Skip to main content

With the advent of the 64 bit computing revolution and that windows based audio editing apps offer the option to record in 64 bits, does anyone make a sound card that allows one to record at 192khz / 64 bits?

Comments

Thomas W. Bethel Wed, 11/26/2003 - 05:09

There are very few recording rooms that are even 24 bit quiet. Most gear is not 24 bit quiet. 24 bits translates into 144 dBu of S/N ratio (or dynamic range). 32 bits translates into 192 dBu for S/N ratio and 64 bits translates into 384 dBu of S/N ratio. My question would be WHY would you want a dynamic range of 384 dBu? Just curious????

Really good recording rooms are NC 15 or 20 (approximately 20 dBSPL). A symphony orchestra playing at full volume can reach 120 dB. Hence a dynamic range of approximately 100 dB (+/-)

There is really no reason for more bits than 24 unless you are going to have the equipment and the rooms to support that S/N ratio. Unless there is some other reason you are wanting 64 bit resolution.

sserendipity Mon, 12/01/2003 - 14:00

With the advent of the 64 bit computing revolution and that windows based audio editing apps offer the option to record in 64 bits, does anyone make a sound card that allows one to record at 192khz / 64 bits?

The word length of your computer and the sample resolution of your audio have absolutely nothing to do with one another.

Originally posted by Yon:
maybe not a 'sound card' like you would buy at a store, but there are definately devices to record at that resolution - and have been for years. They are audio cards, but are used to for other applications, like balance engineering and crazy stuff like that.

What's a balance engineer?

Also, could you name any 64 bit cards? I've never seen anything running at that resolution.

anonymous Tue, 12/02/2003 - 06:19

well my dad was a "balance engineer" and designed computers and machines used to balance things like crankshafts, wheels, other parts for bigger machines and engines (other then cars)... These are something like in the millions of dollars and wiegh a few tons. Shipping them to china is a big deal :) yeah, it's completely off topic and not relevent to us humans playing music... I just thought it might be interesting.

sserendipity Wed, 12/03/2003 - 11:33

Originally posted by GRP:
Hey just thaught you guys might be interested in this as a format to mix or record to but more importantly master from or even to. DSD (direct stream digital) the stuff that SACD is based on. 24bit but the sample rate is 28224Khz and its very different from pcm based formats.

http://www.genex.co.uk/index.html

http://www.sel.sony.com/SEL/consumer/dsd/dsd.pdf

Yes, you might want to do a search for SACD and DSD on this board. There's been some discussion. We'll see this before we see 64 bit PCM.

I looked through the articles you link to, but couldn't find anything that referes to the maxiumum dynamic range at 24 bits. Makes sense though - anything more is overkill unless you are going to be processing the signal. Could you point us to it though?

sserendipity Mon, 12/15/2003 - 13:52

Originally posted by Yon:
DSD is 1-bit!

That's not relevant to the question at hand - the -Arbitrary- maximum dynamic resolution of a DSD stream. One bit data transfer, yes, but maximum dynamic resolution does not correlate to this in the DSD realm, since you are no longer passing along a 'snapshot' of the entire waveform.

So - where does it say that DSD has 24 bits of resolution (as was mentioned earlier), and if it has something different, what is it? It can't be infinite - neither my stereo nor my eardrums could handle that :>

sheet Mon, 12/15/2003 - 21:53

Originally posted by Thomas W. Bethel:
There are very few recording rooms that are even 24 bit quiet. Most gear is not 24 bit quiet. 24 bits translates into 144 dBu of S/N ratio (or dynamic range). 32 bits translates into 192 dBu for S/N ratio and 64 bits translates into 384 dBu of S/N ratio. My question would be WHY would you want a dynamic range of 384 dBu? Just curious????

Really good recording rooms are NC 15 or 20 (approximately 20 dBSPL). A symphony orchestra playing at full volume can reach 120 dB. Hence a dynamic range of approximately 100 dB (+/-)

There is really no reason for more bits than 24 unless you are going to have the equipment and the rooms to support that S/N ratio. Unless there is some other reason you are wanting 64 bit resolution.

I think that some of you are a bit confused. The 64-bit capability of the new G5 and the new PC's is not the same thing as the sample rate if digital audio converters. We are talking data throughput here, not sample rates. There is software now that is forward and bakward compatible to 64-bit. Truely things will blossom when we have companies like Digidesign write software for it.

The other thing is that the theoretical dynamic range given above (1 bit = 6dB of dynamic range) cannot be directly correlated to a S/N ratio. S/N will be determined by converter topology, construction quality, etc.

But, I will say that most rooms are not digital friendly, because they were built for noisy analog.

The average Jim-Bob's hearing resolution equals 16bit/48kHz. A learned ear can obviously hear a difference in soundfield, blah, blah, blah from recordings made at higher sample rates.

The issue for me is still the microphone. Why would we increase sample rates, extending the top end above what analog was yielding, and above what any current microphone can capture.reproduce linearly? Most microphone manufacturers DO NOT include the preamp in the testing stage for S/N. The ones that do use the dB-A scale, which means "add another 10dB to get the real ratio." The only company that includes the whole circuit and does not use the A scale is DPA.

Additionally, while we are talking transducers....what consumer has the speaker resolution to hear this tweak crap? Who really cares? The average Best Buy customer doesn't, because they are buying music with no dynamic range and limited bandwidth. That is the music that pays the bills for the labels.

As for DSD and PCM. It is severely flawed. The high-end is funkified with all of the filtering of the distortion. It is "different" but not the best that the industry could do.

We need to develop the technology that drives the consumers instead of the other way around.

anonymous Mon, 12/15/2003 - 22:43

Nice post, sheet.

The current topic looks like the bit depth equivalent of the Dave Collins 'one-ninety-who?' sampling rate thread on the Glenn Meadows mastering webboard recently, where Dan Lavry's comments were reposted saying, among other things, that more is not necessarily better (regardless of what salespeople might say), with some solid arguments to back it up.

It puts me in mind of the old joke 'why does a dog lick its balls?' . Answer: 'because it can'.

Thomas W. Bethel Tue, 12/16/2003 - 04:50


I think that some of you are a bit confused. The 64-bit capability of the new G5 and the new PC's is not the same thing as the sample rate if digital audio converters. We are talking data throughput here, not sample rates. There is software now that is forward and backward compatible to 64-bit. Truly things will blossom when we have companies like Digidesign write software for it.

The other thing is that the theoretical dynamic range given above (1 bit = 6dB of dynamic range) cannot be directly correlated to a S/N ratio. S/N will be determined by converter topology, construction quality, etc.

But, I will say that most rooms are not digital friendly, because they were built for noisy analog.

The average Jim-Bob's hearing resolution equals 16bit/48kHz. A learned ear can obviously hear a difference in soundfield, blah, blah, blah from recordings made at higher sample rates.

The issue for me is still the microphone. Why would we increase sample rates, extending the top end above what analog was yielding, and above what any current microphone can capture.reproduce linearly? Most microphone manufacturers DO NOT include the preamp in the testing stage for S/N. The ones that do use the dB-A scale, which means "add another 10dB to get the real ratio." The only company that includes the whole circuit and does not use the A scale is DPA.

Additionally, while we are talking transducers....what consumer has the speaker resolution to hear this tweak crap? Who really cares? The average Best Buy customer doesn't, because they are buying music with no dynamic range and limited bandwidth. That is the music that pays the bills for the labels.

As for DSD and PCM. It is severely flawed. The high-end is funkified with all of the filtering of the distortion. It is "different" but not the best that the industry could do.

We need to develop the technology that drives the consumers instead of the other way around.

Nice post. But re read the original post. He is asking about recording in 64 bits (not real clear about whether he is talking about computing 64 bits or soundcard 64 bits.

You are of course correct about the S/N ratio. Maybe I would have been better off saying dynamic range but even that would not be totally correct.

I think this whole push towards 192/24 bits is being driven not by consumers (many of whom listen to music that has a dynamic range less than 16 dB and listen to it in cars with road noise reaching 85 or more dBSPL) but by the manufactures who want to sell people more gear and know that they have to create a need before they can supply that need. So they push higher and higher sampling rates and higher and higher bit rates saying that it is sooooooooo much better that you will have to replace all of your gear. (and they mean ALL of your gear since everything will have to be 192/24 bit clean)

We get a lot of clients in my mastering studio who want to master at 24 bits. We normally do this on a day by day basis so all of our equipment is designed for this. Recently I have had one or two clients that are asking if I can master their material in 192/24. Well no because none of my equipment including my Weiss EQ1 MKII will sample at that rate. And my sound card will not do it either nor will most of my software plugins. Am I going to go out and buy all new equipment for one or two projects I think not. Are there other mastering studios who have that equipment - not around here. Unfortunately the clients that want the 192/24 typically bring in material that is so squashed prior to mastering that it may have a dynamic range of 10 dB and there is so much guitar amplifier noise that the s/n ratio is maybe the same.

Where will this all end...not sure. Will we be do recording and playback at 192/24 on a daily basis maybe but not in the foreseeable future.

Again nice post.

anonymous Thu, 12/18/2003 - 18:43

Originally posted by GRP:
Hey just thaught you guys might be interested in this as a format to mix or record to but more importantly master from or even to. DSD (direct stream digital) the stuff that SACD is based on. 24bit but the sample rate is 28224Khz and its very different from pcm based formats.

A little off there. DSD is one bit. The sampling rate is 64 times 44.1k or 2.822MHz.

DC

golli Thu, 12/18/2003 - 20:52

I can't master at 192k because none of my gear that's worth anything can do it at this rate. plus I can't hear a difference if it could. Oh well, I guess i'm a loser.

And still you have the ability to make your clients happy with your work.
My point is that there have been many threads on the web, regarding "48 vs 96" or 96 vs 192" and people with self proclaimed "golden ears" trying to undermine others (us normal hearing folks).
Anyway, good to hear from a pro, that good music can be made without robbing a bank first.

anonymous Mon, 12/22/2003 - 13:59

Are you kidding? 24/192 is an SRC away, my friend - at the end of the chain, as well. Grab a DVD audio file (not DVD-A, just standard DVD) sometime and look at the frequency response of the source material. You'll be lucky to find anything exceeding 23Khz. Why? Because it was recorded on a 24/48 system, and SRC'ed to 24/96 to meet the DVD spec. These people that ask you to master at 24/192 wouldn't know the difference anyway - for the precise reasons you mentioned - no dynamic range or HF material in their source files. If they want a final output at 24/192, give it to them - but don't change out every piece of your signal chain to do it!

anonymous Wed, 12/24/2003 - 08:23

64 bit 192kHz AD or DA converters does not exist nor will they in any near future for the simple reason that it's not physically possible. To get low enough noise floor one would probably need power in terra watt range and make converters from resistors in µ ohm range, and invent a new kind of low noise transistors.

Now the best converters today have a DR of around 123dB, that is well below 144dB that 24 Bit could provide, so using 64 bit data rate would be a total waste.

Today mixing is often done in 32 bit, which gives good headroom, I don't see any point in increasing this to 64 bit. Now 64bit computers can probably be of good use anyway if the programs i written in a good way, to exploit the advantages of 64 bit processing without using 64 bit data words.

TEG

Michael Fossenkemper Wed, 12/24/2003 - 09:42

Until someone comes out with better processing than weiss, I'm sticking to 24/96. There are so few systems that can even reveal the possibilities of higher bit and sample rates that i don't even know 1 person that has one. Now I've listened to some really great systems but I feel that it has more to do with the construction than the resolution of the material being played through them. I bought a pioneer SACD player with 24/192 D/A converters. It's about a $700 player. doesn't even sound as good as cheap 16/44.1 digidesign converter. I'd be interested on how many people are using 192k in their recording chain.

anonymous Sat, 01/03/2004 - 04:52

Originally posted by Michael Fossenkemper:
I bought a pioneer SACD player with 24/192 D/A converters.

And there, sir, is exactly why it sounds like crapola.

What you bought was one of the Pioneer combination DVD-A/SACD players - the ones that internally convert the 2/2.8Mhz SACD signal to a 24/96 PCM, then engage those 24/192 converters to send it to analog.

The dumbest idea I've ever seen.

Hell, I've got a cheapy little Sony 5-disc SACD player (paid about $300 for it) that whoops those Pioneer combo units - but it doesn't play DVD-A.