Personal View site logo
Z-CAM E1 Camera 100 Mbps hack for 4K DCI (fw v0.31)
  • 255 Replies sorted by
  • Temperature issue can be solved by removing outside plastic box, just keep the internal circuit and an external fan.. They did not design for airflow so that could be the issue.

  • @zcream

    Well, you can see our Z1 internals photo in main topic, it really has lot of shortcoming sin thermal design :-).

    I think few holes and small heatsink already can do a lot.

  • @zcream, ok. I think, it's not a temperature problem. Because A9 is a SoC with ARM as a main CPU. I mean if temperature interruption comes, ARM will receive it (not DSP)

  • @Vitaliy_Kiselev, is it possible to get info about DSP assertion codes?

  • at 4k30 and 100mbs1gop hack the temperature displayed in the ios app rises to 41 degrees and remains stable there, at 1080p30 the temperature is lowest 35-38C , at 1080p60 it is almost back to 40C, this comes from the way the sensor is read out. The sensor is cooled over the front and the bottom plate, I taped the hoset and the ground with 1/4 thread on a large file plan and screwed with thermal grease on a cage for cooling . the chipset sits behind the screen, which seems to produce less heat

  • Does anybody need a E1 fw unpacker/packer?

  • @storyboardcreativity

    Make sure to make it public. And also look for E2 firmware :-) may be some knowledge can be used.

  • @Vitaliy_Kiselev, I still don't have a E2. I started reverse-engineering E1 firmware because I wanted this camera and I saw that 4K24P 60 Mb/s is too low for this camera. It was just interesting for me. But I don't have E2 because it's too expensive for me - maybe later, when I can buy it with ~$400-$600...

    And what do you need to edit in E2 firmware? Is not it good?

  • If you can find an older E2 firmware package, they didn't bother with any sort of encryption. You can just unzip it and look around. The newer ones use zip encryption - so it's not entirely unbreakable, but much more time-consuming if you care to do it.

    The E2 firmware is mostly just a ucLinux OS. The camera-specific stuff (mostly) is in a single executable file. I haven't seen any reason to mess with it, though, since the current quality/features are already fine.

  • The newer ones use zip encryption - so it's not entirely unbreakable, but much more time-consuming if you care to do it.

    Well, due to incremental nature of firmware it is easy breakable :-) As you can just use older, latest firmware before encryption to find proper keys.

    I haven't seen any reason to mess with it, though, since the current quality/features are already fine.

    Usually such things have few small surprises :-) May be some stuff stuff can be just disabled.

  • I actually looked around the older firmware before encryption and couldn't find anything resembling a key (which is not to say it doesn't exist). It may actually be in a part of the camera not changed by firmware (which would mean breaking it would be pretty much permanent).

    If it's just spelunking for features, though, I am already able to do that without breaking the firmware. There's another backdoor in. Is there one in particular that you're interested in?

  • @Vitaliy_Kiselev, what about DSP assertion codes?

  • Hey, guys! New info about E1 firmware internals: when you choose "720P 240 FPS" mode, actual readout from sensor is 1536x288! Than frame is interpolated to 1280x720 and encoded. So, it looks like a foul play from Z Cam :)

  • The sensor has extra a 3x3 readout mode?

  • I got my E1 with the Izugar Fw 0.28   after the update to the then official 0.29 fw a few features disappeared company policy?

  • "720P 240 FPS" mode, actual readout from sensor is 1536x288! Than frame is interpolated to 1280x720 and encoded.

    No wonder it looks so bad, thanks for the info - little point in bothering with 240fps on that camera :-(

  • @rjhinde, I wrote a letter to Z Cam with some questions about slow motion on E1. It will be better if it's possible to read only top sensor part (still 1536x288, but without subsampling). It can make quality much better.

  • This is the actual resolution of the Sony RX100 VI when it shoots 240 fps that is nominally 1080P: "240fps/250fps (1,824x1,026) ." Close, but not there. (480fps/500fps (1,824x616)). So, they all cheat, although Sony makes its real specs available.

  • Hi! New here. Thank you so much for this!!

    Can anyone tell me if this affects the HDMI output at all?

    I use my E1 connected to a capture card on my PC so higher bitrate HDMI out would be amazing!

  • @BerryJP

    Camera does not compress HDMI signal (except may be for color sampling) :-)

  • @ Vitaliy_Kiselev

    I see! Thank you.

    Do you know the statistics for the HDMI out? Is it 10bit from the sensor or 8bit from the chip? Could I use a 4:2:2 external recorder?

    Sorry for the silly question.

  • I need some help from E1 users. Do you have any problems with HDMI out? On my 0.31 version I have frozen picture on LCD when HDMI is connected and after some idle time HDMI out signal is frozen too. Is it ok? :D

  • Today got Blackmagic HyperDesk Shuttle 2 and... it's a fail. Green screen. Any ideas?

    yRgQ2v-gZCo.jpg
    2080 x 1560 - 490K
    v0XCxQu_4Ps.jpg
    2080 x 1560 - 394K
  • GUYS!!! GREEN SCREEN IS TOTALLY FIXED! Blackmagic Shuttle 2 wants YUV 4:4:4 signal out from camera's HDMI, but in E1 it's forced to RGB. I've made it to use YUV 4:4:4 through HDMI - result you can see in pictures below. Some little things left, like make sure it's real 10-bit color. I'll upload fix to my repository: http://github.com/storyboardcreativity/Z-Camera-E1-Reverse-Engineering

    oX4bAzmLQyI.jpg
    1776 x 1082 - 384K
    FqVwPVil-3Y.jpg
    2080 x 1560 - 476K