Opus support? {Implemented with a 3rd party plug-in by thinktink)

[thinktink]

...supported only via on the fly resampling)

Yes.

Opus Custom is not just unrecommended in this situation (plugin inside a media player.) The specifications for Opus inside the OggSquish container format (the standard, and only known used as of this post, container format for Opus files) does not accommodate for the Opus Custom API. It is only for specialized applications (of which Winamp is not one of them) where a developer's custom container format is used, which is not suitable at all in this situation.

[MrSinatra]

forgive me, I can be thick...

but if you are saying 44.1 is not supported at all natively, and only by on the fly conversion during playback, then why work to make the encoder able to encode at 44.1?

that's what i'm asking, assuming I understand the facts right?

[thinktink]

...why work to make the encoder able to encode at 44.1?
...

It's not about making the encoder able to encode at 44.1KHz. It's about adding an on-the-fly resampler to the encoder to convert 44.1KHz to 48KHz so that people can rip CDs directly to Opus using the encoder instead of having to rip to a different format to resample it to 48KHz before converting it to Opus.

[MrSinatra]

ok, its now finally coming together for me! thx for dragging me across the finish line.

so, when ripping to opus, the 44.1 will be converted on the fly to 48, and the end result will be a file that is 48. my confusion was in thinking this was about making CDs.

while I would not want to rip a normal redbook CD to anything other than 44.1, I guess I can see a scenario where someone demanded to have Opus rips of their stuff, and in that case that seems like the best way to do it I guess.

[Kaminari]

44.1 to 48 kHz upsampling is non-destructive.

http://hydrogenaudio.org/forums/index.php?showtopic=97051

[thinktink]

Yes, I've seen that thread more than several times. The issue isn't whether it's destructive or not. The issue is that the only references to an actual resampler in the current library is one that only does mono channels. I haven't found any other yet. To use it would require the daunting task of creating an encapsulating API that would need to deinterlace the incoming audio into each channels' separate buffer, chuck each buffer to each channels' own resampler (one for each channel), and then reinterlace it all back into a single buffer before passing it on to the encoder. Much easier said than done.


The thread also mentions a GPL resampler but fails to mention where to find it in the source. Kinda annoying.

[Aminifu]

while I would not want to rip a normal redbook CD to anything other than 44.1 ...

I think it is pass time for that standard to be updated. A larger bit depth and sampling rate would allow for better sounding files. The files would be larger so less would fit on a CD, but that is solved by using DVD discs instead of CDs. They cost the same in bulk. Just saying ...

[Kaminari]

Mmm, careful what you wish for...

http://people.xiph.org/~xiphmont/demo/neil-young.html

[Aminifu]

Mmm, careful what you wish for...

Thank you for the link to that article. I read it quickly, have to go back and reread some parts. But one thing is clear, facts are stubborn things (the human ear works the way it works). The distortion aspects are also key. Looks like the standard 16/44.1 got things mostly right.

I don't know why, but upsampling my mp3s does make them sound a little better to me (along with a little DSP doctoring, which helps even more). Going forward, I had already decided to use a lossless format (wavpack) for my personal ripping.

[innotect]

I am agree to your comment.

[Omega X]

Version 1.1.15.36 is being recognized as malware by Avast.

[thinktink]

Be specific please.

[thinktink]

Version 1.1.15.36 is being recognized as malware by Avast.

Unable to replicate:

Plugins/in_bpopus.dll

https://www.virustotal.com/en/file/f565d532ffad634c75cb94dfc4551b792a43bb64c4ac7f270c453479420f6a2b/analysis/

Plugins/enc_bpopus.dll

https://www.virustotal.com/en/file/c3ac9177dacfa5a9481497655ce32624b8652b92bde47f4f0266df32b6bcdc9b/analysis/

opusfile/libogg-0.dll

https://www.virustotal.com/en/file/d1f55969cbce51cbb17ceeea26604591ad22f97534195cc5dc379633c351e026/analysis/

opusfile/libopus-0.dll

https://www.virustotal.com/en/file/21e44692d3a73497c232d12d7f734bab65c5450bb3edc3682b9c8e55faa51730/analysis/

opusfile/libopusfile-0.dll

https://www.virustotal.com/en/file/87a8ae9759ea3f33fc814b371dc0846991bf3e64e448e3bcdd5b7f2b05864260/analysis/

BogProgOpus1_1_15_36Install.exe

https://www.virustotal.com/en/file/901134f2e946c02914ebf68096dea78e54cf941d39421a7debe66df090ca03d3/analysis/

[Kaminari]

Avira doesn't report suspicious signature either.

[Omega X]

The zip attachment was throwing off a warning.

What ever it was, its not triggering today. Might have just been a false positive.