ELVIS AT SUN


An Overview of 
the Audio Restoration 



With all the well known and some less well known problems of lost master tapes and poor digital mastering it is a relief to find that the result of the 'Elvis at SUN' project is much better and more uniform than really ought to be possible to achieve. This is due to mostly lucky circumstances regarding what actually has been found in recent years, combined with some recent, unusually well calibrated tape transfers by Sony and perhaps most important of all, the huge manual restoration effort carried out by Kevan Budd. For those who care, even the relative amplitude in terms of RMS power distribution has been carefully measured and set by hand for consistency and for a uniform impression of vocal amplitude regardless of source. This is very much the opposite direction of normal mastering of vintage music of today, which seems to be focused on pushing up the level as much as possible over the 0 dB limit.

Harbor Lights, I Love You Because, 
That's All Right
and Blue Moon of Kentucky


Being an early take of an abandoned recording, possibly from the first session on July 5, 1954, the opening track Harbor Lights is the weakest performance of Elvis' SUN recordings. Unexpectedly it also stands out as being one of the few recordings where the actual SUN master tape has been recovered. The most audible difference is that the out-of-place 1976 'Legendary Performer' echo is gone. The new straight transfer actually reveals Sam Phillips' own, more subtle use of locally miked tape echo. In this case, a delayed signal from Scotty's guitar is mixed with the live signal coming from all microphones. This leaves the vocal almost dry, but with a touch of leakage into Scotty's microphone. The result is a very full sound overall and a very natural sound of the vocals.

The above is best illustrated by comparison to the second track, thought to be from the
same session but recorded dry using just one tape machine.
Recording of country ballad I Love You Because was abandoned after 5 "takes". RCA spliced "takes" 3 and 5 for their LP master in 1956, but the first complete attempt, "take" 2 is the only consistent full take without major errors and therefore appropriate to include on a genuine SUN collection like this. Here we get a very recent Sony transfer of the original SUN tape. The main characteristics seem to be in the 1-track head of the tape recorder picking up signal from the entire width of the tape. This also results in a high signal/noise ratio and may almost be interpreted as truncation of noise, since random noise is likely to vary along the width of the tape. Apart from that, neutral amplifiers and careful NAB tape curve calibration seems to characterize the Sony transfers. The technique is thus optimised for flat transfers of original album masters or anything else that was mastered for release in the first place.
I Love You Because is a dry, flat recording and as much as the Sony transfer of theoretically is as good as it gets, some might find it sound a little hard compared to the full sound of Harbor Lights that actually was transferred by BMG. 

Sadly, the SUN master tape of That's All Right is lost forever. It was used to produce the original single in July 1954 and then once more in November 1955 when RCA transferred it. Since I Love You Because and That's All right were recorded on a rehearsal or session that had failed so far, there either was no time for Sam to setup or rewind the slapback tape before recording to produce the characteristic SUN echo. (Only the latter would support Harbor Lights actually being the first recording)

The original R&B side of SUN 209 was dry, just as the outtake that survived on the same session reel as I Love You Because. To prepare it for re-release RCA added compression and an entirely different type of echo when SUN Box #2 was played back for the last time in November 1955. Nevertheless, the RCA tape copy that still exists, is by far the best remaining source in terms of signal. The only other option would have been to transfer an inferior source, a 50-year-old SUN 45 or 78 RPM. Something that couldn't be justified on a major release aimed at the general market now that a tape source does exist. Thankfully, That's All Right is the only alien on 'Elvis at SUN' and as such it is considered a slight improvement over what has been released before.

The C&W side of SUN 209, Blue Moon of Kentucky is one of the recordings where no master tape was turned over to RCA in the first place. A November 1955 RCA tape transfer of a SUN 78 RPM has been used up until now. Now, Blue Moon of Kentucky is one of the worst recorded/mastered SUN masters with compression and overdone echo, but it isn't nearly as bad as the RCA tape implied. RCA compressed it further (to death actually) in November 1955. A manually cleaned up, new transfer from a selection of original SUN 78 RPM's has rectified the situation on 'Elvis at SUN' and the improvement should be obvious. 

Blue Moon, Tomorrow Night and 
I'll Never Let You Go


Next up is a group of three recordings that perhaps were considered rejected masters or potential album tracks by Sam Phillips. They ought to have been recorded at the same session some day between August 19 and early-to-mid September 1954 because they share the same ambience and warm, original tube sound, all carefully miked for the reverb chain in a way that can not been identified on other SUN recordings. 

As could be heard on the outtake as released on 'Sunrise' back in 1999, the original SUN Blue Moon tape appears to be very worn. It was recorded at lower level as well and this does result in a lot of tape hiss that has been left intact on 'Elvis at SUN' in order to preserve the signal and ambience. On the other hand, the dropouts have been repaired. The result is a big improvement over the 'Sunrise' outtake and of course an even bigger improvement over the previous BMG master.

The 50's Box digital masters of the nice 1956 RCA tape copies of Tomorrow Night and I'll Never Let You Go have been replaced by flat Sony transfers on 'Elvis at SUN'- well suited here since these masters were finalized by Sam Phillips in the first place. This new transfer lacked a little high end and this was only compensated for by as much as the source could handle.

Just Because, Good Rockin' Tonight and 
I Don't Care if the Sun Don't Shine


A completely different sound was recorded on the September 1954 session that produced the second SUN single Good Rockin' Tonight (R&B side), I Don't Care if the Sun Don't Shine (C&W side) and the rejected master Just Because, to be used for the first RCA LP in 1956. The early generation master tape of I Don't Care if the Sun Don't Shine has finally been identified as a recovered master this time of course and the glitch is repaired as good as could be done. A pitch analysis implies that the recovered tape runs a little too slow on 'Sunrise', according to Scotty Moore's highly accurate guitar tuning during the SUN days. This observation is in accordance with the pitch of original SUN single, but yet Elvis' vocals may sound slightly too fast. On 'Elvis at SUN' the pitch has been set to that of the original SUN single and the tuning of Scotty's guitar since most threads point in that direction. 

On 'Elvis at SUN' Good Rockin' Tonight is from a new Sony transfer from the recently found RCA 30 ips tape. Just Because is a Sony transfer as well, but from the only, badly filtered, 1956 RCA copy that exists, apart from the album master of LPM 1254. Regardless of the now existing, best sources of each of these songs, traces of the same original tape curve problems are evident on all the three uptempo September 1954 recordings. They were recorded with way too much signal in the high end, especially in the 16 kHz region and with too little signal in the lower, so important 100-200 Hz region. Possibly a noise reduction curve on top of the NAB curve, or simply poor calibration, but it should be pointed out that this 16 kHz peak can be identified on some other SUN artists recordings between late 1954 and early 1955. 

RCA must have detected this but missed the target a little and filtered more in the 14-15 KHz region instead on Good Rockin' Tonight, which is close actually and not a problem today. On Just Because, RCA filtered out just about everything in the 9-10 kHz region, where there were no problems at all, leaving a black hole sacrificing much of the damped acoustic guitar that was recorded via the reverb tape machine. On 'Elvis at SUN' all three songs have been calibrated with 16 kHz reduction and just a slight lift of the 100-200 Hz to compensate a little for what is considered to have been lost in the first place due to the wrong tape curve calibration. The black hole on Just Because has been carefully raised, but not all the way of course, while the 14-15 kHz reduction on Good Rockin' Tonight was found best left untouched after several blind A-B tests.

Milkcow Blues Boogie, You're a Heartbreaker, 
I'm Left You're Right She's Gone
(slow) and 
I'm Left You're Right She's Gone
(master) 


The recording sessions that produced the third SUN single Milkcow Blues Boogie (R&B side) and You're a Heartbreaker (C&W side) are likely to also have resulted in the final master of I'm Left You're Right She's Gone (C&W side of fourth single) as well as the slow, rejected bluesy version of the same song. 

The master tapes for Milkcow Blues Boogie and You're a Heartbreaker were never turned over to RCA, possibly because they had been recorded over by mistake. RCA made dubs of a Mint 78 RPM copy that were used to produce replacement masters for a late 1955 single re-release and for future use by RCA. Both were processed during this process and many artifacts from the shellac source have remained. All restoration efforts for later CD releases have been extremely poor and actually mainly had the opposite effect with introduction of extra distortion and elimination of ambiance. It is therefore a relief that it has been possible to replace both masters of the third single with fresh transfers from a huge selection of other 78 RPM's. If anybody thought that restoring old sources like this is equivalent to sitting in a nice air-conditioned office being served grapes and dinners while pressing buttons of a lot of luxury software to automate the process, think again. The result from a tedious, manual job on Milkcow Blues Boogie is stunning and simply has to be heard to be believed. 

The pitch is not that easy to solve, with guitar tuning in contradiction with original 78 RPM pitch and the 60 cycle hum distribution over time (such low frequencies needs ridiculously high FFT values to even measure with decent accuracy). It is presented at original speed, which if correct, could be the only SUN recording where Scotty's guitar wasn't carefully tuned. Going by guitar tuning only, the pitch would have had to be set slightly slower.

You're a Heartbreaker
is a big improvement as well, but not quite as good as Milkcow Blues Boogie. The reason for this is that few copies were pressed, as it was the least successful SUN single and that the You're a Heartbreaker side appears to have been played a lot more than the R&B side by the few people who bought a copy. If an Unplayed/Mint copy of this side could be located, You're a Heartbreaker could have been presented with slightly less surface noise.

The rejected, slow I'm Left You're Right, My Baby's Gone is of course improved here as well with dropout repairs and absence of the digital BMG mastering with filtering and added echo that has haunted us until now. The other outtake that was released on 'Sunrise' merely suffered from dropouts and poor eq resulting in a 'hard' sound. 

I'm Left You're Right, She's Gone, the C&W side of the fourth single that wasn't released until 5 months after being recorded has never sounded anywhere near this good before. The main source was a Sony transfer of the recently found 30 ips RCA master tape that is at least one generation down, but not subjected to any bad processing at all. Since the song on the best tape isn't complete (last song on compilation tape), it was necessary to combine more than one source to produce the best result throughout. The now restored and beautifully matched original uncut ending is a thrill to hear - a personal favorite on 'Elvis at SUN' considering the final, uniform result from combined sources.

Baby Let's Play House

The February 1955 session that produced the R&B side of the fourth SUN single, Baby Let's Play House comes from the same Sony transfer of the recently recovered RCA 30 ips master compilation tape and there certainly were a few problems with this recording in the first place. The overall high-end had been reduced in three steps soon after the start of the song. This is something that can be traced back to the original SUN single, but is not as evident behind Bill Putnam's high end dynamic processing of the original single. It was possible to compensate a little, but perhaps not all the way within the strict, but necessary rules set out for 'Elvis at SUN'.

I Forgot to Remember to Forget, Mystery Train and Trying to Get to You

There exists a fantastic original 15 ips SUN tape copy for the masters of the fifth and last SUN single I Forgot to Remember to Forget (C&W side) and Mystery Train (R&B side). This tape was handed over by Sam in January 1956 and is therefore referred to as the 16th SUN Tape Box. The tape was first used for the 50's Box, but has now been retransferred flat by Sony with even better results in spite of the flat, original, unprocessed sound. This is partly due to better NAB calibration.

The new transfer of I Forgot to Remember has already found its way to two recent releases. First the 2003 upgrade of 'Great Country Songs' and later the same year it was used for 'Second to None'. The introduction of noise-shaping when converting back to 1-bit DSD native after first converting to 8-bit PCM-narrow in the Sonoma environment to allow equalization isn't very impressive on either one of them, especially considering the result was aimed at 16 bit PCM only. However, the slight above 10 kHz raise done by Sony for these two releases isn't considered out of place at all - it's right on target actually. It's just the method that seems a little complicated. The distorted clipping of vocal peaks on 'Second to None' is simply devastating since this source has so much dynamic range. The 'Great Country Songs' version was clipped as well, but not nearly as much. The 'Elvis at SUN' version of I Forgot to Remember is a slight improvement with total absence of clipping being the only major audible difference compared to 'Great Country Songs'. 

The equally great new Sony transfer of the unprocessed, dynamic Mystery Train makes its debut here on 'Elvis at SUN'. The original, full-ending-to-the last-note version derived from lost Sun Tape Box #1 (that was reported to us by Sven Adamski in 1999) has been recovered from the best available source (tape copy with compression and severe generation loss) and matched with precision during the fade-out, using all the tricks in the book and for once stretching all the 'Elvis at SUN' restoration rules. The result of this impossible task is stunning considering the circumstances. 

According to Sam Phillips Trying to Get to You would have been the R&B side of Elvis' sixth SUN single. Unfortunately the surviving RCA copies suffer from more compression than usual and not even this Sony transfer of the recovered RCA 30 ips tape can do anything about that. It is a shame how the compression takes over during the loud vocal parts, but apart from that the new source and version isn't all that bad, at least not compared to what has been released before.

When It Rains, It Really Pours

We've had it for years without being able to tell! Exceptionally poor compression has hidden the fact that When It Rains, It Really Pours actually is from the dry reverb tape source. On this recording only Elvis' vocals and acoustic guitar were miked and sent to the mounted tape machine to produce the typical SUN echo on the other tape. The electric guitar didn't have to take this path in late 1955 and the echo of Scotty's Echosonic, purchased on May 24, 1955 (www.scottymoore.net), can be heard leaking into the dry reverb channel, as can Johnny Bernero's drums and Bill Black's bass. They certainly were miked through and recorded directly on the primary console machine, mixed with the delayed signal of what we have here. It is a revelation to hear it as it is. Since it is the last recording its inclusion does not distract the overall impression at all. The unfinished dry reverb tape recording of When It Rains, It Really Pours can be considered a bonus track more than anything else.

If Sam Phillips were alive he might have called this a chronological collection of 11 Masters, 5 rejected master takes or potential album fillers, 2 abandoned recordings and 1 unfinished recording. To the rest of us this is an important piece of 20th Century history.


TRACKS 

01. Harbor Lights (Abandoned)
02. I Love You Because (Abandoned/take 2) 
03. That's All Right (Master)
04. Blue Moon of Kentucky (Master)

05. Blue Moon (Rejected Master)
06. Tomorrow Night (Rejected Master)
07. I'll Never Let You Go (Rejected Master)

08. Just Because (Rejected Master)
09. Good Rockin' Tonight (Master)
10. I Don't Care if the Sun Don't Shine (Master)

11. Milkcow Blues Boogie (Master)
12. You're a Heartbreaker (Master)
13. I'm Left You're Right She's Gone (Slow/Rejected Master)
14. I'm Left You're Right She's Gone (Master)

15. Baby Let's Play House (Master)

16. I Forgot to Remember to Forget (Master)
17. Mystery Train (Master)
18. Trying to Get to You (Master)

19. When It Rains It Really Pours (Unfinished) 





A-B SAMPLES
2x30 seconds, 256 kbps MP3 (LAME), 1.8 MB each.



Harbor Lights  
A-B SAMPLE (1)
A. Previous BMG Master (50's Box)
B. ELVIS AT SUN
When it was first released in 1976 Harbor Lights suffered from unnecessary generation loss, bad processing and rather poor 1970's style echo, different on each stereo channel. Then in the early 1990's more poor mastering and spatial echo was added to the 1976 LP master and that remaster has been used ever since 1992.



Harbor Lights spectral plot, 
A=Sunrise, B=Elvis at SUN
Loss of signal since 1976 and a horizontal pattern in the filtered high frequency range that appears to be the signature of the evil NoNoise algorithm since 1992. 

Harbor Lights  
A-B SAMPLE (2)
A. Today Tomorrow and Forever (alt. take, 2002)
B. Elvis at SUN
When the original session tape, that had been lost for some years, finally was recovered a few years ago this resulted in another take being released on the 2002 BMG 'Today Tomorrow and Forever' set. The transfer of the old mono tape was done by BMG on what appears to be a playback of a 1-track mono tape on 2-track stereo machine. That technique would result in playback of unique, narrow parts of the tape in the Left and Right channels respectively, resulting in more dropouts and tape noise than would be the result of Sony's more exact 1-track mono transfers. 


I Love You Because  
A-B SAMPLE
A. Sunrise (BMG, 1999)
B. Elvis at SUN
The absence of the usual mastering is actually what makes I Love You Because take 2 stand out as better sounding than ever before, although this was one of few recordings that has been released in decent sound on Sunrise (1999).


That's All Right  
A-B SAMPLE (1)
A. Upgraded BMG master, Second to None (2003)
B. Elvis at SUN

The recently found RCA tape is recorded at 30 ips, which may not be very rock'n'roll in terms of bass response when properly calibrated during playback. However, if that is to be compensated for, the entire balance between frequencies must be carefully figured out and calibrated as well as opposed to the rather misplaced bass lift on 'Second to None'. So identical Sony transfer, but with the absence of poor mastering levels, not fully figured out eq work done in the Sonoma environment and resulting bit-depth-reduction noise-shaping, should make the more faithfully presented, flat ELVIS AT SUN more appealing to some.



That's All Right waveform, 
A=Second to None, B=Elvis at SUN
The philosophy behind Second to None is obvious: "If less than 10% of the samples are clipped no human being can tell"...
The price of the relative amplitude approach of Elvis a SUN is that the loudest, compressed songs will have to be mastered at about -3dB.



That's All Right  
SAMPLE (2)
1973 'SUN' 45 RPM copy of 78 RPM SUN 209
The original, dry sound. Tape is lost, but for historical reasons, there is all the reason in the world to get back to this source in a different context in the future.


Blue Moon of Kentucky 
A-B SAMPLE
A. Previous BMG Master, Great Country Songs (2003)
B. Elvis at SUN
RCA's 78 RPM transfer from 1955 has finally been replaced. One of the worst recorded SUN masters, but the new source has much better dynamic range. 



Blue Moon of Kentucky waveform, 
A=Great Country Songs, B=Elvis at SUN
The new transfer reveals the extra RCA compression done in 1955. 



Blue Moon 
A-B SAMPLE 1
A. Previous BMG Master, 50's Box
B. Elvis at SUN


Blue Moon 
A-B SAMPLE 2
A. Outtake from the same tape (Sunrise)
B. Elvis at SUN
It is restoration that makes the difference here.



Tomorrow Night
A-B SAMPLE
A. Previous BMG Master, 50's Box
B. Elvis at SUN


I'll Never Let You Go
A-B SAMPLE
A. Previous BMG Master, 50's Box
B. Elvis at SUN


Just Because
A-B SAMPLE
A. Previous BMG Master, 50's Box
B. Elvis at SUN
A difficult restoration job to reverse RCA's processing. The intro has always been cut and is thought to have been so on the SUN tape. Same thing happened to the intro of Charlie Rich's Rebound and the ending of Johnny Cash's Wreck of the Old '97.


Good Rockin' Tonight 
A-B SAMPLE
A. Previous BMG Master, 50's Box
B. Elvis at SUN


I Don't Care if the Sun Don't Shine
A-B SAMPLE
A. Previous BMG "Alternate" (Sunrise)
B. Elvis at SUN
Listen for the "living stereo" fluctuations and 16 kHz ringing of the unfinished 'Sunrise' mastering.


Milkcow Blues Boogie
A-B SAMPLE
A. Previous BMG Master, 50's Box
B. Elvis at SUN


You're a Heartbreaker
A-B SAMPLE
A. Previous BMG Master, 50's Box
B. Elvis at SUN



You're a Heartbreaker spectral plot, 
A=Sunrise, B=Elvis at SUN
The response of the recent 78 RPM transfer (B) may not look fantastic and there is a lot of noise that had to stay. Still, this should be compared to everything that has been released before.


I'm Left, You're Right, She's Gone (slow)
A-B SAMPLE
A. Previous BMG Master, Sunrise
B. Elvis at SUN


I'm Left, You're Right, She's Gone
A-B SAMPLE
A. Previous BMG Master, Sunrise
B. Elvis at SUN


Baby Let's Play House
A-B SAMPLE
A. Previous BMG Master, 50's Box
B. Elvis at SUN


I Forgot to Remember to Forget
A-B SAMPLE
A. Upgraded BMG Master, Great Country Songs (2003)
B. Elvis at SUN



I Forgot to Remember to Forget waveform, 
A=Great Country Songs (2003), B=Elvis at SUN
The philosophy behind Great Country Songs seems to be: "If less than 5% of the samples are clipped no human being can tell"...





I Forgot to Remember to Forget zoomed waveform, 
Great Country Songs (2003)
A closer look at a clipped vocal peak as mastered by Sony on 'Great Country Songs'.


Mystery Train
A-B SAMPLE
A. Previous BMG Master, Sunrise
B. Elvis at SUN


Trying to Get to You
A-B SAMPLE
A. Previous BMG Master, 50's Box
B. Elvis at SUN


When It Rains It Really Pours
A-B SAMPLE
A. Previous BMG Master, Sunrise
B. Elvis at SUN



When It Rains It Really Pours spectral plot,
A=Sunrise, B=Elvis at SUN






 
                    
   Recently recovered RCA 30 ips tape



     
                            
                           Scotty and Bill credit on RCA...




 


     






                                 ADVERTISEMENT
                        A good friend is exploring new markets. Direct link to Elvis at SUN, but works only if Apple software is installed.


FOOTNOTES

SUN TAPE ECHO


"By 1954 Sam Phillips had upgraded his equipment and installed two Ampex 350 recorders: one a console model and another mounted behind his head for the tape delay echo, or slapback". See http://www.history-of-rock.com/sam_phillips_sun_records.htm

The slapback/echo tape machine (#1, the mounted model) would typically record a dry signal from a local mike. The tape with the local signal, recorded less than a second ago, then reaches the playback head of the same machine. (Length of delay is governed by the tape speed of the echo machine.) The played back dry, local signal then goes into the console where it is mixed with signals picked up by all the microphones in real time, including the one also recorded on the reverb tape machine. The console tape machine finally records the mixed signal on its only track.  

At least one example exists where the tape played the through loop survived as well as the mixed result recorded on the other machine. This example is for an outtake of Get Rhythm (Johnny Cash & The Tennessee Two), combining a Bear Family source erroneously labelled "mike test" (actually #1: the reverb tape) with a Charly Box source (#2: the mixed result of the same take).

Johnny Cash could apply enough pressure on the strings with his left hand to damp the tones of his acoustic guitar (not necessarily done with paper/dollar bills attached to the strings as the story goes - video proof from 1957 exists). Mixed with bass slaps, delay, compression etc., it sounds like a train moving. In this example the vocals would be delayed as well:

1 = Dry as found on Bear-Family Box, vocals + acoustic guitar (other instruments bleeding slightly)
2 = Quickly added compression+delay to the dry track-just as an example-not the right proportions
3 = The result as found on a Charly Box
4 = [Left=3, Right=2]

Assume a distance between the recording and playback head of 4 inches and echo tape speeds of 
7.5 ips: Delay = 4/7.5 = 0.53 seconds (most likely the most common echo tape speed).
15 ips: Delay = 4/15 = 0.27 seconds (When It Rains It Really Pours).
(Update July 2004: 7.5 ips as the common speed of the echo tape machine has been confirmed by Roland Janes of Sam Phillips studio - thanks!)

So the distance between heads would correspond to the longer path if the wet signal is delayed with about half a second (7.5 ips).

Speed of sound at 25 centigrades (77 F): 347 m/s. It would take a distance of 180 metres to get the same delay (as 7.5 ips) that way. This means that it would be difficult to measure if a signal was recorded twice from both amplifier and microphones in a studio, but on the other hand the delay ought to be too small to be worth that extra procedure. One can but wonder if RCA thought of this in January 1956 when Heartbreak Hotel was recorded - perhaps they found out the hard way when they had to use the stairwell eventually.

In the case of the Johnny Cash example, the delay applied to the microphone picking up Johnny's voice and guitar only. Although this was the most typical setup with the Cash recordings, the Elvis Presley recordings sometimes used delay of other microphones. Probably forgotten by now, That's All Right was actually recorded and released entirely dry using just one tape machine. In late 1955 RCA added echo to it during transfer of originally received SUN Tape Box #2. Unfortunately Box #2 was one of the tapes destroyed in Indianapolis after 1957 so no tape source of the original, dry That's All Right is thought to exist anymore. However, the released outtakes from Tape Box #13 as well as the original 78/45 of SUN 209 reveal the original, dry sound. We can guess that there wasn't time to set up the two machines and this adds credit to the story about the July 5 session being planned as just a rehearsal. On one of the following days, Blue Moon of Kentucky seems to have been recorded with full reverb, but it's hard to tell since it appears to have been processed afterwards as well. Perhaps further echo was added by Bill Putnam during mastering.


MYSTERY TRAIN MYSTERY

Background:
The complete song, including the last guitar note played, before the fade-out is over, was found on a cheap European CD release, 'Elvis Presley - The Legendary', The Entertainers CD 254 (reported by Sven Adamski). Several variations exist (1987-1990, France Italy etc.)
Before that it was used on the UK LP's "Rock 'n' Roll", RCA 1972 and HMV 1956 (reported by Kevan Budd). Recently it was traced back to US and the 1955 RCA 78 RPM release (reported by Ernest Boyes) and now it becomes interesting.

Summary of versions:
1. Unprocessed master with very good dynamic range (50's Box, earliest fade).
2. The processed Sun 45/78 master (Sun 45/78 & When All Was Kool, ends suddenly during fade)
3. The processed RCA version (1955 single, complete with last note before fade is over)

Problem:
In late 1955 it appears RCA hadn't received "Sun Box #16" and it has been confirmed that this is the tape pictured in Sunrise ("Bill: Give me 'hot' level..."). If this is the case, then it can be assumed that RCA had to use Sun Box #1 in 1955 and that Box #1 had a version with the complete ending. This might also explain why the long version was sent to HMV, with UK being one of the first export markets (somebody ought to check Canadian and German releases for the same reason).

RCA did some processing of their own on the "long version". By comparing the dynamic range of the long RCA version to the original Sun version it may appear as if that the original Sun compression was on the 1955 RCA version as well. Drop of vocal amplitude due to compression appears to occur at the same points and of the same amount (L/R comparisons). Still, RCA probably was able to produce the same amount of compression - or my method of comparing may not be valid enough.

The task of "Bill" (Bill Putnam) was to produce acetate masters, probably not a tape. Yet, a source that corresponds to a "Bill tape" is available on 'When All Was Kool'. This version is the processed Sun master with the cut ending, likely to have been derived from Box #16.

Possible scenario:

-Box #1 ought to have been the original session tape (with ending intact)
-Box #16 could be a copy of the unprocessed master with dynamics, ending possibly cut due to so early fade.
-Unknown tape: The processed Sun master with ending cut - this shouldn't exist, but it does. Bill might have made a tape and sent to Sam that we don't know about.

This scenario will change as we learn more. Some things still don't make sense, such as, why RCA would choose to use Box#16 for their own master copies in 1956 if this had the ending cut, was just as dry as and a generation down compared to Box#1? In reality things don't always make sense on the other hand.