The best way for Owncast to be something that's useful to the most people is to hear what people have to say, so here's an attempt to create an opportunity to do that.

Are you a current or prospective streamer who is interested in providing thoughts, feedback, or have questions to ask 1 on 1? Or you're interested in other aspects of the project from an open source, integration, or Fediverse prospective?

Find a few minutes to chat about Owncast!

@owncast wait... like... talk... like... with other humans?

Not gonna yell at folks to put in a PR or just write it themselves?

Like... we're real humans to you? =-O

Sorry for the thick sarcasm, but holy shit, it is actually really surprising and refreshing to see devs wanting to have active conversations with folks that might not be able or willing to just talk at you through a github incident.

So thank you very much! I will be doing this eventually.

@owncast Any chance you'll also federate with Peertube instances?

@lispi314 Hope so! It took a while to get a Federation incompatibility fixed with Peertube (, but now that is (hopefully, but I haven't seen the fix released yet) resolved I'd like to some day have full federation between the two. But admittedly it's been lowered on the priority list, since there's so many more important things that can help streamers that can be focused on. But once I see PeerTube able to follow Owncast we'll be able to push further.

@owncast I got plenty of questions / concerns, but they probably won't fit in a single post.

1.) At the risk of stating the obvious, streaming on Owncast needs to be profitable for the streamer, or at the very least, not a financial burden; otherwise there's no incentive to use it over Twitch, YouTube, etc.

Mastodon, PixelFed, etc. are only viable alternatives to Twitter, Instagram, etc. imo because someone with a lot of money is generous enough to host instances for us to use for free.

@owncast Having taken a quick look at web hosting packages, the costs easily exceed "hobby funds" territory unless I plan on just privately entertaining a few close friends (unless I'm gravely overestimating the requirements of sustaining say 20, 50, or 100 concurrent viewers).

Also what happens if more people than the instance can sustain try to watch? Do they get an error, does it just fail to load for them, does the stream crash or degrade in quality for everyone?

@owncast That all being said, how much would it cost to host an instance that can accommodate 20 or 50 or 100+ viewers? Or perhaps I should be asking what specs would such an instance require? And if I attempted to self-host on my own PC (instead of renting a web host) with say a 200 Mbps upload speed from my ISP, then is that also my bandwidth cap?

To help you answer that, let's say I'll continue using my current OBS settings for Twitch on Owncast (720p, 60 fps, 5000 Kbps CBR bitrate).

@SapphireDrew You're not having a discussion in good faith. I understand you really like Twitch, and you feel obligated to them as an affiliate. I'm happy you're happy with them! But some people like Owncast for their specific use cases. You don't have to like what they like or understand why they like it.

@owncast How is wanting to understand the costs, requirements, and sustainability of taking migrating to Owncast seriously not "good faith"? There's plenty I dislike about Twitch, but if you can't even make it not seem like a financial black hole without accusing me of bad faith then best of luck I guess.

@owncast Streaming (other than social security) is my only source of income. I can't afford losses from using Owncast. And let's be real, cost is a barrier to entry for streamers.

You want to do a Q&A to get more streamers, contributors, etc. but we can only use it as streamers if we can make a profit, or at least break even, so you kinda need to address that without accusations. Entertaining people on a regular basis isn't charity work as long there's bills to pay. You know?

@SapphireDrew I'm not sure if you're being serious or purposefully fueled with trolling negativity. But I'll bite and clarify what Owncast is and isn't.

It's software. People who have a use for it make the decisions on where to run it, and how. People who love Twitch should use Twitch, use the tools that are best for you, 100%. But while you can run it as an alternative to Twitch or Youtube it's just as often used as streaming infrastructure for internal, corporate, or private events.

@owncast Let's try this again. I understand that it's software (honestly all I understand about it). What I don't understand are the requirements per viewer and how that scales (bandwidth, cpu, etc.) in order to either self-host or select three correct web hosting package (which will most likely be prohibitively expensive for me) since there's no (to my knowledge) instances I can join like here.

I'd love to switch, but I need hosting it to support any viewers I get without breaking the bank.

@SapphireDrew I think you should just take this a red flag and leave. If you can't get a straight answer on 'what do I even need to use this software' you're likely not going to get any answer about anything.

@SapphireDrew Full disclosure, I asked about accessibility support and said it was a major blocker for using it as a Twitch alternative and got the same canned 'this software might not be for you, go use Twitch' answer

@jookia @SapphireDrew Likewise, my advice is to avoid Owncast altogether. They don't care about any issues but the ones in their own heads, and don't want good faith feedback or suggestions.

They also seem to think that you can't possibly like any percentage of Twitch other than 0% or 100%.

@jookia @SapphireDrew Also full disclosure, I asked for Twitch-based OAuth support because I didn't want to manage or store passwords or make it hard for my users to switch.

@jookia Exactly, I see them thirsty for people to use it, wanting to do a Q&A with prospective streamers and then I get accused of trolling when my first questions are heavy hitters: "How badly will owning my own instance (which is required to use Owncast) affect my bottom line?" and "What are the requirements for streaming X Resolution at Y FPS with Z Kbps per viewer to serve them live video?"

@SapphireDrew TBH I don't think that's a heavy hitter, and probably easily calculated using a back of an envelope. Take your bitrate, say 6000 kbps - then multiply it by viewers. So 100 viewers means you need around 600 megabytes a second of bandwidth. If you stream 6 hours a day 30 days a month, that's 6328 TB of bandwidth a month.


@SapphireDrew However, to actually use this you'd likely need something like S3 which charges by upload and download separately. So your VPS would send a 100th of that to S3 and S3 would send the full load. So you'd need a VPS that can upload 68TB per month to S3, and S3 would cost around 6,000$ USD a month. Assuming I did this calculation right.


@jookia @SapphireDrew Wanna add that, if you self-host, latency can be a real killer even if the bandwidth is fine. This is particularly painful if your viewers are very far away from you - an experience I _definitely_ have.

@koz_ross @SapphireDrew Yeah there's also other issues, for instance CPU usage. Owncast will choose to add latency to your stream instead of dropping frames, so if your CPU is underpowered you'll end up with people like 30 seconds behind.

@jookia @koz_ross @SapphireDrew one other alternative to Owncast could be PeerTube live streaming. Because of the P2P nature, it could save you some bandwidth as the load would be shared between all the viewers. However, I didn't test with a big number of viewers. The drawback is that there is no official chat plugin.

I am using OwnCast in a very cheap server, but with no more than 10 viewers a time and it is usable, but I am not a frequent streamer, so the bandwidth costs doesn't affect me.

@rennerocha @jookia @koz_ross Hmm... that definitely seems more useful with medium to large audiences. Honestly, I average around 10 to 20 myself, but I've had viewership spikes from raids up to a little over 100 which is why I'm setting such a high target spec to cover for that growth. Also, streaming is my only source of income besides social security so the more I can stream, the better.

The social aspect is what makes it so much fun though, so hopefully PT can implement chat eventually.

@SapphireDrew Now to be fair, this is going to happen with ANY streaming service. You can't get around with the fact you have to take in some amount of data and multiply it to viewers.

I'm guessing this is what the Owncast developer doesn't want to tell you, and maybe what they mean by 'this isn't for Twitch streamers'.


@SapphireDrew That said, Owncast could fill a niche for small private streams or irregular streams that don't fit in Twitch ToS (ie, nudity or sex work) but it just doesn't seem to really want to do that. So it's unclear what it's scope is.

The vibe I get is that Owncast is a solution to a technical problem that ignores all the socioeconomic issues involved. So techbro stuff? I dunno. Maybe it should be put on a blockchain, that usually helps solve these things.

Ok I'm done, sorry for rant

@jookia @SapphireDrew That feeling when two critics give more meaningful and useful answers than the project itself.

@jookia No need to apologize, I appreciate the information.

Right now I only stream weekly, but I tend to do bonus streams too, so the hamsters in my brain were crunching what you said with my number... 5000 kbps x 6 ~ 10 hr streams x 3 streams per week x 4 weeks x 10 ~ 100 viewers (Honestly I only average around 10 to 20 mostly lurkers, but I have friends who are Partner that have raided me with 100+ viewers so I like to be prepared plus who doesn't want to grow their audience)...


@jookia So that's like, if I'm doing my math right... 600 Mbps per month per viewer... times 10 viewers minimum, so a minimum of 6 Gbps per month but could spike / grow as high as 60 Gbps per month?

And if my CFS (Chronic Fatigue Syndrome) ever magically goes away, I'd aim for 5 days a week instead of 3... which instead range between 10 Gbps ~ 100 Gbps per month?

It's almost 3am here, so I'm probably doing horribly bad math. 😅

The easiest way to work it out is get the 'seconds streamed per month' then times that by the bitrate and viewers. Make sure to add an extra viewer or two for yourself.

So 3600 seconds an hour x 10 hours x 3 streams x 4 weeks = 432000 seconds

5000 kbps x 432000 seconds x 50 viewers = 108 terabits (divide by 8 to get 13.5 terabytes)


@SapphireDrew To calculate this with cloud storage, go to , click create estimate, type in S3, click the configure button, and in the S3 standard section put in the values in BYTES, not bits.

S3 standard storage: The amount not multiplied by viewers (so 5000 x 432000 divided by 8, divided by 1024 kilobytes divided by 1024 megabytes = 257.49 GB

Multiply that by viewers to get value for 'Data returned by S3 select', which would be 12874.5.


@SapphireDrew In this case it shows up as 15.60 USD. So I think I fucked up with my previous AWS calculation by not diving by 8? Maybe things are actually cheaper than I thought? I can't figure out how I did my previous calculation to get 6000$, it just comes up at 43$ USD. Maybe I'm just really bad at math.

@SapphireDrew Ok I see where I fucked up: I assumed bitrate was kiloBYTES, so it was off by a factor of 8.

@jookia $15 a month would be amazing for that many viewers. I already make more than that in subs if I could convince them to support me elsewhere... Either way, I wish @owncast could've just said that instead of accusing me of bad faith and trolling. 😂

@jookia I think I see where you got the much higher number from. You probably used the Transfer field and I have the bad feeling that's probably what we need, no?

@SapphireDrew Hmm. Not sure. Thats in and out of EC2, so it shouldn't matter.

@jookia I found this Too tired to see how much of it applies to using Owncast but it's there for anyone who needs it.

Show newer
Sign in to participate in the conversation

Fosstodon is an English speaking Mastodon instance that is open to anyone who is interested in technology; particularly free & open source software.