Hacker News Reader

Comment by metaphor | Hacker News Reader

@walterbell submitted this, so intuition tells me I should be paying attention, but all I read is someone struggling with single-link digital video implementation on a FPGA and a sponsored advertisement for a tool that doesn't quite make sense in context.

I think I may have had one too many drinks tonight. Did I miss something?

View on HN

HDMI capture on an FPGA board that does not have an HDMI input. Over the six-part series of blog posts, the author explores the 2015 limits of FPGA hardware performance and IP segmentation for HDMI processing around $150 price point, e.g. http://zedboard.org. Other video projects with less connector hacking:

$250 Spartan-6 HDMI Light: https://news.ycombinator.com/item?id=18216910

$350 Numato Opsis Spartan-6 HDMI2USB (2015): https://blog.mithis.net/archives/category/timvideos-us/hdmi2... & https://github.com/timvideos/HDMI2USB-numato-opsis-sample-co...

$500 Artix-7 Digilent Nexys, https://store.digilentinc.com/nexys-video-artix-7-fpga-train...

$900 Zynq MPSoC (Arm Cortex A53 + Xilinx FPGA) with H.264 video encoder, https://www.xilinx.com/products/boards-and-kits/zcu104.html

Author apparently ran into issues while developing hardware with hdmi output, found out that proper logic analyzers capable of wire-speed capture and analysis of non-video-related data (=transmissions during screen blanking time) are expensive and decided to roll with an fpga based solution as well as document it and how HDMI works.

Quite interesting, especially for those of us designing e.g. host boards for RPi Compute Modules.

This is the introduction page, follow the links at the bottom for more, there is a lot of interesting detail including things like writing a kernel driver to interface with the HDMI receiver.

Reply on news.ycombinator.com