What approach did you settle with for your exports?
The next main thing will be improvements to the timeline. We want editing on Tella's timeline to feel much more approachable than "traditional" timelines.
Interesting, ended up using FFmpeg. It's heavy to run and it takes time to write all the code for overlaying elements on top of each other and drawing text etc - but the quality is great.
- We don't have this. But we have been thinking about a more dedicated recording "mode". However, because we're in the browser we have some limits on what you could annotate. Inside of Tella would probably be fine, but outside of Tella would be tricky (e.g. you're recording another window).
- We support exporting to mp4. So you'll get version of the video that you can distribute (beyond just sharing the link to Tella). We don't have plans on a slides-only export though.
- We use ReasonML, Mux (for video), and Roomservice.dev (for the real-time component).
Just curious about how you technically do the export to MP4. Can you share any details? Do you play it back in real time and record it, or do you have a way of exporting it quicker than real time?? What technical approach do you use if you don't mind me asking? Thanks!
Quickly summarized we use puppeteer headfull with docker to do a recording of the video. So you’re right, it’s playing it back, not faster than real time. I might write more about the specifics in the future. It’s mostly open source code, you can start here: https://github.com/Ventricule/html2screen
Thanks for the reply. I was wondering if you were using the headfull chrome approach! I for one would definitely love to read some more specifics on this in the future when you have time! (And if there's anything you can open source please do!). I did a quick side project proof of concept of automating recording a webpage with video and other elements a few months ago. It's a shame there's not a way of just using headless chrome but my research let me to the headfull docker approach, so I'm happy to hear I landed on the same approach you did! Did you discover any novel tricks beyond what's been written about/open sourced already? What was the trickiest part for you?
No worries! Html2screen was a very good start, I modified it to support audio recording and got it to a pretty good state in a couple of days. After that the hard thing was just to run it automated on AWS. I used AWS Batch and Step Functions for that. When it’s a bit more stable I’ll definitely write about it and might open source it later. If you want to discuss more, feel free to email me at michiel@tella.tv
I've been using Reason for a couple of years in personal projects and I really like its fast compiler, type system, things like variants, modules. I didn't have OCaml experience before but to me it feels like being able to program in a very mature programming language while still being able to target JS. The community is great as well. Also refactoring code has never been easier in my opinion. That's why I also decided to use it for Tella.
I've used TypeScript a lot in the past as well, and I'm not against it at all. You can do a lot of the things you can do with Reason in TypeScript. But I do have the feeling Reason makes me a lot faster than when I worked with TypeScript.
It sounds like Tella could help with at least some of that already. We definitely don't have everything you'd find in a desktop editor yet, but it's the plan to have more!
Tella is free during beta (which will likely be for a few more months). We'll introduce paid individual and team plans after that (and keep a free tier). However, we don't have figures on the pricing we can share yet.
"But the demo video doesn't look like much beyond a web based PPT tool. I'm struggling to see why I would use that over a generic tool like Google Docs."
Appreciate this observation. For people who've not done much video editing before we want Tella to feel as familiar as possible (e.g. editing a slide deck). Equally, we'd eventually like people who do know their way around a video editor to feel like they can create what they want in Tella.
Yes, and I definitely got that from your write up. It's a huge opportunity.
I'm not sure how easily you can add the things I'm excited about into a demo video, namely the idea that you can post process video in a browser. The things you describe, synchronization, etc are really time sinks even when you know what you are doing and have the right tools. If you capture 10% of the functionality and make it automated and simple, you'll have a winner. Maybe you already have this, I haven't played with it, sorry if I'm ignorant about the true offering!
Yeah the demo doesn't illustrate the MediaSync aspect very well. But if you give the product a go you ought to get a sense of what we described in the implementation part of the write-up.
Awesome. Tella should be fine on Chromebooks, but we have an issue with the performance of recording in the past. If anything comes up drop us a line on livechat.
If it's worth anything, my impression is that the CPU (and probably GPU) performance of chromebooks varies widely between models. It seems to be particularly dramatic when comparing ARM models against Intel and AMD models.
I don't doubt that low-end CPUs could cause dropped frames during video recording.
If I had to offer any advice, it would be to test as many models as you can and then decide which ones to officially support.
Yeah we have had reports from one other user that his Chromebook sometimes has issues recording both webcam and screen at the same time. I definitely intent to test as many devices as possible and put up a more official page about that.
Not yet, I'm afraid... we'll be free while we're in beta (which could easily be a couple more months). From there we'll keep a free tier, and introduce paid individual plans and a team plan.
I personally find lack of pricing a big barrier of entry for products I'm even willing to evaluate. Even if there's a good chance they'll be free for my use case.
I definitely understand how hard pricing is, but I think you'd be much better off taking a stab at something than leaving it completely empty.
The most fair/honest strategy in my opinion is to set some pricing in there, and grandfathering people at this price point. It gives you plenty of room to change pricing later without upsetting your early adopters (who can also be your best marketing channel).
Just my 2 cents as a co-founder of a company that introduced paid plans pretty much from day one.
The next main thing will be improvements to the timeline. We want editing on Tella's timeline to feel much more approachable than "traditional" timelines.