DTS encoding - rear surround

[MaroonMike]

First of all, I'm a newbie to DTS encoding. I have been working on getting the 6 mono waves files sounding right for the DTS mix from a 2 channel stereo music WAV file.

My question is about the rear channels - should the rear surround channels include BOTH left and right stereo channels or should they be separated into distinct left/right mono mix channels (like I did for the Front channels.)

I think I have read folks doing it both ways and am a little unsure at which is best (or proper protocol.)

I also added a slight Hall Reverb effect to the rear channels which sounds great. Is it normal to add effects to the rear surround channels, or am I doing something completely unorthodox?

Thanks,
Mike

[jorel]

hy MaroonMike,
welcome!
:)

unorthodox?
no,...question of taste,
some people like Eric Clapton, others Jimi Hendrix(me include)!
:D

seems logical(only my opinion)that the surround
should be separated into distinct left/right cos
after the "slight Hall Reverb effect" that you use
give more emphasis to each channel.
if you include in both channels
it will decrease the "separations"!

but how do you mix the center and lfe channels?
(i have "my way"....just curious)

[Umma]

There IS no "proper protocol" when it comes to doing this. :D

Whatever sounds good. You may find that some CDs sound better with the left/right stereo in the front and back, or some sound better with the surround fields of ambisonics. Pink Floyd, for example, works both ways sometimes. THE WALL sounds great either way. The reverb in the back should not sound bad as long as it enhances the music. Some people do not like having the music sound the same (with equal gain/mix) front and back, but that is a matter of personal taste.

Getting a great sound experience is the ultimate goal.

[MaroonMike]

Here is what I did using Goldwave starting with a 32bit stereo WAV file:

FL = 100% left channel, mono 32bit convert

FR = 100% right channel, mono 32bit convert

Center = 50% left and 50% right mix (I actualy used a command called
left/right channel average mix to do this), mono 32bit convert

LFE = 50% left and 50% right mix (left/right channel average mix), lowband filter at 120hz, volume reduction by 2%, mono 32bit convert

SL = 100% left channel, highband filter at 100hz, slight Hall reverb effect added, mono 32bit convert

SR = 100% right channel, highband filter at 100hz, slight Hall reverb effect added, mono 32bit convert

This sounded pretty good on my system. I will eventually hear it on others to see if it is a good template.

[TimBoz]

My Method
Aimed at creating a general acceptable 5.1 production from stereo sound to play on an average DVD player with an average surround sound system in an average room - there is no such thing but you know what I mean - I don't want to spend a week perfecting the sound so much so that I am sick of it and never want to hear it played again.
It sounds okay and is based on MaroonMike's suggestions. I use Goldwave with a similar procedure which decompresses and then I use PCM only, compression is only done again for the final 5.1 ac.3 mix which I believe minimises quality loss.
To maximise video & audio sync I usually use Goldwave to extract audio from the source AVI and resave it at signed 16 bit PCM (CD quality?). I then use AVIMux to patch the original AVI and the CBR audio together which allows editing of the new AVI with sound in VirtualDub.
Using Goldwave to open the new AVI file I extract the audio. From here I:

Cleanup original audio:
Effect>Filter>NoiseReduction>Preset>Hiss removal
Effect>Resample to 48000khz (if not current format)
Effect>Volume>Maximise>Preset>Full Dynamic range
Save as STEREOMIX.wav PCM signed 16 bit stereo

The settings I use are below, though a little time spent creating presets with the Effect Chain tool will speed up the process though should be placed in the same order of processing.

Create 5 channels:
Left front & rear channels (from STEREOMIX.wav):
Front:
Effect>Stereo>ChannelMix>Preset>MonoLeft
Effect>Filter>Equaliser>Preset>Equal loudness
File>Save as>left.wav PCM signed 16 bit mono

Rear:
Effect>Filter>Low/Highpass>select Highpass radio button>initial cutoff 100hz
Effect>Reverb>OK
File>Save as>rearLeft.wav PCM signed 16 bit mono


Right front & rear channels (from STEREOMIX.wav):
Front:
Effect>Stereo>ChannelMix>Preset>MonoRight
Effect>Filter>Equaliser>Preset>Equal loudness
File>Save as>right.wav PCM signed 16 bit mono

Rear:
Effect>Filter>Low/Highpass>select Highpass radio button>initial cutoff 100hz
Effect>Reverb>OK
File>Save as>rearRight.wav PCM signed 16 bit mono

Centre channel (from STEREOMIX.wav):
Effect>Stereo>ChannelMix>Preset>Double Vocals
Effect>Stereo>ChannelMix>Preset>MonoMix>OK
Effect>Volume>Change>-8db
Effect>Filter>Equaliser>Preset>Equal loudness
Effect>Bandpass/stop>From:150hz To 12000hz
File>Save as>centre.wav PCM signed 16 bit mono

I then use Premiere Pro to mix the 5 wav files and auto-generate the LFE and create the ac3 5.1 audio. I have found that Surcode gets 'stuck' if the wav files are not of the exact same time length (to milli seconds) so use Goldwave to Trim or Add silence so that this is the case before using these 5 files with Premiere. Again Surcode has an upper limit on bitrate which is why 16 bit has been used as a format for each of the surround channels.
I then import the created ac3 5.1 audio to Encore, delete the original soundtrack and place the ac3 in its place for the soundtrack timeline.

Ideally I would like to remove vocals from all but the centre.wav, though I think this removes some of the quality/dynamics.