While we're here, does anyone know a screen recording application that automatically takes into account display changes, audio device changes, camera resets, etc.?
I'm making do with OBS, but I lose more video to "plugged in external monitor and forgot to update OBS config" than I actually record.
I too frequently plug and unplug dock and external displays. It's a little painful to keep checking and sometimes re-add display source to get the default screen and voice recording working.
I wonder if the "modern" way for problems of this nature is to fork the open-source repo and ask your favourite agent to strip it down to bare essentials that you need (performance gains) and add the specific feature enhancements you want (needs to work only for your configuration -- may not pass the quality standard needed for a PR)
This could probably be automated using either an OBS plugin (if you really want deep integration and not running another program) or a standalone program (written in whatever language you want) that listens for display changes and pokes OBS using its websocket API. The hooks are definitely there to update source configs.
Surely the main function of a capture program should be to capture video, no matter what, even if that means capturing too much / uninteresting sources some of the time?
Except it's really designed to composite video from a bunch of different places and broadcast the result, whether they are windows, screens, cameras, etc. And a rather lot of people don't want every screen to be captured, just the ones that they have selected. It's not a "screen recorder" program, really. And it's explicitly designed with an API to be remote controlled if hardware changes out from under it, so you can encode exactly your own policy logic on what to do. (External display plugged in? Disable my laptop source at full resolution and enable my laptop and monitor sources, side by side, with one scaled as necessary, and in this layout. But if I'm currently recording or streaming, don't do anything. - this is user-specific logic)
Yes, please, that would help so much. OBS seems to be designed for static setups and a single output stream. In my case, I have a setup that changes often (for example, connect a USB video device that processes input from a microscope). I thought I would be able to record multiple video streams at once without burning them into a single output stream (e.g. to multiple files, or a multi-stream file), but this turned out to be difficult without hacks.
Have you considered using GStreamer pipelines for this? And taking it one step further, if you want to use OBS generally, there's a plugin for OBS that allows you to integrate pipelines too.
The common case for the plugin is to do some preprocessing before importing the video into OBS, but you can also use the pipeline as a video filter, which is really powerful, because you can inject a filter into not only a single source, but also at any level of a nested scene. So you can scale/composite several inputs together and then the video filter can copy the source stream and encode/tee it to a file output. Might still be considered a hack of sorts.
The Branchoutput plugin[1] is what you are looking for for multi recording/streaming. It allows you to add recording (or streaming destinations) to individual sources, scenes and outputs via the video filter settings. It is really quite powerful.
Thanks -- this does look like it would do this. But I'm wondering if this entire setup isn't completely overkill. All I need is to record multiple video streams simultaneously. I currently do this by launching multiple Quicktime Player apps and clicking "record" in each one of them...
Is it really that hard to fix this in an imperfect way by letting you save a profile? If everything is the same as the profile, go. If things have changed, and all of the options are available, prompt about resetting to match the profile. When there are missing options, ask if you want to go fix that.
Did they finally fix the bug where if you have less than ~40gb storage the recording will just randomly die for no reason with no notice, it is unimaginably annoying
Is this on macOS [0] or another platform? I don't think I've heard of this issue - if you have a log of such a session, please create a bug report [1].
Man I get exhausted just by reading other people's complaints (without environment description or steps to reproduce). I am puzzled by what keeps OSS maintainers going.
- use obs any of the latest versions on macos any version
- have reasonably low storage
- start recording
- couple seconds or minutes later the video on the preview gets frozen and stops being recorded, while audio continues; recording status unchanged
- the freeze usually happens in the moment of switching between apps, allthough this isn't required at all
yes maintainers do it for free for years which deserves recognition but this exact issue existed for years and has been reported in many ways, without any change, and no other apps have ever behaved this way on my devices
Are you sure that's not time machine stomping all over it? You system might be snapshotting and having trouble at low disk space. Or if you're swapping a lot, then that could also be a reason?
I'm not denying you're experiencing this, I believe this issue occurs for you and likely has occured across OS and/or application versions.
My personal MacBook had 8GB of free storage (4.2GB after my test recording). I experience issues with programs like xcode failing to compile or crashes when I open too many files, pages hasn't been able to save a file since the machine had 15GB+ free.
I just download, installed, and recorded 10 minutes of desktop footage with a webcam overlay and audio. Nothing unexpected happened. I watched all 9 minutes and 28 seconds of it and it looks and sounds exactly as I would expect.
I spent my time doing this and writing this comment not to belittle you or your experience, though at this point I would ask you what do you expect to happen?
You did not post an existing bug report, or write one yourself. You've provided no evidence that anyone who contributes meaningfully to the project has experienced or even heard about this issue, let alone anyone who operates the project or consumes it from macos has any notion it is an issue.
Again, I believe this is a real issue. But there is no information here to go off of, especially with my experience, that points to anyway to meaningfully improve an OSS project in regards to your experience. If nobody with the requisite experience is asking the right questions, this will not get fixed. It starts with those people having any information to start with.
How does one operate with only 8GB of free storage? Are you not constantly receiving OS pop ups about low disk space? Are all of your USB ports dead so you can not attach external storage to move some of you (likely) unused files off? Maybe you like living dangerously? Wonder what other aspects of your life you choose to live so dangerously close to the edge as well?
I'm probably showing my age here, but I'm astonished to read that 8 GB of free space could be considered so exceptionally low that you'd receive OS alerts about it or that apps would start preemptively failing.
The most my homedir has ever grown in one year is 3.2 GB, and I don't really go out of my way to keep things tidy. I'd be very pissed off if things started pestering me to clear stuff up years before it actually became a problem.
lol! This comment made my night. In short: low available disk space, lots of projects requiring disk space.
No I don't see OS notifications about low disk space (at least on my macbook), some 3rd party applications give me notifications about low disk space though. I can confirm that all of my USB-C ports work! I have 2 local drives and a remote NAS that backup all the 'critical' data from my machine.
Unlike other operating systems I've had experience with, macOS is the most capable of operating in a (near) 'disk full' state. Where something like Windows will go from crying about the available disk space, to failing opening explorer, macOS will continue running just fine for general usage.
> Wonder what other aspects of your life you choose to live so dangerously close to the edge as well?
As for this: I've spent the last year living in India, where most people I've met think they are invincible, so I roll the dice every day!
> macOS is the most capable of operating in a (near) 'disk full' state
I used to get the OS out of space warning a lot from macOS. I used to have a 128GB SSD, and Adobe would fill up the drive with cache in the Application Support folder frequently. My newer MBP has a larger SSD, but I also manage the storage much closer now.
Makes sense, I don't run a lot of applications other than developer tooling that keep very small overwriting caches. So day to day build up isn't a huge issue.
This experience I have of low disk space is only on Sequoia which I haven't seen give an OS notification until 5gb left, which again doesn't happen often.
I bought a Mac mini a couple years ago to develop an app and went with the cheapest model: $600 for 256GB version. Once I had Xcode, Visual Studio, and the rest of my development setup installed, I was down to roughly 60GB free. After every build, I moved installation + build/debug files to an external drive to save about 1GB.
Over the following year, roughly 20 releases later, my disk was full. Xcode generates a ton of hidden cruft in the DerivedData folder during every build/release that I wasn't even aware of! I spent a couple days googling and trying free & paid disk space analyzers (sidenote: I didn't realize how good we have it in Windows-land, there are so many good free tools but in Apple-land they are all so frustrating to use and most of them are totally useless for this particular case). Eventually, I found an Apple forum thread explaining where Xcode hides its generated build data, it was like 8GB of stuff per build! I set a calendar reminder for every 12 months to clean this folder.
I can't move anything else to an external drive really because what's left is necessary for build environment. I don't use the Mac for anything except development, so I've literally removed everything and it's down to barebones but it's still a struggle. If I had known I would have invested another $150 for the 512GB version.
This specifically applies to macOS Screen Capture sources freezing, not recordings on any platform.
Further, it's an inexplicable issue with the API; any consumer of SCK is vulnerable to the issue. I'd regard it as Apple's issue to fix, since there is no good reason for frames to randomly stop streaming just because you have less than ~40 GB of disk space free.
Any solution on the OBS side that tried to detect a frozen stream and restart it would not only be a hack and quite possibly lead to greater issues, but could also have privacy implications if we cannot positively identify that the stream was not stopped intentionally (which we cannot).
The plugin-loading overhaul and improved source snapping look like useful updates. OBS continues to show how much attention to detail matters in open-source software. Excited to see how creators and plugin developers adapt to version 33.0!
OBS is much more aimed at livestreaming to platforms like Twitch and managing multiple cameras incl one of you, one of your screen etc. Well beyond what Cap is for.
But yes, use the most specialized tool for the task at hand.
Cap completely ate several recordings for me. Admittedly they were over an hour long, and I should’ve broken them up a bit, and perhaps it’s not what it’s used for. Switched back to OBS, while it’s not as easy to use it’s pretty reliable
I tried a bunch of "open" tools last week, but turns out they are a paid tool in disguise most of the time. I eventually paid for CleanShotX for my work and have peace of mind because it just works. OBS is IMO much more stable and aimed at live streaming anyway.
ShareX has been around for ever but just updated the UI and now super easy to do capture work with annotations and cam output etc (no studio but it's light weight)
btw I use OBS virtual camera its one of my fav features with masking etc for a video avatar vlogger effect
There are a handful of comments here rueing buggy behaviors. These days with nearly the same amount of effort to post a complaint you can delegate the task of reproducing the issue, diagnosing the root cause, and proposing a fix. Yes, it may incur a small to moderate cost, but it's a contribution to an open source project that apparently provides some value in return.
If all you do is complain you'll find yourself left out, for good reason. The internet is really not all that different when it comes to this, if you consume other people's emotional energy without ever giving anything back it would be unhealthy to interact with you, regardless of where or why.
The bigger issue is perhaps whether one views someone else politely pointing out something they would like to see improved/changed as a complaint. I suspect a not insignificant portion of users of English unfortunately would interpret it that way. However, the ubiquity of the LLMs might give us some hope that the sending side over time ceases contributing to the problem by saddling the people they wish to freely sacrifice their lifetime for them with their frustration.
> "Appeal to personal responsibility" is not a fallacy.
It is one when you deflect criticism instead of addressing it. Maybe you should have read more than the first sentence after you googled what a fallacy is.
Okay, I'll bite the bait. Assuming that I am a regular user (think gamer), I have a Windows PC, no dev tools, no AI subs or accounts, no GitHub. How specifically can I do the steps required to propose the fix they way you suggest?
Out of interest- what is your recommended path for this? I mean the path of least resistance where I don’t even have to check out the code locally or similar.
Officially and unofficially, the most you need to do is create an Issue report with valid reproduction steps. You'd be surprised how quickly some issues get fixed just because someone finally provided the missing link in an Issue.
Yeah, I don't agree that using an LLM to generate a PR in a code base is the best line of action. I could see using an LLM to generate information to provide to create an Issue with would be potentially useful. It doesn't generate code in a way that may not be in the project's style. It doesn't hide anything from the devs. Instead, it points the devs to the specific problem and allows the devs to grok what/why/how of the issue. As we've all seen before (I'm assuming), fixing a bug in one place might not be the best fix place for the fix or a fix might reveal an issue somewhere else.
I think this is what the whole "AI killed software" fallacy came from.
>nearly the same amount of effort to post a complaint you can delegate the task of reproducing the issue, diagnosing the root cause, and proposing a fix.
Is the effort the same though?
And even it is the same, making a meaningful contribution to an open source project doesn't mean just fixing your problem, but you also need to stay in line with the contribution guidelines, the project's direction and many more, which takes much more effort than just vibe coding a version for yourself.
Building software for one and for a bunch are two different things. AI is good at the former and can be used for the later but not without much effort. Models will get better, true, but that doesn't mean we need less engineering, in fact we need more to stay up-to-pace with AI. Imagine everyone submitting a PR to fix a bug, how fast can you review it manually? And if you use AI to review/verify them, the cost just shift from person who submitted the PR to the open source maintainer.
That doesn't mean AI is bad but that we need new systems and rules.
> All text and content submitted to our project, which includes code, descriptions, issues, and comments, must be human written.
> The use of GitHub Copilot and other assistive AI technologies is heavily discouraged.
> Low-effort or incorrect submissions that are determined to have been generated by, or created with aid of such systems may lead to a ban from contributing to the repository or project as a whole.
> There are a handful of comments here rueing buggy behaviors. These days with nearly the same amount of effort to post a complaint you can delegate the task of reproducing the issue, diagnosing the root cause, and proposing a fix. Yes, it may incur a small to moderate cost, but it's a contribution to an open source project
That would violate OBS's rules - they disallow AI-written code (or write ups) being submitted. Additionally, if you put nearly the same amount of effort into it as a comment mentioning the original bug - i.e. not much effort - they say that may result in a project ban.
Finally... The CEF they used was 2 years out of date, and after RCE exploits [0] occuring in the wild getting big enough that they got picked up in the media [1], we finally have a fix 1 month later.
You can find why it was complicated [0] and why it took so long [1] on GitHub. I do wish I had been faster on it. That said, the RCE required the user to be using a vibe-coded web-based chat overlay that didn't properly sanitize chatter inputs, so has a (hopefully) limited scope.
Thanks for chiming in and linking that, that was an interesting read. As much as I understand CEF's reasoning for making this shift, it's baffling that they don't provide a universal way to implement what you now had to implement yourself.
Unfortunately, most streamers don't have the best IT security knowledge, and with AI tools now allowing them more and more than what was previously possible, it also opens the door for more vibe coded apps that could have serious security holes.
Should the streamers' apps be more secure? Absolutely, but the underlying tool should also strive to be secure enough for all use cases.
Anyway, thanks for the two links, they were interesting reads.
I love that I can stream from my iPhone directly into OBS on my Mac. I don't love that Metal shaders aren't supported, and I have to rely on janky extensions to get them working.
Is this something new? I've seen streamers produce a https://en.wikipedia.org/wiki/Droste_effect with OBS self-capture many times. (Unless that only happens with screen capture?) It doesn't seem like this is something you'd want to do intentionally, anyway.
I haven't spent much time in the video processing/streaming space, but I was delighted to discover OBS Studio this week while putting together what has turned out to be an incredibly cozy little entertainment setup by feeding the two gaming consoles and geriatric gaming desktop I've acquired over the years into an HDMI switcher and USB HDMI capture box that in turn streams the feed into an OBS Studio preview (launched from Steam) running on a dummy display.
Launching OBS Studio through Steam lets me stream that preview to my VR headset over a 6GHz WiFi dongle and play games wirelessly in what feels like my own little private theatre. OBS Studio was the puzzle piece I hadn't figured out when the headset showed up last weekend, and I've been really impressed by its performance so far... I think this setup even shaved some latency off when playing online, since consolidating the consoles next to the router let me hardwire them.
Many many thanks to the OBS Studio team for making software that's helping me live out a small part of what I remember hoping technology could be like back in the 90s. I am grateful (:
What's the negativity? Anything wrong with OBS in particular? I haven't used it in a while, but recently saw DHH saying it's awful. I reckon that's the usual Twitter hyperbole/rhetoric that attracts attention, so just wondering.
That's not what he said. OBS decided to not support streaming to Rumble for political reasons. So Omarchy is gonna ship a leaner alternative soon-ish, that's all.
Must be quite comfy to live in such a simplistic reality. Imagine not having to think through all the complexity of life all the time by just crying fascism whenever people disagree with you.
Was envious for a second there. Then I was reminded that I quite enjoy thinking.
This is disingenuous. You chose to word it in a ridiculous way: "crying fascism whenever people disagree with you". If you enjoy thinking, please think a bit more and be more charitable. Maybe it's not "crying" at all? Maybe it's not because of a simple disagreement?
I saw somebody else besides DHH talking about their decision not to accept PRs with AI written code briefly. I believe there’s a Rust fork in progress.
It's been years since I needed to use it, but: lack of an undo button. Not being able to undo stuff was very annoying, and felt out of place for an otherwise nice piece of software.
A lot of negativity about OSS is deserved, and doesn't necessarily come from supporters of proprietary software, but for long-time OSS users and programmers, who want it to be better.
But many OSS programs are awesome, and OBS is one of them.
I believe a lot of negativity comes from people who believed the OSS preachers "come come, you don't need that proprietary crap, we have everything free and better"
And then most people are in for a surprise, that usually the above is only true in theory. But in reality there are driver issues, missing firmware, outdated repositories, broken updates ... critical missing features and in general a need to understand technical details with no polished UI.
And speaking of, with OBS I really liked the UI to just get started, but performance/quality of produced video was way worse for me, then simple screencapture with ffmpeg.
It was very heartening to have this experience! I tried to refrain from harboring overly positive or negative expectations while setting this up and am quite content with how it turned out (especially since every part of this stack can run on Linux; this morning, I swapped out the gaming desktop part of this setup out for my Linux dev laptop + eGPU just for fun, and it all works identically).
Funny how you went the huge app route for such a narrow scope where I would have just used a lightweight tool like ffplay given the setting would never really change.
Quite so! I'm sure I could have fiddled with ffmpeg commands, set up a custom Steam launcher, etc. I probably would have if OBS Studio's preview performance introduced too much overhead.
I also discovered that OBS Studio can be installed directly from the Steam store, which makes makes the whole thing feel just a little more seamless (I'm streaming to the Steam Frame). I don't mind trading the extra install space for that feeling, especially since OBS is OSS. It was really pleasant to have this all "just work" end-to-end on both Linux and Windows stacks.
Yes in contradiction to what the downvoters might think, I am not judgemental. By all means use what works for you. I was just pointing out how one thing can be done different ways and I might not even have thought about using OBS for that purpose while being a user myself of the program.
Surprisingly low. Steam reports about 12-15ms latency over the 6GHz dongle, and the Internet seems to think that the capture card preview adds about 49 ms; I haven't tested the latter claim myself, but ~65ms feels like a reasonable number. I don't usually seem to notice <100ms latency when I play games (which could be a skill issue on my part), and indeed I don't notice delay between the VR headset image and the direct HDMI passthrough that goes to my desktop monitor (the non-dummy display)
I had initially only planned to use this setup for turn-based games, but after this pleasant surprise, it looks like I get to explore some other kinds of games :D
> Great, another project with an asinine and backwards-thinking AI policy
Nothing asinine about it. Just because many in the industry have decided they are willing to accept the downsides of using output generated by LLMs in their product doesn't mean that everyone in the industry has to. People are still free to do things differently than you do.
I'm making do with OBS, but I lose more video to "plugged in external monitor and forgot to update OBS config" than I actually record.
I wonder if the "modern" way for problems of this nature is to fork the open-source repo and ask your favourite agent to strip it down to bare essentials that you need (performance gains) and add the specific feature enhancements you want (needs to work only for your configuration -- may not pass the quality standard needed for a PR)
People streaming things mights disagree.
The common case for the plugin is to do some preprocessing before importing the video into OBS, but you can also use the pipeline as a video filter, which is really powerful, because you can inject a filter into not only a single source, but also at any level of a nested scene. So you can scale/composite several inputs together and then the video filter can copy the source stream and encode/tee it to a file output. Might still be considered a hack of sorts.
[1] https://obsproject.com/forum/resources/branch-output-streami...
sigh why does Apple name things so wrong. How do you edit a pdf? In Preview, of course. How do you record a video, with Quicktime Player, of course.
(sorry for the rant)
[0] https://github.com/obsproject/obs-studio/issues/10596
[1] https://github.com/obsproject/obs-studio/issues
- have reasonably low storage
- start recording
- couple seconds or minutes later the video on the preview gets frozen and stops being recorded, while audio continues; recording status unchanged
- the freeze usually happens in the moment of switching between apps, allthough this isn't required at all
yes maintainers do it for free for years which deserves recognition but this exact issue existed for years and has been reported in many ways, without any change, and no other apps have ever behaved this way on my devices
My personal MacBook had 8GB of free storage (4.2GB after my test recording). I experience issues with programs like xcode failing to compile or crashes when I open too many files, pages hasn't been able to save a file since the machine had 15GB+ free.
I just download, installed, and recorded 10 minutes of desktop footage with a webcam overlay and audio. Nothing unexpected happened. I watched all 9 minutes and 28 seconds of it and it looks and sounds exactly as I would expect.
I spent my time doing this and writing this comment not to belittle you or your experience, though at this point I would ask you what do you expect to happen?
You did not post an existing bug report, or write one yourself. You've provided no evidence that anyone who contributes meaningfully to the project has experienced or even heard about this issue, let alone anyone who operates the project or consumes it from macos has any notion it is an issue.
Again, I believe this is a real issue. But there is no information here to go off of, especially with my experience, that points to anyway to meaningfully improve an OSS project in regards to your experience. If nobody with the requisite experience is asking the right questions, this will not get fixed. It starts with those people having any information to start with.
The most my homedir has ever grown in one year is 3.2 GB, and I don't really go out of my way to keep things tidy. I'd be very pissed off if things started pestering me to clear stuff up years before it actually became a problem.
lol! This comment made my night. In short: low available disk space, lots of projects requiring disk space.
No I don't see OS notifications about low disk space (at least on my macbook), some 3rd party applications give me notifications about low disk space though. I can confirm that all of my USB-C ports work! I have 2 local drives and a remote NAS that backup all the 'critical' data from my machine.
Unlike other operating systems I've had experience with, macOS is the most capable of operating in a (near) 'disk full' state. Where something like Windows will go from crying about the available disk space, to failing opening explorer, macOS will continue running just fine for general usage.
> Wonder what other aspects of your life you choose to live so dangerously close to the edge as well?
As for this: I've spent the last year living in India, where most people I've met think they are invincible, so I roll the dice every day!
I used to get the OS out of space warning a lot from macOS. I used to have a 128GB SSD, and Adobe would fill up the drive with cache in the Application Support folder frequently. My newer MBP has a larger SSD, but I also manage the storage much closer now.
This experience I have of low disk space is only on Sequoia which I haven't seen give an OS notification until 5gb left, which again doesn't happen often.
Over the following year, roughly 20 releases later, my disk was full. Xcode generates a ton of hidden cruft in the DerivedData folder during every build/release that I wasn't even aware of! I spent a couple days googling and trying free & paid disk space analyzers (sidenote: I didn't realize how good we have it in Windows-land, there are so many good free tools but in Apple-land they are all so frustrating to use and most of them are totally useless for this particular case). Eventually, I found an Apple forum thread explaining where Xcode hides its generated build data, it was like 8GB of stuff per build! I set a calendar reminder for every 12 months to clean this folder.
I can't move anything else to an external drive really because what's left is necessary for build environment. I don't use the Mac for anything except development, so I've literally removed everything and it's down to barebones but it's still a struggle. If I had known I would have invested another $150 for the 512GB version.
We need the money.
Further, it's an inexplicable issue with the API; any consumer of SCK is vulnerable to the issue. I'd regard it as Apple's issue to fix, since there is no good reason for frames to randomly stop streaming just because you have less than ~40 GB of disk space free.
Any solution on the OBS side that tried to detect a frozen stream and restart it would not only be a hack and quite possibly lead to greater issues, but could also have privacy implications if we cannot positively identify that the stream was not stopped intentionally (which we cannot).
Answer this thread if you struggle and I can post when I get to the office
Much happier with Cap for screencast recording https://github.com/capsoftware/cap
But yes, use the most specialized tool for the task at hand.
Even if the preview player shows it correct, once it's rendered the cursor just seems to drift way off making the video useless.
I've gone back to using QuickTime.
btw I use OBS virtual camera its one of my fav features with masking etc for a video avatar vlogger effect
https://getsharex.com/
* Appeal to personal responsibility;
* False equivalence.
I'll strike those off my bullshit bingo card.
The bigger issue is perhaps whether one views someone else politely pointing out something they would like to see improved/changed as a complaint. I suspect a not insignificant portion of users of English unfortunately would interpret it that way. However, the ubiquity of the LLMs might give us some hope that the sending side over time ceases contributing to the problem by saddling the people they wish to freely sacrifice their lifetime for them with their frustration.
It is one when you deflect criticism instead of addressing it. Maybe you should have read more than the first sentence after you googled what a fallacy is.
But at https://obsproject.com/help we see OBS has a Discord (regular gamer), a Forum, a Feedback page, and a Mastodon.
>nearly the same amount of effort to post a complaint you can delegate the task of reproducing the issue, diagnosing the root cause, and proposing a fix.
Is the effort the same though?
And even it is the same, making a meaningful contribution to an open source project doesn't mean just fixing your problem, but you also need to stay in line with the contribution guidelines, the project's direction and many more, which takes much more effort than just vibe coding a version for yourself.
Building software for one and for a bunch are two different things. AI is good at the former and can be used for the later but not without much effort. Models will get better, true, but that doesn't mean we need less engineering, in fact we need more to stay up-to-pace with AI. Imagine everyone submitting a PR to fix a bug, how fast can you review it manually? And if you use AI to review/verify them, the cost just shift from person who submitted the PR to the open source maintainer.
That doesn't mean AI is bad but that we need new systems and rules.
> All text and content submitted to our project, which includes code, descriptions, issues, and comments, must be human written.
> The use of GitHub Copilot and other assistive AI technologies is heavily discouraged.
> Low-effort or incorrect submissions that are determined to have been generated by, or created with aid of such systems may lead to a ban from contributing to the repository or project as a whole.
That would violate OBS's rules - they disallow AI-written code (or write ups) being submitted. Additionally, if you put nearly the same amount of effort into it as a comment mentioning the original bug - i.e. not much effort - they say that may result in a project ban.
Finally... The CEF they used was 2 years out of date, and after RCE exploits [0] occuring in the wild getting big enough that they got picked up in the media [1], we finally have a fix 1 month later.
[0] https://blog.scrt.ch/2026/09/22/how-one-twitch-chat-message-...
[1] https://news.ycombinator.com/item?id=49852143
[2] https://www.scworld.com/brief/vulnerable-obs-overlay-and-chr...
[0] https://github.com/obsproject/obs-studio/discussions/3853#di...
[1] https://github.com/obsproject/obs-studio/discussions/3853#di...
If you don’t know what you are doing any tool can be dangerous
Whew, thanks for putting my fears at rest. You're only pwned some of the time, not always.
Should the streamers' apps be more secure? Absolutely, but the underlying tool should also strive to be secure enough for all use cases.
Anyway, thanks for the two links, they were interesting reads.
Launching OBS Studio through Steam lets me stream that preview to my VR headset over a 6GHz WiFi dongle and play games wirelessly in what feels like my own little private theatre. OBS Studio was the puzzle piece I hadn't figured out when the headset showed up last weekend, and I've been really impressed by its performance so far... I think this setup even shaved some latency off when playing online, since consolidating the consoles next to the router let me hardwire them.
Many many thanks to the OBS Studio team for making software that's helping me live out a small part of what I remember hoping technology could be like back in the 90s. I am grateful (:
Was envious for a second there. Then I was reminded that I quite enjoy thinking.
https://obsproject.com/blog/obs-studio-27-released
"There are the live video compositors everyone complains about and there are the ones nobody uses."
But really, it is great to see a piece of open source software as the default choice in a field where they tend to suffer, desktop applications.
But many OSS programs are awesome, and OBS is one of them.
And then most people are in for a surprise, that usually the above is only true in theory. But in reality there are driver issues, missing firmware, outdated repositories, broken updates ... critical missing features and in general a need to understand technical details with no polished UI.
And speaking of, with OBS I really liked the UI to just get started, but performance/quality of produced video was way worse for me, then simple screencapture with ffmpeg.
- size on disk
- cpu usage
- size in memory
- impact on user lives
- impact on developer lives
- impact on our industry as a whole
- <insert other criteria>
The thing is, ffmpeg is a dependency of obs, so if you install obs you have ffmpeg installed while the opposite is not necessarily true.
I also discovered that OBS Studio can be installed directly from the Steam store, which makes makes the whole thing feel just a little more seamless (I'm streaming to the Steam Frame). I don't mind trading the extra install space for that feeling, especially since OBS is OSS. It was really pleasant to have this all "just work" end-to-end on both Linux and Windows stacks.
I had initially only planned to use this setup for turn-based games, but after this pleasant surprise, it looks like I get to explore some other kinds of games :D
> All text and content submitted to our project, which includes code, descriptions, issues, and comments, must be human written.
> The use of GitHub Copilot and other assistive AI technologies is heavily discouraged.
Use this instead: https://github.com/obs-rust/obs-studio
Nothing asinine about it. Just because many in the industry have decided they are willing to accept the downsides of using output generated by LLMs in their product doesn't mean that everyone in the industry has to. People are still free to do things differently than you do.