When did studios begin 24bit recording ?

[Robin L]

Another way to look at bit-depth for those heads that are spinning on "dynamic range" vs. "resolution"... it's kind of two ways of saying the same thing when it comes to digital.

Take an analog signal and digitize it. For now don't worry about how many bits are used. You only have so many "steps" to work with as you create samples, so there will be a random difference (error) between the original value and the digitized value, for each sample. You take the number that is closest to the real value, but the analog signal will rarely land exactly on the equivalent digital value.

You can now think of the digitized signal as the original analog value, plus a random error signal.

That should sound a bit familiar ... a constantly varying random signal? We call this noise. And in fact, that's what you get when you digitize something. And that's actually how it sounds when you listen to a 44.1khz+ signal but at a very low bit depth - just a really noisy recording.

If you only have a few steps, you get a ton of noise because you can't approximate the original signal very well. But what's really interesting is the original audio is pretty recognizable even if you only have like, 2-3 bits. 8 bits sounds a lot better than you'd expect and 12 bits is a little alarmingly good compared to what you'd guess.

I've got Denon 33CO-1586, a CD issued in 1987 but recorded in 1975 using an early 13-bit system. It's a recording of Beethoven's "Archduke" trio for piano, violin and cello. The sound is just fine.

https://www.vintagedigital.com.au/milestone/fully-digital-pcm-recording/

 

[Ropeburn]

Well uh. Guys. Of course you're right.

I was thinking arsebackwards all this time. All those dynamic ranges in my head, 96 vs 144dB and stuff... that's not the cause of bitdepths, but the result of them. And a merely useful scale for logical purposes - the real physical signal has just a voltage range that's always the same, for comparison anyway. Of course then that range gets quantised in a number of steps determined by bitdepth.

Now that was one hell of a brainfart. I take back everything I said above and claim the opposite. Thanks for y'all's persistence and patience.

 

[voodooless]

@Ropeburn , next time visualize something like this:

 

[Ropeburn]

@Ropeburn , next time visualize something like this:
View attachment 515924

It's kinda embarrassing. I know how this stuff works! Just got stuck in a weird mental loop there. Haha!

 

[Matt S]

It's kinda embarrassing. I know how this stuff works! Just got stuck in a weird mental loop there. Haha!

We all have brainfarts from time to time. All good in the end eh!

 

[voodooless]

It's kinda embarrassing. I know how this stuff works! Just got stuck in a weird mental loop there. Haha!

It happens to everyone at some point, don’t worry about it..

 

[sergeauckland]

Well uh. Guys. Of course you're right.

I was thinking arsebackwards all this time. All those dynamic ranges in my head, 96 vs 144dB and stuff... that's not the cause of bitdepths, but the result of them. And a merely useful scale for logical purposes - the real physical signal has just a voltage range that's always the same, for comparison anyway. Of course then that range gets quantised in a number of steps determined by bitdepth.

Now that was one hell of a brainfart. I take back everything I said above and claim the opposite. Thanks for y'all's persistence and patience.

You're not the only one. I too thought it the wrong way round. I was confused by looking at the AES-EBU or S-PDIF data format that has 24 bits available for audio data, but for 16 bit streams, the last 8 bits are zero, 20 bit streams have four more active bits, with only the last four bits at zero.

S

 

[Ropeburn]

There must be a word for this kind of fallacy.

You start with a wrong premise: oh look, here's the 24bit dynamic range of 144dB, and 16bit is 96dB... therefore the upper 16 bit of 24bit data have the same resolution as 16bit data, the "rest" of the resolution, the lower 8 bit, are resolving anything below -96dB.

Then you take that premise and run with it, applying logic - any thought that follows is logically correct so you think you're right. Of course you aren't because the original premise is wrong.

 

[DVDdoug]

here's the 24bit dynamic range of 144dB, and 16bit is 96dB... therefore the upper 16 bit of 24bit data have the same resolution as 16bit data, the "rest" of the resolution, the lower 8 bit, are resolving anything below -96dB.

That's basically true. The 16 most significant bits are the same but the other bits usually aren't zero.

I can write $1234 but the real balance in my bank account might be $1234.10.

 

[ZolaIII]

98/99 on consumer side 02/03 scratching 18 bit margins and getting to 19 in 04 regarding DAC's and still late regarding ADC's. Talking about DAW's mostly.

 

[krabapple]

Only if you lower the volume. With integer formats there is no headroom over 0dB. ...You could say there is more "footroom"

In addition to more headroom for recording, the benefit of working at higher bit depths when there's a lot of digital processing steps in the production chain, is that it keeps quantization errors inaudible.

That benefit is part of the final product, baked in. Meanwhile, 24 bits at the consumer playback end isn't necessary, but it is common, just 'because'.

tl;dr: CD rate was/is a very good standard for home playback. The cases where it is audibly inadequate there are, let's be very generous, rare at best.

 

[danadam]

There is a subtle detail that may be not so obvious (or maybe it wasn't only to me )

When we say that we have this or that many steps available, it may seem that we lose the information about the signal level up to the nearest quantization level. That is of course true when we talk about a single sample. We get, for example, either 32767 or 32766. But the signal is a series of samples and that gives us averages. Even without dither different samples will have different amount of error and it will average out to some extent. Well, most of the time, I guess. Higher frequencies that are integer divisors of the sampling frequency won't have that many different samples.

Here are 9 versions of 1 second, 100 Hz tone in 16-bits, 44.1k, without dither. The amplitude of each one is 32 quantization steps lower than the previous one. They were then converted to 8-bits (just rounding, without dither). The max positive/negative levels in the files are:

            max +/- 16-bit         max +/- 8-bit
         dec           hex         dec       hex
32639 -32639 0x7F7F 0x8081    127 -127 0x7F 0x81 - 0
32607 -32607 0x7F5F 0x80A1    127 -127 0x7F 0x81 - 32
32575 -32575 0x7F3F 0x80C1    127 -127 0x7F 0x81 - 64
32543 -32543 0x7F1F 0x80E1    127 -127 0x7F 0x81 - 96
32511 -32511 0x7EFF 0x8101    127 -127 0x7F 0x81 - 128
32479 -32479 0x7EDF 0x8121    127 -127 0x7F 0x81 - 160
32447 -32447 0x7EBF 0x8141    127 -127 0x7F 0x81 - 192
32415 -32415 0x7E9F 0x8161    127 -127 0x7F 0x81 - 224
32383 -32383 0x7E7F 0x8181    126 -126 0x7E 0x82 - 256

FFT (44100 points, rectangular window) shows that the 100 Hz level of the 8-bit version doesn't diverge too much from 16-bit version, certainly not by 20*log(126.5/127) = 0.03 dB. The tics in the zoomed-in view are every 0.002 dB.
, , , , , , , ,

And here are the samples at the peak:
, , , , , , , ,

 

[valerianf]

In the 80's many studio recorders were coming from Sony.
The 1984:Sony PCM-3324 digital DASH 24-track tape recorder could be one of the first 24 bits studio recorder.

 

[Mnyb]

Interesting question bussiness wide adoption would likely happened when DAW software was wide spread and in use ?
So when did most DAW suites use for example 24/48 which seems common ? Before year 2000 ?

Specialised equipment for the high end studios seems to have existed very early as pointed out by other posters in this tread .

 

[ZolaIII]

Interesting question bussiness wide adoption would likely happened when DAW software was wide spread and in use ?
So when did most DAW suites use for example 24/48 which seems common ? Before year 2000 ?

Specialised equipment for the high end studios seems to have existed very early as pointed out by other posters in this tread .

Early 90's (Cool Edit) but they also begin to get suport and mature when the actual hardware came. Adobe Audition (which is reworked and expand Cool Edit after Adobe bought it) correlate with first really full consumer space capable true 24 bit PCI card (Creative Audigy 2) 2003. Had ZS Platinum Pro and that's the DAW I used most of my life.

 

[Ropeburn]

In the 80's many studio recorders were coming from Sony.
The 1984:Sony PCM-3324 digital DASH 24-track tape recorder could be one of the first 24 bits studio recorder.

That's 24 track, not 24bit. Big difference. It recorded in 16bit.

 

[Eric Natural]

So many of my recordings from the 1980s all read at least -3db from anything late-90s to current releases. Perhaps those engineers gritting their teeth to avoid catastrophic overs w/that vintage's digital tape.

 

[Eric Natural]

>1997: Pro Tools|24 was released, offering 24-bit, 24-track capability

 

[sergeauckland]

So many of my recordings from the 1980s all read at least -3db from anything late-90s to current releases. Perhaps those engineers gritting their teeth to avoid catastrophic overs w/that vintage's digital tape.

It was always considered good practice to leave a decent amount of headroom. The BBC (bless them!) still leave around 10dB headroom on their digital feeds as they still seem to maintain 0dBFS = +18dBu and peak to +8dBu (PPM6), although using digital mixing desks, this is more notional than actual analogue levels.

Many of the CDs I have from the early 1980s have decent amounts of headroom, and no brick wall limiting (Dire Straits and Moody Blues are good examples of this) but it all went to rat-shit by the mid 1990s at least for Pop/Rock, less so for Jazz and classical.

It's always nice to see the meters on my main system staying well under 0dBFS, but equally distressing when they go to 0 and stay there, as so many later recordings do.

S.

 

[Pluto]

I'm curious about digital studio technique and when did "high-def" begin - maybe call it second generation post 16 bit ?
What year did 24 bit recordings get made ?

The 24bit format, as a wrapper, has been there more or less from the beginning of the digital era. Perhaps the question you really ought to be asking is the stage at which 24bit A to D converters started to become widely available or, more precisely, when they started producing useful accuracy in the few least significant bits. The answer to this question is inextricably bound up with the DSD question because in the late 1990s it was quite likely that DSD was capable of better overall performance than an analogue to analogue PCM loop of the time. I say “quite likely” because unless we could examine and test converters of that era without the degradation that a further quarter of a century has wrought upon them, it's difficult to answer that question with sufficient confidence.

By the end of the first decade of the new millennium we had PCM converters that were usefully accurate to bit 20 or below which left DSD in the dust, mainly because of the complicated noise-shaping arrangements necessary to counter the inherent noise problems of DSD. I'm not sure how low today's best A to D converters can go but I would reasonably expect around bit 21 or 22.

Another way of asking the Q is how many of my(common pop & jazz) recordings have been up-sampled ?

It can be quite difficult to determine the recorded bit depth of product delivered to consumers. Regardless of the bit depth of the original recording, the mastering will (or ought to be!) dithered to 24bit. I'm told that a thorough statistical analysis of the few least significant bits can determine the degree of correlation between them; if there is no or low correlation it is likely that they represent little more than random noise. The higher the correlation, the more likely it becomes that they bear a relation to the music content but it is nonetheless very difficult to make sense of data that refers to content some –130dB below peak.

The main value of 24bit technology lies in the headroom it provides when recording the unexpected. Throughout the analogue era the real skill was keeping the modulation sufficiently clear of the noise while not getting caught out when the tympani suddenly struck a triple fortissimo. The arrival of 16bit digital eased this problem by 10dB, give or take. But with 24bit, there is no excuse. You set levels so that the absolute loudest expected sound hits no more than around -15dBFS, which gives you all the latitude you need to adjust in post. In an unrehearsed situation, you can increase this to 20dBFS or more, if necessary.

If I had this luxury when I was working full time, I probably wouldn't have the high blood pressure I have today but I still maintain that 16bit formats are sufficient for consumer delivery when correctly mastered.