- This topic has 79 replies, 14 voices, and was last updated 12 years, 3 months ago by
.
-
Posts
-
John has been a busy coder in the past few weeks! He’s made some major changes to the edcast tool. Some of those include a new limiter (currently hard coded to -3db) and a new pre-emph (coded to 55us). Those two are combined at this current release but he plans to separate them and make them adjustable. He’s also re-written a lot of the internal audio-routing code to make things much quicker and stable.
My concern is if/how the limiter/pre-emph will ultimately drop the quality of Breakaway Live. I think most of us are here for sonic purity and low-to-no distortion and I wonder if this is counter-productive.
I have no equipment and don’t have a very good ear for sound nuances. I’d like some of you guys to get the latest beta version from his site and try the limiter/pre-emph feature and see if you can get some feedback to us here on how it affected your quality.
Here’s the link: http://users.tpg.com.au/rioradio
Thanks!
the limiter is only used for pre-emphasised mp3s, since emphasis destroys peak control. 😉 and (hopefully) nobody actually uses pre-emphasized mp3s. so those are changes that shouldn’t effect anyone.
Jesse, in Breakaway broadcast, Leif has suggested setting pre-emph to 15us for streaming. There was some reason for this that I’ve forgotten. However, wouldn’t turning on 55us pre-emph do something to the audio regardless of the source files? This is aside from the limiter.
What Leif suggested is doing that within the confines of his clipper, with the de-emphasis enabled so that the mp3 actually has a tad *less* peaks in the treble. Running pre-emphasis flag with pre-emphasis (and a limiter) in mp3 is the opposite, because it’s trying to get rid of noise that basically doesn’t exist, at the expense of loss of peak control, into a limiter to regain peak control (adding distortion), and then encoded into a codec that already has issues with high frequencies.
The other reason is you can’t be sure the decoder will properly handle pre-emphasized mp3s, since they really aren’t used ever, that i’ve seen, at all anymore. It’s a nice trick feature to have a fully capable mp3 encoder for streaming, but it doesn’t mean you’ll even want to ever do everything it’s capable of within the mpeg spec. 😉
Hello again
I don’t pretend to understand the why’s of limiter and/or pre-emphasis. I received a request to add these functions by someone who has a lot of "poor quality" mp3’s. Limiter was easy to code. Pre-emphasis, hell, I didn’t even know what it meant – so I looked it up, and coded it.
The guy with the problem swears his "poor quality" mp3’s are now "significantly improved" at the receiving end.
Hey, who am I to argue.
Anyway – I’ve added controls to edcast to play with limit level / pre-emphasis … limit level is probably WAY over kill down to -30dB 😉 pre-emphasis is 0 to 75us … I chose 75, because all my reading of pre-emphasis mentioned values up to about 75!!
Anyway – it’s there to play with – or not http://users.tpg.com.au/rioradio – I’ve significantly lowered CPU usage – the audio "paths" within edcast were over complicated with way too much malloc’ing going on for my liking
P.S. sorry jesseg if you received this in a private message … don’t know how I managed that!!!
Is it possible to have:
2 streams (mp3 and aac for example). MP3 stream gets audio from pipeline2, AAC stream gets audio from pipeline 3?
My reason for this being I would like to have one stream processed (from pipeline 2) and the other stream un-processed (from pipeline 3).
In a nutshell, selectable audio routing for the various streams I may have.
I could of course be missing a trick here and its already available! I know I could run another instance of edcast but have found this confusing and unreliable (confusing the windows are titled the same).
Any feedback welcome
In edast? yes it is – sort of
I have an ASIO version which does that, mono only as that was the request I received – very specific audio card I have to say – the guy can play 8 different mono audio channels to it – each one available for input via an ASIO channel – each encoder in edcast can be configured to read a different channel – I’ve never tried to make it stereo – though, with the recent changes I’ve made, it would be easy to do so.
that’s the "standalone" version
the DSP version … can you send more than 2 channels via winamp DSP API? if so, I can say that it would be perfectly possible to do what you need. Of course, the source has to be able to send that way
I just tried the new version and it kept throwing integer between 0 and 30 errors. And the sliders for the limiters are all jacked up showing some huge numbers. And the integer errors fill my screen if I slide those sliders!
Anyone else try this and have that experience?
I today updated to http://users.tpg.com.au/rioradio/edcast … 1.1012.exe
All seems ok. (for now)Hi radiorio,
I am using Sam Broadcaster and its set to output to pipeline1. The same principle could be applied to Winamp though – only one output is output (!) but its what you do with it that gives you two streams.
So…
Pipeline1 is unprocessed (straight from Sam, Winamp etc)
Breakaway processes this and spits out the processed audio onto Pipeline2In Edcast (ideally!):
Stream 1 (AAC) = the unprocessed Pipeline1 (untouched straight from Sam, Winamp etc)
Stream 2 (MP3) = the processed Pipeline2Make sense?
[quote author=”Boki”]I today updated to http://users.tpg.com.au/rioradio/edcast … 1.1012.exe
All seems ok. (for now)[/quote]
dsp version 3.22 is badly broken – stupid sliders 😉version 3.23 is good
[quote author=”knigget”]
In Edcast (ideally!):
Stream 1 (AAC) = the unprocessed Pipeline1 (untouched straight from Sam, Winamp etc)
Stream 2 (MP3) = the processed Pipeline2Make sense?[/quote]
Not really … how does one get two pipelines to one edcastedcast (DSP version) can
a) get its input from a sound card
b) get its input via winamp dsp API – i.e. encode_samples(struct winampDSPModule *this_mod, short int *short_samples, int numsamples, int bps, int nch, int srate)as I said, from my end, I can "make" edcast able to handle multiple different sources – the issue is, what the sending end can do, i.e. HOW it intends to send two different streams of audio to edcast
[quote author=”timmywa”]I just tried the new version and it kept throwing integer between 0 and 30 errors. And the sliders for the limiters are all jacked up showing some huge numbers. And the integer errors fill my screen if I slide those sliders!
Anyone else try this and have that experience?[/quote]
fixed in 3.23Up until recently (I was running the original Edcast) I was able to press a button within Edcast and it would stream audio from the soundcard OR from a pipeline of my choice by a dropdown list (whichever I chose would be applied to ALL streams). I could verify this by turning the Breakaway output up and down and this would reflect on the stream levels.
HTH
Attached is a screenshot of the best example I could find – see how this user has Virtual Audio Cables available as a source for the stream
[quote author=”knigget”]Up until recently (I was running the original Edcast) I was able to press a button within Edcast and it would stream audio from the soundcard OR from a pipeline of my choice by a dropdown list (whichever I chose would be applied to ALL streams). I could verify this by turning the Breakaway output up and down and this would reflect on the stream levels.
HTH
Attached is a screenshot of the best example I could find – see how this user has Virtual Audio Cables available as a source for the stream[/quote]
You’re saying the dsp version doesn’t detect your soundcard? Someone else had an issue with that. turns out the latest BASS.dll’s are at fault – out of my hands I’m afraid – though, I could use the older BASS.dll I guess – but then the older ones have issues in vista/7having said that – IF I can resolve the soundcard issue – I can fairly easily add the ability to select soundcard or "dsp" per encoder
- The forum ‘Breakaway Professional Products – [discontinued]’ is closed to new topics and replies.