Tintin in Tibet

Viewing 15 posts - 1 through 15 (of 16 total)
  • Author
    Posts
  • #8191
    SongJiangSongJiang
    Participant

    WARNING: to use this layout, you need a > to 80 version of magicDosBox (with the Pixel Trap function). Version 80 and under will not working. This version could be unreleased at this moment.

    A very great (but very hard) action/platform game, inspired by an awesome comic book. Wikipedia
    This .mgc is just an edit of my generic gamepad .mgc.

    Walk with dPad, run with A, jump with B.
    You can also run by Double tapping right or left arrow key
    Buttons opacity is set to 50% but you can change this by opening the Options menu (swipe from up to down)
    Download .mgc file
    Don’t forget to read the user guide first:
    English Manual
    Manuel en français

    screen

    #8192
    Al exAl ex
    Keymaster

    Very nice, thanks for sharing. ?

    #8193
    adminadmin
    Keymaster

    I own almost all tintin comics

    #8194
    SongJiangSongJiang
    Participant

    I forgot: this layout use the Pixel Trap function for showing level password at the right moment. But should work even with non-beta versions of the app. I hope πŸ™‚ No, it should crash. Wait next version of the app to use this layout.
    And I own all the Tintin’s comics πŸ™‚

    #8197
    Al exAl ex
    Keymaster

    Those were my favourite comics back then, along with Asterix. ?
    Pixel Trap will hopefully be included in the next update. It’s running pretty stable, and it’s such a great feature, incredibly useful. ?

    #8198
    adminadmin
    Keymaster

    It should crash with official release, you must add warning.

    #8199
    SongJiangSongJiang
    Participant

    @Al ex. Yes this feature is just awesome. I secretly wait for it since the first time I use magicdosbox πŸ™‚ The only dark side is the performance sacrifice. It’s difficult to say but I’m pretty sure that before using the beta, Tintin in Tibet ran flawlessly. With the beta I have some slow down one or two time. Nothing unplayable and almost unoticeable, hopefully. My S7 edge is a bit old now πŸ™‚


    @admin
    . Ok. thanks. I was too lazy to test it with the official release XD

    #8203
    Al exAl ex
    Keymaster

    Interesting. I’ll need to do some side-by-side comparison, but on my S9+, I can’t notice any real slowdowns. Something like 1-3%, but that’s almost completely negligible.

    #8218
    SongJiangSongJiang
    Participant

    @Al_ex : I just test Grand Prix 2 (yes, I choose the extreme) : on my S7 edge, the game was playable in vga with almost all graphics options ON, or in SVGA with almost none graphics options ON.
    Now the game is only playable in vga with all graphics options OFF (svga slow down to death when more than two or three cars on screen). So yes performance impact is noticeable, and I guess the more the game is big and your phone is weak, more the impact is noticeable. But Pixel trap is just awesome and I didn’t plan to play GP2, so… don’t care πŸ˜€

    #8220
    adminadmin
    Keymaster

    It must decrease performance from two reasons.

    1. we must run synchronization mode (game thread and video thread are synchronized)
    2. we look for pixels in big array. Let’s say game has resolution 320×200, this is 64000 pixels, each pixel has two bytes. So we search in 128k big array – every frame. This search is executed for every “trap pixel”.

    Both operations are time hungry. Is possible skip point 1 and run traps without synchronization mode, but my rough guess is around 5-8% failure if we set one pixel in pixel trap and maybe 50% failure if we set two or more pixels.

    #8221
    Al exAl ex
    Keymaster

    That makes sense. SVGA with higher pixel count is more demanding. I’ll run some Quake SVGA benchmarks, and see if there’s a difference (= start Quake, set res to 640×480, them run “timedemo demo3” from console with and without Pixel Trap enabled).

    #8222
    Al exAl ex
    Keymaster

    So, I ran the benchmark three times for both configs (Pixel Trap ON & OFF):
    Pixel Trap enabled = 13.5 fps / 11.7 fps / 11.1 fps
    Pixel Trap disabled = 13.6 fps / 12.7 fps / 9.9 fps

    Interestingly, while the best result for both was basically identical, I got the worst result by far running the timedemo without pixel traps. But that was the last benchmark run I did, so all the lower values are obviously just a result of cpu heating. I kept running those benchmarks in rapid succession, and that really squeezed the crap out of my phone. ?

    If you want to do some more in-depth benchmarking, you can also try the DOS benchmark suite from Phil’s Computer Lab: https://www.philscomputerlab.com/dos-benchmark-pack.html

    #8224
    adminadmin
    Keymaster

    Maybe I can speed up point 2, I must do some caching, but I think I can speed it up, but first we should fix all bugs around.

    #8225
    Al exAl ex
    Keymaster

    I can do some more testing, still got my S7 edge here. @zebumper is yours an Exynos or a Snapdragon phone?

    #8226
    adminadmin
    Keymaster

    Is not necessary do some special tests, waiting for another thread is biggest bottleneck

Viewing 15 posts - 1 through 15 (of 16 total)
  • You must be logged in to reply to this topic.