Prelude
Well, it’s been a while again, hasn’t it. I am sorry aobut it as usual. I have done a couple of videos. But not that much writing on the website. I am, again. Sorry about it.
The VOP project itself has kind of stalled for a bit. I probably could go further with the AI coding tools I have available. But while I will say that that is not the way I want this project, or any of my projects to go forward at all. I started the VOP project as a way to learn coding. And while I have learned SOME coding. I have not learned enough to really take ownership of the codebase that grew out of the efforts. I have a rough idea of what signals go where with what commands. But it’s too much of a black box for my comfort. I want actual hands on knowledge. I want to basically rebuild things from scratch to see if I can make it work and prove to myself that I actually know what I am doing.
I can’t really do that with the VOP project as it is right now.
What I however CAN do. Is to start a new project from scratch. And I CAN do that without using AI or LLM’s to write the code. And of the projects I have in my head. The PAniCam is the one I want to pursue as it fits most of the boxes I want to tick.
So what is the PAniCam project?
The PAniCam is, in essence a rewind to what the VOP was BEFORE I decided to make it a Video based Optical Printer. I actually started what was later to become the VOP by brainstorming things I could do with a Pi. And one of the thoughts that came into my head was “Animation Camera”.
The PAniCam will be as the name suggests. A Pi based ANimation CAMera. And really. Nothing more. Nothing more at all. I want it to be an emulation of a real animation desk camera to unreasonable lengths. You advance frames. You hit the expose button. The camera exposes. It looks to see if the current frame in the gate already has a previous exposure. And if so adds the new exposure to it. If not. It’s a fresh exposure on virgin frame. It’s basically the LIME system with a frame selector. And the first release will be just that.
What does it NOT do:
- No exposure sheet. That’s up to you as an operator to keep track of.
- No undo’s. If you mess up. Tough luck. Start over or live with it. Heck.
- Not even an onion skin will be considered. Instead. I may just let the user doodle on the preview window with a virtual grease pencil to do visual markings. But only one virtual cel sheet at a time. No auto switching of them.
This project really should be just that. Expose frame. Save it. Merge with existing frame when needed.
Now, OF COURSE. I will add things like exposure controls and white balance. I also want to pipe through live video from the camera when real destructive exposures are not actively being taken. Just to line things up and help in the way a real viewfinder or ground glass would function on a real animation camera.
So on top of the above. What else does it not do that the VOP already does?
- No hdmi projector system - No virtual backgrounds or anything like that. If that’s what you want. the VOP does have the ADM mode that you can look into.
- No optical printing - Unless you actively rig it up in front of a monitor. This thing is designed to capture images of things on a desk.
- No bipacks -
Ok enough of what it’s not. What do I actually want it to do?
- LIME System - The PAniCam will inherit the full Latent Image Multiple Exposure system from the VOP. This is so I can still do in camera composites and effects. It just won’t be able to control anything in front of the camera itself.
- Live Viewfinder - When not actively taking exposures. A live viewfinder feed will show what the camera is seeing so you can set up the next frame.
- Frame Advance forward and backward - To select which frame we’ll expose.
- Expose button - This will do the whole expose cycle. But won’t auto-advance to the next frame. You can use this if you need to add in another exposure.
- Expose and Advance - Does a full expose cycle. And advances the frame in the gate to the next one. For first iterations. I will limit this to single frame advance. If you want to animate on 2s or 3s. Just hit the Expose and Advance button two or three times before moving on.
- Strict one file per frame system - While it is a bit wasteful. I want the PAniCam to be completely unaware of any animations on twos or threes or beyond.
- Test Exposure - This is the one concession I will make for convenience. This button will do the full expose and LIME cycle. But only show it on the preview monitor. Not save it.
- 16bpc TIFF files - Like the VOP. The files in the CamMag are 16bpc tiffs with linear encoding.
- JPEG SideCar files for quick access in GUI - This is something the VOP doesn’t have that I want to experiment with. All TIFF files will get their own JPEG sidecar files so they can be quickly cached and flipped through
- Workprint generation - Like the VOP. I want the ability to generate a quick little h264 file to preview and show people.
- ProRes generation - Aslo like the VOP. I want the ability to make ProRes files that can be pulled into the NLE of one’s choice for editing and sound.
- FlipBook - Using the jpeg sidecar files. I want the ability to quickly flip through the frames in the Cam Mag. Even cache them and play them back.
- Two monitors - A departure here also is that as this is stripped down in terms of what can be done during exposures. I instead want a setup where we have two virtual monitors. One showing the ongoing realtime video feed from the camera. And the other showing the current frame in the camera gate or the output of the flip-book.
- Grease Pencil - Ok. I know that term has somewhat been coopted by the excellent animation and annotation toolset of Blender. But it did get its name from the grease pencils that artists used on glass to mark and plan shots on cameras and CRT screens. I want that. Just a simple annotation tool that can doodle on the screen. One image buffer per reel, though. Not more. This should not devolve into a rotoscoping tool. It’s just to make something that let’s you plan a shot. Maybe I should call it “dry-erase” to get away from the Blender thing? Maybe add a button to move the plastic sheet between the monitors?
- FLASK Web GUI - Like the VOP. I want the GUI to be hosted on the Pi. Accessible through the LAN.
Enough yapping. Where am I in the project?
Actually. I have not actually started, technically. I mean. I have the plan all ready. I even have sourced and had Amazon deliver a fresh new Pi 4B with 2GB RAM. The one thing holding me back at the moment is that I don’t have a camera.
The Pi Camera HQ seems to have stock issues at the moment. But if things go as I want. I can probably start tinkering with this thing to get a first iteration of the webGUI up and running and As soon as I can order the camera. I will.
I know, there are replica cameras. But I want the hardware to be very close to, if not identical, to my VOP prototype.
I have gotten the lower RAM version of the Pi mostly because it was slightly cheaper but also… Even on the VOP. When I check it during full runs. it doesn’t even use 1GB of RAM anyways. So a 2GB RAM model should be just fine for the PAniCam as it’s just streaming the live video and doing image captures. I think I can offload the flipbook to the client side web browser even.
So that’s what I am doing in terms of tinkering projects.
Other projects?
Well, as usual “I have several projects in various stages of development”. But maybe the closest one to start to focus on outside of the PAniCam would be the advent calendar. I did start one last year. But things got in the way so nothing really got published. I do want to see what else I can do with Herr Nicht Werner and his exploits though… hm… but time is short. And all the things I keep thinking of takes so long to do per episode.
Anyways. That’s my blog update here. I may need to return to ammend things in further blogposts.
Maybe even something I can do with pictures?