Phil Rhodes

I notice those are squeegee mounts, not the rubber blades themselves. They seem intended to allow the use of other branded parts in Arri processors. I'd be incredibly cautious about trying to create something like that which (I presume, not knowing much about film processing) actually contact the film. I mean, I'd do it, I'd just be massively cautious about testing it to make sure I wasn't going to ruin everything.

This is why I don't get the Marvel movies. Sorry, anyone who worked on them, but if you're going to do that sort of movie - and there's nothing wrong with a decent popcorn movie - at least make it look - well - like Transformers or Fast and Furious. I happened across a bit of urban exploration video (see here) the other day where some people explored an old (very old, 1920) power plant, the Port Richmond Generating Station in Philadelphia, which by my eye was used quite extensively on one of the Transformers movies (see here.) It's an amazing looking place, but I'm not quite sure how people make that stuff look... like it does. Track right, pan left, sprint, shout, air cannon, blue, yellow? Maybe it's because I grew up in a place where everything is either grey or brown, but I must admit to a certain liking for this sort of thing. P

Nikon E series. Get the 28, 50 and 100. You'll be all in for £150-200 and equipped for most stuff, and you can add an optical reducer for ludicrous speed.

What can you get a Blackmagic Pocket for these days? I'd go for that in a heartbeat.

That you'd have to test. They're both a similar bitrate but the relative cleverness of the codecs will differ. You'd have to pick up Nanoflash used, I suspect (which is fine, they don't go bad) and shoot some example stuff. Make sure yo ucover a lot of bases - interiors, exteriors, overexposure, underexposure, motion, etc.

You can get an F3 for that!

Atomos also have a little recorder in the Ninja range that has no display and is intended for this sort of situation. Or, yes, hyperdeck shuttle, or even Nanoflash, which will give you 50Mbps MPEG-2, which is super easy to handle for more or less everything if not particularly wonderful quality-wise.

I wrote a piece about this some time ago. Film is conceptually simpler. You shoot it with whatever production design, lighting, filtration and lenses you can muster. You process it, potentially using special techniques but generally not. You colour time and you release. Nearly a century of (theatrical) filmmaking was done this way. Most of the best films ever made were done this way. There are way, way more variables in digital workflows and that is not a good thing. Much of it is down to manufacturers trying to create saleable intellectual property; is it really necessary for every manufacturer to have at least one of its own special log encodings? Is it really necessary for every manufacturer to spin its own codec, especially when may of those codecs (H.264, AVC Intra, XAVC) are basically MPEG-4 with different settings, and many others (ProRes, MJPEG, DV(CPRO)(50)(HD), HDCAM, DnXHD) are technologically near-identical. There are really only two ways of doing lossy compression on video images and they are used by more or less everything. The endless patent fight over codecs is manifestly unhelpful and has made much good and well-intentioned work unusable (H.265.) Do we really need a flash card format per camera, or, as Red recently showed, a completely off the shelf flash card in a pretty and expensive hat? Do we need every editor to have its own timeline description file format? Do we need an HDR distribution format for every manufacturer? Do we need five different ways of characterising the colour quality of white light? I could go on, but no. No, we really, really don't. All of this is completely unnecessary and could trivially be standardised. There is no caveat to this. It should happen right now, and the fact that it's got this bad, in the face of a century of much less complicated filmmaking, is an absolute travesty. It creates extra work and constantly encourages serious, expensive mistakes to occur. A film-style workflow can be implemented with current tech, but generally isn't. On the basis the one per cent need their ferraris, it is of course massively unlikely to happen. P

I think there's some confusion happening here between what HDR tends to mean in still photography, and what it means in moviemaking. What you've posted there is not what people are generally going for in moving image HDR.

I'd get very friendly with my bank manager. There isn't really a good, cheap solution to this. I think that some of the most keenly-priced stuff is probably from Core SWX, but you're right to be suspicious of used batteries. That said, the only alternative I'm aware of involves buying old ones and recelling them. If you're handy, it's not too bad. Some are easier than others, and some very modern types may have internal circuitry that may become unhappy if you try this. That said, if you really want the inexpensive solution, look around for some old nickel cadmium or nickel metal hydride packs and recell them. You could possibly do lithium-ion, but it may be better to look around for something you can get a nickel metal hydride charger for - you don't want to be building new cadmium-based batteries at this point - and figure out what type of cell they take and the mechanics of doing it. I have some old PAG stuff that I did like that. It's not very pretty and I tend to use it for lighting rather than cameras, but at some point a box full of chemical goop is a box full of chemical goop. P

Most people will just look at it and see a very historical-looking image. For anyone who's into this stuff, it very clearly look like early monochrome photography, exactly as you'd expect. The thumbnail they're using for a lot of the YouTube stuff reminds me of the early photos you see of engineering types standing proudly in front of huge bits of stuff they've built. Only, you know. Sharper! P

Sound Forge has some great click and crackle removal filters.

Best to post some video, or at least stills, so we can have a look at the problem. Shooting that sort of stage lighting is notoriously tricky, for several reasons. What looks good on camera often doesn't look good to the naked eye. Easier to say more once I've seen it. P

There's a wonderful halo on the shot of the lighthouse in the trailer. One daft question if you'll indulge me - there's a rather artistic shot of the lighthouse's lens system rotating which is used under the closing title of the trailer. Is that something that was shot, or is it something the title people came up with? P

Is it me misremembering or does 5222 not have the anti-halation backing?

I thought that! The two you can see are the viewfinder power and picture which are as designed on the Ursa. There's one more patching the camera to the TX that you can't see. The problem is that you really want a right-angle BNC adaptor on the input to the transmitter, but I'm not quite sure if that's a great idea with 12 gig SDI for 4K pictures at 60 frames!

I'm not sure if this is really supposed to be a venue for gear porn, but I happen to have a bunch of stuff lying around at the moment which allows me to create this monstrosity. It comprises an Ursa Mini Pro with the grip kit and viewfinder, Wooden Camera D-Box power distribution and UBM-1 mattebox, Teradek Bolt 4K transmitter and Anton-Bauer Titon 150 batteries. All it needs is a decent lens. How much is a Fuji XK6x20 now?

Coemar Centros are physically smaller than most lekos, and they're not the only small type. There are miniature Source Fours, too. Small optics do start to suffer a quite noticeable efficiency falloff, though. P

Mired shifts better represent the apparent difference between colour temperatures. Divide 1,000,000 by the CT to get the mired value of any light source; compare two by subtracting them (order the calculation so that negative values mean bluer.) The exposure value required to properly expose a scene is the binary logarithm of the light level in lux multiplied by the ISO, divided by C, where C is the meter calibration constant usually equal to about 330. For 1000 lux at 100 ISO, calculate: 1000 × 100 ÷ 330 = ln ÷ 2 ln = 8.24 The exposure value of a camera setup is the binary logarithm of the the aperture squared over the shutter time in seconds. For 1/48s at f/2.8, calculate: 2.82 ÷ (1 ÷ 48) = ln ÷ 2 ln = 8.56 (The calculations above are given in reverse polish notation, that is, that's the order you'd hit the keys on a scientific calculator.)

I'm not sure if there's a practical difference between a beam projector and a profile (er, ellipsoidal.) The results are arguably fairly similar, but a beam projector like a Molebeam where the entire optical system consists of a single parabolic reflector (and perhaps a few baffles) is, I'd think, likely to be quite a bit more efficient. Less flexible, though. From what I've seen, a Molebeam or similar is designed to project as near to a parallel beam of light as possible. Most Source Fours, Dedos, etc, won't really do that, although I suspect that in many situations that might not matter too much. However, I have often seen them used to do things like sunbeams through windows, often in haze so it's visible. In this situation it is much more correct to have a near-parallel beam, given the effectively infinite distance of the sun. The limiting factor is of course the diameter of the reflector; for really correct results the reflector should be larger than the window (for instance) you're projecting the light through. For that reason I'm not sure if a small beam projector really makes that much sense unless you have a specific application in mind - you might just as well go and buy some Source Fours. I bought old (late 90s) Coemar Centro and converted them to 150W ceramic. P

Saw it today. Noticed that it does actually look like a film, and I mean that in the best possible sense of the word, rather like a film from the period in which it is set. There's actual contrast, a bit of bite. Frankly, it reminded me how gutless a lot of movies look these days. There's darkness without suffering mush and muddiness. Many technical things I could obsess over, but in general hats off to all involved.

I'm always quite amused by the concept of 500-speed 35mm and it somehow having lots of resolution. It doesn't. Then push it? Ouch. P

My view on this is that if a production company kept on doing this, there would be something to complain about. If it was clear that this sort of situation were being created deliberately, it would be worth raising a question about it, politely and gently in the first instance. If it were to go on long-term, then fine, if you insist, if you really must, if you can't solve your own problems, that's what the union is for. To give someone a hard time for solving an immediate-term problem, in the absence of these things, is absurd, asinine, and the behaviour of a clock-watching apparatchik whose sincerity I doubt on principle.

Isn't it really the production's problem more than yours?

I like the first image, but the second one looks clipped, both in saturation and gamut (yes, you can clip a camera by going outside its colour range, as well as its brightness range.) This can be hard to avoid with LEDs which tend to create colours by mixing red, green and blue primaries. This means that you can theoretically mix pastel shades, especially with the addition of white, but in the end you're still usually mixing red, green and blue spikes and they may still be out of gamut no matter how dim they are. This is the dirty underbelly of colour mixing LEDs which isn't much discussed. In my view if you want this to look its absolute best, gel white lights, or at least compare the results you get with various types of gels and LEDs. Some LEDs use phosphor-converted colours and are better. P