ARD/ZDF Tackles an HD Summer
Higher capacity recording comes at the right time for the broadcasters
October 6, 2012
ARD/ZDF EVS Administrator Patrick Jung amongst some of the 17 EVS XT3s at the IBC.
LONDON -- German public-service broadcasters ARD and ZDF take technology
and the advantages it brings to a broadcast seriously; but when broadcasting
something as important as the London Games, technical changes are taken
In Beijing, they made the jump to
tapeless, though still working in standard definition. Less than two years later the full leap
to HD tapeless production took place in Vancouver. In London, HD production met
the sheer tonnage of events that comprise a Summer Games.
Servers were again central to the operation. They were used in ingest,
playout and production at the International Broadcast Centre, and mostly
in a highlights capacity at various venues where the broadcasters had a
“It’s our first HD Summer Games,” said Patrick
Jung, EVS administrator for both ARD/ZDF’s Vancouver and London broadcast
teams. “It’s HD files with the volume of a Summer Games.”
Two things helped ARD/ZDF manage the volume of all those recorded HD
files. One was simply an increase in the SAN built up around the core,
17-machine EVS XT3 infrastructure in the IBC.
the XT3’s themselves were the very latest units. Jung said the latest versions
of the machine had 900 GB hard drives in a box with the same footprint as the
old machines, which had 300 GB drives. Jung said the new machines could hold
300 hours on each machine: “This is three times the hours in the same physical
volume of events and managing all those big HD files, ARD/ZDF’s workflow was
not radically changed from Vancouver. In London, ARD/ZDF recorded and logged
incoming feeds on the EVS systems, with material transferred on to Avid/Isis
storage for editors to create finished pieces. If the programming schedule
required speedy airing, material was played out directly from EVS servers.
The broadcaster used the latest version XT3 machines, which have 900 GB
drives capable of storing 300 hours of HD material.
There were incremental changes to the workflow and
infrastructure. In Vancouver a few EVS channels had recorded a 24-hour loop of
all incoming feeds, a way to give the production team a second chance in case
they needed material they had missed. In London ARD/ZDF continued this
practice, but also deployed five stations of the IPBrowse service from host
broadcaster OBS. This service allowed rightsholding broadcasters (RHBs) to look
at lo-res browse copies of content before picking only needed content for
HD-res delivery to an EVS server in their facility.
What other software was deployed in London? EVS IP Director
software had browsing and administrative functions, and was also used for
The Germans had an interesting use for IP
Edit software, which is often used as a quick “rough cut” editor by
broadcasters. At ARD-ZDF, however, it was used in an audio capacity as a
supplement to fast-turnaround linear editing using traditional EVS edit
controllers. The video playlist is taken in to an IP Edit and the audio -- off
tube commentary and more -- is quickly added.
One software improvement proved quite helpful in
managing a “big event” broadcast like the London Games. Jung -- whose title is
EVS Administrator -- said that the administrative software to manage and
configure a networked EVS infrastructure had been improved since Vancouver.
Before, an administrator had to go to each machine to check parameters.
“Now, one software can check all the machines for stream
status, temperatures, all the things that need to be monitored,” said Jung.
“It’s peace of mind.”
As for other “up and coming” host
broadcaster services, Jung said the host broadcaster service called the
Broadcast Data Feed was not used in London, but would be given serious
consideration for next time. This feed had within it all the logging data from
the massive OBS pool of EVS servers, and more, as metadata tied to the video
timeline. It even provides clip duration and captures key frames.
Tracking and searching for content in a high-volume big event
operation is critical, and the increased visibility on a clip-by-clip basis
adds value to host assets, making desired content easier to find and providing