"Technically augmented stereo"

[Restoration73]

One early version of stereo on tape used two staggered heads, and the recording would
exhibit channel delay if played on a standard in-line stereo head.

[Ted Kendall]

The error from playing staggered tapes on a stacked head is of course gross, but in practice staggered-head machines didn't give very satisfactory performance either - the manufacturing tolerances on head spacing ensured that there was nearly always substantial phase shift between machines. Add in the varying tape stretch caused by tension variation in the transports of the time, and the technique was doomed to failure. EMI's first stereo rig used modified BTR-1s with staggered heads, and suffered both these problems. Early stereo systems seemed to be far more concerned with channel separation than time alignment, regardless of the fact that much over 30 dB is superfluous - the "crosstalk" of human ears is worse than this.

[barretter]

Originally Posted by Station X

I'm sure I have some LP's somewhere which are labelled something like "Mono recording electronically enhanced for stereo". I'll have a look and a listen.

You will find many of the recordings reissued in the Decca Eclipse series fall into this category though, paradoxically, some of the stereo recordings which Decca had released only in mono before the advent of stereo LPs in 1958, were released in stereo for the first time (in the UK) on Eclipse. A notable example is the Knappertsbusch version of Wagner's Flying Dutchman.
At least Decca were honest about what they were doing.

[Restoration73]

Capitol records actually developed a fake stereo system with the most famous release
being "Pet Sounds" - a proper stereo version from the original multitrack studio tapes
only appeared some 3 decades after the original release.

[barretter]

What do you mean by "proper stereo"?

[TIMTAPE]

I believe Pet Sounds by the Beach Boys in 1966 was mixed to mono for more than one reason. Brian Wilson was deaf in one ear. He couldn't hear a stereo mix. He didn't want the final sound changed due to someone accidentally hearing only the left speaker for example. Then in the day, AM radio where most of the air play happened was mono. That was the listening market they targeted.

But in production of Pet Sounds, many separate tracks of voices and instruments were recorded on 4 and 8 track tape machines. Once those separate tape tracks were mixed to another recorder, some were discarded but apparently many survived. That is what made a true stereo mix possible many years later if they still had enough separate "stems" from which the original mono mixes had been made.

This is true of many other recordings where as normal procedure the multitracks tapes were stored in vaults.

Another help was digital tools. Even though tracks on separate tapes were not 'synced' to each other, their lengths could be digitally stretched or shrunk to sync up, although not always perfectly.

Sometimes it's also possible to extract say a single voice by subtraction. If you have a mix of voices A and B and you have a tape with just voice B, subtracting voice B from A + B, gives you voice A. I believe that technique was also used, but it's a lot more difficult if the tracks are on separate reels as timing problems can be critical.

Today we can buy "music rebalance" software designed to allow some degree of separation of tracks from a mixed recording but it's only a partial solution. I don't think it's possible to create a true stereo mix from a mono original. We still need all or some of the separate tracks to work with, as they had with Pet Sounds.

Mark Linnet was the engineer who helped with this work in I think 1997.

https://en.wikipedia.org/wiki/Pet_So...and_production

[Ted Kendall]

Originally Posted by TIMTAPE

I don't think it's possible to create a true stereo mix from a mono original.

Quite so - it isn't. The information simply doesn't exist.

[rontech]

My experience of playing 78 rpm records using a stereo Shure M55 with 78 stylus may be of interest. The music content appears as a well defined central mono acoustic image. The surface noise appears to be completely separated and emanate from the left and right loudspeakers.

Nowadays I use a proper Ortofon mono 78 cartridge and everything is centally located.

I think the separation effect with a stereo cartridge is psycho acoustic,

[bluepilot]

The noise on a mint shellac record is surprisingly low. Most of the noise you hear is damage caused by the needles when the record was played on an old gramophone. There's no reason why the damage should be the same in both channels. If you use a stereo cartridge you will hear stereo noise whereas the music will be in the middle. If you use a mono cartridge then obviously everything will be in the middle.

[Herald1360]

Would this effect provide a useful way to reduce the surface noise level in a mono remix?

[Sideband]

Going back to the DES system of 'spectral editing' I was very impressed by this stereo remix of The Animals 'House of the Rising Sun'. This was never released in stereo. This is very convincing and if no-one knew any different, would think it was a stereo master.

https://www.youtube.com/watch?v=qqVRx3Zz7KY

I believe Eric Records have also done a stereo remix of Procul Harum's Whiter Shade Of Pale also never released in stereo.

[TIMTAPE]

Originally Posted by Sideband

Going back to the DES system of 'spectral editing' I was very impressed by this stereo remix of The Animals 'House of the Rising Sun'. This was never released in stereo. This is very convincing and if no-one knew any different, would think it was a stereo master.

https://www.youtube.com/watch?v=qqVRx3Zz7KY...

Maybe it is a stereo master, originally recorded in the studio to at least two tracks.

[ms660]

So far as I know the Animals House of The Rising Sun was a one take mono recording, I think there was a fair amount of wow in the original 45 rpm pressings.

Lawrence.

[TIMTAPE]

Yes that's also what I just read in an interview with Mickie Most the original producer. House of the Rising Sun, straight to mono although he mentioned two takes, not one. The guys travelled down to a London studio for the session.

So I guess the stereo version on YT was done with a "Music Rebalance" type of processing. It certainly has separation of the four instruments and the vocal.

It'd be good to compare it directly with the mono source version to hear what if any side effects were generated. I heard some nasty distortion in the vocal from the stereo version but there was some distortion in the mono versions I heard as well. I believe there are limits to the quality that can be achieved depending on how ambitious is the task.

I believe Music Rebalance was a development of an earlier tool called Melodyne which had the ability to recognize the related harmonics in a sung voice or musical instrument. Speaking of wow, there's a related tool called Capstan which can help remove wow from a recording so long as it's music, not speech.

[Ted Kendall]

Originally Posted by Herald1360

Would this effect provide a useful way to reduce the surface noise level in a mono remix?

Packburn and others played a mono groove with a stereo pickup, using a sidechain to select the quieter groove wall at any given instant, to great effect.

[cheerfulcharlie]

Originally Posted by TIMTAPE

Maybe it is a stereo master, originally recorded in the studio to at least two tracks.

By that time record producers were starting to produce a mono mix for 45s and a stereo mix for stereo LPs..simply because (as you say) they had 2 or 4 tracks to work with.
It seemed with the early Beatles that they were mainly interested in the mono mix as the stereo versions usually just had vocals plonked on one side and say a guitar on the other side.

This is an example (I think) of an ephemerial pop tune that spent it's early life in mono, but being a multitrack recording it could be appreciated in stereo in the CD age .

https://www.youtube.com/watch?v=tLuEhYElzrs


I think 'Nights in White Satin' was one of the first 45's to be issued in stereo in the UK?

[dave cox]

I rather suspect it is possible to create fake stereo from mono recording using neural-network processing. It is true that you can't extract something that isn't there from a mono recording but I don't see any reason why a network can't be trained to recognise instruments and extract them to separate 'tracks' and remove them from the original by subtraction. Then you'll need a good sound engineer to mix that down to stereo

I don't know if anyone has actually done this ...

dc

[rontech]

Originally Posted by bluepilot

The noise on a mint shellac record is surprisingly low. Most of the noise you hear is damage caused by the needles when the record was played on an old gramophone. There's no reason why the damage should be the same in both channels. If you use a stereo cartridge you will hear stereo noise whereas the music will be in the middle. If you use a mono cartridge then obviously everything will be in the middle.

I agree about mint shellac records.As a teenager ( 1956) I had a vacation job in an amusment arcade. They changed the juke box from a78' to a 45's machine whilst I was there. I "rescued" all the redundant 78's. The "B" sides of these records were unplayed and were surprisingly low noise.

[rontech]

Originally Posted by Herald1360

Would this effect provide a useful way to reduce the surface noise level in a mono remix?

That is what I thought when I first noticed the effect. However I feel that the central image is a construct of the ear / brain system and has no real existence.

Discuss!

[bluepilot]

Originally Posted by dave cox

I rather suspect it is possible to create fake stereo from mono recording using neural-network processing. It is true that you can't extract something that isn't there from a mono recording but I don't see any reason why a network can't be trained to recognise instruments and extract them to separate 'tracks' and remove them from the original by subtraction. Then you'll need a good sound engineer to mix that down to stereo

What is real stereo? Nowadays instruments are recorded individually on a multi-track recorder at different times and maybe even in different places. (Remember Ebony and Ivory where supposedly Stevie Wonder and Paul McCartney never even met). Everything is then cobbled together by a mixing engineer to produce the final mix. The sound field you hear existed only in his head and never in reality. (If you hear Ian Anderson singing front right and his flute playing rear left you'll know what I mean). In that respect I guess that whatever you make out of a mono recording is just as stereo as anything else although the process is admittedly much less flexible.

I have often thought that at least with simple music you ought to be able to get a neural network to recognise the individual instruments. For example I was recently processing a record by Bar Trio, a German group from the 1930s. The instruments involved were an accordion, a guitar and a piano. They all have their own distinctive sounds. I thought that if you could create say a midi file for each instrument then you could get sound fonts for the original instruments and mix them however you wanted. I don't know whether I'll live long enough to get around to trying it.