First Step: Color Meter? How can you not want one?

[Diego Treves]

I have little knowledge and basically no experience in light and scene lighting. But I want to learn, so a few question arose. First of which: how can you not want to use a color meter?

 

I know I could be wrong (and that’s the reason why I am asking, because I am sure I am missing something) but how can you light a scene exposed to natural sun light if you don’t have a color meter?

In the library I usually attend there is a very large room with ceiling fluos and very large windows (ten of them).

 

I was thinking: say you want to shoot your scene here, even if you can guess the right corrections to make in order to match fluorescent lights (which are old and quite greenish, but this is another issue) with your Hmis or Kinos AND with the windows (or whatever solution you may choose ok) HOW can you DISREGARD the shifting light coming from the windows at different hours in the day ?

How can you simply state that daylight is 5600K and be satisfied with that?

 

What happens if you shoot the same scene at 12.00 am before launch break and then again at 15:00? I guess you’ll have two different color temperatures, so you need to check that and apply the necessary corrections I guess..

 

And how can you do that without an instrument of any sort? Can you just guess? Will you be satisfied with this approach, will you consider it professional?

Or can you say that a certain amount of experience on the field is all it takes to get it right?

 

And even if you go ahead and decide you can have some uncontrolled mix in the color temperature, how can you know how much is enough?

 

A, say, 100k shift in the 3200k range of the nm scale will be more or less noticeable than a 100k shift in the 5500k range?? And why is that? How can you monitor that without a proper instrument?

 

So again: I am conflicted because at this stage I am I consider a color meter as a crucial tool, basically because I don’t like that kind of street-wise attitude which disrespects deep and technical knowledge. I want to build a solid expertise and I believe that building your own vocabulary in terms of typical situations and settings (and what color temperature natural light usually provides in given situations) should be a right first step to take.

 

But I am conflicted because color meters are really very very expensive. So my point is: either everyone approaching scene lighting buys a color meter (and I don’t think this is the case), or they mostly disregard it. But then once again, how can they be sure of what they do on the set? How can they meet precise directors’ guidelines and requests in terms of look and color if they can’t exactly predict and govern the color of their lights (I am always talking about situations where you have to deal with natural light mixed with your own lights of course). How can they be accurate?

 

I know light “color” is just a part of the whole “dealing-with-light” scenario, but it seems unforgivable to me that you are a dp and you cant have full control of your tools (light being one) on your scene..

 

So at this point I’d appreciate your advice on this.

Thanks for reading and sorry for bad english

[David Mullen ASC]

A color temperature meter is very important, which is why my gaffer always carries one... so I don't have to! :)

 

Seriously though, there are some color issues that can be addressed by eye and/or how it looks on a monitor and it isn't always necessary to be precise in terms of matching, you can some leeway depending on the shot and how much blending of different lights is going on.

 

And overall color tints can be adjusted in post as well.

[Satsuki Murashige]

Diego don't apologize, your English is exceptional!

 

I think color meters are important tools. But personally, I just can't justify owning one for $1500 right now! This is partly because in the digital age, a calibrated monitor is already an excellent color meter and usually there is one already on set. I would invest in one when you have the means to do it, but it's not critical in the beginning.

 

In the example you mentioned, the one light source that you have no control over is the sun itself. So I would expose for the sun-lit background consistently with a waveform and adjust the lights around it to match. If the sun has moved from a direct overhead position to side-lit noon time, then matching the direction, intensity, and quality (soft/hard) of the light is just as important as the color. So if you work on getting those correct, you will 90% there already. When you are losing the light and scrambling to grab as many setups as possible, no one is going to wait for you to pull out your color meter and gel the lights accordingly! :)

[David Mullen ASC]

Some of this stuff can be done by eye and monitor - if you have one part of the frame lit by real sunlight and a closer subject lit by an HMI meant to simulate the sun, you can add some degree of CTO or CTS to get close to the color of the sun. After all the correction gels don't come any finer-tuned than in 1/8 increments anyway.

 

Besides, one reason for lighting a room with HMI's is to bring consistency -- the more that the HMI's are doing the work, the less critical it is to match whatever is going on outside hour to hour until it gets very different, like at sunset. So you might have a big HMI "sun" with 1/4 CTO on it all day long coming through the windows and fill with ungelled 5600K Kinos or LED's all day long -- it won't matter if the sun outside is drifting by a few hundred kelvin from mid morning to mid afternoon, or even if the sun goes away due to overcast.

 

Where color temp meters are more useful is in dealing with partial spectrum sources that have a green spike, which can be hard to see by eye but will show up on the monitor. You may find that it is easier to add, let's say, 1/4 Plus Green to your windows to "green up" the daylight to match the color of the overhead lights in the location, if those can't be swapped or turned off. You may find that an LED light needs 1/8 Minus Green on it to match an HMI you are using, or a daylight Kino.

 

I remember an old interview with Gordon Willis back in the 1970's where he said he didn't use a color temp meter much because most color mismatches he could see by eye, and if he couldn't, then they didn't matter. But of course this was back in the day before LED's and compact flos, etc. And he had well-trained eyes -- he could see the difference between 1/4 CTO and two layers of 1/8 CTO put together...

[Diego Treves]

Thanks for your replies, I would like to answer in better detail but I am on mobile right now.

 

So I got it, it's not that crucial if you have a deep expertise and a wide set of tools like gels,wave monitors and the like.

 

The problem being that filter gel are appr. 80$ a roll (!) So I guess a beginner like me cant rely on a wide array to choose from.

 

in the end I should probably buy one of those "master location pakcs" with different colors in smaller cuts (like 50x60) and start practicing with that, making this one my attempt at building your personal vocabulary.

 

let me ask you something: I understand that the strenght and the quality of your diffusion comes from the size of the light source: so if you want to have an even diffed light you cant certainly apply a small frame of gel to your lights; but on the other hand such a rule doesn't apply to color correction/conversion gels, meaning that you can surely apply a small frame of gel to your light and get the same result if you had applied a large one.

 

So .. it means I can buy Master Location Pack and use the small gels for color correction, while I need a larger area for diffusion, am I right?

 

I need to understand this to get a figure of how much I need in terms of money to start practicing, keeping in mind that learning to light isnt all and that I'll still need to learn how to move camera accurately and confidently, so probably also other basic staff (like a fluid head and a small camera) should be included into the list .. (and having really no cash) ..

[Diego Treves]

Diego a calibrated monitor is already an excellent color meter and usually there is one already on set. I would invest in one when you have the means to do it, but it's not critical in the beginning.

 

In the example you mentioned, the one light source that you have no control over is the sun itself. So I would expose for the sun-lit background consistently with a waveform and adjust the lights around it to match. If the sun has moved from a direct overhead position to side-lit noon time, then matching the direction, intensity, and quality (soft/hard) of the light is just as important as the color. So if you work on getting those correct, you will 90% there already. When you are losing the light and scrambling to grab as many setups as possible, no one is going to wait for you to pull out your color meter and gel the lights accordingly! :)

Thanks I didnt think of that. While I knew waveform monitors existed I wasnt aware at all of how a calibrated monitor could help.

 

-------------------

[Diego Treves]

Some of this stuff can be done by eye and monitor - if you have one part of the frame lit by real sunlight and a closer subject lit by an HMI meant to simulate the sun, you can add some degree of CTO or CTS to get close to the color of the sun. After all the correction gels don't come any finer-tuned than in 1/8 increments anyway.

 

Besides, one reason for lighting a room with HMI's is to bring consistency -- the more that the HMI's are doing the work, the less critical it is to match whatever is going on outside hour to hour until it gets very different, like at sunset. So you might have a big HMI "sun" with 1/4 CTO on it all day long coming through the windows and fill with ungelled 5600K Kinos or LED's all day long -- it won't matter if the sun outside is drifting by a few hundred kelvin from mid morning to mid afternoon, or even if the sun goes away due to overcast.

 

Where color temp meters are more useful is in dealing with partial spectrum sources that have a green spike, which can be hard to see by eye but will show up on the monitor. You may find that it is easier to add, let's say, 1/4 Plus Green to your windows to "green up" the daylight to match the color of the overhead lights in the location, if those can't be swapped or turned off. You may find that an LED light needs 1/8 Minus Green on it to match an HMI you are using, or a daylight Kino.

 

I remember an old interview with Gordon Willis back in the 1970's where he said he didn't use a color temp meter much because most color mismatches he could see by eye, and if he couldn't, then they didn't matter. But of course this was back in the day before LED's and compact flos, etc. And he had well-trained eyes -- he could see the difference between 1/4 CTO and two layers of 1/8 CTO put together...

Thanks very helpful advice, I didnt know you could detect green spikes with a color meter, I thought more of a specific tool for that

[Mark Dunn]

 

let me ask you something: I understand that the strenght and the quality of your diffusion comes from the size of the light source: so if you want to have an even diffed light you cant certainly apply a small frame of gel to your lights; but on the other hand such a rule doesn't apply to color correction/conversion gels, meaning that you can surely apply a small frame of gel to your light and get the same result if you had applied a large one.

 

So .. it means I can buy Master Location Pack and use the small gels for color correction, while I need a larger area for diffusion, am I right?

 

 

Yes.

CC only needs to cover the front of the light. Diffusion is effectively creating a new light source as big as the frame, assuming you actually light all of it.

[Robin R Probyn]

The softness of the light also really depends on how close you are to the subject,as well as the size of the light.. i.e. closer the subject the light source gets bigger and the light softer.. if that makes sense .. the sun is a very big light source .. but its well.. light years away.. so it can be very harsh on a clear day.. but cloud cover can be like a big sheet of diff.. and its close to us. and big... so the light gets soft..

 

ie a huge frame diffused light wont be soft if its far from the object/subject..

[Diego Treves]

Ok thanks,

 

I need to practice all this, so that it becomes immediate knowledge.

 

And by the way I was thinking:

if I buy a roll of diffuser (I've been watching that demo clip on YT by Lee Filters performed on that beautiful model and I have to say that I like very much the 216 full white and the 129 full frost if I recall correctly- they have a very similar effect in that case.)

 

So if I buy a roll of diffuser(lets say the 216 because the frost feels to me somewhat more specific) I will be able to draw cuts of maximum "X" by 1,22 metres.

which is good, you can have a 4x4 feet frame which you can use to play with.

 

At that point the problem would be finding enough light to fill it all, and at this stage all I have is three 500w construction projectors, which won't probably be enough.

 

Adding a full CTB AND some white diffuser to a 500w projector I am afraid will result in such a small amount of light that wont be enough even for a photografic portrait ...

[Satsuki Murashige]

216 is a good diffusion to start with, it is a full white plastic diffusion so it is very smooth and even. 250 would be 1/2 diffusion and 251 is 1/4. Also very common is Gridcloth fabric which also comes in Full, 1/2, 1/4 and has a little texture to it.

 

You can put a large roll of diffusion on a horizontal c-stand arm and pull it down like a backdrop to make a simple large "frame." Then you can backlight it. Home Depot has some $15 heat lamps that are normal household bulb fixtures with tin reflector dishes and a little clamp. The large ones can take up 250w bulbs. So if you gang four of them together behind the diffusion you'll have enough light.

 

Even cheaper than diffusion is white muslin fabric which you can buy by the yard at any fabric store. There is bleached (pure white) and unbleached (warm-off white). You can place them on the floor or on tables, tape them to walls and windows, and even make roll out frames by stapling large sheets to a 2x4. Grab each side of the 2x4 with a c-stand and Cardellini clamp and you basically have a 12x12 frame. You'll need something punchy like a 2K open face or a few 1K parcans to push enough light through it though.

[Diego Treves]

Thanks very much very helpful

 

I was afraid you didnt reply anymore as I had put too much in this topic probably.

 

Now that I think Ive gathered enough information, what I need to do is take a while and realize what I really want in this field, what i really want to achieve.

 

thanks for the moment

[John E Clark]

But I am conflicted because color meters are really very very expensive. So my point is: either everyone approaching scene lighting buys a color meter (and I don’t think this is the case), or they mostly disregard it. But then once again, how can they be sure of what they do on the set? How can they meet precise directors’ guidelines and requests in

 

People who read or post in this group would probably use a color meter effectively. I've been lusting for the Sekonic C-700 color meter which gives a spectrum graph, which allows for oddities in the light 'color' to be more easily detected (green spike, lack of 'reds', etc...).

 

This figures into how to correct the light, or even if it is possible, given the light's spectrum.

 

Unfortunately the unit is $1500... There are iPhone or Android apps these days that give a rough estimate, and probably could be good enough for some level of users...

 

But it has been my experience that few people at the 'entry/low' end of cinematography actually use, or know how to use, meters... they look at the histogram, perhaps a wave form, or use 'zebras' or 'false color' displays...

 

Because I'm from the olden days, where one had 2 Film film color temperatures to work with, 3200 K for tungsten light, or 5600 K for daylight... I have tended to use my eye to adjust up or down from those basic settings. In the olden days, any residual 'color correction' was done in the printing process. Since I never printed my own color, I would check the printed still image, and send it back with 'notes' if it didn't turn out the way I expected. These days there are color correction tools that one can learn to use to develop a sense of what can be done to better balance the color of the image relative to one's expectation.

 

As for selecting color filters, the problem with modern non-tungsten lighting is that the filters are designed for a continuous spectrum, and with the spikes or holes in the spectrum that some lights have, the filter result may not deliver...

 

Hence the need for knowing the spectrum, and also the need for new designed filters as well...

 

That or just use tungsten no matter what...

Edited October 6, 2015 by John E Clark

[Diego Treves]

 

Unfortunately the unit is $1500... There are iPhone or Android apps these days that give a rough estimate, and probably could be good enough for some level of users...

 

 

Which iPhone almost costs the same ah ah

 

Bo, I have one of those apps and it completely does not work.

 

Thanks anyway for the answer

Edited October 6, 2015 by Diego Treves

[Stuart Allman]

John,

 

I noticed that the gel companies started making gels specifically for non-continuous spectrum lights. On Lee's website they have "LED Filters" as one of their major categories now. How well they work? Dunno. I had to return my C-700 demo unit. It's an ultra cool toy, but unfortunately the stuff I work on doesn't justify the cost of that tool.

 

Now if only the filter companies could work on adding phosphors to the gels to "fix" the LED spectrums! That would be Academy Award winning.

 

Stuart

------------------------------

illuma.blogspot.com

[David Mullen ASC]

Trouble is that a gel is a filter, i.e. it can only remove wavelengths. A gel cannot add missing colors. So a gel can cancel excess green from a spike in the spectrum but it isn't going to fill in any gaps.

[John E Clark]

Trouble is that a gel is a filter, i.e. it can only remove wavelengths. A gel cannot add missing colors. So a gel can cancel excess green from a spike in the spectrum but it isn't going to fill in any gaps.

 

Yeah, that's why Stuart added the 'if only there was a phosphor'... that would 1) absorb say the 'green' spike and 2) reradiate in a color where the light is 'weak'... thus 'filling' a spectrum hole...

 

My experience in such devices involved scintillators for converting X-Rays and Gamma Rays into 'visible' light, with photomultiplier tubes, and either vidicon tubes or silicon sensors, and writing software to analyze images derived from such... not exactly applicable to film production involving live humans...

 

Well... there was this device which used low dose x-rays for full live human scanning, but I wasn't part of that development... it's now in many airports around the world... at the time, the company couldn't 'give them away', except to prisons... and even then... but I digress...

Edited October 7, 2015 by John E Clark

[Stuart Allman]

David,

 

LED lights already have a blue "pump" built in. That's why there's always a spike in blue. Typically they are lacking in cyan and deep red.. So if the phosphor could fill in those cyan and deep red gaps you'd have something that could perform almost as well as tungsten for color reproduction. Essentially, the gel would be like the remote phosphor panel like you find on Cineo, BBS, and the new Arri Skypanel lights. I don't know if the phosphors exist that target those particular wavelengths.

 

The trouble is that LED vendors and their phosphor formulations are all over the place. You could tune it for one LED vendor, but it would be off for other vendors. It's probably too much trouble and too much of an interim solution for Rosco or Lee to deal with.

 

Stuart

---------------------------

illuma.blogspot.com