Quote:
Originally Posted by ninjikiran /img/forum/go_quote.gif
Bit my finger and ordered one, can't wait to try it out.
Nice write up, looking forward to seeing your work. Would also like some of your wisdom being you are a more experienced developer
I would assume following your method might see some improvement in the audio stream considering it might only be re-sampled once in windows rather than once in windows (16/44.1k) then again in the device (24/92k). Just an uneducated assumption. Since any form of processing has the potential to corrupt the stream
|
Thanks for the compliment. There are quite a few more details in what I have learned in playing around with Microsoft's newest implementation of the DirectX (DirectSound in this case) replacement called "Microsoft Universal Audio Architecture (UAA)" which is designed to simplify connecting the audio file to the device interface.
I believe that setting the USB device in an application such as J.River's Media Center to 24/96 it is helping to bypass the restriction imposed at the hardware level to 16/48 as Direct Sound in Vista or greater (2008 or Win7) the data stream may be able to send at 44.1,48, or 96 as required since we are bypassing the 16/48 limitation set through Control panel/Device Manager.
I have some USB packet analyzer code that I am going to play with next that will help to confirm or reject that assumption and I will have time later this week to do this.
As a concept you should start to think of the hardware layer as being a separate entity altogether from the software (device driver/filter/codec) level as MS has been pushing to separate the two once the divorce from KMixer took place in going from XP to Vista. The new specifications for the UAA design calls for up to 32 Bit resolution (up from 24 Bit) and ability to operate at 32 OR 64 Bit (helps with HD Audio with many streams).
They have moved from the restrictions of having a device driver tied to a PCI or USB bus to an abstraction above this allowing XBox, Win, Maybe Zune, etc. to be utilized with the same code base.
I will let all know what I find going across the USB DeviceID for the D1 with the 24/96 settings with a Redbook 16/44.1 file being played very soon.
I am still thinking that it would help if I were to write an Audio Driver for Windows that a person would install if using Vista or Win7 and the D1 would be assigned to it upon audio stream initialization bypassing the WASAPI/ASIO/Direct Sound choice you would normally make in your media players and instead you would pick the D1 output instead.
Sorry if this is not detailed enough for some. It would take quite a bit of writing to detail this at this time and I need to do the first steps to see if this is even achievable before investing in the detail. Right now this is just wild theories without evidence to back it up. But, I promise to share success or failure along the way, and give details as more is known.