Tired of this shit
Americans don't use it because they get a scary letter from their ISP
Thirdies will only have trash quality encodes uploading at dial up speeds
>cd >not "high res" >bar graph (real commerical DACs don't produce such a signal) >analog >no noise whatsoever
Maybe educate yourself about basic concepts before falling for audiophile lies.
Also maybe hard to swallow pill: The music you listen to doesn't sound bad because your medium/download doesn't have enough bits or whatever, but because the music actually IS bad and comes from a shitty loudness war master to begin with.
No amount of oversampling and wasting space will fix that.
graphs like this are so stupid, if gives off entirely the wrong conclusions, that is that analog is always more complete, and also that you can only /approach/ analog by increasing the resolution of digital sampling
also that's not even the correct way of drawing a sample-and-hold graph, for one they aren't discrete bars, and for two, the samples should be on the analog line, not below it
of course, high res audio is a complete meme for listening purposes anyway (as opposed to mixing/editing purposes, where it can be useful), so it's not surprising they use such misleading graphics to sell it
Great video!
I stand by my sperglord assessment, the "stairstep" in the OP is not continuous, it perfectly illustrates discrete sampling and has none of the issues of a stairstep. >inb4 lines are wider than one pixel
they show it like this because a screenshot of 3 identical waveforms just doesn't have the same selling power
i understand illustrations don't have to be perfectly accurate to be useful, but that's just completely misleading, so many people believe digital audio comes out of speakers with "stairsteps" because of how prevalent these misleading pictures are, so yes, it's important to point it out
I legitimately cannot wrap my head around sound. It's always drawn as a single curve like this, and on analog media such as vinyl records the movement of the needle is somehow sufficient to generate fully textured music.
But how can that be right? In a recording of an orchestra, there can be well over a dozen different instruments, multiple players of most of those instruments, and perhaps singers as well. Even if they're all holding a single note, there are probably resonant frequencies from various instruments, and then all of the texture that comes from the timbre and distinct characteristics and personality of each of those instruments.
So wouldn't the curve really be, if one zoomed in on it, composed of a highly variable set of dozens of curves layered upon one another? How can all of that depth be reliably and (mostly) accurately transcribed as grooves on a piece of vinyl? I feel as if someone told me they invented an algorithm that consistently and losslessly compresses any given already-compressed 100 MB video file down to 1 MB. Surely there's just no way?
any complex wave can be viewed instead as a sum of infinitely many sine waves, you can perfectly convert between the two
humans can only hear up to around 20kHz absolute peak, so you don't need an infinite set to capture audible sound
any instrument can emit any combination of frequencies (sines) and amplitudes thereof, which all combine together through constructive or destructive interference, and the end result can be conveyed as a single waveform, or preferably at least two, since we have two ears (stereoscopic hearing is more complex than sound imo)
in a vinyl record, this waveform is equalised with a standard curve for practical reasons, then cut directly into the plastic
in a digital system, the waveform is sampled at or above nyquist (double the frequency of the highest source frequency you wish to capture), this has been mathematically proven to perfectly capture all frequency information, amplitude is quantised into a fixed number (the bits), which defines the captured noise floor
it all may seem oversimplified or reductionist, but it's true, the only reason a live performance may sound different is because of /how/ you hear it, not /what/ you hear, that is, using your own ears and moving around causes you to pick up slightly different waveforms than what the microphone(s) would, since they aren't in the same positions and shapes as your head would be
while a single combined waveform is simple to conceptualise, capture, and reproduce, an infinitely many number of them bouncing around a 3D space isn't so much, that's where the complexity lies
a sound recording is more like a photograph in that sense, it's a capture of that specific moment and angle, and while it may convey what you can see if you stood right where the camera was, it can't convey what it's like to be there and move around
Thank you, this was very informative. I like the photo analogy. I didn't understand this part, though: >equalised with a standard curve
Does this mean for volume purposes across certain bands of frequency ranges, e.g. treble/bass? Or something else? Because I still have trouble seeing how the physical movement of a needle on vinyl can be precise enough to recreate slight variations in a combined waveform on the order of up to 20k oscillations in a single second.
Fourier transform? I remember doing something related to this summation of sine waves to approximate other periodic functions back in high school.
Torrents or usenet, but pirating music just isn't worth the hassle anymore.
Torrents will eventually lead you to the sheer maximum autismo that is private trackers. They gamify the tracker "ladder" and its rungs consist of a pointless grind.
Usenet will probably cost as much as a basic Spotify subscription.
Even if you have a good source of content, you still have to deal with download automation (Lidarr).
tl;dr download music if you have no life
It really has become a dubious proposition. If I wasn't already invested in it, the pros-and-cons aren't as clear as they used to be. As you say, audio streaming saves a ton of time (way more than the equivalent amount of money you'd fork over for it), and makes it easy to share with friends and use multiple devices. There are benefits to a private collection, obviously, but given the limitations it has as well, I don't think I'd recommend it to anyone new to the game.
Damn near everything else should be pirated, though.
2 months ago
Anonymous
>Does this mean for volume purposes across certain bands of frequency ranges, e.g. treble/bass?
yes, i'm referring to the "RIAA curve"
because low frequencies produce larger waves than higher frequencies, this in turn also produces wider grooves for bass and smaller grooves for treble in a record, so the riaa curve reduces the power of bass frequencies, and increases the power of treble frequencies before cutting the discs, to try to even them out a bit in terms of cutting amplitude, then in players the inverse is applied to restore the original audio, this is why you need a phono preamp specifically before you can amplify the sound from a turntable, if you tried to amplify it directly it will be extremely tinny >Fourier transform
yep
Has anyone else been getting very low quality audio through NewPipe lately? Specifically over mobile data.
I have all my settings to reasonable values, 720p or over, but it's still very bad and it used to be much better a month ago with the same settings and same NewPipe version.
Maybe your phone's data limit defaults/rationing settings changed? I.e. it's not NewPipe, but your actual network settings. At least your NewPipe works. Mine has to sit for several minutes straight before it can first connect to YouTube, every time I open the app.
for the quickest solution, https://doubledouble.top. supports a decent amount of services, but it goes down often.
if you have your own ARLs/accounts, from somewhere like https://rentry.org/firehawk52, then just use orpheusdl and modules for whatever platform you wanna grab from.
if you want more niche stuff, vinyls, cd rips, others, do soulseek with https://nicotine-plus.org/ or https://rutracker.org/ if you find something there that isn't anywhere else and want to use torrents
if you want to get into the rabbit hole of private torrent trackers, an interview for https://redacted.ch is pretty easy if you read https://interviewfor.red
>graph
stairstepping does not exist in any competent DAC, nor is it possible to hear any frequency above CD quality
why can't people into nyquist-shannon
Torrents or usenet, but pirating music just isn't worth the hassle anymore.
Torrents will eventually lead you to the sheer maximum autismo that is private trackers. They gamify the tracker "ladder" and its rungs consist of a pointless grind.
Usenet will probably cost as much as a basic Spotify subscription.
Even if you have a good source of content, you still have to deal with download automation (Lidarr).
tl;dr download music if you have no life
>pirating music just isn't worth the hassle anymore
Maybe not for everything, but it's still nice to have a personal copy of any favourites. With licensing bullshit the way it is, no guarantees it will always be easy to find a given song legitimately and in decent quality.
https://nicotine-plus.org/
Tired of this shit
Americans don't use it because they get a scary letter from their ISP
Thirdies will only have trash quality encodes uploading at dial up speeds
rutracker
>cd
>not "high res"
>bar graph (real commerical DACs don't produce such a signal)
>analog
>no noise whatsoever
Maybe educate yourself about basic concepts before falling for audiophile lies.
Also maybe hard to swallow pill: The music you listen to doesn't sound bad because your medium/download doesn't have enough bits or whatever, but because the music actually IS bad and comes from a shitty loudness war master to begin with.
No amount of oversampling and wasting space will fix that.
That goes for most modern shit music. If you listen to classical music, the recording quality is often far better, at least for some labels.
it is obviously a simplified illustration that still captures the basic gist of it anon
>he thinks 16 bit 44.1Khz is "high res"
Let me guess, you need more?
you literally and physically can't hear the difference on anything higher
graphs like this are so stupid, if gives off entirely the wrong conclusions, that is that analog is always more complete, and also that you can only /approach/ analog by increasing the resolution of digital sampling
also that's not even the correct way of drawing a sample-and-hold graph, for one they aren't discrete bars, and for two, the samples should be on the analog line, not below it
of course, high res audio is a complete meme for listening purposes anyway (as opposed to mixing/editing purposes, where it can be useful), so it's not surprising they use such misleading graphics to sell it
Holy sperglord
It's called illustration, moron. As in, it illustrates a concept for the audoence as opposed to being an oscilloscope screenshot.
And it illustrates the concept wrong.
I recommend watching this:
Heh, this video kills the Hi-Res tard. I love this one.
Great video!
I stand by my sperglord assessment, the "stairstep" in the OP is not continuous, it perfectly illustrates discrete sampling and has none of the issues of a stairstep.
>inb4 lines are wider than one pixel
they show it like this because a screenshot of 3 identical waveforms just doesn't have the same selling power
i understand illustrations don't have to be perfectly accurate to be useful, but that's just completely misleading, so many people believe digital audio comes out of speakers with "stairsteps" because of how prevalent these misleading pictures are, so yes, it's important to point it out
I legitimately cannot wrap my head around sound. It's always drawn as a single curve like this, and on analog media such as vinyl records the movement of the needle is somehow sufficient to generate fully textured music.
But how can that be right? In a recording of an orchestra, there can be well over a dozen different instruments, multiple players of most of those instruments, and perhaps singers as well. Even if they're all holding a single note, there are probably resonant frequencies from various instruments, and then all of the texture that comes from the timbre and distinct characteristics and personality of each of those instruments.
So wouldn't the curve really be, if one zoomed in on it, composed of a highly variable set of dozens of curves layered upon one another? How can all of that depth be reliably and (mostly) accurately transcribed as grooves on a piece of vinyl? I feel as if someone told me they invented an algorithm that consistently and losslessly compresses any given already-compressed 100 MB video file down to 1 MB. Surely there's just no way?
any complex wave can be viewed instead as a sum of infinitely many sine waves, you can perfectly convert between the two
humans can only hear up to around 20kHz absolute peak, so you don't need an infinite set to capture audible sound
any instrument can emit any combination of frequencies (sines) and amplitudes thereof, which all combine together through constructive or destructive interference, and the end result can be conveyed as a single waveform, or preferably at least two, since we have two ears (stereoscopic hearing is more complex than sound imo)
in a vinyl record, this waveform is equalised with a standard curve for practical reasons, then cut directly into the plastic
in a digital system, the waveform is sampled at or above nyquist (double the frequency of the highest source frequency you wish to capture), this has been mathematically proven to perfectly capture all frequency information, amplitude is quantised into a fixed number (the bits), which defines the captured noise floor
it all may seem oversimplified or reductionist, but it's true, the only reason a live performance may sound different is because of /how/ you hear it, not /what/ you hear, that is, using your own ears and moving around causes you to pick up slightly different waveforms than what the microphone(s) would, since they aren't in the same positions and shapes as your head would be
while a single combined waveform is simple to conceptualise, capture, and reproduce, an infinitely many number of them bouncing around a 3D space isn't so much, that's where the complexity lies
a sound recording is more like a photograph in that sense, it's a capture of that specific moment and angle, and while it may convey what you can see if you stood right where the camera was, it can't convey what it's like to be there and move around
Thank you, this was very informative. I like the photo analogy. I didn't understand this part, though:
>equalised with a standard curve
Does this mean for volume purposes across certain bands of frequency ranges, e.g. treble/bass? Or something else? Because I still have trouble seeing how the physical movement of a needle on vinyl can be precise enough to recreate slight variations in a combined waveform on the order of up to 20k oscillations in a single second.
Fourier transform? I remember doing something related to this summation of sine waves to approximate other periodic functions back in high school.
It really has become a dubious proposition. If I wasn't already invested in it, the pros-and-cons aren't as clear as they used to be. As you say, audio streaming saves a ton of time (way more than the equivalent amount of money you'd fork over for it), and makes it easy to share with friends and use multiple devices. There are benefits to a private collection, obviously, but given the limitations it has as well, I don't think I'd recommend it to anyone new to the game.
Damn near everything else should be pirated, though.
>Does this mean for volume purposes across certain bands of frequency ranges, e.g. treble/bass?
yes, i'm referring to the "RIAA curve"
because low frequencies produce larger waves than higher frequencies, this in turn also produces wider grooves for bass and smaller grooves for treble in a record, so the riaa curve reduces the power of bass frequencies, and increases the power of treble frequencies before cutting the discs, to try to even them out a bit in terms of cutting amplitude, then in players the inverse is applied to restore the original audio, this is why you need a phono preamp specifically before you can amplify the sound from a turntable, if you tried to amplify it directly it will be extremely tinny
>Fourier transform
yep
This makes me miss being in school, made lots of plots like that on my hp 50g in DiffEq 2. Such fun.
https://rentry.org/firehawk52/
thanks anon, i love you so much
>Spotify
Nah just fricking around, umm archive.org
i had no trouble downloading music from youtube with ytdlp except thatvit doesnt download metadsta like artist, album, genre and other stuff like that
>listen to compressed shit on youtube music
>on my iphone
>with airpods pro
good enough
Has anyone else been getting very low quality audio through NewPipe lately? Specifically over mobile data.
I have all my settings to reasonable values, 720p or over, but it's still very bad and it used to be much better a month ago with the same settings and same NewPipe version.
Maybe your phone's data limit defaults/rationing settings changed? I.e. it's not NewPipe, but your actual network settings. At least your NewPipe works. Mine has to sit for several minutes straight before it can first connect to YouTube, every time I open the app.
https://doubledouble.top
Omg, please Chinkhawk, stfu annoying af
If you give Deezer $10, they let you download FLACs of everything they have. I just found it easier bc couldn't find songs torrenting
for the quickest solution, https://doubledouble.top. supports a decent amount of services, but it goes down often.
if you have your own ARLs/accounts, from somewhere like https://rentry.org/firehawk52, then just use orpheusdl and modules for whatever platform you wanna grab from.
if you want more niche stuff, vinyls, cd rips, others, do soulseek with https://nicotine-plus.org/ or https://rutracker.org/ if you find something there that isn't anywhere else and want to use torrents
if you want to get into the rabbit hole of private torrent trackers, an interview for https://redacted.ch is pretty easy if you read https://interviewfor.red
Nyaa.si for japanese / asian stuff
Rutracker.org for everything else
Yes, you will not find *everything*, for that you'll have to start looking at private trackers.
A name to go off of is jpopsuki, do your own research for more. I can't tell you how to get into them because I've never bothered trying myself.
>graph
stairstepping does not exist in any competent DAC, nor is it possible to hear any frequency above CD quality
why can't people into nyquist-shannon
youtube-dl
Unironically https://freemp3cloud.com/
Still works
Torrents or usenet, but pirating music just isn't worth the hassle anymore.
Torrents will eventually lead you to the sheer maximum autismo that is private trackers. They gamify the tracker "ladder" and its rungs consist of a pointless grind.
Usenet will probably cost as much as a basic Spotify subscription.
Even if you have a good source of content, you still have to deal with download automation (Lidarr).
tl;dr download music if you have no life
>pirating music just isn't worth the hassle anymore
Maybe not for everything, but it's still nice to have a personal copy of any favourites. With licensing bullshit the way it is, no guarantees it will always be easy to find a given song legitimately and in decent quality.
Rutracker usually has everything. DSD/TR24, literally everything.