Old vinyl recordings

[Glenn Davidson]

So now on my Pro-Tools set up I have a RTAS plug in that will remove record scratches and surface noise and one that will add them. I does seem a little ironic that I have digital plug-in's that produce hiss, hum, distortion, scratches and wow and flutter. Thanks Greg for the tip on the free record simulator.

[Zulkifli Yusof]

hey!

That's a nice little find! Thanks alot for the link.
I'll actually be doing audiopost for a short film set in the 1940s. There's gonna be 2 original songs recorded for the film and I need to have 2 different mixes going on for each song; 1 for "live" performance and another for the gramaphone.
Not much problems that I forsee for the "live" mix having already heard the compositions but having not done a vinyl mix and a period film before, this project proves to be quite a challenge sound design wise!

This is gonna save me alot of time tweaking knobs to get the right sound! Now I can just get the desired sound off the right preset and concentrate on the overall sound mix instead!

Thanks again!

(p.s. Do continue the interesting discussion on old records though!)

[Bob Hart]

Vinyl as a common distribution medium was not around but had been developed and proven I think from 1935 or thereabouts.

In the 1940s music recordings were being done direct to a disk master. This was then used to create the mould for the mass produced disks. Earlier ones were created by laying down a a thin coat of shellac on the mould, (there is a correct name for this). The two copies of the recording were bonded to the body of the disk. I dont know if this was a one-step process or in stages.

Later evolutions saw the recording and the bakelite body of the disk made in one stage.

What we hear now from older recordings does not necessarily reflect how they did sound then.

What you hear now is generally reproduced via what is called RIAA equalisation. Before this uniform equalisation curve was introduced, major record companies had their own versions.

So playback on a modern turntable to RIAA equalisation tends to introduce a more bassy playback of 78rpm records than was the original character of them.

Played back with matching equalisation applied, except for the surface noise and being mono, the fidelity and dynamic range of 78rpm records can be surprsingly good.

The main reason we hear scratches and distortion now is that nearly all specimens we hear, have been subject to wear and tear. A "new" 78rpm record will have a higher noise floor depending on how good the original surface finish of the record was but can be relatively noise free.

Apparently and quite surprisingly against expectation, the best quality of some bakelite "HMV" record mass production was achieved in India.

In the context of your project, you might like to consider the origin of the recording. Is it a "popular" form of music contemporary to the location and time?

"Parlophone" recordings common to Europe, have a sound of their own, seem to have a better overall sound quality and can hold up well to more recent vinyl. There seems to have been more large orchestral work done on them.

"Decca" recordings, especially post WW2 seem to have a very strong dynamic range and modern playback systems can have a problem with playback without distortion.

Others, such as "HMV", "Brunswick", "Capitol", later "Vocalion" recordings seem to sit more or less in the middle. Some older "Vocalion" recordings had a lower signal level relative to the noise floor which makes recoveries difficult.

Postwar, tape technology from Germany and early forms of multitracking were introduced. Recordings created using tape were more finessed and the performances perfect. Vocals seem to have been closer miked and more warm.

Sound mixing was more finessed by then whereas in earlier times, performers moved towards or away from a single mike to achieve balance and level control between performers and dominent instruments.

In some older recordings, little performance errors can sometimes be heard, such as something striking the body of a string bass. The vocal performances can also have a greater sponteneity and life to them.

Modern stereo turntable reproducers or pickups with two separate stereo channels can be more sensitive to the surface noise of old mono recordings as the nature of the stereo pickup is to detect and reproduce vertical as well as horizontal movements of the sylus following the groove.

My own personal preference would be to record the performances to modern standards, then for the "recorded" version, on a software equaliser, roll off the frequencies below 120Hz, perhaps in dynamic effects, apply a little compression to the low to mid-range, leave the mid-range to high end alone because is where a lot of the "new" recording noise floor for an old technology recording is going to be.

I would then get hold of an old 78rpm record or maybe a few. I would record the silence from the outer edge and centre of the disk when the lead-in and lead-out sections are. These will give you the surface noise and fairly rapid cyclic repetition of surface defects which are the signature of a 78rpm recording which you can then loop and apply to an effects track. More importantly, it will give you the actual lead-in and lead-out of a 78rpm disk playback.

The older 1930s or thereabouts shellac surfaced records often have a cyclic low frequency defect which comes from a coarse open weave fibre binder in the bakelite subtrate beneath. It manifests as a quiet "woop-woop----woop-woop----woop-woop" or "woop---woop---woop---woop" during lead-in or lead-out or silences in the recording, usually cannot be heard when the recording begins and often does not persist through the entire recording.

A characteristic of 78rpm records is that the apparent noise level and dynamic range of noise and also the recording itself seems to diminish as the tonearm moves toward the centre of the disk. You actually see this in visual form on a Cool Edit Pro waveform display.

Another characteristic is that noise introduced by surface defects in the structure of the bakelite or shellac is produced by defects of the same physical size.

The surface speed is higher at the outside of the disk, probably the reason why the surface noise seems to be of a higher frequency and intensity than when the tonearm is at the end of the recording towards the centre.

Surface damage to the disk from handling and use is also more apparent in the earlier portion of the recording than when the tonearm reaces the centre.

Lastly but most importantly, one of the most glaring plausablity/authenticity errors which occurs in movies depicting the times is a couple making love or smalltalking with a single 78rpm disk visible and playing as an audio backdrop.

This is fine except for one small detail. The playack has been on a portable gramophone or portable record player which was only single disk capable.

A 78rpm disk runs out after three minutes, so the parties either have to change the record or leave it running silent and clicking in the end groove while they do their thing, oblivious to the fact it has stopped, or have somebody else change or restart it in the background.

Autochangers might have been around but I think they were still a fairly new innovation. When I was a kid, use of autochangers for multiple disks was not encouraged to due possibility of damage when one disk dropped onto the other.

For processing recoveries from 78rpm recordings using a modern turntable and RIAA standard reproducer, I have made and saved an equalisation profile in Cool Edit Pro which I named as "deRIAA".

For several popular labels, I apply an equalisation profile unique to each which I generated from charts published on the web. This puts you in the ballpark of what the recordings may have sounded like but you still have to do some additional correction.

Sound engineers of the times would also apply their own personal touch to an individual session so there can be variation with individal record brands.

The tinny playback sound is a long established cliche derived from the depiction of playback of older acoustically recorded disks or rolls through acoustic reproducers. It is so common that now, the humble public expects to hear it that way.

Acoustic recordings were typically very low level, poor dynamic range and high noisefloor. The electric microphone changed all that in 1929 or thereabouts.

In reality, the older record players and mantle radios of the times you intend to depict used valve (tube) amplifiers and larger diameter speakers. They actually had a warmer sound than the modern small radios and players of the sixties through to eighties, which we now associate with antiquity

Another audio characteristic of the times was that the audible frequency of the AC power supply, 60Hz in the US and 50Hz in other counties was often to be heard as a humming sound in the playback of some appliances.

It was especially evident during the warmup of valve appliances, particularly radios from the moment they were turned on to the point where the audio signal began to fade in and the hum became subordinate. Mains filtering was not then as cheaply achieveable as it is now.

I doubt you want or need to go this this level of obsession with authenticity but the point I make is that on the higher quality playback systems of the timesor even the low-end consumer radios, which played the horseracing calls in nearly every hardressing shop in the land, the older records did not sound as tinny or as bad as they are often now depicted.

[Greg Boston]

Originally Posted by Bob Hart

In reality, the older record players and mantle radios of the times you intend to depict used valve (tube) amplifiers and larger diameter speakers. They actually had a warmer sound than the modern small radios and players of the sixties through to eighties, which we now associate with antiquity

I can vouch for this as my parents still have a working console radio and as you indicated, about a 12 inch speaker. Has a lot of low end to it. I love turning it on and listening to it when I visit.

-gb-

[Zulkifli Yusof]

Bob!

Thanks for the in depth reply! Definitely alot of new things I've learnt about old records. You're right that I may not go that deep for the authenticity of the sound but what you've given me are the principles which I can use to achieve the desired sound.

I already intend to do the "recording to modern standards" as a first step. Glad that someone else chipped it on it. I dont think I need a highly accurate reproduction of the sound. I just need something that the general audience can easily identify with.

You mentioned about recording the lead in and lead out of a record. That is something I need to consider as well. How does the sound actually starts from the moment the needle is dropped onto the record?

[Bob Hart]

The sequenceof events for a 78playback is the same as for a vinyl LP but happens a lot faster.

There is an initial soft click or thud as the playback operator manually lowers the tonearm onto the disk. If the groove is missed, there is a second click and then commencement of surface noise as the stylus drops into and follows the lead-in groove. Depending on how far out on the lead-in groove the operator lowers the tonearm, a split to a full second might pass, then the stylus will run onto the recorded area of the disk.

From this point where the cutter operator is satisfied it is running and cues the bandleader or performer, there is a delay for the time it takes for the person to react to the cue and begin the performance, usually about two or three seconds.

When the performance ends, there is usually a delay of about three or four seconds left on the end of the recording to allow reverbrances to decay then the lead-out groove commences. When the stylus runs onto the elliptoid loop at the end, there is sometimes a repetitive click or double-click as it comes back onto the start of the loop and crosses the junction of the lead-out groove.

On some specimens, a faint 60Hz or 50Hz, + or - harmonics can sometimes be heard during the silences. The origins are the usual likely suspects which exist to this day, drive motor noise acoustically conducted to the cutter via the disk due to hardened rubber mounts, stray inductance due to inadequate shielding or bad earthing and the acoustic environment of the performers.

From the frequency of the hum, the recording can be guessed as US or British Commonwealth in origin.

On rare occasions, a recording was made where there was a "wow" or rapidly cyclic frequency shift embedded in the recording.

The origin could be attributed to worn equipment such as a gear on a shaft secured by a grub screw and key. If the grub screw loosens and the gear moves with time on the shaft and key and wears the joint, it will begin to rock on the key and introduce a slight speed fluctuation per revolution of the shaft.

If you can send me an email to provide a return address, I can hunt through some recoveries to see if I have a raw version from which I have not trimmed the lead-in and lead-out and attach an mp3 file to a reply.

[Glenn Davidson]

Hey Bob...You obviously know A LOT about vintage records. Any specific turntable, cartridge or stylus that you like. I am using a Dual turntable with Shure cartridge and stylus run through a Sony vintage pre-amp and then to Protools with Waves restoration software. I am thinking of up grading some of this. What are you using for 78 RPM playback?

[Bob Hart]

Turntable.

Currently I am using a modern direct drive Stanton with the shorter stright disco tonearm. The short straight arm is probably not right for this job but it was what I could get at the time.

Older vintage record players are quite adequate if the internal noise is not too obtrusive but I find they nearly always produce a lot of low frequency rumble, 50Hz or 60Hz due to deteriorated rubber or bad design.

There is often a half-cyclic thumping as the turntable rocks on its centre bearing which is often a hybrid of plain bearing and ball bearing. This is a product of normal wear and tear.

This can be quietened a little by use of silicone high viscosity lubricant but this can be a bit draggy and introduce other problems with roller wheel type drives which were never the most inspiring piece of engineering destined for the consumer marketplace.

Oftentimes, with old players, the mechanism gets locked up. 90%, it is due to hardening of the lubricant on the spindle of a slave gear which has beneath it a snailcam and roller lever arrangment for actuating the automatic tonearm lift and autochanger.

For audio recoveries this is best removed altogether but as you may lose the autochanger's mechanical start and stop facility and switching functions in the process you might have to renovate it and leave it in the system.

A frozen changer gear may continue to operate baulkily as the fixed axle shaft then turns its rivet mount in the platform and will eventually ruin either the platform or the axle shaft. It is a farly easy fix but should not be ignored as a lot of unwanted speed variation can come from it.

The modern Stanton has a single button selection of 33.3, 45, and with two buttons pressed together, 78. There is a enough variable speed control referred to as pitch control available to go up or down for the few non-standard rpm variations on the 78 rpm theme.

On the tonearm is a Ortofon cartridge with 78 stylus. All tonearm motions are manual with a soft lift and drop lever. the Stanton normally comes with a Shure cartridge. I used the Ortofon because I could not find a 78 stylus for the Shure at the time.

The Stanton machine actually exists also as Pioneer. This has a longer double bent tonearm which presents the stylus to the groove better across the entire arc of movement across the disk. I think both may have come from one OEM manufacturer as essentially they are identical.

The Stanton provides unprocessed output, RIAA equalised output, both via a conventional analogue RCA unbalanced pair of leads. There is also a digital output available for those who can go straight to digital.

For superior 78rpm record recoveries there is available from a niche supplier in the UK, a special stylus which is designed to cope with worn record grooves.

The best mono recoveries for surface noise suppression, for me actually came from an old HMV turntable with an early generation transistor pre-amp and amp system. It used the long established BSR crystal cartridge pickup.

It has a specific "mono" selection which seems to apply a cancellation effect on surface noise but enables the wave form to emerge intact. But the waveform is quite bizarre and other turntable noises are too hard to eliminate.

Stereo cartridges are not the ideal for recovering old mono as they combine detected movements of the stylus in both the horizontal and vertical planes. But for dabblers like me they are adequate.

Crystal or ceramic cartridges have good initial gain but are regarded as being inferior due to distortions of the waveform compared to dynamic pickups which are based in induced magnetic fields. The Ortofon is dynamic.

I understand the Shure cartridge to be comparable with the Ortofon so probably no need to change anything there. It is just a matter of finding a good 78rpm stylus or one of the more recent specialised ones designed to cope with worn and wider record grooves.

If you have a good old quiet turntable with a good pickup and low system noise, I don't think there is any real need to change it unless you want to go that last 15% to perfection in which case you are also talking about Cedar systems and lots of money.

You can sometimes diminish wow and flutter of older players by placing a disk cut out of lead sheet on top of the platter then a thin sheet of felt on top of that for the record to sit on.

To preserve the stylus, after thoroughly cleaning the record, I had been using a witches brew of wax based furniture polish and Mr. Sheen silicone based furniture polish.

However Mr. Sheen was taken over and the formulation changed from a silicone based to a wax based product so for record recoveries it is now pretty much useless.

There is a method for getting about a -8db improvement on the wear and tear noise floor involving, wax polish, Mr. Sheen and olive oil but it is a long and involved five pass process which would take too long to discuss here.

I have been using Cool Edit Pro with good results. Sometimes for the most faithful rendition of the original recording, you have to leave some of the surface noise in there.

[Glenn Davidson]

Thanks for the info. Some of the disks that I transfer are 'recordable' paper or aluminum with a coating. I would like to be able to plug in different styli into my Shure cartridge and determine what tracks and sounds best. Some of these home recordings sound great, but some are really low level with lots of surface noise. It probably relates to the original recording, but it would be good to try an assortment of styli, especially the type for the worn groove. I will search the internet for a supplier. I have never tried lubricating the disk, but it sounds like a good idea. I have a great thrift store in my town and am alway picking up 'vintage' electronics. It is amazing how many 'extinct' record and tape formats their are. Anyway, thanks again

[Bob Hart]

Some of those coated aluminium or paper substrate disks need to be treated with care relating to putting stuff on them.

The material was sometimes a composition which comprised in part castor oil which over time dries off. By now, especially if they have been stored in a ot or dry evironment, they have begun to shrivel and crack and are more tha likely to be unplayable.

There may be conservation and prreservation methods for them but I do know this.

The signal on these can sometimes be found to be recorded at 33.3 or 78 rpm. The amplitude and dynamic range of the signal will be found to be much lower, closer to 33.3rpm recordings and buried deeper in any noise floor.

Fortunately, the surface the signal is cut in has a finer structure than most bakelite or shellac coated 78rpm disks. If the material has not shrunk or cracked, this noise floor may be lower.

For some transcription disks, and apparently for the huge vinyl transcription disks used by the US Armed Forces Radio Network??? the stylus to play these disks was a unique shape and was made of bamboo. I do not know anything about the stylus profile or if the disks are damaged by attempts to recover from them with modern styli.

Something I have forgotten to mention is that there were other recording media prior to WW2.

A forerunner to the tape recorder was the wire recorder, which was a fairly portable field recording appliance.

Another was optical recording direct to motion picture film. Because this system used sproketted film, synchronisation across copies for mixdown was possible.

I think some film recording may have been a source for some musical releases on 78 rpm disks, such items as "from the motion picture ----". Some of the MGM disks have a distinct tonal quality and the waveform appears to be compressed.

[Garrison Hayes]

What i'd do is play around with the tracks speed (You know how on old record players theres the little switch that makes the song speed fast or slow?...Well that.) Try to make it sound a little warped...but not to much!

And i'd take most if not all the bass out of the track. That will give it an old radio sound.

Hope this helps...
G. Hayes

[Zulkifli Yusof]

Hey all

I've recently completed the recording of the 2 songs needed for the project and I must say, working with professional demo singers and vocal trainers really help smooth your work by ten folds!

I've also begin working on mixing one of the songs. At the same time had a lot of fun tweaking izotope's vinyl plugin and I have learnt the following:

- The stereo/mono switch is a very useful tool and it's not limited to its use as just a vinyl plugin.
- Using the plugin on an electric guitar solo can definitely give a much needed "crunch" to its sound and it cuts through the mix better. Try it!
- Most of the presets I've used tend to make vocals sound like the telephone effect. They also make the whole mix sound more like an old radio than a gramophone.
- Using izotope vinyl plugin on its own is not enough to achieve the gramophone effect I needed. Doing further EQ'ing after applying izotope's vinyl, adding some compression and old tube effect makes it sound much closer.