My associate who referred this customer to me, when asked, tells me the customer is “no bullshit” and to “take him seriously” and has cited some of my customers prior jobs which are extremely impressive.

So anyway, over the last 2 months the customer has been talking about a business he is intending to establish in Los Angeles. He requires the assistance of someone like myself, who in his own words “is fluent with technology to the point new unknown challenges are of little issue”. Quite a compliment on my skills – essentially he’s intending for me to operate an IT department. He would like me to come to Los Angeles for 3 to 6 months to assist building his technology driven business. He has stated the offer isn’t rolling yet, as they’re seeking financing as it is a large scale operation. He has hinted at the financing they’re receiving, although very unspecific, it is a figure in the tens of millions to start-up.

I met with the customer earlier this week and he told me his business partner is in Los Angeles right now and only last night was approved financing. They are now in panic mode and are drafting up plans for everything. The customer will come to me in the coming weeks with an offer.

So tonight I did a bit of research into my actual expenses and determining what I think is fair. I worked out that intending to live in Los Angeles very close to my work will cost a minimum of $50,000 USD with expenses like accommodation, travel, a car, my necessary tools (computers), etc.. I also have to consider the job is intending to place me under much stress and pressure so I should be well compensated – and during the trip I would intend to travel to New York, Canada, and Mexico. Also, I’m being asked to close up a business in Australia for an extended period of time that’s taken me over a decade to establish. But maybe I can keep my Australian business rolling. If I can convince my partners brother to cut his very long hair, it may work out that I can employ him to conduct my Australian operations while I’m away and rely on Skype to supervise and train him from LA.

It seems very fair to ask for amounts upwards of quarter of a million dollars for the totality of this job.

The job is also a film or motion picture orientated business. My customer appears to be well connected to well known Hollywood based talents. The phrase “going Hollywood” pops to mind since if this job pans out, that’s what I will be doing.

I’m also fascinated by the idea that LA has extremely low cost fibre optic Internet, and its very likely my office in downtown LA would be sitting on a GigE connection with the intent of streaming real-time high definition video and providing free Internet access to customers of the business. I’ll ultimately be responsible for a couple racks of equipment, which likely will be in my office.

My partner is really thrilled by the idea that if it pans out, we’ll be jumping off the plane and unloading $50k almost immediately. The job is also interesting because my partner is prepared to give up her newly acquired job (only today in-fact) if this pans out. I can keep her occupied obviously by having her work directly for me. Alternatively, my customers business model to me obviously requires someone with her qualified skills – so either way she can come and shop and will also have a job to keep her occupied while I am. I’ll also end up proposing to my partner overseas somewhere really special.

I would also like to have my parents come visit and my partners brother, both of which I’d have to finance. My partners parents I’d like to visit too but they could likely finance their own trip. With all of them I’d intend for them to stay at our home in LA. I would use Skype to stay in touch with them too – the office should be able to saturate any Australian broadband connection with high definition video of us.

I would intend on returning to Australia, maybe, depending on how financially rewarded I can remain in LA. Ultimately I’d want to be very financially rewarded to remain in LA because I’d be distanced from my family, and have fears of differences in politics like no free healthcare in the US while we have Medicare in Australia. If I stay in LA or if I returned to Australia, either way I’d want to invest a large chunk of any income from this job into my business – I figure if my business can generate that kind of income, its clearly deserving of further investment, and takes priority number one for how to spend that money. If rolling a few tens of thousands can give this kind of reward – what can I achieve from investing $100,000? What kind of job offers will I get in another 10 years time? This kind of investment though – I’d want to do very very carefully, but at the same time, I’d have to make sure I’m not too careful and don’t retain it too long where its ultimately wasted over a long period on useless purposes.

However you have to take everything with a grain of salt. I’m not expecting this job to pan out, but yes I am really excited about it. To me it seems like an achievement that someone who does not bullshit and is the real deal is offering me such a job. If only it could be put on my business portfolio – but right now I don’t think it can since the entire project is top secret (oops, did I blog a little?).

Icecast supports metadata in MP3 & AAC+ streams so this metadata can be used to represent whats being sent to air such as the name of the program or the name of the track.

I needed a databased calendar for EPG. So I turned to Google Calendar because of its easy interface to modify the calendar data. It also has an API which can be used to both perform lookups of whats presently on so we can send that to Icecast, and we can use the API to pull the whole program guide and apply our own formatting on our website.

Today I made a script to pull the current entry from Google Calendar and to publish it to Icecast over HTTP. I had always thought this would be difficult to pull off, however it ended up to be much easier than I ever imagined. I wrote this script in PHP so that its modular with the same lookup functions I’ll later use on the web based player.

Now all that remains is to convince the web developer that the program guide should use the Google Calendar data so we’re not maintaining 2 different program guides. I’m still waiting for him to implement my web based player. And its been 2 months since I first requested admin privileges to his WordPress installation. So my hopes on having anything done any time soon aren’t high.

To achieve this they had an FM tuner attached to the sound card on their server.

We got some new mainboards and we were upgrading the server anyway. I realized I had quite a few Brooktree TV tuner cards.

So I installed one in their server so the FM tuner can be controlled by software. Setting up the card wasn’t so straight forward as there wasn’t much documentation but it was fairly easy.

Debian did not detect my card correctly and did not identify the FM tuner component. So I had to unload the kernel module and load it with a few parameters. So I executed:

rmmod bttv

rmmod tuner

modprobe bttv tuner=2 radio=1

modprobe tuner

And then I had the driver loaded properly and /dev/radio0 exists. So I ran “aptitude install tuner” to install the tuner command line utility. Once installed I could then tune the FM receiver with the command:

tuner –c /dev/radio0 –f 96.1 -q

And then the radio was tuned.

To make it always work I then added all the rmmod, modprobe & tuner commands to /etc/rc.local so its all setup on boot.

Today I’m writing about using VoIP in a radio broadcasting environment and why its really awesome for a radio broadcaster over conventional PSTN lines.

A radio broadcaster will typically have 3 unusual telephone devices that most people wouldn’t need.

They’d have a “telephone interface” to connect a telephone conversation to their broadcast which has volume balance capabilities and retains a studio quality audio feed for the studio microphone (instead of phone quality). This can be replaced with a PC running a VoIP softphone and a couple hardware based audio mixers to interface the PC to the studio’s main mixing deck.

They’d also have a recording device that would also retain studio quality audio for the studio microphone so that pre-recorded interviews can be produced. This can be replaced with Asterisk’s Monitor() function without care of studio quality audio. Or, it can be replaced with a PC running a VoIP softphone capable of recording. If the VoIP softphone can’t record in high qualities, you’d just run Audacity simultaneously and later use Audacity to mix the two ends of the conversation together.

They’d also have an external broadcasting unit that uses a specialized modem to encode studio quality audio on a regular phone call so the broadcaster can broadcast outside of their studio. This can be replaced with Audio over IP technologies which are also used in STL (sometimes called STL-IP).

Typically each of these devices are worth several thousands of dollars or more and every replacement can be done entirely with free software.

On top of this, some broadcasters will also have a PBX system provided by their teleco so they can play their broadcast as music on hold and in place of ringing noises when someone calls one of their phone numbers. Asterisk of course can match this functionality.

VoIP is really beneficial for a radio broadcaster for a whole range of reasons, the main one is that expensive hardware can be replaced with free software. The low cost of computer hardware makes it possible to build redundancies into a telephone system that essentially did not have these features to start with.

So now all I’m left waiting on is board approval to begin some migrations to VoIP at my first signup. At a later time I will have to revisit this blog with some more technical detail.

I was interested to find that community radio stations won’t need to buy any new transmitters to enable digital radio – as digital radio is “multiplexed” so numerous stations are squeezed into a single signal with 1000kbits of bandwidth. There are approx 30 digital stations in Sydney which all come from just two transmitters. Because of multiplexing, the law was made that any commercial digital radio transmitters must allocate two ninths of their capacity to community radio stations.

So all we need to do is get backhaul into the transmitter site (whenever it becomes available).

Its also interesting to note that DAB+ uses AAC+ encoding. There is a new uprise in IP based STL (Studio to Transmitter Links). Many STL-IP solutions can encode in AAC+. So with an AAC+ STL-IP, there is no encoding required at the transmitter site.

My customer will be attempting to seek funding for STL-IP even for their FM transmitter but this will cost at least $10,000. Their studio is probably 200 metres from their transmitter, and is currently serviced by FM modulated analogue audio over microwave. There is only one pathway to the transmitter site and if this goes down the FM transmitter goes off air (or transmits nothing).

So the thought is by implementing STL-IP we could continue to use microwave as our primary STL, and we’d need to migrate the microwave link to something IP based. But if the microwave goes down we could utilize the studio’s ADSL2+ link and the transmitter sites fibre optic link to provide a secondary IP pathway. Migrating to STL-IP only makes sense as we could later use multicasting to send the same feed to any digital radio transmitter that pops up in the area.

But something curious about this is that CBAA’s policies on digital radio state that commercial transmitter sites only have to provide one pathway and don’t have to offer any secondary pathways. Meaning a migration to STL-IP would make our FM transmitter more redundant, but digital radio would be more susceptible to failure.

I’m now wondering if the local council or the tree hugging hippies can be made liable for any injuries or deaths in the area where people have been unable to call Emergency 000 with their mobile phone.

I’ve voiced my protest in the form of rotten eggs to the front doors of the houses on Woodlands Rd where the tree hugging hippies live. Enjoy.

Emma Schofield stated “We’ve got a school probably 350 metres away, a hospital not much further I think with so many inconclusive and unsubstantiated results and studies, we should be really careful”.

I found it interesting that an article was published without inspecting the geographical facts. North Katoomba Public School is more like 400m from the Colless Foods site and the Hospital is 350m away – something that can easily be measured with a tool like Google Earth.

There also is in fact already a radio site at the Hospital offering several frequencies of a power higher than the proposal by Telstra for this new mobile phone tower. A further two radio sites for Integral Energy and the NSW Rural Fire Service are on Showground Lane some 400m from the Hospital and 600m from the School.

While we are concerned about inconclusive and unsubstantiated results I also find it interesting that Barton St and Victoria St which both surround the North Katoomba Public School have high voltage power lines, lines that equipment used by Integral Energy is subject to routine quarantines to remove magnetic energy – yet no such quarantine system is employed by the telecommunications industry as the energy exposure is much less.

Optus currently considers the Hospital and North Katoomba Public School to be in strong signal meaning both sites are already exposed to significant amounts of mobile phone signals.

Even without electricity we are exposed to radiation continually from both our Sun, other stars, and the microwave background.

The study of safe radiation dosages is based on observations of the health effects from the Hiroshima atomic bomb which was a high dosage. Health effects from lower dosages was assumed to be linear and its not exactly known if the assumptions are correct – which is something that is potentially being disproven after the Chenobyl nuclear accident as there has not been enough deaths to line up with the radiation dosages chart.

While the reception issues that Telstra wish to clear up will effect hundreds, if not thousands of homes, and provide mobile phone reception in a bushland area not already covered which will provide improved public safety.

So whats all the fear about? I’m not so sure. If mobile phones scare you, maybe you should move west of Bathurst, because you will be exposed to electromagnetic energy if you like it or not – and there is no conclusive evidence to suggest this exposure is harmful in any way.

I was quite surprised by a customer, with no real technical capabilities, had specifically requested Linux. We ended up using Debian, NIS & NFS for ‘account roaming’ across multiple machines, and the Gnome Desktop.

The Internet stream initially seemed simple to solve but proved to be quite a challenge. Reliability was the major issue. I setup an Icecast stream and found due to my DSL connection dropping out, the stream would drop out some 6-12 times per day. Once every 2 hours did not seem acceptable.

The solution to that problem now encompasses two physically separated sites with their own independent Internet connections uploading the stream to two public Icecast servers. In order to allow for redundancy, the stream uploaders run Icecast servers and the public servers are configured to relay streams from these machines. With stream fallbacks, one of the stream uploaders can go offline, or one of the public servers can go offline, and the stream will remain in a working state. Even still, its important enough to warrant SMS notifications in the event any equipment in this pool goes offline – so we monitor it around the clock.

The only form of Internet available to us is ADSL from Telstra and their resellers. There simply are no other affordable options in our area since the OPEL project was canned by Krudd. This limits us to 384kB/sec at each stream uploader site. We can only afford one line at each uploader site, and there are other uses of the Internet connection, so the realistic available bandwidth is even less.

The bandwidth limitations meant we could only upload one high quality stream. However, we required several in different formats and bitrates to suit listeners requirements. Stream transcoding at the server end seemed the only possible solution – and it seems the Icecast developers aren’t interested in stream transcoding at all due to ‘quality loss’. There are very few applications available to handle this task. I ended up making numerous watchdogs for the buggy streamtranscoder package from oddsock. This enabled us to transcode the stream into another 3 formats.

Playback in web browsers seemed to be another drawback for Icecast/Shoutcast. Two methods are possible – use HTML <embed> tags to request a plugin the user may or may not have installed, or use Adobe Flash. Adobe Flash seems sensible, but its lack of handling for live streams makes it the worst choice – playback is distorted, or non-existant, in many mainstream browsers. I’m still to develop a reliable method to play the stream in a browser that “works everywhere”.

Another problem the station faced was broadcast recordings. By law, ACMA requires the broadcast is recorded and copies are retained for 6 weeks. The old solution was huge piles of VCR machines which would fall over frequently for many different reasons including tapes only lasting 6 months with the duty cycle the station imposed on them.

A couple shell based tools made recording possible with just a shell script executed from cron. But like always, it took some time to perfect. We had also added a terabyte of RAID5 so there was plenty of storage for 2 months of recordings at a high bitrate which seemed an easier number to archive. The implementation of this quite simplistic shell script has been a huge success – with many presenters pulling their own shows, for their own archiving purposes, from the recording system I setup. It is the primary method for recording radio shows today.

Better software for broadcasting from the PC is in the works at last. It seemed the station was initially interested, but then failed to locate a presenter who was willing to use such a system. We now have one – and hopefully can use this presenter to turn more. Its still unclear how we can wire it to the FM broadcasting equipment while retaining reliability.

Given the special requirements for a radio station, and the costs of licensing software to handle it – seems you’d be crazy to not take the Linux route. But it isn’t an easy route – but that seems true for any PC solution for public radio broadcasting.