Defcon Glasgow Youtube Playlist
Obviously today I am making this page there is just one video in the Glasgow Defcon playlist. But here is the link to what should become the ongoing saga:
Mission Statement
In May 2019 I started to record videos for Glasgow Defcon (DC44141). Mostly this was because I was inspired by the legend that is Cooper who I saw flitting around at BSides Scotland for the last two years. Check out all the videos he records for the infosec community here:
I saw the wealth of talent on display from his videos of Securitay, Le Tour De Hack, and BSides Scotland. Thought maybe Glasgow Defcon can attract more speakers if they know they have the option to get it captured?
There is great talent in the student societies and a link to a video on a CV is, as a hiring manager, very much of interest. Speaker spots at big conferences are harder to obtain than local meetups, and people need to practice somewhere first right? These are the reasons why I looked into the bare minimum technology required to capture an audible talk. The rest of this page is about how I did it. I hope others can replicate or improve on it.
Note: I know absolutely NOTHING about how to do any of this or what most of the AV terms mean. I am a genuine noob who just fancied trying it. Always find new things to be a noob at and you will always be learning 😀
I want to do the same, how did you?
I had three goals:
- Capture the HDMI source from the speaker’s laptop for crystal clear visuals of slides and any audio that is played. [required]
- Capture an audio feed of what the speaker said and the general noise of the room when people ask loud questions. [required]
- Capture video of the speaker in the room. [optional]
Seeing the speaker in video is nice to have but being able to read and hear the content they provided was my priority.
What hardware was used?
Here is the hardware that I used with indicative first hand costs. Beg, borrow, and ebay and you will get the price down significantly from there. Even first hand the kit is obtainable at around £200:
| Item | Cost | Why? |
| AVerMedia Live Gamer Portable Lite | £80 | Plug HDMI from Speaker’s laptop into this, USB into your laptop, and an HDMI out to the projector/TV at the venue. It man-in-the-middles the HDMI basically. The crucial bit is it hardware encodes the stream to save your CPU and Disk Space on the fly. This is last generation tech and is USB 2.0 even but it worked fine. |
| Snowball Blue Microphone | £50 | I already had one so I didn’t pay out for this. The mic has a couple of modes with one being conference calling. It should pick up someone speaking behind the direct focus in this mode. Or you can pop it full forward to catch just the speaker. By putting this on the table next to the speaker who is closer by around a metre than anyone else in the room it performed adequately for my needs. Not professionally crystal clear but pretty good. |
| Logitech, C920 HD Pro Webcam | £50 | Pretty popular webcam for youtubers making it simple. It films in 1080p and happened to have microphones. If my primary microphone messed up then this offers a poorer but better than nothing audio feed. |
| Camlink TP330 Tripod | £15 | To elevate the webcam off a table with more stability. Allows more control over the angle of the shot and you can easily move sideways if you have to. |
| 2x 5 metre HDMI Cables | £20 ish | One from Speaker’s Laptop to the capture card and another from the capture card to the projector. |
| 5 metre USB cable | £5 ish | To connect the Snowball Blue at a distance far enough away from my laptop. |
| No Residue Gaffa Tape | £6 | With all those long cables crossing the floor of a pub it was a no brainer to try and tape things down. By paying for no residue stuff you don’t wreck the cables or the floor of the venue. This was a direct tip from Cooper’s website and I do love this stuff now that I have used it! |
With this setup you also need a laptop from which you control the recording process. Due to the HDMI capture card you do not require a super fast CPU. While my laptop is an i7 beast, the recording software said I maxed out at about 15-20% utilisation during the recording.
There are several other ways to do things, but this is what I was comfortable paying for at the time.
What software was used?
One of the best parts of doing this was discovering the excellent open source projects which enable so many stories to be told. I did not pay for a single bit of software. Everything was intuitive and nothing frustrating happened to me! I could be lucky… Hope you are too!
Open Broadcaster Software (OBS) – Recording Software
To record the actual event I used Open Broadcaster Software. When you open that up you can create various “scenes”. Each scene can include various visual, audio, image, or textual components. Want to add an audio feed? Hit the “+” and select a microphone that your OS can see. It really was that simple. I setup three scenes:
- Full speaker view – showing the webcam maximised with a text insert for the talk title and speaker’s name.
- Full HDMI view – the inverse where only the HDMI source was visible and the webcam was not but otherwise the same.
- Half and Half – a mix with the HDMI source approximately 2/3 of the screen and the webcam making up a smaller portion.
Common to each scene was the audio sources from the AverMedia capture card, the snowball blue and webcam microphones. No matter what was on display this captures consistent audio.
While you are recording you literally click on the different scene names to cut between them. Feel like a director for the day eh?
The “gotcha” for this software for me was to work on ensuring that all audio streams are being captured on different channels. I found this tutorial which set me on the correct path:
I wanted to capture all audio at a good level and then worry about mixing them later.
My settings ensure that I record to “.mkv” with all audio sources in separate channels. The GUI tells you to flip to .mkv when you add multiple channels.
Someone, somewhere on the Internet said use “.mkv” over “.mp4” because the “.mkv” is more robust should the software crash you still have what was recorded to that point. I didn’t have a reason to question that logic and it worked for me so far. I will absolutely trot this sentence out in real life to make it look like I know what I am talking about.
As a side note this software is fully capable of live streaming to Facebook and YouTube. It was really easy to set this up so I have played with that. The problem is you then open yourself up to having to mix the audio correctly during recording.
I chose not to live stream Glasgow Defcon since the ability to do so relies on the venue’s WiFi and that is a moving part! But this is a great capability of the software that you might be interested in if you are following my steps.
Audacity – Audio Trickery Software
After hitting “stop recording” I used the “Remux Recordings” option in OBS to convert the .mkv to .mp4. You can then use Audacity to extract all audio channels you recorded into individual tracks which you can then select and edit. This process is shown here:
I picked out only the Snowball Blue stream as the HDMI source had no audio for my first recording adventure. While the webcam’s microphone did have an audible version of the talk it was far worse than the microphone right next to the presenter. Given the distance from the camera this was as expected. A not completely rubbish backup though.
I didn’t need to clean up the feed much as the talk was audible throughout but audacity lets you lower or boost parts of the audio as required and you can try fancy tricks to remove background noise etc. All I needed to do was mute the other audio tracks and leave the snowball blue microphone unmuted.
Go ahead and export the audacity project to “.mp3”. You will be guided to install “lame” mp3 encoder to do this. Note the warning that the “file -> save” is NOT an mp3 file but an audacity project. Find the “export” function to do this.
OpenShot – Video Editing Software
By this point you have a “.mp4” version of the talk and a “.mp3” of the audio. These should be exactly the same length and therefore in sync. If you have edited the length of the audio then good luck to you!
What I did in openshot was to import these two files into the list of content. Then dragged the video onto a track. Right click on that track to find the mode to split out the audio channels and then deleted those (since we have in theory mixed a single mp3 stream we want to use).
At this point drag the “.mp3” into the track under the video. These should be the exact same length.
Use the scissors icon to snip the start and end of the talk as appropriate.
I then exported the video as “.mp4” with the settings predefined for YouTube HD. This process is where you need a good CPU and it will take some time to complete. Probably could optimise this part somehow but I don’t know anything about this bit.
I would not dare volunteer to record a conference like Cooper. But a free event down the pub with your mates where the worst that happens is someone who couldn’t make it missed a talk they missed anyway? Probably alright, the stakes are not high.
Hopefully this makes sense and can be of use to you. If you are involved in a local meetup I would say to you; if this noob can do it, so can you.
cornerpirate.com 