ings to Consider when Choosing a Cloud Encoder

Encoding.com was the first media processing service to see the promise of cloud computing. The scale and flexibility that cloud computing delivers is a perfect marriage for compute intensive use cases such as video encoding. Since we started this journey a lot of competitors have joined the fray with their own offerings. While there are a lot of similarities, there are also a lot of key differences. When selecting the right service, these are the elements we found most important to look for.

Workflow compatibility How do you plan to use the service? Programmatically through and API? Make sure the solution has a well-tested and mature API. Across an entire department with point and click simplicity? Look for a solution with a watch folder. From time to time for the occasional encoding job or batch of jobs? An easy-to-use desktop uploader is critical for this. These are important questions to answer before you make a selection. The most powerful encoding engine in the world won't be very helpful if it doesn't match the way in which you plant to use it.

Format Support What formats do you plan to be using as inputs and outputs? What devices are you delivering your content to? Are adaptive formats such as HTTP-Live-Streaming or Microsoft Smooth Streaming a requirement? Even if you understand your format requirements there are still a tremendous number of variants and controls that affect compatibility and support.

Broadcast Codecs DNXHD, ProRes, XDCAM, DV,DVCPRO,DVCPRO*, DVCProHD, IMX, XDCAM HD, XDCAM HD422, XDCAM EX*

Adaptive Formats HTTP Live Streaming, Microsoft Smooth Streaming, Adobe HTTP Dynamic Streaming, MPEG-DASH

DRM PlayReady, WideVine, Marlin, Native, Secure Player

Closed Captions

Inputs MP4withSidecarinWebVTT, MP4 with Caption Track in CEA-608, MP4 with Caption Track in CEA-708, MP4 with Caption Track in MPEG Part17, 3GP with Caption Track in 3GPP TT, MPEG-2 with Caption Track in CEA-608, ASF with Caption Track in CEA-608, MOV with Sidecar in SCC, FLV with Caption Track in SRT MP4 & DFXP MP4 with Sidecar in DFXP, MP4 with Sidecar in SAMI, MP4 with Sidecar in SCC

Outputs MP4withSCCSidecar, MP4 with a SRT Sidecar, MP4 with CEA-608, MP4 with MPEG-4 TT 3GP with 3GPP TT, HLS with CEA-608 MPEG-2 with CEA-608 MOV with a SCC Sidecar

Speed The typical cloud media processing workflow has 4 key steps.

Ingest The amount of time it take source content to reach the data center for processing. Make sure the solution has some form of network acceleration in place, such as Aspera's FASP. FASP leverages UPD transmission which offer greater speeds over typical TCP/FTP delivery. In addition, global distribution of data centers is critical, ensure the solution has a data center with geographic proximity to your source content. Lastly multi-threaded support for services like FTP or Amazon S3

Max Queue time This is the time it take for processor cores to become available to complete your encoding tasks. Very few cloud encoders even measure this let alone guarantee this will happen efficiently. Look for a solutions that ensure your task won't be waiting very long to begin processing.

Processing This is the time it actually takes to encode your source content. Make sure the solution guarantees minimum numbers of processing cores. This ensures that when processing begins, it happens as quickly as possible.

Delivery/Egress Much like ingest, look for UDP acceleration, multi-threaded delivery, and global proximity of data centers.

Service Level Agreement A guarantee on the level of service delivered is a critical component to an Software-As-A-Service offering. Depending on the service the components of the SLA will look very different. As mentioned before, queue times, processing cores are guarantees to look for in encoding solutions. In addition, uptime guarantees are another important component of a SaaS SLA.

What are the requirements for your media processing workflow? Let us know in the comments.

LINK:	http://www.encoding.com/blog/2015/03/31/things-to-consider-when-choosi...
	See more stories from encoding

Most recent headlines