Digital audio - Is it sliced and diced?
Vložit
- čas přidán 31. 05. 2024
- Analogue audio enthusiasts hate it that digital audio is sliced and diced, and then reassembled for listening. Digital enthusiasts will deny this but, in my video, I demonstrate with incontrovertible evidence that digital audio truly is sliced, and truly is diced. [Includes my irritating slicing and dicing music]
Digital Show & Tell - xiph.org/video/vid2.shtml
Monty Montgomery - en.wikipedia.org/wiki/Chris_M...
00:00 Introduction
00:49 What's so perfect about the analogue signal?
02:16 Where does 'slice and dice' come from?
03:11 Where's this going?
04:10 Going back to Monty...
04:51 Going back to Monty again...
05:58 Over to you...
06:30 Notes from the Mancave
DAVID MELLOR'S MUSIC
David Mellor's music on Bandcamp - davidmellor.bandcamp.com/
David Mellor's music on Spotify - open.spotify.com/artist/6OkaD...
Available on all good streaming services
EQUIPMENT USED TO MAKE AUDIO MASTERCLASS VIDEOS
CAMERA - Sony Alpha A6600 amzn.to/3uj7Dtq
COMPUTER - Apple Mac Mini M2 Pro amzn.to/47tW6Wg
MICROPHONE - Sennheiser MKH 416 amzn.to/3QWWvuO
ALTERNATIVE MICROPHONE - AKG C414 amzn.to/3SLqCad
BUDGET MICROPHONE CHOICE - Rode shotgun amzn.to/3G0ni36
MICROPHONE PREAMPLIFIER - Comica LinkFlex AD2 amzn.to/40SMHFT
TELEPROMPTER - Padcaster Parrot alternatives amzn.to/40A55TJ
TELEPROMPTER SOFTWARE - PromptSmart Pro promptsmart.com/
AUDIO INTERFACE - Behringer UMC202HD amzn.to/3ugj6db
HEADPHONES - Sony wired amzn.to/3sEjlht
SPEAKER CABLE - KabelDirekt amzn.to/3ucgQ6T
POWER AMPLIFIER - Fosi V3 amzn.to/3vRumNH
KEYLIGHT- Amaran 100d amzn.to/40A5Gor
FILL LIGHTS - Switti RGB amzn.to/3ufhOPu
SOFTBOX Phot-R alternatives amzn.to/3QGoTjJ
LIGHTING STANDS Phot-R alternatives amzn.to/3QWXOKe
MONITOR - Samsung UJ59 amzn.to/3MEoBsk
MONITOR ARM - Invision alternatives amzn.to/40SLqi5
MIC ARM - Neewer amzn.to/49Ob0sD
TRIPOD - Velbon alternatives amzn.to/46gu2oa
GREEN SCREEN - Elgato alternatives amzn.to/3SG2F3L
DESK - FlexiSpot E7 Pro amzn.to/3QVIWvS
VIDEO EDITING - Apple Final Cut Pro www.apple.com/final-cut-pro/
IMAGE EDITING - Adobe Photoshop amzn.to/3FZ7Ww0
AUDIO EDITING - Avid Pro Tools amzn.to/3QA5Wz7
NOTEBOOK - Leuchturm 1917 hardback amzn.to/47zonKX
PEN - Sakura Pigma archival pen Sakura Pigma archival pen amzn.to/47zz3cP
OSCILLOSCOPE - Hameg HM 203 alternatives amzn.to/47tWN1O
GROOMING - Philips 3000 nose hair trimmer amzn.to/40BdLJo
GROOMING - Nose plier amzn.to/41IOZaR
HAIR COLOUR - Just For Men amzn.to/475OgS1
HAIR CONDITIONER - TRESemmé Silky & Smooth amzn.to/41AdWFd
Disclaimer: Videos on this channel may include affiliate links and may feature software that has been received free of charge with an NFR (not for resale) licence, equipment that has been loaned or donated by a manufacturer or distributer, or equipment that has been bought for less than full retail price.
![My First DAC! With FOUR important digital filtering options and audio demonstrations [iFi Go Bar]](http://i.ytimg.com/vi/CdHNEq6C9TM/mqdefault.jpg)
![My First DAC! With FOUR important digital filtering options and audio demonstrations [iFi Go Bar]](/img/tr.png)
I was a bit of an audiophile back in the '70 -'80s and I had excellent hearing back in the day. I was in the top independent stereo retailer back then when the first digitally-mastered album came out - Ry Cooder's "Bop til you Drop" - and they were very excited to play it, and they played it all the time. Why? Because it sounded phenomenal! No noise floor at all, every detail revealed, incredible dynamics and the biggest, cleanest bass you ever heard.
What was crystal clear was that digital was superior to analog recording, and CD's were going to be superior to vinyl records. And they were - even in my system with a lousy CD player versus a Thorens TD-160 turntable with a top-flight Shure cartridge.
Are there poorly mastered CD's? Absolutely. But the technical superiority of digital recording and playback are painfully obvious to those of us who only had analog for many years.
Indeed. People forget that in the 80's when CDs became widespread we were (nearly) all blown away by how much better the recordings sounded in this new format. I have plenty of vinyl but I don't kid myself that those LPs sound better.
Agreed. CD has been the biggest audible advancement in my lifetime.
I like occasionally playing my Vinyl records but I prefer my CD's for all the reasons you mentioned, plus the convenience of playing up to 80 minutes, or more on a multi disc player, without having to turnover or change the disc. They are also easier to maintain and keep clean than records, which unlike CD's, wear a bit every time you play them.
My first mid-80's cdp was an akai. It was abysmal. It skipped if you sneezed in the room. It sounded tinny and noisy with the host of bad cd pressings I got from Japan. In 88 I got a much better cd player that worked. I got some pressing from Germany, and I finally had faith in the format.
When slicing and dicing onions, the true audiophile would be in tears when it happens to their precious music, so it was appropriate after all.
That's very funny. So funny in fact I wish I'd thought of it.
I love the warmth of my 22,05khz, 8bit depth, DAC
It's interesting how I cannot hear any effects of slicing and dicing when I listen to my CD's, be they AAD, ADD or DDD recordings. I believe that the reason for that is found in the explanation by Monty Montgomery on his excellent video about digital audio, that everyone should watch.
Sound enters our outer ears, our eardrums vibrate, this gets mechanically amplified through the bones in our ears, and excites fluid in our cochlea, when nerve hairs pick up the vibration, and send nerve impulses to our brains. Impulses aren't continuous, they are pulses. They are essentially electrical signals sent along the nerve fibers, with each pulse representing a specific level of activity. The rate and pattern of these pulses encode information about the sound, such as its pitch and loudness.
So, the nerve impulses our brain interprets as sound are indeed a series of pulses, not a constant signal.
We slice and dice sound to hear it.
🎧
Yes a really good explanation. !!!
Maybe then we should encode digital audio in a similar pattern.
That was one of my first thoughts. We don’t hear a continuous stream. Everything is converted into small electrical pulses by our body anyway. Our ears are converting analogue audio into digital anyway.
@@mwalmsley72 Exactly - our ears are an analog-to-digital converter and our brain likely is a DAC, converting this signal back to analog...or so we think...
It is even worse. It makes us think what we hear. But in reallity it can be different from what somebody else is hearing.@@chrisk9890
Watched Monty. I thought of reconstruction of digital to analog like dot-to-dot but at such a high resolution that you can't see any straight lines and even if you could there's a stylized curve between the dots from the analog nature of drawing it. What I came away with from Monty was: Reproduction from digital is a smooth analog characteristic. So choose your noise. No available analog medium can reach the low noise floor of the digital accomplished for the masses with the introduction of the CD.
And that's why I am sticking to vinyl, VHS tapes, cassettes, mechanical watches, AM radio, dial micrometers, stethoscope cardiograms, paper maps, and everything else that is analog.
There's nothing like a well-made, accurate vernier caliper.
And don't forget your analog computer to watch and comment CZcams videos.
I am pretty sure that VHS tapes are a digital format.
@@donjohnstone3707 VHS is not digital
You must be big fun at parties.
A brilliant and easily digestible analogy, thank you so much :)
The courses are can not be found in the website.Can not we enrol now?
Love the new minimalist background.
The knife used for slicing & dicing the onion, however, looks like it is in desperate need of sharpening.
It’s a fish knife, too thin for the job. Looked better than anything else I had handy though.
As I understand it: One overlooked aspect of the digital audio process is the reconstruction filter -- the low pass filter after the DAC. The "infinitely thin slices" are very sharp pulses that hit the low pass filter and ring like a bell. Each pulse rings out for many samples afterward, and all these ringy pulses overlay on top of each other. The end result -- the sum of it all -- is the original analog signal with all the missing slices filled in (minus the high frequencies above twice the sampling rate).
The reconstruction filter is what fills in the areas between the slices, er, samples and gets you back to the original signal. And any low pass filter can serve as a reconstruction filter, including the natural rolloff of your ears.
Reconstruction filters are not without their issues. Not for this video, but perhaps in the future.
Actually most DACs have a "set and hold" output... so there is no actual gap between samples.
Feed it 7FFFF ... get 1 volt out forever, unless you feed it a different value.
@@Douglas_Blake_579 That's true. My point addresses the ideal or theoretical case of infinitely thin apple slices and lollipops. It's easy to picture what happens in the frequency domain: The high frequency images are filtered out with a sharp low pass filter and leaves the original analog signal. It's not intuitive what happens in the time domain and how the smooth sine-wavey missing parts are produced between the sample points.
As you said, in actual practice, set and hold (or zero-order hold) is also used to effectively interpolate between the sample points. That first generates the well-known staircase pattern. I believe that reduces the amplitude of the high frequency images and likely makes the work of the low pass filter that comes after easier (at the expense, I believe, of some high frequency roll off that may have to be compensated for).
Once you filter at the input and output the signal has one choice, beeing exactly the same. We are left with quantization distortion. At 96dB - i can’t hear it.
@@melaniezette886
Sadly that's not exactly correct ... for example a DAC or amplifier's internal distortion will appear on the output but not the input. In fact, that's how we measure it.
The ADC/DAC sampling (slicing and dicing) isn't the problem. The problem is the digital processing (baking the apple crumble). Some audio technicians (chefs) know what they're doing, some don't.
They know what they're doing. Unfortunately they're paid by people whose ambition in life is to make the world a worse place.
@@AudioMasterclass I agree, the good audio techs know what they're doing. And their bosses probably do too. But money talks, and they have to repeat what sells, no matter how it sounds. This reminds of the DJ speed wars in the 60's when analog processing went wild. To brighten up the sound, the capstans on the record or tape players would be slightly changed to make the playback speed and pitch a little higher. Eventually it got so bad, you could barely stand listening to it if you weren't preconditioned. The same thing goes for volume compression. Some things never change!
@@bobkitchin8346 I remember reading an interview with Chuck Berry who complained that his songs were pitched up on transfer to vinyl, and then he couldn't sing them live and sound the same.
Hence my long held opinion that most audio _engineers_ would be far better suited to driving trains.
@@Douglas_Blake_579 The minimum age to drive a train in the UK has just been proposed to be reduced to 18. I doubt that competency in audio would often be achieved by that age.
As there are many admirers of High-End and/or Hi-Fi sound here, I ask... Has anyone had the opportunity to listen to Bang Olufsen's “Beolab 90”? If so, what were your impressions?
Very good analogy
Would have loved to hear what 'Audio Phil' has to say on this topic. Maybe it is time for a new character 'Digital Dave'
Phil contends that the parts of the signal that aren’t encoded contain more information than those that are. He extracts them and sends them to his walk in wardrobe system to listen to while he’s selecting his shoes.
24 bit should cover all possible air movements of a wave in our atmosphere, and 24 bit FLAC is amazing. Still, I love my vinyl distortion.
Captured and then reproduced analogue audio isn't infinite, it is bandwidth limited depending on the storage medium (sometimes VERY limited), much the same as video. The original source is infinite but any analogue recording is anything but infinite.
Luckily I just enjoy the music and don't look at the oscilloscope to tell me if I should be enjoying it or not.
Good point ...
I use my scope to make sure it's working right...
I use my ears to see if I'm enjoying it our not.
Well no matter how high the sampling rate is and no matter how many bits per sample are used, the end result can only be as good as the microphones.
Or the performers, the acoustic, and the writing. If they are all good then you only need a half-decent mic to make a hit record.
I enjoy watching sliced and diced analogue films, and listened to a lot of sliced and diced analogue sampled FM radio in my younger days as well.
About forty years ago, Stereo Review conducted a blind comparison study to see if the listeners could discern the audio difference between Monster Cable and 14 gauge zip cord in speaker connections. The listeners could tell no discernable difference. At least that study had those results. There was a fair amount of outrage presented in the letters to the editor by the users of Monster Cable. I am sure there is a lot of beautiful sound emerging from the sound systems of analogue enthusiasts. However, I wonder how such a thing would stand up in a blind A/B study. Would the wee pops of missed dust motes give the nudge? Which would score better as a result?
Having in the distant past seen how crap analog circuitry is when looking at a signal on an older analog Tektronix oscilloscope, square waves, for instance, are treated rather poorly, one can only imagine how floppy large mechanical low frequency drivers image bass frequencies, let alone the desperate jittering of a tweeter's diaphragm "reproducing" those inaudible overtones. We shouldn't be satisfied with anything less than direct neural input!😳
Funny thing is I might have a serious video on that coming up in future.
As Long as the samplerate is over double the frequency of the signal and the filters are good. You should be good.
Thank you for the very important topic.
Digital signal is not sliced/diced at all, Here is the explanation an difference from an analogue signal reproduction:
analog: A Typical analogue channel is based on an analog source, amplification and a transducer (Speaker)
Digital: A Typical Digital Channel is based on a Digital source, Digital to analog conversion and an analogue channel....
The analogue channel involves a closed loop"" or "Negative Feedback" gain stage(s) to reach precise analogue level at any given time.
A Digital channel produces a predicted Analogue signal using an open loop method for achieving the predicted value.
the slices and dices referred on this video are only check points for adjusting the output level to reach its next predicted level.
If you ""zoom in" into a digital signal with the best magnifier you will never be able to find any gap (slice), the converted to analog digital signal is as continuous as any analog signal.
Analogue amplification using a feedback can be often challenged by an impulse signal response, when the input change is faster that feed back response then a glitch happens in the output.
Digital converters which works on prediction do not suffer from such behavior.
To summaries: the digital numbers stored on an audio digital file represent a control signal level points that assures an accurate reproduction of a continuous pure analogue signal.
So no apple crumble?
@@Astrodicted smoothy instead.
Bonus points for spelling analogue correctly. Like most analogies, the taped recording of a sound wave is rather different from the "real" thing. The athlete ran "like the wind" is a good "image" but isn't real. The magic & crucial determinant is "like". Analogues are "like" something and metaphors THE thing...even is mystically unreal. The athlete ran like the wind; she's a stallion.
Metaphor audio... it could be a thing.
@@AudioMasterclass it already exists because Audiophiles insist their isn't LIKE but IS.
@@AudioMasterclass Don't we have that already? Read the notes and let the music play in your head....🙂
Who wants to listen to music that has to be reconstructed?
Apparently a lot of top drawer audiophiles, since vinyl sound requires a RIAA reconstruction filter to boost bass and limit treble by as much as 30db.
Well said!!!. A huge number of vinyl 'purists' believe that they're listening to a perfect reproduction of any given track, not realising that there is plenty of mastering required to make that track work on vinyl.
I can see why some dislike digital though. The loudness wars and poor mastering, reduced dynamic range etc etc, just meant that CD tended to end up with a poorer version of any song, more suitable for radio perhaps.
Regardless, nobody is taking digital and my Roon away from me :-)
@@MattSB2588
The irony of all this is that the medium with the best range (CD) ended up being used for music with the least.
@@Douglas_Blake_579 t'is rather annoying indeed. But, when you've a well produced song, digitally, you have the best version.
There is a database of dynamic range against many hundreds of songs. I don't recall the website, but it was quite interesting.
@@MattSB2588
Yep ... the "Dynamic Range Database" ... with over 10,000 albums listed.
You seem very manic in this presentation, even more than usual.....cheers.
Probably in a hurry to get to his apple crumble!
And I don’t think I’ve peaked yet.
@@AudioMasterclass :)
All Solid State 😊
All Day Digital 😊
I had a friend go to Munich high end audio show last week, armed with possibly the most horrible anti music track I could possibly make, and play it on some of those million dollar systems. Just to take the piss out of everybody and get a few laughs.
I had a demonstrator play one of my tracks on a pair of £20,000 speakers. Still didn't sound all that good though.
You produce a fantastic factual and wonderfully cool channel. I just love any music or expression of the human spectrum. I am being influenced by your lectures.
Well, flac files created from a digital source as a CD are identical. I don't have access to master tapes. Not relying on the DAC on cd player to play music but instead a really nice DAC on my computer usb port (Douk Audio).
A little off topic but relevant to something mentioned early in the video. A five piece band in their tour van crashes on the way to a gig. Luckily none were killed but the local paper reported the next day that four musicians and a drummer were injured. 😊
People who think they are musicians should try playing the drums. Then they’ll know. Source: Me.
Lovely. Hope the crumble was as good as the vid 😉Cheers 🍺
It was!
I wait with bated breath and ears for the next episode.
Enjoyable video, as always. Personally, I mostly care about whether I like the song.
That's how we ended up with Taylor Swift.
Magnetic tape also has distinct magnetic particles. They are just a bit more random than digital samples.
There’s a deeper level to this which would be magnetic domains. I don’t see any comparison with digital at the moment but I might look into it for a future video.
Oh dear. Is that what you have to eat in Britain? I have lived a fortunate life. Nice video as usual. I do enjoy your work.
We also have cheese. Cheddar. Not American.
How did I know you were a Waitrose shopper? I was shocked to find, after moving to West Wales, that my nearest store is two and a half hour's drive away. Oh well..
Got to be Waitrose or Lidl/Aldi. No half-measures.
You didn't peel the apple first! Enjoy your crumble with added peel.
Who doesn’t like a bit of peel? I eat the core too - it’s the best bit. A treat at the end.
I was going to ask about what would happen if you used an onion, and then you did! I haven't got much more to say really, except that maybe we should try measuring the effect of digital and analogue audio on the listener rather than confining ourselves to analysing the audio alone. Of course, this has been attempted - perhaps successfully, I don't know. There was, I believe, an experiment with cows listening to both digital and analogue audio which was conducted back in the mid-'80s. I can't find the paper to reference it, but it was reported in a CD magazine of the day (I would name it if I could remember). I can't recall much about that article, so I don't know if it was a well-designed experiment, whether it led to meaningful findings, or if the findings could be replicated. But there is that question about what you choose to examine or measure. If you analyse what a tape is, you could measure the composition of the strip of film, the binder, the magnetisable coating, and the extent of its magnitisation, but that won't necessarily tell you that The Beatles are in there somewhere (or maybe it would?), or, importantly, what effect it would have on you if you heard the recording (or consumed the diced onion - once cooked, mixed and mastered). Just a thought - and not a very original one.
Truth being told, you probably listen to digital audio all the time and don't even realize it.
@@Douglas_Blake_579 These days it's very rare (for me) to hear purely analogue audio. Even the stuff I record to reel-to-reel is almost entirely digital in origin.
@@IndigoDavei
Yep, that's about right.
My music collection is entirely digital, stored on hard disks or streamed.
In the early 2000s I transcoded big pieces of my Vinyl collection to MP3/192 which was the best bit rate at the time. I made sure the sound quality was equivalent, then I sold off the collection and the turntable. These days the collection is upped to MP3/320 and curated to -16lufs using MP3 Gain, so I can listen without clipping or loudness jumps. Works a treat for me.
Purists will jump on me for using MP3 ... but the software and space to do this with FLAC doesn't exist.
Arnt all modern recordings digitally recorded and mixed?
Most.
@@AudioMasterclass So for most (I think it will be almost all) new vinyl, the audio was sliced and diced at some point before being pressed?
Do you think anyone would be able to tell?
@@ltmund There are much wiser minds than mine trying to find the answer to this question.
@@AudioMasterclass Ah, but by recognising there are wiser minds makes yours a wiser mind
@@ltmund And if you can recognise that I may have a wiser mind then that makes your mind wiser still.
Is that a proper brown onion? Or one of those extra large onions? The latter I find is less tasty. I like a small brown onion, the type that makes you cry, in the way that audiophiles, cry when listen to music on their system. Thanks for the video!
It quit mattering to me, if it ever actually did. There is enough time allotted in life to enjoy both. I'd have to question all that happened to the "mixes" in studio and every other alteration that happens to it and that would drive me mad. It all sounds good, and better, to me.
Where to start, the linked video with "Look at my 8bit scope, the wave looks the same". :- ) THD is a distortion, but there are a lot of other distortion mechanisms. Amplitude/compression, intermodulation of any tone with another tone/s, PSU interaction, third order intercept, spurious free dynamic range, channel separation It’s a growing list. Take the 8 bit reduction example in the linked video, apparently it only affects the noise floor. Take to an extreme, say three bits, where is your amplitude accuracy to the original? Stuck between bits. It’s distortion, but not THD. Good job we can’t really perceive amplitude a low level. Digital convertors are not perfect, the space between bits isn’t equal. Your 44.1KHz clock _generally_ isn’t accurate, as a timepiece, my external soundcard is off by ~5s / day. Thankfully, with 16bit audio and dither, all these distortions will fall well under perception, they are still there, we just can’t hear them, as the final reproduction link, it’s fine. If you want an idea of how well the electronics is doing, you will need a SysID test, and stop doing all the now very old steady state tests in isolation.
You who were there at the time, surely most of it was recorded on tape? Wide bands, if I remember correctly. How did it work? Did you connect a number of microphones? could the tape recorder record ten or more microphone sounds? On occasion, you can give a history lesson.
I have Cuisinart Beyerdynamic Shure microphone.
You'll be wanting a Vitamix Neumann next.
The biggest problem I have with digital audio are when older recordings are remastered to CD. My favorite recording from vinyl has more detail than the digital CD. I even tried a download from HD tracks and same result, missing the cymbals at the start of the first track. So the hunt for new vinyl original is almost impossible unless I want to spend 200 bucks. Next step is to try and do a deep clean on the remaining two vinyl records I have. One is basically worn out and the other damaged. My fault I should have transferred them to CD with my recorder with a 24 bit 96 Khz chip. The sad thing is when I bought a greatest hits CD of Simply Red and watched my VU meter slam constantly at 0 Db on the CD recorder playback, Harsh and irritating with DIGTAL GLARE ! But the CD wins first place in loudness. The other thing is many vinyl records are mastered by digital recorders and a good recording in digital or analog can happen, so you pays your money and you take's your chances.
I wouldn't be too worried - It's only Simply Red after all.
It is a shame that many classic recordings have been remastered to death. Of course this isn't a problem with the digital audio medium, it's whoever instructs the mastering engineers to do this.
Your audio system can differentiate between analogue (vinyl) and digital (CD or HD download). Regardless of the source the master recording was probably digital. So, if CD or HD download sounds inferior this down to a poor-quality DAC. Book a listing session at a good HiFi store, take your CDs and audition a DAC with a separate DSP or FPGA based digital filer. All of the detail you claim to be missing will return. Unfortunately, processing power for this type of DAC makes the design more expensive.
@@user-mu2dp9hg4o The original recording was made in 1977 so I would assume it was analog. But thank you for the comment. Now I know that original recording in analog beats a remaster in digital. Instead of paying five grand or more for a D to A converter I will go online and find a used album in good condition. Now I hope the corporate big wigs can come up with a blue tooth speaker with three 2 inch drivers they can sell for a thousand or more. Some remasters on digital did not take into account the RIAA eq. curve and are horrible.
That sure isn’t the Monty that was Patton’s nemesis! Lol
Neither is he the mysterious cowboy or the second guardian in the reptile room.
I'm wondering whether I was one of the people who suggested Monty's video to you. I know I suggested it to someone in some comment section or forum.
As for what I think of the analogy, the audio may be sliced and diced, but the apple sauce used to fill in the gaps is really good. The MQA people think their apple sauce filler is better, but I think they added too much sugar.
I saw his video years ago and then forgot about it. Fortunately a few commenters reminded me. It's a good video - clear and concise. Whether audiophiles will like it is an open question. Whether apple crumble works better with a bit of rhubarb added is another.
@@AudioMasterclassI vote for no rhubarb
@@AudioMasterclass It's one or the other for me, some things don't mix.
Why no link directly to CZcams video? czcams.com/video/UqiBJbREUgU/video.html
Because that's not the original. Monty's video is licensed under Creative Commons and several channels use it. No problem with that but I prefer the original source.
As a pure Audiophile i have my cake and eat it, don't care how many bits or crumbs. 😀
Enjoy your apple crumble
Yes analog to digital conversion does indeed slice and dice an analog signal.
An Analog To Digital Converter (ADC) samples the analog signal from a microphone (or other transducer). It is essentially a recording voltmeter that gathers the voltage of the signal as a binary number at a fixed sampling rate.
This results in a whole slew of measurements recorded a whole mess of times per second.
At 44,100sps (samples per second) that is one sample every 22 microseconds.
At 16 bit resolution for a 2 volt signal each step in the measurement is 30 microvolts.
Whatever happens between those samples is indeed lost... but it would also be so far beyond our hearing we'd never notice it... 44khz or higher. Not even a bat could hear it, even if it was there.
Whatever happens to the levels between samples is on such a small resolution, less than 1/960th of a decibel, that nobody could ever hear it.
The samples are stored in a computer file as a sequence of binary numbers, where they are no different than any other computer data. Stored, manipulated, and communicated just like any other computer data, with bit perfect resolution.
The big secret is in the reconstruction filter on playback. Here the gap between samples and the differences in amplitude are smoothed by "reconstruction filters" that give back a nearly exact version of the original analog signal.
Very simply ... it's so close as makes no difference... and with higher bit size and sample rates it gets even closer.
I spotted the word 'nearly' in your comment.
@@AudioMasterclass
Until it captures things like RF noise and power spikes, it will be "nearly" ... but if you think about it, that's actually a good thing.
What a load of bullocks. The filter perfectly reconstructs the Fourier response of the transfer function to complete precision.
It's up in the air what sampling frequency is sufficient (48k is fine to 23,999Hz) or what word length (referred to as bit depth) 16 bit gives 96dB s/n ratio. We use 24 bit in the studio because each 6dB of attenuation drops one digit from the word length, hence a -48dB signal referenced to Zero uses 16 bits of word length, still 96dB s/n
It only takes 2 data points to construct a perfect sine wave or an absolute representation of the transfer function after the Nyquist filtering which is exactly half of the sampling frequency, which is passed as an analogue signal of full bandwidth, DC to half the sampling frequency.
It cannot be argued.
Cannot be argued? I would argue with that but since your comment addresses precisely nothing that I said in the video I don't think I'll bother.
2 data points can reconstruct a perfect sine wave .... but that won't even get close to reconstructing a complex harmonic structure in music. Music, contrary to many depictions is not "just a bunch of sine waves" ... it is a complex mess of harmonics and sidetones, virtually none of which even vaguely resemble sine waves.
@@Douglas_Blake_579 What do you mean "none of which even vaguely resemble sine waves"? All signals can be decomposed uniquely into sine waves. That's what Fourier taught us: en.wikipedia.org/wiki/Fourier_transform
@@nicksmith4507
The answer is in your own rebuttal ... _"all signals can be decomposed"_
Fourier *Transforms* are mathematical abstractions that attempt to break a signal down and display it's frequency components. They are not the signal itself.
Serious question ... Have you ever actually used an oscilloscope?
I can guess from your response you have never seen a raw audio signal displayed as it goes down the wire.
It's a basic fact of physics that any circuit segment can have only a single voltage present at any given instant of time. This includes everything from the pins of transistors, the leads of a resistor or capacitor, to the output of a DAC chip or amplifier, even the wire in the cables we use to connect things together.
This means that an audio signal is actually a single voltage that varies over time. There are no sine waves, no frequencies, no separate elements ... there is only one voltage bouncing around like crazy.
Audio signals are down right messy things. If you track this voltage over time with an oscilloscope, you will see a waveform that bears absolutely no resemblance to those nice tidy sine waves or frequency graphs all these test reports show us.
The truth is that all the way from the microphone that captures it to the speaker that reproduces it you are actually dealing with a single voltage (per channel) that is nothing more than a map of how to move your speaker cones... and by extension your eardrums.
"IS LIFE ANALOG OR DIGITAL? By Freeman Dyson [3.13.01]" -- Edge. That's metaphysics...
Dyson was an idiot.
Lol, very entertaining video, but your analogy doesn't extend very well to the d2a portion of this signal processing. You mention Nyquist, but that is the critical piece as to why no information is lost at the appropriate sample rate. If you accept Nyquist, and it is universally accepted from a physics and engineering basis, then the real debate becomes what sample rate is needed to fully capture all the frequencies that audibly effect the final sound. The front end a2d "slicing and dicing" is not in question and the question about losing information from this dissection is only relevant if the entire data flow from a2d to d2a is considered.
16/44.1 is all that is necessary.
@@simonzinc-trumpetharris852 True, if you accept that anything above 20KHz (or 22KHz) is inaudible, but that is not universally accepted. Anti-aliasing is also an issue if there is frequency content above the nyquist frequency. Low pass filtering and higher bit rates then come into play.
@@Esprits4s Whether or not anything above 20kHz is audible, it mostly didn't get recorded anyway on analogue recordings and if you analyse everything above 20k from a vinyl record being played back, it's mostly noise and distortion products from the recording and cutting process.
Quite right about anti-aliasing though - if by chance your mic or preamp picked up much sound over 20kHz and you didn't have good anti-aliasing filters in the ADC, that ultrasound would reflect back into the audio spectrum and become audible.
yeah - slice it up
Word up, Aardvark!
I don't think the apple/onion analogy adds anything: sampling is not that hard to understand. Of course sampling is not a problem if done at sufficient rate/depth. The only question is what is that sufficient level: that's an experimental question dependent on content/situation/listener.
the rate/depth of CD is chosen for a reason, it should be suffiecient (for almost everybody)
@@thepuma2012 have they ever found someone for which it was insufficient?
@@JustEnjoyingLife73 not really, I guess. But sometimes I do believe that streaming 24/192 gives just a little bit extra.... but also if that is true (or is it a feeling?) it is something you don t really miss if using 16/44
Rotating motor Vinyl or Tape the motors have a rotation speed just like a digital clock, Technically there is no difference just speed
This would be the cogging effect that some audiophiles can hear. Other audiophiles merely worry about it.
Ngl, Monty needs a makeover.
As long as the crumble is good, aardvark doesn’t care how it’s made. Aaaaaaaaasaaasaaaaaaardvark.
it was in the beginnging
I happen to like my music "sliced and diced" 🙂
You, possibly unwittingly, make an interesting point. VInyl enthusiasts, and cassette, seem to enjoy the problems and limitations of the medium, so why not digital listeners too? [Edit: Inability to hold down the shift key to type a question mark]
@@AudioMasterclass 😛
All CLASS-AB amps
slice the audio into 2 parts, positive and negative, literally 2 halves, amplified via 2 completely differently manufactured devices PNP & NPN (Fet p-channel and N-Channel),
and then combined back together with a ton of negative feedback to make it all work, but the fools never complained about sliced which it IS!, lets complain about digital.....
This is the answer to give to anyone complaining about slices and dices, ask them about their audiophile crapintosh amp $$$$$
It's a good question whether slicing into two is better or worse than slicing into 65,536 or more. I might ask it, but probably not answer it, in a future video.
you're cooking apples???
maybe even in a microwave?? good gods...
So, we should take it then, you've never had Apple Pie?
Sliced and diced is not a correct phrasing. Because it’s literally just a value with two parts, no space occupied, not even in a 2 dimensional space. The information is recreated by giving that digital information to a converter that turns these values into a changing voltage. And because of physics and the limitations of the digital information, there’s no chance for the output to be any different from what the input has been.
"Digital audio is just a value with two parts, no space occupied, not even in a 2 dimensional space." I wonder how many views that would get as a title.
Aardvarks never killed anybody.
Tell that to the ants
For the real purists, as audiophiles and hi-fi enthusiasts behave, there is no such thing as true analogue. The electric current in analog sound equipment is not exactly analogous to the air vibrations of sound. There is distortion and background noise.
Digital sound is not necessarily worse or better than analog sound. If the digital resolution is too low, the digital distortion will be more audible than a good analog system. If the digital resolution is high enough, the distortion is virtually inaudible compared to an analog system.
And for the sake of convenience, let's ignore the fact that two different people never really hear the same thing, not with the same sound equipment and not with the real sound itself. Even more so, the same person does not hear exactly the same sound at different times.
Truly analog is a utopia. It doesn't exist in the real world.
arguably the only TRUE REAL SOUND (tm) is live from completely acoustic sources.
What was the actual purpose of this video?
To prompt you into making an idiotic comment.
@@AudioMasterclass I think a lot of people watch this and wonder what it all means as evidenced by the comment "So? Digital audio isnt good?"
Sounds like he talking about lossy audio.
No, this is purely PCM.
Funny guy, try a cooking channel also!
I rather doubt it, but what I do know is how to poach an egg better than all the nonsense on CZcams.
So? Digital audio isnt good?
Question is, what do YOU think?
The myths about digital recording feed voraciously on ignorance. Always insist someone fully explains their bizarre theses. It will soon get comical.
I often make up a story when asked about a recording of mine in a leading context. Once there is agreement, lots of pseudo scientific justification, I either tell the truth or make up another story. I’m sick of the ignorance and don’t need extra friends.
Perfect? Absolutes are unattainable.
'you have your own science'. That's an oxymoron any way you slice it. :)
Well there's Copenhagen vs. Many-worlds and plenty of other places in science where each can have their own. I don't see why not in audio.
Aren't all capturings of physical data reduced to a finite rate of capture? The physical space is theorized to be infinite in scope and detail/depth... isn't it?
I have very elementary understandings here. But I think a vinyl record and casette tape should have a capture rate.
I don't think they're capturing as rapid as things actually occur, are they?
What about our ears, or another animal's? Do they capture all that light at the same rate it exists?
Wouldn't it be like... if you could capture physical data at its natural rate, your ears or something would vibrate at like an infinite rate of speed?
I am so wildly out of my league here. Let me go hide under a brick because I am clueless. Small ant.
Hmmm, I sense a video coming along about this. My premise would be that the reason digital audio, and analogue too for that matter, works so well is that we have reduced music to so few dimensions, basically the varying amplitude of the waveform over time, and then two channels for stereo. I shall ponder...
Analog doesn’t exist, everything is quantum and discrète.
Ummm ... I don't know what you've been studying, but I think you had best look for more credible sources of information.
Analog signals absolutely do exist, as anyone who's ever used an oscilloscope knows perfectly well. In fact even digital signals are analog, distinguished only by the way we treat the information they convey.
Human hearing is not perfect.
To all audiophiles: can you see the frames in a film? Just 24 FPS are enough to make you believe that they are not different “chunks”; in monitors, generally, anything over 60 Hz is considered as a stable image. I don’t know what the ear frequency is - maybe someone will: in any case, it cannot be more than 70 Hz, the maximal frequency of human brain (gamma waves) - I don’t think it’s possible and, anyway, it would be completely useless.
So, summing up: that despicable CD-Audio, with only 44100 Hz is… 630 times faster than your senses. And yet, some of you can hear it… Are you superhuman?
24 might be the 'film look' but I certainly prefer 60 fps for my CZcams videos and I'm not imagining the difference. Any of my videos that include iPhone footage or effects, I make at 60 fps even though my camera maxes out at 29.97.
You're being evil!
It's OK. The knives have now been taken from me and placed in a locked cabinet.
@@AudioMasterclass You're a tease Mr. Masterclass! Hey, I've seen an analogue signal generator hooked up to a DAC, convert signal to digital, then converted back to analogue and fed to analogue oscilloscopes - the waveforms looked identical - good enough for me. :)
@@davids4610 You won't see much detail in an oscilloscope trace, either in terms of distortion or noise. They're useful devices but they have their limitations. But if you did a similar test just using your ears, comparing input to output, most people won't hear any difference and that includes me.
@@AudioMasterclass Thanks my Master Class Guru!
sharpen that knife, please.
If you were a little more clever than you think you are you would see that sharpness is not the problem.
Dear David, with all due respect by sticking to this stupid "slice and dice" analogy and trying to visualize it by cutting apples you simply proved that you lack the proper understanding the basics of the discreetly sampled and quantized audio reproduction systems and the underlying physical and mathematical law of nature, described by the sampling theorem. When a signal is SAMPLED it is NOT SLICED to pieces of equal though infinitesimally small width but simply MEASURED regularly. It is the mathematically proved feature of the universe we are living in that when we measure the amplitude of a continously changing phenomenon at a certain frequency we'll have all the necessary information to reconstruct the original signal though bandwidth limited to the half of the sampling frequency. No information is required for this between samples, unless we are happy this limitation of bandwidth. If we record the measurement values in a digital system of a certain "resolution" (number of levels) defined by the number of bits we introduce a certain level of so called "quantizing" noise due to the inherent roundings of the values. Even in a bread and butter 16 bit system this kind of noise is way below of anything you could achive by the state-of-the-art analogue technology. Sad to see that any professional who is acting as influencer for such a big community could make such a misleading video, which could turn into "reference information" of audiophiles to refer later on.
Of course you could ask what is the difference between a "slice" and a "sample" or "measurement". Well while on the surface they look the same but there is a very important difference. The "slice" model implies pieces which include information about the signal's behaviour in time-domain, suggesting that it contains information about the timbre or other features, therefore in digital systems we'll drop a huge portion of the musical information by dumping most of. The truth couldn't be further as the samples contain no more information than the actual amplitude at the moment of measurement, which is sufficient for the recontruction.
If you watch my video again you won't find anything that's factually wrong. What you're complaining about is stuff you've made up in your own head. My only regret in this video is that it would have been a damned sight easier to make my crumble if I'd peeled my apple first.
I enjoy your videos but your sarcastic chortles are grating imho
Grated digital audio would be quite another thing than sliced and diced.
Report all you like. It won't make any difference to me or what I do in my channel.