Getting the transcript
Reading the captions from YouTube. A video nobody has opened here before takes 10 to 30 seconds; this page fills in on its own.
Getting the transcript
Reading the captions from YouTube. A video nobody has opened here before takes 10 to 30 seconds; this page fills in on its own.

Roblox Learn · @RobloxLearn
Where viewers went back to watch this video again, from YouTube's public Most replayed graph, lined up with what was said at that moment.
Most replayed moment #1
4:0311.7x the video's typical replay level
future. When you enable slim on a model, Roblox automatically generates many meshes from high complexity to low complexity and chooses which one to render based on distance from the camera. Here, my character is outside the streaming radius. As the camera
Said at 3:57
Most replayed moment #2
2:252.0x the video's typical replay level
performance gains in a scene on a target device and decide if the visual trade-off is worth it. The more interesting settings come when you enable streaming, particularly level of detail. Rather than just reduced quality, streaming lets you choose between having models outside of the streaming radius not appear at all,
Said at 2:18
Most replayed moment #3
2:512.0x the video's typical replay level
streaming lets you choose on a per model basis to achieve the visual outcome you're looking for. First, the automatic or disabled setting. This one is pretty straightforward. The model initially isn't there, and when you enter the streaming radius, it shows up. This setting respects the render fidelity properties, so
Said at 2:43
The graph counts replays. It does not show where viewers stopped watching.
Words
1,095
Runtime
6:03
Speaking pace
181wpm
Reading time
5min
181 words per minute, the same as the 181 median of 349 measured videos. That distribution comes from the 349-video hook study.
Opening (first 30 seconds)
This video is about managing level of detail settings in Roblox experiences. And yes, I hope to cover the subject at an appropriate level of detail. I'm so sorry. Let's let's get right into it. Hi everyone, it's Andrew from Roblox. Level of detail refers to reducing the complexity of a model as it gets farther and farther away from the character or camera, which can have huge performance benefits for your experiences. If you've ever been really close to a painting at an art museum, you might have noticed that
91 words, the words spoken in the first 30 seconds at 181 words per minute.
Free, no signup. See how the first 30 seconds hold attention, with rewrites.
Sentence shape
| Measure | This transcript |
|---|---|
| Sentences | 64 |
| Average words per sentence | 17.1 |
| Longest sentence | 50 words |
| Questions asked | 0 |
| Sentences containing a number | 3 |
Most used terms
Filler phrases
3 in total: actually 2 · like 1.
A literal whole-word count of the same phrase list the Prepublish browser extension uses, so a phrase inside another word is not counted and a phrase used in its ordinary sense still is. It is a count and not a judgement.
Free, no account. See where attention is likely to drop, with a rewrite for each weak line. The free check shows the scores and the one issue costing the most. Or run it on the words above first.
Free · No login · See a sample audit first if you prefer.
What this transcript is
Every word below is the caption track YouTube publishes for this video, pulled from the video itself and reproduced unchanged. It is not Prepublish's writing, not a summary, and not a re-transcription: it is the video's own published captions. English captions, generated automatically by YouTube, in the video’s original language. Source: the video on YouTube. A channel that would rather this page did not exist can ask for its removal through the contact page, and it is removed.
This video is about managing level of detail settings in Roblox experiences. And yes, I hope to cover the subject at an appropriate level of detail. I'm so sorry. Let's let's get right into it. Hi everyone, it's Andrew from Roblox. Level of detail refers to reducing the complexity of a model as it gets farther and farther away from the character or camera, which can have huge performance benefits for your experiences.
If you've ever been really close to a painting at an art museum, you might have noticed that what you recognized as a beautiful bushy tree from 5 meters away is actually just a brown blob for the trunk and a few expert taps of the paintbrush for the foliage. Our eyes are helpless without our brains. So at a distance, our brains interpret that blob as a tree in a process called perceptual completion. This process, humans interpreting reality rather than seeing it, is what allows level of detail systems in games to lower the quality of models at a distance while still providing visuals that are comparable to the originals.
Ideally, ideally comparable. I say ideally because it's easy to imagine a game going too far, oversimplifying to the point that our brains can't connect the dots, and we just see an ugly smudge. A slow, methodical game might need to use conservative LOD settings since the image on screen is more stable, whereas a fast Twitch experience might be able to get away with more aggressive settings. If you remember our performance optimization video, link in the description, everything you render has a cost in terms of draw calls and triangles.
Even if an object is really far away, that cost is roughly the same. So, level of detail systems take advantage of perception to delete a bunch of triangles as the camera moves away. That now lower cost means higher performance. The most rudimentary setting in Roblox is the render fidelity property on mesh and solid modeled parts. At automatic, Roblox renders them at high, medium, and finally low quality as the [music] camera moves away.
You can strategically change to precise if visual quality suffers or performance for less important parts. See how these three trees behave as we move back and forth [music] along my 500 stud ruler. I personally can't tell the difference between automatic and precise at this size and resolution and against the skybox, but maybe your mesh is more complex or your background less busy. It's up to you to measure the performance gains in a scene on a target device and decide if the visual trade-off is worth it.
The more interesting settings come when you enable streaming, particularly level of detail. Rather than just reduced quality, streaming lets you choose between having models outside of the streaming radius not appear at all, using lowresolution impostor meshes, showing lightweight slim meshes, or in rare cases, displaying the original model. Each choice has trade-offs, so streaming lets you choose on a per model basis to achieve the visual outcome you're looking for.
First, the automatic or disabled setting. This one is pretty straightforward. The model initially isn't there, and when you enter the streaming radius, it shows up. This setting respects the render fidelity properties, so quality might decrease as you move away. When you eventually leave the radius, the engine can remove the model either opportunistically, which I've enabled here, or when your device needs the memory.
In terms of rendering fewer triangles, you can't do much better than zero. But this setting can be quite the visual cutback. Some models though, you might not care if they disappear, probably because the object is small enough or insignificant enough that it doesn't negatively impact visuals outside of the streaming radius. Next, impostor meshes. Choose streaming mesh to get a chunky, lowresolution mesh that looks pretty crude up close, but like our painting from earlier, might look fine at a distance.
It depends on the model, its size, and the rest of the scene. Our general guidance is to use these with a streaming target radius of at least 1,000 studs. Any closer and the cracks start to show. Well, not not literal cracks because impostors are smooth meshes, but you get the idea. Slim meshes are our newest option and one that we view as critical to enabling the large highfidelity worlds of the future. When you enable slim on a model, Roblox automatically generates many meshes from high complexity to low complexity and chooses which one to render based on distance from the camera.
Here, my character is outside the streaming radius. As the camera moves away from the tree, you can see quality gradually degrading. Or maybe you can't notice at all. If you can't, that's the idea. Broadly speaking, slim meshes offer much better visuals than impostors at a similar performance level. We think they're the sweet spot in many situations. If you absolutely must have a model on the client at all times, likely because a client script relies on it for some gameplay logic, you can set model streaming mode to persistent so that it never streams out.
The interaction with the level of detail setting is a bit complex. Automatic is essentially ignored because the model never streams out. Streaming mesh is ignored as well. the model never switches to an impostor mesh. Slim meshes do support this setting, however, so you can have a persistent model with slim enabled. Our overarching advice is to only use this option when absolutely necessary for gameplay purposes, not for visual purposes.
Setting too many models to persistent defeats the purpose of streaming. To close things out, I want to talk about visual quality. While it's true that level of detail settings render distant objects at lower quality levels, appropriate use of this feature can actually increase overall visual quality. When you save performance on distant objects, you have more processing [music] power to devote to nearby objects. Higher frame rates mean that players who have graphics quality set to automatic are more likely to stay at higher quality levels.
So rather than thinking about what you're losing, think about what you're gaining. the ability to have denser scenes with higher fidelity assets and an overall better appearance for the things that players are most focused on, all without tanking your frame rate. For more information on working with the various level of detail settings from this [music] video, check out the documentation at create.roll.com. Thanks for watching.
The words are the caption track's own and nothing is reworded or re-transcribed. Paragraph breaks are placed between sentences so the text reads as prose.
Free tools for your own script: paste a draft and see where it stands before you record it.
Paste your draft and see where viewers are likely to drop off, with a rewrite for each weak line.
Paste the first 30 seconds of your own draft for a hook score and rewrites.
Check your draft against YouTube's advertiser-friendly guidelines before you record it.
Read this channel's public videos and transcripts, and download a writing brief for it.