I am using audioweaver with a dante network to feed audio to a system. When I play audio through the system without audioweaver, the songs sound normal. When I use audio weaver the songs end up sounding extremely distorted and sounds like it is pitch shifting down but not affecting the playback speed. Even on quiet volumes, the audio sounds very deep and like it is clipping. Even when I create a fresh audioweaver project with NO processing at all, just simply input to output, the same effect is happening. I suspect this could be sample rate related, but I am unsure. The audio track is 48k sample rate, and audioweaver is also set to 48k sample rate. I have tested multiple different audio files with multiple different sample rates.
Any ideas?
2:09pm
Generally, it's good to check that audio samples are being properly imported into AWE Core for processing, and then exported properly for playback.
To check your input, a good first step is to play a sinewave on your source device (e.g. 1khz, full-scale) and then look at what's coming in on the AWE Core's input using a Sink Display module. (To expand the time-window, you can either change the block-size for the layout (Input Pin properties) or you can add a Buffer-Up module in line before the Sink. - or you can adjust the sine's frequency such that multiple periods fit into the existing blocksize.)
To check the output, drop a Sin Gen module in and feed that to your output pin. (Even though you're ignoring it in this test, don't forget to connect your input pin to a sink or a meter or something - Designer doesn't like floating pins.)
2:46pm
It sounds like you're referring to file playback? I am actually using line-in and feeding audio from audio mulch if you are familiar with this. Does this principle still apply to line in or just for file playback?
2:50pm
No, it applies for any AWE Core integration. At some point, audio samples are being sent into AWE Core from the application/firmware. There's a few different things that could go wrong here, so just looking at the waveform of what's landing inside AWE Core helps to figure out what might be going wrong.
For example, AWE Core expects signed, 32-bit fractional data. If one has 16 or 24-bit samples, they need to be left shifted before sending them into AWE Core. If one doesn't do this, it's very evident in a Sink Display inspector, as your sin-wave data will be the wrong amplitude, and be rectified (as the sign bit will always be 0).
3:06pm
---
3:21pm
Okay I was able to successfully do the tests you mentioned. When I feed a 3k sine wave in from audio mulch it gives the very strange, horrible distortion sound. When I use a 3k audio weaver sine wave directly into the output, it gives me a clean 3k sine wave playing in the system.
We have a different audio weaver project running a different set of speakers that uses these same sound files being played out of audio mulch into audio weaver and they do not behave this way? I simply copied the files over to a new PC. Do you think this could be a 32-bit fractional data versus 16 or 24 bit? Are there any other possibilities for why I just got those results from above? How do you suggest I check what type of data I am receiving from these sound files in audio mulch?
3:27pm
I just found something else interesting that I can't explain. When I use the file player within audio weaver and play the 48k bach test track, it does the same distortion as sounds coming from audio mulch. Don't we know that file should be 32-bit fractional data?
3:38pm
Also, the only difference I can find in audioweaver settings between this not working project and our other working project is the input and output channels are 128 on this project, while only 32 on the other project. I am only using 32 channels in both systems, but wondering why 128 are showing up on the not working system and if this could have the distortion effect i'm experiencing.
3:42pm
Could this be an issue with the type conversion block that the input feeds into in audioweaver?
4:19pm
1 last thing, I made comparisons between our AWE server preferences working system and our not working system. The working system's "Basic block size" is 32 samples, while the not working system is 16 samples. The working system's input/output "pins" are both 32 channels, while the not working system are both 128 channels fract.
Just trying to give you as much info as possible to help guide you to a possible solution.
Thank you!!
9:55am
Hello,
Wow, 128 channels! You may be running into a limit on the number of input/output channels that we support for certain audio devices in the AWE Server. We plan on increasing this limit in a future release of Designer which should enable up to 256 input/output channels on the Native target.
Thanks
-Axel
10:52am
I am only actually using 32-38 channels of the 128 available. The I/O seems to have automatically detected and set itself to 128 channels. How do I manually change the I/O to only be the number of channels I need? Hopefully this will fix this issue.
10:55am
The number of channels shown in AWE_Server is inherited from the audio device's settings. You will have to configure your audio device (maybe this is your Dante Network) to limit the number of channels.
-Axel