I'm a bit confused. Is there any sound quality penalty for buying and using the low-delay ASIO version opposed to the normal Breakaway Broadcast product? Also, I have access to a 2.39GHz Core 2 Duo computer with 4G of memory in it. The processor is lower than the recommended specs. Would it likely work OK? Thanks!
-
Get involved.
We want your input!
Apply for Membership and join the conversations about everything related to broadcasting.
After we receive your registration, a moderator will review it. After your registration is approved, you will be permitted to post.
If you use a disposable or false email address, your registration will be rejected.
After your membership is approved, please take a minute to tell us a little bit about yourself.
https://www.radiodiscussions.com/forums/introduce-yourself.1088/
Thanks in advance and have fun!
RadioDiscussions Administrators
Breakaway question
- Thread starter OKCRadioGuy
- Start date
No sound quality penalty as long as you don't actually use the low-latency mode! The ASIO version can run in phase linear mode as well, but then it is of course no longer low latency.
Low latency does incur a sound quality penalty compared to phase linear, but it should still be cleaner than any processor with similar latency. Breakaway still does have a first generation true psychoacoustic distortion masking clipper (Omnia.9 is second generation), even in low latency mode. It just doesn't have as much time to fine tune the result with the latency restrictions.
Your CPU is plenty fast enough. Core 2 Duo 2.4 GHz is not lower than Pentium 4 3.4 GHz. It's not the megahertz, it's what you do with it! Core 2 Duo was an enormous improvement over the Pentium 4, both in efficiency per clock cycle and power / heat efficiency.
Best,
//Leif
Low latency does incur a sound quality penalty compared to phase linear, but it should still be cleaner than any processor with similar latency. Breakaway still does have a first generation true psychoacoustic distortion masking clipper (Omnia.9 is second generation), even in low latency mode. It just doesn't have as much time to fine tune the result with the latency restrictions.
Your CPU is plenty fast enough. Core 2 Duo 2.4 GHz is not lower than Pentium 4 3.4 GHz. It's not the megahertz, it's what you do with it! Core 2 Duo was an enormous improvement over the Pentium 4, both in efficiency per clock cycle and power / heat efficiency.
Best,
//Leif
Leif,
Does the psychoacoustic clipper deal well with mp2 and mp3 source material since they are also applying a different psychoacoustic algorithm to the audio? I wonder if there could be interaction that would result in a sound effect similar to double compressing. In a perfect world, of course, the source would be uncompressed, but some stations just don't have that option.
Does the psychoacoustic clipper deal well with mp2 and mp3 source material since they are also applying a different psychoacoustic algorithm to the audio? I wonder if there could be interaction that would result in a sound effect similar to double compressing. In a perfect world, of course, the source would be uncompressed, but some stations just don't have that option.
Kmagrill, that's an interesting question!
Perceptual codecs use psychoacoustic analysis to figure out what they can get away with removing (to reduce the amount of data). Distortion-masking clippers use psychoacoustic analysis to figure out what they can get away with adding (to reduce the absolute peak level). The two do not interfere in the slightest.
Double compressing (MP3->MP3) is bad, but imagine what would happen if converted MP3 to a format which randomly throws away information (rather than analyzing and throwing away the least audible parts). I guarantee it would sound worse.
Likewise, cascaded clipping is bad (which is why the loudness war sucks), because the first stage (clipping during CD mastering) essentially adds random uncontrolled distortion. At least it will seem random after we've changed the frequency balance during multiband compression, as that changes the levels, brings some previously clipped peaks down to where they no longer needed to be clipped, as well as brings some previously unclipped peaks up to where they will need to be clipped before they hit the airwaves.
If the music industry used a distortion masking clipper, although that would still obviously be worse than not clipping at all, it would be better than what they are currently doing, because at least the distortion wouldn't be random crackle applied to the audio.
So, even if you cascade several stages of psychoacoustic distortion masking clipping, they will not interfere. The result will still be much better than cascading traditional clippers.
Also.. Although cascading MP3s are bad, think about how bad things used to be outside of the professional world. How about cascaded Cassette->Cassette copies with random head azimuths?
When I was little, me and a cousin played a version of the whispering game this way. One of us cascaded Cassette->Cassette copies using the built in microphones and speakers of two randomly selected boomboxes, and then the other got to guess what song it was :-D.
Best regards,
///Leif
Perceptual codecs use psychoacoustic analysis to figure out what they can get away with removing (to reduce the amount of data). Distortion-masking clippers use psychoacoustic analysis to figure out what they can get away with adding (to reduce the absolute peak level). The two do not interfere in the slightest.
Double compressing (MP3->MP3) is bad, but imagine what would happen if converted MP3 to a format which randomly throws away information (rather than analyzing and throwing away the least audible parts). I guarantee it would sound worse.
Likewise, cascaded clipping is bad (which is why the loudness war sucks), because the first stage (clipping during CD mastering) essentially adds random uncontrolled distortion. At least it will seem random after we've changed the frequency balance during multiband compression, as that changes the levels, brings some previously clipped peaks down to where they no longer needed to be clipped, as well as brings some previously unclipped peaks up to where they will need to be clipped before they hit the airwaves.
If the music industry used a distortion masking clipper, although that would still obviously be worse than not clipping at all, it would be better than what they are currently doing, because at least the distortion wouldn't be random crackle applied to the audio.
So, even if you cascade several stages of psychoacoustic distortion masking clipping, they will not interfere. The result will still be much better than cascading traditional clippers.
Also.. Although cascading MP3s are bad, think about how bad things used to be outside of the professional world. How about cascaded Cassette->Cassette copies with random head azimuths?
When I was little, me and a cousin played a version of the whispering game this way. One of us cascaded Cassette->Cassette copies using the built in microphones and speakers of two randomly selected boomboxes, and then the other got to guess what song it was :-D.
Best regards,
///Leif
Thanks for the clarification Leif. On the subject of dueling algorithms, I have some friends that have a station that is feed with a fractal T-1. The throughput on this link is only 512 so linear audio isn't an option. Their site is some 60 miles away and rolls through two different phone companies to get there. We used to have a Omnia 6 on the station and recently swapped it out for the 8600. It seems to my ears that there is something in the 8600 that actually does a fine job of masking dueling algorithm issues as this station is really bad about using mp3 audio given to them by record companies as downloads. I personally wouldn't ever allow that crud on the air and I would figure out how to get a real T-1 out there as I feel links shouldn't be compressed if you're a music station. I just find it a pleasant supprise that it seems to my ears that the newer processor seems to handle things better. Are designers of audio processors these days trying to take worst-case senarios like this in to consideration for real-world challenges from what is truely not the best engineering practice?
The format is Urban...
OKCRadioGuy said:
I can't speak for anyone else, but for me the answer is absolutely yes.
For example, Undo is all about that.
I test with roughly equal parts lossy and lossless, because that's how my own library happens to be, much like the typical radio station's.
//Leif
I watched Leif demo the de-clipper function of the Omnia.9 at the NAB. Indeed, it does seem to work. I didn't ask how it performs this magic, but I assume it starts with a FFT analysis of the harmonic content of the clipped waveform and goes from there. The result is you end up starting with source material that is better than what came on the CD so now the source can be processed very aggressively without hearing the audible distortion that was already present at the beginning. Nice trick.
In using the 9, there are only two situations where I've "broken" the distortion masking. One is with a Beatles record, where the vocal is off to one side and the other is quiet, you can hear some of the "junk" that is usually masked. I spoke to Leif about this awhile ago and he's working on it. The other is with songs that have a Hammond B3 or equivalent alone on a track... like the Band's "Chest Fever". I'm guessing the inherent distortion and harmonic content that is in that instrument makes the perceptual masker have to work hard to hide the clipping distortion among the harmonics already there.
Frank, Leif and Corny have all be super responsive and helpful setting up Ones, 11's and 9's. And they all LISTEN to feedback from their customers, and if it is valid, apply it to the next release. If it isn't valid, they will carefully explain why it won't apply in this case, but still keep an open mind if it comes up again.
How often do you get to give feedback directly to the creators of products? Try that with your TV, car, stove, computer, software, etc. It's like sending Bill Gates a message about Windows 7 and having him respond directly. Thanks again to all the folks at Omnia.
* Also to Wheatstone/Vorsis, I have found the same level of communication. And to Bob Orban for also answering questions and concerns on this board and others!
Frank, Leif and Corny have all be super responsive and helpful setting up Ones, 11's and 9's. And they all LISTEN to feedback from their customers, and if it is valid, apply it to the next release. If it isn't valid, they will carefully explain why it won't apply in this case, but still keep an open mind if it comes up again.
How often do you get to give feedback directly to the creators of products? Try that with your TV, car, stove, computer, software, etc. It's like sending Bill Gates a message about Windows 7 and having him respond directly. Thanks again to all the folks at Omnia.
* Also to Wheatstone/Vorsis, I have found the same level of communication. And to Bob Orban for also answering questions and concerns on this board and others!
Amen to that. Other broadcast manufacures could take some notes on real customer service found at Telos-Omnia-Axia, AudioArts-Wheatstone, Orban, and Continental, and to some degree Nautel.
Which begs the question : If you don't need the Low Latency mode, is there any advantage to the ASIO version compared to the standard BBP?konbaasiang said:
BobOnTheJob said:
Yes, BBP ASIO still has low latency studio output. And, perhaps more importantly: When I release the next version, the two will be combined (simultaneous ASIO + WDM with flexible routing in the same app, and low latency studio processing), and the base Breakaway Broadcast version will be discontinued. The new version will be $299, so it will be a free upgrade from BBP ASIO (no price increase!), whereas the upgrade from standard BBP to the new combined BBP will incur an upgrade fee.
If you elect to buy BBP and not buy the upgrade, you will of course still be able to use the BBP you paid for as is. We're not pulling support, we just won't sell any new licenses once the new version is out.
The combined version will offer several advantages, particularly for low budget installations. For example, you'd be able to run Input + Studio Output on an old M Audio Delta 44 you have laying around, with 192 kHz Composite out of the motherboard! Realtek chips with 192 kHz support performs beautifully electrically -- especially if you get a motherboard so low-end that they didn't bother with coupling capacitors, giving you a DC coupled output
.///Leif
Thanks Leif! I could be like everyone else and ask when the new BBP will be released, but I'm going to take the high road and be patientkonbaasiang said:BobOnTheJob said:
Yes, BBP ASIO still has low latency studio output. And, perhaps more importantly: When I release the next version, the two will be combined (simultaneous ASIO + WDM with flexible routing in the same app, and low latency studio processing), and the base Breakaway Broadcast version will be discontinued. The new version will be $299, so it will be a free upgrade from BBP ASIO (no price increase!), whereas the upgrade from standard BBP to the new combined BBP will incur an upgrade fee.
If you elect to buy BBP and not buy the upgrade, you will of course still be able to use the BBP you paid for as is. We're not pulling support, we just won't sell any new licenses once the new version is out.
The combined version will offer several advantages, particularly for low budget installations. For example, you'd be able to run Input + Studio Output on an old M Audio Delta 44 you have laying around, with 192 kHz Composite out of the motherboard! Realtek chips with 192 kHz support performs beautifully electrically -- especially if you get a motherboard so low-end that they didn't bother with coupling capacitors, giving you a DC coupled output
///Leif
. I'm playing with putting BBP ASIO on a low end machine just to see if it will work and learned 2 things:
1) 3.2ghz Single Core is iffy at best (strangely, it works better on Low Latency(75% CPU) than on Low CPU or Phase Linear(95-100% with burps). It seems to sound OK on Low Latency, but not something I'd risk on air given how close to the edge the CPU usage is.
2) When installing a Marian Trace Alpha, newer drivers are not always better! The current 2.36 version doesn't pass audio in or out of the Alpha (at least on this computer). The 2.11 that Marian suggested passes audio in and out of the card but BBP doesn't see it. My proven BBP computer had 2.01 in it, which I placed on the 3.2ghz machine and BBP recognized it fine.
Bob
Core 2 > Core > Pentium D > Pentium 4 HT > Pentium 4konbaasiang said:
Not to mention a Duo has 2 full cores (no HT tho).
It's capable of detecting clipping (and more importantly, what looks like clipping but actually isn't) on a per-sample basis, so... re: FFT, you're not on the right track as far as the root "effect" is concerned.Kmagrill said:
It's algorithmic in nature. As far as it being "a neat trick", I'll agree, and add that I think it's the biggest innovation in broadcast audio processing since the multi-band compressor. I still remember the day we first heard it fully automated, after years of me begging Leif to check out multi-band expansion (after most of a decade of using my own mid-side multi-band relative-threshold expansion in my mastering). We were both STUNNED not only at how well it worked, but how it also adds consistency (!!!) in a way where it sounds like every instrument is playing each note at the correct level for that particular beat/bar of music (in comparison with the more heavily clipped/limited original)."Ken Ruzicka Joins Nautel Sales Team" announced todayOKCRadioGuy said:
"an RF expert with 32 years' experience in broadcasting"They are certainly beefing up their stock of humans at Nautel which is great. Some things I really like about Nautel is that they support stuff from their first on up to now. Their stuff is pretty reliable ( think the Z series Harris is slightly better in that regard). Parts are more reasonable and available. I'm going to get a bit of ragging here, but at times there are a few guys at Nautel that aren't as knowlegable and polite as they could be. There can be a bit of attitude there. I'm not saying they are horrible by any stretch, but I'd give them a satisfaction rate a bit below Telos-Omnia, Orban, CE, etc. I've never recieved any foul attitudes from them ever.
- Status