BenMMusTech
 I will restate, that mixes and in particular mixes that use analouge emulation, time based and spatial effects benefit from a 64bit fp work flow through to the master.

I assume this statement is from empirical evidence i.e. your senses as I have seen no mathematical explanation to support your claim.

I can't think of any good reason why bit depth would be pertinent to time-based effects specifically. Bit depth has no significance in terms of the time domain. At a stretch, we could say that as reverbs and delays tail off into silence, then greater bit depth means greater resolution at low volumes. But 1/ I am certain that's not what's being claimed, and 2/ how significant that is within a general mix, where there's not going to be any silence to tail off into, and signals will have fallen below the masking effect of other signals way before that, well, I'm skeptical, myself.

BenMMusTech
Hmm if none of you can hear the difference then that should tell everyone not to trust your advice for one.

A predictable move that I'm surprised it took you this long to make. You, apparently, have golden ears, and we are all completely tin-eared, and can't even hear the excellence of what you're talking about. Sure. That's what's happening here.
 
Anyway, for me, I've pretty much said my piece. It bothers me that anyone reading this forum might have their time wasted by some of the nonsense being spouted. I think for the time being, we've headed that off as much as possible. If you want to continue to make a fool of yourself, it's not my job to talk you out of it.

∞

Amused that you reference Nicholas Negroponte who (I personally think wrongly) is generally opposed to putative "HD" media technology. I wonder how closely you read him. I also wonder why you think he has any credentials in audio; he doesn't, and doesn't claim to. But this is par for this boring course you've set us on. Are you familiar with the term "appeal to authority"?

dj squarewave
Info about the wave file container format:

https://www.loc.gov/prese...ts/fdd/fdd000001.shtml

Indeed it is. What point are you making, though?

I swear I'm going to leave this alone soon, but there are so many ludicrous claims in that massive wall o text up there, it's hard to pick them out. But this one's a doozy.
 

BenMMusTech
The beauty of 64bit fp audio though is you get a wide fat mix in the 1st place, and then it doesn't matter about the delivery stage.

 
Create master files in 64 bit. Then deliver over a walkie talkie. Because it doesn't matter about the delivery stage. 64 bit will magically make it awesome.

BUT THEN!

And as for the person saying his LG has **** speakers...all handheld computers have pretty crap speakers...they're the size of a very small coin. It's when you connect the handheld computer to decent speakers...like Harmon Kardon for instance you get the benefits.

Oh, look, the delivery stage, mattering.
 
That'll be those amazing arguing skills there.

Okay, at the risk of sounding stupid, in the midst of all this debate I remain confused what the right answer really is to one particular facet of my inquiry:  To clarify I'm using Sonar 5 (which supports a 32 bit file container - wasn't until S6 that 64 bit was supported).  So, when I am mixing a project with audio files that are recorded at 24 bit, and I wish to export a mix hoping to maintain the highest quality audio, should I select 24 bit or 32 bit file depth in the export dialog? Or does it simply not matter ??   If it doesn't matter, are there any pros or cons to selecting 32 over 24 bit? 
 
    Sonic
 
 

Sorry, I've not been helping at all have I?
 
I think the question's been answered a couple of times in the thread. There are no gains in exporting to 32 bit. What's coming off your master bus isn't 32 bit, and it's going into your RME, to be played back at not 32 bit, but 24 bit, and that's what you're hearing.
 
There's been some back and forth in this thread between different parts of the process, and I think it's all got a bit muddied up (I'm as much to blame for that as anyone). So here's a quick overview.
 
1/ Initial recording - there is no point recording into 32 bit files, as it's currently not possible to capture audio of greater depth than about 21 bits (and may never be possible).
 
2/ Internal processing in the DAW. In your case, this is happening at 32 bits. Nothing to concern us here.
 
3/ Bouncing / freezing operations - Sonar sets it's default render bit depth for these operations at 32 bit. The principal benefit of this is that it makes it very unlikely you'll accidentally create a clipped signal when doing a freeze or a bounce. The secondary benefit is that you're very unlikely to suffer signal degradation even across multiple bounces of the same material.
 
4/ Playback - your interface (and the majority of interfaces) is running at 24 bit. The mix coming off the master bus will be 24 bit, regardless of any intermediary processing going on in steps 1 and 2.
 
5/ Final export - same as the playback. You do not now, nor at any point in the process, have music that's playing at higher than 24 bit.
 
 
 
 
 
 
 
 
 
 
 
 

Thanks for the clarifications John.  What you explained was my proper conclusion then. I was just starting to get confused reading the discussion and then started to doubt aspects of what i thought I already knew, as well as some aspects which thought I learned during the discussion.  As you said, things kind of got a bit muddled up in the twists of the debate.
 
Sonic