Hi:
 
I've finished mastering my project using Sonar X1 Producer Expanded. There are 50 audio tracks, all of which are routed to the master bus. Each audio track contains some effects and automation and the master bus also contains some effects. Overall, everything sounds the way I want it to. I would like to export everything to a wav file, but have a few questions about some of the settings.
 
1) For the source category, should I choose entire mix, main output or master bus? I want to ensure that the exported audio file sounds exactly like what I hear in Sonar X1 and, in particular, I want to ensure that all the effects contained in the master bus are picked up. What is the difference between these sources?
 
2) What should I choose for the bit depth? I understand 16 to be usual, but is 32 better?
 
3) What setting should I choose for dithering? My project contains a lot of harmonic nuances and I want to ensure that they are all captured accurately. Would Pow-r 3 be the best option?
 
4) Is there any benefit to de-selecting fast-bounce? Will I get a more accurate export file if I de-select this?
 
5) I have been using a 64 bit engine for the entire project. Should I select this export audio option to be consistent?
 
Any other tips or suggestions would be appreciated.
 
Thanks to all those who reply.
 
Best,
 
RM

A safe bet is to select the entire mix, although you could select buses and just the main bus. I see 'entire mix' as a sort of a short cut instead of further selecting/deselecting busses.
 
Bit size: depends on where your going with the output. If you are going to use the output file as a master copy for use in another application to produce, say, mp3's or another file type, go with 24 bits, otherwise go with 16 bits.
 
Dither: as the help file says, go with triangle, unless you know for sure this is a final copy, than consider powr 3.
 
Fast bounce: 6 of one, 12 dozen of another. No reason to deselect.
 
64 bits on export: Not needed unless you know have a unusual technical reason for it. Doesn't buy you anything sonically.

I've understood it this way:
If you're about to master the exported wav in another program using 32 bit floating point (or SONAR) export it in 32 bits (without changing the sample rate) which is SONARs internal depth. That way the project stays 100% untouched. If I've misunderstood it, someone wiser will correct, I'm sure.
Dithering is only needed if you lower the bit depth.
If you just want to burn the wav to a CD and/or make MP3s,  use 44,1 kHz/16 bit and Powr-3 dithering (IF you have used 24 bits in recording). OTOH, I wonder if anyone really hears differences between different dithering algorithms. The differences are measurable, though.
 
If all tracks are routed to Master Bus, as you describe, use Master as the source. That's the only way to be sure you export what you hear.
It's always possible, theoretically, that you have abandoned tracks which are not routed to master. Choosing Entire Mix as the source might export them, too.

Using Entire Mix might cause a problem if you have more than one set of outputs routed to your interface.
 
For instance, I have a headphone bus which is routed to outputs 3/4 on my interface.
If I export using "Entire Mix", the resulting mix will be 6dB higher than expected, because "Entire Mix" is the sum of all active output pairs, so for this reason, I always use "Main Outs"

1) Entire Mix
2) 16 if you want to burn it to a standard CD
3) Your choice. There is a section in the manual that describes each type and what its best for. 
4) Fast bounce doesn't affect the quality of the resulting WAV file.
5) I wouldn't change it.

Note that you can only use Fast Bounce if all instruments are either frozen or recorded to audio.  For example, you can export with external MIDI synths and internal soft synths that are not recorded or frozen, but you will have to select Include Live Input which means no Fast Bounce.  (I do this routinely.)
 
I don't know if you have to bounce Melodyne or not to use Fast Bounce--I don't recall.

konradh
Note that you can only use Fast Bounce if all instruments are either frozen or recorded to audio. 

 
That's not strictly true, unless I'm misunderstanding you.
I routinely use fast bounce on all of my soft synths and have never included Live Input in order to get them to render properly.
The only one that doesn't play well with fast bounce is BFD2, so a real time bounce is necessary whenever it's in the project and unfrozen.

OK, I may be incorrect on exactly what is and isn't included.  I will test a bit.

You haven't mastered your project - you've mixed it.  May be pedantic but I guarantee the right terminology will save you grief in the future.  Mastering is the next step of assembling several projects/songs into a coherent unit while limiting the individual songs.
 
What I do with the "mix" is export the main output.  That should be the only thing going to your monitors so what you hear is what you get.  64 bit engine should be selected - this runs the computations  at twice the depth as 32.  I don't know if it makes better sound but it should theoretically have less rounding errors and it just makes the computer work a little harder.  A 32 bit rendered file is fine, or 24 bit at whatever the project rate is (tho you may want to bump that up to 96 kHz).  Anyway, I use the project rate/depth.  Then I put that master mix back into the project on a track labeled Mix and save.  That way the highest quality mix stays w/ the project and you can always go back to it.  And it also allows you to compare your mix w/ the mixed version - just use the hardware "main out" rather than master in the project so you hear both versions w/o interference (as your project does w/ effects on the master).
 
Mastering is done from a copy of the Mix file.
 
One thing that few in the home recording business talk about is getting a system for naming numbering the different files of the same project and mixes.  There is the project, there is the master, there is the cd copy and mp3 copy (and professionally you might see main vox +1, -1, no vox, etc. for remixes).  Set up a file naming system that makes sense to you and stick to it.  If you do a lot of songs, you won't remember why this version is unlabeled and just sitting there, mucking up your system.  I have songs that have 20 or more variations and mixes - and that is before you get to the mp3/CDs.  If you work at an audio/video house, they will have a system but at least you'll already have the modules to plug into their system already firmly in your mind.
 
Have fun.
 
@