Comments

TheHappyFriar wrote on 1/12/2004, 8:50 AM
If the source is stereo, you probley don't want to mix it specificly into surround (it was ,after all, designed for stereo and sounds best in it). If you just want to use the rear 2 speakers in surround with conjuction with the front two (front/rear left and front/rear right aproduct the same sounds), just do what you said. You don't really need a signal delay or reverb. That would be determined by the sound system (my sound system lets you input the feet between the speakers and listener. It then delays sounds approperatly).
TorS wrote on 1/12/2004, 9:58 AM
Here's some info:
The Dolby tech page
Tor
Maverick wrote on 1/12/2004, 11:05 AM
Have a look at this link which gives some detail on how a mono track was converted to 5.1.

http://freespace.virgin.net/anthony.clark2/recon.htm

Perhaps the ideas can be adapted for stereo tracks.

HTH.

jeremyk wrote on 1/12/2004, 11:30 AM
This is from the Dolby tech page that Tor gave a link to. Basically, they discourage most up-conversion from stereo for reasons that seem fairly plausible...
___

It is best to transmit two-channel matrix-encoded material as such to the consumer, and not decode it in the studio to derive an "artificial" discrete multichannel mix. For one thing, the results can vary from the intent of the original producer/director, especially if extra processing has been added on top of the surround decoding in the studio. Also, the sound will fail to meet listener expectations for true 5.1-channel material.

In addition, many consumers do not yet have a full 5.1-channel Dolby Digital playback systems, and connect their DVD players' two-channel outputs to a stereo or Dolby Surround Pro Logic system. Under these circumstances, the Dolby Digital decoders in the players "downmix" multichannel soundtracks "on the fly" to mono, stereo, or Dolby Surround matrix-encoded stereo. If the multichannel soundtrack was derived from a matrix-encoded two-track master, there may be downmix quality problems due to phase effects from the original matrix encoding. In nearly every case the original Lt/Rt mix will give the best results.

When upmixing from an Lt/Rt master is absolutely necessary, it is best to use a matrix decoder without steering or even a simple "Hafler" matrix. Any steering logic, including that used by Dolby Pro Logic, adds some degree of "swimminess" that is not welcome, particularly for pure music applications. Of course, due consideration must also be given to downmixing playback situations, suggesting, for example, that no time delays be used in the surrounds.
Spot|DSE wrote on 1/12/2004, 3:26 PM
While it's not a recommended practice if you have a finished, existing stereo mix, there is a good way to do this, if you are insistent on having 5.1 that won't mess with your downmix on older systems.
Take the stereo mix, duplicate it on a new track.
Change project to 5.1 vs stereo.
Find the KEY points of the mix you'd like to have in the surround. This might be jets, ambiences, stereo music at key points that dialog doesn't exist.
Use the "S" key or a volume envelope to cut these desired sections from the rest of the second stereo track. Delete unwanted audio. Use volume envelopes or fades on the remaining pieces, and use the panning keyframes to move those pieces from side to side, rear only, or rear to front if you need motion.
Duplicate the original track again.
Cut out the MAIN dialogs on the new track. Use 5.1 location window to locate in center only, usually at -3dB.
Duplicate again.
Use the 5.1 location window to make this an LFE only track. Use compression and EQ to enhance the overall low frequencies.

Be sure you have a good monitoring system to do all this with. Otherwise, it's a waste of time no matter what.