- This topic has 22 replies, 7 voices, and was last updated 16 years, 1 month ago by
.
-
Posts
-
Hi guys!
Long time no see — I’ve been busy — developing Breakaway FM!
It looks a lot like regular Breakaway, but the similarity is superficial — under the hood, most everything is new.
This is a real FM processor.
Abridged feature list:
- Advanced Distortion Cancelled Clipping back-end, for unprecedented clarity on FM, while being competitively loud. It’s significantly cleaner sounding than anything on the market — $10000+ boxes included.
- PHASE LINEAR from beginning to end, to maintain waveform fidelity even with pre-distorted CDs
- Pre-emphasis (75 / 50us for FM, 25 / 15us for webcasting)
- Adaptive asynchronous sample rate conversion (Input 44100 – 192000, Output 44100 – 192000)
- FM Stereo Encoder (MPX), with Pilot+RDS input for using an external RDS encoder, or Airomate RDS software.
- Line-up tones: 400hz, 1000hz, 60hz Square, Sweep, and Bessel Null tones
- Adjustable Tilt correction, separate for L/R and MPX outputs
- Phase Linear Parametric EQ for the MPX output, to compensate for sound-card frequency response
- Defeatable L/R output de-emphasis — so you can preview the audio with your regular sound card
- Three breakaway pipelines — for easy interfacing to any software on your computer, both input and output.
This is a BETA version. It’s not ready for release yet — but hopefully will be soon 🙂.
This version is also usable for webcasting.
Breakaway Web will follow shortly after release of Breakaway FM — but the two will not be the same.
Breakaway Web will have selectable Phase Linear or Low Latency (Breakaway Personal is always Low Latency mode), and use look-ahead limiting for peak control. It will be very CPU efficient — roughly the same as Breakaway Personal.
Breakaway FM, on the other hand, uses an advanced distortion-cancelled clipper. It uses much more CPU, but when selecting 15 or 25us pre-emphasis, and enabling de-emphasis, it will allow you to make an even LOUDER webcast than Breakaway Web will. So, if absolute loudness is important — if you want your stations to jump out of the speakers — then Breakaway FM will be the best choice, even after Breakaway Web is released!
Breakaway FM will be $199. Breakaway Web will be less.
Enough talking. Check it out! 🙂
http://bredband.leif.cx/browse/bafm <– download here
If you’re interested in how Breakaway FM stacks up against market leading hardware processors, head on over to:
http://mpxtool.com/site/torture-test.html
…where you can download recordings of different processors, and A/B compare them using MpxTool (the free demo version).
///Leif
Thanks Leif,
looks and sounds fine.
But I didn’t test it today.
Hope to find some time at the weekend.Ran Breakaway FM through a Bext 30 watt FM exciter/stereo generator (used L/R audio without composite).
Used my Inovonics FM modulation monitor and is sounded great and got a decent modulation level. Breakaway FM is On The Air!
I hope Breakaway FM is available by November 1st when I launch our radio statation.Leif,
You have once again created a masterpiece.
I love the plutionium preset; it makes my web stream louder than any other, and it still sounds good.
Any idea when you will release the "official" version? I’d love to get rid of those trial commercials 😉
– Oele
Awesome, Oele!
What pre-emphasis setting are you using? For streaming I recommend 15 or 25us.
Feel free to post the URL to your stream 🙂.
///Leif
Hi Leif,
Would you recommend Breakaway FM to enhance the satellite uplink audio (256Kbit MPEG II) that feeds our FM transmitters across the country? My problem is our FM transmissions broadcast in mono I have Aphex Dominators at the TX side that has increased the loudness while being legal, it is no chop compared the Omnia’s and Orbans in the area. I have ran various DSP software over the years before the satellite and before the FM transmitter with some limited success but your new Breakaway series sounds the best so far. Maybe you could make a pre-set to enhance the sound of MONO broadcasts?
Sam
Hello Leif:
I have been doing tests in the air, fm, and the sound is very good (I especially liked the preset "cleveland" 😮 ). Congratulations, you’ve returned to surprise with a phenomenal job. I hope that some day, not far away, can have some way to eliminate or minimize the delay time, to listen in real time without discomfort 🙄 . BeakAway Fm is a product to be proud.
Fernando.
Sam, this is a very interesting problem!
Difficult, but interesting nonetheless.
Bit-reduced codecs aren’t designed to deal with peak-controlled processed audio.
I tried an experiment with MP2 (using the TwoLame command line encoder, version 0.3.12b).
At 256 kbps, it introduced lots of overshoots – up to 17%. This is a big problem, because dealing with that much overshoot with any clipper other than Breakaway FM itself, would cause audible distortion.
On the other hand, at 384 kbps, I only saw 3% overshoot. So, if it’s possible to increase the bitrate, and feed the encoders digitally, you may be able to run Breakaway FM into the encoder (pre-emphasized), disconnect the Aphex and plug directly into your exciters at each transmitter site.
Every preset is completely channel-discrete, so the current presets would work just as well in mono.
Otherwise, the very best solution would be to replace each dominator with a server grade pc, and run Breakaway FM at each transmitter site. You could get by with an inexpensive server PC and an inexpensive sound card, and if you used the MPX output of Breakaway FM, you’d gain stereo sound too!
Fernando, thank you very much! Man, I knew there had to be someone to like that preset 😉.
The delay, unfortunately, cannot be removed. The algorithm delay alone is over 200ms. A clipper this advanced simply needs time to process the audio — long, complicated phase linear filters add delay, and there are several such stages in Breakaway FM.
What we will recommend, is to also run a low latency version of Breakaway, and to listen to that one instead of the FM.
///Leif
Leif,
This is a very interesting and extremely frustrating! I have spent countless hours tweaking etc.. What I should of mentioned is that the satellite link is in stereo so that only leave us with 128k audio. Today I will enquire on the possibility in reconfiguring the link into mono at 256k or even better 384k.
When you say feed the encoders digitally, Do you mean via AES/EBU or actually encoding the mpeg then streaming to the sat uplink?
Mpeg II is a very normal way of distributing broadcast audio, Is it predictable? As in, can you make your DSP know that this piece of audio is going to have problems in MPEG so Breakaway compensates somehow? or is the compression just completely random? Know what I mean?
Sam
Hi Sam!
I understand you completely. Tweaking is extremely frustrating when you throw unpredictable variables into the mix!
By digital I meant AES/EBU into the encoder. That *should* do the trick — but you’ll have to look and make sure.
The best way would be to play Breakaway FM (pre-emph) into AES/EBU into the encoder, and then take AES/EBU from the DECODER, and feed that to a sound card with digital input, on a computer running MpxTool. Then you’ll be able to see exactly what overshoots you can expect from a given bitrate.
Mpeg II is predictable if you can jack up the bitrate high enough. I’d be comfortable running it at 384, but below that, overshoots get too large and unpredictable.
It’s not possible to compensate for these types of overshoot — they’re basically completely random. It comes from certain things being thrown away, and if the thing that is thrown away happened to be a clipping harmonic that was keeping a peak within 100%, part of the peak comes back.
How come you’re broadcasting mono, by the way? And, what country are you in?
///Leif
[quote author=”Leif”]Awesome, Oele!
What pre-emphasis setting are you using? For streaming I recommend 15 or 25us.
Feel free to post the URL to your stream 🙂.
///Leif[/quote]
Pre-emphasis is currently on 15us.The stream URL is http://muziek.oele.net:8000/muziek . This is just a "hobby" stream for me and a few friends. Don’t be surprised if you hear weird combinations of musical styles or weird music in general 😉
Breakaway is now on the "plutonium" preset. Although i’m impressed by the loudness i will probably change it to a more "sensible" preset one of these days 😉
Breakaway runs inside a VirtualBox Windows XP Virtual Machine on the linux server that hosts the stream.
On linux: mplayer & perl scripts play mp3 and flac files -> jackd -> oddcastv3 -> icecast (flac)
On XP: mplayer plays the flac stream -> breakaway fm -> simplecast -> icecast (mp3)(Yes, this is one of the reasons why i’d love to have a linux version of breakaway 😉 Maybe i should try to move the other stuff to Windows too. )
We have over 40 sites, not all the TX sites have the Dominator 722 installed. I am in the process of purchasing some more, as well as some more Tandberg 1220’s, then running pre-emphasis on the Sat link will be a possibility…
In the mean time I have made some more tweaks and now disabled the Aphex Compellor at my local TX its now sounding more natural though not as loud but in spec, but only about 98% of the time! I have observed these overshoots on other networks too. One in particular is running Orban 2300’s but there content is news with audio around 3-5khz I think the 2300 is struggling and in some cases perhaps causing issues? Just wondering if they should have used all there old 8100’s they replaced from there other station.
We are running mono due to our ERP restrictions and the area we need/want to cover, I’m in Australia. When will Breakaway FM be available for purchase? I would like to give it a go on the uplink, as I don’t think Breakaway standard rolls off @ 15khz or does it? As much as I would like to plug your software on the station I don’t think the boss is going to like hearing the Breakaway FM demo adds on air every 10 mins or so lol 🙂
Sam
Hi Sam!
40 sites, not bad!
Breakaway FM has an extremely sharp dropoff at 16 (sixteen) KHz. At 16.5 KHz it’s down 60db (!!), so it provides impeccable pilot protection AND extended audio bandwidth.
The Breakaway FM advanced distortion cancelled clipper takes great care NOT to introduce distortion beyond the available audio bandwidth. If you feed 5 KHz audio into the clipper, you get 5 KHz audio out of it — no matter how much clipping you’re doing. This is in stark contrast to every other product, which load up the spectrum with harsh sounding distortion — that’s why phone calls sound so bad on the air!
Breakaway Standard has no low pass filter, nor does it have pre-emphasis or clipping. There’s simply no need for it in a personal audio processor.
I understand why you run mono now — makes sense.
To be honest, there’s not really anything stopping you from purchasing Breakaway FM right now — the beta version is essentially a release candidate, and it can be already be registered and activated, and you can of course upgrade to the release version when it’s out. I’ll check with Keith and see if he can get a webstore going.
The ad actually only comes on average once every half hour — but it’s randomized. Good idea though, wasn’t it? 🙂
I couldn’t find any relevant information about the Tandberg 1220 (other than that it’s a satellite receiver of some kind) – what does it do?
When using any FM audio processor, whether it’s a Breakaway FM, an Optimod or an Omnia, the key to loud, clean sound in the air is modify the audio *as little as possible* between the output of the processor and the input of the exciter.
Once the audio is tightly peak controlled, any modification, no matter how minute, will cause overshoots. Non-linear phase response? Overshoots. High frequency rolloff? Overshoots. Low frequency rolloff, any kind of EQ’ing at all? Overshoots.
The more modifications, the more overshoots there are. Thus, if you were to run Breakaway FM audio into any other processor, you would be combining the WORST aspects of both processors, and end up with something quite awful sounding.
I’ve played extensively with 8200’s and 8500’s, but never actually had my hands on an 8100. From what I hear, the 8100 is clean and smooth sounding, but not very loud and bright. That might indeed have worked well for news, but extended high frequency readroom is also important for intelligibility, which is a rather crucial point for an news station 🙂.
In short, if we manage to get Breakaway FM on the air, with as little deterioration as possible between Breakaway FM and the exciters, it could be an enormous jump in audio quality.
Do your exciters have any kind of protection clipping built in? If so, you could use those deal with the overshoots — since you’re not running stereo, you don’t have to worry about either pilot contamination or stereo separation issues, meaning you have great freedom in how to beat overshoots into submission. If we could get 384k over that satellite link, I think everything else can be solved.
///Leif
Leif,
First off, I love Breakaway FM! It’s probably the best so far without hours of tweaking. Your "Reference Settings" setting is bang on the money!
Now a quick question: Is it at all possible to remove the audio ID from it during the beta phase? I have no problems with a local pop-up nor would I have a problem posting a link to your site on my website. But the audio ID takes away from the programming we have and walks over programming (we’re mostly a talk format station). Hell I’d even put in a 30 second announcement that we’re beta testing Breakaway FM in lieu of the audio ID.
Another suggestion would be a 60-90 temporary key system like a few other processing DSP’s use.
Looking forward to seeing how well this grows!
Thanks,
Rick
- The forum ‘Breakaway Professional Products – [discontinued]’ is closed to new topics and replies.