What were the model numbers of the Intel motherboards ?
*******
Your motherboard has an Analog Devices HDaudio codec, either AD1988A or AD1988B.
It isn't RealTek.
One of the few motherboard chipsets to support real time hardware AC-3 encoding,
was the NForce2 (SoundStorm). That is a low latency solution to compressing 5.1
audio to fit on the two digital channels of S/PDIF. (S/PDIF can support four
channels at reduced resolution, but it not normally used that way.)
http://en.wikipedia.org/wiki/Dolby_Digital
http://en.wikipedia.org/wiki/SoundStorm
There are some sound cards with a function like that. At least one Korean
sound card had real time encoding in hardware (it was one of the few reasons
to buy the card). And the above article mentions some Creative cards as well.
More of the sound solutions on motherboards, may have included a software
encoder for doing AC-3 over S/PDIF. Such a scheme had a latency of 0.5 seconds,
making the solution useless for general PC usage. It would be no good for gaming.
It would be acceptable for video, as long as the video was delayed by 0.5 seconds
as well. The delay was likely fixed at some value, to make it easier to compensate
later. Otherwise, the encoding delay would likely be shorter but more variable.
Maybe you had a software encoder on your other motherboards ?
I see no evidence in the AD1988A/AD1988B documentation, that there is any
active support for the functions. If a motherboard company pays for a Dolby driver
(effectively paying for a Dolby license), then the function may exist. But
Dolby Labs survives, partly by means of the licensing fees it collects.
I don't know what the terms of the license for each of their trademark
technologies entails, but my bet would be that an "encoder" function
costs money. And if a software encoder is used, there has to be a way to
trace that a per-unit license fee was paid.
That doesn't stop you from finding a free-software encoder and installing
that. But any legitimate companies would be approached by Dolby lawyers,
if they did encoding functions without paying a license for it. I can't
predict what the lowest latency of a software encoder could be, with a
modern processor. I don't think the latency value has to be 0.5 seconds.
The practical limits are likely lower than that. But I don't know if
they're low enough, to support good gaming audio over S/PDIF.
*******
From the comments section near the bottom.
http://www.anandtech.com/mb/showdoc.aspx?i=2358&cp=4
"The only difference between a Sound Blaster card and an audio card that
has a real time encoder, is that you can make a one-wire, digital connection
from your audio card to your home theater receiver and enjoy discrete
multichannel sound from the game. However there will be a continuous,
slight delay, known as "latency", as the encoder is creating and
transmitting the bitstream, and of course the compression scheme
being used is "lossy" (i.e. not bit-accurate)."
Based on that, converting everything to AC-3, if connecting to another
sound editing system, seems counterproductive. Analog transmission
might well be better, with shielded cables. While not bit-accurate
(as the other system may have to use its ADC), at least there isn't
a predictable loss of accuracy due to the compression algorithm.
Sending two channels of audio over S/PDIF, should be lossless and
bit-accurate (unless, of course, the playback chain has been altered
by the driver, which apparently has happened in some implementations).
It actually takes a lot of work, to prove that digital audio sent from
a computer, to an external digital audio device, has not been tampered
with. Sometimes, there are unnecessary re-sampling functions or
unnecessary truncations of the least significant bits. A lot of consumers
are blissfully unaware their digital audio path is not bit-accurate. Apparently
they can't hear the difference :-)
Paul
It's been a while since I looked for a software encoder, but I expect
there is one out there somewhere.
At one time, Google would allow me to search for my own postings,
now over 20,000. Google is now broken, with respect to personal
searching. So there is no hope of me finding a previous post
I might have made, about AC-3. So I cannot even review my previous
notes, using Google.
If a Creative Labs sound card continues to have such a capability,
that would be another option. Add a sound card and use whatever AC-3
encoder function it might employ. The advantage there could be, that
the latency would be low enough for general usage.
Or you could go in search of that Korean card.
This will give you some idea how "transparent" AC-3 is :-)
It also explains why there has to be some latency in it.
The encoder has to view the dynamic content to pick the
right coder options at any instant in time.
http://www.mp3-tech.org/programmer/docs/ac3-flex.pdf
*******
I tried looking on Newegg for a card, and noticed HT-Omega was
using CMedia chips. So I looked here, to see if any have encoder
functions. A couple of chips do.
http://www.cmedia.com.tw/pci_audio_new.html
HT-Omega card - seems to have the odd driver issue.
Hard to tell what you're getting for the money.
Has the chip that can encode AC-3, but you'd need to
find a better product description to understand whether
it is working or not. or alternately, look for other sound
card companies using one of the CMedia chips.
http://www.newegg.com/Product/Product.aspx?Item=N82E16829271001
This Creative card may have the capability. This one is PCI Express x1. $99
http://www.newegg.com/Product/Product.aspx?Item=N82E16829102024
Windows 7 will likely add more variables to your situation,
as some of the smaller companies making sound cards, may not
be able to "keep up" from a driver perspective.
Paul