Thanks tons Noel.
 
Michael

Now that's quick service.

since I first replied I've done some more reading on the subject. It just shows. I've been involved in audio nearly all my life and wasn't aware that such files existed. Is there an advantage other than a form for archiving, digital transporting, compression. For example is the single wav smaller or larger than the sum of the individual audio tracks/channels it contains. It doesn't seem likely that any device or software can play the file back as is.

Hi Mike,
 
This is a little known and even less understood approach to musical composition. I basically program the computer to do the composing, which I then edit. I feel it is the next step in musical composition. Because I can do so much more when assisted by the computer, the music has an exponentially large range of possibility. So the music sounds foreign to people right now.  I use AI to extend the possibility of music. (AI in this case stands for Augmented Intelligence, not Artificial Intelligence.)
 
I render these files "natively" from Csound in multi-channel formats. I write algorithms that, among many other things, create spatial relationships amongst the sounds. For home playback I often mix these compositions down to 4.1. But they are intended for the concert hall where they are played back, usually, using Pro Tools. I also play the multi-channel files for art installations and for that I use Sonar or Ableton Live. 
 
One of my areas of focus is 3D spatialization, where I not only have speakers left and right (1d - stereo), front and back (2d - quad) but also top and bottom (3d). So all three spatial dimensions are used. I programmed a system of Cartesian Coordinates in Csound that stipulate where, within the 3d Cube, a sound exists and if it is to be perceived as moving, how it moves... 
 
So, as you can see, I cannot render a 144 channel file in separate channels any more than one could do so with a stereo file. It is not really practical.
 
You can find out more about my work at http://www.perceptionfactory.com
 
Thanks for your interest and I hope that answers your question,
 
Michael

I looked into it. It turns out that this is a bug in Microsoft's multi channel wave file handling. For multichannel wave import we use a component from Directshow called a wave parser. This is not handling more than 24 channels for some bizarre reason and reporting a mono stream instead. Thats why it gets a mono file.
There is some discussion on the web about this as well. I've reported it also to MS and we'll see what they say.
I can work around it using our native import but it will take some rework to do so. Will keep you posted.
 
One thing you can try is rendering the files as WAVEFORMATEXTENSIBLE wave files. Microsoft tends to handle these files better (Even 24 bit normal waves are not understood by Windows without this header) so it might work in that format. Windows normally identifies multichannel audio files as being surround format, so even with WAVEFORMATEXTENSIBLE they could have some limits based on what they think is a valid surround config.
 
I'm not sure if CSound supports export to WAVEFORMATEXTENSIBLE  but you can look into it.

That all sounds spot on Noel. I did not realize that Sonar uses an aspect of Directshow to import sound files. That explains why ProTools can do it... they probably use their own protocols.
 
I am checking into Csound supporting WAVEFORMEXTENSIBLE but from what I see in the manuals and and other documentation it does not look like it does. Sounds like it would be a workaround.... but if it worked that is all that matters.
 
Thanks a lot.
 
Michael
 
 

We use directshow only for a few functions as a fallback since it has support for a bunch of formats. It turns out that the multichannel wave case relied on that code. I tried implementing the import natively and its definitely doable.
It requires a bunch of changes to the import code to handle deinterleaving to mono files. Not sure which update I can do it but it should definitely be possible.

This was very indicative of why I have stuck with CAKE since the late 80s.  You guys always go the extra mile.  This feature does nothing for me but the fact that you (Noel) jumped on this as an opportunity is really impressive.
 
Thanks a TON for all you do from one end user,
Jim

Michael I was able to rework the code to handle this natively. Your 48 channel file now imports ok.
It will require a bunch of testing so it might take awhile before you see the fix but it works now...

Wow. That's great. I wonder though with so few affected by this issue, actually just the one person we know of, why this issue has got such immediate attention. Just asking for understanding!

There are hardware vendors who also write multichannel audio files like TASCAM and Roland. 
Also we try and remove dependencies on problematic 3rd party API's as far as possible and this item was already on my dolist. In general 3rd party compatibility esp with file I/O is a high priority item with us.