1. 93
  1. 9

    I love everything about this. What a great solution. :-)

    1. 4

      Yep it’s really nice somehow to see these amazing results!

      Sad thing is since an hour of an average engineers time is worth more on the open market then this TV is, so rationality would say that these efforts are rarely worth the time spent, but yet we do it, there’s a less tangible recreational value in it.

      I’ve got an 10 year old flaky tv myself, it’s probably not worth anything but it’s good enough for me. Lately it’s had problems with powering on properly and a friend using the same model had the same problem, we fixed both our tv’s by screwing them appart and surely finding the same cheap capacitor building like there’s no tomorrow, procuring new replacements and replacing it (and the rest of them while at it) combined with time spent unmounting it from the wall etc must have largely out-weighted both it’s value and the price of a new tv, yet there’s a rewarding feeling in fixing it and knowing that I’ll be able to use it a couple of years longer.

      1. 7

        If the time did not displace working time, then there’s no loss of income for you from doing this.

        Electronics are also likely priced too cheaply because environmental and labour costs are discounted through poor living standards for the workers and lax environmental regulation. If we lived in a world with a flatter income distribution and better environmental controls then we’d probably reuse, repair and recycle a lot better.

      2. 2

        I want to have friends doing stuff like this!

        1. 0

          For me, it’s more at the level of “not bad”. Here’s the disappointing bit:

          It would be nice to apply the corrective filter to the whole screen instead of just a video playing in an application, but I couldn’t think of a way to do it.

          If the author had managed to get the filter into the TV’s firmware or something like that, I would be truly impressed.

          1. 4

            It would be indeed sweet to have the correction running all the time. An FPGA devboard with two HDMI ports could be a realistic solution here.

            However I can’t even begin to imagine the toll such hardware hacking would take on my free time. Sometimes a “80% there” solution is good enough.

        2. 1

          As the first figure under “Finding the inverse” is a relatively simple closed-loop system, wouldn’t a typical control theory technique have worked here, without having to manually mess with that one image?

          (EDIT: yes, sorry, I’m too lazy to properly work something out, I know.)