This is a combined news update regarding both our hardware and software development progress. As you can see, our work with Axiom Alpha prototype is coming along quite nicely. We’re sure you’ll all be happy to hear that raw still image capture has been working well on our testing board for quite some time now. However as we are still finishing the prototype enclosure, we have not yet carried out many other calibrations and tests - especially considering that the image sensor we’ve installed in the only prototype built so far is (for cost purposes) a defect sample better suited for prototyping. A second Axiom Alpha prototype unit containing slightly improved internal components is now being assembled with a full grade image sensor. This second unit will allow us to start more extensive testing soon.
Black lens mount for second prototype unit with full grade image sensor
IT8 Test Chart, raw image captured with Axiom Alpha prototype under incandescent light with some post fixes for dead pixels applied
IT8 Test Chart, with preliminary post calibration for tungsten light
As you will notice, our white balance presets are still a work in progress. If you’re interested and posses the required skillset to work in this area, please contact our development team to assist with fine-tuning.
The following images have been taken from our HDMI output tests:
HDMI signal “ghosting” visible in the images - still some timing bugs to fix
After Debugging - What test images currently look like when coming out of our Axiom Alpha prototype’s HDMI port
What we’re now working on is the FPGA logic to supply the HDMI feed with real-time Full HD image data. HDMI video out from memory is already working, however we experienced a number of delays and problems with the memory interfaces, which did not behave as documented when using both interfaces to feed data into the memory from the image sensor and feed the data from the memory to the HDMI interface simultaneously.
Converted DNG in UFRAW
In parallel, we have started coding a program that converts the raw data captured from the image sensor into a DNG file. Whilst this is already working pretty well, things like additional metadata, camera profiles and hot/dead pixel definitions still need to be implemented or improved.