Because the audiophile is broke, and will have to listen to some music on a lowly device, but the craving for some placebo is still there.

EDIT: btw, the bitrate is missing a k in your command 😉

Not audiophilic enough.

ffmpeg -i in.flac -ar 48000 \
            -af aresample=resampler=soxr:precision=28:cheby=1:dither_method=shibata \
            -c:a libopus -b:a 224k out.opus

There is no need to talk about an imaginary version of IPFS. GNUnet already exists. You can add that to the list of actually superior technologies that long predates IPFS.

As I mentioned, IPFS is nothing but very basic tech that got overhyped to junior/uninformed developers, and crypto scam victims.

Besides being overhyped basic tech where way more useful and practical solutions existed for decades (Freenet existed since year 2000 btw, and Tahoe-LAFS since 2007), there is nothing private about IPFS. This is a dangerous message to purport.

IPFS is as practically useful as NFTs. No wonder the two crowds connected well!

iroh is an attempt to create a useful and practical IPFS. But none of the bigger practical features is implemented yet. And the design itself doesn’t appear to be finalized. I’m willing to give iroh a chance, although the close proximity to the IPFS crowd doesn’t fill one with confidence.

The codec is basic, uses decades-old tech, and was trivially REed.

Lemmy instance choice does not check out ;)

and for testing you get a wackload of SBCs and Bluetooth chips and test that

I asked because I wasn’t aware of any consumer buds supporting Opus. I wasn’t aware of PineBuds, thank you for mentioning them.

keep in mind that it’s hard to get real numbers on LDAC because decoding is proprietary

I used to think the same. But as it turns out, a decoder exists. Maybe some people don’t want anyone to know about it to keep the myths alive ;)

EDIT: Also, as a golden rule, whenever anyone sees the words High-Res in an audio context, they should immediately realize that they are being bullshitted.

After testing LC3Plus, Opus, and AAC personally for bluetooth, LDAC claims are BS

How did you test Opus for bluetooth?

latency is significantly better then AAC (tested against libfdk) and marginally better then opus

In case you didn’t know, you can use 10ms (or even 5ms) frames with Opus instead of the default (20ms). 10ms should roughly match LC3plus’s default latency while still retaining high quality.

LDAC claims are completely bullshit.

LC3plus is worse than AAC quality wise (to be expected). Lower latency is the only thing going for it. And that’s just because AAC is a very high-latency codec. Opus (as a format) would win on both fronts, although there could be issues with creating a high-quality encoder for it that is not too complex, and power-efficient.

Or “Hyphanet” as it’s now called.

wtf, I missed that news.

Yes. That was what I’m alluding to when I wrote:

that architecture didn’t see large scale success before, except in Japan

Perfect Dark is a major network in Japan. Freenet is a network most people in the globe are not aware of. Hell, Perfect Dark may have a larger Japanese user-base than Freenet’s global one.

It’s worth mentioning that the former leader of the Freenet project wasn’t the most competent. Combine that with him spending years trying (and failing) to cater to the needs of imaginary dictatorships’ defectors (anyone of them using Freenet instead of Tor is the imaginary part), instead of focusing on maximizing the reliability and performance of the network to help its actual users. So it’s not just the ignorance of the masses that was at fault. The default FN user experience was often a horrible one. And users needed to ignore the officially-recommended microblog/forum applications, and even use a patched FN version, to get a decent performance out of the network.

Anyway, Freenet is the past and the present. And as I wrote in the parent comment, I hope a Freenet-like network would become a major success in the future, but I’m not holding my hopes up.

I do think it is the future of filesharing

In internet years, Torrenting is old. I2P is old. Even torrenting in I2P is old. Nothing about this is “the future”.

Ideally, the future of file sharing would involve a fully/natively integrated anonymous network with content-addressable distributed filesystem.

But this will probably not happen, as that architecture didn’t see large scale success before, except in Japan where at least some elements of this architecture are used in their popular P2P networks.

The I2P crowd themselves tried with Tahoe-LAFS, but that was never really a network, even aMule over I2P had more traction, and by traction I mean tens or hundreds of users, not thousands or beyond.

Ironically, the one content-addressable distributed filesystem that gained some attraction (outside Japan) is IPFS, which doesn’t offer anonymity, or replication, or anything special really. Yet for some reason, some hype-susceptible techies liked it, together with the NFT crowd, a great fit.

The future of file sharing will depend on where most content will land where it will be easily accessible and quickly grabbable. How those networks will look like? Nobody knows.

People really do have the memory of a goldfish.

Soulseek is an old-style P2P network. It has nothing to do with my parent comment. I personally don’t use it (see my other comments in this thread).

If you want to grab a non-reencoded file from YouTube, you can use a tool like yt-dlp

# see what formats are available for a YT vid
yt-dlp -F <youtube-url>
# format 251 is usually available as the highest quality Opus format
yt-dlp -f 251 <youtube-url>

That last command should grab you an Opus stream in WEBM format.

If you’re not a CLI guy, others should be able to give you a good GUI recommendation.

YouTube has audio in Opus format@~150kbit/s. Opus is a much better format than MP3. Almost all audio is completely transparent at that bitrate, where with MP3s, there are cases where audio is not transparent without using non standard >320kbit/s bitrates (a lot of content is transparent @320kbits/s though).

Now, sites/tools like the one you mentioned take the Opus (or AAC) file/stream from YouTube, and lossily re-encodes it again, probably to a file that is larger than the original, with at best the same quality, but probably worse quality. You obviously can’t get better output than the input in lossy compression.

So, the disk space argument is weird if you can play Opus/AAC (should be playable on every device nowadays).

This is the valid part for why you shouldn’t use YT-to-MP3 converters.

But there are also invalid reasons why people will tell you it’s shit:

• They think all MP3s sound like the shit ones from a decade (or two, or three) ago, using low bitrates and/or created with shit encoders. In reality, not all MP3s sound like shit, but vigilance is needed at every encoding step, as is the case with all lossy conversions.
• They are conflating the quality of the conversion, with the quality of the source, and think the bad quality of some user-uploaded YouTube content is due to the lossy conversion done by YouTube, and/or the MP3 converter re-encoding from YouTube. Content uploaded by the copyright holders (assuming basic competence) does not have that problem at all.

What hubs are you recommending that are better than Soulseek?

I hardly use DC++ anymore. I mentioned it because I didn’t find anything unique about Soulseek when I tried it last a few years ago. But I did grab plenty of classical music in lossless format from DC++, using public hubs.

So, it’s the “are better than Soulseek?” part of the question that intrigues me. What’s good about Soulseek? For lossless collections, it doesn’t (didn’t?) have much. For lossy stuff? There are better (in selection, availability, and quality) places to grab lossy files from (e.g. YouTube). And Torrents (with or without DC++) would probably have you covered there too anyway.

That is/was my experience with all these platforms/networks. I’m open to learn something new if I’m missing something.

Just as the other user said, using YouTube Audio for this stuff is the way to go.

Just look for <Artist Name> - Topic channels and check the playlists (not the uploads). You should find full albums uploaded directly by copyright holders. Use a VPN if you don’t find anything. Sometimes stuff from your region will not be available in your region, but available if you appear to be somewhere else ;)

Also, if we are going old style P2P, and not using torrents for some reason (RuTracker deserves a special mention), then DC++ should come before SoulseekQT/Nicotine+ anyway.

Mostly lossless grabs from torrents + YouTube Audio (edit: using yt-dlp), and you have a selection with guaranteed high quality*. Definitely better than whatever scattered MP3s in SoulseekQT/Nicotine+

* Opus@150kbits/s is transparent, except for some killer samples heard by a trained ear.