Str4y & SpeedySpawn
SpeedySpawn
Hey Str4y, I found a glitch in the Braid speedrun that could shave a full second off, but the timing is razor thin. Need a crystal‑clear step‑by‑step to test it—can you break it down into a puzzle‑like verification?
Str4y
Sure, here's the puzzle. 1. Lock your controller into a fixed input buffer. 2. Trigger the first jump on frame 0 of the second lap. 3. Wait exactly 120 frames, then press the button that normally slows you. 4. Observe the timer—if it ticks down a full second, you hit the glitch. 5. Repeat with a 1‑frame offset to confirm. 6. If it fails, back up one frame and try again. The secret lies in that tiny frame window. Good luck.
SpeedySpawn
Nice! That’s a tight window—120 frames is 2 seconds on a 60fps run, so any slip is brutal. Make a macro script that logs every frame, or use a frame counter overlay. If the glitch is consistent, the timer should drop exactly 1000ms; any deviation means you’re one frame off. When you back up one frame, watch the lag spike—if it still ticks down, that’s your sweet spot. Good luck, champ!
Str4y
Okay, you’ll need a frame logger that writes to a text file, each line a frame number, and a separate log that records the timer ticks. Then parse that file: look for a line where the timer changes by exactly 1000ms between two frames that are exactly 120 frames apart. If that line exists, you have the spot. If not, the glitch might be probabilistic, or you’re reading the wrong timer. The key is to isolate the frame where the game’s internal counter lags by exactly one cycle. Once you find that, repeat to confirm. Good luck.
SpeedySpawn
Got it—grab a text‑editor, dump the frames, and run a quick Python script to flag any 120‑frame gaps with a 1000ms tick. If it pops up, that’s the spot. If it’s a probabilistic glitch, just crank the buffer up to thousands of trials. Either way, once you lock it, you’ll be shaving seconds off everyone else’s runs. Let's crack this!
Str4y
Sounds like a logic puzzle wrapped in code. 1. Make the log a simple CSV: frame, timer. 2. In Python, iterate over rows, subtract current frame from previous, check if difference equals 120. 3. When that condition holds, see if timer difference equals 1000. 4. Flag the frame number. If none show up, try increasing the trial count until you get a hit, then re‑verify. Keep the buffer quiet, the noise is what kills precision. Good luck.