The actual process of making no-commentary speedrun videos

Most people think no-commentary speedrun videos are just raw gameplay with the audio turned off. That's technically correct but completely misses what actually separates a watchable run from a boring mess. The difference is timing, camera work, and a thousand tiny decisions made between frames. I recorded my first legit run in 2019 for a mid-tier puzzle platformer. I cut the commentary track and uploaded it. It got twelve views. Nobody commented. The problem wasn't the absence of talking — it was that the video had no structure. Just sixty minutes of identical input. People click away at forty seconds because nothing holds their attention.

Setting up your Gameplay Speedrun No Commentary recording You need OBS or a comparable capture tool, a stable home folder for raw footage, and a way to verify your inputs before posting. I use a simple naming convention: game title, run ID, segment length, and date. Without this, finding a specific take during editing becomes a nightmare. I spent three weeks once looking for a clip where I had nailed a tricky sequence, only to realize I'd overwritten it two days earlier because I didn't label anything. Record in 1080p minimum at sixty fps. Lower resolutions make input shows illegible, and input shows are the backbone of any no-commentary submission. Viewers need to see exactly what you're pressing at the frame level. A 30 fps recording smears those frames together and looks like garbage on close inspection. Keep your HUD on unless the category explicitly requires HUD-less. This isn't about aesthetics. HUD elements give viewers anchor points — timer, current section, any relevant counters. Without them, people don't know where they are in the run. They leave. < h3 >Why input shows matter more than anything else

An input show displays every button press, controller stick movement, and key event as an overlay alongside your footage. Tools like SHInputShow or the built-in input recorder in TASVideos-compatible software will generate these automatically if configured correctly. I've seen runners submit runs without them and get rejected within minutes. It's a hard requirement for almost every major leaderboard. The input show shouldn't be cluttered. I keep mine minimal — just the controller inputs, no unnecessary data. Too many overlays fight for screen space and distract from the actual gameplay. You're showing the run, not your debugging tools. Sometimes your input show won't sync properly. I ran into this once with a retro emulator build where the frame timing drifted by roughly two frames over a fifteen-minute segment. The inputs looked correct but landed slightly ahead of the visual feedback. I fixed it by recompiling the emulator with a fixed frame pacing flag and re-recording. It took me about twenty minutes total. Most people would just assume the run was invalid and move on.

Editing without killing the flow

No-commentary doesn't mean unedited. You still need to trim the dead air. The intro cinematics, loading screens, menu navigation, and deaths outside of recorded segments should all be cut. Most category rules allow for intro skips and death cuts depending on the split system being used. I typically edit using FFmpeg command-line tools for the heavy lifting. It's faster than a GUI editor for batch processing segments. My workflow takes a raw sixty-minute run and compresses it down to roughly forty-five minutes of actual content, sometimes less for games with shorter optimal routes. The goal is density. Every second should show something that matters. Color correction isn't necessary. I've seen people spend hours grading footage for no reason. The original game output is fine. Don't add filters, transitions, or music. Music changes the pace artificially and can invalidate certain categories. Some leaderboards explicitly ban background tracks. Check the rules before you post.

Common mistakes that sink runs

The biggest mistake I see is runners treating no-commentary like a lazy option. It's not. It requires more precision in timing and awareness because there's no voice trail to fall back on. When you're narrating, you can talk through mistakes, explain why you made a certain choice, recover from frustration. Without that safety net, every error is exposed. Viewers notice things you might overlook. Another issue is not validating your timer. I've submitted runs where the timer was off by a few seconds because I started it at the wrong moment. Some games have frame-perfect timer triggers. If you begin counting a frame early, your time is invalid regardless of how fast the run actually was. Always check the split times against the official category requirements before exporting. A third problem is ignoring the community standards for file format. MP4 with H.264 is the standard. WebM works too. Don't upload MKV files or anything exotic. The validation teams and viewers don't have specialized codecs. Your run gets flagged or ignored.

Where no-commentary runs fall flat

This format doesn't work for every game. Narrative-heavy titles with heavy dialogue segments become excruciating to watch when stripped of explanation. A three-hour RPG with minimal gameplay breaks down into repetitive grinding footage that nobody wants to sit through. Speedrun categories for those games usually thrive on commentary because the talk bridges the gaps between mechanically shallow sections. Games with opaque mechanics also suffer. If a run depends on subtle knowledge about frame data or memory manipulation that isn't visible on screen, viewers have no way of understanding what's happening. Commentary fills that gap. Without it, the run looks random even though every input is deliberate. If your game falls into either of those categories, consider a hybrid approach. Keep the commentary minimal during high-skill sequences but allow narration during transitions or explainer moments. Some runners do this successfully. It's not pure no-commentary, but it serves the content better than forcing a format that doesn't fit.

What to do before you upload

Watch your own run all the way through once before uploading. You'll catch timing issues, clipped frames, or audio artifacts that you missed during recording. I usually watch it at 1.5x speed on the second pass. Anything that feels off at that pace is definitely worth fixing. Check your split times against existing top runs. If your times are significantly slower in certain segments, figure out why before you post. It could be a routing error, a missed sequence break, or simply not executing a technique optimally. Running the same segment five times and comparing outputs helps isolate where you're losing ground. Post in the right community. Leaderboard forums like TASVideos, SpeedRun.com, or dedicated Discord servers have specific submission guidelines. Follow them exactly. Skipping that step results in rejected posts and wasted time.