Steve Albini

What was his point? Link or summary?

>>69686117
https://www.youtube.com/watch?v=LaVC6FGSHVg

>>69686349
5 mins in, agree so far.

>>69686349
i don't know if i agree, but i definitely am enjoying this lecture

>>69686349
He sounds right, what's the counter argument?

File: 05f1e187784a1660a06b7697d9fc7175.jpg (33KB, 450x393px) Image search: [Google]

33KB, 450x393px

whys he so fucking based

he's like on the borderline between cringe and genius

>>69688370
Cringius

A modern computer can emulate basically any old OS. It's 2016 and now we can emulate MS-DOS on a phone. There will definitely be a way to access any of today's daws 50 years from now.

Audio files never stop working. Lossless formats can always be converted to any other lossless format without losing quality. You can always export all tracks from a project into separate audio files. Like a tape, but much more efficient.

>>69686027
Tape wears out.
Digital files don't, storage systems can (like hard drives and CDs) but the files can be stored on the cloud and shit, and be transferred to new HD or whatever if needed, much easier and cheaper than transferring tape.

>>69688673
Fake

someone should set fire to his old tape just for irony

>>69688937
kek.
i wonder if he at least backs up his tapes digitally.

Why does /mu/ call him an edgelord? curious

Digital recording is about work flow, consistent recordings, time management, and keeping prices reasonable for all clients

>>69689508
'Pop music is for idiots and children' - Steve Pepperoni

>>69689508
he makes noise rock and has opinions

>>69689508
why we call the guy who named his band rapemen an edgelord?

>>69686027

It's fine if he prefers recording on vinyl and likes the way it sounds, but digital files can capture the entirety of the frequency spectrum as far as humans can hear it (roughly 20Hz-20kHz) and due to that, a digital recording will have more clarity (assuming it's not intentionally recorded or mixed completely poorly). Vinyl can't reproduce a lot of low end, and a lot of the high end will get lost when reproduced on vinyl too. However, the natural compression/saturation you get when recording to tape is a sound a lot of people enjoy. I do as well, but being able to record digitally is so much more convenient and cost effective as well as sonically having many more options to choose from.

>>69689508
>I think I fucked your girlfriend once... maybe twice. I don't remember. Then I fucked all your friends' girlfriends... now they hate you

>>69689758
>>69691394
I thought there was more, that's not all that edgy

File: 1481792029616.png (443KB, 960x948px) Image search: [Google]

443KB, 960x948px

>>69686349
>muh resolution
jesus christ, he's fallen for the high resolution meme

File: tpae.png (60KB, 771x340px) Image search: [Google]

60KB, 771x340px

>>69691206
Tape, not vinyl.

>>69688700
>Audio files never stop working
>Failing to account for rotational velocidensity

>>69691459
>implying hi-res isn't better
poon.

It's possible to kind of reproduce the sound of vinyl by increasing the 200-250hz area by a few db in the EQ and cutting some high frequencies, you will never be able to reproduce or even get close to the sound of a digitally mastered sound of a CD on a vinyl.

The winner is pretty obvious here, only memers think that even if the CD is perfectly mastered vinyl wins.

>>69691537
i don't think you know what high res for music means

it's the idea that music will sound better if you play it above frequencies we can't hear the 44.1 sample rate is as good as it will ever get because we can hear anything above that, hell, adults can't hear over 16khz anyways.

plus most equipment can't record in frequencies higher than 20khz, only electronic instruments can.

File: teces_164.gif (2KB, 352x180px) Image search: [Google]

2KB, 352x180px

>>69691627
It's not about the higher frequencies, it's about increasing the resolution of the frequencies that we can hear

>>69691673
i don't think you understand that either
stair stepping means nothing because it's an analog signal in the end and will be the same as the top.

there are couple videos clearing the miss information of stair stepping

>>69691714
>it's an analog signal in the end and will be the same as the top

Not at a low resolution. The example of the pic of the low sample rate definitely won't sound like the input signal. Obviously the sound waves won't actually come out of the speakers with unnatural vertical and horizontal, but the result is approximating the input signal. The higher the resolution, the closer the approximation.

>>69691804
This guy gets it.

>>69691804
unnatural vertical and horizontal lines*

>>69691823
thanks I agree with me

>>69691673
>>69691804
Jesus Christ people still believe this shit?

https://en.wikipedia.org/wiki/Nyquist%E2%80%93Shannon_sampling_theorem

tldr if we sample at 44.1khz (and brick wall filter all frequencies above 22.05khz) we can reproduce the original analog sound 100% perfectly

>>69691832
You're doing gods work, anon

>>69691832
Sum up what the wiki is saying. I'll wait

>>69691886
read it yourself, cunt

>>69691886
I did already but I'll add more detail if you want, warning it's been 2 years since I took this class so sorry if it takes a min.

Information in the analog domain is stored like a sine wave, volume (or power) over time, in the Frequency domain information is stored volume (or power) over frequency, so a single sine wave in the frequency domain is just a vertical line at the point on the x axis for the frequency of the sine wave, and the height is the power.

Great, so what does that mean? Well a Fourier transform allows us to translate between the analog domain and frequency domain, and if we take a brick wall filter in the frequency domain (basically delete all information in the audio file above a certain point, we choose 22.05khz because humans can hear up to 20khz) and take the reverse fourier transform of it we get something called the Sinc Function. Well if you take a digitally sampled audio file (so literally just volumes and timestamps) and then take the sinc function, centered at every point in your sampled file, scaled so it's peak matches the volume peak of each timestamp one at a time and then add it all up, you will get a perfect representation of the original audio file.

The only catch is aliasing, which is what occurs when there are more than one possible analog outputs for one set of digital inputs, that's where the 44.1khz sampling rate comes in, Nyquest says that you need exactly double the sampling rate of the highest frequency in the original analog sample to perfectly reproduce it with no aliasing

>>69691886
>>69692165
also for reference, this is an example of aliasing
https://upload.wikimedia.org/wikipedia/commons/thumb/2/28/AliasingSines.svg/675px-AliasingSines.svg.png

and this is the sinc function
https://upload.wikimedia.org/wikipedia/commons/thumb/5/59/Si_sinc.svg/670px-Si_sinc.svg.png

>>69692165
there is actually no aliasing
https://m.youtube.com/watch?v=cIQ9IXSUzuM

>>69686027
he thinks digital should be an archival format, but never explains why he can't just record digitally with open source formats/codecs and make a tape copy of the high resolution digital recording for archival, along with a digital copy attached to information about its file encoding
I don't believe he hasn't thought this through, he probably just likes the process of recording with tape

>>69692277
there is no aliasing at a sample rate of 44.1khz if we have no audio at a frequency above 22.05khz, but if you didn't brick wall filter and have signals above that you would absolutely get aliasing, and aliasing in audio is not like aliasing in the visual, there isn't any stair stepping. Aliasing in audio is just a signal of the wrong frequency showing up when the source material was a different frequency, like that image I linked here >>69692231 the red signal is the one that is above 22.05khz, the blue one is what is re-constructed when our sample rate is at 44.1khz

File: Screen Shot 2016-12-15 at 11.11.14 PM.png (59KB, 1037x463px) Image search: [Google]

59KB, 1037x463px

>>69692157
I've outlined a spot on my scrotal sac that I would like you to kiss

>>69691832
That's what I was going to bring up, doesn't that just prevent aliasing? You may be preserving the exact frequencies, but does that necessarily imply that the frequencies are sonically exactly the same as the input signal? I mean here's like 13k at 44.1k. It's maintaining the frequency but that's clearly not what 13k should look like

>>69692351
that image is not what audio aliasing looks like, visual aliasing is a different beast because we are not used to discussing frequency in the visual/2d fields in the same manner as the audio fields. I just typed this out above for the other dude.

please at least read the intro to that wiki article on Nyquest it gets to the gist of it, you don't have to understand all the math

"The theorem also leads to a formula for perfectly reconstructing the original continuous-time function from the samples."

>>69692351
for reference I should also say that your DAW or whatever that screenshot is from does not do the proccess of taking the sinc function and adding it at every sample to simply display it to you, that is done by the DAC

>>69692433
I understand, there is no aliasing at 44.1k. What I'm asking, is just because you're keeping the "wrong frequency from showing up" does that necessarily imply that the reconstructed frequency will sound sonically the same? How is it anything more than just an approximation that is exponentially getting closer to the real thing?

>>69692478
I get that the signal approximates the spaces between samples. Will an approximation based on that 13k sound wave sound the same as one with 4x more samples? or an analog signal?

>>69692534
Yes, Categorically the Nyquest-Shannon Sampling theorem guarantees 100% perfect reconstruction of an analog signal from a digital storage medium if the sampling rate is at least 2x the maximum frequency stored. It will sound exactly the same as one with 4x, 8x, or even the original analog signal itself.

The reason it's not an approximation is the "magic" of the Sinc function and the fact that it is derived from taking a reverse Fourier transform of a frequency domain brick wall filter

File: 1411799844903.gif (3MB, 412x304px) Image search: [Google]

3MB, 412x304px

>>69692619
Well I don't quite get the math so I'll just have to agree that I'm right

>>69692704
Assuming you aren't just being sarcastic, you don't have to agree with me, random internet person, this one paragraph from the wiki article says the same thing. That if there is no information in the signal above a cutoff frequency, your reproduction is perfect.

"Strictly speaking, the theorem only applies to a class of mathematical functions having a Fourier transform that is zero outside of a finite region of frequencies. Intuitively we expect that when one reduces a continuous function to a discrete sequence and interpolates back to a continuous function, the fidelity of the result depends on the density (or sample rate) of the original samples. The sampling theorem introduces the concept of a sample rate that is sufficient for perfect fidelity for the class of functions that are bandlimited to a given bandwidth, such that no actual information is lost in the sampling process. It expresses the sufficient sample rate in terms of the bandwidth for the class of functions. The theorem also leads to a formula for perfectly reconstructing the original continuous-time function from the samples."

>>69692771
I see that it claims to produce perfect fidelity, but I don't understand the math well enough to know if that's true, and I'm tentative about accepting it without understanding it

>>69692843
it clicked for me when I did a step by step reproduction on a test signal by hand, drawing out the sinc function and watching it all match up nicely.

also this is not a new theorem, it's been known since as early as 1959, it's incredibly well studied and the basis of the entire digital audio industry.

File: IMG_1582.jpg (2MB, 4032x3024px) Image search: [Google]

2MB, 4032x3024px

>>69692843
Also I hope you are happy I just drew this for you

>>69693061
>>69692843
shit sorry I thought it was rotated, I guess 4chan didn't like the orientation...

>>69689818
The band was named Rapeman after the japanese superhero.

for a guy that has had his name on a million albums, his bands sounded like shit until he started Shellac, so i don't really give a shit what he thinks