Hey! Alex here from Connectify! IRL Live Streaming is a lot of fun, but it
can be expensive to get the equipment necessary to keep a stable HD stream going while you’re
out and about. One of the most popular pieces of hardware
for this (and on the low end of the potential budget range!) is the $995 LiveU Solo, featured
in the official $1600 Gunrun IRL Streaming Backpack. Let’s see if we can use new Speedify for
Linux technology to assemble a similar working setup for a fraction of the price- For this IRL streaming setup, we are using
the Raspberry Pi as the centerpiece. We want to stay pretty close to a standard
IRL streaming setup on the outside, keeping cables as concealed in the backpack as possible. We also want to be able to use a standard
HD video camera, so to make it work with the Raspberry Pi we’re gonna need a capture
card capable of transcoding the raw camera footage into a format that the Pi can upload. Then, we’ll need to program a button that
can start and stop the stream. Finally, to keep the stream stable and reliable
while moving around outside, we want to incorporate channel bonding so we can send data out over
multiple 4G ]connections from different carriers at the same time. And the plan is to keep the total budget as
low as possible. Let’s get started! We decided to go for the Sony AS-300 for the
camera, since it’s a nice portable, wide angle, HD camera and a popular choice for
streamers. At $298, it will take up the biggest chunk
of our budget. You should be able to swap it out for any
other camera that works with a USB live capture card. We’ll hook it up to the pi with a micro-HDMI
cable plugged into an Elgato Cam Link capture card, adding $116. Next, we’ll use a Raspberry Pi 4 for the
computer, for an additional $55. To keep it cool and protected in the backpack,
we purchased a $10 GeeekPi Acrylic with Fan RPi 4 case. We’ll also need a battery to power the whole
setup, which we purchased for $38. To control the stream, we’ll use a Circuit
Playground Express as a remote, which we picked up for a total of $30 including the plastic
case. We’re holding everything together in a Cocoon
GRID IT! Accessory Organizer we got for $11, and we
decided to mount both the camera and the Circuit Playground remote on a $28 Smatree telescoping
selfie stick. For the backpack, we’re just using a Jansport
backpack we already had laying around. Finally, we need our mobile connections. You can borrow some phones from friends to
tether or plug in a few cellular adapters. Here, we’ll be using a couple of phones
plugged in via USB. This brings the total budget to about $600,
which is still about $1000 cheaper than the Gunrun backpack, and you may not need to purchase
all of this equipment if you already have similar items available. To get everything working for IRL streaming,
we installed Speedify and FFmpeg on the Pi. I’ll link to more detailed instructions
in the description below. We programmed buttons on the Circuit Playground
Express to start and stop ffmpeg streaming with a Python script that runs this hefty
command: This command lets us stream live to Twitch,
in 720p at 24 frames per second, which is about as good as the Pi can handle. We’ll have our code to set up the streaming
buttons available in a link in the description below. Speedify is set to automatic start so it runs
as soon as the Pi boots up. This way, we don’t need a UI for the Pi
once we get everything together. So let’s test it out! How long are we? What Time is it? 1:26 44 minutes. All right. Hey, this is Alex from Connectify, so this
is the first test, just streaming live, of our new IRL Speedify streaming backpack. So we’ve got a Raspberry Pi, we’ve got
2 tethered cell phones, and we are just walking around the city of Philadelphia. So I guess we are 45 minutes in - I can’t
see it but I’m hearing from people that it’s been working well - so we’ll see
it looks like a successful test so we’ll be both uploading the video clips as well
as the instructions on how we built this thing and how to use it. Let’s go back to Connectify’s office! And we’re back. I gotta get back to
the lab. Thanks for watching!