Submitted by Sebastian on Thu, 05/23/2013 - 11:21
apertus° is opening up all the internal discussion about the Axiom Alpha prototype to the public and is currently seeking additional hardware developers. Read on to learn why and how.
On his blog, Philip Bloom recently wrote: “Please stop bringing out new cameras and stop announcing them. Just go back into your top secret bunkers and concentrate on refining and perfecting your current ones and ideas for your future ones. [...] I am a big fan of announcing your new product on the same day as you can order it”.
So what does it all this have to do with apertus°? Well, it was a reminder for us to again push in the absolute opposite direction. While we understand the desire for everyone to acquire and use new gear right away, this mostly relates to the classic commercial world where announcements are made by marketing divisions and the actual development happens behind tightly closed doors.
But apertus° is very different.
First Axiom is much more than just “a product”. It’s an open camera platform - ready to evolve into things we haven’t even dreamt of today. Many people are currently afraid of buying a camera because of the speed at which new “products” are now being released, making a lot of devices very quickly outdated and obsolete. We want to stress that Axiom will not be outdated anytime soon - it will be constantly evolving - as we rely on FPGA based designs rather than fixed “set-in-stone systems”. We and the entire community will deliver new interfaces, codecs, firmware and significant degrees of newer functionalities with a simple software download to the camera. It is our intention to use this to create a very powerful ecosystem around Axiom, providing support alongside the development of new features.
But Axiom will also create an open knowledge library: the result of everything we have learned on the path to creating it - freely open for everyone's disposal.
With Axiom we decided to keep a lot of stuff to ourselves for now which we explained here. We still think that was and is the right decision to not disclose the full plan yet - as technology changed quite a bit in certain areas that are very relevant to Axiom. Now as time has passed we will likely be able to do some things with technology we did not think we would able to a year ago now. But as we stated ourselves in that article, while we see it’s necessary, we actually hate not being able to share everything - so with the Axiom Alpha prototype development we have decided to transition to a more open approach: Sharing the good and bad times - we know that some people will find it discouraging to read about obstacles, problems, delays in addition to progress, milestones and success but this is what happens in real life. We think that everyone can learn as much from problems as they can from success - probably even a bit more.
That's why we decided to open up all the internal discussion about Axiom Alpha on the internal apertus° forums and move them to the public part of the forums: Full Axiom Alpha Progress Forum Thread
.Notice: The following paragraphs are rather technical.
The initial plan was to build a custom image sensor board hosting an FPGA and some interfaces (HD-SDI, USB, etc.). Basically we would have built the entire prototype from scratch. A huge effort and while we had a bit of money from sponsors and partners we were not sure it would be sufficient to actually complete this prototype.
Then the Xilinx Zynq (Series 7 FPGA SOC) started to ship around Christmas 2012 and we reevaluated our options. The idea was to use an FPGA evaluation board which already features most of the connectors/interfaces and the accompanying hardware and software drivers for them. This suddenly meant we could build upon an existing foundation rather than reinventing the wheel and starting from scratch. We were thinking of starting off with a Xilinx Zynq evaluation board for $ 1,500 USD, however we then discovered the Zedboard which is pretty much the same thing spec-wise but costing only $400 USD. Both evaluation boards feature a so-called FMC connector (a high density PCB to PCB connector) that can help us reduce the amount of work required for a functioning prototype down to “only” designing a PCB hosting the image sensor. The lens mount is already finished as you can see on the Axiom Alpha minisite.
We are now putting all the money the apertus° company raised into creating this image sensor front-end PCB and to make it act together with the Zedboard. This is Axiom Alpha.
But since Axiom will use the same FPGA SOC and image sensor all the work will be mostly reusable for the final Axiom camera. So nothing we do now has to be thrown away further down the track.
With the working Axiom Alpha prototype outputting an uncompressed Full HD video stream from the Super35 4K sensor, we will then shoot sample footage and start to re-evaluate our final Axiom plans and concepts. When we reach this point, we’ll arrange everything online and present the complete plan to the public, all shortly before we launch the crowdfunding campaign.
With money we can speed up the development process by prioritizing apertus° work for our developers, paying them instead of asking them to work for us in their sparse spare time. You can donate to apertus° here.
Alternatively people can donate their skill and worktime. We currently seek additional hardware developers (for helping us design a high frequency circuit PCB or working with embedded systems and FPGA/VHDL/Verilog). Even if you yourself don’t have such skills, maybe you know someone who could be interested? Spread the word!
10 Comments
I want to participate.
Hi,
I want to participate as a hobbyist VHDL for FPGA design and as hobbyist MCU developer for Dictator (with Teensy 3.0) but I couldn't found infos about VHDL/FPGS side. For Dictator I already found here https://www.apertus.org/dictator.
Thank you
Great! We are already in
Great! We are already in email contact.
Regards Sebastian
HDSDI
http://solutions.inrevium.com/products/fmc/serial_connectivity/index.html
"First Axiom is much more
"First Axiom is much more than just “a product”. It’s an open camera platform"
Thats why i like axiom :)
ZedBoard
Cool project. The ZedBoard only has an LPC FMC (Low Pin Count FMC). The CMOSIS sensor needs 64 LVDS channels to transfer data at full framerate. Can an LPC FMC connector handle that? LPC FMC on the ZedBoard has 68 lines that can be configured as 34 differential pairs. Won't the ZedBoard fall short (you need a high density FMC connector)?
Correct, we will only reach
Correct, we will only reach half performance with this LPC FMC from the sensor. Half performance still means 75FPS at 4096x3072 and we only need 25FPS for the prototype so thats more than enough for now. For the full Axiom we will use all LVDS lanes of course.
participation
Hi
Is the project alive ? I have skills in high-speed FPGA design, I designed several boards that are available at OHWR, recently I and my colleague managed to connect CMV4000 to the ZYNQ and implemented RSTP streaming over Ethernet.
I think our experience could help your project, I am also very interested in hi resolution cameras.
Regards,
Greg
Sounds great, any help would
Sounds great, any help would be greatly appreciated, email is on its way :)
Wher is the shutter located
Wher is the shutter located in the camera? I saw the pictures and saw no shutter. The shutter would be in between the sensor and the lens. I did not see a shutter.
There is no mechanical
There is no mechanical shutter.
Add new comment