Hacker News new | ask | show | jobs
CameraTag: Simplified Webcam Video Recording (cameratag.com)
32 points by cdanzig 4832 days ago
Just about to launch a new service to simply the process of recording videos from your users' webcams. Would love your thoughts.

-Chris

p.s. Right now we're using flash to record with an HTML interface. As soon as WebRTC works well enough we'll switch over to a full HTML solution.

10 comments

Couple of questions:

  1. do you do transcoding at all?
  2. where do the videos live, and in what format?
  3. any size/time limit on the recordings?
  4. you compared with some competitors, but not Tokbox; how is it different?
  5. is your solution dependent on Flash?
  6. is there a demo recorder we can take a look at?
Nice job!
Hey guys,

Sorry about the typos in the site copy (we're fixing them now). (this is why we ask for feedback before launching) :). To answer questions:

1. Transcoding. We transcode the videos into h.264 and generate thumbnails. At the PRO level you can have as many outputs as you would like at as many resolutions / frame rates as you want.

2. We will host the videos on AWS for you for free as long as you maintain an account.

3. 5 mins

4. Always think of TokBox as a live streaming service (i.e. video chat) not so much a recording service. I know they allow you to record chats to a server, but it's not their focus and integration is a bit more involved as a result.

5. No. Unfortunately, WebRTC isn;t quite ready for primetime when it comes to recording. As soon as it is- we'll go 100% HTML.

6. The homepage has a recorder built in but the best way to play around is sign up for a free account and put a recorder in you site (no CC necessary to record up to 20 videos)

Thanks for the interest!

I made something similar for recording HTML5 video tags which you can pipe getUserMedia to - https://github.com/wearefractal/recorder output is WebM though so transcoding is needed
Very cool!
Cool product! We've been working on this stuff (and a lot more) over at VideoView.com. Right now we're working on an entirely new video interview recording application which is kind of like a very extended version of this.

Our experience has been that one of the biggest challenges is actually in getting the average (non-technical) user to be guided through the webcam and audio setup process. There's nothing as frustrating to a user as thinking they've recorded a video, only to discover that the audio didn't work or the video went wrong.

How do you handle that with CameraTag?

Hey,

We auto detect activity (change in frame-to-frame) in the the microphone and camera. It means the user has to spend 2 seconds waving to their camera before recording but it also means they don't have to mess with camera / mic settings.

-c

Typo on lower right of front page:

"Build on Amazon Elastic Beanstalk" should be "Built on"

Also:

"All HTML interfcae is 100% customizable." should be "Our HTML interface is 100% customizable."

I'm getting a popup on the homepage saying "Please upgrade your account"
LOL.. Forgot to upgrade our own account :)
Nice, I'm looking for a solution like this for an app I'm building. Didn't like the current offerings too much. Will check it out when I'm ready to add in video recording.
If I can be of any help when you're ready let me know. Chris@

-c

Love your work. Had to work on a video capturing project 18 months ago that nearly killed me. Zencoder reduced some of the pain but we still blew months on the project.
We started off building this infrastructure for another project we were working on. We quickly realized that we could / should encapsulate our pain / learnings into a simple SAAS solution so others wouldn't have to waste their time the same way. :)
Nice work. Can you talk about the technology stack behind this? I am very interested in knowing about that.

Thanks

Hey Obaid,

Sorry for the delayed response. Don't know how I missed you question. Happy chat about the tech:

For now the recorder base is a flash app that has a JS API. We built a stock HTML interface that consumes that JS API but our users are free to replace or exclude that HTML if they so please. As soon as we can reliably record a WebRTC stream we will swap out the Flash :)

The server side is built in Rails running on AWS (Beanstalk) hosted on Passenger and NGINX. In addition to the web interface, There is a REST API for accessing info about your recordings. We can generate h.264s and png thumbnails of your recordings in any resolution and frame rate you want- our transcoding is handled by Zencoder and/or Amazon Elastic Transcoder (depending on requirements).

Hope that helps- did you have any specific questions?

It would be helpful to see a browser support list, with mobile browsers listed too.
Hey Lucas,

CT should work in all major browsers that can run flash. Right now (because of the flash dependency) most mobile devices are not supported. As soon as we get WebRTC polished we'll be 100% HTML and will work in most modern mobile browsers.

-c

doesn't record audio?
You probably had line-in or something else selected.
supports video + audio