new pro tools is not as good

[zion]

Originally Posted by mightyduck

Well, I'm sure you can see that a lot of that is double-speak and cya. They still don't come out and say what they mean by 64-bit compatible.

Anyhow, personally I think a lot of the decisions were made in favor of increasing track count. That's what they're on about, and that's what a lot of users probably want. More tracks. So that's what they are concentrating on.

The more track count the more gain (volume), thats what they are referring to here. Thats were a 32bit float and 64bit float mixer summing going to come into play here in a major way. As well as a better sound quality.
"Quote A 64-bit floating-point mix bus provides more than 1,000 dB of headroom, which is more than enough to handle the huge track counts that Pro Tools|HDX can deliver"

[zion]

Originally Posted by mightyduck

Well, I'm sure you can see that a lot of that is double-speak and cya. They still don't come out and say what they mean by 64-bit compatible.

We created AAX for the future, as it's a 64-bit (ready) format

[mightyduck]

Originally Posted by zion

The more track count the more gain (volume), thats what they are referring to here. Thats were a 32bit float and 64bit float mixer summing going to come into play here in a major way.

They're trying to say that the 64 bit summing bus will allow a larger track count. But that's a red herring, because its still going to have to fit onto a 24 bit fixed path for output. And that's where it may get weird.

As well as a better sound quality.

Not with what they are doing.

[Emcha_audio]

Originally Posted by mightyduck

Well, I'm sure you can see that a lot of that is double-speak and cya. They still don't come out and say what they mean by 64-bit compatible.

Anyhow, personally I think a lot of the decisions were made in favor of increasing track count. That's what they're on about, and that's what a lot of users probably want. More tracks. So that's what they are concentrating on.

AAX is 64 bit ready. It's not a may or may not, they are.

[mightyduck]

Originally Posted by zion

We created AAX for the future, as it's a 64-bit (ready) format

What is your point?

64 bit what?

[Emcha_audio]

Originally Posted by mightyduck

What is your point?

64 bit what?

Memory access 64 bit ready. I doubt well see the engine (as in floating points through and through) change soon, it's taken long enough for it to evolved to what it is now.

[mightyduck]

Originally Posted by Emcha_audio

Memory access 64 bit ready. I doubt well see the engine (as in floating points through and through) change soon, it's taken long enough for it to evolved to what it is now.

Right, I don't know that there will be much in the way of 64 bit [double precision] aax or rtas plugins, if past is prologue, and given that the bus is 32 bit float.

Who knows. 64 bit app. will kick butt for virtual instruments and stuff, though.

[zion]

Originally Posted by mightyduck

What is your point?

64 bit what?

TDM has no future in the 64bit world. The bottom line is it ends here at
Pro Tools 10 software.

[Avid]

Some clarification from Engineering:

The progression to a 64-bit app has nothing to do with the mixing DSP, only with the memory space/OS used (and it breaks compatibility with older code including plug-ins).

We have no plans currently to change the mixing structure in HDX (or Native) – it is very similar to TDM systems in that the internal mixing is done to a 64-bit floating point accumulator (even if a mix bus spans DSP chips, similar to how a TDM system will keep a 48-bit fixed point mix bus across DSP chips). There are no cases in a TDM system where we keep 48-bits but don’t keep 64-bit floating point in the HDX system.

Floating point is more convenient for programmers, but it also has some cool advantages. Gain stages don't strongly affect precision. You can gain something down by 60dB (for example) and back up again with no loss in precision. That is where the dynamic range comes into play. Depending on the actual number, you could have some rounding in the lowest bits, but nowhere near the potential signal loss of fixed. This allows you to be a little more care-free with gain-staging. Single Precision floating point does have some disadvantage in precision compared with 32 bit integer, but it has no precision downsides that I'm aware of compared with 24 bit integer. In all places where we need a lot of precision, we have converted to double precision floating point. Double precision floating point, which is what we converted the 48 bit parts to, actually has 52 bits of precision along with the increased dynamic range, so it is clearly better in all regards.

Dither is indeed a different beast with floating point – a fixed-level dither signal doesn’t “smooth out” quantization distortion like fixed-point dither does. That said, there are several dither plug-ins available which provide the right level of dither such that the eventual quantization to 24/20/16 bit is properly “smoothed.”

Avid Engineering