Saturday 5 May 2012

Defining ISO for cinematographers



I realise that this may be unnecessarily technical, but frustrated by the lack of definitions of ISO useful to a cinematographer, I went and worked one out myself:


The ISO scale is calibrated so that at 25fps and T2.8, a luminance of 14 foot-lamberts (equivalent to a grey card under approximately 84 foot-candles) will be rendered by ISO 100 as Zone V, and ISO changes are arithmetic, so that a doubling of the number corresponds to an increase of one stop and a halving of the number corresponds to a decrease of one stop.


We all know what foot-candles and foot-lamberts are, and we have a box that can measure them by making use of photoresistance. Easy. But then that box tells us what ISO, T-stop and shutter speed to use. Whaaaat? Isn't it refreshing to try and make the jump that would otherwise be forever hidden in the depths of the black box?

So bust out your lightmeters and give me some feedback.

Wednesday 18 April 2012

The myth of the uninflected camera


There is a certain school according to which there is something neutral, standard, uninflected that you can do with the camera. Something approaching objective. I don’t subscribe to this school and never have. I don’t think most people really believe in the logical conclusions of the theory, but they still bandy around phrases like ‘Sometimes it is important not to do anything with the camera’ – as if that was possible.

Camera movement or lack thereof, focal length, camera position; there is no neutral, standard, uninflected approach to any of these. There is no objective place, dear Bazin, to film a scene from; there is no objective focal length; there is no objective camera movement or lack thereof. Not even a lockoff. How is a lockoff ‘objective’? Do we stand still all the time and look at things? No.

“Camera movement should be motivated.” True but incomplete and therefore misleading. All camera behaviour should be motivated. The mapping is not a textbook mapping, it isn’t mathematical, it isn’t always intellectual. It can be instinctive. But let’s not misconstrue the above dictum and think that we should stick to a lockoff until we find enough ‘motivation’ to move the camera.

The search for a default, neutral, standard camera behaviour seems to me to be just a form of laziness, an attempt to establish a firm starting ground for every project, so you can say “The ‘standard’ style is a locked off 35mm at eye level, now let’s see if I can drum up the mental energy to decide whether to veer away from this today.” No. I reject this approach to our craft. We need to go back to square one for every film, every job, every project and ask what we need to do with the camera on this specific day to bring this specific material to life.

L.

London, 18th April 2012

Sunday 12 February 2012

Technicolor Cinestyle LUT curve

At the start of the year I shot a music video using Technicolor's Cinestyle loaded into a Canon 7D. I graded the video myself in Apple Color using the S-curve display LUT provided by Technicolor here. Instead of using the LUT as a display LUT, I rendered the grade with the LUT applied. The theory is that passing Cinestyle material through this LUT gives you a pretty standard DSLR-like image, but playing with the information in the image upstream of the LUT unlocks more shadow and highlight detail than you could gain access to without shooting Cinestyle. I did this by using the primary out curves.

It did slightly bug me not knowing what the LUT was doing exactly so I took the liberty of extracting the info from the LUT and plotting it myself. I hope it is of help to some.

PS By those alarmed by the 'DR theft' debate, it is unfounded, and you can read why here.

Saturday 11 February 2012

Lighting by candlelight on the Red MX

I recently shot the feature Nightmare Box for American director Jon Keeyes, and we had a long candlelight sequence round a table. When the subject of dollying 360º round the table was breached, I started worrying about how to light it all in a convincing fashion. Then I stumbled upon a camera test by David Fincher, filmed on the same camera as the one we would be using, the Red One Mysterium-X. The clip, which is shot at ISO 2000 and lensed with a Master Prime at T1.3, features Leonardo DiCaprio lighting a cigar with a match. The background is lit, but the key on DiCaprio’s face is provided exclusively by the match and cigar.





This clip brought home the total feasibility of using the Red MX with a T1.3 lens to light our scene exclusively with candlelight. After all, ISO 2000 / T1.3 corresponds to a 1.5 footcandle key. Production designer Eric Whitney sourced a large number of candles and very cleverly dug out the top and popped in tea lights, saving me hassle with changing light levels and saving himself hassle with candle height continuity.

We spent the sum total of a full shooting day dollying 360º around this table seating four characters with 10-14 candles (I forget the exact number) as the only light source*. We were able to get hold of MkII superspeeds thanks to Lightning Media in London. This was my first time using their superspeed set and I was astounded by the optical performance wide open – much more solid than other superspeed sets I’ve seen.

Below is an H.264-compressed HD quicktime of our camera test using a crew member as stand-in. Red MX @ ISO 2000 / 24fps 180º / 4500K / Superspeed T1.4**




A slight contrast curve is applied to the above file. For anyone interested, here is a .zip file containing the raw camera file for your scrutiny, as well as full-res TIFFs.

Check out the film when it comes out to see the full sequence!

* At the very end of the sequence a character stands, and I had to carry her with a dedo for that segment, but the rest of the sequence is lit purely by the candlelight.

** Although this test is 24 @ 180º, the actual scene was shot with a 225º shutter, i.e. 1/40th @ 25fps. We went to 25fps one day into the shoot for flicker reasons, and I opened the shutter on this scene to help the exposure out.