- This topic has 83 replies, 18 voices, and was last updated 7 years, 9 months ago by
.
-
Posts
-
quote :I certainly do provide free updates and enhancements! Future versions of Breakaway Broadcast Processor will of course be free to everyone who has bought it.
///Leif
Again one of the questions i had is answered 🙂
I can tell you that when it comes time to buy another audio processor…we’re not buying one…we’ll be going THIS route, and we’re a ‘big professional’ radio station.
Stability is not an issue. I truly believe a good PC, running offline from the internet, is about as stable as any other hardware digital audio processor. The best part about BBP? You just need a computer to run it on.
I mean, think about it…if your $13,000 audio box dies (AND THEY DO!!!)…you can count on being without it for how long?!? (while it’s off at the factory) With BBP…you just need your settings, and another computer…and you’re up and running again.
BBP wins, in my book…and latency is a non-issue for us.
I love our Omnia 6, but I think BBP sounds cleaner..and the new plugins are intense. Since we just bought both of our Omnias, it’ll be awhile before we need a new processor. But WE WILL be adding additional streams and an HD2 channel…and again, since latency won’t be an issue, I can tell you what we’ll be using to process the audio, guess! 😀
[quote author=”luke-san”]I would like to see what kind of issues it would cause because of filtering on the audiocards[/quote]
All DA converters have filtering, weather they are in an Omnia6 or a $10 special from Newegg. Breakaway already doesn’t have a problem with MPX output, or even asymmetrical output for AM, with many "prosumer" converters. Clipping the MPX instead of L/R wouldn’t change that at all. That being said, care to explain what kind of issues you are referring to anyways? 🙂
JesseG,
Indeed all DA convertors have filters, sure, but none are the same and that is the problem.
If I measure a multi-tone signwave on my ancient Terratec EWS-64, M-audio or or a simple soundblaster (to make it evil) it will give different results which will cause a signal to become compliant or non-compliant and that is just one issue.
As a matter of fact I did do a test a couple of years ago and I had trancient peaks of 6 to 8 dB, and that is something you just don’t want.
To make matters worse measure a square wave on that $10 card or the Omnia6 and it will have show a different output.
A clipped signal looks like a square wave in worst cases.
I think you are refering to clipping on the output due to the limitation in voltage, but that is hardware clipping.
It all needs to be ‘correct’ when we are in this game and suddenly DDS transmittors become very attractive to me.If Leif could develop, or to make it more simple, adopt an existing audiocard which would comply and would be part or his processor package that would be a big advantage to sell.
I believe the Esi Juli@ would make a nice candidate because of its stability and top grade components.I wonder which professional radiostation is going to buy a pc based solution, I truly wonder.
When I spoke to some establishments, lets call it that way, in Europe they smilled and looked the other way.
Most hardware processors are 19 inch in a rack and most of them are 300 to 400 meters up in a tower.
So if you would go the pc way you need to buy a 19inch one with top grade components etc.
On a stripped version of XP or 2000 with a SSD drive.
How do you handle this in a professional way?Leif, maybe best to move this flow of thoughts in a seperate topic.
Hi Luke!
The 6 to 8dB of overshoots probably happened because the signal was not properly clipped. (If it wasn’t that, perhaps the audio card had some effects left on, like reverb or CMSS). When clipping a signal, it’s not enough to just clip all the samples to a certain value (this is a common misconception) — you have to make sure that the waveform, when reconstructed by a D/A converter, stays inside the lines. This is usually done through oversampling. Scope the output of Airomate (on the output of the sound card) and then switch to BBP, I believe you’ll see what I mean.
When clipping is done properly, the D/A filtering is a non-issue — the highest frequencies in the signal are FAR from nyquist for any sampling rate. More specifically, L/R 44.1 KHz, Nyquist is at 22 KHz, highest audio frequency 16 KHz. MPX 176.4 KHz, Nyquist is at 88 kHz, highest signal frequency 54 KHz. The signal is completely within the passband, and as such it is completely unaffected by any D/A converter filter — there is simply nothing to filter out, the signal is clean already.
A squarewave on any proper FM processor (such as an Omnia6) is obviously not a perfect squarewave. Perfect squarewaves are *illegal* on FM, because the signal must be bandlimited.
I am indeed looking into having a custom audio card built. However, it is absolutely not a requirement for the current incarnation of Breakaway Broadcast Processor. It performs perfectly even with on-board sound (provided the on-board sound supports 192 kHz and has a reasonably low noise floor, such as realtek 883 or 888).
I believe it is customary to install audio processors at ground level, down by the transmitter, not 400 meters up in the tower by the antenna. 😉
///Leif
Hi Leif,
I do not 100% agree about the audiocards.
A very simple test is to do a single tone sweep that goes from lets say 20 Hz to 70KHz, now lets see the -1 dB points and how this can actually happen.
Now check what happens with the phase shift.
All because of filtering and that was my statement between the Omnia and the 10$ card. With that squarewave on the input the output on the Omnia will be FCC clean and the 10$ card might not be.Some links I found on the net in regards to the difference in audiocards, and no some of them I don’t want to connect to a transmitter.
I think the spectrum analyser would show some not so nice stuff.
http://www.clarisonus.com/Research%20Re … Hz-1dB.png
http://techreport.com/articles.x/13874/15Btw I don’t want 400 meters of coax going to antenna’s with about 20KW going thru it not even talking about the loss.
I don’t know in which part of the world this is being done?Hi Luke-san,
I do notice you don’t 100% agree 🙂.
First link shows high frequency spikes on an m-audio card. Indeed that wouldn’t be ideal, but don’t forget to pay attention to the level of these spikes in dB’s. The very tallest one is at -65dB. At that high frequency, it will very likely be masked by the RF triangular noise, even at audibly good reception conditions. Note that -65dB dB corresponds to an injection level of 0.06%, or 0.045 Hz deviation. It’s not much of an issue — but I can see where you’re coming from, it wouldn’t hurt to use a cleaner sound card.
On the other hand, the second link, come on. You missed a VITAL piece of information right there on the page — the test was done at 44.1 KHz sampling rate!
Those cheap consumer cards are obviously locked to 48.0 KHz (as most are), and do a poor job of sample-rate converting 44.1 KHz sources, while they would do MUCH better with 48.0 KHz sources.
It’s not a coincidence that the default output sample rate for all the Breakaway products is 48000hz.
///Leif
A friend of mine has a new tool from a company called Microgen…it’s a usb device that simultaneously monitors every FM station on the dial: reads the rds, deviation, features l-r graphics imaging…scope, etc, its nuts! Anyway he hadn’t heard breakaway yet, he’s got a killer soundsystem…so we sat and played lots of stuff and were able to race a couple stations…
We ended up using a newer exciter that gave us much better peak control…and sounded cleaner.
This preset literally blew us away today:
Plutionium NR
+.2
50
45
50
Normal Bass
-13 bass shapebass efx plug in: 2
impact/clunk: slam 2, clunk 2Crazy crazy crazy sound.
[quote author=”luke-san”]Btw I don’t want 400 meters of coax going to antenna’s with about 20KW going thru it not even talking about the loss.
I don’t know in which part of the world this is being done?[/quote]I just read this and chuckled.
First of all, I have never ever in my life heard of a processor being mounted on a tower. The logic behind doing something like that doesn’t even begin to line up. What engineer do you know that is going to climb a tower to change out a cable or a dead box? AND the processor comes before the transmitter in the airchain, so are you telling me that the audio comes in from the STL, is fed up a tower to the processor, and then back down to the transmitter, and then back up the tower to the antenna?
Also, I don’t know if you have ever seen the kind of cabling that feeds the antenna from the transmitter, but it’s not coax that you find at RadioShack. It’s 6-inch thick, copper-lined, high voltage cabling. And it is designed to carry audio up a 1500′ tall tower without crapping out.
Just sayin’.
Adam
Looks like we scared him away already 🙂.
The simple truth is that the only two instruments you need, to verify an FM processors output (including the output dac, i.e. the sound card), is an oscilloscope and spectrum analyzer.
*Any* and *all* issues will be detected. Phase problems? If they’re bad enough to cause overshoots, the scope will tell you. Frequency response? Same. Non-linearities? The spectrum analyzer will tell you — you’ll see it as garbage around the pilot, where BBP normally has none at all (80dB pilot protection).
///Leif
Adam and Leif ,no, you didn’t scare me at all well actually you guys scare me.
I’m just laughing about all the comments that you guys just made here.
Maybe you guys use simple towers where you have to climb into, poor guys. 😆
95% of the transmitters here and other places in lets say in a radius of 1000km’s are mounted in concrete towers with elevators.
Much easier to get electric up there than messing around with RF pipelines and combinators downstairs.STL , how about fiber what we use, lol
I don’t know where you guys in the world live but in my part it seems to make more logic doesn’t it?Well, we use T1 to feed the transmitter from the studio, but ‘STL’ is a common generic term used to explain the technology that moves audio out to the transmitter.
And, unless you’re mounting your stuff at the top of a skyscraper, I can’t imagine seeing 1500 foot tall concrete towers in more than a few places. They’re not practical. Or maybe we just haven’t stumbled across that technology here in the US 😉.
Adam
Fiber directly from the broadcasting building to all the towers here.
The only STL we use is when we are broadcasting from mobile newsfeeds etcA nice example of one of them http://www.degregorio.be/MiniSite/Pagin … filter=&p=
300 meters high / about 1000 foot good enough for a couple of FM transmitters, DVB-t, DAB. Analogue TV transmittion was turned off last year.
I don’t understand why you tell me why they aren’t practical. Do you guys move towers over there?
Nothing beats going up in an elevator and having all the equipment safely and RF as close to the antenna’s as possibleRegular towers were used here for AM and short wave transmitions but 90% was deactivated at the end of last year.
I suppose building a tall concrete tower and putting the transmitter + processor closer to the antenna would reduce transmission line losses, but I can’t imagine the losses of proper high voltage cabling (run from the BOTTOM of the tower and up) would be so severe that this solution would actually SAVE money in the end.
I have never, ever heard of this being done.
///Leif
sgeirk,
[quote author=”sgeirk”]This preset literally blew us away today:
Plutionium NR
+.2
50
45
50
Normal Bass
-13 bass shapebass efx plug in: 2
impact/clunk: slam 2, clunk 2Crazy crazy crazy sound.[/quote]
I tried your setting and it gave a great light show as the all the BBP meters flashed on and off to red. I guess somethiing’s overloading something. I also noticed it made the studio Correlation Meter dance a lot more.Sounded great though.
Scott
- The forum ‘Breakaway Professional Products – [discontinued]’ is closed to new topics and replies.