http://www.dailyrindblog.com/technical-supporting-high-quality-audio/?utm_source=Sailthru&utm_medium=email&utm_campaign=Client%20Newsletter:%20week%20of%20May%2016%2C%202016&utm_term=All%20Clients%20%2B%20All%20Orchardites
 
Hey there
 
I find it very important as well as confusing. I distribute my music via The Orchard. Far as I remember, the only was to upload music to Orchard was MP3. I uploaded my last album in 2010, so I do not remember what I did- it is possible I uploaded 16 bit/44.1 .wav file and they converted it to what ever MP3 standard they used at that time.
 
Now I am confused. This article about their new specs is contradiction. It says The Orchard’s brand new standard audio specs are: WAV (.wav) / 16- or 24-bit / 44.1 kHz / Stereo. But then it says, it accepts other formats (see article). Note that it says NOTHING about MP3. 
It also says, "Fun fact: most audio releases are in fact mastered as 24-bit, so it makes sense to offer to the consumer a record in its most native state."
 
I am wondering: Lets say I am mixing a song which is mixed 24 and 16 bit tracks. (I record at 24 now but my older songs were at 16 bit, and I am not sure what software synths use which samples). I mix the song first, and then master it in the new project with Izotope Ozone 7. Before I master, I export my song as a stereo track. Until now, I was exporting it at 16 bit and then do the Ozone 7 mastering on ut and export at 16 bit again.
 
The question is... if they do accept 24 bits now, does it means I should export stereo file after mixing in 24 bits for mastering, master it in 24 bits and then truncate one copy for CD format and leave another copy at 24 bit for digital distribution?
 
What do we do now? Thank you. 

Sounds like a question for the Orchard site support.
 
 
That said, there is no advantage to distributing 24 bit audio as a final product. No reproduction device (except perhaps a sonic weapon) will use more dB than the 16 bits standard will encode. Any concerns about computer processing truncation or rounding errors or digital clipping become irrelevant once the final mastering is done. So this is just a gimmick they are using to try to make people think they are getting some additional bits that they can use somehow to make the sound superior. If you think there are enough ignorant music downloaders out there that will not buy your work unless it contains useless garbage bits, then you will probably need to make a 24 bit version available. I am pretty sure you can also safely reduce the final bit depth from 24 to 16 using Sonar without sacrificing any audible quality if that is your decision. 

+1
 
Except that I would recommend staying at 24 bit until the FINAL export. It's much better to leave Ozone working at 24 bit. 
 

What specs do you use when you export from Sonar for Mastering and what are your final specs? I would also ask who you distribute with, thank you. 

I export SONAR files at 24 bit, 44.1 or whatever the project uses.  I put that wav into the project audio file and add it to the SONAR project timeline (muting, of course).  I then import that file into Sound Forge and  either use digital to master or send it out through outboard for mastering.  Then top and tail, and save as a 16 bit CD file and then mp3 (tho the best practice is to mp3 from the "project master" at 24 bits.  I put those finished files into another folder (mixes).  That pretty much covers all bases.  And again, it is best practice to master from the highest quality copy.

I master with Izotope Ozone and it offers a war to change 24 to 16, but I was not sure which program- Sonar or Izotope doing it better, so I was exporting from Sonar in 16 bit and then apply Izotope to Master bus on the new Sonar project. I guess I could still master at 24 bit and then use Sonar Master project to export it as 16 bit. For some reason I felt that if I apply all the effects of mastering to 16 bit project, some how it would cover up missing bits. (I am thinking about it as a sample rate, which is not). It feels like if I want to take something out of the file, it is better to do before all effects applied. But then again- this is why I am asking.
 
 AT wrote, "I put that wav into the project audio file and add it to the SONAR project timeline (muting, of course)."
 
I am not following this- what is project audio file and project timeline? You mean- you put it on a track?   

I export my mixes at 32 bit 44.1KHz.
 
I then re-import each one into the originating project file on it's own track. this makes it easy to verify that the export went as expected by comparing the export (track) with the mix of the project.
 
When all is ok, I open up my mastering template in sonar and import the same file again and feeds my mastering chain.
 
Remember this is still all at 32 bit!!
 
Finally, I export at 16 bit, 44.1KHz with dithering applied in order to burn a CD.
 
Your final requirements might be different depending on what you want!

I noted that Sonar is set to export .wav at 32 bit if you dont change anything, which is annoying, because soooo many times I overlook it and then my song would not play because it is in 32.
 
But why would you export in 32 bit? Do you record in 32 bit? Most of samples libraries are in 24 bit audio. I hear people record at 96 bit- not sure what advantage it gives other than huge file sizes.