster with these cams to give me a little headroom.
As for metering slog the onboard tools are basically not acceptable. I don't really consider the 5/55 as being fit for purpose. Sony should be embarrassed and certainly lost one F5 sale (me), not on image quality but their total failure to make a professional shooting tool.

David.
The fs100 histo is far too small and not rgb so a pretty poor tool - you get better histos in $200 Dslrs, I believe Sony have a huge fail there and should have a decent WFM or histo under the menu that bring up the nice chunky sound bars - which are excellent. It is almost acceptable on the FS100 price point but at the F5 level, just unacceptable.
A simple way to meter using a histo is point the camera at a grey card and you will see a lump appear in the graph - change something until that lump sits at the desired point generally the middle. In time the lumps will make sense to you. I prefer wfm but using a little Dslr as a meter you are only concerned with a single lump so it works without interpretation. I mainly like pointing an rgb histo at a grey card - or directly at a source (stop down) and you can see the colour spikes match or not - really handy with flouros in practical locations.
 
Last edited:
morgan_moore said:
David. I thought this was a pretty light (hearted) thread.

Being serious, of course I can see the value of a light meter for pre lighting.

Genuinely over my 20 years with images I have used them or not. Shooting strobe flash and stills on transparency a meter was critical.

Once I moved to raw stills the thing got dropped, I then moved to crappy codec motion where exposure is more critical the main problems I had with choosing a camera setting is choosing the colour settings under mixed light

As an aid to that I certainly found the rgb histo a fantastic tool on the 5d and seriously missed it on the ex1 and fs100 - my displeasure with the colours on those cemeras has now 'forced' me to move to the f3 to get a more colour flexible codec - partly to make up for the info I don't have on set.

If I was going to invest in a light analysis tool above $1000 I would certainly want it to tell me something of both temp and tint..
Click to expand...

Josh Bass said:
How would you use a vectorscope on set except as an elaborate white balance tool? It could tell you if your red was true red, your skin tones bang on etc. Is that it? Can cameras record illegal colors (i.e. too saturated?) What would you do if saw the red light in a scene was illegal? Mess with cam matrices to bring it down?
Click to expand...

I think that a vectorscope on set would be mostly helpful in quantifying the color of what you are shooting. In long form projects, especially fictional narrative where color is one of a cinematographer's most powerful creative tools, an on set vectorscope would be very handy in keeping your color consistent. If the vectorscope shows that there are lights in your shot that ain't legal, well... at least you know. With that knowledge you can decide to deal with it, and my first suggestion would be to turn the mofo off, if possible, or not. This would, IMHO, cut down on those confidence bashing, "Ahhh sh*t" moments when dealing with your footage in post.

Now a vectorscope is really on my, "It would be cool if I had one" list but pretty much near the bottom. I do not anticipate being able to work with one regularly on set anytime soon. However, I do see value in having one.
 
I don't know the cost of such a device. I know color meters are costly. I know stills cams with rgb histos are about $200 or free if you have one. I have an old Nikon d80 which is worth nothing. I have started using it on set. If it gets dropped stolen or lost - no bother!
 
hscully said:
Thanks for the reply. Do you have any references you can share on the steps necessary to determine my camera's ISO? I'll look around. Here's something I don't get: how is it that a camera has a single ISO? The ISO is variable. I've hear that the Alexa is rated at 800 ISO but you can set the ISO to ... well whatever so is it actually an offset that you want to get?

It was actually 2 stops from 60 to 100 IRE, f2 to 4. I wrote the wrong value in my post.
Click to expand...

Watch this video by Andy from Abelcine. http://blog.abelcine.com/2011/03/11/abelcine-tests-the-sony-f3-iso-rating/

A camera doesn't have a single ISO but when you rate a camera's ISO you should set the camera to the lowest ISO possible. You are basically just establishing your "base" ISO and later can work around that.

Remember, the main reason you are doing this is so that you can align your light meter to exactly how much light would give you 50% middle grey on a waveform monitor.
 
I wrote a very lengthy procedure for rating a camera, etc.. but honestly it's not entirely worth posting as they say "a picture is worth a thousand words" so it is best to watch Andy's video and see how he does it.

Remember to pay very special attention to your camera's settings of gain, shutter speed, frame rate and not to have any ND filters engaged if they are built in. Some cameras will do -3db gain so I recommend sticking with 0db gain.

Also a couple notes worth noting: most "cine" meters are defaulted to 180 degree shutter so you dial in the frame rate and the shutter speed is automatically DOUBLe your frame rate speed. 24fps with 180 degree shutter is 1/48th. If your meter doesn't let you enter frame rates you will have to put in the shutter speed and again make sure the camera is setup the same way. Most photography meters won't have 1/48th shutter speed but you can use 1/50th with marginal difference.
 
morgan_moore said:
A simple way to meter using a histo is point the camera at a grey card and you will see a lump appear in the graph - change something until that lump sits at the desired point generally the middle.
Click to expand...

I usually refer to the 'lumps' as 'haystacks'... and for a 18% grey card, where most of the field of view is taken up with the card, the 'haystack' should have it's peak in the middle of the graph display.

One can also look at the histogram graph, and for various f-stops, note the 'profile'. As one goes from full open, to closed down, the histogram will change, and if one wants to have 'all the scene' range, there should be no pile up on either side of the graph.

Scene contrast can be estimated as well... all the scene values piled up in the middle, with no apparent data on either end... 'low contrast'... etc.

But I've only used histogram for when I'm shooting events and have no time to meter, and want to know I have no blank whites, or blank shadows... unless of course the situation is just 'that way'...
 
One consideration with metering gray cards, no matter the method, is metering depends on angling the card even with matte cards. This can be a hassle. Incident readings are generally significantly faster (for me, at least). If I am judging exposure off an on-camera tool, I go WYSIWYG - I'd mostly look at some scene reference (skin usually) and put it where I want it.
 
morgan_moore said:
in monitor design it is super important that a histo is reading the input signal from the camera before any lut is applied or at least that is an option.

Histos are great but Reds implementation - which is I guess where most cinamatographers met them is a total mess up as it reads after the "iso lut" has been applied and leads one not to understand the core data one is recording.
Click to expand...
Don't get too mad at RED, they do it the same way everyone else does. That includes Sony, ARRI, and every monitor maker out there.

Oh, except one. Us.

As soon as we implement our programable LUT feature, we will also give the ability to set the various image analysis tools in the Odyssey7 & Odyssey7Q to measure the LUTted image or the original source (there will be a clear indicator so the user is not tricked into improper readings).

Think of something as simple as shooting in Log but viewing in Rec709. I might want the picture in Rec709 to set my lighting but I don't need the waveform to tell me a window is blown out in REC709 -- I can see the damn thing clipping. What I want from the meter is to know if I'm retaining the highlight in the LOG signal. So you can set your waveform to Log separately from the image which might be set to Rec709.

No one else can do this. Hell, most monitors get screwed up if you turn on frame lines or peaking, as those are often happening upstream from the waveform or histogram, so they become false measurements in the meters. Crazy. The Odyssey is able to do this because we have dual, separate signal paths for the image and the analysis tools. This way an adjustment to one does not alter the other.
 
Mitch Gross said:
As soon as we .... This way an adjustment to one does not alter the other.
Click to expand...

This is all very good.

If you have a 'focus check/magnify' mode showing just the histo/wfm of that would help for 'spot metering'

I know its all more development time and getting the GUI so you know your mode is important - but most of all it is great to hear a company think (if you can hear thought)

--------

Yes I think the Red (and BMC) implementation is ultra dumb. People sticking the camera on 800 or 1600 getting a full histo and in post they realise that they have shot only in the 'dirty thirty' zone of the file and cos they were under tungsten the blue was not even hitting thirty. Caused huge problems for camera/file understanding - probably kept film alive for another three years. Bloom did huge early damage to the BMC rep by shooting a bunch of shots no higher than 20 on the (post) histo and rating the camera as noisy - slow maybe but not noisy.

S
 
Last edited:
morgan_moore said:
Unless you are shooting Slog or ETTR raw, it could be from 38% t0 70%

S
Click to expand...

Yes, I was going to mention that the 'gamma' of the camera (or setting if available, or selection of 'color space'... etc.) will impact where the '18% grey card' falls, as well as what to expect from reflectivities above and below 18%...

Since I don't have 'fancy' cameras, these sorts of features are absent, or not clearly stated in the various 'style' settings.

But that's not unlike Film film days, where one had to take the gamma of the particular negative material in to consideration of 'am I going to blow out the highlights or drop out the shadows'... and either mitigate with ND filtering, screening, lighting, or in post, negative processing or print manipulation...
 
Last edited:
I set up a gray card and had it sit exactly on 50 IRE. I have a slate with color bars and made sure that white was just kissing 100 IRE. I tested that with the WFM on the AF100 and the TVLogic monitor and they both agree. I tested with the Voigtlander 25 mm and a Zeiss ZF 50 mm.

At this given exposure the meter suggests 2.8. On the lenses it actually exposed properly at closer to 3.5 on the AF100. So the AF100 is a little more than 1/2 a stop "hot" at ISO 400 shutter speed 1/60. I think that's livable and it is consistent. The GH2 was almost dead on at 400 ISO 1/125 shutter speed but with average exposure. I don't know what the histogram is telling me. I have to work on that some.

That's what I have so far. All these are done with incident readings. I'm not going to expose with the meter. I will use the WFM finally but I want to learn to light with the meter. Next stop, contrast ratios, reflected readings.
 
morgan_moore said:
If you have a 'focus check/magnify' mode showing just the histo/wfm of that would help for 'spot metering'

I know its all more development time and getting the GUI so you know your mode is important - but most of all it is great to hear a company think (if you can hear thought)
Click to expand...
The waveform & histogram can work concurrent to the focus assist (peaking) rught now, but not with the pizel zoom feature. A bit too much horsepower on that one. But we have a separate spotmeter function for this in the works. Think of a spot meter type tool with information for focus, exposure, color temperature and other information. It'll take some time to make it into the unit, but I think it's pretty cool.
 
I carefully rated my F55 according to Dennis' steps, Alister's S-Log2 exposure points, gray card 32% white card 59% (F5/F55 specific), rated the incident mode as well.

Then I set out on a corporate industrial shoot on the factory floor with impossible contrast ratios, no key or fills, and a need to not interfere with the production activities. I went about logging the light levels on paper for each shot beforehand with only the light meter and gray/white card.

Lastly, I made the shots quickly and entirely according to the notes taken earlier, using only a monopod.

In post, when I unwound the S-Log2 curve with an ACES input device transformation, I was shocked how utterly perfect the exposure came out on each of the shots, and the consistency. I was basically left with little to nothing to grade. Very elated with this process.
 
So what would advantages/disadvantages be to working with a meter measuring in footcandles, lux, etc. vs fstops? If you didn't have to relay the info to anyone else, that is. Do most meters do all those?
 
Josh Bass said:
So what would advantages/disadvantages be to working with a meter measuring in footcandles, lux, etc. vs fstops? If you didn't have to relay the info to anyone else, that is. Do most meters do all those?
Click to expand...

One of the 'stub your toe' things about moving from still photography to 'moving picture' photography, is the use of continuous lights... which are most often listed in footcandle/lux, if given at all... (Ok, some ebay sellers don't even bother with that, and just list the power consumption of the bulb...).

Quite a bit of the discussion of 'light' in cinematography uses fc/lux as well...

So, for me, I bought the Sekonic L-308DC just because it did have a footcandle read out, and now I use it to learn to 'see' light in terms of those units...

'footcandles/lux' are really only directly readable from an 'incident' situation. But not all incident meters have an 'fc/lux' readout option.

For reflective forms of meters, the actual units measured are 'foot-lamberts' (for the US... candela-meter^2 for the world...), so in order to get the foot-candles of 'incident light' one would have to know the reflectivity (and there is a discussion of what type of surface, and 'reflection profile'...) of the object, and calculate... so just simpler to get a meter which reads fc/lx directly...
 
As someone who is still in the early learning stages, I prefer to read f/ stops. Among other things, because it's a universal measure (fc is a US measure, lux is most-of-the-rest-of-the-world). I know it's easy to convert one to the other. Just personal preference, though.
edit: and possibly caused by laziness (i.e. avoiding having to learn what each lux measurement means)

(the one thing that pushes me to use something like fc/lx is that my meter gives me readings such as f/5.6-7, which means "seven tenths of a stop darker than f/5.6"; whose brilliant idea was it to mix f/ stop and decimal info in the same measure? still getting my head around this...)
 
Last edited:
My meter is pretty old timey but there is a conversion table on the back of it for fc. I am just getting my head around absolute values of light. Mitch mentions that he works in foot candle values but I think a lot of pro gaffers and DPs work in stops, usually T stops which from my research are measured for each cinema lens to maintain consistency across brands where f stops are "calculated" for each lens and can be somewhat variable. One thing is absolute, and that's the amount of light hitting that meter however you communicate it.
 
You must log in or register to reply here.
Back
Top