Slashdot Mirror
Digital and Analog Audio's Curious Coexistence (cnet.com)
Steve Guttenberg, writing for CNET: It's a funny thing, the ongoing turntable sales surge shows no signs of slowing down, but nearly all new music is recorded digitally. It seems like a contradiction, turntables and LPs are purely analog in nature, but nearly all new (not remastered LPs) made over the last 30+ years were recorded, mixed, and mastered from digital sources. Older, pre 1980 LPs were made in an all-analog world. Today's LPs are hybrids of a sort, the grooves are still analog, but the music was probably made in the digital domain.
Be that as it may, LPs, regardless of vintage, can sound great. While pre-1980s records may be richer in tone and warmth, there are lots of more recent albums that sound just as good or better. In other words vinyl's sound quality or lack thereof has mostly to do with the quality of the original recording, and the choices made by the recording, mixing, and mastering engineers.
Despite the overwhelming number of digital recordings, there is still a tiny percentage of all-analog recordings being made. To cite one mostly analog studio, the legendary Electrical Audio, which owner Steve Albini told me records and mixes around 70 percent of all of its sessions on tape.
Be that as it may, LPs, regardless of vintage, can sound great. While pre-1980s records may be richer in tone and warmth, there are lots of more recent albums that sound just as good or better. In other words vinyl's sound quality or lack thereof has mostly to do with the quality of the original recording, and the choices made by the recording, mixing, and mastering engineers.
Despite the overwhelming number of digital recordings, there is still a tiny percentage of all-analog recordings being made. To cite one mostly analog studio, the legendary Electrical Audio, which owner Steve Albini told me records and mixes around 70 percent of all of its sessions on tape.
In other words vinyl's sound quality or lack thereof has mostly to do with the quality of the original recording
No, if everything comes from the same digital master, then vinyl's difference in sound quality comes from imperfections in the medium itself.
Just got an advertisement in my email for a bluetooth input/hybrid vacuum tube output amplifier. Like, what's the point?
Give me CDs any time. I'm glad to be rid of hiss, pops, scratches, wow, flutter, 5% total harmonic distortion, stretching, rumble.
The dynamic range compression required to stop the needle jumping out of the groove plus the non linear frequency response of the needle itself and the also non linear way the actual dynamic range changes as the needle gets closer to the centre (and so is effectively moving slower) give vinyl a particular feel/sound which is what some people like. They fool themselves into thinking its better reproduction of the original source that digital - its anything but.
However music is subjective and its what you like that matters, not how true it is to the original.
Haha! Allow me to scratch your eyeballs out and you tell me if your earballs think it still sounds great! This is just stupid people in the world doing things that moves the needle just a bit. Kind of like fucking crack whores - you know it's never going to end well, but you do it anyway because you don't know better. You are, after all, republican.
This article on Myths of Vinyl has some interesting facts
lol. Hello soulless communist. Your life is measured in experience and not by marketing specifications. There is more to life than efficiency and five year tractor production targets. Do not forget the ceremony of dropping the needle and the perusal of the gate-fold sleeve.
P.S. Republicans are deluded children, scared of everything.
Facts are history now plebs have politics for religion on social media.
And these are the same LIES that the record industry (and 'audiophiles') were using when CDs first came out - describing CDs as 'clinical' and 'sterile'...
Good luck with your rapidly degrading, irritatingly difficult to use and handle vinyl records.
This is just another example of how gullible people are nowadays, and how they will jump on bandwagons in order to signal to other people how 'superior' they think they are. Vinyl records have no advantages whatsoever over digital recordings, and loads of disadvantages.
I'm happily using my Sandisk Clip Zip with 32GB of songs, and my PC for song playback when I'm on my PC.
A regular 44khz audio CD can't capture the full resolution of a digital master done at e.g 96khz.
The thing is, human ears can't capture it either.
Physics/physiology has a nasty habit of popping in the way.
More seriously, there's a point in the digital domain (basically when it has reached and overtaken the limitation of the human ears you're targetting) beyond which you can consider the sound perfect and all the problem coming from the medium. And as you point out :
But imperfections in the medium are more likely to cause differences you can actually hear.
(Perfect: it's not actually. But unless you have a few bats and dolphins that managed to hide among your public, you can ignore safely the difference).
(Also, hoping that the digital to analog conversion isn't horribly distorted).
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
The obsession with analog audio stems from a gross misunderstanding of what digital audio is. People see digital sampling as a partial capture of the analog waveform, and thus conclude analog must be superior. Digital sampling is not a partial capture. It's an exact capture of the analog waveform within the frequency range (22 kHz in most cases - well beyond what most people can hear). The part that's not intuitively obvious which trips most people up is that if you take a digital sample of an analog waveform, there is only one possible analog waveform which passes through all those digital samples while not exceeding the frequency cutoff. So the digital sample ends up being a perfect reproduction of the analog waveform (within the frequency range of interest).
You can demonstrate this by taking an analog waveform, feeding it into a digital sampler, then converting that digital sample back into an analog waveform. The beginning and ending waveforms will be identical despite the latter one having been converted to digital and back to analog.
All the "warmth" and "richness" of analog audio is nothing more than distortion.
I'm glad to be rid of hiss, pops, scratches, wow, flutter, 5% total harmonic distortion, stretching, rumble.
You can gladly exchange them for saturated over-loud mix, where your equalizer's "frequeccy analyser display" has all the display bars permanently stuck to the top, with frequent pops and clicks due to range-clipping.
(More seriously, there is a key difference :
- Vinyl's defect come from limitation (and fagility) of the medium.
- CD's biggest problem come from the idiot at the mixing table who tries hard to get more attention by attempting at being louder than the others
But these defect might be also a reason to why people might try to avoid digital media : not because inherent flaws, but because they are fed up with the type of mixing that ends up being done on those media.)
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
Nobody would really argue too much about your last statement. It's what that 'distortion' adds is the question.
Pointless unless it's to (re)ignite the tired old debate about digital vs.analog & recycle some gold monster cable jokes.
Anybody who worked with the old analog studios will tell you how "noisy" they were; harder to use too, and the damn tapes always seemed to break at a critical moment, or strangely erase themselves, or just get plain lost or stolen...
Digital also allowed many more people (for better or worse) to record and mix cheap & fast.
Anyway, the main reason why a lot of "digital" music on CD sounded crap was the way it was mastered, not the way it was recorded or replayed.
A lot of the music itself was pretty poor to start with too, which didn't help...
Worse, many people's introduction to CD was in a bundle with crap (cheap) amps and speakers; this sounded not good compared to (Grand) Dad's audiophile setups with massive class AB amps and speakers the size of iceboxes.
Listen to a decent CD (remastered pop / rock, jazz or even easier to find, classical) on a decent rig and see what I mean; it sounds a LOT better than an LP.
Of course, that does not mean that people (and include me too, please), don't love a "noisy", growling rock recording with overdriven amps and speakers howling & distorting all over...
But I don't want "snap, crackle & pop" in the middle of a quiet section of an opera aria, thanks.
Now get off my lawn etc.
Also, people had gotten used to the artifacts that recording, mixing and vinyl playback
"But I don't want "snap, crackle & pop" in the middle of a quiet section of an opera aria, thanks."
Play 'em wet. 50/50 distilled water and ethanol, add a drop or two of ethylene glycol (or dishwashing liquid), wet a cloth with it, wring it out until it just stops dripping, and wipe that over the vinyl. Play, then dry before putting away.
Reduces heat from friction, too.
They sentenced me to twenty years of boredom
It turns out that music recorded on a potato will produce copies that sound like they were recorded on a potato.
If you gave me a choice between a printer and a giraffe with explosive diarrhoea, i'll get my ladder and my raincoat
I hate to even say it, but there are two limitations to your (completely correct) statement:
The first is that you need to filter the signal, stripping out any frequency components above the sampling frequency (half the sampling rate). This is a necessary condition of the exact reproduction characteristic of Nyquist-Shannon sampling.
The second is that this filtering necessarily discards some inter-track phase information from multi-track recordings. Inter-ear phasing is an important part of the way we process sounds in our brain. The difference in time it takes a sound to reach our ears and crawl through our hearing mechanisms tells us about where a sound is in three dimensions, and the difference in time taken by various reflections and echos of that sound tells us what kind of room we are in.
It turns out that none of us can hear anything over 22 kilohertz, which is why they picked that sample rate. And the minimum phase difference in a CD corresponds to roughly a quarter inch difference in the free air path of the sound (napkin math) .
I don't know if the minimum phase difference that our brain can respond to is known or even knowable. But if it is smaller than a quarter inch, I'd be amazed.
At any rate, that phasing information was lost in the recording and mastering process anyway, for nearly all recordings.
See that "Preview" button?
The analog waveforms we are speaking about are sound waves.
The microphone that translates these to electrical signals does not do this perfectly.
The cable that transmits the electrical signals to a microphone preamplifier does not do this perfectly.
The microphone preamplifier that amplifies the signal for further treatment does not do this perfectly.
The analog-digital converter that converts the signal from analog to digital does not do this perfectly.
If someone attempts what you describe in this context, taking an analog waveform (sound), recording it through a microphone, converting it into a digital sample, then converting it back to analog and emitting it, the beginning and ending waveforms will NOT be the same.
A Scarlett Solo single combined microphone preamplifier and AD-DA converter costs $100 at retail. The Avalon M5 microphone preamplifier costs $1665 at retail, with far less functionality. Clearly, anyone who pays for the latter is simply an idiot fooling themselves, because they never studied enough maths to do the appropriate equations.
Really? Me, and the parent "flamebait"? Someone's drunk.
They sentenced me to twenty years of boredom
Go to your local symphony orchestra for a full analogue performance as many of these are done without microphones or speakers.
This is the only way to get a full analogue experience of the sound of most instruments.
This reminds me of the early days of CDs with people looking for DDD on the box.
https://en.wikipedia.org/wiki/...
Keep the Classic Slashdot.
The real points here are not with the recorded sound quality, but with the listened sound quality.
If you insist in using earpieces, noisy environments, low attention and MP3-like sources, the vinyl ain't any better than a 192 kHz MP3. And it loses quality play after play.
If you instead use high end speakers in adequately insulated rooms and keep your attention to the sounds you're are listening to, than maybe the first plays on a vinyl will make some difference.
But also SACD and DVD-Audio can do the same while retaining its quality forever (thanks to the copies you can make).
So, in the end, why vinyl? Just because of marketing hype!
Sent as ripples into the electromagnetic field. No single photon has been harmed in the process.
Many people think analog is more natural and sound loses something else in its digital conversion. Especially when you compress it or lower itâ(TM)s sampling. Also I think something more is lost with digital and that is the physical media of vinyl and sleeve it came in. There is something good about not being perfect in audio.
Aside from the only truly valid reason to own a turntable, which is, 90+ % of all the music produced prior to about 1990 will never be released in a digital format ... in other words it's about the software, not the hardware. It is the fundamental reason for owning a vinyl playback system, or a cassette deck for that matter. All this hardware talk is just noise. Sure, some people want better playback of these analog formats, but focusing on that is a huge Red Herring. For some reason Tech writers can't get past a focus on hardware, and that goes for digital as well as analog audio.
But, we live in an analog world when it comes to music. It starts analog, and it ends analog (playback). A very, very long time ago I learned that with electronics, every time you make a translation ... whether that's simply recording live to tape or Digital Audio Workstation, or a change in format, or any number of ways to do a job with the electronics ... and there is always the final translation to moving air in a room, you lose something. Maybe not much, but something.
The other thing is you use the best tools for the job. Recording on a DAW is better than recording on magnetic tape, the only real viable alternative option. Yes, you can record direct to (vinyl) disk, but that's hard and doesn't lend itself to large quantity replication, so it's a niche example. It is better than mag tape, but it's also severely limiting, an "old-school" technique, live to final mix, that was happily abandoned when multitrack recording technology came along.
So, whatever tools were used to *create* an album, when it's final form is finished, that's your product. It doesn't matter if it was recorded, mixed and mastered on a DAW anymore than it matters that the artist used a toy piano or a concert grand to make the music. Once in finished form, then it matters how it's played back, because a vinyl record doesn't sound like a CD, and it shouldn't sound like a CD, otherwise there is something seriously wrong going on (with the CD, probably).
So, a phono cartridge is a transducer. Like a dynamic microphone, like a loudspeaker. What distinguishes transducers from other parts of the playback chain is they are not powered devices. A phono cartridge has no power supply, it generates it's own voltage through movement. If you push on the cone of your subwoofer, it generates a back-electromagnetic force on the power amp. And so on.
And although it's not obvious to most people, when you listen to music through a modern sound system, you are listening to the power supply, modulated by a music signal. So the quality of the power supply is paramount to the sonics.
CD player? Power Supply modulated by a music signal.
Amplification? Power Supply modulated by a music signal.
But not a phono cartridge. There is a vast array of issues to deal with when you have to use a power supply driven by mains current from the wall. It would not be an exaggeration to say that almost everything in audio that has developed since the early 20th century is the story of power supply technology and ways to modulate that supply.
So, it would be unusual if vinyl *didn't* sound different, even if the final product (the shipping software, in LP or CD or whatever form) was created exactly the same way.
Not audiophiles but Nyquist deniers !
People on here who say something is lost really really haven't read up to Nyquist or watch the excellent
"D/A and A/D | Digital Show and Tell" video on YouTube.
They are true science deniers. They say it's better but can point to no measurement of why this is. The best they can come out with is frequencies above 22KHz, which are likely noise and even if not, most cutting heads cut ultrasonics to avoid overheating the cutting head anyway. Yet they still claim their medium that is crackles, gets worn out, is likely mono at low frequencies to avoid the needle jumping out of the groove (above the subwoofer cut off frequency) is better.
A few reasons to like vinyl, the art work, avoiding the loudness war and nostalgia. Best to digitise vinyl of first play and never play again, this digital recording will always be the best one.
The analog is always better people need to ask themselves, so why is our DNA is digital, simple, to maintain fidelity across copies.
There is no helping some hipster people.
And the minimum phase difference in a CD corresponds to roughly a quarter inch difference in the free air path of the sound (napkin math) .
There's no loss of phase information in a quantized bandwidth-limited signal.
See this video https://xiph.org/video/vid2.sh... at 21:00
Wrong. A square wave requires infinite sampling frequency and cannot be reproduced from 22kHz sampling.
(let's assume 15 kHz here as that's in the audible range, unlike 22 kHz).
If your original 15 kHz signal was a square wave, it would have had frequency components above 15 kHz - specifically, it would have components at 15 kHz, 45 kHz, 75 kHz, 105 kHz etc. Nobody has claimed that a 44 kHz sampling can capture those. But then your ear can't capture them either, so no harm done - a 15 kHz square wave will sound exactly the same as a 15 kHz sine wave (providing the 15 kHz component has the same amplitude).
As an aside, before sampling you'd also need some low-pass filtering to cut off everything that isn't below 22 kHz to prevent aliasing.
but of course the filter has no way of knowing whether the original signal WAS a square wave , or sawtooth or triangle or anything else so to say it can reproduce it exactly is incorrect.
I said "provided they are below 22 kHz". A 22 kHz square wave has higher frequencies (all at odd multiples of 22 kHz, lowest at 66kHz and 110kHz), so it violates that condition. If you take a 22 kHz square wave, and you limit bandwidth to 0-22 kHz, you get a sine wave as the output.
None of this matters, as your ears cannot pick up the 66 kHz harmonics either, so you cannot tell the difference between a 22 kHz sine wave, square wave, or any other waveform with 22 kHz fundamental frequency.
After purchase of very special audio equipment which allows precision tuning of channel delay on microsecond level I was frankly surprised that changes of 10 microseconds in delay between two channels are actually audible (although difference is very subtle). Quick thinking would lead to believe that there might be signals with transients which would benefit from >44.1 kHz sampling, but thinking of human hearing range largely cancels out such assumptions - to matter much, perceived frequencies involved should be on range of close to or over 10 kHz, which again is beyond the range of sound localization by interaural time difference.
Thus I mostly believe that although there are subtleties on channel delays and whatnot (you can really hear a shift of a vocalist in the aural field if you change delay between channels by 10 microseconds), they're not really relevant regarding limitations of 44.1 kHz sampling. Also, for all sensibly practical purposes outside a mastering room, 16 bits of dynamic range should be sufficient if it's employed sensibly.
Saddest part about many of those audiophiles enthused to care about things is that often their listening rooms are the ultimate bottleneck in faithful reproduction. Even when they're well designed, which they often are not. Go and measure the frequency response of a listening room from the listening position, and look at the high end of unfiltered result. Unless you're doing the measurement in an anechoic chamber, it looks essentially like a bar code. Any minor differences on high frequency end of source quality are likely to be completely washed out by the listening room characteristics, which often change significantly just by moving head by an inch...
I don't think we could recognize different waveforms at 22kHz. From how I think the ear works I would say it does a Fourier transform and recognizes the frequencies of the elementary sine waves, but I could be wrong.
It's a funny thing, the ongoing turntable sales surge shows no signs of slowing down, but nearly all new music is recorded digitally.
The only people who care about vinyl records are young people who never had to grow up with vinyl records so they lack an appreciation of what a pain in the ass they are or old farts with an overdeveloped sense of nostalgia or hipsters who want to show off. No the sound is NOT appreciably better especially after you have actually played the record more than a few times. Vinyl records are fragile, readily damaged, and generally sound like shit after any appreciable amount of use. Even if you are absurdly careful with a vinyl record it's still almost certain to get damaged at some point. Sharp needles and soft vinyl tend to be a bad combination. Whatever minor advantages they might possess are quickly lost with actual use. I don't really buy the arguments that vinyl somehow sounds better but even if it does the differences are so marginal as to be meaningless.
Be that as it may, LPs, regardless of vintage, can sound great. While pre-1980s records may be richer in tone and warmth, there are lots of more recent albums that sound just as good or better. In other words vinyl's sound quality or lack thereof has mostly to do with the quality of the original recording, and the choices made by the recording, mixing, and mastering engineers.
Sigh... "Richer in ton and warmth"? That sounds like typical audiophile bullshit to me unless you are talking about some corner cases. I'm old enough that I predate the CD. Vinyl records and cassette tapes were the only options in my childhood. No the sound was not better. Mostly worse if anything. It was just what we had at the time and we dumped vinyl records almost overnight for CDs because vinyl records SUCK to use in the real world. Any issues with digital music not sounding a particular way have NOTHING to do with analog vs digital and everything to do with engineering choices.
The placebo effect is real, too. You can actually cure illness with sugar pills! Bottom line: If analog sound is better in their heads then it really is better (for them).
Umm, no. That is not an objective truth we can agree upon. The placebo effect is real but you don't design a medical system around it either. Similarly, just because a single person claims they prefer an "analog sound" you don't go and pretend the laws of physics somehow are different for that one person.
If anything, vinyl lends itself to highly compressed mixes by default, because it has a narrower dynamic range than tape. Compressed mixes in digital are not a result of the medium, they are a result of INTENTIONALLY applying processing to make them compressed. You can have an extremely compressed all-analog recording if you do choose. Conversely you can have a full dynamic range mix with an all digital recording, if you choose to.
you can now get the same turntable arm vibrations, speaker feedback, channel overhear and clicks/hiss/scratches with a digital player. Good thing is that you can decide exactly how bad you want the sound. You are not forced to accept what you 200.000USD shitty sound system gives you.
I have one good thing to say about vinyl... The LP covers are an awesome format.
I have had people suggest that maybe they can "feel" the higher frequencies somehow.
People also suggest that homeopathy actually is something more than a placebo. Doesn't change the fact that it is a bullshit claim that has yet to withstand any scientific scrutiny.
In the digital domain, a 'square wave' can be reproduced from a ten hertz sampling. The timing will probably be off, but the edges will be at a right angle.
Stop being a dork.
Are you outing yourself with this off-topic statement followed by your sig?
"P.S. Republicans are deluded children, scared of everything."
"plebs have politics for religion on social media."
"Worse, many people's introduction to CD was in a bundle with crap (cheap) amps and speakers; this sounded not good compared to (Grand) Dad's audiophile setups with massive class AB amps and speakers the size of iceboxes."
My own "introduction" was when I added a $500 Sony CD player to my ~$2k stereo system back around 1983. The first thing I noticed was that some CDs sounded great while others were crap. After a while someone pointed out that almost all the CDs were being produced in either Japan or Germany, and it turned out that the German ones mostly sucked...though I'm not sure why.
"Now get off my lawn etc."
I'm older than the dirt under your lawn sonny.
Just another day in Paradise
Discrete levels (0-65535), but not a square wave (0-1).
No, you are incorrect. You would see a 22KHz sine wave as you would expect. Why not a square wave? Because a 22KHz square wave has spectral components greater than the bandwidth of the channel. The same goes for a triangle wave and a sawtooth. Those all have spectral components outside of the bandwidth of the channel.
The reconstruction filters on DACs will reproduce all signals in the passband nearly perfectly. It doesn't have to 'guess' at what the signal was.
You would see an exact copy of the original.
See the video illustrated above for examples addressing this exact point.
Briefly, most drawings of digitized waveforms, and some low-quality DACs, use what's called a "zero-order hold" output filter, outputting the value sampled at time t until the next sample time at time t+1/f.
This common illustration is highly misleading. A zero-order hold filter is known by everyone who works with digital audio to be inaccurate and it is not used in any application where accurate reproduction is desired.
> filter has no way of knowing whether the original signal WAS a square wave, or sawtooth or triangle or anything else
And it's irrelevant. The only difference between those waveforms is the presence and quantity of overtones at 44, 66, 88, etc. kHz. If you're only reproducing up to 22 kHz, there is no difference.
Human hearing can't detect overtones above 20 kHz (and even that limit is generous), so a human being can't hear any difference between different waveforms at 12 kHz. As long as the fundamental amplitude is the same, the rest is thrown away by your ears. They all sound just like a sine wave.
I am what people here call an old fart with an oversized sense of nostalgia. The turntable is my perpetual analog loophole, forever protecting me against the DRM shit content owners would like to force upon me.
Analog = Unlimited ownership of your physical medium.
As digital first, and bit lover, I have to attest, analog has something old-school, retro, vintage, slowing down, chilling and relaxing effect - Revox B77 reel-to-reel: https://www.youtube.com/watch?... https://www.youtube.com/watch?...
Then why does the BluRay spec have audio sampling at 192KHz? Why is "CD" quality fine for music but movies need sampled an order of magnitude deeper?
Only the State obtains its revenue by coercion. - Murray Rothbard
Can you give some figures on this claimed distortion? Start with %THD and then IM.
Only the State obtains its revenue by coercion. - Murray Rothbard
You're mkaing the flawed assumption that they're mastered the same. The master on the CD will most certainly be compressed to the point that anyone under 30 should be able to hear the distortion, particularly to the high hats. If the vinyl is mastered separately, and it generally is, then it will sound a whole hell of a lot warmer because it's not compressed nearly as brutally.
I've often thought the same way, and it seems like a reasonable way of modelling it. I suspect, however, that the actual nature of it is far weirder (though I have nothing to go on here).
There is no XUL, only WebExtensions...
That was my point you pillock.
So pure digital output from a sampler running at twice the frequency in question can produce a sine wave before filtering can it? Wow, you should write a white paper on this, clearly you have a nobel prize waiting!
He has no facts, no understanding, and yet he's posting on CNET. Genius!
On the high end, a vinyl is 22KHz/12-bit, a 44/16 CD easily surpasses the sound quality. None of this matters to idiots listening to Spotify/Pandora/Amazon/iTunes/Google on a portable Bluetooth speaker.
So pure digital output from a sampler running at twice the frequency in question can produce a sine wave before filtering can it?
Normally, you would first increase the sampling frequency in the digital domain, and then convert to analog. The residual error can be pushed to the high frequency domain (> 100 kHz) where it can be easily removed with a simple analog filter.
0-1-0-1 and 0-100-0-100 are both square waves.
APK quotes people (including myself) without context and should not be trusted. Just thought you should know.
A 22KHz signal captured at 44KHz would only have a 0 and a 65535 with nothing in between. Not much different than a 0 and a 1. Let's say it's perfectly aligned with the sample rate for simplicity. The first sample is at the peak (or 65535). Because it will be back to the peak again two samples later, the next sample will have to be at 0. And the third sample would be at 65535 again. Any waveform shape is imaginary at this limited resolution - there is only peak/valley/peak/valley.
It isn't possible for there to be literally no loss of phase, but I may have overestimated the minimum increment. Thinking about it some more, I think the minimum time resolution varies with the frequency of the input wave, and inversely with the product of the sampling rate and the noise floor (bit depth).
My bad for posting while still half asleep.
He covers the whole topic of timing in about 30 seconds and does so in very scant detail, and with an infinite wave very very far from the filter frequency. It would be interesting to see the demonstration done with a precision delay line, a 20 kHz signal and various bit depths. If I'm right, and I may not be, even though I'm fully awake now, you can't accurately reproduce an 88 nanosecond phase delay on a 20 kHz signal using 8-bit PCM at 44.1 kHz.
I should hedge that a little more. I actually calculated that for a 22.05 kHz signal, right at the limit, where all of the phase information necessarily has to come from the bit depth. If we drop to 20 kHz to get under the filters, the resolving power of each bit gets a little bit stronger. 50 nanoseconds is probably too small to register digitally - but still trivially resolvable on a decent analog scope.
See that "Preview" button?
My daughter wants a USB turntable, which is the stupidest thing in the world. It takes an analog signal, digitizes it to send it over a cable, then your amp does a digital to analog conversion to send it to the speakers. In other words, it offers the worst of both analog and digital playback. (Admittedly it would be useful for ripping old LPs to MP3 format. My daughter doesn't have any records that are only available in vinyl.)
I've abandoned my search for truth; now I'm just looking for some useful delusions.
I think all recording now are digital.
The production is so much easier. The days of splicing tapes and track hopping or 24 track limitation have long been over.
Only crazy motherfuckers like Dave Grohl bury themselves in garages to record on old tape equipment to get "The Sound"
Vinyl does not magically convert a Digital recording to analog. It just compresses the signal to the bandwidth available to the stylus.
p
Listen to an analog mastered recording from the early 70s, and compare it to one from the early 80s, right around when CDs came out, but before digital recording was the norm, and the quality difference is huge. I would expect a modern analog recording to sound really, really good if the technology continued along that path, and it probably sounds great on vinyl without any DA/AD steps in between.
I grew up with vinyl, and then transitioned to CD audio, and for the most part I am happy with the sound (on a good system, with a good DAC). I've started to dabble in 24/96> but really don't find the improvement that huge. I still have a high end turntable and cartridge, but so far have felt no need to hook it back up. A true complete analog signal path created with the most modern technology might give me a reason to revisit that idea.
Tape saturation / aka over biasing. People forget that magnetic tape was the only way to record in the days of vinyl LPs. I was a studio engineer at the time of introduction of digital tape recording. It was common then for the studio engineer to over bias teh tapes, like +4 or even +7, to get a "hotter" or "fatter" sound. This is really the origin of the warmth that people here in vinyl. I did a session where the drums, vocals and bass went to a reel to reel digital tape recorder, and the guitars went to a +7 biased analog 2" and the DTR at the same time. The guitar tracks sounded much better on the analog tracks, hands down. There are other factors involved, too, like the deterioration when the original mother is cut on the lathe, then that is used to make multiple negative metal duplicates, for pressing, which, in turn are used to press the disks themselves. The mother is cut from lacquer or acetate, then sprayed with metal paint, make what is called teh daughter, the negative, which is use to press the blobs of vinyl into the LPs. so there is degradation there. Wheat is recorded onto the mother and winds up on the LP does not have a true EQ. It has the top and bottom rolled off, to reduce needle bounce and mistracking. This is done through the use of what is called the RIAA curve, and the loss of top and bottom is restored at the phono input stage of the amplifier. A number of studios are returning to muti track analog recorder to get that over biased sound. there are digital plug ins that will hopefully duplicate that sound, but, and this is coming from someone who had one and tracked on one, a 2" , 16 track Ampex flat top machine, biased to +4, just sounds perfect, and digital will never be able to get that sound..
Republican leadership = Idiocracy
There are significant advantages in the studio to tracking to tape, even knowing full well that the material is going to be digitized during mixing, mastering, or manufactureing.
The logarithmic diminishing returns of +inputsignal:+recordedsignal as the metal oxides below spun by the the write head's magnetic field exceed 80% of available produces a naturally responsive dynamic range compression that is more pleasing to the ear than that produced by most stand-alone dynamic processors.
This quality allows the recording engineer to safely use the more of the available dynamic range than when tracking digitally, which sounds harsh ( sudden square-wave-ish clipping) when pushed too hard. In contrast, the distortion that occurs when pushing tape both subtly fades in and out with the signal and also leans more towards triangle-wave shapes rather than square-wave shapes, making it more pleasing to the ear.
Tape also exhibits a gradual fall of in reproduction signal strength as frequencies climb. This can also be pleasing to the ear, both by taking the edge off of super-trebley sources like crash cymbals or some kinds of guitar distortion effects without the noise & distortion introduced by many (finite impulse response) EQ processes - and also by passing less signal into the ugly-responding cut-off range of the anti-aliasing filter when the material is digitized downstream. Moving the cut-off of the AA-filter farther away from our range of hearing and allowing gentler knees (and thus less distortion as input signal approaches the cut-off frequency) on the filters as they approach the now solidly-ultrasonic Nyquist limit is actual the primary benefit of 88.2/96kHz audio).
Those factors combined lead to the feel of "warmth" people describe. Many of those things can be accomplished similarly in the digital realm with the right tools, practices, & knowhow; but with tape, they just happen naturally.
Wrong, these are sawtooth. They describe only peak and neutral. For a sine wave approximation you need at least 2 additional points between peak and neutral. A sine wave is more like 0-0.4-0.9-1-0.9-0.4-0-(-0.4)-(-0.9)-(-1)-(-0.9)-(0.4)-0
The reason old vinyl records sound good is because the music and musicians were better! Also, the recording engineers didn't compress the life out of the recordings ...
Why not? If we can help some people without using drugs, why wouldn't we?
Seriously? Are you trolling or stupid? I can't tell...
Because placebos by definition do not do anything. The effectiveness of placebo is the bar we use to determine whether a medicine does anything. If it isn't more effective than placebo we don't use it because it doesn't do anything. Furthermore prescribing a placebo introduces dishonesty and potential fraud into the relationship between the doctor and patient.
I get that the ethics of lying to a patient are complex, and skeptics likely are immune, but there is a large group of people we could be helping with placebos and I see no reason not to start there.
Sigh... BY DEFINITION you cannot help large groups of people with placebos and you sure as hell don't design a treatment system around them. That's called snake oil and the administration of false remedies is a real problem. The entire point of placebos is that they have no mechanism of action. They don't do anything by definition. That's not medicine or science. That is faith and prayer.
Wrong, harmonics do not exist in analogue world.
In the analogue world, square waves, sawtooth, or triangle waves don't exist either, so I guess we're even.
to add: I left this out - in mixdown, the old way was to go to 2 track with 0 bias. That went to the mastering lab to be tweaked and used to cut the mother. I got to listen to several audio production paths: 24T (or more) analog> 2T analog> LP mother>etc. 24TA>2TA>digital (no processing) 24TA>digital. multi track digital? digital. My ears liked 24T analog dumped as 24T to digital, best. While I have heard several types of digital tape emulation plug ins, the real thing beats it hands down. I will qualify this, though: Some types of music, like live orchestra, sound best done full digital - you need all the nuances that digital captures. But for "pop" music, the warmth that analog tape saturation gives adds to the recording. I see that most of the posts here address recording resolution, like 44 / 48 / 96 and the higher rez does give a better sense of air and space, as it captures those frequencies that we don't "hear" but sense". But it's all pissing in the wind if the result is converted to MP3 or played back on less than top level audiophile gear. It is nuts to track on 96 when you know yur audience will be listening to your tracks on spotify, through $100 headphones or a car audio system.
Republican leadership = Idiocracy
It's not about the laws of physics, it's about what that person prefers. Think of it as black coffee vs coffee with cream and sugar; the same laws of physics that allow you to prefer one allow me to prefer the other. To you, one is better; to me, the other is.
If you cannot measure the difference there is no difference. That is the only way to have a rational discussion about any of this. People like to pretend they can hear/see/taste things that aren't actually there all the time. Audiophiles claim all sorts of absurd things that they cannot identify under rigorous conditions. Wine lovers have been shown to imagine differences between wines based on price tag or even the same wine with food coloring added. I'm not talking about preferences, I'm talking about actually measurable differences based in physics. You have to have those first to have a justifiable any preference. Otherwise you are just making shit up and wasting everyone's time.
You can demonstrate this by taking an analog waveform, feeding it into a digital sampler, then converting that digital sample back into an analog waveform. The beginning and ending waveforms will be identical despite the latter one having been converted to digital and back to analog..
Funny then that you can hear such a huge difference between different DA converters then.
Simple in theory, but maybe not so much in actual practice.
You are correct, I should have used 0-1-1-0-0-1-1-0 and 0-100-100-0-0-100-100-0.
APK quotes people (including myself) without context and should not be trusted. Just thought you should know.
you can't accurately reproduce an 88 nanosecond phase delay on a 20 kHz signal using 8-bit PCM at 44.1 kHz.
Correct. And I misspoke above when I said "quantized", I meant "sampled".
You can capture any delay with a 44.1 kHz sampled signal, in a bandwidth-limited source, even down to 88 ns, provided you have enough bits. .
Of course, in practice you don't have unlimited bits. The net effect of using fewer bits is that you add noise to the signal. With 16 bits, and no noise shaping, you could get 90 dB Signal/Noise ratio if everything else was perfect.
So, while you cannot achieve a perfect 88 ns delayed copy of the original 20 kHz signal, you can get that perfect 88 ns delayed signal summed with -90 dB noise. As long as you can't hear that noise, you can't hear the difference between the original signal and the reproduction.
The problem isn't that the source is digital. Studio quality digital has a high enough resolution that it captures the nuances lost in translation to mainstream distribution methods. Be it MP3, or CD there is loss, and that loss makes for a poorer recording. That's the attraction of vinyl is that you get back what's been lost in digital copies. Vinyl has problems. It wears with every play, surface noise, portability, etc. It's not the most desirable means of distribution. Now DVD-A, or BluRay audio is much better, but it's not as common. I believe the reason it's not common is because you'd have a near perfect copy that will last a lot longer than a record, and record companies want you to keep repeatedly buying the same recordings.
I don't believe in karma, I just call it like I see it.
"Wrong, harmonics do not exist in analogue world."
Did you even take a music class in school or are we still doing like Jethro Bodine and still sitting in 5th grade?
*plays a natural harmonic on a guitar*
I mean, it's not like the term hasn't been in use for several fucking centuries before digital anything came into existence.
Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
knowing whether the original signal WAS a square wave
It doesn't need to. The original signal was a sine wave, as every periodic signal is composed as a sum of sines and we can't heard sines above a certain point. So while a 200Hz square wave and a 200Hz sine wave sound worlds apart, a 20khz Square wave and a 20khz sine wave firstly sound the same as the limits of our hearing is the fundamental and wouldn't even make it to the first harmonic.
There's no waveforms other than sine waves when discussing range limits like human hearing or nyquist. 10kHz is the highest frequency square wave that a human can hear as something other than a sine wave, and a human with perfect hearing will hear that as a combination of a 10kHz sine wave and a 20kHz sine wave, both of which fit just fine in the constraints of a CD. ... Unless you're a dog.
By the way probably just an oversight since you at least had some clue, but if you look at a 22kHz unfiltered output from a 44.1kHz system you'll see a triangle wave, not a sine wave.
You seem to think that producers/engineers will create two cuts of a recording - one with heavy compression for digital, and one unrestricted for vinyl. How quaint!
Doe *all* engineers ? It's very likely :*NO*.
Do *some* studio make an new mix for vinyl ?
Some will. And that contributes to the myth that vinyl sounds better.
(The sound *currently stored on this vinyl* does indeed sound better than the sound *currently stored on that CD*.
By conscious choice of the engineer doing the corresponding mix.)
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
Except you can't, because there aren't enough samples to represent 22KHz at 44,1KHz. It would indeed have to be 0-1-0-1 (or 0-65535-0-65535, depending on volume level). You only get two samples per wavelength at the upper limit (the nyquist limit).
Also correct, but my point was that a square wave doesn't just have to be 0 and 1.
APK quotes people (including myself) without context and should not be trusted. Just thought you should know.
And my point is that the distinction is meaningless at 22KHz at a 44.1KHz sample rate. A sine wave, square wave, sawtooth wave - all get the same exact digital representation when quantized to digital at this rate.
There's a huge distinction between a value range of 0-1 and a value range of 0-65535, though, which affects all waveforms. Square waves can have a range of different volume levels, as can any other, while RoccamOccam's representation made it seem as though they must be full volume. Your point is congruent to mine, not counter to it, but also completely irrelevant to the conversation that was being had.
APK quotes people (including myself) without context and should not be trusted. Just thought you should know.
That does not move the speaker membrane in a sine wave pattern. Membranes move in linear fashion. This patterns move the membrane in a sawtooth patten. That is why you need sampling frequency 4 to 8 times the frequency which sine wave you want to capture. Most music instrument's sound waves are very close to sine, regardless of frequency, because they are created by mechanical vibration.
Regardless, that is still the claimed limit of the Nyquist theorem to accurately reproduce 22KHz sound.
Vinyl is physical, collectible. You buy it and you have it. There's also a bit of a ritual involved with pulling it off the shelf, out of the giant sleeve, and setting it on the turntable and setting the stylus and all that. Even if people don't think it sounds better, I suspect it lends a feeling of ownership and permanence to music buying and listening that downloads or even CDs, which are so easily ripped and copied, do not.
I just do not have the time, space, or budget to participate in all that at this point in my life, nor do I know if I would want to. But I think I get some of the appeal.
The Daddy casts sleep on the Baby. The Baby resists!
I don't think you understand fully what argument you are making, which is my point exactly.
A 44.1khz sampling frequency can accurate represent the shape of the sound waves up to 5.5125Khz (1/8 of sampling), and semi accurately represent the shape of the sound waves up to 11.025Khz (1/4 of sampling). Anything above 11.025khz cannot be accurately represented at 44.1khz sampling rate.
The human ear is a pressure sensitive organ and can detect the difference due to the different sound pressure the different wave forms represent. This is the integral of the waveform function in a quarter period. A sawtooth has a relative sound pressure of about 0.5, the sine of about 0.63 and the square of about 1. That is why the shape of the wave is important to the ear and the sampling frequency matters.
You need a sampling frequency of at least 160hz to be able to accurately describe the shape of the waveform at 20khz, in order to convey the proper sound pressure to the ear.
It is not compression at all. The RIAA EQ curve is an industry standard, in which the top and bottom are EQ'd down, then the reserve EQ curve is applied on playback at the phono input stage. Thinking that compression is the same thing as EQ'ing is stupid.
Republican leadership = Idiocracy
but also completely irrelevant to the conversation that was being had.
Actually, you're the one that threw the thread off the original subject. The original subject was that a 22KHz waveform is represented the same as a square wave (technically sawtooth might be more accurate) at a 44.1KHz sample rate. Lower frequencies would have more sample points along the curve and a cleaner sine wave shape.
There's a huge distinction between a value range of 0-1 and a value range of 0-65535,
The value range is only the amplitude of the wave, not the shape of it, when you have only two samples per peak/trough. This specifically only applies to 22KHz and near it. At lower frequencies, the details of the amplitude helps actually define the waveform shape. At 22KHz specifically, the amplitude (sample value) only defines the volume and nothing more.
Actually, you're the one that threw the thread off the original subject.
Actually, go back and read the comment to which I was replying; I was not the one who brought up the value range issue, I was merely replying to it.
The value range is only the amplitude of the wave, not the shape of it, when you have only two samples per peak/trough.
I never claimed otherwise and, in fact, stated that you were correct the first time you made this point.
APK quotes people (including myself) without context and should not be trusted. Just thought you should know.
Jeez, people at least google this stuff before you post. The RIAA curve is a standard EQ curve inserted between the output of the mastering console (or 2T tape machine) and the lathe, that reduces gain at the top and bottom. When the LP is played back, the amp's phono input stage restores the roll of, applying the original, only reversed.
Republican leadership = Idiocracy
A 44.1khz sampling frequency can accurate represent the shape of the sound waves up to 5.5125Khz
Nope. It can accurately represent any waveform, as long as all frequency components are less than 22.05 kHz.
https://en.wikipedia.org/wiki/...
Actually, a 22KHz signal at half intensity would have 0 and 33768 (16384 and 49152 if the DC offset were removed). Hell, 0 and 1 could well be the limits, even if the range was 65535, if the amplitude were small enough and the DC offset great enough. I think l4ko was correct, elsewhere in the thread, when he said you don't understand your own point; indeed, it seems you don't understand much of the topic at all, while I've personally been working with audio for 2/3 of my life and understand it quite well.
APK quotes people (including myself) without context and should not be trusted. Just thought you should know.
You might be making a point of how much information is needed to convey an accurate signal. That is described by Nyquist's Theorem, which does state that 22KHz can be reproduced from a 44.1KHz sample rate. It's a lot more complicated of a mathematical proof than I have time to learn, but it's held up for a very long time.
Still, my point is mroe about what actually happens to 22KHz when quantized at 44.1KHz. Which was the point of this whole thread. You don't get a sine wave shape encoded. Playback is a different story, since I'm sure that playback equipment tries to fit a curve to it anyway.
When you said:
You are correct, I should have used 0-1-1-0-0-1-1-0 and 0-100-100-0-0-100-100-0.
I was only saying that wasn't possible to have this for 22KHz at a 44.1KHz sample rate. Which was what the thread was on. It might be more correct to call it a sawtooth wave.
Maybe we got confused on different interpretations of this:
There's a huge distinction between a value range of 0-1 and a value range of 0-65535, though, which affects all waveforms.
Which I took them to think meant that there were gradiating values between peak and trough when there can't be. And it seemed like you were agreeing with that.
I spent a big chunk of my morning by reading almost all posts in this thread. The condescending general attitude is amazing - it's digital - so shut up. Well - as usual - it's not that easy. I own a record label and we create vinyl records. Some with and some without the help of DAWs. First and foremost: We are specialized in vintage audio. So all our source material is either analog or derived from analog masters and transferred to DSD (if the master owner doesn't entrust us with the original masters). As long as there is no restoration work necessary (the source material is not damaged) we won't need digital. We use our fully analog tool chain in the studio to bring the sound up to par. That is tedious work - and it requires passion, know-how and determination. We don't simply crank up the big bottom on the APHEX - before we even start the work we have listened to other recordings or performances by the same artists to get a feeling how THEY wanted it to sound. The difference in a performance of Beethoven's "Eroica" by Bernstein or Karajan are critically important. Karajan, compared to Bernstein, is much more subtle not as overwhelming and the audio engineer has to understand that. So is it analog? Or digital? I'd like to start by asking: What equipment do you have and where are you listening to it? If you're sitting in front of your computer and listening to you Logitech speakers, you're perfectly fine with mp3. If you have decent speakers and decent equipment AND the time to enjoy the performance, you may want to listen to a medium that has never been in the digital domain. In all other cases, you're ok with a CD. Sound is analog. Go to classical concert, close you eyes and listen. That is the measurement for true sound. That is the sound you feel, you experience - without amplification or any kind of technology. That is the sound that Beethoven, Mozart or even Benny Goodman or Duke Ellington wanted. It can't be captured by any means. From the microphones to the tapes, mixers and other equipment - we can only try to approximate it as close as possible. And in order to be able to appreciate the performance, you need the right equipment, the right amplifiers and speakers. You need a recording that has been worked on by people who cared. And yes - I truly believe that analog sources should not be converted into the PCM digital domain. So finally - our customers appreciate our analog only approach. Yes - there's a difference between vinyl, CD and .. say .. DSD. But - to be dead honest - most people don't even have the right playback equipment or even the time to appreciate it.
I already posted on this, but yeah. Just saying that peak/valley/peak/valley is all the samples you get at 22KHz. Regardless of what the peak or valley values are, there's no gradiation along the waveform curve like there would be for lower frequencies - the information is not captured at that sample rate.
I was only saying that wasn't possible to have this for 22KHz at a 44.1KHz sample rate. Which was what the thread was on. It might be more correct to call it a sawtooth wave.
That was never in dispute, but thank you for clarifying. You are correct, those would be 11.025KHz square waves.
Which I took them to think meant that there were gradiating values between peak and trough when there can't be. And it seemed like you were agreeing with that.
It does seem you've found the source of the confusion.
APK quotes people (including myself) without context and should not be trusted. Just thought you should know.
For what it is worth, I don't believe that people can hear the 22 microsecond phase delay (aka quarter inch) from my initial calculation, much less a 50 nanosecond one. But I also don't think it could be comprehensively measured in the mind either way.
A lot of dubious and overpriced gear gets sold in the gap between "can't prove" and "can't disprove".
The interaural phase difference really does get thrown around as a reason to buy very expensive analog gear. My initial estimate, despite being off by a few orders of magnitude already strained credulity, and was very simple math. The overlap between "has money" and "does math" is apparently not great.
See that "Preview" button?
Nobody would really argue too much about your last statement. It's what that 'distortion' adds is the question.
And that's the problem. A true audiophile would never "add" to the original performance that is being reproduced. It's impossible for a vinyl medium to not introduce additional sounds or alter the sound on playback. Pops, clicks, wow, flutter, etc. Yet, these vinyl fans are almost all self proclaimed audiophiles.
Very strange.
"A plan fiendishly clever in its intricacies"- Homer Simpson
I see people arguing the digital is better than analog. They are really just different and are treated as such.
Things like classical recordings can sound much better digital. This is because you want the most accurate reproduction possible. Stick two nice mics over the front of the stage and record. Edit as little as possible.
With Rock and Roll or Punk, or even country etc all of the parts become the instrument. Drums are close miked with mikes that are more designed to be fat and big sounding but not 100% accurate. Cymbals are miked with more accurate condenser mikes but all the low end is pulled out. Guitar amps are close miked etc. Vocals are recorded with large diaphragm condenser mikes. Then digital, analog and tube processing is used to alter and even distort the sound. Art and science becomes mixed. You create and mix for a sound, not accuracy. All of the products used, mikes, mixers, tape, tubes, analog processors are like brushes, palettes and colors to a painter. Different artists, both musicians and engineers, will have favorite tools and will be known for their sound using them.
This continues to the mixout and even the final product. Engineers and the musicians in the 60s, 70s were thinking about the final output being on tape or a record and recorded for the best sound on those devices. If you take the the final mix of a 70s recording and just transfer it to digital you will not get what the engineer in the studio was after. He expected those imperfections and in some cases is counting on them to give the sound he/she wants.
Yes, you can duplicate many of the analog effects used in the digital world but its not the same. They feel different when you are working with them. When you listen to a record from the 60s or 70s on a good stereo you are hearing what the artist and engineers want and expect you to hear. I have worked for years with both digital and analog. Both are great and useful but things like a kick drum really sound great on 2" tape,but that sound does not transfer to digital, god knows I have tried .
Records are not better or worse than CDs, just different. Both have their place.
AdFuel
I also don't think it could be comprehensively measured in the mind either way.
Measuring phase delay is often simple as asking the participant to point at the source of the sound. Phase delay is a major component of how we locate an audio source in 3D with our two ears. (It's been a while, but I seem to recall hearing things "above" you is entirely caused by phase delay.)
The DSP in my car handles phase delays down to 8 microseconds (overkill) specifically to let the DSP "shift" the sound stage so the driver hears the sound image as "front and center." The idea is to compensate for the driver being closer to one set of speakers than the other, and it does work -- for the driver.
Various "spatialized" audio techniques that have been made over the years also depend on it (such as SRS, QSound, A3D, which are all variants of head-related transfer functions).
-- Sometimes you have to turn the lights off in order to see.
Yes, I am familiar with them. I still mourn the loss of my original pre-soundblaster A3D card and the rumble headphones that came with it. I ganked a lot of noobs with that card.
I said "mind" specifically because audiophile robbers are absolute scum. We are talking about distances that are hard to see, much less hear. There is no way that someone could point out a 22 microsecond delay, much less a 50 nanosecond one. But that doesn't keep the scum from claiming that it is "better" in some nebulous way.
Even failing A-B testing doesn't invalidate the "more pure" theory.
See that "Preview" button?
50 ns is definitely too small.
But a 22 microsecond phase delay can be detected*
(*) for a set frequency range, and even then the tone will merely sound like it came from a few degrees further the side. Woo!
Don’t get me wrong - I hate audiophile charlatans as well. It’s a 21st century version of patent medicine & snake oil. Patent medicine is a good analogy, though — the placebo effect isn’t just for drugs.
Except with patent medicines, we did get a number of tasty soft drinks.
The “Loudness War” is the best reason to use Vinyl — Vinyl’s deficiencies made it impossible to abuse the signal past a certain point.
-- Sometimes you have to turn the lights off in order to see.
That xiph.org video is great for theory, but not in practice. The Digital-Audio converters introduce their own inconsistencies in conversion. Not to mention the steep low-pass filter that needs to be used to remove the aliasing frequency (which is less steep at higher sampling frequencies). These create a noticeable difference in sound quality between CDs and high resolution formats.
Don't believe me? Studies show that people *can* hear a difference between 16 bit 44.1/48k and high resolution audio:
http://www.aes.org/e-lib/brows...
I worked in a recording studio in the late 1990's, and I was disappointed at the time hearing the sound quality of a mixdown from multitrack to 16bit, 44.1 DAT (the standard for mixdowns at the time). Quite a bit of the original sound was lost in conversion.
As for vinyl vs. CD quality, I think the difference is more subjective; both formats have their faults. For me, vinyl tends to have a clearer sound in the high frequencies, while CDs have better low frequency reproduction. Since most of the music I listen to doesn't need solid low frequencies, I like listening to vinyl better. Overall though I would prefer digital files in high resolution format (24 bit, 96k or higher), but since most music can't be bought at that quality, vinyl is the best substitute.
I think you've got that backwards. Oversampling allows MUCH steeper anti-alias filtering.
> That xiph.org video is great for theory, but not in practice.
Huh? It is a vivid demonstration of the excellent fidelity of digital audio in practice, using analog test equipment.
I was 23 when I bought my first audiophile piece of gear: a pair of speakers I bought from the factory next door to the dictating equipment company I was working at in 1974. I forget the brand; they were small and regional. The speakers had a dowdy old dark walnut wood appearance, had something similar to an infinite slope crossover (see Joseph Audio), and a dome tweeter that could be 'aimed' by rotating it. I've no idea how good or bad they were. But I got a good deal on B-stock.
Next, a NAD 50 watt integrated amp and a separate tuner, bought in 1981. Turntable was a Systemdek IIX. A few years later I acquired a Sony small CD Walkman player. When the amp died I replaced it with a Golden Tube Audio SE-40 amp and associated pre-amp, and eventually Linaeum 10 speakers. The turntable was replaced with a Well Tempered Turntable. The CD player was upgraded to a Rotel-855, and then a Rega Planet. By the time I turned 60 some years ago, I could tell my hearing was not as acute, and I discovered Youtube with close quarter monitoring, using a Monsoon Planar Magnetic speaker system. The amp was replaced with a Panasonic SA-XR55. I sold the Rega CD player and WTT and relied on an Oppo DV97H1.
I'm sure other audiophiles will recognize some old friends.
At one point I had over 3000 LPs. As someone above mentioned the subject, many never made it CD. At some point I had about 2,000 CDs.
During the 80's up to about 2000 I also went to a lot of high end audio shows and had friends who had much better systems, such as large Soundlab speakers with Accuphase amplification. I also regularly attended classical music concerts from small to chamber orchestra to large orchestra, Indian music concerts, Jazz, and even some rock. Eventually I started playing hand drums and shakuhachi, merely an amateur, but it gave me a greater appreciation of music. Ah, yes, I mostly stayed in the game for the love of music, but was seduced by the love of tech at times.
The major bit of information (pun by happenstance) I want to relay to many posters above is this: electronically recorded and reproduced music, or merely amplified music, rarely sounds convincing at replicating the the sound of live, unamplified music.
One of the major changes during the post-digital era, is that many people involved in creating recorded music stopped putting musical quality (as in sound reproduction, not performance) that you could listen to and hear as a worthwhile goal. Measurements and mathematical/engineering replaced listening to the end result, be-it an LP or a CD.
Of course, to many, listening is of lesser value than measurements. Because we can believe in numbers.
Is it? Can we? Consider this: the piano, violins and strings, acoustic guitar, lutes, horns, woodwinds, percussion, etc. etc, were all invented and developed prior to the invention of electronics. But the moment someone puts a mic in front of one of those, the ear isn't good enough?*
Me? I can still sometimes listen to a recording and recognize the sound of a Shure M57 instrument microphone. In a pinch one can also double as a small hammer.
*Thanks to pianist James Boyk who made an earlier version of that argument about 15-20 years ago. It went something like this:
as a college professor and expert in the piano, I'm expected to be able to tell a Yamaha from a Steinway, Van Clybourne from someone else, when a piano is in tune or broken, but if someone puts a mic in front of one my knowledge is scorned as a personal opinion.
"Most music instrument's sound waves are very close to sine,"
Hardly any musical instrument tones are close to sin waves at all.
Here are some examples:
http://www.audiomisc.co.uk/asymmetry/asym.html
Where are they getting the information that things over the last 30 years have all been recorded digitally? Almost every CD I own is AAD - meaning it's analog recording and mixing, and digital mastering. I have a few that are DDD, but that might be a dozen or so. Granted today digital recording and mixing is much more common, but going back to 1988? I am skeptical of this claim.
You don't understand sampling theory or filtering. The group delay introduced by a digital filter can be any amount; it is not in any way quantized to the sample rate - there is no smallest value. Think of the filter as an averager of a sort; it can delay a waveform by an arbitrary amount. Fully awake doesn't seem to be helping. More math.
Listener preference for LP over CD-or-better digital—is not based on LP being a better medium; it is not, by any objective and any fidelity-based subjective measure. I suppose others on this thread have commented on nostalgia or faux-nostalgia—think "millennial" or "hipster"—but there is a better reason that some prefer LPs. That reason is that LPs in many cases are not created with the same signal as the corresponding CD. The LP signal is _better_! It is a documented and audible fact that many LPs use a signal that has been subjected to less dynamic range compression and less peak limiting, both of which are used extremely heavily to horrible effect on most recordings of the last 25-30 years. Look up "loudness wars." I have personally seen histograms of the same tracks taken from LP and CD by a colleague and the differences are striking. The track was "300 m.p.h. Torrential Outpour Blues" from Icky Thump by the White Stripes. And here https://www.soundonsound.com/t... is an interview with the recording folks involved, including this quote from the second sidebar:
"Jack wanted the CD to sound loud and aggressive, so it was cut as hot and exciting as possible, whereas the vinyl was cut in a more traditional way. The vinyl version has more size and dynamics and air, all the things about vinyl that we love. Was the CD version brickwalled to compete in the loudness wars? Let’s hope not!”"
Go get some graph paper and draw a full cycle sine wave. Call that 20 kHz. Calculate the time per division of the paper for 44.1 kHz sampling and mark the sample points. Measure the voltage at the three sampling points. Make a table with quantizations of those three sample points at various bit depths.
Now, draw a second full sine wave offset to the left (delayed) by 1 nanosecond. Measure that new wave at the same three times. Add those samples to your quantization table. Figure out how many bits you need to get different values from the first wave.
Congratulations, you now have a chart that shows that there indeed is a minimum phase delay that can be represented by a given sample rate and bit rate, and for a given input frequency. Feel free to repeat the exercise at different input frequencies, different sample rates and different phase offsets if you want to explore all of the dimensions of the limit.
See that "Preview" button?