2013/7/4 Florian Jung <[email protected]>:
> Am 03.07.2013 22:59, schrieb Robert Jonsson:
>> Hi Florian,
>>
>> 2013/7/2 Robert Jonsson <[email protected]>:
>>> Good morning Florian,
>>>
>>> 2013/7/2 Florian Jung <[email protected]>:
>>>> Hi guys!
>>>>
>>>> good news for you: My AudioStream changes do now actually work!
>>>>
>>>> There's still lots of work to do, occasional segfaults, audio glitches
>>>> due to seeking bugs, and a HUGE lot of integration with the rest of
>>>> MusE. But basically, the hard part is done :)
>>>>
>>>> Try it out:
>>>>
>>>> git checkout audiostreams
>>>> maybe git pull
>>
>> Didn't manage to get the code this way, not sure what I'm missing..
>
> you have to be inside a cloned working copy of the repo, of course.
>
> git clone <whatever> first. then cd whatever, and THEN git checkout
> audiostreams. should do.

Ok, I need to experiment more with this. I actually tried to make a
commit some weeks ago but failed, have to try that again.

>
>
>> There was an option to download a zip file of the branch which worked fine.
>>
>>>>
>>>> ensure that LibRubberBand is installed and used by cmake
>>>> build, start
>>>>
>>>> then create an audio track
>>>> import a wave file
>>>> (do not move the created part! moving is not fully implemented yet, as
>>>> noted above)
>>>>
>>>> tinker with the tempo map, then play the song.
>>>>
>>>> While you are not yet able to set a wave-file-internal tempo map, you
>>>> see and hear how the audio is stretched. Pretty good quality, as i
>>>> think, even if you go for pretty extreme ratios (50% or 200% of original
>>>> speed)
>>
>> Got it working, got it to crash too :)
>> Fun to manipulate the tempo map! Does take quite a bit of cpu on this
>> old laptop though.
>> Also, I'm hearing artifacts in the audio, I hope that can be improved
>> or just a problem for me, it doesn't sound that usable yet.
>
> what kinds of artifacts? A strong noise? That happens with mono files,
> ATM, due to lazyness/"not implemented yet" ;) stereo should work

It was stereo and sounded like a pitching artefact. I'll try to
provide some comparison sound files.
It was there even though I didn't change the pitch, which I think
should be possible to make _very_ transparent even if the processing
is active so I'm hoping it is fixable.

Regards,
Robert


> however, i found out that neither aligns the played back audio with the
> displayed wave form, nor does one of both align with the beats like it
> should. gotta adjust...
>
>
>
>>
>> Great work nevertheless!
>
> thanks :)
>
> greetings
> flo
>>
>> Regards,
>> Robert
>>
>>>>
>>>>
>>>>
>>>>
>>>> By the way:
>>>>
>>>> it already supports (should support, not tested yet) sampling rate
>>>> conversion-only (for people who don't want rubberband for any weird
>>>> reason), and i've planned that it will do "naive" stretching (i.e.
>>>> modify both tempo and pitch). That might create funny effects :)
>>>>
>>>> Don't get too excited however, there are lots of infrastructural changes
>>>> needed to get this into master.
>>>>
>>>> But it will come. And it will be great :)
>>>
>>> Sounds awesome! I'll check out the repo when I get home!
>>>
>>> Regards,
>>> Robert
>

------------------------------------------------------------------------------
This SF.net email is sponsored by Windows:

Build for Windows Store.

http://p.sf.net/sfu/windows-dev2dev
_______________________________________________
Lmuse-developer mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/lmuse-developer

Reply via email to