When did studios begin 24bit recording ?

[Eric Natural]

super helpful notes guys, thx

 

[Dan Lavry]

super helpful notes guys, thx

Digital electronics evolved around the concept of a byte (8 bits). Hardware (comunication, memories...) were oriented that way.
16 bits required 2 byte, but at 17 or more bits, you have 3 bytes for 24 bits.
But even a 20bit converter has 4 bits unused. It is not a 24 bits converter.
There is no real 24 bits converter. The last bits can be zero, noise or garbage.
How many effective bits? Take the dynamic range (no A Weighting) and divide by 6. Example: CD offers 16 bits so 6X16= 96db dynamic range. It is a close aproximation.

 

[danadam]

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.

Wasn't delta-sigma architecture becoming more popular in general, regardless of what format (PCM or DSD) was delivered to the end-user (the computer)? I thought that DSD was actually the idea to store and distribute the output of the DS modulator directly, instead of converting it to PCM. And it appeared a few years after DS converters were available. The Audio DAC and ADC History (by @AnalogSteph AFAIK) shows DS chips from 1993 and 1996, while DSD/SACD appeared only in 1999. AFAIK most (if not all) contemporary ADCs are still DS.

On that page there's also a quote from an article about Deutsche Grammophon Recording Centre from March 1995:

4D goes to third generation

The Deutsche Grammophon Recording Centre has developed a third generation upgrade of the Stage Box system central to the 4D recording chain. All recordings made by the Recording Centre since October 1994 have used the new DG AD III technology, whose convertors feature the new Crystal CS5390 delta-sigma 20-bit A-D convertor ICs to provide 23-bit digital-floating delta-sigma A-D conversion.
...

 

[Ropeburn]

Digital electronics evolved around the concept of a byte (8 bits). Hardware (comunication, memories...) were oriented that way.
16 bits required 2 byte, but at 17 or more bits, you have 3 bytes for 24 bits.
But even a 20bit converter has 4 bits unused. It is not a 24 bits converter.
There is no real 24 bits converter. The last bits can be zero, noise or garbage.
How many effective bits? Take the dynamic range (no A Weighting) and divide by 6. Example: CD offers 16 bits so 6X16= 96db dynamic range. It is a close aproximation.

I'd argue, if you can process 12bit audio on an 8bit system, I'd argue you can also process 24bit on 16bit systems. It just wasn't necessary, because when 24bit became a thing, 32bit processors were already standard.

 

[Dan Lavry]

I'd argue, if you can process 12bit audio on an 8bit system, I'd argue you can also process 24bit on 16bit systems. It just wasn't necessary, because when 24bit became a thing, 32bit processors were already standard.

I think we are talking about linear pcm digital audio.
I am an EE electronic design engineer of audio converters. I am pointing at hardware, chips and more. The transition to over 16 bits was a big step. The third byte had only 1 out of 8 bits.

 

[Ropeburn]

I think we are talking about linear pcm digital audio.
I am an EE electronic design engineer of audio converters. I am pointing at hardware, chips and more. The transition to over 16 bits was a big step. The third byte had only 1 out of 8 bits.

Yes, we are talking about PCM.

12bit audio absolutely was done on 8bit CPUs, if that's what you questioned. Various sampling instruments in particular, for example the E-Mu SP-12 that was based on the good old Z80.

 

[Dan Lavry]

Yes, we are talking about PCM.

12bit audio absolutely was done on 8bit CPUs, if that's what you questioned. Various sampling instruments in particular, for example the E-Mu SP-12 that was based on the good old Z80.

I guess we are not connecting.
Linear PCM such as AES/EBU for pro audio and SPDIF for consumer audio consist of 64 bit words. 24 bits for ch 1 audio, 24 bits for ch2 audio. The other 16 bits are for information.
If you want to store it in memory, it takes 3 bytes.

 

[Dan Lavry]

I guess we are not connecting.
Linear PCM such as AES/EBU for pro audio and SPDIF for consumer audio consist of 64 bit words. 24 bits for ch 1 audio, 24 bits for ch2 audio. The other 16 bits are for information.
If you want to store it in memory, it takes 3 bytes.

I meant to say each sample takes 3 bytes for each sample. A stereo signal takes 6 bytes.

 

[Ropeburn]

I'm not talking about transmission protocols, but about the processing itself.

You process 12bit data on an 8bit system simply by using two bytes one after another. Likewise, you could do the same with 24bit on a 16bit system.

 

[Dan Lavry]

I'm not talking about transmission protocols, but about the processing itself.

You process 12bit data on an 8bit system simply by using two bytes one after another. Likewise, you could do the same with 24bit on a 16bit system.

The original question was " when did studio started using 24 bits". I was there in the early days. I was the first to break the 120dB range in early 1990. I called the AD122 (122dB dynamic range) a 20 bit converter. The next AES convention people were selling 24 bits converters, most under 18 good bits...

You are taking about processing. You can do all kinds of things with real fast 8 bits, but that was not the question I was answering.

 

[Ropeburn]

Nice for you. I will shut up then.

 

[Dan Lavry]

Nice for you. I will shut up then.

 

[Dan Lavry]

I don't want you to shut up. We were going in different directions. It does not invalidate what you say. We were talking about different things...

 

[Palmer]

For something maybe more enjoyable - anybody remember the Publison Infernal Machine? I saw and heard one of the first ones, end of 1983 beginning 1984, granted - 16 bit, floating... but a harbinger of things to come, back then. Wish I had one, sort of unobtainable, now.

Cheers

 

[AnalogSteph]

The original question was " when did studio started using 24 bits". I was there in the early days. I was the first to break the 120dB range in early 1990. I called the AD122 (122dB dynamic range) a 20 bit converter. The next AES convention people were selling 24 bits converters, most under 18 good bits...

Side note, the original AD-122 seems kind of lost to the sands of time, I can't even find any specs beyond its 122 dB (unwtd) dynamic range. I did find out that they (and the MkII) like to run hot, as things on the bleeding edge often do, and that early units generally cannot be serviced due to parts unavailability.

I've come across a 32-bit ADC with a 90 dB dynamic range spec in some modern-day headset codec thingy, thought that was kind of hilarious (50% of bits being pure noise).

By 1990, the best mainstream IC converters were good enough to warrant more than 16-bit (which meant 18-bit) output, and by 1993 the first 20-bit jobs popped up. 20 bits was found in converters like the Focusrite Blue series by 1994 and became all the rage in prosumer soundcards in 1996, and then the next year the first 24-bit jobs popped up (of which even the best ones barely needed more than 20, but I guess people found handling 22 bits awkward and the data stream already supported 24 anyway).

The '90s were a wild time. Imagine going from barely 100 to almost 120 dB of dynamic range in like 7 years. By the end of the decade, performance approaching custom, hot-headed cost-no-object designs like the AD-122 became available in volume production (still expensive but no longer involving a kidney or first-born). By 2004, any reasonably well-heeled enthusiast could buy an EMU 1212M for like 200€.

(Mind you, in the same time a new PC probably went from a 386 with 1-4 MiB of FPM DRAM to a 233 MHz Pentium II with 64-128 MiB of PC66 SDRAM, and then a 2.8 GHz P4 or equivalent Athlon64 with 512 MiB to 1 GiB of DDR/DDR2 SDRAM. That's "a bit of a difference", too.)

 

[Ropeburn]

The '90s were a wild time. Imagine going from barely 100 to almost 120 dB of dynamic range in like 7 years. By the end of the decade, performance approaching custom, hot-headed cost-no-object designs like the AD-122 became available in volume production (still expensive but no longer involving a kidney or first-born). By 2004, any reasonably well-heeled enthusiast could buy an EMU 1212M for like 200€.

(Mind you, in the same time a new PC probably went from a 386 with 1-4 MiB of FPM DRAM to a 233 MHz Pentium II with 64-128 MiB of PC66 SDRAM, and then a 2.8 GHz P4 or equivalent Athlon64 with 512 MiB to 1 GiB of DDR/DDR2 SDRAM. That's "a bit of a difference", too.)

Exciting times indeed. Early 2000s you could also get a Delta 1010 for 400, or 500-600 when it came out in 1999. Not bad for the time for 8ch i/o, not bad at all. Products like these popped up in numbers, and finally meant just about anyone could suddenly afford professional features and quality.

 

[Pluto]

…Products like these popped up in numbers, and finally meant just about anyone could suddenly afford professional features and quality.

And the actual quality of recordings has diminished ever since.

You can give anyone a saw, mallet and chisel. That does not make them a carpenter…

 

[Ropeburn]

And the actual quality of recordings has diminished ever since.

Has it, though? I dunno man, ever since then I keep finding tons of great music that certainly would've sounded worse in the 90s and before, because the only way to get great quality was renting or building a 50,000$ studio. Doing away with that and democratising quality recording was the last great musical revolution in the late 90s and 2000s. Since then, all you need is a cheap laptop and audio interface, all that matters is skill and talent, not money. I much prefer that to the old ways, when poor kids couldn't do anything substantial.

Music hasn't gotten worse, especially not in recording and production quality, it's the exact opposite. There's just way more now, which inevitably includes the 99% crap that always existed. You just don't remember the old crap, typical survivor bias. Find the 1%, it's continuously been getting better and better.

 

[EJ3]

I'd say analog stuff really does benefit. De-noising and otherwise digitally massaging older recordings is better when playing with 24 bits. There's a lot more headroom.

I'll say TRUE: it "COULD" benefit: IF done right and not making other changes (for instance: the loudness wars messing up the dynamic range that was there originally)
It also "COULD" muck it up and make it worse.

 

[Robin L]

I'll say TRUE: it "COULD" benefit: IF done right and not making other changes (for instance: the loudness wars messing up the dynamic range that was there originally)
It also "COULD" muck it up and make it worse.

I mostly listen to classical and some older pop/rock. I'm not regularly encountering brickwalled recordings or remasters. With classical reissues, most recent reissues have some improvement over older remasters, there's a lot of them at Tidal, many as high-rez FLAC.