Hacker News Reader

> But what about all the hi-def audio stuff you see these days? The good news and bad news is that it's indistinguishable from good old 44.1khz/16bit audio like you'd find on a CD. Those engineers were very smart; 44/16 audio more than covers everything the human ear can hear.

Yes, but more channels are better. The spatial resolution of human hearing is pretty impressive, and stereo+headphones doesn't really do it justice.

You only need two channels to provide complete spatial information through headphones if it's a binaural recording.

https://www.youtube.com/watch?v=Yd5i7TlpzCk

Dolby Atmos is the latest multichannel audio format; rather than mixing down to a fixed 5.1 or 7.1 master, an Atmos master carries spatial metadata that can be decoded and rendered to any array of loudspeakers, from a binaural mix for headphones to a 24.1.10 home theater setup. Windows and Xbox One both support Dolby Atmos support; Windows also has a native spatial post-processor called Windows Sonic.

https://www.dolby.com/us/en/professional/content-creation/do...

https://msdn.microsoft.com/en-us/library/windows/desktop/mt8...

> You only need two channels to provide complete spatial information through headphones if it's a binaural recording.

I don't really believe this, given the shape of the human ear and my own experience trying to reproduce 3d sound. Humans can judge direction and distance remarkably well up to a few dozen meters.

Parent mentioned a binaural recording.

Those needs specialized equipment to be recorded -- and can indeed reproduce the 3d spatiality of the sound (sic).

In fact there are 3d sound processors that can do very impressive things even without headphones (e.g. some from Roland).

Here’s the funny thing: humans can discern vertical positioning as well as horizontal positioning. Binaural recordings can’t reproduce that, since it’s highly dependent on the shape of our ears.

I have a friend into binaural recording, he has a rig shaped liked a head with 3d-printed ears. He wanders around the woods recording birdsongs with it. I would imagine that provides a fair degree of verisimilitude.

It would be interesting to check, indeed. Especially since ear shape is highly individual.

Any chance at a soundcloud link?

Yes, but the way our nervous system does this is by two mechanisms:

1) The inter-aural time difference. Sound takes time to travel, so any sound coming from left of the head will arrive at the left ear sooner than the right ear.

2) The head-related transfer function. The shape and auditory characteristics of the head and ear canal change the sound slightly depending on the frequency and the direction of the incoming sound. Since this direction is different for the two ears, your auditory system can compare the 'shape' of the sound it gets from each ear and use that to position the sound in space.

The inter-aural time difference is, by far, the more effective of the two, but it only gives horizontal bearing. It also can't give 'in front' or 'behind'. The head-related transfer function can give both horizontal and vertical bearing, but it's not that accurate. Together, these explain why humans are far more accurate at localizing sounds horizontally than vertically.

Binaural recordings try to capture both of these mechanisms through stereo recording with a mock head. That's why listening to a binaural recording gives you the feeling of sounds coming from a point in space, because it's replicating the same experience you'd get from actually being in that space.

Try listening to aphex twin, maybe selected ambient works 2 or analog bubblebath 5

The whole point of binaural recording is to exploit the shape of the human ear to encode spatial information.

A binaural recording setup uses an artificial head with artificial ears, with the microphone capsules located where the eardrum would be. The microphones capture the phase and magnitude effects of the head and outer ear. Unless you have really weird ears or really bad headphones, the effect is extraordinarily lifelike.

https://en.wikipedia.org/wiki/Binaural_recording

https://www.youtube.com/watch?v=kTkVnb3TXO8

It's also possible to process sounds digitally to create a binaural effect using convolution. Using a mathematical model of the ear and head (a head-related transfer function), you can position sounds in a 3D space around the listener. Many games offer HRTF audio and it provides remarkably accurate spatial information. HRTF audio is a key aspect of the immersiveness of VR experiences. It's possible to generate a unique HRTF based on your own anatomy, using either an audio capture from a microphone placed inside the ear or a 3D scan of the user's head.

https://en.wikipedia.org/wiki/Head-related_transfer_function

https://www.youtube.com/watch?v=_dxIK3-vqpc

https://www.youtube.com/watch?v=cMDVJWExcxY

> A binaural recording setup uses an artificial head with artificial ears

It can also use a real head with real ears and e.g. the Roland CS-10EM binaural headphones (which I've had reasonable success with.)

https://soundcloud.com/zimpenfish/amarcade1

That's pretty darned spatial through headphones (recorded with Roland CS-10EM binaural microphones.)

I did a thing as a kid that basically proves it out.

I recorded my home, an afternoon with the family from the speakers that it would be played back from. After some fiddling with a pre amp to get the recording balanced well enough to play back, the effect was pretty effective.

I didn't expect it, but when that recording was played on that system and speakers, it was possible to sit in my room and basically hear the house! TV in the front room, dog walking around, doors closing.

Brought people in late at night to play it. Was spooky to hear midday sounds, and believe them. The illusion was that good.

To be clear, that experiment reproduced the sounds of home near perfectly. Up, down, left, right, front, back.

Two channels.

>You only need two channels to provide complete spatial information through headphones if it's a binaural recording.

That only works for headphones though...

DIY speakers, if you get the basics of speaker design right, can be really good. It's because if you build them yourself you can build an enclosure as big as you like, as thick and heavy as you like, with enough dampening as you like and as long as you balance the drivers and crossover right they can sound fantastic and for not much money. Also amplifiers since the 1990s have been great, they just need a service with some fresh capacitors and some hack (e.g. a smart plug) to deal with the fact that they suck power when on/standby but not in use, so getting a good one second hand is worth it. That said IC amps have improved a lot too and kits with these are also now easier since getting a clean power supply is not as difficult as it used to be.

A word of caution. I looked into this and it seems extraordinary complex to design your own speakers. Even build them seems tricky.

Here is a German shop that sells kits: https://acoustic-design-online.de/de/

They had once kits to build your own electrostatic speaker (warning, high voltage, danger!): http://silberstatic.de/

IMHO the Acoustic Design ones look like DYI, the C-Note doesn't.

48khz please! Not because of the extra frequency, but because of compatibility with video :)

And RCA? What's wrong with balanced audio on an XLR?

I saw a lovely post on Facebook about how a freelancer had brought in something odd, that could only get the audio out of via analog.

He proudly came out with his crappy £50 RCA cables. We had to tell him that we only dealt with balanced audio. My friend went to the stores and found the most beat up converter she could find, and some ropey looking cables :)

>And RCA? What's wrong with balanced audio on an XLR?

There's nothing wrong with it, it's just totally unnecessary for line level signals over short cable runs in domestic environments. Balanced cabling doesn't sound any better, it just has better EMI rejection.

Sure, and you never get EMI sources in a domestic environment, say mobile phones.

If you're building a system from scratch, you may as well use balanced audio, the driver costs under a fiver.

> Balanced cabling doesn't sound any better, it just has better EMI rejection.

Unfortunately for me, when you live downtown near a radio station that leaks into all your audio equipment, better EMI rejection definitely sounds better.

Reminds me of dropping off a composite/svideo video mixer as a favour loan to someone running an event and their first question being "so how do I plug my Macbook pro into this?"

Is £50 a typo?

I agree with you on 16/44.1 audio, I have a transparent DAC that can output 24/96, and on repeated blind tests I have failed to identify which is which, using files from the same source (a 24/96 song converted to 16/44.1). For audio production and mixing 24/96 makes a lot of sense though, because of dynamic range.

> For audio production and mixing 24/96 makes a lot of sense though, because of dynamic range.

Obviously only the bit-rate matters for dynamic range, and a 16 bit signal has 96 dB of dynamic range. That is more than enough for even the most dynamic of audio signals.

A larger bit rate is useful for lots of digital summing as I mention here [1] and I assume what you allude to, but for most home applications nobody needs 16bit+ for anything other than improving their noise floor (which is still borderline inaudible at 16bit).

[1] https://news.ycombinator.com/item?id=16995020

> A larger bit rate is useful for lots of digital summing as I mention here [1] and I assume what you allude to, but for most home applications nobody

Yeah that's why they said audio production, and not home use. 96db of dynamic range sounds like a lot, but not when you're stacking 45 tracks. Also compression and saturation later in the chain will further bring up that noise floor.

And you don't get that full dynamic range because you want to prevent clipping in a recording, so you give yourself something like 12db of headroom away from the maximum, and now you've lost a good chunk of your dynamic range. 24 bit recording lets you keep plenty of maximum headroom while also staying far from your noise floor.

> Yeah that's why they said audio production, and not home use.

Yes, but the comment was specifically about dynamic range in audio production, not about higher bit rate's prime benefit of reducing artefacts in digital summing. The noise floor at 16 bit is -96db, if you're down there, you're doing it wrong.

> 96db of dynamic range sounds like a lot, but not when you're stacking 45 tracks. Also compression and saturation later in the chain will further bring up that noise floor.

It's more than enough if you're doing proper gain staging - the number of tracks is irrelevant.

Any mastering compression of more than 3db gain reduction is mostly excessive, but let's say 6db. So a final master will bring up a noise floor 18db at most (if gain staging was done properly and as you say you left plenty of headroom for the mastering engineer - although clearly you wouldn't be able to use all the headroom). That's still a noise floor of less than -78db, but more realistically around -85db.

Saturation usually only add harmonics not gain (obviously depends on the kit you're using).

> And you don't get that full dynamic range because you want to prevent clipping in a recording, so you give yourself something like 12db of headroom away from the maximum, and now you've lost a good chunk of your dynamic range. 24 bit recording lets you keep plenty of maximum headroom while also staying far from your noise floor.

Myth. Watch this [1] and you'll realise that there's plenty of range with 16 bit. People were producing with 16 bit for years perfectly fine. Before that high end studio tape machines (like the classic Studers or Telefunkens) were equivalent to about 14-16 bits (based on their noise floor) and we had decades of music produced and recorded on that format.

I am not saying that 24 bit isn't better than 16 bit. Of course it is, there are tangible benefits in summing and it gives you more headroom to work with. But, I'm saying it's a myth that there's not enough dynamic range or that you're somehow just bumping along above the noise floor. If you are then you're doing it wrong.

[1] https://www.youtube.com/watch?v=cIQ9IXSUzuM

> Yes, but the comment was specifically about dynamic range in audio production, not about higher bit rate's prime benefit of reducing artefacts in digital summing.

But audio production often involves lots and lots of digital summing. I don't see how that's not related.

> Any mastering compression of more than 3db gain reduction is mostly excessive, but let's say 6db. So a final master will bring up a noise floor 18db at most (if gain staging was done properly and as you say you left plenty of headroom for the mastering engineer - although clearly you wouldn't be able to use all the headroom). That's still a noise floor of less than -78db, but more realistically around -85db.

That's only counting mastering compression.

What about NY style parallel buss compression on drums where you want to hit -20db GR as an effect, which will inevitably add some saturation too. 6-10db of compression on an individual drum track is common too if you're doing metal or electronic music. So you've got 10 drum tracks, each with several effects and a parallel track with distortion/compression.

> Saturation usually only add harmonics not gain (obviously depends on the kit you're using)

If it's not adding amplitude to your signal it's almost certainly bringing up the noise floor. Saturation generally works by applying gain until the device/plugin distorts. Just because amplitude isn't going up doesn't mean gain isn't increasing.

> Myth. Watch this [1] and you'll realise that there's plenty of range with 16 bit. People were producing with 16 bit for years perfectly fine. Before that high end studio tape machines (like the classic Studers or Telefunkens) were equivalent to about 14-16 bits (based on their noise floor) and we had decades of music produced and recorded on that format.

This is true for when you're operating on friendly signals like a sine wave signal generator with a fixed amplitude. Not so nice when you're tracking with 20db of headroom and the drummer some how manages a flam 24db f*cking louder than everything else which clips the overheads (this actually happened using 24bit, but now with crazy dynamic drummers I give myself TONS of headroom because there's no reason not to). At least when tape "clips" it doesn't ruin the take.

I do agree with you that 16bit can be good for recording (and for home audio I agree you'll never be able to tell a difference) but I don't believe there are literally zero differences between 24bit, and hard disk space is so cheap that there's no reason not to use the extra bits.

Another much appreciated advantage of 24 bit is that it allows so much extra headroom:- when recording, levels can be set so the peaks hit maybe -18 -- -12dB(FS) which helps eliminate any chance of clipping in unexpected circumstances.

I would love to get into this, and I'm sure I'm not alone.

Any specific beginner-friendly resources that you would recommend?

I thought of starting with something small and defined, like upgrading my TV's sound but even that comes with a variety of bewildering options.

If you are into headphones, you could start hanging out on the subreddit. It's a big community, and you'll see many of the current technologies. It could get really expensive, I mean the new Sennheiser closed-back HD820 has a ~$2k price tag (ouch!), but certainly you could call yourself done with a $400 dollar setup, including DAC and Amplifier.

EDIT: You could also try IEMs, Etymotic makes some of the most clear, coherent, natural and neutral sounding. Their TOTL (Top-Of-The-Line) is the ER4 line which has a MSRP of $350, but you could get the same value with the ER3 line at $179.

The first thing to do if you want the best quality sound is to sort out the acoustics in the room you're in. Do that before buying or building any high end gear, otherwise you're just throwing money away. Shit speakers in a well treated room will sound better than the most expensive ones in a poorly treated one.

Get yourself a reference/measurement microphone so you can measure the room, then use REM [1] to analyse the results (it's free). The Behringer ECM8000 is fine ($50), you don't need to spend a lot of money on that. You put the reference mic at your ideal listening position (you're not going to make an entire room acoustically perfect without needing a second mortgage, even high end studios aren't).

What you want to get to is a flat response across the frequency range. You won't get it, and you will need bass traps, mid traps, etc. to fix the problems. If you're doing it for your TV then you can get attractive looking traps that don't look ugly. But if you're doing it for music, and you want lots of bass, then you may end up with big bass traps in the corner of the room. Although the PSI Audio AVAAs [2] are another (more expensive) option for trapping bass without taking up any more space than a regular speaker. I trialled a couple of these and they were exceptional.

Depending on how bad the room is you can use room correction systems to fix it. I use a Trinnov ST2 Pro [3] (on top of a treated room) in my studio and the difference is astonishing.

This is my frequency response before and after the Trinnov [4]

Once you have got your room acoustics to a pretty good place, then the world's your oyster in terms of audio gear and speakers. I would advise against getting trapped in the super expensive audiophile world, you won't get much difference once you're above the $10k speakers range, and actually you can get results which will blow your mind with $1k speakers. Building your own is tough, because the dimensions are so critical, but if you're up for the challenge, it could be worth it!

I've spent a large amount of time in this world because I've built a high-end music studio in my house. And going much beyond this advice isn't worth it. You need to go and listen for yourself and work out what you like. But acoustics will always win over gear, always.

[1] https://www.roomeqwizard.com/

[2] http://www.emerginguk.com/?portfolio=psi-audio-avaa-c20

[3] https://www.youtube.com/watch?v=dBpEpvmo-GA

[4] https://imgur.com/a/LTXinuL

Unfortunately ~100$ shipping to Germany. :-/

> But what about all the hi-def audio stuff you see these days? The good news and bad news is that it's indistinguishable from good old 44.1khz/16bit audio like you'd find on a CD.

Not true at all..

Just because 44/16 well covers the frequency and amplitude domain of human hearing, it doesn't mean that the representation of the waveforms within that domain are precise. Other digital systems like 96/24 and variable encodings are much better at this because of the higher sampling rate and precision, and have much less error due to quantizing jitter.

Anyone with a decent set of ears and a bit of patience/training can learn to tell the difference in quality, and once recognized, the difference is quite clear.

I will agree that although clear, the difference is still subtle, and 44/16 is generally pretty good.

> Not true at all..

Not true at all.. the bit rate will only affect the noise floor, you'd be hard pressed to notice any difference between 16 and 24 bits (you won't). And when it comes to the Hz - (width of the time-slice, 44 / 96) that will only make any difference on super high frequency sounds, but again, even with that the vast majority of people will perceive no difference. Band limited sound waves can only solve to one solution, and making the slices smaller in the time domain will have relatively little effect because you can perfectly represent all signals in the range of human hearing at 44khz sample rate.

Higher bit rates are useful for people using DAWs like Cubase, ProTools, etc. where your 'signal' is going through myriad processing and summing stages because of the use of multiple tracks, plugins, etc. (sometimes hundreds) That's when you don't want the artifacting from summing multiple approximations. I use 32bit / 96khz in Cubase for this reason.

The human brain is very good at convincing itself [3] that it can tell the difference, but in a double blind test you'd have no chance.

Useful info and reference:

[1] https://www.youtube.com/watch?v=cIQ9IXSUzuM

[2] https://en.wikipedia.org/wiki/Nyquist%E2%80%93Shannon_sampli...

[3] http://www.image-line.com/support/FLHelp/html/app_audio.htm

How about percussive sounds, such as a cymbal crash? These show very high frequencies in a Fourier analysis. Wouldn't that invalidate the application of the Nyquist-Shannon theorem?

It's worth watching the video in my references [1] - it's a proper hacker's view of digital audio and how it translates to the analogue domain. Even with a 20khz sound you're not going to get any aliasing at 44khz sample rate. Most people's hearing starts to fall off significantly after around 12khz or so (depending on age). Most cymbals have a fundamental frequency around 2-5khz in my experience.

Also, because cymbal sounds are pretty close to being noise, it's arguable whether it would matter or not. Although if you listen to a 128 kbps MP3 then you can often hear the top end frequency sounds (like cymbals) becoming 'mushy'. To the vast majority of people once you get to 192 kbps it's inaudible, and then above that most people with 'perfect ears' would struggle. And that's with a lossy format, never mind the lossless formats of CDs, WAV or FLAC which is the discussion at hand.

[1] https://www.youtube.com/watch?v=cIQ9IXSUzuM

The ADC (Analog-digital-converter) will always have an anti-aliasing lowpass filter to remove any frequencies above the nyquist and prevent aliasing in the recording. Sometimes it's analog, nowadays it's often a 2-stage system with oversampling and a digital anti-aliasing filter. But the result that goes into the recorder is the same, any frequencies above the Nyquist (which would become aliased and mirrored) are removed or reduced to inaudible levels before being finally digitized.

So while cymbals do have lots of high frequency content, in a 44.1khz recording any frequencies at or above 22.05khz should be removed first. (In reality it's not a brickwall filter at that frequency, but tends to steeply roll off once it gets nearer).

Unless I'm misunderstanding what you're asking.

So you say 'not true at all' to my repudiation, and then go on to say:

"you'd be hard pressed" and "the vast majority will percieve no difference"

which is it?

bit rate: yes, as sounds are at the lower end of amplitude within the bit depth, the more 'stepping'/'zippering' occurs to the input waveform since less bits are available to represent the shape of the wave at that amplitude, and so things become more distorted, and the signal noise goes up. similarly, for higher amplitude sounds, the less need for compressing/limiting to remove clipping artifacts, so again, less distortion (though the latter isn't inherent to the signal encoding per-se).

sampling frequency: irrespective of the nyquist theorem, the shape of the wave is different. even if you can't explicitly hear at a base frequency of the overtones above a certain frequency, the waveform is not the same and your ear is not being moved in the same way, period, and relative distortion of overtones with respect to the base frequencies within the nyquist spectrum increases with frequency, so the tonal effect of the sampling frequency on the sound (and net distortion) is diminished as the sampling frequency increases. aka overtones become increasingly distorted to a lesser extent as the sampling frequency increases; banded spectrum (e.g. FFT display) is much cleaner, as a visual example.

both of these can be masked somewhat with dithering, but again, this is introducing noise into the signal, which you can also hear; all other things equal (which they arent), the dither free signal will inherently have less noise and so be cleaner sounding.

I agree that bit rate is harder to perceive, and for most things, this is more important w/r/t multiple generations of distortion introduced by software.. but just because something is 'hard' to percieve doesn't mean it is imperceptible and therefore a net zero, which is what the original post is claiming. whether a net 0.0001 is relavent is a different matter (as I said previously, 16/44 is generally fine, so I'm not in the 'I can totally hear the difference from my monster cable digital edition' camp), but it is not a net 0, and when listening for tambre rather than frequency, quite clear once you know what to listen for.

also yes:

I did a blind listening test for this while in an audio engineering program. Was able to clearly tell 44 vs 96 and vs 192, as well as 16 vs 20 and 24 (though 24 vs 20 was pretty difficult). That was a while ago so not sure if my ears are still as sharp.