InsideAMX Tertiary

Horses for courses – one size doesn’t fit all


Software v. hardware video distribution.

If you look around some recent projects in Australian universities, the biggest or most visible ones seem to involve collaborative learning spaces in some form or other.  The challenge of how to shuffle content around these spaces is worth exploring. 

It is interesting to note the different approaches taken by different universities – and the robust debates about the solutions they deployed.

The fact is there is no single right or wrong way that will suit everyone.  Some have adopted pure software solution; some a pure hardware solution; and some a hybrid solution. Each has its own merits and drawbacks. 

In the table below, I’ve listed some features of the software and hardware approaches – hybrid solution will obviously sit somewhere between the two. 

table1.JPG

 

In practice there are situations where a software solution is entirely appropriate – eg from low quality YouTube or Skype video to quite good quality video distribution.  But there are other applications where it is not - eg medical, engineering or media studies applications which will most likely demand the highest image quality and intolerant of any spatial or temporal artefacts.  

Remember that image quality is not just about resolution – there are several other factors involved: frame rate, latency, colour space and compression artefacts need to be considered.  It’s fairly easy to compress and transport 1920x1200 (or higher) at low frame rates and one or two seconds delayed.  Compressing and transporting that resolution at 60fps with correct RGB colour space and minimum latency is much harder.  As a general rule, the higher the quality, the higher the cost.

As mentioned above, hybrid solutions sit somewhere between pure software and pure hardware systems – in terms of cost, quality and reliability.  Most involve proprietary hardware boxes using the corporate network as the transport.  Some are standards-based (eg H.264 for ubiquitous videoconferencing), some are not.  Regardless, the image quality will be determined by the capacity of the network connection between source and destination – big pipes will allow high quality; small pipes and/or high traffic will mean degraded quality.  It’s all simple maths, not rocket science.