-

2025-06-30 22:36:45

jrp@hub.kliklak.net

Anyone else having issues with the streaming sound on #Mirlo? My uploaded files tend to sound boomy, small, distorted. I am not experiencing this to this grade with other sites (e.g. Bandcamp) otherwise.

2025-07-01 07:59:07

Herr Irrtum! 🔊

herr_irrtum@s.basspistol.org

@jrp@hub.kliklak.net

Am here with my phone, so I can't do much of an analysis.
First Idea:
When you say "streaming sound": If you download those tracks again from Mirlo, are they fine then? *)
Because this sounds to me like there is processing happening right at your uploaded files. If so, that would proof it. *)

I have only one Track with a bigger dynamic range (not anything maxed out from the beginning ;) ). And here with my phone that sounds louder at Mirlo compared to the local file at my phone. Not sure if that says something, as I may have uploaded a normalized variant. On the phones speakers however it doesn't sound thinner or something. Just the volume is raised. But sound wise I shouldn't rely on phone speakers. Also again: I'm not sure, if I uploaded the variant that rests on my phone. Will check tonight.

*) Just realized there is no option to do this (should I enable this or do I have to remove the pay option?). Maybe with the "Media" function from Firefox or a Firefox-Fork? But then that would be the streaming variant. Sorry for the bad advice then.

#mirlo

2025-07-01 21:10:10

Herr Irrtum! 🔊

herr_irrtum@s.basspistol.org

Two times the same song. The above two tracks are (the stereo) representation as played on Mirlo. The two tracks below are the (stereo) Wav master from 2001. They look more or less identical to me. Doesn't change when switching to a fft spectral view btw <br />(not shown at the picture).

@jrp@hub.kliklak.net

So I looked into it. I played my tune "Gagarin" directly from the Mirlo site and recorded the internal monitor of my soundcard - so I have a 1:1 digital representation.

Then I put the original "master" recording directly below. Without doing anything to any of those tracks - they look (and sound) 1:1 similar to me. I know, there will be fine differences due to lossy compression, but in general: This is fine. In my case.

However:
@arod@social.coop - I know for a fact that Jayrope has a very fine and sensitive hearing. He certainly doesn't make this up. Do you have any explanation how such a thing may happen perhaps? Or know someone who does? TL'DR: His tracks suddenly sound "boomy and thinner" at the same time on Mirlo.

Maybe the Mirlo servers do some kind of bit rate reduction, when there is too much load on their servers? That would be a possible explanation.

2025-07-01 22:43:39

*_jߍyrope

jrp@hub.kliklak.net

@Herr Irrtum! Have you done the phase reversed comparison? (Simple math for audio engineers)

And of course they will not sound exactly the same, because you'll compare an uncompressed source with a compressed source. My question is, how do they differ exactly? Method to compare is align sample exact, then reverse the phase on either one of them. The result is the mathematical difference of the two files. Provided, that they actually run at the exact same volume one should be able to hear and see the difference. I'm going to have to do that myself, then i'll know!" Gimme a minute...

2025-07-01 11:53:23

*_jߍyrope

jrp@hub.kliklak.net

@Herr Irrtum! Danke, Du hast's probiert :) Ich habe einfach Bandcamp versus Mirlo laufen lassen, und Mirlo klingt schlechter, kommt generell mit hoch verdichtetem bzw. verzerrtem Material nicht gut klar. Das ist natürlich ganz doof für experimentelle Musik, wo's halt öfter an die Grenzen der Akustik gehen kann.
Mirlo sagen dazu: "we convert to HLS using 320kbs libmp3lame using ffmpeg. We keep the original file uploaded to use when someone requests an album to download, and then convert that file to whatever the requested format is (again using ffmpeg)."
HLS ist adaptive bitrate streaming, da könnte das Übel liegen.