Master tape deterioration science

[levimax]

There is a lot of discussion on the internet about “original master tapes” of music from the past and their condition and availability as these tapes are used for “re-issues”, “re-masters”, and “re-mixes”. Some people claim original LP’s and even CD’s made from “fresh” master tapes decades ago can not be equaled today due to damaged or lost master tapes. To me it seems that while these claims are in general overstated in some cases they do seem to have some merit (for example I have an original Jimi Hendrix LP which sounds better to me when level matched ABX against more recent “re-issues” CD’s or LP’s of the same).
My question to the experts is what is the “science” behind the deterioration of master tapes over time? How much fidelity is lost when a “safety copy” is made as compared to the original? What about “baking” of tapes I read about, if tapes have to be baked do they lose fidelity and how much? Do master tapes lose fidelity every time they are played and if so, how much? Do master tapes lose fidelity over time just sitting in a box? I have also read “early” tapes from the 50’s and 60’s hold up better than newer formulations in the late 60’s and 70’s… any truth to that? Finally, I have read that aligning the heads to a master tape made on a different machine than it was originally made on creates issues… can this be quantified?
Thanks for any insight.

 

[gene_stl]

I don't claim any high degree of expertise but there are lots of different things that happen to master tapes. A lot of very good questions.

First of all they were recorded "back in the day" which imposes some limitation on them by virtue of the loose specifications in some recording studios. ie. when digitized there is only 10 or 12 bits of dynamic range. and maybe more distortion (very evident in my favorit Jimi Hendrix tune in spite of being recorded at Electric Ladyland)

Even absent the above if a reel to reel master tape is thirty of forty (or even more!) years old it can suffer from several things that decrease its fidelity.

The first is print through. This is where the recording on one layer of tape probably at a higher level imposes itself and becomes audible on an adjacent layer of tape. It can sometimes be heard as a pre or post echo, sometimes even on LPs. (Just as an aside it seems to me that this would be something that would be a candidate for some software writing hotshot to figure out a correction process for if it hasn't been done already) LPs can have a similar phenom of adjacent groove cross talk but I don't think that happens unless its a really old cutting.

Then there is the phenomenon of self erasure. I think this is a statistical process of magnetic domains (correct me if I am misteaken) and iirc affects the high frequencies more.

The above are problems of the magnetic layer.

There are also problems of the tape backing layer.

One of these is adhesion failure which causes the magnetic oxide to fall off the tape. It the piece that falls off is big enough there is "data loss" or possibly S/N degradation. If a tape is shedding oxide badly it can mess up playback machines which certainly should be cleaned and demagnetized before playing back a valuable master tape.

The mechanical backing layer of tape is a thin plastic. often acetate or mylar. These are selected to be stiff so as not to stretch when on a tape transport. However when they get to be 50 years old atmospheric oxygen may have contributed to additional polymerization of the backing layer which may make it stiffer and brittle and more prone to snap if the transport buttons are not pressed very judiciously and the transport brakes not adjusted correctly.

Another backing layer problem is sticking together. I believe this is the problem for which they bake the tapes. I am told by a recording engineer friend that TDK tape does not have this problem but that Ampex Grand Master does. I think it also depends on how long it has been since a tape has been rolled and how tightly packed it was at the time. There may also be effects from whether it was stored "tails out" ie played and not rewound at high speed. The pack if you leave it tails out is prettier and probably excludes atmospheric oxygen better than if you rewind it at high speed.

Baking a tape probably increases some self erasure though it may not be enough to be audible.

If the tape deck is not demagged then playing the tape may cause partial erasure which might be hard to distinguish from self erasure.

Just playing a tape probably does not reduce it's quality in a measureable amount. Some of the magnetic layer may wear off and some might be erased if there is a magnetic spot on the transport. But I am sure the engineers that design tape decks would pick non magnetic alloys for hardware that was going to be near the tape.

Of course the worst thing that happens to master tapes is when the vault they are in catches fire and they have not all been digitized already. Even 16 bit RBCD audio is better than any master tape and is easy to duplicate and store on multiple sites and so that the temporal limitations of magnetic tape don't cause loss of quality.

It probably is better to play a tape every few years than to let it sit in its' box and "take a set" or get stuck together. Similar to how its good to exercise the focusing mounts of cameras and microscopes so they don't freeze solid or exercise other mechanical things.

 

[MakeMineVinyl]

This is a complex subject with lots of variables, not the least of which is how the master tapes were stored.

In general, there was a period from the late 1970s up through the late 1980s (and possibly longer) when Ampex mastering tapes especially degraded over time because of what was dubbed 'sticky shed syndrome'. This was due to the urethane used in the tapes binder formula, and caused all such tapes made in this period to become unplayable unless baked for about 8 hours at 120 degrees F. Once baked, the tape became playable for at least a few plays while a copy could be made. I have a shelf of masters I recorded during this period and all of them suffer from this.

Tapes made before this do not suffer from this, but they have their own problems. One such problem is with acetate based tapes which sometimes suffer from 'vinegar' syndrome where the backing shrinks (and emits a vinegar smell) and thus becomes unplayable. Not all acetate tapes from this period suffered from this, and I have a good number of such tapes which play fine. It seems that the better sealed the container in which these tapes were stored, the more likely they were to suffer from vinegar syndrome. Another problem with acetate based tapes is that the backing 'curls' and because of this can lose intimate contact with the playback head.

Tapes made with polyester backing (Mylar) do not suffer from this and generally are perfectly playable today.

Unless a master tape is played to death (multiple thousands of times), there is really no reason that it shouldn't have the same sound quality as when it was recorded - provided it was well stored. Tapes stored near magnetic fields (motors, speaker magnets etc) can have their signal partially destroyed; sometimes this can be heard as the high frequencies fading once per revolution of the reel.

So there is a vast number of variables which can determine how a master tape sounds today, ranging from tape formulation problems to storage problems.

By the way, to answer your question about 'safety copies', this depends if the tape is a copy of the master, or if it was made simultaneously as the master on another machine. Generally you can get away with copying a tape once but making subsequent copies of that copy run into audible problems, the most obvious is noise buildup.

Most of the leading recording studios practiced very good maintenance of their tape machines, including daily demagnetization of heads, guides etc. and cleaning of heads and the tape path. At the head of each master tape, tones of 1kHz, 10kHz and 100Hz were recorded at '0' level so that subsequent machines could be aligned to play the tape back properly.

 

[Pluto]

if tapes have to be baked do they lose fidelity and how much?

Now there's a question. The need to bake a tape is dreaded and if one (i.e. I) was confronted now with a tape in need of such treatment I would probably advise the client to spend whatever extra it might cost to have the tape baked by someone with a known good track record in this awful field. Once baked, you have to work on the assumption that you will get just one good pass out of the tape before it fails catastrophically. This is when you hope that, above all, the folks who originally prepared the tape were not too mean on the 10kHz tone so you have enough time to get the playback azimuth right – which takes 30"-60" so when you hear the oscillator bump to 100Hz after 15" you get that awful sinking feeling.

When the azimuth is tweaked to your satisfaction, the mission is to get a decent flat digital transfer at a high-ish sampling rate with the best ADC you can lay your hands on. More or less everything else can be adjusted after the fact, but azimuth needs to be right. True, there are tools that perform the equivalent of an azimuth adjustment in the digital domain but it's best to treat such tools on a "use if you must" basis. Once you have your flat digital transfer, you can do anything else necessary – Dolby, EQ to taste, re-edit etc. – in the digital domain or you can convert back to analogue to apply the vintage tools so much in vogue at the moment.

So to answer your question: once a tape is baked, you are grateful for whatever you get out of it. At that stage the prime objective is to capture whatever is on the tape as accurately as you can, and deal with the art once you are safely clear of the pile of powdered ferric oxide now residing within the guts of your precious Studer deck!

 

[MakeMineVinyl]

I've had much better luck with baking tapes, Ampex 456 - the worst offender - specifically. I used my kitchen oven and baked the tape (on a metal reel obviously) for 8 hours at 120 degrees F. These tapes are still playable after almost a year since baking. Now its true that the tapes are being stored in a location which is has very low humidity, so that may have something to do with it.

When I did the music track restorations for the 1998 re-release of "Grease", the head of the sound department took the 1" 8 track tapes home and baked them in the kitchen oven. They played fine on the Studer machine used for transfer into my ProTools DAW, and these digital transfers were used on the dubbing stage for the re-mixing.

Usually with tapes which exhibit stick-shed, after a few minutes of playback, the guides and heads will have so much 'goo' from the tape that the pinch roller cannot pull the tape at normal speed. Trying the rewind the tape will just result in stalled reels. After baking, the tapes play normally.

I have a good number of masters recorded on Scotch 250 tape from the same era; these do not exhibit sticky-shed and play normally after decades of shelf storage.

 

[MakeMineVinyl]

By the way, I have been buying 'vintage' Ampex tapes on Ebay recently and none of the tapes have had sticky-shed despite them being decades old. Note however that these tapes have all been non-backcoated older formulations - Ampex types 631 and 641. These are polyester (Mylar) based tapes. I've also purchased a few reels of 611 which is an acetate based tape, and these have shown tape curl from shrinkage. This is not a serious problem however since these tapes will only be used on a full track mono machine (my avitar is a full track Ampex playback head). If used on a machine with narrower tracks, 4 track stereo or half track stereo, track dropout could be a problem.

 

[Pluto]

I've had much better luck with baking tapes

Please note that what I said was…

Once baked, you have to work on the assumption that you will get just one good pass out of the tape before it fails catastrophically.

I have encountered several that do seem to be OK for multiple passes post-bake, but initially you do need to assume that you will get one pass only, hence the need to make that one pass count. If you get more – happy days!

 

[MakeMineVinyl]

Please note that what I said was…

I have encountered several that do seem to be OK for multiple passes post-bake, but initially you do need to assume that you will get one pass only, hence the need to make that one pass count. If you get more – happy days!

I think it may have more to do with any differences in relative humidity in our respective locations; mine is very dry.

 

[Pluto]

Not so much the geographic location as where/how the tape has been stored for the last nn years!

 

[MakeMineVinyl]

Not so much the geographic location as where/how the tape has been stored for the last nn years!

Uh, yes...where there is humidity more of the year than one which is not. I'd venture to say that the UK is more damp a greater portion of the year than New Mexico. The only thing which would make location irrelevant would be a sealed, atmosphere controlled vault. Even the studios I've worked at don't have that. Heck, even most movie studios in Hollywood don't have that. Not all masters are stored at Iron Mountain. www.ironmountain.com/resources/data-sheets-and-brochures/t/the-iron-mountain-underground-rock-solid#:~:text=Iron%20Mountain's%20“The%20Underground”%20is,naturally%20protected%20former%20limestone%20mine.&text=Vaults%20maintain%20specified%20temperature%20and,and%20extended-term%20storage%20conditions.

 

[levimax]

Thank you everyone for your comments. I have learned a lot and had a lot of things clarified. The reason I asked was to understand if there was any scientific basis for seeking out "better sounding" versions of older recorded music rather than assuming the latest remaster was "the best" and I guess the answer is yes it is possible that recordings sourced from older undamaged sources are better quality but that it is a cases by case basis. Since the music companies are trying to sell the latest version they are not going to give out any reliable information. I wish there was a scientific way to compare recording quality but for music that would be difficult so I am left with conflicting internet opinions and my own AB testing. While the differences I have found are usually subtle at least I can reliably ABX some recordings unlike my attempts at ABX DAC's and amps where I could tell no difference.

 

[MakeMineVinyl]

The differences in sound quality between remasterings have much more to do with decisions which the remastering engineer makes. Remasters are usually made from the same master source, either tape or a digital copy of it, so they all start from the same place. What is different is that the remastering engineer has what amounts to 'tone controls' in the form of EQ and they also have the ability to compress/limit or not. The bass might be boosted a bit, the highs boosted or reduced, or any number of adjustments which are all at the discretion of the remastering engineer and/or original artist. Some remasterings are literally re-mixes - the recent release of Sergeant Peppers is a perfect example of this where they went back to the original work parts and remixed the entire album. Of course this release is going to sound far different than the original LP from back in the day (I am split on whether I like it better or the original)!

There are no scientific / ABX solutions here - it is totally and utterly subjective. Trying to do 'scientific' comparisons of something which was created at the subjective whim of the remastering engineer is really a waste of time.

 

[levimax]

The differences in sound quality between remasterings have much more to do with decisions which the remastering engineer makes. Remasters are usually made from the same master source, either tape or a digital copy of it, so they all start from the same place. What is different is that the remastering engineer has what amounts to 'tone controls' in the form of EQ and they also have the ability to compress/limit or not. The bass might be boosted a bit, the highs boosted or reduced, or any number of adjustments which are all at the discretion of the remastering engineer and/or original artist. Some remasterings are literally re-mixes - the recent release of Sergeant Peppers is a perfect example of this where they went back to the original work parts and remixed the entire album. Of course this release is going to sound far different than the original LP from back in the day (I am split on whether I like it better or the original)!

There are no scientific / ABX solutions here - it is totally and utterly subjective. Trying to do 'scientific' comparisons of something which was created at the subjective whim of the remastering engineer is really a waste of time.

Yes I understand and agree but I am not sure "everything starts from the same place" as the original LP's usually used "brand new fresh master tapes" while the original CD's used tapes that had been stored for decades and in some cases back up tapes or LP cutting tapes etc. In some cases it sounds to me like the original recordings using fresh master tapes have "more information" especially HF information than later versions (In addition to EQ differences). In other cases later releases sound much better. Very hit and miss and in most cases very subtle. I do have to say though that my first press LZ II and Hendrix Are you experienced sound better than any later version I have heard ... of course much of this is EQ choices but those EQ choices of first pressings, made by Robert Ludwig and the like, are part of the "art" of recorded music.

 

[MakeMineVinyl]

Yes I understand and agree but I am not sure "everything starts from the same place" as the original LP's usually used "brand new fresh master tapes" while the original CD's used tapes that had been stored for decades and in some cases back up tapes or LP cutting tapes etc. In some cases it sounds to me like the original recordings using fresh master tapes have "more information" especially HF information than later versions (In addition to EQ differences). In other cases later releases sound much better. Very hit and miss and in most cases very subtle. I do have to say though that my first press LZ II and Hendrix Are you experienced sound better than any later version I have heard ... of course much of this is EQ choices but those EQ choices of first pressings, made by Robert Ludwig and the like, are part of the "art" of recorded music.

There is no such thing as "brand new fresh master tapes". Either the "new" one has to be a copy of the original session tape, or be a remix along the lines of Sergeant Peppers I mentioned earlier. In the 50s and 60s it was common practice to run "A" and "B" master machines at the same time - these were usually 2 track or 3 track. In this case, they could go back to the "B" master which presumably was never played. When mulit-track machines became more common, it became impractical to run two master machines at he same time; keeping track of two 24 track master recorders with things like punch-ins would be a nightmare! When I was a recording engineer at Sound City, I never saw two mix masters being created at the same time, even on popular albums you probably have. Safety copies were always 2nd generation copies of the original (I have a number of these). Both the large "A" studio and the smaller "B" studio had one 24 track master recorder each (a Studer A-80 in the "A" studio and an Ampex MM 1200 in the "B" studio). Both studios had one Studer 2 track for mixdowns, and there was a 'roving' 1/4 track Revox which was used for making copies for clients to play at home. This was the situation in all LA studios I ever saw.

 

[DSJR]

A little anecdote from a mastering engineer pal who used to work at Decca in London's Belsize Road before it was shut some years back - someone who loathed having to eq or 'alter' the masters he worked with unless he absolutely had to as his view was to present on CD what the producer signed off (being a classical musician himself, any remedial work - 'help' with fades due to master faults and so on and so on - was always very lightly done). One of his tasks was 'half' a Tom Jones two-CD compilation, he doing one disc and his boss the other (the stories his boss told me, but that's for another time). I believe most tracks were left unadulterated in the final discs, so what you hear is pretty much exactly what there was BUT, apparently these mid 60's tapes were shedding quite badly and my pal felt they were lucky to get to them when they did (early to mid 90's from memory) as he felt they wouldn't take many more plays or years in storage before dropouts became a serious issue. I have to say the songs do 'sound' very good.

As for LZ 2, the original vinyl sounds awful, with azimuth issues on The Lemon Song, way overloaded Whole Lotta Love when the gain was increased for the quiet part and left high when the band come crunching back in - there's an entire page given over to the deconstruction of this mix but this'll do for now - Oh and that young Plant voice.....

http://www.ajawamnet.com/ajawam2/lzcompare_0006.mp4

Got it! -

http://www.ajawamnet.com/ajawamnet/remixoflz.html

I believe what we're getting now from the 1990 and recent remasters is more of what's actually on the tapes, like it or not, as Jimmy Page himself has overseen the remastering work done and I'm sure he didn't skimp. Thank the lord the multi-tracks are still in existence unlike so very many classic studio session tapes lost in the Universal fire some time back

My pal told me thirty years ago that digitising the contents of tape vaults was a race against time, although an early '50's Decca master (one-sided reel) he showed me was in excellent condition. When the facility was abandoned by polygram as it then was, the tape and earlier material vaults were moved and I don't know where. I believe any digitising and remastering work was out-sourced but no idea if the work on the vailts' contents was continued or just mothballed...

Apologies for any thread drift here..

 

[MakeMineVinyl]

Digitizing has its own problems - some masters were archived to DAT tape, and we know how that turned out.

 

[Pluto]

apparently these mid 60's tapes were shedding quite badly and my pal felt they were lucky to get to them when they did (early to mid 90's from memory) as he felt they wouldn't take many more plays or years in storage before dropouts became a serious issue

This precisely echoes my experience of dealing with such tapes. 60s tapes, for the most part, had their oxide simply fall off while 70s tapes had it pushed off by the tape transport following the necessary baking process

 

[levimax]

As for LZ 2, the original vinyl sounds awful, with azimuth issues on The Lemon Song, way overloaded Whole Lotta Love when the gain was increased for the quiet part and left high when the band come crunching back in - there's an entire page given over to the deconstruction of this mix but this'll do for now - Oh and that young Plant voice.....

http://www.ajawamnet.com/ajawam2/lzcompare_0006.mp4

The original mix is the original made by the band and the mastering engineer working together for better or worse... the remix is nice but still not a sonic master piece and still has lots of distortion and is not what was originally released and sold tens of millions of copies. Since I grew up with the original that is what I prefer but I can see why people would like the remix.

I believe what we're getting now from the 1990 and recent remasters is more of what's actually on the tapes, like it or not, as Jimmy Page himself has overseen the remastering work done and I'm sure he didn't skimp

The Jimmy Page remasters are controversial at best. Part of it is the "original" version thing from older fans but many people think that the EQ choices made we caused by Jimmy's age related hearing loss. Personally I prefer either original vinyl or early Diament mastered CD's.

 

[Cbdb2]

More info.

https://www.richardhess.com/tape/history/HESS_Tape_Degradation_ARSC_Journal_39-2.pdf

 

[MakeMineVinyl]

This precisely echoes my experience of dealing with such tapes. 60s tapes, for the most part, had their oxide simply fall off while 70s tapes had it pushed off by the tape transport following the necessary baking process

Were these 60s tapes EMI or BASF/AGFA by any chance? I ask because I've never experienced oxide flaking off tapes made in the US from the 50s or 60s. The most I've seen is either acetate base curling or in a couple cases tape squeal caused by the lubrication in the tape becoming ineffective.