Forum Replies Created
-
Posts
-
Hi Frank!
I’m working on the next version of Breakaway right now – I will attempt to reproduce this and do my best to fix it.
///Leif
Hi Bassman!
I believe there are a couple of bugs in the Child Process code – I’m currently in the process of trying to reproduce and find them. I’ve had to set up a machine park with every conceivable OS combination to be able to test this – pain in the neck 🙂.
What operating system are you running (and service pack)?
A temporary workaround (until I fix the problem properly) is to turn off Audio Realtime Priority. This completely disables the child process subsystem, and runs the audio engine in the same process as the user interface.
///Leif
Howdy!
We do have a multi-discount option! Click on the Order Now link from the main page http://www.claessonedwards.com/ – all the discounts are listed there.
Broadcast Version is in development 🙂. It’s taking a while, a few things have gotten in the way, but it’ll come.
///Leif
One time payment confirmed 🙂.
When you change computers, just email support with your new hardware ID and we’ll make a new key for you.
///Leif
Hi DarkFade!
I’ve been to Helsinki a couple of times. Went to Assembly ’94, ’95 and ’96 if I remember correctly. Nice city!
Thank you for your comments!
I’d love to do a custom box. I’m currently looking into the possibility of having a custom appliance-type PC manufactured inexpensively. It could be useful for all sorts of audio products!
I never heard back from JohnnyTheFox so I’m afraid I don’t know what the conclusions of the troubleshooting was.
I’ve never seen anything like those dropouts on any of my test systems, so I have no real way of tracking it down myself.
Virtual Audio Cable (and hence Breakaway Pipeline) is synchronized to the system clock, whereas the sound card has its own clock. Because these clocks aren’t locked together, the Input and Output of breakaway are unsynchronized, and run at slightly different rates. The Pipeline clock, as far as the system is concerned, is exactly 44,100.00Hz or 48,000.00hz, but the Sound Card could very well be running at 44093.97Hz or 48120.35Hz (or any arbitrary number you can think of, as long as it’s reasonably close). With a standard FIFO buffer, you would be guaranteed periodical dropouts or skips, as the buffer would regularly underrun or overrun.
However, Breakaway contains an adaptive sample rate converter. In real-time it detects the actual rates, and continuously (and very carefully) adjusts the sample rate conversion ratio, to keep the buffer at medium. If you’ve ever (on a heavily loaded system) noticed an audio drop-out, followed by a change in pitch, that’s likely the sample rate converter in Breakaway getting confused by the drop-out, and resetting itself to recover.
I can’t make any sense of this drop-out occuring after 6-7 hours, though. No buffer in breakaway is as long as a couple of seconds — not even on Huge buffering!
I could really use the help of someone experiencing the problem to run through the troubleshooting checklist I wrote before, to attempt to narrow down the problem.
There’s no manual yet – I’m not sure what to write in it other than what’s already on all the tooltips in the program – but if you have suggestions, I’ll be happy to listen 🙂.
///Leif
Jesse, actually Aeromax-HDFM *does* run on Windows 2000. The audio backend in Windows 2000 (kernel mixer etc) indeed does not support 192khz — like we’d want it in our signal path anyway!
Using either Kernel Streaming or ASIO, 192khz in Windows 2000 is no problem. In fact, KUVO has been on the air like this for two years!
Hard drives crashing is easily solved – use a plug-in IDE flash drive. 😉
///LEif
Hi Dennis!
When the software is ready for sale, I will definitely make a demo version. However, I already have an extensive demo you can play with. PM me your IM address (msn, aim or icq) if you’re interested.
///Leif
Hi Olli!
Radio Magic is actually designed to do exactly the thing you don’t like – that is, bring quiet parts way up, and bring loud parts down. You can counteract this by turning Power down – or use a different preset. Reference Settings retains a lot of dynamics, as does Zenith.
For small speakers, choose the setting with the overall sound you like, and then do the rest with Bass Boost. You should be able to squeeze some "body" into the sound by turning Bass Shape upwards. Remember, Bass Shape only controls what Bass Boost does, so if Bass Boost is at 0, Bass Shape does nothing!
///Leif
The loudness war in radio has been going on for almost 40 years, it’s certainly nothing new 😉.
The problem with overmodulation (and peaks) is not just purely regulatory. Apart from causing the receiver to lose lock (replacing S-sounds with white noise as if the station isn’t tuned in right), it can cause interference to neighbouring stations. Whether the authorities care about pirate broadcasters or not, they certainly will care if the licensed stations are calling in to complain about interference!
Also, if audio quality is not important — why do audio processing at all?
///Leif
Hi Smog!
Bad english?? Dude, you spell better than most native speakers, I can’t find a single error in your post, neither spelling nor grammatical! 🙂
It’s *possible* to write an application to output audio through two devices at the same time.
The problem is that because they’re different devices, they can never be perfectly synchronized. One pair of speakers would be slightly more delayed than the other, and create a strange sounding echo-effect if you tried to turn both on at the same time.
It could also be a configuration nightmare, confusing for a lot of users. Do you know of any other programs that are doing this? If so, please let me know – it might give me some ideas for how to do it sensibly. No promises, but I’ll take a look, and keep it in mind.
///Leif
I understand. That’s exactly the kind of graph you’ll get if you follow the instructions I painstakingly typed up in my previous post. 🙂
///Leif
Visualizing? I don’t know man – I usually listen with my ears 🙂.
Actually, the hard part about visualizing is that it’s not just one function, and it’s also not a static function, but dynamic. Part of the boost happens before the multiband compressor, which means the boost will be program-dependent.
One way to make a graph (even if it won’t be completely accurate) may be to play a slow sweep tone, log 20-20000hz, at a low level (maybe -20dB) through Breakaway, and with Range and Power set to minimum.
This way, the compressor will not affect it much (if at all), and you could record the output through your sound card, and simply look at the level of the waveform you recorded. Make sure you use Reference Settings for this.
Even with Bass Boost turned off, It won’t be *perfectly* flat due to the multiband crossovers, but it should be very close.
///Leif
Fnord,
do you mean "OctiMax"?
I’m well familiar with that algorithm. I’m the guy that wrote it – 8 years ago. Believe me when I say, there is nothing OctiMax does that Breakaway doesn’t do better. 🙂
There is positively no way to be competitive and sound good on FM with straight look-ahead limiting, while remaining within legal 100% modulation. The reason for this is that pre-emphasis causes huge high frequency peaks (at 75us, 15 KHz gets boosted 17dB!) and attempting to limit these with any method other than oversampled distortion-cancelled clipping will either punch a hole in the audio or cause distortion (or both).
If you’re overmodulating (which you would be if you were anywhere near as loud as clipping-based audio processors with a look-ahead based solution) it’s not comparing apples to apples anymore — and if overmodulating is allowed, the guy with the clipper can do just like you can, and kick your stations butt again 😉.
///Leif
Hi Fnord!
Yes, we’re aware of this problem.
The next release will have an update for Vista – it will still have the VAC chain, but with much lower latency.
After that, we will look into making it a Vista-style audio enhancement, but because it’s a longer term project, we’ll improve the current version first – so that we don’t keep Vista users waiting too long. 🙂
///Leif
