The Core Cloud platform is an incredibly powerful IP video tool, but it can feel overwhelming at times, especially if you’re new to streaming video. In this post, we’ll go over some best practices to get the most out of your Core configuration, while highlighting a few key methods for cost-cutting and increasing the reliability of your stream. So let’s get started!
What exactly is Core?
Think of Core as a toolbox for IP video that lives on the Internet. It does all sorts of things, such as recording your live streams, transcoding video feeds, routing streams to multiple destinations, and so on.
What are Engines?
Engines are just servers that offer different functionality for your encoders. We discuss the 3 different Engines available on Core below.
What is a Sputnik and is there a difference between On-Demand and Dedicated?
Sputnik is just a fancy term for our streaming server software that sits in the cloud. Sputnik servers are the basis for many of the things you can do in Core, including multi-destination streaming, remote configuration, and bonding. A Sputnik server is required for general operation on our cloud. We have implemented these servers in a few different ways:
On-Demand: Everybody gets an On-Demand server automatically when their encoder connects to Core. You don’t even have to think about it. These servers are designed to be efficient so they spin up / shut down as the user needs them. The IP address of the server is dynamic so it may change every time a new On-Demand server is spun up. If you require a fixed IP address, take a look at a Cloud Sputnik. On-Demand Sputnik servers cannot use the Share Stream function for CoreTV.
Cloud-Hosted Sputnik: These are nearly identical to On-Demand servers in functionality, but have a different implementation. Each instance supports up to 5 simultaneous encoders, allows the “Share Stream” function for CoreTV, and comes with a dedicated IP address, should you need it. These are generally more reliable than On-Demand servers because they are a dedicated resource just for you. Cloud Sputniks cost $75 / month and run 24/7 during that time. Data streamed through these servers is billed at the same rate as On-Demand servers, with your per GB price determined by whatever plan you’re on.
Self-Hosted Sputniks: Also identical to the other two servers, these are for installation at your data center or facility. Self-hosted servers do not come with data charges because the customer is providing all of the bandwidth. However, these servers are leased for a monthly fee and require separate fees to keep them connected to Core. If you’re a business with a dedicated gigabit fiber line and plan to put a lot of traffic through Core, this is a good option for you. For more information, please reach out to Jon@teradek.com.
When using Cloud or On-Demand Sputniks, you must choose a region for your instance to run on. It is a best practice to choose the region closest to the location of the encoder(s) you’re connecting. This will ensure lower latency and, in general, a more reliable connection. You can always change regions should your local region experience connection issues.
So now you know all about Sputnik servers and how they function as the heart of our streaming technology in Core. Let’s talk a little bit about our Transcoding engine.
Transcoding for VidiU Go
We enable the transcoder automatically for VidiU Go users. We do this because RTMP(S), the only protocol VidiU Go supports, is not compatible with HEVC so the video feed must be converted to H.264 for delivery to an online video platform. Adobe, the creator of RTMP, has abandoned the protocol so no further updates are expected, which means we will likely never see HEVC support. As a reminder, the transcoder is billed at $3 / hour and will spin up and shut down automatically. The bitrate delivered to your destination is a calculation based on several factors - the user does not have manual control over the transcoded bitrate at this time.
Note: if you are using a Cloud Sputnik with VidiU Go and require a Transcoder, you will need to spin up a separate Transcoding Engine (see below). Automatic transcoding only occurs on VidiU Go’s when they are connected to On-Demand Sputniks.
Transcoding for our other codecs
If you have a Prism, Cube, Slice, T-Rax, or Bond, then you need to manually set up a Transcoder engine should you require it. It’s billed at the same rate as VidiU Go’s transcoder ($3 / hour) so if you’re not using it, make sure you turn it off! Of course, you may not need a Transcoder at all if you plan on streaming point to point with HEVC compatible protocols (SRT, MPEG-TS).
Note: If you want to use both a Cloud Sputnik and Transcoder Engine, you will only need the Transcoder Engine. In this circumstance, the Sputnik and Transcoder live on a single instance so you will get both functions. (Yep, it’s confusing, I know). Like the VidiU Go automatic transcoder, the user does not have manual control over the Transcoded bit rate at this time.
Keep in mind that it probably doesn’t make sense to run the Transcoder full time. At $3 / hour, it will be considerably more in 1 month (running 24/7) than a dedicated Cloud Sputnik at $75 / month. Use the Transcoder only when you need it. Remember to shut it off when you’re done (Core will eventually turn it off for you if it detects a long period of idle time, but you will still be responsible for the hourly charge until then).
There is only one Engine left to discuss: Helios. Helios is our latest Engine and offers a set of unique features not available on Sputnik.
What is Helios and why do I need it?
You should think of Helios as a server that provides extra features not found on Sputnik. The two key Helios features are:
- Generic Ingest: bring in video feeds from other encoders via RTMP, SRT, MPEG-TS
- Failover: setup multiple video feeds as a cascading failover should one feed fail
Do you need Helios? Well, that’s up to your unique workflow. It’s a great set of features that can provide reliability (failover) and more flexibility (generic ingest) for your live streams. The cost is $75 / month per type (Generic Ingest OR Failover) and this server runs 24/7. Even if you shut it off, there are no cost savings - you are billed for the month so make sure you make the most of it. There are no additional data charges when streaming video moves through a Helios server.
Note: If you require 4x 3rd party encoders to be ingested into Core then you would need 4 separate Helios engines.
What is the Hotspot data charge I see on my Core plan?
Hotspot data refers to the price per GB of data going through Core from our Link Pro device. This does not have anything to do with cellular data you might be using through your Node modems -- that data you must purchase through Telna or your desired cellular carrier(s). We will be updating the term “Hotspot” to better reflect that this only applies to Link Pro.
We’ve also been asked why we charge so much money per GB with Link Pro ($5 / GB). The reason is that this product was designed for generic data usage: surfing the web, chat, email, etc.. - not for video applications. General web usage generates, in general, fairly low data usage so the price is commensurate with our infrastructure costs. If you need a bonded video solution, there are better options from us, such as Bond or VidiU Go.
CoreTV and Share Stream
Share Stream allows you to securely share your video feeds with colleagues for remote-monitoring sessions. To get started, you must connect your encoder to a Cloud-Hosted Sputnik and then create a new Workspace with any of the video feeds you plan to share (you can create as many Workspaces as you like, each with their own set of video feeds). From the Workspace, select the Share Stream icon near your Workspace name to invite colleagues.
Viewers can monitor the feeds using the CoreTV application for iOS, Mac OS, and AppleTV, as well as VUER for iOS and MacOS. Additionally, your viewers can also monitor using a web browser and VLC, but both options have higher latency. We will have an update for Windows and browser-based viewers later this year that will provide extremely low latency streaming.
Devices compatible with Share Stream: Cube, Serv Pro, Bond, T-Rax, Slice, and Prism
A couple of things to note about Share Stream:
- CoreTV apps can monitor streams in 10-bit, 4:2:2 quality, provided your encoder is delivering that to Core
- All CoreTV and VUER streams are encrypted using the SRT protocol
Codec limitations in Core
Not all encoders are treated equally on Core. The primary example is VidiU Go, which can only do the following things on the Core platform:
- Remote configuration
- Multi-platform streaming (RTMP / RTMPS only)
- Automatic transcoding (HEVC to H.264)
- Compatible with Cloud Sputniks and Transcoder engines
VidiU Go cannot use the SRT or MPEG-TS protocols or make use of the Share Stream function (i.e. CoreTV). Why? VidiU Go is designed for live streaming only (RTMP). That’s it.
If you need full Core functionality, Cube, Prism, Bond, T-Rax, Slice, and now Serv Pro are your best options.
Saving Money with Core
We know Core is complex so here are a few tips to keep your costs down:
- Core will begin charging you for small amounts of data usage the minute your encoder(s) connect to the platform. We do this because the encoder instantiates a server to allow remote configuration from the cloud, even if no video data is being streamed to the platform. Turn your encoder off or disconnect your Core account from the encoder(s) if you do not want to be charged during idle time!
- HEVC will keep your bitrate lower than H.264, which translates into lower data costs. If you’re using SRT or MPEG-TS and your encoder(s) and destinations are HEVC compatible, we recommend keeping HEVC enabled.
- If you are a power user and plan on streaming terabytes of data or for extremely long periods of time, we recommend using an On-Premise Sputnik server to keep your data costs reasonable. To get started with one of these, reach out to our Sales team to discuss your options: Jon@teradek.com
- Paying for a Cloud Sputnik might cost you more monthly, but may end up saving you much more in the long run due to its reliability. We know this option is not for everyone, but if you’re using Teradek for your business, Cloud Sputnik’s will deliver the consistency your customers demand. I like to think of this as an insurance policy; when failure is not an option, I would invest in a server that increases reliability.
- As mentioned above, once you spin up a Cloud Sputnik or Helios Engine, you are charged for 1 full month. Shutting down the server prematurely will not save you any money -- just make sure you shut it down before your month is up to avoid a second month of billings.