Just to further clarify (or confuse) Tom's comments:

There are actually a few different stages which come in to play on a  
typical film mix, some which could perhaps be construed as  
"sound-on-sound", depending on how you want to view the overall process.

It was common practice (even in fairly early days), to do premixes of  
various elements, depending on their complexity. Dialogue pre-dubs and  
FX pre-dubs were the most common. This would allow the mixers to get  
the track count down to a manageable size, and take a lot or the  
tedium of basic fixes out of the final dub. Otherwise, you would end  
up with a track spread that was simply unmanageable.

Depending on the final format of the mix, these pre-dubs could be  
mono, 3 track L/C/R, 4 track L/C/R/S, or 6 track (various flavors).

However, the original pre-dub elements would usually be kept on hand  
if the director didn't like some part of the pre-dub mix, allowing the  
mixers to go back and tweak the pre-dub decisions (although, since it  
took quite a bit of time to re-hang all the elements when working in  
35mm analog, so it was avoided if at all possible).

In the current digital workflow scenario used on most major Hollywood  
features, there is still a pre-dub stage (again, mostly for dialogue  
and FX), but it can usually be "de-constructed" when need be, as all  
the original tracks ae usually kept on the timeline. (It is not  
unusual to have a track count of 500+ on even a routine film nowadays,  
which would have given mixers back in the day a heart attack!).

As Tom points out, the guys who did all this work back in the days of  
optical sound recording were incredible craftsmen (no women ever sat  
in the mixer's position until at least the early 1980's). The pressure  
that these guys worked under was just unbelievable. You could get down  
to 950 feet in a mix, and if you blew a cue, the master negative would  
usually be scrapped, and you would start all over. (There was no  
reverse selsyn system back in those days). Not unsurprisingly, many of  
these guys either left the business, or turned into fairly hard core  

Sadly, while we no longer have to contend with optical sound  
re-recording and forward-only interlock systems, the hours and  
conditions the post-production crews work under currently make the  
1940's look like a cakewalk.

Scott D. Smith

Quoting Tom Fine <[log in to unmask]>:

> Hi Aaron:
> What they were doing in Hollywood, from the early days, was recording
> different aspects of the final soundtrack on different bits of film and
> then mixing together from motor-sync'd playback to a final sound
> master. There were crude mixing consoles from early in the electronic
> recording days, too. One specific example I was told about, and I'll
> ask the guy for the film title because I don't remember it, was the
> final music was mixed from three optical elements, one made from each
> microphone, with each microphone focused on a different musician or
> group of musicians. This would be very similar to live-in-the-studio
> multi-tracking. They were also able to pre-record music tracks very
> early, so a singer on film would be singing against a playback. And
> lip-sync'ing and indeed orchestra play-sync'ing were developed early
> on, too. By the early 1930's, Western Electric (and probably others)
> had developed amplifier and mixer-network systems allowing for mixing
> many different sound elements into a final soundtrack. Also, the whole
> idea of "stem" mixes came out of Hollywood, a way to reduce many
> elements to a few logically organized stems for final mixdown. By the
> 1940s, the major studios' sound departments had big 3-person consoles
> for final mixing (dialog, music, sound effects). Those guys were aces,
> too. Think of the mono soundtracks for some of the big musical
> pictures, that's a very complex sound universe to fit into one channel.
> -- Tom Fine
> ----- Original Message ----- From: "Aaron Levinson"
> <[log in to unmask]>
> To: <[log in to unmask]>
> Sent: Wednesday, June 24, 2009 6:38 PM
> Subject: Re: [ARSCLIST] Recording Innovations
>> I agree that Les Paul takes undue credit for many things but what   
>> Tom describes as multi-track recording in Hollywood is not strictly  
>>  speaking correct. To me multi-tracking means being able to change   
>> separate levels AFTER the process, what he is describing is more   
>> like sound-on-sound as opposed to multi-tracking as we commonly   
>> understand it today. The same is true of Mike Biel's assertion   
>> adding a sound or a voice to an already existing recording, this   
>> involves a generational loss whereas with multi-tracking and   
>> overdubbing as we employ it today it does not. But sound-on-sound,   
>> stereo and a bunch of other so-called modern techniques clearly had  
>>  their unique antecedents which should be accorded their due. I   
>> nevertheless stand by my basic assertion that the reason for so   
>> many alternate takes was the recording process of the 78 era. I am   
>> well aware that some exceptions do exist and I apologize for not   
>> duly noting them.
>> AA
>> Tom Fine wrote:
>>> While the general gist of what Aaron said is true (MOST sessions   
>>> were done live and MOST for-profit record labels did not want to   
>>> pay for elaborate overdub or punch-in stuff if it was avoidable),   
>>> Mike is right about Les Paul inventing very little, by any   
>>> reasonable definition of inventing. However, Paul is indeed a   
>>> superb musician with an innovative mind. I wish he wouldn't "take   
>>> credit" for so many other people's hard work, since he's done   
>>> plenty that he can legitimately take credit for.
>>> Anyway, Mike, how did Edison do "overdubbing"? Did he use some   
>>> sort of acoustic mixing system or just play a cylinder into the   
>>> room at the same time live sound was being made, with the horn   
>>> picking up both?
>>> As for multi-tracking, just about as soon as electronic-optical   
>>> recording hit Hollywood, people were figuring out how to mix   
>>> sprocket-synchronized sounds. There were multiple sound elements   
>>> to some very early optical-sound pictures. At least that was told   
>>> to me by a restoration guy who has done some very high-profile   
>>> films.
>>> -- Tom Fine
>>> ----- Original Message ----- From: "Michael Biel" <[log in to unmask]>
>>> To: <[log in to unmask]>
>>> Sent: Wednesday, June 24, 2009 12:59 PM
>>> Subject: [ARSCLIST] Recording Innovations (was: take numbers on   
>>> emerson records)
>>> From: Aaron Levinson <[log in to unmask]>
>>>> I for one am not at all surprised by numerous alternate
>>>> takes in the 78 era, it makes perfect sense. Anyone that
>>>> makes records, and Tom will back me up on this, knows that
>>>> even in the era of multi-tracking takes can have a very
>>>> different feel if not outright errors. Everything was
>>>> live pre-Les Paul so no "punching" was possible.
>>> I wish people would stop giving Les Paul more credit than he is due.  He
>>> was not the first to do overdubbing, he was not the first to do
>>> multi-tracking, and punch-in editing was not one of his things in the
>>> early years.  He is an extraordinarily talented musician with a
>>> fantastically innovative mind, but his knack is to adapt new technology
>>> and expand on past techniques.
>>> It is not true that everything was live before Les Paul.  Even Edison
>>> did overdubbing on tinfoil!!!!!!!  I am not kidding.  This is the
>>> absolute, well documented, truth.  Just this weekend Dave Weiner showed
>>> a film at the Jazz Bash that showed a violinist playing a trio with
>>> himself in the 1930s -- both sound and picture.  Voice over-dubbing was
>>> common.  Adding instrumental tracks was common.  Editing in and out of
>>> music -- punch-ins -- was common.  I challenge you to show me anything
>>> Les Paul did that had not been done before.  And you have to realize
>>> that by the late 1930s even many 78s by companies beyond Edison and
>>> Pathe (who had done it back to the turn of the century) were dubs, not
>>> recorded direct-to-disc.
>>>> The players wanted it to be right and at that time the only way
>>>> to insure that was to play it again Sam.  AA
>>> It was not the ONLY way, it was just the usual way.  I have been playing
>>> records for sixty years and have been researching the technology of
>>> recording for fifty, and one thing I have learned is to never think that
>>> something had never been done before.  I am still constantly surprised
>>> by discoveries of earlier technologies.  All too often when a statement
>>> is made "This is the first time . . ." it really should have been a
>>> question "Was this the first time . . . ?"
>>> Mike Biel