64-bit audio engines are on the way !

[defjamm]

'do you make money with music?'

I played piano in a Kansas City whore-house for 5 years-does that count?

"we don't believe you. do you have some video footage or pictures?"

[Improv]

'do you make money with music?'

I played piano in a Kansas City whore-house for 5 years-does that count?

"we don't believe you. do you have some video footage or pictures?"

DVD's (And t-shirts) will be for sale in the lobby as you leave. Thank you for flying Hooker Air!

[Omniphonix]

It's funny how much people in this thread are confusing 64-bit mixing bus with 64-bit application running on 64-bit hardware. The 64-bit mixing bus (and internal processing at 64-bit by plugins) is really just a marketing ploy to say "Look how accurate our plugin is, it will sound superior to all", when in reality, as BONES explained, even the ancient golden-ear mastering engineers would not be able to tell the difference.

Now when we start talking about 64-bit CPU/OS/App, we're talking about the processing of the application's instructions and you will get a real, noticable performance gain. It's the same reason so many datacenters are moving to 64-bit solutions for their database systems, more bandwidth through which to move data. This means moving more audio through the CPU in less time and that's really what all this is about. It's not a "flip a swtich to 64-bit and you are now twice as fast" kinda deal, but more like any other kind of technological progress.

When Windows 95 came out and we finally had a 32-bit OS on 32-bit hardware, application were not all of a sudden magically faster. It took years of code optimizing and progress with development tools (not to mention nifty add-ons like MMX and SSE) that 32-bit opened the door to. Now the same thing will happen as developers begin optimizing their code for 64-bit platforms. In 4-5 years, people will be talking about their "old" 32-bit applications and how much it sucks they still need to keep them around (just like when you had to run a 16-bit app on Windows 98).

[maoinhibitor]

Bones, Kingston.

What's with the negativity? Feeling defensive because your hos is missing a few features?

I've done A/B comparisons and I can hear the difference in my mixes.

And, yes maybe I didn't need to mention that my friend had won some grammy awards. But he is the most knowledgeable person I've ever met about these sorts of things both technically and sonically. His ears are golden and he has received recognition from the industry for his work.

You can stay in denial if you want. Let's suffice to say that when your host's engine goes to 64-bit you can make up your own mind.

[spaceman]

What's with the 'I have a grammy award-winning/engineer friend who..' lately?

I used to get Granny Awards when I was young
a pick from the candy box.. woohoo!!!

[spaceman]

'do you make money with music?'

I played piano in a Kansas City whore-house for 5 years-does that count?

I bet you spent more than you made

[tony tony chopper]

I've done A/B comparisons and I can hear the difference in my mixes

If you can hear A difference, then maybe you should consider there's something really wrong & buggy between the 32 & 64 mixing modes?

Plz share some demo clips so that we can hear too, btw.

[Arksun]

If anyone wants to hear the difference the 'accumulative' effect of working at 64-bit compared to 32-bit is, simply fire up ArtsAcoustic Reverb.

Run a vocal through it or something, and flip the little 32-bit switch to 64. That kind of difference is clearly audible to anyone, no golden ears required.

However that is a heavily multi layered accumulative effect.

So ignoring all that dynamic range/noise floor bollocks for a moment (because that's not what the higher bit-rates are really about, but retaining all the bits for retaining the waveshape after a process)

The real question is, is the accumulative effect of simply mixing multiple channels at 64-bits, significantly different than it is at 32-bits to be audibly noticeable?

One question I can't answer because cubase that I use doesn't have a little button I can flip to switch it to 64bit mixbus.

However, high end digital mixers have used extremely high internal bit-rates for many years now, 67-bits, 96bits, 128. I don't believe this is all down to a cheap marketting ploy.

Comparing just 2 channels mixed together at 32-bit or 64bit, of course the difference will be unnoticeable. Anyone who thinks they can hear a difference there is a fool. But if heavy project with many many channels was mixed at a higher rate?. That is the question.

I don't know how sonar works, I only ever tried cakewalk a loong time back. So if anyones made a piece of music in sonar that uses say, I dunno 40 audio tracks, some soft-synths etc. Please do, make a mixdown with the mixbus set to 32bit, and another with it set to 64bit. Mix em down and post up a couple short .wavs so we can hear if there is any diff.

[tony tony chopper]

The real question is, is the accumulative effect of simply mixing multiple channels at 64-bits, significantly different than it is at 32-bits to be audibly noticeable?

I've done tests for an additive synth (adding like 500 sines, it's nothing small), and a 64bit version didn't sound any different.

Of course for other processes, like filters, 64bit is sometimes necessary. But in no way this requires a 64bit output.

However, high end digital mixers have used extremely high internal bit-rates for many years now, 67-bits, 98bits, 128. I don't believe this is all down to a cheap marketting ploy.

I believe digital mixers mostly work with integer math, so it's not directly comparable (and it's bit depth btw, bitrate is something else).
But yes, it's also marketing - whatever sells. If competitor uses 64bit, and people buy, you have to, wether it's dumb or not.

[swampmonster50]

bollox, im goin 2 make a sandwich

[Arksun]

The real question is, is the accumulative effect of simply mixing multiple channels at 64-bits, significantly different than it is at 32-bits to be audibly noticeable?

I've done tests for an additive synth (adding like 500 sines, it's nothing small), and a 64bit version didn't sound any different.

Of course for other processes, like filters, 64bit is sometimes necessary. But in no way this requires a 64bit output.

Really we need someone to do a 32-bit vrs 64-bit mixdown on a very heavy 'real world' project. Lots of mixed sounds, bright percussion etc. Sine waves don't tell us much, its the most basic fixed soft-wave shape and no hard transients are involved.

Any Sonar users brave enough to step up to the challenge and back up their bold claims?

[tony tony chopper]

Sine waves don't tell us much, its the most basic fixed soft-wave shape and no hard transients are involved.

why do you need transients exactly?

Believe me, if it's noise you want to hear (and it's about noise here), you're likely to hear it on a sine, NOT on a final mix.

Just take any of your existing songs, and turn it to 14bit integer, even down to 13, you may not even hear any difference.

[tony tony chopper]

Btw, we've just released a new beta of FL, and it comes with Edison that has a scripting engine. So, providing you know a little pascal, you can make your own tests, adding lots of stuff in 32 & 64bit, and listening to the results.

[Arksun]

Sine waves don't tell us much, its the most basic fixed soft-wave shape and no hard transients are involved.

why do you need transients exactly?

Believe me, if it's noise you want to hear (and it's about noise here), you're likely to hear it on a sine, NOT on a final mix.

Just take any of your existing songs, and turn it to 14bit integer, even down to 13, you may not even hear any difference.

Thats not the same thing, thats taking a finished pre-processed at 32-bit song, and reducing the final mixdown bits to 14.

Or do you mean inserting a bit-reducer on every single channel at the production level?. Even then, forcing each channel to go to 14-bits, Cubase will STILL be mixing them together at 32-bits. As Cubase doesn't have a "switch mixbus to 14-bit mode" I can't create such a situation :/

I recommend some of the articles Bob Katz has written about bit depth, make fun reading. This isn't about noise floor people

On a side note, did anyone know the Emu 1820M soundcard (a card which sounds surprisingly good for the money) uses 67-bit internal mixing engine?

No neither did I until recently. It's not written on their website or on the box, so not for marketing.

Like I say, 64-bit clearly makes a big difference on specific processes like reverbs. As for mixing?. Hmmm, I'm on the fence on this one. Need convincing by a Sonar user with .wav examples. Somehow I get the feeling we won't get any

[tony tony chopper]

Thats not the same thing, thats taking a finished pre-processed at 32-bit song, and reducing the final mixdown bits to 14.

It was just to say that you're not likely to hear any noise in the 16bit range on a final real-case song (you wanted a 64bit mix of a song).
So to prove one's point better, let's use whatever is supposed to show the noise better.