Fantasia's 'Fantasound' Manual Recreation (WIP)

Yet another Fantasia related topic, however after assisting on the Fantasia LD restoration project and reading about what TonyWDA has been up to, I started thinking about Fantasound, and how currently the Dolby Surround matrixed version on the VHS/LD is the closest modern rendition that we have. From what I could piece together from various topics and writing histories regarding Fantasound is that it was essentially 3 channels with a control track, something like LEFT-CENTER-RIGHT (and likely SUR-LEFT and SUR-RIGHT (duplicate sounds of the front)).

Has anyone attempted taking the existing 2ch magnetic audio track and creating a new control track for it, similar to what Tony Porter did in 1990? The center channel (if a true center not just a mix of L+R) seems to have been lost, but could be “recreated” by an OOPS effect or something similar. And if you wanted an LFE track just do a 120hz cut on the L/R channels to create one

In Vegas Studio I can create a keyframed panning track, similar to the TOGAD which will send the sound to any speaker at any time. I have a link here to an .ac3 encoded file from Vegas Studio that is a rough example of how this could be done.

Therefore if this is possible the next step would to be, aside from listening to the 1990 audio over and over, is to find any other reviews,notes,or commentary about the original release and what people heard. For example, apparently the church bells at the end of the Bald Mountain came from the back of the theater, for example. Effects like these could easily be recreated. If there was an easier way to do it, I think it would open an fun world of different people creating their own Fantasound mixes for how they feel a certain piece should sound in surround.

https://mega.nz/file/CmoCwaYI#_bX3D23GU9nT7G9Vnt-Ac_OCNA_eAg3xecr5UEaXUAU
A simple panning example of the finale choir voices with the men on the left and women on the right then coming together again

https://mega.nz/file/viRAGB5Q#orQcNtFys3U3MkiwGzMs3lMnrsp-5RWj37zDoOBMrIo
The three french horns bit from Toccata, one of my favorite parts, the sound pans from L to C to R with each horn note

From reading http://www.widescreenmuseum.com/sound/fantasound1.htm I think there where 3 independent sound channels.
Each channel was directed towards a speaker (L C R Ls Rs) according to a control track. Basically they did the mix on playback.

The 3 original channels don’t correspond to any specific speakers. Channel 1 could be directed to L at one point, and later on to R. So you absolutely need the control track to recreate it.

In theory Dolby Surround can reproduce the original mix perfectly.
Modern signal processing gives you quite good separation, even with complex sounds.

Is it possible to mix Atmos without some expensive Dolby workstation? I bet you could do something really cool with a good Atmos mix to Fantasia, even if it’s not 100% true to the original Fantasound.

stwd4nder2 said:

Is it possible to mix Atmos without some expensive Dolby workstation? I bet you could do something really cool with a good Atmos mix to Fantasia, even if it’s not 100% true to the original Fantasound.

That would be interesting, another way that enthusiasts could mix their own versions. I don’t have Atmos, and I wonder how effective it will be with a stereo source but I think there is potential.

4throck said:

From reading http://www.widescreenmuseum.com/sound/fantasound1.htm I think there where 3 independent sound channels.
Each channel was directed towards a speaker (L C R Ls Rs) according to a control track. Basically they did the mix on playback.

The 3 original channels don’t correspond to any specific speakers. Channel 1 could be directed to L at one point, and later on to R. So you absolutely need the control track to recreate it.

In theory Dolby Surround can reproduce the original mix perfectly.
Modern signal processing gives you quite good separation, even with complex sounds.

Dolby Surround could work as an automated approach with it’s ‘steering’ to focus sound on whichever speaker has the loudest waveforms, which is possibly what was done in the 90s. I think a more live approach could be interesting.
While the technical approach is good for a background understanding, I feel what what he have left today, that a dolby-esque mix is about as good as we could get it. This would be going beyond simply creating a 5.1 mix which is what the DVD/BD did and was unsuccessful.

The gif link here shows how Vegas will pan the sound depending on my keyframes.
https://imgur.com/a/c8pKJv0

Starbond9 said:
This would be going beyond simply creating a 5.1 mix which is what the DVD/BD did and was unsuccessful.

The 2000 DVD audio mixes were 5.0; the 2010 Blu-ray mix was 7.1. Only the 2010 DVD and the Disney+ stream are truly 5.1.

I’d say the 2000 mixes are more faithful to the original Fantasound mix than the 2010 mix, even if the surround channels are less active than they are on LaserDisc. The 2010 mix is, from what I’ve heard, plagued with some pretty egregious noise reduction!

After reading through http://www.widescreenmuseum.com/sound/fantasound1.htm again it sounds like my scope is more of “remaking the dolby surround” track using a standard 5.0/5.1 channel setup. Since Terry Porter was working with the same two channels we have access to this would be like a refresh, trying to emulate the directional sound using an keyframed surround panner. As others have said, the DVD mix wasnt very good and the 7.1 mix wasnt much better, but I feel there is a lot of potential here. Still is a shame the original stems are long gone.

Although I am curious if anyone has recorded a modern classical performance for Surround/Atmos yet, because that would essentially be a modern Fantasound, at least the Mark II version. As said in the write-up “Separate channels recorded close pick-ups of violins, cellos and basses, violas, brass, woodwinds, and tympani. The seventh channel recorded a mixture of the first six channels and the eighth channel recorded a distant pick-up of the entire orchestra.” Seems like that would be very easy on a recording today.

Look:

Mark I - 3 front horns, two rear…essentially a modern day 5.0 system
Mark II - 3 front horns, two rear, two sides, one top… a hybrid 7.0/atmos system
Mark IV - automated TOGAD of Mark II
Mark VI - reduced back to 3 front horns. First Fantasia Dub
Mark X - added back the two rear horns to the Mark VI system, and used on roadshows…again a 5.0 mix

You might be confusing playback with recording tracks.

The 7 tracks were used for studio recording only, and then mixed down to 3 on the actual movie.
So you have 3 tracks + control (http://www.widescreenmuseum.com/sound/fantasound1.htm)
“four photocells which scan three program tracks and a pilot control-track”

The control track was there to extend the dynamic range of the audio tracks. Basically it did volume/noise control.
The manual panning was done on top of that, directing each of the 3 audio tracks towards specific speaker(s).

The 3 program tracks are not LRC, they can be any arbitrary channel combination. For example Ls, C, R or Ls, Rs, R

Yes you are correct and I wasn’t trying to conflate the two. But the idea still stands that for Fantasia they recorded 8 individual tracks to send to whatever speaker they wanted to. In a modern recording a sound engineer might just decide on a permanent playback speaker depending on where the instrument is physically located in the soundspace. But thats not to say they couldnt feed, say a woodwinds mic, through the left then right then rear etc. to create movingc clarinets. Its very possible today but A LOT more work and ultimately based on a creative decision.

So I’ve tried to distill what is missing from today vs then. If you take it down to the most basic elements you had

8 input tracks

• 1-6, instrument pickups
• 7, mix of 1-6
• 8, a distant mic

3 output speakers(or 5 if you used Mark X)

• Left
• Center
• Right

What we have today is a mono track (possibly a mix of input track 7 and 8) and a stereo track of which I don’t know how the mix down was performed from the initial tracks.

Ultimately the point being that the dolby 1990 mix is the best separation we’ll get but there is still potential for a 5.0 true remake of it on modern computers, which would be more accurate than the DVD version.

Im definitely following this. BTW… the original NY presentation had speakers all down the side and back. The live panning could go anywhere in the room.

TonyWDA said:
The “centre” music channel, as it was originally presented, does exist on the magnetic stereo releases. I was given three discrete audio tracks— left, centre, and right— when my print was captured. It’s why the 2000 and 2010 home video releases also have discrete centre channels. Disney didn’t artificially generate that with some plug-in or standalone software, it’s always existed, and it does contain information unique to that channel; it’s not just a cheap sum of the left and right signals.

Those are the 3 original channels that were manually panned towards the speakers.
Just to be clear, those 3 channels are not L, C and R. They are 3 discrete arbitrary sound channels.
The mix would direct them to whatever speaker combination was desired.

So you could have “left” playing simultaneously on your SR and FL speakers, “center” directed to SL and “right” to C. All combinations are possible but you only have 3 discrete source channels playing at any given time.

It’s my interpretation of the system and might not be 100% correct, but hope it helps.