I've been searching high and low for an example on how to use the Speex library's preprocessor for multichannel audio.
The documentation for speex_preprocess_state_init() says that:
Creates a new preprocessing state. You MUST create one state per channel processed.
I assume that means I need to call speex_preprocess_run() on each channel separately, but won't that potentially "skew" the result if the preprocessor happens to remove more noise from one channel than the other?
Also, speex_preprocess_run() indicates whether the audio is considered voice or noise/silence. If I have to call the function for each channel, what happens if one channel is considered voice and the other isn't?
Am I overthinking this?
Voices recorded in stereo typically mix down to mono without trouble. Microphone placement can cause some phasing issues, but that generally isn't an issue.
Once you mix down to mono, you can process the audio as normal.
Alternatively, you can pick one of the channels, and ignore the second. This might not be as reliable though, as the voice might have been off-axis when recorded.
Related
I'm making a game in which there are a series of events (which happens, say, every 30 frames in a 60fps setting) that I want to sync with the music (at 120 bpm). In usual cases, e.g. rhythm games, syncing the events to the music is easier, because human seems to perceive much smaller gaps in music than in videos. However, in my case, the game heavily depends on frame-based time, and a lot of things will break if I change the schedule of my series of events.
After a lot of experiments, it seems to me almost impossible to tweak the music without disturbing the human ear: A jump of ~1ms is noticeable, a ~10ms discrepancy between video and audio is noticeable, a 0.5% change in the pitch is noticeable. And I don't have handy tools to speed up audio without changing the pitch.
What is the easiest way out in this circumstance? Is there any reference on this subject that I can refer to? Any advice is appreciated!
The method I that I successfully use (in Java) is to route the playback signal through a path that allows the counting of PCM frames (audio frames run at rates like 44100 fps, as opposed to screen updates which run at rates like 60 fps). I don't know about other languages, but with Java, this can be done by outputting using a SourceDataLine class. As the audio frame count is incremented, it can be compared to the next item (pending item) on a collection of events that require triggers to other systems or threads. Java has an excellent class for handling the collection of events: ConcurrentSkipListSet. It is asynchronous, and automatically sorts elements via a Comparator set to the desired PCM frame count.
Some example code that showing the counting of frames can be seen in this tutorial Using Files and Format Converters, if you search on the page for the phrase "Here, do something useful with the audio data". They are counting bytes, not PCM frames, but the example does give the basic idea.
Why is counting PCM effective? I think this has to do with the fact that this code (in Java) is the closest we get to the point where audio data is fed to the native code controlling the sound system, and that this code employs a blocking queue. Thus, the write operations only happen when the audio system is ready to receive and playback more sound data, and audio systems have to be very accurate in how they maintain their rate of processing. The amount of time variance that occurs here (especially if the thread is given a high priority) is smaller than the time variance incurred by choices made by the JVM as it juggles multiple threads and processes.
I want to convert a FLAC file to a MP3 (and Vorbis, in a second time) file.
These MP3/Vorbis streams are then transmitted, raw, to a second device that decodes them.
Depending on the quality of the transmission, I want to be able to change the bitrate on-the-fly.
The change must be gapless (hence the "in the PLAYING state" in the title).
The specific encoders are lamemp3enc and vorbisenc (and cannot be changed).
To my knowledge, changing the bitrate while playing is actually not possible with these codecs.
But I guess there are clean and simple ways to change the bitrate without introducing any gap in the stream: I'd like to learn about any of them.
(NB: I did write any, not all, I am not asking for the "best" way, I am not asking for a review, I just want something that works.)
Read through this ..
You will:
block element before lamemp3enc
flush the encoded frames into queue with EOS sent to lame and discard the EOS when it comes out of lame
then set the lamemp3enc into NULL state
change the parameters
set lame to PLAYING or PAUSED - this will preroll it again with new data using new bitrate
check when lame is in playing and then you know everything is working
there should be no gap as queue has lot of old buffers which it sends forward during you are doing the witch
You can inspire yourself with the example from the link above.. However you are not doing any removing and adding new elements.. Do not forget to set it to NULL state as it will discard all internal states (well hopefully if its not buggy). Then You would just change the parameters with g_object_set...
Also I never did this so you may also ask on IRC of #gstreamer at freenode if you are stuck or not sure.
HTH
I have a windows phone 8 app which plays audio streams from a remote location or local files using the BackgroundAudioPlayer. I now want to be able to add audio effects, for example, reverb or echo, etc...
Please could you advise me on how to do this? I haven't been able to find a way of hooking extra audio processing code into the pipeline of audio processing even through I've read much about WASAPI, XAudio2 and looked at many code examples.
Note that the app is written in C# but, from my previous experience with writing audio processing code, I know that I should be writing the audio code in native C++. Roughly speaking, I need to find a point at which there is an audio buffer containing raw PCM data which I can use as an input for my audio processing code which will then write either back to the same buffer or to another buffer which is read by the next stage of audio processing. There need to be ways of synchronizing what happens in my code with the rest of the phone's audio processing mechanisms and, of course, the process needs to be very fast so as not to cause audio glitches. Or something like that; I'm used to how VST works, not how such things might work in the Windows Phone world.
Looking forward to seeing what you suggest...
Kind regards,
Matt Daley
I need to find a point at which there is an audio buffer containing
raw PCM data
AFAIK there's no such point. This MSDN page hints that audio/video decoding is performed not by the OS, but by the Qualcomm chip itself.
You can use something like Mp3Sharp for decoding. This way the mp3 will be decoded on the CPU by your managed code, you can interfere / process however you like, then feed the PCM into the media stream source. Main downside - battery life: the hardware-provided codecs should be much more power-efficient.
I have a program written in C++ that uses RtAudio ( Directsound ) to capture and playback audio at 48kHz samplerate.
The input capture uses a callback option. The callback writes data to a ringbuffer.
The output is a blocking write function in a separate thread that reads from the ringbuffer.
If the input and output devices are the same the audio loops thru perfectly.
Now I want to get audio from device 1 and playback on device 2. Each device has its own sampleclock set to 48kHz but are not in sync. After a couple of seconds the input and output are out of sync.
Is it possible to sync two independent oudio devices?
There are two challenges you face:
getting the two devices to start at the same time.
getting the two devices to stay in sync.
Both of these tasks are difficult. In the pro audio world, #2 is accomplished with special hardware to sync the word-clocks of multiple devices. It can also be done with a high quality video signal. I believe it can also be done with firewire devices, but I'm not sure how that works. In practice, I have used devices with no sync ("wild") and gotten very reasonable sync for up to an hour or two. Depending on what you are trying to do, the sync should not drift more than a few milliseconds over the course of a few minutes. If it does, you can consider your hardware broken (of course, cheap hardware is often broken).
As for #1, I'm not sure this is possible in any reliable sense with directsound. To the extent that it's possible with any audio API, it is difficult at best: both cards have streams that require some time to setup, open and start playing. In general, the solution is to use an API where this time is super low (ASIO, for example). This works reasonably well for applications like video, but I don't know if it really solves the problem in general.
If you really need to solve this problem, you could open both cards, starting to play silence, and use the timing information generated by the cards to establish the delay between putting data into the card and its eventual playback (this will be different for each card and probably each time you run) and use that data to calculate when to start actual playback. I don't know if RTAudio supplies the necessary timing information, but PortAudio does. This document may help.
I am trying to write an application(I'm a gui first timer) for my son, he has autism. There is a video player in the top half and a text entry area in the bottom. When letters are typed sounds are produced to mimic the words in the video.
There have been other posts on this site in regard to playing sounds on key presses, using gstreamer as a system call. I have also tried libcanberra but both seem to have significant delays between sounds. I can write the app in python or C but will likely do at least some of it in C.
I also want to mention that the video portion is being played by gstreamer. I tried to create two instances of gstreamer, to avoid expensive system calls but the audio instance seemed to kill the app when called.
If anyone has any tips on creating faster responding sounds I would really appreciate it.
You can upload a raw audio sample directly to PulseAudio so there will be no decoding and (perhaps save) extra switches by using the following function from Canberra:
http://developer.gnome.org/libcanberra/unstable/libcanberra-canberra.html#ca-context-cache
The next ca_context_play() will use it.
However, the biggest problem you'll encounter with this scenario (with simultaneous video playback) is that the audio device might be configured with large latency with PulseAudio (up to 1/2s or more for normal playback). It may be reasonable to file a bug to libcanberra to support a LOW_LATENCY flag, as it currently doesn't attempt to minimize delay for sound events afaik. That would be great to have.
GStreamer pulsesink could probably get low latency too (it has some properties for that), but I am afraid it won't be as lightweight as libcanberra, and you won't be able to cache a sample for instance. Ideally, GStreamer could also learn to cache samples, or pre-fill PulseAudio...