@brianfauteux@spectraimsim also not true, it's a bad look, sure, and should be avoided. but it's only a true regulatory issue if consideration (money) was paid in conjunction with an ad. just airing an ad from another source without consideration isn't actionable from an fcc enforcement pov
@beeritual_jazz@VivianKesandre not true. fcc doesn't care about content except for very rare circumstances like sponsorship acknowledgements/underwriting, certain political content, profanity/obscenity, drawings/giveaways, simulcasts/SLAs and, for TV, closed captioning accuracy and commercial loudness settings
@Dave_Abear@_expskin@VivianKesandre then that wouldn't be a "master" - you're saying they are transcoding from a non-master recording. idk if that's the case, but when I listen with studio monitors on lossless via USB audio interface or monitoring headphones I hear absolutely nothing objectionable
@Ancients@VivianKesandre broadcasters standardize on a format even when they *could* play out from a variety of formats, it's about consistency, quality, and the huge amount of money involved in accurate live playout of content, commercials, underwriting, etc. PE has also gutted dev work for bdcast tech
@Ancients@VivianKesandre I know you don't actually care to listen but you're talking live vs. production, server-based vs. workstation-based, shared multi-site, multi-city, multi-country hybrid-cloud libraries vs. local (or even indiv. cloud) workstation storage, metadata requirements.
@Ancients@VivianKesandre ok but that's not robust or 100% reliable for broadcast, and avoiding transcoding is not some "weird" broadcast thing, it's a critical broadcast quality measure. plus why risk your workstation or server having a single glitch with the "on the fly" part? could cost station $ 1000s
@Ancients@VivianKesandre yeah i got what you meant but it's more complex than just one-file-in-a-buffer at a time. you'd need a buffer manager layer that would buffer cued playlist files, nearly-instantly decode files as they're loaded for quick turnaround airplay, plus lots of downstream transcode
@Ancients@VivianKesandre you don't want to be managing buffers on a frequency and frame-based codec. pro broadcast systems utilize a great deal of metadata-based ramps/fades, cuts, xfades, etc that will just not work well with MP3s directly. if it was Rivendell, they've always had PCM as an option anyway
@VivianKesandre@pcdkd@testpatterns_ avoiding transcoding and/or perceptual audio codecs whenever possible is, in fact, objectively best practice for any broadcast
@Ancients@VivianKesandre a professional system, open source or not, may/should be able to import/trancode MP3s but will not use them for playout because it is a less robust format for transmission than layer II audio - which is a time-domain compression scheme rather than frequency-domain as with MP3s
@luuk58@VivianKesandre I'd never use MP2 or 3 at less than 256kbps, so i disagree on that front, and overall it's better for broadcast and for re-encoding because it's a less complex, but more robust format for transmission, editing, introduces fewer perceptual artifacts imo, etc.
@VivianKesandre Perceptual audio codecs are inherently lossy, so you're transcoding from another lossy format via streaming. This is already bad, but then you hit a wideband AGC and multiband compressor/limiter, which processes audio levels for consistent audio, and reveals a great deal of GIGO
@VivianKesandre Lots of overconfident wrongness here from someone who doesn't know how audio gets on air.
It goes from the studio to the transmitter, and it's not always AES, balanced analog, linear PCM or even less-lossy MPEG 1 Layer-II audio, it's often MP3 (Barix or similar), AAC/+, APT-X