The Debate
Building a camera is/isn't easy.
Two basic concepts:
1. A philosophy of a constructive mindset to build anything, without regard to challenges present. "While it's challenging it's quite fun and that makes it worthwhile- wait till you see what's coming next!" vs. "It's not easy and progress is slow, maybe you'll see a firmware update next year, maybe not. Sorry for the delay".
2. What makes a great camera or product?
Let's start with Red (since I believe they own this forum?). Out of the gate wild, positive claims of amazing performance. Years later, while it's still over-complicated and has other issues (e.g. not quite reliable enough yet), many people love that camera system and no one can argue that it doesn't produce fantastic quality imagery now, even for skin tones (the most challenging aspect of color science).
Black Magic- similar story at a much lower price point (starting at $3k).
Z Cam- they're just getting started, and at an even lower price point ($2K).
See the pattern- as DLD noted next will be around $1K (and it won't stop there).
What is a camera?
Sensor + computer (hardware + software) + storage + power, and minimally a button to start/stop recording. A lens isn't even needed, a pin hole works. Power can be switched by plug/unplug battery (anyone remember the NanoFlash haha).
Right now I (and anyone with the right software and hardware skills) can purchase an Ambarella A9 development kit: https://www.ambarella.com/uploads/docs/A9-product-brief.pdf , the sensor of their choice (up to 4K), a battery, lens mount, lens, literally some cardboard and duct tape (a la Google cardboard for VR), a switch/button and an SDXC memory card (and a few more minor items) and can build a real, working camera in just a few hours total working time.
Is that easy? From my point of view that's super easy. You can customize it to your heart's content to make it better. Imagine Sony selling you a Venice sensor (provided it can interface with that board), you work with a friend who has a 3D printer and/or has access to a vacuum+oven to make carbon fiber components, and the body can be super sexy, strong, and lightweight, any form factor you can image. Next you can add a simple LCD for settings. What do you really minimally need? ISO, WB, shutter, framerate (start with all manual lenses). Add a WiFi/Bluetooth chip and you can control with iOS/Android and full touch-screen GUI, remotely. Over time you can develop it into exactly the perfect camera. For YOU!
This is the future of cameras. Just like with PCs, it will be possible to use COTS (common off the shelf components) to build an ultra powerful, completely customized camera for anyone's needs. This pattern also happened with cars. In the beginning most people bought factory cars (vs. building them), later people began to modify them, and now we've got many small companies (and some getting bigger: Tesla) building cars that would never be possible if only the 'big boys' were in charge (and the big boys have, and still are, providing a lot of friction to slow down the little guys).
EDIT: already possible to build a tiny modular camera with an HDMI camera+lens ( e.g. https://www.bhphotovideo.com/c/product/1382660-REG/aida_imaging_uhd_100_uhd_4k_30_hdmi_1_4.html ) and e.g. Atomos recorder. For pro use external monitors are already common. Future pro camera with IBIS, usable AF (gimbal, or working fast- docu work etc.) and monitor+recorder and there's your custom camera in a tiny package. Swap out the sensor over time as need or per shot / available lenses. Looks like full frame sensors may be hard to source by themselves for a while. Camera can be controlled via HDMI (technically possible- can be done in the future). SDI is of course better: https://www.bhphotovideo.com/c/prod...dmft_uhd_mr_micro_studio_c inema_camera.html . That market will continue to grow- look forward to seeing S35, FF, phase detect AF, IBIS, etc. That's the future for those needing highly configurable camera systems (pro use etc.): tiny camera block+lens mount and you can use any recorder/display or VF per your custom, personal needs. All in a tiny, lightweight package. Sure, BM has work to do re: design and quality, would love to see Sony etc. develop something like this (Canon has a similar camera though not priced in the same ballpark: https://www.bhphotovideo.com/c/prod..._me200s_sh_multi_purpose_camera_cine ma.html ).
Most pros could work with this camera's quality (that's the closest I could find to their Micro Studio Camera 4k- similar tech/look?):
USB 3.1/Lightning is fast enough to send compressed Bayer to a phone (4K and more). Phones with GPU and AI hardware have plenty of power to do some amazing processing, today (even more tomorrow). So building a custom camera with a killer full frame sensor package with IBIS with an EF mount (or m43 take your pick) which can plug into iOS or Android would make building customized cameras ultra easy. This could create a new market for sensor vendors, as well as folks providing cases/cages to hold the phone and sensor block + lens. People can write their own camera control software for iOS/Android, or purchase one from the store (if this were to become a market). Just like plugging a USB camera into a PC/Laptop and using software to capture/control.
EDIT: examples of USB 3.1 cameras currently on the market, with On and Sony sensors: https://www.lumenera.com/products/high-performance-usb-3.html
On Semi full frame 35mm 6.5K global shutter sensor and active EF mount! Too bad FR is a bit low (perhaps sending Bayer/Raw to a phone would be possible at 24fps?): https://www.lumenera.com/products/high-performance-usb-3/lt29059h.html
A great camera is very personal to one's needs and point of view. The opportunity to custom build a camera for any desired purpose means soon everyone can have their perfect camera if they want to do the extra effort to build one. The market for pre-made cameras isn't going anywhere for the mass market wanting something good enough with the least hassle (and increasingly built into phones). Using an external sensor + lens with a phone seems like the best bang-for-buck in the near future, curious why no one's done that yet.
That's my point of view for taking the position that building a camera is easy.
Is it "cheating" to build cameras this way? Depends on your point of view. Are all the divisions of large camera companies really a lot like many small companies working under the same name? Is ARRI cheating using On Semiconductor sensors, Xilinx (most likely) FPGAs and a myriad of other IC and electronic components, none if which they themselves manufacture? The Japanese camera cartel cameras are produced mostly in China!
Making a great camera or any product takes iteration. The rate of iteration depends on many factors, and reducing friction increases the convergence rate to what a market would call a great product. It's starts with consciousness and how leadership thinks and leads their teams. Freely available technology increases speed (a lubricant) and patents, politics / blocking tactics from competitors, and non-constructive team thinking reduces speed (friction).
Building a camera is/isn't easy.
Two basic concepts:
1. A philosophy of a constructive mindset to build anything, without regard to challenges present. "While it's challenging it's quite fun and that makes it worthwhile- wait till you see what's coming next!" vs. "It's not easy and progress is slow, maybe you'll see a firmware update next year, maybe not. Sorry for the delay".
2. What makes a great camera or product?
Let's start with Red (since I believe they own this forum?). Out of the gate wild, positive claims of amazing performance. Years later, while it's still over-complicated and has other issues (e.g. not quite reliable enough yet), many people love that camera system and no one can argue that it doesn't produce fantastic quality imagery now, even for skin tones (the most challenging aspect of color science).
Black Magic- similar story at a much lower price point (starting at $3k).
Z Cam- they're just getting started, and at an even lower price point ($2K).
See the pattern- as DLD noted next will be around $1K (and it won't stop there).
What is a camera?
Sensor + computer (hardware + software) + storage + power, and minimally a button to start/stop recording. A lens isn't even needed, a pin hole works. Power can be switched by plug/unplug battery (anyone remember the NanoFlash haha).
Right now I (and anyone with the right software and hardware skills) can purchase an Ambarella A9 development kit: https://www.ambarella.com/uploads/docs/A9-product-brief.pdf , the sensor of their choice (up to 4K), a battery, lens mount, lens, literally some cardboard and duct tape (a la Google cardboard for VR), a switch/button and an SDXC memory card (and a few more minor items) and can build a real, working camera in just a few hours total working time.
Is that easy? From my point of view that's super easy. You can customize it to your heart's content to make it better. Imagine Sony selling you a Venice sensor (provided it can interface with that board), you work with a friend who has a 3D printer and/or has access to a vacuum+oven to make carbon fiber components, and the body can be super sexy, strong, and lightweight, any form factor you can image. Next you can add a simple LCD for settings. What do you really minimally need? ISO, WB, shutter, framerate (start with all manual lenses). Add a WiFi/Bluetooth chip and you can control with iOS/Android and full touch-screen GUI, remotely. Over time you can develop it into exactly the perfect camera. For YOU!
This is the future of cameras. Just like with PCs, it will be possible to use COTS (common off the shelf components) to build an ultra powerful, completely customized camera for anyone's needs. This pattern also happened with cars. In the beginning most people bought factory cars (vs. building them), later people began to modify them, and now we've got many small companies (and some getting bigger: Tesla) building cars that would never be possible if only the 'big boys' were in charge (and the big boys have, and still are, providing a lot of friction to slow down the little guys).
EDIT: already possible to build a tiny modular camera with an HDMI camera+lens ( e.g. https://www.bhphotovideo.com/c/product/1382660-REG/aida_imaging_uhd_100_uhd_4k_30_hdmi_1_4.html ) and e.g. Atomos recorder. For pro use external monitors are already common. Future pro camera with IBIS, usable AF (gimbal, or working fast- docu work etc.) and monitor+recorder and there's your custom camera in a tiny package. Swap out the sensor over time as need or per shot / available lenses. Looks like full frame sensors may be hard to source by themselves for a while. Camera can be controlled via HDMI (technically possible- can be done in the future). SDI is of course better: https://www.bhphotovideo.com/c/prod...dmft_uhd_mr_micro_studio_c inema_camera.html . That market will continue to grow- look forward to seeing S35, FF, phase detect AF, IBIS, etc. That's the future for those needing highly configurable camera systems (pro use etc.): tiny camera block+lens mount and you can use any recorder/display or VF per your custom, personal needs. All in a tiny, lightweight package. Sure, BM has work to do re: design and quality, would love to see Sony etc. develop something like this (Canon has a similar camera though not priced in the same ballpark: https://www.bhphotovideo.com/c/prod..._me200s_sh_multi_purpose_camera_cine ma.html ).
Most pros could work with this camera's quality (that's the closest I could find to their Micro Studio Camera 4k- similar tech/look?):
USB 3.1/Lightning is fast enough to send compressed Bayer to a phone (4K and more). Phones with GPU and AI hardware have plenty of power to do some amazing processing, today (even more tomorrow). So building a custom camera with a killer full frame sensor package with IBIS with an EF mount (or m43 take your pick) which can plug into iOS or Android would make building customized cameras ultra easy. This could create a new market for sensor vendors, as well as folks providing cases/cages to hold the phone and sensor block + lens. People can write their own camera control software for iOS/Android, or purchase one from the store (if this were to become a market). Just like plugging a USB camera into a PC/Laptop and using software to capture/control.
EDIT: examples of USB 3.1 cameras currently on the market, with On and Sony sensors: https://www.lumenera.com/products/high-performance-usb-3.html
On Semi full frame 35mm 6.5K global shutter sensor and active EF mount! Too bad FR is a bit low (perhaps sending Bayer/Raw to a phone would be possible at 24fps?): https://www.lumenera.com/products/high-performance-usb-3/lt29059h.html
A great camera is very personal to one's needs and point of view. The opportunity to custom build a camera for any desired purpose means soon everyone can have their perfect camera if they want to do the extra effort to build one. The market for pre-made cameras isn't going anywhere for the mass market wanting something good enough with the least hassle (and increasingly built into phones). Using an external sensor + lens with a phone seems like the best bang-for-buck in the near future, curious why no one's done that yet.
That's my point of view for taking the position that building a camera is easy.
Is it "cheating" to build cameras this way? Depends on your point of view. Are all the divisions of large camera companies really a lot like many small companies working under the same name? Is ARRI cheating using On Semiconductor sensors, Xilinx (most likely) FPGAs and a myriad of other IC and electronic components, none if which they themselves manufacture? The Japanese camera cartel cameras are produced mostly in China!
Making a great camera or any product takes iteration. The rate of iteration depends on many factors, and reducing friction increases the convergence rate to what a market would call a great product. It's starts with consciousness and how leadership thinks and leads their teams. Freely available technology increases speed (a lubricant) and patents, politics / blocking tactics from competitors, and non-constructive team thinking reduces speed (friction).
Last edited:
).