Last time, I tried to shoot some basic camera shots in the style of an Ozu film, with the intention to “re-shoot” a scene from The End of Summer. There were a few issues I had, which in the end was an issue with blocking, and understanding the limitations of different lenses. Given the size of the space, I couldn’t shoot some of the “head-on” portrait shots that Ozu does, even using a 35mm lens, at least not without introducing a lot of distracting distortion.

This time, I shot with a 50mm lens and made sure I had the characters placed in such a way that I could have the distance I needed to get the shots I needed, and I think it really worked out well this time. The only thing I was a little worried about was because I was using a 50mm lens, I would have less leeway to adjust the shot (something that I probably have to do in general, since I can’t actually actively see myself in front of the camera and have to do a little guesstimating. It still worked out though, and I’ve started to develop a system now for how I set up those shots, both in terms of camera distance (subject about 5-6 feet away from the camera) and also camera height and angle (shoulder height, and shooting straight on). The camera height and angle I’m borrowing from my last shot, where I really liked how it turned out, and used it again with good success. The shots feel portrait-like in a way that I want to keep practicing and utilizing for the future.

The only “errors” in these shots are fixable. The depth of field is still very shallow, even though the lens was not wide open (only a few stops down), and in the second shot, the focus is a little bit behind my face, so I’m a little out of focus. Also, the room is very dark, with only indirect lighting from some incandescent lights. This was a good test for the limits of the Blackmagic cameras’ low light capabilities (they aren’t great) but I think everything still looked passable – it just required some color adjustments.

So, in sum, I think I have a “shot” that I will probably be using a lot in the future, that I know how to set up and execute. (And which makes up most of this scene I plan to put together).

I think maybe the next sub-project I will take on is shooting a conversation scene with audio? It will be a little tricky, because it requires me to do some “acting” but that’s fine. Not something I will be using for music videos, but if I ever think about shooting a short film, it’s something I’ll have to learn how to do.

The initial assignment I think was a little ambitious for me, so I’ve broken it down into learning subparts. I was going to shoot a dialogue scene from an Ozu film, The End of Summer. But unlike the previous assignment, I was really trying to get the spirit of the camera angles, which is so distinctive of Ozu’s films. So, for this first subpart project, I just wanted to shoot basically 3 angles – one with both characters in the scene, and one with each of the characters speaking to the other.

The issue of getting the correct shots for each, however, ended up being an issue with trying to balance focal length and field of view, though I know how to try and attack the issue again next time. Basically, I thought with the close quarters, I would try and shoot the scene with a 35mm instead of a 50mm lens. The wisdom being that I would be able to shoot the scene needed without having to back up so far – where distance from the scene is not in great abundance.

You can see the problem mainly in the second shot, where it was not possible for me to move far enough back – but even there, the closeness of the subject causes some distortion that changes the feel of the shot. While the other scenes were shot with a 35mm, I was able to zoom in on the subject to essentially approximate something closer to a 50mm shot – and because the camera was actually further away from the subject in both shots, it minimized the distortion (especially when zoomed in). So, I actually like the first and third shots in this scene a fair bit.

I think the lesson in all this is to just shoot with a 50mm lens, and to simply place subjects in spots where I have the distance to shoot with the 50mm. This, I guess, is the art of what they call blocking. So I guess in subpart 2 of this project, I’m going to shoot these basic shots again, with the 50mm instead of the 35mm, as much as I would love to shoot with my Lomo LC-A Minitar-1. What I will say, is that I think I’m starting to get a feel for these different lenses and how I want to use them in different situations.

One nice thing about moving to the 50mm is that it allows me a much better F-stop. The Minitar-1 has a lowest F2.8 – though I bumped it up to F5 in these shots in order to widen the depth of field that I see in Ozu’s shots. The Takumar 50mm is an F1.4 which lets a lot more light in. And while the lower F-stop generally means narrower depth of field, because the camera is further away, the main subjects are going to be generally in similar depth-ranges, making it a bit less of a concern.

One other nice thing though, I was able to shoot these scenes at F5 indoors at night, and still ended up with a passable image. Even stopping down the Takumar a few notches, I can still shoot at F2.8, making it easier to get the light I need in these night shots.

Finally, watching this progress of three shots, I started to think about how the characters were introduced into the scene. Without an introduction, the transition from the first scene to the second is confusing, since you don’t have many cues to link the subject in the shot to a specific character in the first scene. (He is wearing a different jacket, but I think you need to give the brain more links, is my feeling). How would I go about addressing that? It got me thinking about the scenes I would put in prior to the first scene, and that got me thinking about how I would set that up, and the transition shots for that. This is all to say, I think I’m starting to develop some of the intuitions I need to put together more complicated short films.

It took me a while to get this put together because of other life events, but I had as a project to try and recreate a scene from a movie, shot-for-shot. In this case, I put together a scene from Bong Joon Ho’s movie, Memories of Murder. It was really a good learning experience, because it got me to really understand both the scene itself, as well as the “logic” of how the scene was put together. It also helped me to start to understand the editing process, as I had to make some decisions on my own, based on the footage that I had taken over a 10-minute period in our dining room.

In this scene, the protagonist expresses a viewpoint – which is posted throughout the movie (I think? its been a while since I’ve seen it) that he has “shaman eyes” and is able to see things that are not immediately obvious. The chief of the police department challenges him on this point, pointing to a pair sitting at a neighboring desk, consisting of a rapist and the brother of the victim, who caught him in the act and dragged him into the precinct. The chief asks the protagonist to identify which of the two was the criminal, which invites the viewer to think about the extent of the protagonist’s powers (and the answer as to whether he has shaman eyes or not is not revealed in this scene).

The scene utilizes a few different camera angles – I had to add another one (the closeup of the police chief, which is a bulldog salt shaker) in order to break up the shots a little bit – as I don’t have an actual scene partner to add an interactive element to the scene. (In the movie, the shot of the protagonist and chief conversing takes up almost half of the scene). But I think I am starting to get how things were put together. There were interesting shot choices – the subtle lean-in of the camera when the dialogue comes closer to the “hook” of the scene, and the cut to the protagonist leaning toward the chief’s test subjects (here, a pair of Japanese pagoda lamps). It may seem obvious, but I understand a little better how much the scene-writing is important, and not just a matter of getting pretty camera angles.

The shots are otherwise very sepia-toned because the room has a lot of incandescent-colored light, which makes it more difficult to color-correct. But not really the focus of the project. But otherwise I think I got a lot out of this. I may try a few more shot-for-shot exercises before maybe taking a scene from a script and figuring out for myself how I’d want to shoot it, or maybe even writing a scene myself from the ground up. This scene was just so cool as an intellectual exercise, I hadn’t really thought about it much until I tried to put it together myself.

After the issues with outdoor shooting last week, I cleaned my lens – in this case a Takumar F1.2 50mm, and did some shooting outside. While the universal glow effect did disappear with a clean lens, there were still some dots that appeared in the image, which you can see below. Given this, I took a look inside my camera and saw that there were some spots. While they initially looked like spots on the sensor, they were in fact spots on the IR-cut filter, which covers the sensor itself. I cleaned those up, and when the IR-cut filter looked clean, I shot some additional footage, some of it straight into the sun, and as you can see, it looks like the image is for all intents and purposes free of spots. There are of course lens flares galore with the Takumar, but that’s perfectly fine. And the last few images that you see out of the camera when everything is clean, look really spectacular I think.

Next week, I think I’m going to tackle a series of projects, where I find a scene from a movie, and shoot something similar to it, just to get a sense of how to set things up for a shot or series of shots. This will, I guess, require a bit of acting, but all the more things that I will learn as a part of increasing my skill set.

The assignment for today was to shoot a transition from outside to inside. I also wanted to include a basic match cut.

I haven’t shot film outside in a while, and wanted to get practice with that, since lighting is very different. Here, I ran into an easily-identifiable problem, because you can see the shadows of dust specks in the shot – something that doesn’t show up when you shoot inside. So, I basically need to clean the lens. Cleaning the lens will also, I suspect, cut down on the blooming glow that you see in the shots that include the sky – which while overcast, is nevertheless a significant source of light.

Because I didn’t want to stay outside for long, I also handheld the camera, and wanted to test the stabilization capabilities internal to DaVinci Resolve, which I’m sure are great, but which did not result in super smooth video, at least not without applying a more severe stabilization algorithm. Not the biggest issue, because I was really whipping the camera around, and I don’t plan to do much handheld video work in the future.

That said, the inside video I think looks really great compared to what I’m used to capturing. I think I also learned an exposure concept, which is something that is very different from the audio world. In the audio world, you don’t want the track to distort because that kind of distortion can easily distract from the general listening experience. But in the video world, distortion is not terrible – there is distortion (clipping) in the sky. You want to shoot so that the object of the viewer’s attention (as you intend) is exposed correctly. I think working in audio world for so long, I always tried to correct video work to avoid any kind of distortion (clipping) at all cost. But I’m less averse to that now, and that leads to better exposure and correction.

The match cut worked just fine I guess. In the “narrative” of the shot, I just used the fiction of an object being found outside, and then viewed inside, as a way to create an occasion for a match cut. The goal is really just for the viewer to connect the dots, which is not a big leap for the viewer, so I think it turned out fine. What was interesting to think about is how becoming comfortable with different types of cuts can themselves be occasions for building narrative. As one becomes comfortable with more kind of cuts, the director has more avenues for story-telling?

Of course, there is a cut-on-action also built into this clip, which also worked out reasonably well, and allows an occasion for the match-cut itself, to connect back to the more general assumed narrative of someone finding a watch outside, and bringing it inside to examine.

My assignment was to put together a short scene that involved several transitions and a dialogue, using the more speeded up process that I used before for color correction. Here, I have maybe 4-5 different shots, the first two cuts are what I now know are called “cut on action”, and the remaining cuts are some kind of dialogue cut, or a POV cut? The way I see it, is that the cut to the bottle works, because it is cutting to what the protagonist is seeing. Alternatively, the cut draws on the viewer’s understanding that they are watching a dialogue.

In either case, the color correction presets I had developed worked, even though this is shot in the daytime, which was great to see work. (I made a slight adjustment in a separate node where I adjusted gain but that was it).

This was also an experiment in a concept that I’ve been thinking through, using subtitles to allow me to vary the dialogue of the actors regardless of what they are saying. (Because I do music video work, it sort of doesn’t matter what they actually say, so if I can write my own dialogue independently, that is always interesting).

I have always had difficulty with color-correction, and have gotten in my head about it. What I decided to do with this video was to use some LUTs to make the process of color correction faster. I suppose a while back I might have felt that color correction was an expressive part of the process of filmmaking. Probably because I used to be a painter, and color correction shares so much psychology with the act of rendering in painting.

But part of developing filmmaking skills involves, I think, practice. And in order to get as much practice as possible, I need to streamline the process. And this involves streamlining color correcting. I am sure I am using all of the language wrong, but I used two LUTs. One prepares the LOG film to look more “regular” (REC709?), and the other lightly adjusts it to look “stylized”. Unlike in the before times, I’m saving the color-correction settings so I can re-use them in all future projects.

In the clip below, I decided to shoot two night shots with minimal indoor light, because I have always had difficulty with that. (And BlackMagic cameras are notoriously not great with low light). I also used two different lenses, because I wanted to see if the Lomo Minitar lens (with I got in the last year) rendered colors that differently my Takumars. It’s not that different! And I think these came out so much better than when I tried to color correct everything from scratch! Took me almost no time from shooting the video to color correction to compiling.

I’ve started this blog, using an old website, to work on my film skills while I’m in between albums. Unlike many indie musicians you might have heard of, I began releasing albums when I didn’t have the money to pay someone to make music videos for me. So, I started making my own music videos as part of the press process for an album release. Over the years, I’ve come to really enjoy the process of making music videos. The format is short enough that it doesn’t beg for an expensive production, but you have a lot of freedom over what it is that you choose to shoot. (Despite the fact that music videos have become a cliche in the modern day).

Nevertheless, there is still a lot more that I can learn from the process, so I decided I might try and give myself something like abstract film school assignments, and see what I can shoot in a given week. I’ll come up with an assignment – many will be very mundane – and see what I end up with. Provide some notes on what I learned, and what worked and what didn’t and go from there. Maybe by the end of this process, I will have expanded my skills for when the next round of music videos need to get made.

Cheers.