Some digging around indicate this dashcam is likely powered by Hi3516 V300. Datasheet indicates this SoC released in late 2016 has one ARM926 core running at 800 MHz, hardware encoder for H.265 for up to 1080p45 and 30 Mbps output bit-rate. Not much else interesting about it aside from the ARM926 core is outdated as of 2006 and code-compatible with ARM7 designs. It is however more efficient despite longer pipeline (5 stage vs 3 stage). This SoC is targeted at IP-CCTV applications.
Now something is curious and could explain why the level of detail in the output video isn't so great. The hardware encoder seems to have a limit of 30 Mbps (meaning 3.75 MB/sec). So why is it the files are 10.5 MB/sec? Did the CPU just pad the files to make it seem like it's going to contain a decent amount of details? That would be quite sad because the IMX307 sensor is capable of delivering more details than the encoder can cope with.
In comparison:
DDPai Mini V2 with IMX323 sensor (not Mini 2) uses Hi3518 V200. Hardware encoder is limited to H.264 2MP 30 fps but can output up to 100 Mbps.
DDPai / 9Eagles A2 uses Hi3516 V100. Hardware encoder is limited to 30 Mbps. I'll check it out when I receive this cam, but I expect them to pad it up to 10.5 MB/sec anyway
DDPai Mola Z5 uses Hi3516 V300 as well and would likely have the same limitations. (not supercap! 3.7v 240 mA battery inside)
DDPai Mola N3 uses Hi3556 V200. Hardware encoder is limited to 100 Mbps (12.5 MB/sec) but produced 14.5 MB/sec files.
DDPai Mini 3 uses Hi3556 V100. Hardware encoder is limited to 100 Mbps.
DDPai Mini 5 and X2S Pro uses Hi3559 V200. Hardware encoder is limited to 100 Mbps. Now I'm questioning if 100 Mbps can cope with 8MP/4K sensor in the case of Mini 5, or with 4MP (OS04B10)+2MP sensor in the X2S Pro.
PS: The X2S rear sensor seems to have its own SoC. So yeah the Hi3559 V200 is probably OK to deal with OS04B10 in the front.