Finally (and arguably because I have a bit of an interest in audio processing and loudness control) this article has a good overview on processing for streaming.
I particularly support the comments of Bob Orban that start "If you want to sound like the original recording"
That would be my starting point with PlayIt Live on a remote server.
From my experience, PlayIt Live can run 24/7 and I can't see why it would not run for months on server hardware.
It wouldn't however be my first choice approach.
Taking a step back, what do you want to play 24/7 and why do you want to do it ?
There are various platforms which will "carousel" material to an Internet radio server with some sort of "Auto DJ". That kind of service with included streaming might suit your needs.
If the idea is some live programmes and a "sustaining" carousel then some of the services I've seen in passing seem to do this.
It is possible to back up an Internet connection with a second (say using the mobile network) and some hardware makes that relatively simple but the way I read your question a managed and supported service seems the answer.
I've no connection with the website or any of the firms listed but this looks like a good choice of services in the UK (even if the list is now a little old).
Many thanks for your reply and advice.
My primary idea - I would like to have an internet radio station running 24/7 x 365.
* There won't be any live presenters - I'm not reliant on any studio-based hardware, technical ability of presenters etc.
There will be jingles, idents and sweepers so there needs to be some
sort of playout pattern or "rules" in the playlisting to insert these
into the right place at the right time.
* I would like to run a sponsorship type message at certain clock times. By that I mean, I'd allow a track to play right to its end point and then have let's say a 2 minute message from a sponsor, so the playout times of these need to be "loosely" fixed.
* Any spoken links are of an "interview" type nature and these will be pre-recorded with an iRig out on the streets. I wouldn't go so far as to consider this as VT, I'd treat it just like a musical track. Again there needs to be a playout pattern in place to ensure these are played out at the right time and in the right place.
I had trialled Radio.co - everything is hosted
by them so you pay a monthly subscription fee to you use their servers,
bandwidth etc. Basically, with them you're ready to operate "straight
out of the box", BUT, the big problem with Radio.co is that it's nowhere
near as intuitive as
PlayIt Live when it comes to silence analysis on songs that fade out –
no kind of segue editing with Radio.co to tidy up this shortcoming. I
was literally editing the ends of songs that fade and then finding that
the crossfade was still very messy at best. PlayIt Live is so good a
silence analysis that I'd say on 99% of occasions there isn't even a
need to manually segue.
In terms of features, I'd personally rate PlayIt Live as being as good or better than the likes of Zetta. I've ran PlayIt Live on my computer on a trial basis for days at a time and it has never crashed. However, I'd like everything to be on a server so as to avoid downtime, poor internet connection etc.
You mentioned, "It wouldn't however be my first choice approach", so I'm wondering if PlayIt Live is "too good" for my basic requirements? I'm starting out, but you sound like you know what you're talking about. If my project was yours, how would you do it?
Thanks in advance.
It sounds like you have tried what would have been my first suggestion by using Radio.co or similar. What you are aiming at does seem to need a "proper" playout system and so PlayIt Live is a good choice.
Running PlayIt Live on a remote (well connected) Windows machine at a data centre or similar would do what you want.
I would be quite concerned by security, but people who sell this kind of remote machine service will probably be able to lock down access to your machine.
If your server is "virtual" you need to consider that PlayIt Live will be reasonably demanding of resources and need to work without interruption, I don't know how it would work in a virtualised environment, I think it would be better if you had your own "physical" machine.
I also wouldn't try and run the stream server on the same machine as PlayIt Live.
Shoutcast or IceCast are not very demanding but they need to be on a public facing machine whereas (assuming you have a static IP address at home) your PlayIt Live machine access could be locked down so only you could connect.
As a first try, you could get a second (local) machine (with the same Operating System as your server) and run PlayIt Live on that.
If you find remote access to that local machine allows you to do everything you need, then it would be worth trying it on the remote server.
One other comment (possibly something you have thought of).
Your material should be as processed as you need for transmission because (other than possibly some form of software audio processing) you won't have access to the "raw" audio output of your station (stream).
PlayIt Live does have some processing but it isn't something I would rely on with highly variable source material.
At the very least processing all material to have similar apparent loudness and controlling the dynamic range of your inserts would be my approach.