Undisclosed Project: Experimentation

March 15, 2021June 30, 2021 neilosemanAlexa, Angenieux, Arri, blog, bokeh, cinematography, Cooke, focus, lens flare, lenses, lighting, Panchro, Red, Red Gemini, Sony, Sony F55, spherical, Superspeed, testing, Varotal, video, Zeiss

The main event of last week’s prep was a test at Panavision of the Arri Alexa XT, Red Gemini and Sony F55, along with Cooke Panchro, Cooke Varotal, Zeiss Superspeed and Angenieux glass. More on that below, along with footage.

The week started with Zoom meetings with the costume designer, the make-up artist, potential fight choeographers and a theatrical lighting designer. The latter is handling a number of scenes which take place on a stage, which is a new and exciting collaboration for me. I met with her at the location the next day, along with the gaffer and best boy. After discussing the stage scenes and what extra sources we might need – even as some of them were starting to be rigged – I left the lighting designer to it. The rest of us then toured the various rooms of the location, with the best boy making notes and lighting plans on his tablet as the gaffer and I discussed them. They also took measurements and worked out what distro they would need, delivering a lighting kit list to production the next day.

Meanwhile, at the request of the producer, I began a shot list, beginning with two logistically complex scenes. Despite all the recces so far, I’ve not thought about shots as much as you might think, except where they are specified in the script or where they jumped out at me when viewing the location. I expect that much of the shot planning will be done during the rehearsals, using Artemis Pro. That’s much better and easier than sitting at home trying to imagine things, but it’s useful for other departments to be able to see a shot list as early as possible.

So, the camera tests. I knew all along that I wanted to test multiple cameras and lenses to find the right ones for this project, a practice that is common on features but which, for one reason and another, I’ve never had a proper chance to do before. So I was very excited to spend Wednesday at Panavision, not far from my old stomping ground in Perivale, playing around with expensive equipment.

Specifically we had: an Arri Alexa – a camera I’m very familiar with, and my gut instinct for shooting this project on; a Sony F55 – which I was curious to test because it was used to shoot the beautiful Outlander series; and a Red Gemini – because I haven’t used a Red in years and I wanted to check I wasn’t missing out on something awesome.

For lenses we had: a set of Cooke Panchros – again a gut instinct (I’ve never used them, but from what I’ve read they seemed to fit); a set of Zeiss Superspeeds – selected after reviewing my 2017 test footage from Arri Rental; a couple of Cooke Varotal zooms, and the equivalents by the ever-reliable Angenieux. Other than the Angenieux we used on the B-camera for The Little Mermaid (which I don’t think we ever zoomed during a take), I’ve not used cinema zooms before, but I want the old-fashioned look for this project.

Here are the edited highlights from the tests…

You’ll notice that the Sony F55 disappears from the video quite early on. This is because, although I quite liked the camera on the day, as soon as I looked at the images side by side I could see that the Sony was significantly softer than the other two.

So it was down to the Alexa vs. the Gemini, and the Cookes vs. the Superspeeds. I spent most of Thursday and all of Friday morning playing with the footage in DaVinci Resolve, trying to decide between these two pairs of very close contenders. I tried various LUTs, did some rough grading (very badly, because I’m not a colourist), tested how far I could brighten the footage before it broke down, and examined flares and bokeh obsessively.

Ultimately I chose the Cooke Panchros because (a) they have a beautiful and very natural-looking flare pattern, (b) the bokeh has a slight glow to it which I like, (c) the bokeh remains a nice shape when stopped down, unlike the Superspeeds’, which goes a bit geometric, (d) they seem sharper than the Superspeeds at the edges of frame when wide open, and (e) more lengths are available.

As for the zoom lenses (not included in the video), the Cooke and the Angenieux were very similar indeed. I chose the former because it focuses a little closer and the bokeh again has that nice glow.

I came very close to picking the Gemini as my camera. I think you’d have to say, objectively, it produces a better image than the Alexa, heretical as that may sound. The colours seem more realistic (although we didn’t shoot a colour chart, which was a major oversight) and it grades extremely well. But…

I’m not making a documentary. I want a cinematic look, and while the Gemini is by no means un-cinematic, the Alexa was clearly engineered by people who loved the look of film and strove to recreate it. When comparing the footage with the Godfather and Fanny and Alexander screen-grabs that are the touchstone of the look I want to create, the Alexa was just a little bit closer. My familiarity and comfort level with the Alexa was a factor too, and the ACs felt the same way.

I’m very glad to have tested the Gemini though, and next time I’m called upon to shoot something great and deliver in 4K (not a requirement on this project) I will know exactly where to turn. A couple of interesting things I learnt about it are: (1) whichever resolution (and concomitant crop factor) you select, you can record a down-scaled 2K ProRes file, and this goes for the Helium too; (2) 4K gives the Super-35 field of view, whereas 5K shows more, resulting in some lenses vignetting at this resolution.

How is Dynamic Range Measured?

August 31, 2020June 11, 2021 neilosemanAlexa, Arri, Blackmagic, camera, cinematography, dynamic range, exposure, latitude, lens flare, lenses, sensor

The high dynamic range of the ARRI Alexa Mini allowed me to retain all the sky detail in this shot from “Above the Clouds”.

Recently I’ve been pondering which camera to shoot an upcoming project on, so I consulted the ASC’s comparison chart. Amongst the many specs compared is dynamic range, and I noticed that the ARRI Alexa’s was given as 14+ stops, while the Blackmagic URSA’s is 15. Having used both cameras a fair bit, I can tell you that there’s no way in Hell that the Ursa has a higher dynamic range than the Alexa. So what’s going on here?

What is dynamic range?

To put it simply, dynamic range is the level of contrast that an imaging system can handle. To quote Alan Roberts, who we’ll come back to later:

This is normally calculated as the ratio of the exposure which just causes white clipping to the exposure level below which no details can be seen.

A photosite on a digital camera’s sensor outputs a voltage proportional to the amount of light hitting it, but at some point the voltage reaches a maximum, and no matter how much more light you add, it won’t change. At the other end of the scale, a photosite may receive so little light that it outputs no voltage, or at least nothing that’s discernible from the inherent electronic noise in the system. These upper and lower limits of brightness may be narrowed by image processing within the camera, with RAW recording usually retaining the full dynamic range, while linear Rec. 709 severely curtails it.

In photography and cinematography, we measure dynamic range in stops – doublings and halvings of light which I explain fully in this article. One stop is a ratio of 2:1, five stops are 32:1, thirteen stops are almost 10,000:1

It’s worth pausing here to point out the difference between dynamic range and latitude, a term which is sometimes regarded as synonymous, but it’s not. The latitude is a measure of how much the camera can be over- or under-exposed without losing any detail, and is dependent on both the dynamic range of the camera and the dynamic range of the scene. (A low-contrast scene will allow more latitude for incorrect exposure than a high-contrast scene.)

Problems of Measurement

Before digital cinema cameras were developed, video had a dynamic range of about seven stops. You could measure this relatively easily by shooting a greyscale chart and observing the waveform of the recorded image to see where the highlights levelled off and the shadows disappeared into the noise floor. With today’s dynamic ranges into double digits, simple charts are no longer practical, because you can’t manufacture white enough paper or black enough ink.

For his excellent video on dynamic range, Filmmaker IQ’s John Hess built a device fitted with a row of 1W LEDs, using layers of neutral density gel to make each one a stop darker than its neighbour. For the purposes of his demonstration, this works fine, but as Phil Rhodes points out on RedShark News, you start running into the issue of the dynamic range of the lens.

It may seem strange to think that a lens has dynamic range, and in the past when I’ve heard other DPs talk about certain glass being more or less contrasty, I admit that I haven’t thought much about what that means. What it means is flare, and not the good anamorphic streak kind, but the general veiling whereby a strong light shining into the lens will raise the overall brightness of the image as it bounces around the different elements. This lifts the shadows, producing a certain amount of milkiness. Even with high contrast lenses, ones which are less prone to veiling, the brightest light on your test device will cause some glare over the darkest one, when measuring the kind of dynamic range today’s cameras enjoy.

Manufacturer Measurements

Going back to my original query about the Alexa versus the URSA, let’s see exactly what the manufacturers say. ARRI specifically states that its sensor’s dynamic range is over 14 stops “as measured with the ARRI Dynamic Range Test Chart”. So what is this chart and how does it work? The official sales blurb runs thusly:

The ARRI DRTC-1 is a special test chart and analysis software for measurement of dynamic range and sensitivity of digital cameras. Through a unique stray light reduction concept this system is able to accurately measure up to 15.5 stops of dynamic range.

The “stray light reduction” is presumably to reduce the veiling mentioned earlier and provide more accurate results. This could be as simple as covering or turning off the brighter lights when measuring the dimmer ones.

I found a bit more information about the test chart in a 2011 camera shoot-out video, from that momentous time when digital was supplanting film as the cinematic acquisition format of choice. Rather than John Hess’s ND gel technique, the DRTC-1 opts for something else to regulate its light output, as ARRI’s Michael Bravin explains in the video:

There’s a piece of motion picture film behind it that’s checked with a densitometer, and what you do is you set the exposure for your camera, and where you lose detail in the vertical and horizontal lines is your clipping point, and where you lose detail because of noise in the shadow areas is your lowest exposure… and in between you end up finding the number of stops of dynamic range.

Blackmagic Design do not state how they measure the dynamic range of their cameras, but it may be a DSC Labs Xlya. This illuminated chart boasts a shutter system which “allows users to isolate and evaluate individual steps”, plus a “stepped xylophone shape” to minimise flare problems.

Art Adams, a cinema lens specialist at ARRI, and someone who’s frequently quoted in Blain Brown’s Cinematography: Theory & Practice, told Y.M. Cinema Magazine:

I used to do a lot of consulting with DSC Labs, who make camera test charts, so I own a 20-stop dynamic range chart (DSC Labs Xyla). This is what most manufacturers use to test dynamic range (although not ARRI, because our engineers don’t feel it’s precise enough) and I see what companies claim as usable stops. You can see that they are just barely above the noise floor.

Conclusions

Obviously these ARRI folks I keep quoting may be biased. I wanted to find an independent test that measures both Blackmagics and Alexas with the same conditions and methodology, but I couldn’t find one. There is plenty of anecdotal evidence that Alexas have a bigger dynamic range, in fact that’s widely accepted as fact, but quantifying the difference is harder. The most solid thing I could find is this, from a 2017 article about the Blackmagic Ursa Mini 4.6K (first generation):

The camera was measured at just over 14 stops of dynamic range in RAW 4:1 [and 13 stops in ProRes]. This is a good result, especially considering the price of the camera. To put this into perspective Alan measured the Canon C300 mkII at 15 stops of dynamic range. Both the URSA Mini 4.6 and C300 mkII are bettered by the ARRI Alexa and Amira, but then that comes as no surprise given their reputation and price.

The Alan mentioned is Alan Roberts, something of a legend when it comes to testing cameras. It is interesting to note that he is one of the key players behind the TLCI (Television Lighting Consistency Index), a mooted replacement for CRI (Colour Rendering Index). It’s interesting because this whole dynamic range business is starting to remind me of my investigation into CRI, and is leading me to a similar conclusion, that the numbers which the manufacturers give you are all but useless in real-world cinematography.

Whereas CRI at least has a standardised test, there’s no such thing for dynamic range. Therefore, until there is more transparency from manufacturers about how they measure it, I’d recommend ignoring their published values. As always when choosing a camera, shoot your own tests if at all possible. Even the most reliable numbers can’t tell you whether you’re going to like a camera’s look or not, or whether it’s right for the story you want to tell.

When tests aren’t possible, and I know that’s often the case in low-budget land, at least try to find an independent comparison. I’ll leave you with this video from the Slanted Lens, which compares the URSA Mini Pro G2 with the ARRI Amira (which uses the same Alev III sensor as the Alexa). They don’t measure the dynamic range, but you can at least see the images side by side, and in the end it’s the images that matter, not the numbers.

Camerimage 2017: Wednesday

December 2, 2017 neilosemanArri, blog, camera, Camerimage, cinematography, eye, festival, grading, HDR, lighting, Netflix, pipeline, Red, vision

This is the third and final part of my report from my time at Camerimage, the Polish film festival focused on cinematography. Read part one here and part two here.

Up.Grade: Human Vision & Colour Pipelines

I thought I would be one of the few people who would be bothered to get up and into town for this technical 10:15am seminar. But to the surprise of both myself and the organisers, the auditorium of the MCK Orzeł was once again packed – though I’d learnt to arrive in plenty of time to grab a ticket.

Up.grade is an international colour grading training programme. Their seminar was divided into two distinct halves: the first was a fascinating explanation of how human beings perceive colour, by Professor Andrew Stockman; the second was a basic overview of colour pipelines.

Prof. Stockman’s presentation – similar to his TED video above – had a lot of interesting nuggets about the way we see. Here are a few:

Our eyes record very little colour information compared with luminance info. You can blur the chrominance channel of an image considerably without seeing much difference; not so with the luminance channel.
Light hitting a rod or cone (sensor cells in our retinae) straightens the twist in the carbon double bond of a molecule. It’s a binary (on/off) response and it’s the same response for any frequency of light. It’s just that red, green and blue cones have different probabilities of absorbing different frequencies.
There are no blue cones in the centre of the fovea (the part of the retina responsible for detailed vision) because blue wavelengths would be out of focus due to the terrible chromatic aberration of our eyes’ lenses.
Data from the rods and cones is compressed in the retina to fit the bandwidth which the optical nerve can handle.
Metamers are colours that look the same but are created differently. For example, light with a wavelength of 575nm is perceived as yellow, but a mixture of 670nm (red) and 540nm (green) is also perceived as yellow, because the red and green cones are triggered in the same way in both scenarios. (Isn’t that weird? It’s like being unable to hear the difference between the note D and a combination of the notes C and E. It just goes to show how unreliable our senses really are.)
Our perception of colour changes according to its surroundings and the apparent colour of the lighting – a phenomenon perfectly demonstrated by the infamous white-gold/blue-black dress.

All in all, very interesting and well worth getting out of bed for!

At the end of the seminar I caught up with fellow DP Laura Howie, and her friend Ben, over coffee and cake. Then I sauntered leisurely to the Opera Nova and navigated the labyrinthine route to the first-floor lecture theatre, where I registered for the imminent Arri seminar.

Arri Seminar: International Support Programme

After picking up my complementary Arri torch, which was inexplicably disguised as a pen, I bumped into Chris Bouchard. Neither of us held high hopes that the Support Programme would be relevant to us, but we thought it was worth getting the lowdown just in case.

The Arri International Support Programme (ISP) is a worldwide scheme to provide emerging filmmakers with sponsored camera/lighting/grip equipment, postproduction services, and in some cases co-production or sales deals as well. Mandy Rahn, the programme’s leader, explained that it supports young people (though there is no strict age limit) making their first, second or third feature in the $500,000-$5,000,000 budget range. They support both drama and documentary, but not short-form projects, which ruled out any hopes I might have had that it could be useful for Ren: The Girl with the Mark.

Having noted these keys details, Chris and I decided to duck out and head elsewhere. While Chris checked out some cameras on the Canon stand, I had a little chat with the reps from American Cinematographer about some possible coverage of The Little Mermaid. We then popped over to the MCK and caught part of a Canon seminar, including a screening of the short documentary Kolkata. Shortly we were treading the familiar path back to the Opera Nova and the first-floor lecture theatre for a Kodak-sponsored session with Ed Lachman, ASC, only to find it had been cancelled for reasons unknown.

Red Seminar: High resolution Image Processing Pipeline

Next on our radar was a Red panel. I wasn’t entirely sure if I could handle another high resolution seminar, but I suggested we return once more to the MCK anyway and relax in the bar with one eye on the live video feed. Unfortunately we got there to find that the monitors had disappeared, so we had to go into the auditorium, where it was standing room only.

Light Iron colourist Ian Vertovec was talking about his experience grading the Netflix series GLOW, a highly enjoyable comedy-drama set behind the scenes of an eighties female wrestling show. Netflix wanted the series delivered in high dynamic range (HDR) and wide colour gamut (WCG), of a spec so high that no screens are yet capable of displaying it. In fact Vertovec graded in P3 (the colour space used for cinema projection) which was then mapped to Netflix’s higher specs for delivery. The Rec.709 (standard gamut) version was automatically created from the P3 grade by Dolby Vision software which analysed the episodes frame by frame. Netflix streams a 4,000 NIT signal to all viewers, which is then down-converted live (using XML data also generated by the Dolby Vision software) to 100, 650 or 1,000 NITs depending on their display. In theory this should provide a consistent image across all screens.

Vertovec demonstrated his image pipeline for GLOW: multi-layer base grade, halation pass, custom film LUT, blur/sharp pass, grain pass. The aim was to get the look of telecined film. The halation pass involved making a copy of the image, keying out all but the highlights, blurring those highlights and layering them back on top of the original footage. I used to do a similar thing to soften Mini-DV footage back in the day!

An interesting point was made about practicals in HDR. If you have an actor in front of or close to a practical lamp in frame, it’s a delicate balancing act to get them bright enough to look real, yet not so bright that it hurts your eyes to look at the actor with a dazzling lamp next to them. When practicals are further away from your cast they can be brighter because your eye will naturally track around them as in real life.

Next up was Dan Duran from Red, who explained a new LUT that is being rolled out across their cameras. Most of this went in one ear and out the other!

“Breaking Bad”

Afterwards, Chris and I returned to Kung Fusion for another delicious dinner. The final event of the day which I wanted to catch was Breaking Bad‘s pilot episode, screening at Bydgoszcz’s Vue multiplex as part of the festival’s John Toll retrospective. Having binged the entire series relatively recently, I loved seeing the very first episode again – especially on the big screen – with the fore-knowledge of where the characters would end up.

Later Chris introduced me to DP Sebastian Cort, and the three of us decided to try our luck at getting into the Panavision party. We snuck around the back of the venue and into one of the peripheral buildings, only to be immediately collared by a bouncer and sent packing!

This ignoble failure marked the end of my Camerimage experience, more or less. After another drink or two at Cheat we called it a night, and I was on an early flight back to Stansted the next morning. I met some interesting people and learnt a lot from the seminars. There were some complaints that the festival was over-subscribed, and indeed – as I have described – you had to be quick off the mark to get into certain events, but that was pretty much what I had been expecting. I certainly won’t put be off attending again in the future.

To learn more about two of the key issues raised at this year’s Camerimage, check out my Red Shark articles:

Grading “Above the Clouds”

November 11, 2017November 12, 2017 neilosemanAlexa, Alexa Mini, Amira, Arri, Blackmagic, cinematography, colour, DSLR, grading, postproduction

Recently work began on colour grading Above the Clouds, a comedy road movie I shot for director Leon Chambers. I’ve covered every day of shooting here on my blog, but the story wouldn’t be complete without an account of this crucial stage of postproduction.

I must confess I didn’t give much thought to the grade during the shoot, monitoring in Rec.709 and not envisaging any particular “look”. So when Leon asked if I had any thoughts or references to pass on to colourist Duncan Russell, I had to put my thinking cap on. I came up with a few different ideas and met with Leon to discuss them. The one that clicked with his own thoughts was a super-saturated vintage postcard (above). He also liked how, in a frame grab I’d been playing about with, I had warmed up the yellow of the car – an important character in the movie!

Leon was keen to position Above the Clouds‘ visual tone somewhere between the grim reality of a typical British drama and the high-key gloss of Hollywood comedies. Finding exactly the right spot on that wide spectrum was the challenge!

“Real but beautiful” was Duncan’s mantra when Leon and I sat down with him last week for a session in Freefolk’s Baselight One suite. He pointed to the John Lewis “Tiny Dancer” ad as a good touchstone for this approach.

We spent the day looking at the film’s key sequences. There was a shot of Charlie, Oz and the Yellow Peril (the car) outside the garage from week one which Duncan used to establish a look for the three characters. It’s commonplace nowadays to track faces and apply individual grades to them, making it possible to fine-tune skin-tones with digital precision. I’m pleased that Duncan embraced the existing contrast between Charlie’s pale, freckled innocence and Oz’s dirty, craggy world-weariness.

Above the Clouds was mainly shot on an Alexa Mini, in Log C ProRes 4444, so there was plenty of detail captured beyond the Rec.709 image that I was (mostly) monitoring. A simple example of this coming in useful is the torchlight charity shop scene, shot at the end of week two. At one point Leo reaches for something on a shelf and his arm moves right in front of his torch. Power-windowing Leo’s arm, Duncan was able to bring back the highlight detail, because it had all been captured in the Log C.

But just because all the detail is there, it doesn’t mean you can always use it. Take the gallery scenes, also shot in week two, at the Turner Contemporary in Margate. The location has large sea-view windows and white walls. Many of the key shots featured Oz and Charlie with their backs towards the windows. This is a classic contrasty situation, but I knew from checking the false colours in log mode that all the detail was being captured.

Duncan initially tried to retain all the exterior detail in the grade, by separating the highlights from the mid-tones and treating them differently. He succeeded, but it didn’t look real. It looked like Oz and Charlie were green-screened over a separate background. Our subconscious minds know that a daylight exterior cannot be only slightly brighter than an interior, so it appeared artificial. It was necessary to back off on the sky detail to keep it feeling real. (Had we been grading in HDR [High Dynamic Range], which may one day be the norm, we could theoretically have retained all the detail while still keeping it realistic. However, if what I’ve heard of HDR is correct, it may have been unpleasant for audiences to look at Charlie and Oz against the bright light of the window beyond.)

There were other technical challenges to deal with in the film as well. One was the infra-red problem we encountered with our ND filters during last autumn’s pick-ups, which meant that Duncan had to key out Oz’s apparently pink jacket and restore it to blue. Another was the mix of formats employed for the various pick-ups: in addition to the Alexa Mini, there was footage from an Arri Amira, a Blackmagic Micro Cinema Camera (BMMCC) and even a Canon 5D Mk III. Although the latter had an intentionally different look, the other three had to match as closely as possible.

A twilight scene set in a rural village contains perhaps the most disparate elements. Many shots were done day-for-dusk on the Alexa Mini in Scotland, at the end of week four. Additional angles were captured on the BMMCC in Kent a few months later, both day-for-dusk and dusk-for-dusk. This outdoor material continues directly into indoor scenes, shot on a set this February on the Amira. Having said all that, they didn’t match too badly at all, but some juggling was required to find a level of darkness that worked for the whole sequence while retaining consistency.

In other sequences, like the ones in Margate near the start of the film, a big continuity issue is the clouds. Given the film’s title, I always tried to frame in plenty of sky and retain detail in it, using graduated ND filters where necessary. Duncan was able to bring out, suppress or manipulate detail as needed, to maintain continuity with adjacent shots.

Consistency is important in a big-picture sense too. One of the last scenes we looked at was the interior of Leo’s house, from weeks two and three, for which Duncan hit upon a nice, painterly grade with a bit of mystery to it. The question is, does that jar with the rest of the movie, which is fairly light overall, and does it give the audience the right clues about the tone of the scene which will unfold? We may not know the answers until we watch the whole film through.

Duncan has plenty more work to do on Above the Clouds, but I’m confident it’s in very good hands. I will probably attend another session when it’s close to completion, so watch this space for that.

See all my Above the Clouds posts here, or visit the official website.

The Cinematography of “Perplexed Music”

August 31, 2017July 21, 2020 neilosemanAlexa, Arri, Blackmagic, camera, cinematography, composition, crowd funding, crowdfunding, drone, ISO, lenses, lighting, on set, short film, slow motion, soft light, Ultra Primes

In June I was recommended by a mutual friend to shoot a short drama called Perplexed Music, inspired by the Elizabeth Barrett Browning sonnet of the same name. It’s a passion project from writer-director Mark McGann, with his brother Paul McGann (Doctor Who, Alien 3, Withnail & I) in the lead role of a man grieving for his deceased partner.

Paul and Mark pose with one of the supporting artists between takes.

Mark was keen from the outset to shoot on an Alexa, and I was quick to agree. Arri Rental very kindly gave us an amazing deal on an Alexa Classic and a set of Ultra Primes. As on Above the Clouds, we used a Blackmagic Micro Cinema Camera as a B-cam, capturing two specific angles that were impossible on the Alexa with our limited grip budget.

Throughout July, Mark and I had a very satisfying creative dialogue about the cinematic techniques we would use to tell the story of Paul’s character, The Man, who never speaks. I had been watching a lot of Mr. Robot, and was keen to use unusual compositions as that show does. The visual grammar that we ultimately developed eschewed The Rule of Thirds, either squeezing The Man right into the side of frame – at times when things are too much for him – or placing him dead centre for moments of clarity and acceptance, and for flashbacks to when his partner was alive.

The Blackmagic Micro Cinema Camera is mounted on a combo lighting stand to capture a high angle through a streetlamp.

While testing lenses at Arri Rental a few weeks prior to the shoot, I took the opportunity to shoot some frame-rate tests between 24 and 48fps. Since the film has so little dialogue, I figured there was nothing to stop us using a lot of slow motion if we wanted to. I didn’t want it to look like a music video though. I thought perhaps a very subtle over-cranking, creating languid blinks and slightly heavier movement, would add to the burden of The Man’s grief. Mark agreed as soon as he saw the tests, and we ended up shooting a number of set-ups at 28 and 30fps, plus 40fps for a pivotal sequence.

I also tested various ISO settings on the Alexa (click here for full details, stills and video from this test). Based on these, I decided to use ISO 1600 for the majority of the film, partly for the extra latitude in the highlights, and partly to add grittiness to The Man’s grief-stricken world, in the form of a little picture noise. When we started shooting the flashbacks, on the spur of the moment I decided to switch to ISO 400 for these. A few years back I shot the music video below on a Red Epic and, for reasons I forget, one set-up was done at a lower ISO than the rest. I remember the feeling this gave, when I saw the final edit, of everything suddenly being smooth and hyper-real. I thought that would be a great feeling to give to the flashbacks.

1st AC Rupert Peddle and 2nd AC Ben Davies set up a lakeside close-up under a diffusion frame which will soften the light on Paul.

Much of Perplexed Music was day exterior, but a couple of sequences required lighting. In the opening café scene, I fired HMIs through two windows, but kept their light away from The Man, keying him with a practical to put him in his own little world. Meanwhile, a happy couple he’s watching are bathed in sunlight (sometimes real, sometimes not) warmed up with a quarter CTO, and bouncing beautifully off their table to give them a healthy glow.

For night interiors at The Man’s home, I was keen to rely on practicals as much as possible. Firstly there wasn’t much space in the little cottage, secondly I didn’t want the hassle of having to shift them around to keep them out of frame when we changed angle, and thirdly it just looks more natural. So aside from a tungsten bounce in a corner of the living room we knew would never be seen, I stuck to practical table lamps and exterior lighting.

Setting up for a night exterior shot. Photo: Gary Horton

I had planned to use direct HMI sources for moonlight through the windows, but my gaffer Sam suggested going softer so that we wouldn’t have hard shadows inside which would need filling. I saw that he was right, so we used a kino through one window and a 2.5K HMI bounced into poly through another (pictured at left).

Perplexed Music was shot over five days in Frome in Somerset and Rame in Cornwall. The latter provided us with a spectacular cliff-top and the isolated St. Michael’s Chapel on the peak of the headland. Here we employed the services of The Fly Company, who captured two dramatic, sweeping shots on their DJI Inspire 2 drone. We were all extremely impressed by what they were able to achieve, especially as it was done in very windy conditions, in between rain showers.

We completed the final set-ups of the schedule as the winds began gusting up to 60mph, and poor Paul could barely stand upright! I was certainly glad we picked the Alexa to shoot on, because anything lighter would probably have shaken during takes, if not blown over!

Lining up a shot with director Mark McGann. Photo: Gary Horton

I had a fantastic time working with Mark and Paul, and the whole cast and crew. We were sad to part ways at the end of the week, and we all look forward to seeing the finished film soon. And at this point, dear reader, I ask for your help. Currently a Kickstarter campaign is underway for postproduction. It’s well over 50% funded at the time of writing, but every little helps in our quest to reach the finishing line. Rewards for backers include thank you video messages from Paul and Mark, and tickets to a private screening in December. Even if you can’t contribute, please consider sharing the page on social media. Thanks!

https://www.kickstarter.com/projects/perplexedmusic/perplexed-music-post-production

Anamorphic Lens Tests

August 7, 2017August 18, 2017 neilosemanAlexa, anamorphic, Arri, aspect ratio, bokeh, cinematography, Cooke, focus, Hawk, Kowa, lens flare, lenses, testing, video

Anamorphic cinematography, first dabbled with in the 1920s, was popularised by Twentieth Century Fox in the fifties as CinemaScope. Television was growing in popularity and the studios were inventing gimmicks left, right and centre to encourage audiences back into cinemas. Fox’s idea was to immerse viewers in an image far wider than they were used to, but with minimal modifications to existing 4-perf 35mm projectors. They developed a system of anamorphic lenses containing elements which compressed the image horizontally by a factor of two. By placing a corresponding anamorphosing lens onto existing projectors, the image was unsqueezed into an aspect ratio of 2.55:1, or later 2.39:1.

Since those early days of CinemaScope, anamorphic cinematography has become associated with the biggest Hollywood blockbusters. Its optical features – streak flares, oval bokeh and curved horizontal lines – have been seared into our collective consciousness, indelibly associated with high production values.

I’ve not yet been fortunate enough to shoot anamorphic, but I was able to test a few lenses at Arri Rental recently, with the help of Rupert Peddle and Bex Clives. Last week I wrote about the spherical lenses which we tested; our anamorphic tests followed the same methodology.

Again we were shooting on an Alexa XT Plus in log C ProRes 4444 XQ, this time in 4:3 mode, a resolution of 2048×1536. Since all of the lenses had a standard 2:1 anamorphosing ratio, the images unsqueezed to a super-wide 2.66:1 ratio. (This is because the lenses were designed to be used on 35mm film with space left to one side for the optical soundtrack.) You can see the full width of this ratio in the first split-screen image in the video, at 2:08, and in the second image below, but otherwise I have horizontally cropped the footage to the standard 2.39:1 ratio.

We tested the following glass:

Series	Length	Speed	CF*	Weight
Hawk V	35mm	T2.2	30″	5.6kg
Cooke Xtal	30mm	T2.8	?	3kg
Kowa Mirrorscope	40mm	T2.2	36″	1.15kg
Kowa Mirrorscope	30mm	T2.3	?	?

* CF = close focus

For consistency with the spherical lenses, we used lengths around 32mm, but in the anamorphic format this is a pretty wide lens, not a mid-range lens. We shot at T2.8, again for consistency, but I hear that many anamorphics don’t perform well wider than T4.

We were only able to test what Arri Rental happened to have on the shelves that day. The biggest and presumably most expensive was the Hawk V-series. Next in size and weight was the Cooke Xtal – pronounced “crystal” – a 1970s lens based on the much-loved Speed Panchros. The smallest and lightest, was the Kowa Mirrorscope, with a list price of £1,200 per week for a set of four. (Sorry, I couldn’t find any pricing info for the others online.) Note that there isn’t really a 30mm Mirrorscope; to get this length you put a wide angle adapter on the 40mm. As this extra element decreases the optical performance, we tested it with and without, hence the two lengths.

Here’s the video…

Skin tones

Click on the image to see it at full quality.

To my eye, the Hawk has a fairly rich, warm skin tone, while the Cooke – as with the spherical S4 tested last week – seems a little grey and flat. The Kowa is inexplicably brighter than the other two lenses, which makes it hard to compare, but perhaps it’s a little cooler in tone?

Sharpness

Focus is more critical with anamorphic lenses than spherical ones. From a forum posting by Max Jacoby:

Anamorphic lenses have what is known as a “curved field of focus” that works similarly to the curved movie screens in some large Cinerama theatres. This is one reason that one needs to expose these lenses at a deeper stop. If one doesn’t, the curved field will not be covered by depth of field and either the edges or centre of the frame will be soft.

One day I’d like to re-test these lenses at a lower stop, T4 or T5.6, where they will all undoubtedly perform much better. But in this T2.8 test, on Bex’s face in the centre of frame, the Hawk V and the Kowa Mirrorscope 40mm – both almost a full stop from their maximum apertures – are clearly the sharpest of the bunch. The Cooke Xtal, which is wide open, is unsurprisingly softer. The 30mm adapter on the Mirrorscope completely destroys the image, not only making it very soft but also introducing colour aberration.

Now let’s look at the checkerboard at the side of frame and see if we can spot any differences in sharpness there…

It seems to me that the Kowa, both with and without the adapter, has a greater difference in sharpness between the centre and edges of frame than the the Hawk and Cooke. With the latter two lenses, the checkerboard is reasonably sharp, at least on the lefthand side, with some ghosting/blur visible towards the righthand side. The same thing can be observed on the chart in the flare tests at the end of the video.

Breathing & Bokeh

All of these lenses have a noticeable degree of breathe, which I suppose is to be expected from anamorphics. The Hawk V has roughly oval bokeh, the Cooke’s is more circular, while the Mirrorscope has interesting D-shaped bokeh.

Flare

The Hawk V doesn’t flare much at all, which is apparently due to the anamorphic element being in the middle of the lens, rather than at the front. The Kowa has a nice streak and glow around the light source, with a funky purple artefact on the opposite side of frame. But it’s the Cooke Xtal which provides the most classic lens flare, with a horizontal line across most of the frame and a partial star pattern around the source, despite the lens being wide open.

At the end of the video you can see how the flares develop on each lens as the light source moves horizontally across frame.

Distortion

A bulging effect is very obvious on all of these lenses, due to the focal lengths being quite wide for anamorphic. Notice how at 40mm on the Kowa Mirrorscope this curvature of the image is significantly reduced.

It’s hard to compare the levels of distortion because none of the focal lengths are exactly the same, except for the Cooke Xtal and the Kowa Mirrorscope with the 30mm adapter on. The Cooke’s top right and bottom left corners appear to be stretched away from the centre relative to the other two corners. I suppose that strange and funky stuff like this is exactly why you choose vintage glass.

Interestingly, the Cooke’s image appears a little tighter than the Kowa’s, which combined with my inability to find any evidence online of the existence of a 30mm Xtal, leads me to suspect we may have been given a mislabelled 32mm.