ELV1S 30 #1 HITS:
An Audio Review in the Spirit of Steve Sholes

In spite of more than 400 CD titles being available on for instance Amazon, not too many of Presley's master recordings had been released as expected, in a format decently faithful to source in September 2002 when Elvis 30 #1 Hits was released. Most of the catalogue as of September 2002 was transferred to digital and remastered during the first half of the 1990's. The problems with these remasters were much due to exaggerated faith in new audio restoration techniques. All old ambient audio recordings appears to have been considered corrupt and in need of restoration. If there ever had been more to evaluating sound quality than signal-to-noise ratio, this was long forgotten.

Waking up ten years later, it can be concluded that the untouched, basic digital transfers probably weren't archived and that physical transferring of old master tapes is a large project, because later releases of masters were all derived from these early digital audio restoration experiments. Ironically, this situation eventually resulted in better sound quality of more recently remastered outtakes that by far surpassed the quality of the masters. Not that there is anything wrong with outtakes released in stunning sound, but they can never replace the masters.

With this in mind, the news of ELV1S 30 #1 HITS being remastered from scratch, using a contemporary approach left some both excited and worried. Excited, since a sonic upgrade of these recordings was long overdue. Worried, because of what 'contemporary' might mean. Producer David Bendeth was kind enough to answer this and other questions before the album eventually was released in September 2002:

THE 50's MASTERS
01. Heartbreak Hotel
02. Don't Be Cruel 
03. Hound Dog 
05. Too Much 
06. All Shook Up 
09. Don't 
11. One Night 
12. A Fool Such As I
13. A Big Hunk O' Love

THE HOLLYWOOD SOUNDTRACKS 
04. Love Me Tender
07. Teddy Bear
08. Jailhouse Rock
10. Hard Headed Woman
17. Wooden Heart
20. Can't Help Falling in Love
23. Return to Sender

THE 60'S NASHVILLE MASTERS
14. Stuck on You
15. It's Now or Never
16. Are You Lonesome Tonight
18. Surrender
19. His Latest Flame
21. Good Luck Charm
22. She's Not You
24. Devil in Disguise
25. Crying in the Chapel

THE AMERICAN SOUND RECORDINGS
26. In the Ghetto
27. Suspicious Minds

THE 70'S MASTERS
28. The Wonder of You
29. Burning Love
30. Way Down

2002 Remix
31. A Little Less Conversation (Radio Edit)

M&S: Naturally, modern mastering will mean that the amplitude has been raised between 6 and 9 dB or so over the peak 0 dB-limit?

DB: Ours was very much a contemporary approach and although you might be right about the mastering in 6-9 dB territory. You also have to understand the process chains in mixing the record. This was not a simple noise floor to deal with and louder to us does not mean better, just means louder. Utilizing modern technology was certainly handy, however I think you will find the full dynamic range intact especially in the vocals and the bottom end. Nothing has been added and nothing has been panned in any place it was never before.

"Full dynamic range" after mastering 6-9 dB over the limit doesn't sound possible, but let's take a closer look using a modern consumer approach.

The 50's Masters
First song is Heartbreak Hotel and it makes an excellent choice for examination. Alternate take 6, recorded just before the master can be found in good quality on the 1997 'Platinum' 4-CD box, but the digital representation of master take 7 had been devastating. That is, until now because the new remaster sounds remarkably good and not too far away from source - on first listen.

The three main problems with the 50's masters are gone: 

1. There is no sign of any kind of noise reduction and the benefit of that is ambience. 

2. The high and low frequencies have survived the equalization process. In fact, high and low frequencies have even been increased or exaggerated, as we'll find out.

3. The song is presented in genuine mono. The digital "bathroom" echo of the early 1990's is not present here and who will miss it? The original stairwell echo should be enough.


Spectral plot of 20 second intro of Heartbreak Hotel master. The left source (E1)
has more natural, or even exaggerated frequency response up to 22 kHz and good signal
above 15 kHz. The right source (50's Box etc.) is the previous, noise-free, official
version that we had been living with for the past 10 years.
X-Axis=Time; Y-Axis=Frequency components (black=silent; yellow/white=loud)

With these crucial problems gone, let's go into the details. One such detail is a fluttery noise during the more silent parts of Heartbreak Hotel. This noise is thought to be a heritage from the original recording rather than just the result of processing, although the new equalization and dynamic processing do make the fluctuations more audible. This assumption is supported by re-equalizing and limiting the Platinum outtake to match E1 a little.

The equalization of Heartbreak Hotel as well as all of the other masters on E1 involves increased signal in the mid-to-high frequencies (from 600 Hz to 22 kHz) as well as the deepest bass (below 80 Hz). This roughly can be described as the opposite direction compared to before and this increase of high and low frequencies is thought to be significantly more than needed.

Much of the E1 equalization has been done dynamically, which is funny because dynamic compression of original record masters often involved dynamically raised low level signal of high and low frequencies - typically increasing the ambience, deep bass and of course, noise. The main difference here is that the dynamic high-end raise goes all the way down to the higher midrange and you normally don't compress the higher midrange/lower high end (or say 600 Hz-3000 Hz) this much unless the entire frequency range is processed. This is controversial to say the least. Whether it is brave or stupid is subjective, but at least one vote goes for the latter. It is the reason for the increased vocal presence on the other hand, even on a song recorded in mono. One drawback is that the vocals sometimes sound very hard, but this is also due to the amplitude raise above the limit. Apart from this volume boost sounding a little unnatural it also leaves audible clipping on the following loud parts: 

1:03 "if your BABY* leaves you"
1:30 GUITAR SOLO*
1:44 "although it's always CROWDED*" 
2:00 "they could DIE*"

* = broken signal due to amplitude raise above the absolute 0 dB-limit

Still, an average power comparison indicates that this song only has been pushed about +2 dB above the limit, which is less than many of the other remastered #1's - the worst is yet to come.

Since Hound Dog had been the worst sounding 50's masters during the 1990's, it is a relief to find that a good source tape had been used this time, i.e. a source tape not being the 1957 'Golden Records' album master. This is thought to be derived from a similar tape source as the excellent Rhino version, although the processing is very different. The compression, high-end raise and the amplitude raise about +3 dB above the digital limit gives a loud, even aggressive sound and the pitch is out of place and at least 1.9% too fast. The main reason to search for the Rhino version may be gone, but it subjectively remains the best source to many, simply because of the more careful dynamic compression if nothing else (and it doesn't say 'pop' near the end...). 

The other non-soundtrack 50's masters, Don't Be Cruel, Too Much, All Shook Up, Don't and One Night, have all been processed with similar equalization, compression and amplitude above the limit as Heartbreak Hotel and Hound Dog. Outtakes, instead of masters, were used for A Big Hunk o' Love and A Fool Such As I. These will not be examined. The amplitude of All Shook Up has been raised with as much as +4.5 dB (68%) above the 0 dB limit, sacrificing the slap beat peaks mainly, but there are clipped vocal parts as well. Too Much, that used to run a little too fast is now presented fairly correctly in the key of A_b. 

A decent digital version of Don't, not derived from the damaged 50's Box restoration, was released on the CD 'From the Heart' (PD 90642, BMG 1992) and another, acceptable, but slightly noise-reduced version could be found on 'Elvis Sings Leiber & Stoller' (CD 3026-2, BMG 1991). The remastered version of Don't has only been raised by about +0.5 dB above "what's possible". More couldn't be justified since the recording has a lot of dynamics and overstepping the mark even this little still breaks a few vocal peaks here and there, but the result may be considered acceptable to some. The high-end raise may have been taken a little too far, but this is again a subjective issue and, unlike the broken vocals, it can be adjusted.

The Hollywood Soundtracks
The Hollywood soundtrack recordings on E1 are represented by Love Me Tender, Teddy Bear, Jailhouse Rock, Hard Headed Woman from the 50's and Wooden Heart, Can't Help Falling in Love and Return to Sender from the 60's.

M&S: I hope you had the opportunity to use the stereo tape of Love Me Tender (Aug. 24, 1956) since the dynamic range appears to be a little better on that.

DB: The "Love Me Tender" we worked on was not the one mentioned. However, after multiple hours of figuring out the wow and flutter, we settled on an extremely balanced and diverse mix. There was a problem with the original taping but using the technology we had we were able to make this baby sit down. It is a great performance and vocal take and I think you will be happy with the results.

So the source of Love Me Tender is unfortunately the flat and hissy mono tape and there certainly was a problem with the original taping. A little research might have avoided use of this inferior source and maybe some of the extra work. The restoration job as such is better than expected. The hiss reduction is well done considering the source, but what a waste of studio time.

The Radio Recorders songs were recorded in a different manner than other studio recordings and any attempt at remastering them from the original backup tapes, mono/2-track('binaural') of the 50's or from the 3-track of the early 60's, must be regarded a true challenge because of the flat, dynamic sound, the dry vocals and tape hiss (mainly 50's). The 1990's remasters typically involved overdone digital echo of vocals resulting in decrease of vocal presence and intact dynamics, whereas the vocals were carefully compressed and only had a little (more natural) echo added on the original 60's stereo masters.

The 1957 Jailhouse Rock mono masters from the master reel turned over to RCA had a natural-sounding tube echo and ambience as well as very nice dynamic compression, especially of the bass. One could easily have guessed that the masters on that tape, dated August 22, 1957 were processed live during recording, but that shouldn't be the case. It should also be pointed out that it wasn't until the 1962 Girls! Girls! Girls! soundtrack that RCA really knew how to produce soundtrack stereo masters properly and that the standard of the processing gradually became more uneven or even sloppy after 1964. The original stereo masters of Blue Hawaii and especially G.I. Blues were far from perfect in the first place.

Remastering the Radio Recorders material today will have to involve compromises and we shouldn't expect miracles. The use of refurbished vintage analogue equipment in order to produce a natural echo/reverb as well as careful compression of vocals would certainly be one option. Hopefully it will be possible to do it just as good or better entirely in the digital domain in the future.

When asked about the overall processing on E1:

DB: We brought out the air in the top and the bottom end by using by using analog eq API and Neve 1081, 1082 four band eq. The vocals were mostly suited to Massenberg stereo eq, Neve compressors and Lang tube, LA2a and LA3a tube compressors - always using a lite approach without squishing it all down.

Without being familiar with the equipment mentioned, a quick Internet search implies that at least some of the units were semi-vintage equipment that is being manufactured again these days. It seems some of the right decisions were taken when remastering the three 60's soundtrack masters for E1 and yet the result isn't quite what one would have hoped for. They still sound sharp and flat. It's interesting that an outtake of Can't Help Falling in Love (take 26) recently was released on BMG's 'Today, Tomorrow and Forever' box, that the same hiss problems were encountered and solved in a similar way (i.e. de-esser at 10.5 kHz on TTAF, 12kHz on E1). The job done on the TTAF outtake is in fact considered a little better. 

The 60's Nashville Masters

If some of the 50's masters are an improvement, the same cannot be said about most of the Nashville 60's masters. Some of the original 1960 two-track masters were very good in the first place, processed live with a characteristic tube sound. It isn't obvious that they need to be remixed at all, although any effort always is welcome as an experiment. On E1 the vocal presence is increased due to the nice use of compression on the vocal channel. However, the raise of lower frequencies is out of place. The signal near 50 Hz reaches ridiculous levels, especially on the first song presented in stereo, Stuck on You. This together with an amplitude raise almost 4 dB above the natural maximum gives a "boosted" sound with sacrificed dynamics and even distortion. If this is aimed at the general public, they must have ignored that a cheap modern micro-stereo typically has a set of different loudness buttons and that the chance of neither of them being pushed is close to zero. Then we have a "double-boosted" sound to move the building. 

The Nashville stereo recordings really suffer from the generic E1 processing. Nevertheless, Crying in the Chapel is one of two highlights on this release. It is not necessarily differently processed than the other masters, but the compressed vocals are very nice here. The mean conclusion is that the same processing brings better results on this particular song because of the sparse instrumentation.

 The American Sound Recordings

Next up is probably the best sounding remaster on this release, In the Ghetto. Like on Crying in the Chapel, the compression has increased the vocal presence without adding any unwanted effects. Again, it's the sparse instrumentation that makes this approach work here. It wouldn't be possible to get the same results with a full band/orchestra such as on the louder American Sound recordings. In other words, this would be well suited for a single release rather than on a Memphis Sessions anthology. Suspicious Minds, on the other hand, has several unwanted effects. The song is mastered far too loud. Nearly every peak is clipped and then we have the overdubbed brass causing flanging artifacts of other channels.

David commented this restoration before it was released and the following home-made stereo mix inserting mono vocals and doubled instruments of the soundtrack version of Frankie and Johnny was prepared for this review to illustrate this kind of inevitable flanging problem when combining the same instrumental track from two analogue sources.

DB: Remember on "Suspicious Minds" there were no horns on the master and "It's Now or Never"had no piano on the master. Both tracks have been repaired and are restored to their original overdubbed form... It took me FOUR DAYS!

The problem is that the channel with the overdubbed brass obviously wasn't isolated. It isn't possible to synchronize the other, duplicated channels with the precision needed. Maximum acceptable deviance would be shorter than 1/44100 or 22.7 ms in CD resolution. Even a 1-sample delay will cause some flanging. Analogue tape machines wouldn't come close to that precision of course. 
If the final overdubbed brass tracks recorded on August 7,1969 aren't available, better options would have been to omit the Las Vegas overdubs like on the 1987 Memphis Record, or to use the original stereo master. For an over-produced, major consumer release like this, re-recording the lost brass section might have worked best of all.

The 70's Masters

The 70's masters are represented by Burning Love and Way Down. The tape with Way Down had to be baked to reactivate the binder. A reminder that time is running out for some of the 1970's master tapes.

It appears as if the compression of vocals may have worked well on Burning Love and the new mix may have had some benefits, but all these efforts are more or less in vain due to the catastrophic volume raise. Way Down is even 5 dB louder than the Japanese "24-bit/paper sleeve" version and Burning Love might be the loudest remaster on E1. The result can be described as a pulsating loudness effect as the original peaks try to break the 0 dB barrier about 3-4 times per second. Unnatural and not just occasionally clipped, this is plain distorted. The promised "full dynamic range" occurs during the first 5 seconds and during the fade-out of Burning Love.

Footnote: Dynamic range describes the ratio of the softest sound to the loudest sound


Hard to guess, but this is the waveform of Burning Love with "full dynamic range".
X-Axis=Time; Y-Axis=Sample Value (+/- 0-32768).
 
Summary

An interesting experiment with a large budget, perhaps too large. It is remarkable that the processing has been called so "contemporary" when in fact another description could be "revival of traditional dynamic compression". The use of tube compressors says it all. The major difference is that the upper midrange to lower highend (or 600 Hz to 1500 Hz) has been dynamically raised this much. This increases the vocal presence more than usual, but also results in vocals sounding unnatural. The contemporary part of the project is instead the terribly overdone amplitude raise above the limit, let's call it "static compression". The artifacts introduced due to this spans from a few clipped vocal peaks of Heartbreak Hotel to the totally distorted Burning Love. Going this far with amplitude raise above the limit would not be acceptable on a release for collectors or anybody interested in audio. Bearing in mind that this is aimed at the general public, or a new generation that never has had a chance to listen to music with dynamic range, it probably will go unnoticed. 

Since its release in September 2002, a myth has been created that the sound is so much better than other releases and that one of the main reasons is that the true master tapes haven't been used before or since. Many thought that the early 2003 DSD releases (e.g. "ELVIS 56") sounded dull in comparison. The DSD releases weren't subjected to dynamic compression at all so a simple comparison will not reveal quality of tape sources and digital transfers.

The E1 digital transfers were done using a Sony 3348 in 24 bit/96 kHz PCM and it should be pointed out that the different sound that everybody hears has nothing to do with PCM or DSD. The following A-B-C audio sample of Don't Be Cruel was prepared to illustrate this:

 A. Either the first or second sample is the ELV1S 30 #1 HITS version.
 B. Either the first or second sample is a clone attempt derived from the 2003 DSD release "ELVIS 56".
 C. The third sample is the 2003 DSD version of "ELVIS 56", as released.

Guess which is which of A and B.