Okay /g/, I made this picture last night comparing different codecs(FLAC, OPUS, MP3, and AAC).
For the first test, I encoded all the lossy coded in 320
I chose a song with with an original flac file with a sampling rate of 48000khz to test the encoding of these different formats.
As the picture shows, AAC is the superior format for high bitrate. Out of all the lossy encodes OPUS was the largest in filesize, but only by 10kb
but all three lossy encodes seem to not differ from flac
next will be low bit rate test
>>51770913
Fair enough but no human can hear beyond 20 kHz.
And now the low bitrate test
somehow my Opus encoder kept on using a sampling rate of 48000khz and there is no way for me to change this, but it shouldn't hurt anything.
You can clearly see that Opus is much better than AAC at lower bitrates. I didn't use MP3 in this test because its just pointless
What exactly is your point? what do you want to convey to us? what format do you want us to use? which format do you use? and what is it that itches you the most about other formats that finds them unacceptable?
>>51770913
You should consider using a logarithmic scale for the frequency; it's a much better match for how human hearing works.
Right now almost half of your graph is made of stuff that makes up like 5% of the experienced sound.
>>51770956
I know, i just wanted to test how far these encoders could go. Sampling rates above 44100khz is pointless
>linear scale on frequency axis
>>51770973
>I didn't use MP3 in this test because its just pointless
How about no?
You're doing an analysis of audio formats and you're excluding the single most common format of all time, are you retarded?
OP is a certified idiot.
First off, why not start with a song that in lossless, contains audio data above the 22kHz ceiling? The entire point of lossless is that it "sounds the best".
C'mon OP, you're sooooo close.
>>51770913
You don't hear with your eyes, anon. It just doesn't work like that.
ABC-HR or ABX, or GTFO. I bet you fail ABX with Opus at 160.
>>51771081
I have a machine girl album in 96000hz, even then its still limited by the sampling rate
opus 128kbps best lossless
>>51771240
>96 kHz
You can't heard beyond 20 kHz anyway.
I'm practically deaf because I always have my earphones/headphones on max.
I like 320 CBR for Electronic music & V0 for everything else. I have a few albums here and there in flac but I can't notice the difference between them and other mp3s I have because I don't own a high end sound system.
P.S: LAME master race.
>>51771282
Get a AMP.
These tests are all pointless since none of them prove audible transparency. Also most adults cannot hear above 16 Khz let alone 18 Khz.
Still Opus has better compression efficiency than AAC and can be used in bitrates lower than 32kbps for things like streaming or voip with very low latency. Therefore it's still the best Audio codec out there now that support is growing and can be played on android and apple devices.
>>51771263
>implying higher sampling rate doesn't affect the form of lower frequencies
>>51771314
That will make him go deaf faster you twat.
>>51771350
It doesn't.
Reproducible proof:
https://www.xiph.org/video/vid2.shtml
>>51771263
Sampling theorem and multiple channels, though you're correct that 96k is pointless.
>>51771350
Found the retard.
>>51771367
>Sampling theorem and multiple channels
>although
So what's wrong with the statement?
96 kHz sampling rate yields 48 kHz frequencies, far beyond what a human can hear, 20 kHz.
>>51771363
I can just look at my frequency analyzer on foobar and can tell
>>51771397
Must be a real quality FFT there.
>>51771383
Pointing out that frequencies up to 42-48k do actually make sense, as people may be misled by the "can't hear past 20k" statement everyone is making. People may not realize the difference between playback and encode rates being discussed.
>mfw I spent $2000 on a 32 bit 96 kHz DAC and monster cables
>mfw you're all plebs
>>51771440
>Listens to it on headphone
>>51771440
At least our intelligence is still open for debate.
>>51771440
4/10 bait, you need to be more subtle.
>>51771440
>32 bit 96 kHz DAC
Bruz we at 64 bit 5,644,800 Hz now.
I have not seen any fucking hardware supporting Opus so it's moot anyhow. AAC, MP3 or FLAC if you are lucky.
Besides. Anything above 192kbps MP3 you can't fucking tell the difference unless you are 10 years old with 'golden 'ears' listening on a $10,000 hifi system.
Anyone above the age of 18 is a bald faced liar.
>>51771519
Nobody gives a shit about opus hardware decoding anymore. Software decoding already exists and uses extremely little cpu resources.
>>51771560
>like I am gonna convert my 100k songs to some shitty unsupported software algorithm.
>>51771578
mx player can play opus files just fine.
>>51771578
It literally shits all over AAC though.
>>51770913
AAC is commended for being nearly transparent in terms of compression, but it's still lossy and built on an non-free encoder. MP3 320 has a high cut-off at about 19.5-20kHz but preserves more data up until that point. Give V0 a try - it has more dynamic range.
>>51770995
Trolling is against the rules outside of >>>/b/. Take it there.
>>51771578
>100K songs
Bet you never listen to 90% of those songs ever.
Why the fuck would you NOT rip .flac?
>>51771658
Not all of us have autism like you.
>>51770956
lol
>>51771263
lolol
>>51771678
>>51771692
Rosen, Stuart (2011). Signals and Systems for Speech and Hearing
Rossing, Thomas (2007). Springer Handbook of Acoustics. Springer. pp. 747, 748. ISBN 978-0387304465.
The Scientist and Engineer's Guide to
Digital Signal Processing
http://www.dspguide.com/ch22/1.htm
>>51771678
reeeeeeeeeeeeetaaaaaaaaaaarrddddddddddddddddddddddddddddddddd
>>51771678
A few kids MIGHT be able to but once you become a teen you lose the ability to hear above 18 Khz. Then once you hit 30 you are guaranteed to be deaf above 16 Khz due to aging.
This is all irrelevant to music anyway as most musical tones are spread throughout 100-10,000 Hz.
>>51771712
then why am i downloading 24/96 vinyl rips???
>>51771733
Simple, you have Autism.
>>51771670
Autism is irrelevant. FLAC has it's uses. Lossy encodes can't be reencoded for example.
>>51770913
Are you retarded or what? The point of these codecs is not to preserve the inaudible range of audio but to keep the most of the rest with small size. Also how did you convert opus to 320 if its only VBR? The common problem of codecs in terms of quality is pre-echo, you should check that on high bitrate instead.
>>51771733
So you can convert them to a more suitable format without losing any of that vinyl quality in the process?
>>51771755
You would only lose inaudible quality anyway.
24 bit is even sillier. Vinyl records are physically unable to reach even 16 bits of SNR.
>>51770956
>2011 still clings to a human-centric worldview
I seriously hope you guys don't do this
>>51771755
If he'd convert it to 44.2 kHz then he would probably lose some because 92000 and 44100 are not dividable, so it would interpolate which means losing more quality then he should. Reripping with 44.2kHz and converting would be the real solution
>>51771750
FLAC is completely useless. You encode to lossy once and call it a day. Once you go past 128 VBR Opus, music becomes audibly transparent.
For people who have MP3 rips, they don't need to move to Opus since their 320k MP3 rips sound great and still don't take as much space as FLAC.
FLAC is a dead format except for autists. Even music studios don't want to touch that shit. They use uncompressed WAV instead.
>>51771794
*44.1 kHz
>>51771733
Because you're an idiot.
>>51771806
So why dont you use the old and huge and unsupported windows wma s instead of mp3s?
>>51771806
And what if a certain codec like 320 mp3 stop being supported by mayor manufacturers and you have a lot of files encoded on it? If I had to download my whole collection again it would take years.
>>51771806
>You encode to lossy once and call it a day.
Maybe you do, but I've got a lot of albums that I was able to switch from mp3 to Vorbis only because I had lossless sources for them. Shrunk my lossy library by over 20 GB that way while actually adding more albums in the process.
>>51771835
Because I never used it. In fact nobody with common sense did. WMA was a piece of shit. Everyone used MP3 even in the 90's.
All my music is 320k MP3 rips and I've only recently started to convert any new CDs I get with 192 VBR Opus. I never reencode though. As for the CDs, I either sell them on ebay or give them away.
>>51771794
>>51771817
I don't think any decent resampler would produce relevant artifacts, at least in the audible frequency range.
This is probably one of those cases where it pays to have just a bit more sampling rate than what your hearing needs.
>>51771350
>i don't know anything about sampling but i'll comment anyway: the post
>>51771846
Not gonna happen.
>>51771914
>I can predict the future
>>51771891
You missed the point. What if mp3 dies? Streaming services and games already moved on to either AAC or Opus or Vorbis. Do you thing mp3 will live forever?
>>51771891
So you're quite literally projecting your habits onto everyone else and declaring FLAC to be useless because of that.
>>51771806
>Even music studios don't want to touch that shit. They use uncompressed WAV instead.
Because you can't record directly to FLAC and you have to uncompress it to edit it, so there's no point in using it to simply save a bit of space during the recording and mixing process. Studios by and large do not distribute their recordings themselves, which is the only point where FLAC would come into play.
>>51771733
>vinyl rips
i couldn't tell you. those are pointless to have in anything above 16bit and 44.1kHz ever.
>>51771902
I know, i was just pointing out that technically it would.
MP3 does not support multi dimensional tags.
You can't have multiple ARTIST, TITLE or GENRE tags inside an MP3.
>>51771733
>24/96
Probably useless unless you edit your tracks.
>Vinyl rips
For the better mastering.
Storage is so cheap nowdays you may as well get all your shit in 24/96 FLAC just in case you ever want to mess around with it.
>>51771866
You didn't need to though. 320k MP3 is just fine. With 256GB microsd cards coming soon it won't matter if you use 320k MP3 vs 192 VBR Opus. I encode my new CDs to Opus because I want to support a free format, I could have easily stuck with 320k MP3 if I wanted to.
>>51771953
Even if you edit the tracks, vinyl does not even reach 16 bits of SNR. It's entirely pointless in all cases.
>>51771962
Why not? Is it a limitation of the medium?
>>51771932
Games have never used mp3(at least console as far as i know) they used their own format for stuff like sound effects. And im not talking about Midi
>>51771955
Soon is not now, and certainly not some months ago when I wanted to be able to fit as much of my library as possible onto a 128 GB microsd card.
>>51771972
Yes. Tape doesn't even quite reach 16 bit equivalent unless you use noise reduction from what I recall.
>>51771955
>FLAC is just fine. With 256GB microsd cards coming soon it won't matter if you use FLAC vs 320k MP3. I encode my new CDs to FLAC because I want to support a free format, I could have easily stuck with 320k MP3 if I wanted to.
>>51771973
>Half-life and mods
>Half-life 2 and mods
>>51771972
Not him but yes. Vinyl barely reaches 12bits of dynamic range in the best cases.
>>51771972
Yes
Even if the record was made of frozen hydrogen, the highest possible SNR would be 19 bits.
4-bit precision gives you about 16.77 million values. Assuming a total groove width of 50 x 10^-6m, the maximum movement of the cutter is physically bounded at about half that. Much more and the cutter will be in the space for an adjacent groove. Thus, 50 microns width divided by 16.77 million gives us about 3 x 10^-12m, i.e. ~0.03 angstroms.
The diameter of a hydrogen atom is 1.0 angstroms (1 x 10^-10m). That would make the resolution of a 24-bit digital signal equivalent to an analog cutter whose resolution is just about 1/30 the width of a hydrogen atom. Sadly, this seems to be physically impossible, as none of the particles smaller than atoms are stable enough to be used in records.
Of course, records aren't made of hydrogen, they're made of the polymer pvc. One molecule of pvc is about 100,000 angstroms. This means that, if the cutters were actually removing single pvc molecules the vinyl records would have about 11 bits of resolution. Sadly, they don't get even that precise, though I'm not sure the actual precision. To get down to a record made of hydrogen atoms (possible under very low temp/very high pressure I suppose) one would need 19 bits. Anything beyond that is useless as long as the laws of physics hold.
PVC molecules however are 100000 times larger than hydrogen.
The sampling rate can be estimated with a similar thought experiment:
A 33rpm record spins at 0.503 m/s at the outside of the record, then gradually slows to a meager 0.196 m/s at the center.
A PVC molecule is about 100000 ångström.
Even if the pressing AND the needle were accurate to single PVC molecules, which they aren't, the sample rate would only be
(0.503m)/(100000 ångström) = 50300 Hz
at the beginning of the vinyl album and
(0.196m)/(100000 ångström) = 19600 Hz
at the end of the vinyl album.
>>51772035
>4-bit precision
Typo; meant 24 bits
>>51772035
>His records aren't made out of degenerate neutron matter
>>51772035
Cool, thanks.
>>51771806
>FLAC is a dead format except for autists.
Even normie-tunes offer FLAC downloads what are you talking about stupid bitch?
>>51771920
>>51771932
Software media players will never lose support for MP3 (what I give a shit about). Even when some other codec replaces Opus as the kind audio codec. MP3 is eternal. It's been around since the 90's and it's not going anywhere.
I know Opus is still better than MP3 though which is why any new CDs I lay my filthy mitts on gets ripped to 192 VBR Opus. Eventually over half my library will be Opus some day.
>>51772073
It's called trolling, friend.
>>51771993
>Autism is just fine. With 256GB microsd cards coming soon it won't matter if you use Autism vs 320k MP3. I encode my new CDs to Autism because I want to support Autism, I could have easily stuck with 320k MP3 if I wanted to.
ftfy
>>51772081
Stop predicting the future, many formats have died before and many still will, mp3 might be one of them. There are downsides to mp3 but non to FLAC and you and your pleb mp3 collection will have to deal with it. Fag.
>>51772116
>If I don't approve of it it's autism.
Yeah, ok.
>>51772073
Normies download music from itunes/google play store. Guess what? MP3 and AAC is used in those places. Like I said, nobody except for a few autists use FLAC.
>>51772118
MP3 can be as transparent as flac at 320
it isnt going anywhere for a LLLOOONNNGG time
>>51772147
>Guess what? MP3 and AAC is used in those places
You can download FLAC from itunes as well dumbass.
>>51772118
In 2005 they said MP3 will be kill, in 2010 they said MP3 will be kill, in 2015 they said MP3 will be kill.
They will also say the same thing in 2020 and 2025 and so on.
Most of my music will be Opus by then though.
>>51772151
>it isnt going anywhere for a LLLOOONNNGG time
A long time in this days may be only a few years and then you'll have a collection of unsupported files worth shit. Good luck with that.
>>51772162
But normies don't' you dipshit, especially applefags with 8-16GB of extremely limited storage with no way to expand storage space without some shitty adapter.
>>51772172
>Muh predicting the future!
Fuck off.
>>51772196
If normies didn't then they wouldn't be offering it dumbass. But they know that normies do download FLAC because everyone knows it's the superior format.
>>51772126
That wasn't me but even when phones accept 2TB micro sd chips, FLAC will still be nothing more than a meme. The problem isn't storage space, the problem is FLAC takes up unnecessary space. Most people will never use FLAC for the same reason nobody will ever store lossless copies of movies/tv shows - they all waste unnecessary space.
>>51772198
Not him but even if MP3 was no longer supported in 2030 there is nothing stopping him from using legacy software for MP3 playback.
>>51772249
You can transparently re encode flac when you put the music in your phone. And it's not unnecessary FLAC has enough information to make re encoding possible without a massive loss in quality, something that lossy formats like mp3 don't.
![VistaWMAIcon[1].png](https://i.imgur.com/UwSzrmb.png "VistaWMAIcon[1].png")
wma
Archival Tier: FLAC
Everyday Listening Tier: AAC
Embarassing Tier: MP3
>>51772249
>nobody will ever store lossless copies of movies/tv shows
These are literally never distributed so nobody does even now.
I download FLAC as I see it as a hassle free format with freedom to do whatever I want with them. I do transcodes for my mobile devices, sometimes I might want to put a song into a video which will then(usually) have to be transcoded to a format allowed by the container if I want to distribute it somewhere. I'm usually safe from poor transcodes by having most of my music in lossless formats and whatever formats we get in the future, I can always convert my music to those if I want to. Storage space is an issue in some cases but storing the songs on my HDD it's really nothing for me to worry about. I'd rather have the lossless original and minimize the possible impact on fidelity a poor lossy encode might do and get the freedom lossless file gives me.
>>51772212
They let you download FLAC just in case you have autism. Apple barely has 20% mobile market share right now and will do everything they can to not disappear like Symbian, even if it means catering to autists. Normies by large (including android and apple users) do not download FLACS. Apple users are a tiny minority of normies.
>>51772268
He would still be sitting on a collection of files encoded in a dead format without any choice about it. Meanwhile FLAC collections will still be able to be re encoded at will to whatever new better format there is, to eternity.
He himself admitted that he was migrating or would migrate to Opus, I don't need to because I have a FLAC collection.
>>51772322
Whatever the percentage is it's still big enough to make offering the format worth it. That means that normies do use FLAC even if not the majority (everyone knows that the majority are but plebs anyway).
Apple music and car subs.
Fuck all of you, when I finish the math on a quad-10 ported folded horn box, bitches will cum when the drop hits.
>>51772283
Which is all completely pointless. Encode to lossy once and call it a day, no need to become autistic about encoding music.
I do agree that re-encoding lossy to lossy audio is stupid though. This is why I still have 320k MP3 rips.
>>51772360
Sorry but some of us like fidelity.
>>51772320
You could have skipped all that autism by encoding to lossy once. MKVs can easily remux MP3 /Opus audio streams.
>>51772378
>completely pointless
For retards sitting on an inferior mp3 collection, yes. Meanwhile I will never have to re-download or re-rip my collection because it's encoded and compressed on a superior format.
>>51772417
>encoding to lossy once
Says the pleb with files encoded in a dying format. Top kek.
>>51772429
Neither will I, only difference is my hard drives and micro sd chips will be less bloated than yours. That and I won't be labeled an autist.
>>51772417
mkv can do literally anything. not every place accepts those and yt for example just re-encodes everything you upload to webm with vp9+opus now(+bunch of other formats for compatibility). it's no autism. if i don't personally have any issue with storing lossless files on my computer, i don't see a single reason why i would use anything else. if storage space space isn't a concern, flac is a better format even if the cases where this might actually benefit are rare.
>>51772444
>dying format
nice trips but [citation needed]
>>51771263
>>51771002
You guys realise that you have to sample at double the rate of your highest frequency (or more), if you sample a 1hz sine at 1hz sample rate, you'll have 1 point of sampling for a complete wavelength and while it won't work at all, it will create lower frequency artefacts that will fuck everything up. On the other hand, sampling at 2x or 2,5x (or more) allows for a perfect reproduction of the sine by a good DAC after filtering.
That's why there are brickwall filters in ADCs and Smoothing filters in DACs. But if you filter the sound above a certain frequency it will fuck the phase up and the sound around this very frequency.
44K1 is an old standard that works okay but it can damage the high, but as a sound engineer I always prefer 48K, it allows for a slightly bigger filtering headroom and 24kHz is plenty, (most people will feel stuff above 20K but can't hear a damn thing above 19k, this is my max I believe).
While recording on the other end, I tend to go with 88K2, so I can really play around with the sound. I often Master at 88K2 too so it doesn't have to do weird interpolations when converting to 44K1 (Altho I never heard any audible artefact, it my be my OCD)
>>51772464
If you upload to YouTube regularly then you are using very high bitrate video with uncompressed WAV audio. FLAC will only slow you down as encoding/decoding it takes unnecessary time.
>>51772324
But he gets to keep copies of music to listen to for life. You on the other hand keep FLACs that you must re-encode to new lossy formats for life. That seems pretty autistic to be to be quote honest.
>>51772571
By for life you mean while you can find something that can still play a by-then dead format? Sure..
>>51772249
>nobody will ever store lossless copies of movies/tv shows
Remuxes are a thing.
>>51772622
He can still use legacy media players though. Assuming support for MP3 is ever dropped.
>>51772397
Never said I don't either, just that this thread is so full of autism it hurts.
First off, debating audio format is moot. MP3 is here to stay as the normie format, AAC for Apple, and FLAC for all else. Even then, half of these noobs here are using gear that already colors the sound so any gains with format are vaporized.
I literally laugh at people who have DAC's on M50's on the go. Only an autistic person could be bothered to go through all that trouble to listen to tunes on the go, when the alternatives are so good.
>>51772657
So is autism.
>>51772485
Holy shit the one guy in the thread who actually knows what he's talking about thank God I lurked for this long, I thought everyone was retarded.
>Qobuz is about to get shut down
Good, fuck their lossy Masters
they should say that on the fucking release
music is for children
>>51772690
>hurr hurr autism
It's ok to be proven wrong sometimes.
>>51772732
You didn't prove anyone wrong, you literally are autism.
>>51771363
>>51771375
>>51771904
>what is aliasing?
>>51772775
>what is automatic oversampling?
I never get tired of the "I know more than experts" meme.
>>51772775
Not present on any decent gear.
>>51772657
Those are not lossless.
>>51772657
Yeah but they are completely pointless. Just give me a 16 CRF movie rip with opus audio god damnit.
>>51772775
>>51772485
Good to see someone knowing his shit.
Also yeah, higher than 48khz are useful for audio editing.
>ATH-M50/ATH-M50x users in the topic
>>51770913
AAC is non-free though
>>51773176
There is still zero reason to have source material in 96 kHz, not to mention ridiculous 24/96 vinyl rips.
>>51773316
What if you're compressing or extending the duration of a sound sample? Isn't the extra information of 96khz useful for more accurate calculations?
Also, scientific studies often require higher than human frequency sounds.
>>51770973
Opus is supposed to be transparent at around 170 kbps so you probably could lower the filesize even more.
And yes, opus can't into 44.1kHz but it is not an issue since 44.1 fits in 48.
>>51773316
refer to
>>51772485
and read this
http://recording.org/threads/oversampling-explained.48087/
basically if you record at 44.1kHz and an instrument produces a frequency higher than 22.05kHz you get distortion in the audible range
>>51772485
>I often Master at 88K2 too so it doesn't have to do weird interpolations when converting to 44K1 (Altho I never heard any audible artefact, it my be my OCD)
it's just your OCD, the PCM stream gets converted to a PDM-like stream before getting converted back to a PCM stream of lower frequency, so even if the end sampling frequency is half of the starting one the converter does the same amount of interpolation that it would do if the ratio wasn't so "nice"
>>51773647
It can not, thats not how these work. It will get interpolated because they are not divisible and lose quality. It probably cant be heard, but technically it still loses quality
>>51770913
>but all three lossy encodes seem to not differ from flac
get a better listening medium (speakers, headphones, converters...) or a better pair of ears
>>51771282
>because I don't own a high end sound system.
your sound card and a pair of $100 headphones will work as well
>LAME master race.
why even
try fraunhofer
>>51771440
>posted from my KRK rokits
>he bought a 32 bit DAC
>he actually fell for that meme
>mfw
also I'll explain because I get the feeling that you're stupid
32bit converters are useless because 32bit (integer) audio files don't exist, and if you were thinking about the "infinite headroom" of 32bit (floating) then that only works when the signal is digital - your converter will have a fixed amount of headroom because of its analog design
>>51771519
>Anything above 192kbps MP3 you can't fucking tell the difference
>people really believe this
>people with ears believe this
>people think it only has to do with frequency
>>51771794
>I don't know how sample rate converters work
>>51772147
>calls normies the people who download aac and mp3 audio
>calls autists the people who use flac
are you by any chance both?
>>51772162
these would be ALAC
tho really?
I didn't know, I might actually start using iTunes to buy music now
>>51772775
aliasing affects higher frequencies much more than lower ones
say you're sampling a 10k frequency 4 times per cycle, so at 40kHz
now, with 40kHz you can sample a 200Hz wave 200 times per cycle
do you think it will make any difference if that's 200 or 400 times per cycle?
you should stop replying, you're only making things worse
>>51772781
what are you on about?
you're even more stupid than him if you think sample rate converters do that
>>51770913
Why do you think AAC is the best? Because it holds on to the useless, inaudible frequencies above 20 kHz rather than discarding them like it should?
Also, how the fuck do you encode Opus at 320 kbps? It's VBR.
>>51773992
>aliasing affects higher frequencies much more than lower ones
>say you're sampling a 10k frequency 4 times per cycle, so at 40kHz
>now, with 40kHz you can sample a 200Hz wave 200 times per cycle
>do you think it will make any difference if that's 200 or 400 times per cycle?
That's not what aliasing in audio has to do with. It's interference from frequencies higher than the nyquist frequency of the sampling rate you're using.
>you're even more stupid than him if you think sample rate converters do that
I'm not sure what you mean by "sample rate converter", but from what I know most ADCs oversample to make it so their anti-aliasing filters don't have to have as hard of a rolloff.
>>51774093
>That's not what aliasing in audio has to do with.
according to wikipedia, aliasing is an effect that causes different signals to become indistinguishable (or aliases of one another) when sampled, so what he was referring about may have been aliasing, even if he still was pretty off
>I'm not sure what you mean by "sample rate converter"
a program that converts a piece of audio data from one sampling frequency to another
like weiss saracon
>but from what I know most ADCs oversample to make it so their anti-aliasing filters don't have to have as hard of a rolloff.
ADCs are usually sigma-delta
they first convert the audio into PDM, then that gets converted into PCM, which is the digital output stream
all of the filtering is made between the PDM and PCM stages, and the high frequency is due to the fact that a 44kHz (or even 384k) PWM file would be useless - they work at around 2.8M and 5.6MHz
anyway the high frequency at which they work doesn't have anything to do with the filtering
>>51771383
>>51771367
>>51771263
What are you guys talking about? Sampling rate is that number of samples the ADC takes in a second and it doesn't have anything to do the hearing range.
Audio with higher sampling rate is more detailed, that's about it
>>51774648
>Sampling rate is that number of samples the ADC takes in a second and it doesn't have anything to do the hearing range.
no and no
sigma delta ADCs don't "take" audio to convert it directly in PCM streams, and they work at frequencies in the order of 5MHz
sampling rate determines the highest frequency a device can sample, so it has to do with hearing range if the sampling frequency is lover than about 40kHz
that's why you don't see music encoded in 22.05kHz audio files
![22201_700x[1].jpg](https://i.imgur.com/P14IBvl.jpg "22201_700x[1].jpg")
>>51775782
>tailor swift
>>51771013
I never understood why it made a difference
ITT: OP is a faggot who discovered spectral analysis yesterday and now he thinks he is an audio engineer
Analyze this bitch
>>51778324
Op here, all i see is a cat
>>51778324
Top kek
