Control Group Blog

First off, Buzz has some serious potential. Google gained an incredible amount of ground on Facebook and Twitter with this launch, and I do sense a shark-jumping moment for Facebook. Once the Google App ecosystem takes off, and social games and e-commerce get integrated, there will be a huge erosion in Facebook market share.  People want one thing, one place to go, and Gmail is already mandatory.

1. Lets start off easy here: Mobile. Google, you own the platform, how hard would it be to launch with an Android app? In the time it took to do the marketing piece on the mobile site, Google could have developed a full-fledged app. Instead I have an “above ground only” slow-loading mobile web page. HTML5 isn’t quite here yet – and Android 1.6 is not supported.
2. Two way integration! Getting Tweets in Buzz is great, but I still have to go out to Twitter or TweetDeck to post. If I had the option to choose which networks my updates appeared on from within Buzz (Facebook, Twitter, LinkedIn, etc), I would never have to leave Gmail… except for…
3. Where is Facebook? They have a strong API and a straightforward authentication service. Connect to Facebook and never make me go there again. Aggregate and publish (see above).
4. Buzz for Biz. I know it’s coming for Google Apps, but get it going already! How about full integration with LinkedIn – a CRM app would be a really interesting mash-up, as well as bringing all my connections into my address book. No more stale email addresses or phone numbers.
5. OK, I saved crazy for last. Google Profile. I am the strongest believer in an open information society – if everyone knows everything about everyone, then we are all equal. I understand that’s a little overboard, but why should I hide my information when it will only make the web more relevant to me, and get us to our ultimate destination quicker.  But with Buzz, your Google Profile went from obscurity to super relevant. So, quick inventory: Google has information about my friends, my browsing/search history (not to mention DNS info), my purchase history, my communications, the news I’m reading, and my documents. Now they want to know my age, sex, where I grew up? Are you crazy? Google is holding ALL the cards now. They keep repeating “don’t be evil” but you know what they say about absolute power… (did I mention my location?)

If history is any indicator of the future, Google will evolve and add features and services. They have done so consistently since inception, and people will choose convenience above all else, including quality and privacy. And I guess I’m one of them. See you on Buzz.

And if you disagree, see you in the comments!

Control Group has rolled out a number of Call Manager Express systems for our clients, and one issue that frequently comes up is what to do with callers on hold. While most companies – ours included – work to avoid putting people on hold, it does happen. The easy options offered by the Call Manager Express are a double beep at regular intervals or a ‘Music on Hold’ file that you can store in the host router’s flash memory. I have received many complaints about the beeps and even a single repeating music file can be insufficient for some users.

When I started searching for a way to connect an audio feed to a gateway router, I found that there were no great turnkey solutions. What I did find is that there is a way to make use of an available FXO port to bring a live feed into the system with some inexpensive parts. The solution I have deployed requires a surface mount RJ 45, a chassis mount RCA jack, and a stereo 1/8″ mini to RCA adaptor. All of these pieces can be found at a nearby electronics parts store and purchased for under $20. The FXO port and an iPod will set you back some but if you want easy control of your hold music it may be worth it.

Build the Adaptor

To make the adaptor, open up the surface mount box and figure out how you will attach the RCA jack. In one case I used a drill bit to make a hole for it, in another case I found a form factor where the notch for a cat 5 cable fit the connector perfectly. You may also find a general purpose surface mount with modular fittings where an RCA “snap in” would snap right in.

Once you’ve sorted that you will need a pair of wires from either a solid or stranded piece of cat 5. Punch these down to pins 4 and 5 of the RJ45 and solder the other side to the RCA jack.

When you are done it should look like this.

Adaptor wiring detail. Though the jack in this configuration is an RJ 45 you can use a standard RJ11 phone cord to connect between the adaptor and the FXO port.

Configuring the Voice Gateway

Configuring the voice gateway is pretty straightforward:

1. Create a DN for the MOH call out with an unused multicast address

ephone-dn 274
number 1 no-reg primary
moh ip 239.12.13.1 port 2000 out-call 899

2. Configure the available FXO port remain open to the iPod

voice-port 0/3/3
signal loopStart live-feed
input gain 2
description To MoH Live Feed

3. Create a dial peer to connect to the FXO port

dial-peer voice 55 pots
destination-pattern 899
port 0/3/3

4. Configure CME to use the multicast for MOH

telephony-service
multicast moh 239.12.13.1 port 2000

Then Shut/No shut the voice port to invoke the change.

I have been using one of the built in USB ports on our gateway here at Control Group to power the iPod which turns out to be very convenient.

Safety and Other Considerations

There are some very important caveats.

1. The adaptor I describe is potentially dangerous. If you were to connect an iPod to a POTS line or an FXS port feeding a battery signal you could damage the iPod or the device connected to it. You could also receive a serious electrical shock.
2. The use of copyrighted material for Music on Hold can be a violation of the rights of the copyright owner.
3. The adaptor will only work with one channel of audio from the iPod. If you use a mono 1/8″ to RCA cable it will short out the other channel.

These issues need to be minded if you would consider this solution. Also, keep in mind that people’s taste in music varies widely. Implement Music on Hold and you will find out just how much!

Miles Green is a Senior Network Engineer at Control Group and worked as a Musician and Recording Engineer in a previous life.

Apple’s venerable Apple Software Restore (asr) tool includes the insanely useful ability to image a nearly unlimited number of network clients. It accomplishes this via a router’s ability to broadcast data to any number of clients simultaneously from a single IP address. Known as multicasting, this allows even a modest computer to image a hundred Macs with 35GB images in a single fell swoop.

An ASR Restore Image in Disk Utility

The disk images asr works with are the same format used by Mac OS X’s Disk Utility. This means you can do a rollout over the network and keep the master file on hand in your re-imaging kit, should one of your workstations run into trouble and need to be re-imaged over FireWire. To ensure the sanctity of the final result, disk images include an embedded checksum which is automatically verified during the deployment process. This can be a significant advantage in using asr over of Apple NetInstall, which requires its own folder-based setup of restore source files.

Disk images are also hardware agnostic for the most part. You can build your image on a Mac Mini and apply it to anything from a PowerMac G5, to an Macbook Pro so long as it can get on the same subnet as the asr host.

The usual caveats of disk imaging apply unfortunately. You’re going to have to sweat individual serial numbers if you don’t employ network or volume licensing. Settings like hostnames and non-ubiquitous local users will require individual workstation visits without centralized management. However, having asr around to do the heavy lifting means you may be able to turn a strenuous two day deployment into a breezy one day affair.

Last week, I wrote about leveraging Final Cut Server as the core of a platform for asset management, approvals, and delivery. I wanted to append that post with some recent thinking and findings.

First, a quick discussion of semantics:

When you ingest a media file into Final Cut Server, it creates an asset. This asset is actually a collection of files, including the original media file, or primary representation. If the file being uploaded is a graphics or video file, Final Cut Server creates additional representations: a poster frame and a thumbnail. These are used to represent the asset’s media within Final Cut Server. If the file being uploaded is a video asset, a clip proxy representation is also generated, used for viewing the file within Final Cut Server.  This representation is created by transcoding the primary representation file to a lower-resolution codec.

Asset Representations in Final Cut Server

So a video asset is actually a container that’s made up of a bunch of files. This is pretty cool, and it’s mostly transparent to the end-user in Final Cut Server. However, while these representations are customizable (codecs, quality, etc), they all generated on ingest. Final Cut Server doesn’t currently support a way to create a new representation of an asset on-the-fly and have that representation become part of the asset container.

I’m struggling with this limitation as I explore Final Cut Server/Episode Engine integration. It would be ideal if I could setup Episode integration via a Final Cut Server copy response to an Episode watch folder, and have the resulting transcoded file copied back into Final Cut Server and made a representation of the original asset. Right now there’s no way to do this, so instead we are only able to re-ingest the new transcoded file as a new asset. And there is no relationship between this new asset and the original asset from which it was created.

Frustrating, but we’re thinking about ways around this. More to come as our tests and thinking solidifies. More info on Episode Engine/Final Cut Server integration in this pdf from Telestream’s website.

As an Enterprise Consultant at Control Group, I help lead our Broadcast, Media, and Entertainment technology consulting group. While CG employs experts in a wide variety of technologies, my area of experience and expertise is focused around Apple solutions for professional video and design. A few months ago, I wrote an article for O’Reilly on Final Cut Server that was targeted at consumers and covered the basics for professionals in the production industry. I wanted to expand on that article and share some thoughts on Final Cut Server as part of an integrated media workflow.

Final Cut Server client

Apple markets Final Cut Server as a tool primarily for Final Cut Pro users. Apple’s focus is on enabling users to manage their FCP projects and related files, providing them with automations to save time, and giving production teams a centralized system for collaboration. Beyond Apple’s sales pitch, we think Final Cut Server has real potential to play a central role in a production and distribution platform made up of a framework of connected systems.

Here’s an example of a typical broadcast infrastructure:

• Storage area network (SAN) for storage
• Media asset management system (MAM) for organizing and versioning
• Editor/artist workstations
• Producer workstations
• A transcode system for delivery

Now here’s an example of how those pieces might fit together in a broadcast workflow:

1. Content from tape or tapeless media is ingested through the MAM. Metadata is added at ingest – both technical metadata (shot, tape, take, etc) and possibly contextual metadata (actor, object brands, locations).
2. Data is saved to the SAN, where it is cut and crafted by editors and artists. As the content comes together, project files and new assets are saved to the SAN by editors and artists, and reviewed by producers.
3. As content is completed, it is transcoded for delivery to television, tape, and the web. Web distribution might include delivery in a custom player, and/or Hulu, YouTube, Vimeo.
4. Once content is in the wild, content owners need analytics tools to understand who is watching what and where.

Beyond its out-of-the box capabilities, Final Cut Server can be customized to play a key role in workflows like these. While it doesn’t include an API, developers can leverage custom responses in Final Cut Server to read and write XML and run external scripts. On its developer website, Apple provides an example of Final Cut Server integration with an external Rails application that enables the viewing and commenting of movie clips in a web browser. An example like this serves as a useful starting place when exploring building the middleware to connect Final Cut Server to other applications or platforms in a workflow.

Episode Engine Admin

For example, Final Cut Server leverages Compressor for all of its out-of-the-box transcoding. However many existing infrastructures exist using established transcoding systems, such as Telestream’s Episode Engine. By combining metadata subscriptions and watch folders in Final Cut Server with custom responses that leverage external scripts, Episode can easily be integrated as the transcode delivery component for a Final Cut Server workflow. Similarly, the ability to read and write XML to assets in Final Cut Server makes possible the development of web applications that interact with assets, and can even store information in their own disparate databases, populating Final Cut Server when appropriate.

What excites us about this are the many Final Cut Server integration possibilities that are not currently being talked about. And since Control Group marries Apple video integration expertise with a team of developers under the same roof, we’re excited to continue to innovate in this area. Give us a shout if you’d like to open up a dialog on how Final Cut Server might fit into your broadcast or production workflow.