Forum Replies Created
-
Posts
-
It’s a false positive.. It seems to be just random, and NOD seems to be particularly bad, it’s not the first time it has happened.
///Leif
Hi Martijn!
Yes, BBP ASIO is a completely new product.
People who own both Breakaway Live and Breakaway Broadcast will get it for free, and we’ll keep this promotion alive for a while, along with upgrade options.
///Leif
Argh! these stupid %#(/%!# antivirus scanners, I’m not even going to try to keep up.
Haven’t tried it on W7 but perhaps someone has? 🙂
///Leif
quote :Is there any benefit to running Speakers mode for streaming, or is it best left at Broadcast mode?There is, actually! The Encoder plug-in gets called before the speaker controller, so by running in Speaker mode, you can use the speaker controller to correct the sound of your studio monitors, while still broadcasting the signal pre-correction (as it should).
However, the PEQ in the speaker controller should not be used for broadcasting, because it’s after the final limiter of the Breakaway core. There is another limiter in the speaker section, but that’s a protection limiter just to prevent clipping the sound card, and it is not as clean sounding as the dual-band limiter in the core. Even if it was, it would still be double limiting and double IM-distortion due to the first limiter limiting some things it shouldn’t have, and missing some peaks, which means the second limiter has to do it all over again, and all that remains of the first limiters job is the distortion it added! It’d be a really bad idea for broadcasting. This, incidentally, is why chaining two processors is a catastrophically bad idea. The limiters in the first processor will be vigorously limiting peaks (and adding distortion) which will then be undone in the second processor, while the distortion will still remain. It’s easy, really — peak limiting has to happen at the very end of the chain to be effective.
I will add parametric eq to the input of future versions of BA Live and BBP though. Probably at the same time I add stereo enhancers. After that, I’ll be satisfied for quite a while 😉.
///Leif
quote :Does this version feature the same (improved) presets as the new non-ASIO version?It does!
///Leif
quote :10 seconds off of the air. try to flash the firmware & restart a DSP-based hardware unit that fast!LOL, excellent point, I never thought about it that way 😉.
///Leif
Stuart:
quote :I noticed your tooltip for the output meters (still?) refers to dark red that indicates peak level before clipping. I think that is an unintended holdover from BBP.Agh! You’re absolutely correct. I fixed the scale problem but completely forgot about the tooltips. Thank you — I’ve already fixed it in the source code, it’ll be in the next release.
Next release probably won’t take very long, I’m sure there’ll be more little bugfixes over the next few days.quote :What do the blue/green input meters below reference measure? The dynamics of them seem a bit different than the output meters. Is there some kind of averaging going on?Yes there is! The input meters are now ITU BS.1770 loudness meters. The colours are subjectively chosen to help the operator get the input level right (even though Breakaway will of course normalize the audio even if you don’t). There’s a slider in the settings window where you can adjust your ITU target level, which is the level you’re planning to feed into the processor. This actually adjusts the downward expanders and other parameters inside the core. It also moves the white markers around so that you can easily see where you want to be.
The meters are colour-coded according to how the input sounds if you don’t have a Breakaway core following it, which is indeed rather backwards, as they are sitting right at the input to the core. They’re coded to mimic how levels are perceived by humans, according to loudness standards for television. Green is good, yellow is a little too loud but okay for transients, red is too loud — it’s where someone would have to reach for their volume control to turn it down. Blue is dropping off the dial.
Breakaway will take basically any input level and get the output level in the green.
Here’s a screenshot from during development of my ITU BS.1770 implementation, including long-term graphs:
Left is input (before Breakaway), Right is output (after Breakaway). Standard "Reference Settings" were used!
The Red-all-the-way-to-the-top indication means clipping on the input. The source material here were different songs from different CDs.
DjSmooth:
There is no appreciable treble loss between 320 and 256kbps, you just get more artifacts. It can’t be fixed with EQ. However, if you want a sweet detail-boost EQ, try 15000hz, 1.0 octave, adjust gain to taste. Maybe +3.
///Leif
quote :Leif, can you explain what do u mean by using the words : full quality ?Sure. BBP ASIO, in low latency mode, sounds almost as good as regular BBP. Latency in the low latency algorithm has been cut by a factor of 50 — something has to go. The low latency mode is not phase linear. In fact, low latency phase linear is impossible if you have bass processing!
I would say the low latency mode is 90% of the quality of the full latency, full quality mode, on most program material, which still makes it much cleaner than the competitors. 🙂
What are the differences?
During heavy processing, the phase linear Breakaway Core is significantly more transparent than the low latency core. You can easily try this in Breakaway Live, where you can switch between Phase Linear and Low Latency, without changing anything else.
For the clipper, the full latency clipper is better at masking distortion, because it has much more time to analyze the audio, and many more stages to do the processing in. The low latency one has to hurry everything, so that it doesn’t delay beyond the human echo-fusion threshold.
That being said, I am very satisfied how clean the low latency version came out. When I first developed the BBP algorithm, I never would have dreamed it would be possible to make it this clean at realtime latency levels!
It is still xylophone safe, still telephone safe, and still the only family of FM processors which can claim this feat 😉.
To make a long story short, in Full Quality mode, BBP ASIO is not low latency, but sounds just as good as regular BBP.
Even if you use this mode, BBP ASIO still has the advantage of giving you a low latency studio monitoring output through the same sound card, from the same program! No need to combine BBP and BA Live anymore.
In fact, even if you have just a 2-channel sound card (such as ESI Juli@ or M Audio AP192), you can still take advantage of this, by using the Left output for MPX, and the Right output for mono studio output.
You could probably also use the S/PDIF output to get stereo studio output, but you’d need a DAC that can accept 192k s/pdif input.
quote :DOES THIS ASIO VERSION…HAVE MPX CLIPPER??No. Sorry!
Actually, during my experiments I’ve been unable to make the composite clipper perform acceptably with less than 50ms latency or so, even using all the tricks I learned during development of the BBP ASIO low latency. I don’t have a lot of hope that it will be able to be fast enough. The clipper has to have less than 10ms latency (L/R input to finished MPX output) to be able to function realtime, because ASIO also adds latency, and multiband processing also adds latency.
///Leif
quote :I have 4 ‘zones’ that can each play different music or audio streams. Basically 4 instances of winamp running concurrently. I suppose then I’d need to have 4 BA ‘cores’ running concurrently as well. Is that doable?Yes — and it’s cheaper than buying 4 copies of Breakaway Live (Live can actually run four stereo pairs in one instance of the program if you have a 4-instance license). However, it make be beyond the budget of a home system – a 4-instance license is $379. It depends on how serious you are. If you do use PEQ for each zone, each speaker separately, then it’s one hell of a bargain, for getting truly excellent audio quality. A single M Audio Delta 1010LT would take care of all 8 channels in a low latency fashion.
However, it currently offers EQ only on the first core. I think I’ll change that, and offer 2-channel eq (one pair of speakers) on every core.
quote :The speaker EQ is a very interesting concept. The fact that each channel is fully parametric offers HUGE flexibility. Are you going to allow the user to shoot pink noise out from the program directly and read its results (after a quick pinknoise loopback sound card calibration?)Shoot pink noise from the program directly? Heck yes! Quick loopback calibration? No, I’m afraid not, because I don’t believe it can be done well. Yes, there’s hardware that does it — note that I included the word well. 🙂
The thing is, when you’re measuring, you get cancellations and echos all over the place, which affect the measured frequency response. The brain is very good at compensating for this, so we don’t even notice it — but nothing escapes the microphone. When calibrating manually, you naturally move the microphone around to get a good average, and get a "feel" for what’s a real peak and what’s just a temporary thing right where the microphone is. The former should be EQ’ed, the latter should not.
So, calibrating a system, for me, is an iterative process that takes about half an hour for a pair of speakers, maybe more.
I’ve released a free program called Breakaway RTA which makes PEQ-calibration possible. Check the Breakaway Professional Products forum!
I’ve also actually just released Breakaway Live 0.90.74 with all these new features (Except EQ only on the first core). Check it out.
///Leif
Hi Stuart!
Man you’re quick 🙂. It does not have the toolbar yet, simply because I haven’t had a chance do do it. It will in the next release though, for sure. It does have the volume control popup (when you single left-click on the tray icon) already!
Riplee, the RTA is not built in, because it’s really not very convenient to have the RTA built into the audio processor.. That’s why I released Breakaway RTA completely free!
The PEQ, on the other hand, is built in. 🙂
///Leif
I understand what you mean.. But hey, regarding the crossfading: Unless the input and output are widely disparate level, shouldn’t crossfading be smoother with the processor after than with the processor before the crossfader? With processing after, you get a highly accurate blend at the exact point of transition, but with processing before, you get a true fadeout before the next song starts, with nothing to counteract the fadeout. I guess it’s a matter of preference — I personally prefer the processing after the crossfading 🙂.
Regarding 4 zones, that all depends! By 4 zones do you mean playing different music, or playing the same music with different processing settings, or just 4 amps hooked up to one output so that they can have independent volume?
Breakaway Live (next version, any day now, i promise) will have some things no other of my processors have, including 8 speaker output with individual 24-band (yes that’s 24-band) PARAMETRIC eq, allowing for extremely accurate speaker equalization and a real speaker quality improvement. One could say that the Breakaway core itself subjectively improves the quality of the music, but speaker equalization objectively improves the quality of the speakers!
Speaker equalization takes patience, skill, pink noise, and a $50 Behringer ECM8000 microphone. It is SO worth the effort — you can expect a night-and-day improvement on almost any speaker, because speakers are very rarely flat frequency response.
Here’s 12 bands of EQ for one pair of speakers. Scroll down (in the actual program) for more. Four pairs of speakers are supported, with separate eq, and all for the same price as before — $129. All are fed from the same Breakaway processing core, though, so it’ll be the same music in all four pairs.
///Leif
I think so.. It doesn’t make much sense to spend the development time to only work with one player, when Breakaway Audio Enhancer already works great with Winamp as well as other programs. Also I believe Winamp’s popularity is probably going down, not up. If there’s one player to target it’d probably be iTunes.
(I personally use Winamp!)
///Leif
Actually, it seems I have to officially withdraw my recommendation of the ASUS Xonar DX!
I’ve used the card a lot, and it has worked great for me — but I only ever used the output.
Running with the ASIO driver, output is flawless, down to buffer size 96 (2ms output latency)! I’ve used it both for MPX (Breakaway Broadcast) and for 5.1 output for a different project. Worked flawlessly — but at no time did I ever use the Line In jack. I always fed audio through some other method, such as my own proprietary Virtual Audio Link protocol (for uncompressed audio over TCP/IP).
…until today, when a customer asked whether the latency on this card is any good, before he buys it. So, I figured I’d just hook up a mic and headphones and see how low the latency really is.
And the result is: crap!
Even though the output latency is only 2ms, the input at those rates is completely garbled. In fact I had to increase latency to 40ms before the input cleared up! I have NEVER seen an ASIO driver behave in this way before, so I never would have dreamed this would be the case.
I also did a firmware update on the card and installed a newer driver. And tried the card in a different computer. And tried another card in yet another computer. Same exact problem.
So, my recommendation for this card is officially withdrawn — there’s no way to make it work!
So, my experience actually mirrors Diekgait’s with the more expensive Xonar D2X, which means BOTH of them are crap.
My recommendation of ESI Juli@ and M Audio AP192 still stands though. 😉
///Leif
Heh heh 🙂. I actually do care about safety even though it may not look like it. Unlike common practice in Thailand, all my equipment is grounded, and protected with 4 stages of surge protection (two inside power strips, one in the UPS, and one (the first stage) in the form of dual 6A breakers, cutting both live and neutral, followed by a surge arrestor. There’s also about 20 knots in the supply cable before the first set of breakers.
In the rainy season in Thailand, there’s rain and thunder absolutely every day. Having to shut down and unplug the server every day turned out not to be feasible. Not only is it no good for the drives to be shut down that often, but while the server is down I can’t work.
A nice advantage to having a big cabinet instead of a small normal server case is that I can use 120mm fans instead of the Soloviev-like microscopic fans that are common in servers, which would make it unbearable to be in the same room. This server actually makes no more noise than the air conditioner! 🙂
///Leif
I didn’t feel like wasting the money on a server case, so I had a metal worker build me some nice disk towers instead.
I am using a server grade board, 3ware hardware raid-5 controllers, and two enermax supplies (650 and 550w) though. Just not a "normal" server case. Hell, they’re too loud anyway 😉.
///Leif
