POST Archives - Budget-stretching tips for filmmaking, animation, motion graphics & visual effects

Sharpening your footage in post — in a non-crummy way

This one’s quick and easy, and there’s a good chance you already know this if you’re a seasoned post professional. But I see poorly sharpened footage all the time, so it must not be totes common knowledge.

Maybe your old no-name lenses are a little soft, or your exterior establishing shot needs to look a little more detailed, or you’re matching different lenses, whatevs. Do dis:

Don’t use Sharpen. That’s like a race car driver using an automatic transmission.
Use Unsharp Mask.
Right off the bat, set it to 100% Amount, and 0.3 Radius.
If that doesn’t do the trick, then try upping Amount to 200% at the most, or Radius to 0.5 maximum. But generally, going beyond that will cause negative byproducts.
Lastly, adjust Threshold as needed… for video there are many factors involved in deciding whether to set it high or low (types of skin, lighting, how much geometric stuff is in a shot, how noisy your footage is, etc), so really just slide it around and eyeball it to taste.

So there ya go… Unsharp Mask, 100% / 0.3, or adjust to taste within ranges 50-200% / 0.3-0.5. And generally, this applies to still photos as well. The settings should be pretty universal whether you’re doing it in Premiere Pro, Final Cut Pro, After Effects, Sony Vegas, or whatever else.

Before & after color grading comparison of flat scanned 35mm on the Outpost DVD

The British independent low-budget (£200,000 according to wikipedia) feature Outpost directed by Steve Barker achieved something pretty extraordinary– it rose above the endless flood of made-for-straight-to-DVD movies that belong to what I call the “post-Alien subgenre” (a group of soldiers/technicians who don’t like each other are stuck somewhere isolated/dark/claustrophobic, waiting it out for a specified amount of time until rescue/safety, while being killed off one by one by a mysterious & obscured creature/entity)… by being bought up by Sony and given a limited theatrical release in the UK.

Ok cool story, brodawg. Yeah so on the DVD there are deleted scenes that appear to be from an assemblage edit of the flat transfer film scan footage. Which means you can directly compare that to the look of the color graded final feature, if you’re into that kinda thing. It was shot on (Fuji) 35mm, so the flat transfer holds 13-14 stops of dynamic range (good prosumer video cameras and DSLR/DSLRish cameras hold 7-10 usable stops). The color grade’s bleach bypass/ENRish look is kinda par for the course considering the genre, but it’s definitely well crafted. And if you wear an eye patch, have a parrot on your shoulder, and use the word “booty” in a non-hiphop way, you can even rip the deleted scenes and try mimicking the final result… which would be pretty useful for anyone who’s learning to color grade. BUT THAT WOULD BE ILLEGALZ.

Here be the Outpost DVD on Amazon.

Efficient footage filenames when using an external recorder

If you haven’t noticed, most of the tidbits I put on this here blawg are born-ed of my trial & error, my woeful miseries, and my occasional tiny logistical victories over the ruthless gremlins of Murphy’s Law. Wait, that makes it sound like this article is going to be a lot more exciting than it actually will be — you’re all “dayumm this gonna have gremlins an’ stuff in it, like some Ghostbusters 3 shii ri here.”

Nope, boring. Just filename suggestions for recording footage to an external recorder like a Sound Devices PIX 240i.

Recently I was DP on a feature, shot on a Sony F3 with one. I assume other external recorders allow for flexible filenaming with automated take number advancement, just like the PIX 240i does — which is great. Because of this, I recommended to the producers that we shoot without slating. It can save significant time on a production, especially one like ours where we had a language barrier, chaotic locations, and a tight schedule… along with all the other factors of a low budget production.

So here’s the formula I recommend for the filename structure:
(project acronym)_(scene number)_(date)__(take number)

If your film’s called “The Buried Dirtball“, you’re shooting scene #26 on July 10th, and it’s the fifth take, then itsa gunna looka lika thissa:
bd_s026_07.10__005.mov

Though there’s one caveat… I don’t have a 240i in front of me, so I’m not 100% certain you can get all that stuff into the filename. But as I recall, you can manipulate the naming pretty heavily.
Including the date may seem cumbersome and maybe even unnecessary at first, but to me it is now essential. We didn’t use it when shooting the aforementioned feature, and that’s exactly how I learned that we should have. You may end up shooting shots from a particular scene many days apart from one another, so trying to figure out what the last numbered take was for that scene can be time consuming, if even possible at all when on location. But you probably can remember if it was earlier that day. So anyways, this can help prevent you from having shots with the same filename, which can be disastrous if one file happens to overwrite the other when placed into the same folder.

Note the three digit numeral for the scene & take numbers. That’s so if you put all your .mov files into the same folder, they’ll stack in proper order, making it easy for you to find stuffz.

Also, don’t use spaces instead of underscores… yes, I know it’s not 1997 anymore, but if for some reason the spaces cause some kind of problem for like your sound designer or colorist, then you’ll get to send a text message to yourself with nothing but a saddyface emoticon in it. Also, I’ll get to laugh loudly in your face. And I’ll probably make sure I eat something stinky immediately before. Like dog feces. That’ll teach you a lesson. Haha, in your face, bro. Jaykay, that’s just simulated schadenfreude– I would probably send you an upbeat, encouraging emoji to make you feel better, like that iPhone one of the twins in cat suits dancing.

The extra underscore before the take number is just there to visually scan better.

Also, you can use s000 (scene #0) for random unassigned stuff like 2nd unit exteriors, etc. I prefer this over other methods since those shots will all be easy to find in one place.

If you have a script supervisor on set, they can notate the “o’clock” time for shot numbers if you’re not slating (ie. shot3A was shot 2:16 – 2:38pm), and the assistant editor can then check the “date created” and rename the clips in the editing software, or just organize them into folders if you don’t need to go that hard.

Yeah, so anyways, based on my experiences on productions big and small, this format should work as a catch-all, with minimal time, energy, or headache during shooting and post. You can add more stuff to the filenames I guess, but going into the recorder’s interface to add shot numbers can be a slow and confusing hassle, especially if you’re not a fully crewed production with a dedicated script supervisor and 2nd AC. I would recommend the sorting of footage by shot be done in editing software, where it’s pretty simple & painless.

Using Super 16 lenses on the Blackmagic Cinema Camera: It works for a 1080 HD crop

There are Super 16 lenses that have been sitting around collecting dust for the last 5-10 years, and you can usually scoop some up at bargain prices. That’s because, aside from a minority of Red users, no one’s been using them, due to HD sucking all the wind from its sails like some sort of gigantic cloud vampire. No, I’m not drunk. Yeah so anyways, that all might change now that the Blackmagic Cinema Camera is starting to ship in quantity.

SUPER 16 ZOOM LENSES

The BMCC sensor size is a good chunk bigger than Super 16 (abbreviated “s16” from now on… I’m too lazy to type it out the bazillion more times I’ll need to reference it). But it’s close enough that some s16 zoom lenses will cover the sensor at the higher end of their focal length ranges, probably half to three-fourths of its zoom range, depending on the particular lens. That’s because the image circle of a zoom must be big enough to fill a format’s picture area when at its shortest focal length, ie. zoomed all the way out. And then by principle of a zoom lens, that image circle will get larger as you zoom in. As that image circle gets larger, it can hence fill a larger picture area/sensor. Here is a diagram featuring completely random and arbitrary imagery:

But here’s one thing to keep in mind: some still photography zoom lenses maintain their small image circle due to their particular design, so the same may apply to some s16 zooms as well.

***A LIL’ UPDATE***

Cinematographer John Brawley

The results don’t look too great for full BMCC sensor coverage. The Angenieux maintains a small image circle, and the Canons show noticeable chromatic aberration outside the s16 picture area, even when they cover the sensor. You can see all the DNGs via his dropbox link in this Blackmagic Forum thread.

A few paragraphs down, I’ve added examples of how these lenses can work for the BMCC via a 1080 center extract.
***

SUPER 16 PRIME LENSES

I have no idea if they’ll fill the BMCC sensor. Do long focal length primes have larger image circles than their short focal length counterparts? Maybe. Image circle sizes vary from manufacturer to manufacturer, and product line to product line. So the only way to know is to test that specific lens.

HOW TO USE ANY AND ALL SUPER 16 LENSES ON THE BMCC

This might be a big deal for some people. And it’s so simple: just crop into the 1920 x 1080 center pixels of the 2432 x 1366 frame of BMCC footage. Actually, you’ll only have to do that for footage that needs it. Like I mentioned earlier, some lenses will cover the whole sensor. But as a boilerplate, unilateral policy, it’ll work for all s16 lenses, regardless of focal length. Here’s why…

You can click here or on the picture to see a full res version of this BMCC 2432 x 1366 frame diagram. Notice how the Super 16mm picture area is just big enough to cover the extracted 1920 x 1080 HD frame from the full BMCC frame? Boom, there ya go. So it’s all good with using s16 lenses.

***UPDATE***

this Blackmagic Forum thread

John Brawley

***

I haven’t tested this myself because I’ve yet to get my filthy hands on a BMCC, but numbers don’t lie. There are a few things to keep in mind if you choose to do a 1080 center extract:

You’ll have to shoot in 2.5K RAW mode. Obviously the downsampled 1920 x 1080 ProRes or DNxHD modes won’t work for this.
You’ll need to mark the 1920 x 1080 center on your viewfinder. This is actually not that big of a deal. You just get some clear touch screen protective cover, and then make your markings via trial and error… ie. set up a tripod & chart/whatever, import footage & perform the extract, then mark the cover. I do this exact process for certain kinds of VFX shots all the time, it’s easy.
You’ll have to do the 1080 center extract in post, but it’s hecka easy. Do I even really need to explain this? I will, just in case. In whatever software, make your timeline/composition/whatever 1920 x 1080. Set your BMCC footage to be at 100% scale/size, that way it’s only showing the center. If a particular shot doesn’t need the 1080 extract, then change its scale/size to 79%.
It’s obviously not going to look as good as the full sensor 2400 x 1350 image that’s been scaled down to 1920 x 1080, because of Bayer filtering mumbo jumbo that you can google. Whether or not the optical resolution is good enough is up to you… but if your basis of comparison is a DSLR/DSLM, then it likely will. From the samples I’ve seen, a 1080 center extraction looks pretty great. And if you do see de-Bayer artifacts, try extracting from a slightly larger area than 1920 x 1080 if your lens’ image circle allows for it.

If you’re thinking it’s just weird and strange to capture image areas that are going to be discarded or unseen by the viewing audience, just keep in mind that it’s standard procedure with most non-anamorphic lensed 35mm shooting formats. Also, cropping in to 1080 on the BMCC is very similar, in principle, to the crop resolution modes of the Red cameras. Plus, really, your work is 99% likely to ultimately be seen in 1080 HD anyway. Is the 1080 center extraction worth all the extra work? Probably only if you just really wanna shoot on a fast cine zoom that’s affordable.

Also, if you’re content with shooting for a 720 HD finish, you can use standard 16mm format lenses… by cropping to 1578 x 888 and then downscaling to 1280 x 720.

Using a car interior as an audio booth

If you need to record clean ADR or voiceover, etc, and don’t have access to an audio booth, try recording inside a car (that’s not running, duh)… I guess they’re actually designed to have great acoustics. The only problem is that the actor can’t stand up, which often makes a big difference in delivery, especially if it’s an untrained actor… so maybe just always use really really short actors, I dunno. I can’t remember where I got this tip from… either a book, website or an audio person.

As I’m writing this, I remember a related experience from before I had heard of that. There’s a scene with two actors exchanging dialogue in the back seat of a parked car, which I shot from the front seat with a Sony VX1000.

As we were setting up mics, I was astonished to realize that the audio from the camera’s built-in stereo mic actually sounded better… which is the first and only time that has ever happened. But I guess it makes sense– that sedan had nice acoustics and we were in the middle of a quiet wooded area. Also, they were moving around a lot and keeping a directional mic properly aimed at them within such an enclosed space would’ve been difficult.

You can also get one of these relatively inexpensive thingees… I have one and it works great, especially for the price.

Color-saturated fill light for a better composite

Here’s a useful technique that I came up with for shooting certain kinds of greenscreen stuff…

Make the fill light a saturated color (that’s different from the screen and the subject’s clothes etc, duh).
In post, isolate that particular color and then tweak it to match the hue of whatever the natural bounce light would be in whatever environment you’re compositing them into.
Though keep in mind that it’s always better to just use a colored fill light that matches your particular background plate. This technique is more for cases where you have to use one setup for several varied backgrounds.

For example, you can use a saturated magenta for the fill, and then use the secondary color correction in Color Finesse to tweak just the saturated magenta to appear bluish if they’re supposed to be standing on the belly of a giant, naked sleeping smurf or whatever. Another great thing about this technique is that you not only can control the hue and saturation of the fill color, but also the luminance of the fill, and quite easily at that. Of course, this technique will only be useful for certain shots, depending on the lighting and blocking.

Here’s a straight forward example of this technique from a shoot I was DP on. Because of last minute schedule changes, I had to shoot numerous bits of Snoop Dogg in just one setup, even though they took place in a multitude of settings. We could’ve shot him in neutral color and depended on conventional color correction to help color him into the background, but that usually looks pretty rookie & unsavory in my opinion, since any color adjustments to affect the “bounce/ambient” light will also affect the entire subject’s color, even the areas being illuminated by the key light. So with a magenta fill, we had some tweakable “ambient” light to match the backgrounds in post.

You can see from the original footage on the right that not only was I able to easily change the hue & saturation of the magenta fill, but also the luminance. More examples from the exact same lighting setup below…

Also, note that in these examples (which I composited & color graded myself) the tweaking of the magenta fill was done with just the “Hue and Saturation” effect in After Effects due to what was available on that particular workstation. Theoretically, Color Finesse’s secondary color correction would’ve given even better results.

Miniatures/models episode of Movie Magic

Movie Magic was a really great nonfiction/documentary cable show in the 90s that gave actually useful insight into film special & visual effects, even though it was made for a general audience. I remember being a teenager and seeing the segment showing the multiple lighting passes of the photography of the Star Trek ship model and feeling the sensation of revelation as something clicked in my brain. Yeah so anyways, some of the episodes are on the ol’ youtubes, so check them out.

This episode on miniatures & models has a total booya quote:

“The great thing about shooting models as opposed to computer graphics is that you can get into a happy accident. A light can hit something a particular way, and you take advantage of that, and with computers you don’t necessarily do that, you have to plan for that accident.”
– Rob Legato, VFX Supervisor

youtube.com/watch?v=dEmXA0Itv6A