User Tag List

Page 6 of 25 FirstFirst ... 4 5 6 7 8 16 ... LastLast
Results 51 to 60 of 243

Thread: Fusion 2.5 Microstuttering (moved from Fusion 3 status thread)

  1. #51
    Clicker Fusion 2.5 DeveloperAndroid Export ModuleHTML5 Export ModuleiOS Export Module
    Fusion 2.5 (Steam)
    schrodinger's Avatar
    Join Date
    Nov 2014
    Posts
    3,155
    Mentioned
    27 Post(s)
    Tagged
    1 Thread(s)
    @ Prominent: I seemed to notice the same, though I'm not very sensitive to the issue so was not sure about it.
    We could try making it flash every length2 non-0.016 sequence (just change sequence length >=3 in line 7 to >=2),
    I had tried but that seemed too much flashy to me, and maybe the flash could be a bit misleading for the eye too..

    Also, while tinkering, I noticed a strange thing:
    setting a different framerate in the application properties (say, 40)
    and then forcing a framerate of 60 at start of frame (so it should be the same as the basic test mfa)
    gives very different and unexpected results, with extremely long non-0.016 sequences.
    And the ball seems to be more jerky to my eyes.
    Not a clue what this could mean though...

    Was thinking we could record every relevant data possible (like in those test mfas and more)
    and then have some good eyed (and good willed ) clickers press "spacebar" when detecting a stutter,
    then look back if there was any noticeable pattern of some kind in data recorded, near the spacebar pressed.

    Same here for salamanderpants video: I can see stutter very well there,
    and recognize having seen the same other times while tinkering with Fusion,
    though running that mfa isn't "almost" stuttery at all here,
    not sure if because of my machine, my monitor, or my eyes
    it's a very evasive issue


    I wonder how much would be difficult testing Phi's intuition, I guess only devs could tell.

  2. #52
    Clicker Fusion 2.5 Developer

    Join Date
    Feb 2014
    Posts
    976
    Mentioned
    16 Post(s)
    Tagged
    1 Thread(s)
    another nice example schrodinger

    the ball is flashing like crazy on mine. & it still also flashes in full screen mode.

  3. #53
    Clicker Fusion 2.5 DeveloperAndroid Export ModuleHTML5 Export ModuleiOS Export Module
    Fusion 2.5 (Steam)
    schrodinger's Avatar
    Join Date
    Nov 2014
    Posts
    3,155
    Mentioned
    27 Post(s)
    Tagged
    1 Thread(s)
    does your perfectly synced long-time-player eyes seem to recognize a match between the stutter and the flash of the ball?

    trying to understand when exactly a stutter is perceived,
    if it's a sequence of "not-homogeneous" frame durations or we should look elsewhere...?


    edit: do you notice stutters when the ball is not flashing /
    and if so,
    changing >= 3 to >=2 in event line 7 makes any difference?
    (though it might be still more difficult to get then, with the ball flashing all the time...)

  4. #54
    Clicker Fusion 2.5 Developer

    Join Date
    Feb 2014
    Posts
    976
    Mentioned
    16 Post(s)
    Tagged
    1 Thread(s)
    putting my " perfectly synced long-time-player eyes " to the test

    its hard to see coz the visual shock of the flash makes it seem like a jerk lol, i'm gonna make something else flash next to it instead so i can monitor it better.

  5. #55
    Clicker Fusion 2.5 Developer

    Join Date
    Feb 2014
    Posts
    976
    Mentioned
    16 Post(s)
    Tagged
    1 Thread(s)
    made a ball flash underneath instead & removed the background as it visually makes it harder to see.

    in this example i can clearly see the juddering & also prominent is correct, the ball does not flash with every stutter.

    & it still happens even wen changing the number to >=2

    i think most shud be able to spot it in this example
    Attached files Attached files

  6. #56
    Clicker Fusion 2.5 (Steam)

    Join Date
    Apr 2017
    Posts
    32
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    What I would do is spawn a vertical line at the position of the ball each frame (so you have a series of new lines as the ball moves across the screen), and based on whatever your method to calculate the stutter, change the new line's color to signify the discrepancy.. Then you can just look at the series of lines and check where the color changes and if the distances are any different from the other lines.
    This assumes your method of catching the stutter is correct. But it also could serve as a way to find the correct method if it isn't.

  7. #57
    Clicker Fusion 2.5 DeveloperFusion 2.5+ DLCiOS Export ModuleInstall Creator Pro
    Julian82's Avatar
    Join Date
    May 2012
    Location
    outbuddies.com
    Posts
    943
    Mentioned
    47 Post(s)
    Tagged
    0 Thread(s)
    I put Schrodinger's flashy ball (sounds a bit like Schrodinger's cat, doesn't it ) in my really heavy project (Outbuddies) cuz I was curious. Thanks for making this, it's an excellent tool to check how much your game micro-stutters and especially handy if you're not sensitive to it (like me). The strange thing is Outbuddies does microstutter, but significantly less than your example (res adjusted). Tested on 60FPS and VSync on. This is very strange I think, I would have expected it the other way round.

    Also made some tests with my game with the following findings:
    x Lower resolutions stutter less (even when scaled up). I found this not true in Schrodinger's example however, it always stuttered like hell on my system.
    x Disabling the text overlays did not reduce stutter for me, both in my game and in the example
    x Smaller rather empty rooms in my game seemed to stutter less or show no stutter at all
    x Rooms with heavy enemy count (= a lot of collision detection for the AIs) stuttered significantly more
    x Once the game was out of sync (stuttered a lot) it kept stuttering more even in small rooms. Resyncing in-game by disabling/ enabling Vsync seemed to solve this somehow (but not always).

    To summarize there seems to be some dependency on the CPU load as well as on the rendered resolution/ at least in my game. But anyway, this micro-stuttering seems to appear quite randomly and I cannot explain why this example stutters more than my heavy game (about 6000 lines) and why changing resolution of the expample does not seem to help at all.

    edit: I've also read that multi-core setups cause a lot more sutter than singles. When I run Outbuddies on my system it's blocking 100% of one of my cores, while the example .mfa does (of course) not. Could this be relevant to the finding that a heavy game may even stutter less? Really don't know...

  8. #58
    Clicker Fusion 2.5 DeveloperAndroid Export ModuleHTML5 Export ModuleiOS Export Module
    Fusion 2.5 (Steam)
    schrodinger's Avatar
    Join Date
    Nov 2014
    Posts
    3,155
    Mentioned
    27 Post(s)
    Tagged
    1 Thread(s)
    well done Shinbaxter!
    It's indeed clearer this way,
    watching it run, I seem to catch a connection between detected sequences >=4 and noticeable microstutter
    though it also occasionally happens in other non-detected situations, so there's probably other cases to consider


    @ Prominent:
    that's a nice idea, unfortunately this:

    if the distances are any different from the other lines
    won't hold true though, as frames are anyway drawn at end of their (varying) timespan
    to see changes in lines distance (and it would be indeed super useful)
    we should draw lines at perfect "microtimer" intervals >>> "inbetween" frames, which is not possible

    color changing lines could be interesting, or we could make them higher or lower, to actually make a "bar graph",
    I'll try tinker a bit more and see if I can come to something useful


    @ Julian:
    so when you open the mfa you can see it stuttering and not suttering at the same time
    mmm what you noticed is very interesting,
    I wonder if the fact your game is heavy makes the test less prone to stutter,
    because Fusion can "push" less in the accelerator, and thus have less stutter on the "faster" side of the spectrum...?

  9. #59
    Clicker Fusion 2.5 Developer

    Join Date
    Feb 2014
    Posts
    976
    Mentioned
    16 Post(s)
    Tagged
    1 Thread(s)
    it just seems completely random to me lol.

    i run the mfa & its not stuttering so bad one time, close it run it again & its really stuttering.

    its almost like its breathing, like its alive....like it's human!

  10. #60
    Clicker Fusion 2.5Fusion 2.5+ DLC
    casleziro's Avatar
    Join Date
    Mar 2013
    Location
    United States
    Posts
    644
    Mentioned
    11 Post(s)
    Tagged
    0 Thread(s)
    Interesting thread.


    I tried Schrodinger's example and i'm getting an almost rhythmic stutter of 0.017s and 0.015s:
    stutter results.png


    I definitely notice the stuttering in my own projects. At one point I downloaded and tested out Scirra Construct, just to see how it differed from Fusion. In the end, I liked it less than Fusion and it was very clunky in some critical areas. But unlike Fusion it had no noticeable stutter. It was butter smooth when scrolling. So maybe this is something Fusion-specific.

Page 6 of 25 FirstFirst ... 4 5 6 7 8 16 ... LastLast

Similar Threads

  1. Replies: 21
    Last Post: 25th September 2017, 08:50 AM
  2. IOS music pass through fusion app - can't find the thread
    By Tuna in forum iOS Export Module 2.5
    Replies: 6
    Last Post: 15th June 2017, 07:07 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •