DP's internal mix engine?

[kassonica]

Just a clarification, is DP's internal mix engine 32 or 64 bit?

If 64 when did this happen?

I have a reason why I am asking, but I will reveal after the answer.

Thanks

[cuttime]

AFAIK it is and has always been 32 bit.

[kassonica]

AFAIK it is and has always been 32 bit.

That is what I thought as well, But I didn't know if it had changed with 8 or 9.

Maybe someone can confirm in the know.

[cuttime]

How about p.232 in the DP9 User's Guide?

[kassonica]

How about p.232 in the DP9 User's Guide?

Ok, I haven't even looked at the PDF...

But thanks, it is what I suspected...

[mikehalloran]

32bit floating point processing has nothing to do with DP running 32bit or 64bit.

[kassonica]

32bit floating point processing has nothing to do with DP running 32bit or 64bit.

Hi Mike

Yep I'm well aware of that.

[bayswater]

Says it's 32 bit float in the descriptions of V9 on the MOTU site. In the table of general features in "Feature Highlights".

[kassonica]

NOw from memory it went 32bit in 7, is that correct?

[Phil O]

A quote from their website:

With internal 32-bit floating point precision throughout the mixing path, combined with 64-bit precision during the mastering stages, audio quality is unparalleled.

What does "during the mastering stages" mean?

Phil

[mikehalloran]

NOw from memory it went 32bit in 7, is that correct?

No.

I don't remember when 32bit floating point processing was introduced but it was mentioned in reviews of DP 5.
http://www.soundonsound.com/motu-digital-performer-5

and DP 4.5

https://reverb.com/item/1231037-motu-di ... torial-dvd

And DP 4.

https://books.google.com/books?id=oMcqB ... ng&f=false

[kassonica]

Ok I'll level...

I recently recorded a EP at another studio for a very talented young lady (her debut recording) using Pro tools 11...

Now the Mic kit alone is worth around 30k, top of the line pre's etc etc, Apollo, conversion UAD plugins etc etc, So taking this into account, I noticed something else, The mixes sound really wide, really clear and have a amazing definition, Even bouncing out 44/16 (project was recorded at 96/24) this is retained, I asked myself why...

Now accounting for the gear/space/talent etc it left still something unaccounted for, IMHO, so I looked into PT...

Before I go on I would like to add I have used PT in the past and noticed a slight difference in the way it sums and bounces but it was nothing major....

So Looking into it, it seems since they introduced 10 HDX Avid changed PT from 48 Fixed point to 64 floating point and also changed the way it sums internally...

I've always noticed a collapse of depth when bouncing In DP or any other program, some collapse more than others BTW, even just a bounce to the same SR and BD it will change, but NOT this time.

It retained all the depth, Lots of verb was used as well and it was retained, The soundstage was left the same, I couldn't believe it TBH

Ps I'd like to add this is not a go at DP, I've been a registered user since 4, it is an observation.

If anyone would like to add their opinion I'd be happy to hear it.

PS I mainly use all DAW's for audio like a tape machine, I don't use much MIDI or software stuff.

Thanks for reading

[Robert Randolph]

Very likely has to do with the difference of the pan law between DP and PT. Especially since you comment about soundstage and depth, which pan law would affect.

Extremely unlikely to have anything to do with the bussing data width.

[kassonica]

Very likely has to do with the difference of the pan law between DP and PT. Especially since you comment about soundstage and depth, which pan law would affect.

Extremely unlikely to have anything to do with the bussing data width.

The pan law is the same AFAIK...

the 64BIT mix buss, though means DB the resolution...

[Robert Randolph]

Very likely has to do with the difference of the pan law between DP and PT. Especially since you comment about soundstage and depth, which pan law would affect.

Extremely unlikely to have anything to do with the bussing data width.

The pan law is the same AFAIK...

the 64BIT mix buss, though means DB the resolution...

Pro Tools defaults to -3.0db pan (-2.5 for V8 and lower), DP is fixed 3.5db. PT is variable as well, so it's possible someone could have changed it to another setting.

32-bit floating point gives you a scaling 192db of dynamic range. The possibility that your signal exceeded that DR requirement and was audible is very unlikely. You would have had to record signals at -72dbFS and ad 72db of gain before you'd maybe hear the noise floor above your converters.

There's other things to consider too, such as how PT handles automation interpolation, downsampling, crossfade types/rendering, etc... A pretty long list actually.

Then consider of course everything else you mentioned. Gear, space, talent..

The chance that somehow it was a 64-bit mixing engine difference is basically none. Simply moving your head 5mm to the side has a greater difference on the perception of sound than increased bit depth beyond 32-bit float (24bit int).

I'm not questioning that you heard a difference, but you're barking up the wrong tree. There's quite a laundry list of other things that it's significantly more likely to be.