Missing from the ‘golden’ chain was the Ampex Valve Analogue Tape machine used for these recordings. Waves acquired what they believed what was the right machine and set about modelling…knowing it would be difficult based on what was already on the market. As it turned out the task was much more daunting than anticipated and Waves first attempt, the initial Kramer MPX, was withdrawn shortly after being introduced into testing. It turned out that this third piece of the Kramer Olympic chain was going to be the hardest.
Waves discovered (along with the guidance of some of our Beta Team) that they were missing a number of things, each one extremely complicated by itself, but in total represented a huge challenge. Thanks to the guidance of Bob Ohlsson and John Haeny, Waves were ultimately directed to the Ampex 350 Transport with Ampex 351 Valve electronics. Thanks to Bob’s help one of these rare beasts was found in Florida owned by Eric Shilling. Eric kindly agreed to let us do some preliminary testing to ensure that this machine was producing the sound that both Bob and John had missed in our first effort.
To clarify what appears to be a contradiction regarding the Ampex 350/351 Tape Machine being an American product, it should be made clear that at the time Eddie Kramer was working at Olympic Studios in London, it was Ampex 350/351 tape machines that Olympic and their clients were using as master recorders.
The Ampex 351 Valve tape machine was the mainstay of the recording industry (particularly in America) and was used for literally thousands of hit Albums and Singles over more that two decades of recording. For example in 1954 a recording studio equipped with an early Ampex reel to reel tape machine and an unknown truck driver named Elvis Presley recorded his historic first single, “That’s All Right” at Sun Studios in Memphis. The Ampex 351 was also the backbone of the earliest days of multi-track recording. The first of the Ampex 8-track recorders was custom built for and sold to Les Paul for $10,000 in 1957and was installed in his home recording studio. It became known as the “Octopus”. Ampex 8 track serial number 3 was sold to Atlantic Records at Tom Dowd’s instance in early 1958. Atlantic Records was the first record company to use a multi-track recorder in their studio on a regular basis. Just pause for a moment and contemplate all the great recordings that came out of that Atlantic Records recording studio. If they ever were to induct a tape recorder into the Rock and Roll Hall of Fame, it would be the Ampex 351!
With the modelling project still sounding reasonably workable, in the end it turned out that the recording pre-emphasis curve used, tape emulsion, it’s thickness, the Flux or level recorded on the tape (more later), and the Bias settings being used greatly impacted on the final result, so again, with Bob and John’s help there were endless discussions and tests made regarding tape types and alignment techniques until we settled on what we were going to model. Over time and with much experimentation, we finally made sample recordings that both Bob and John agreed had established a solid baseline from which to proceed. It would be these initial samples of the test recordings that would be used for detailed comparison to ensure our model was performing accurately.
Once again Eric Shilling came to our rescue and agreed to do the massive testing and modelling runs required to model the two primary tape speeds, a number of tape emulsions plus variations in Flux and Bias settings and approaches to alignment techniques.
With the modelling files in hand, Waves began the excessively complex task of creating models, not only of the variety of analogue tape recordings, but the variety of Bias, Flux levels (Saturation) and Speed settings. Once Waves began evolving early stage Alpha plugs, they required painstaking subjective analysis by comparison to the original samples recorded on the original tape machine and then though feedback to the development and engineering team at Waves, the models were honed. As we could only develop for one platform at a time, Waves chose to do their initial development for Macintosh, at which time Bob took a back seat and John Haeny essentially make it a full-time job by becoming a co-developer of this product.
What did we model and how does that affect the GUI and the operation of MPX?
Waves modelled 3M Scotch 206/207 tape as it was considered an almost an ideal match for this machine. There were other earlier tape types that were exceedingly popular as well, for example 3M Scotch 201/202/203 was used extensively by Motown, but sadly it was no longer available in a sufficient and durable enough quantity for our project. Of the 3M Scotch 206 (a 1.5 mil base) and the 207 (a 1.0 mil base), 207 was selected because the slightly thinner base, although perhaps subject to greater print-through (clearly not a issue for a tape modelled plug), provided a more intimate recording and playback head contact (called Tape Wrap) and thus produced a more extended high frequency response. Fortunately John Haeny had a sufficient supply of virgin 3M Scotch 207 stock for our purpose so off the shipment flew from Australia to Florida.
The 350 Transport has 2 speeds. 15ips (Inch Per Second), and 7.5ips. 15ips was the default professional standard providing the best high frequency response and the lowest noise. 15ips has a gentle roll off at around 16khz. 7.5ips was the minimum reasonable professional quality speed for studios and there was also a fair amount of equipment for the home that operated at 7.5ips. 7.5ips has quite a high frequency loss with a roll-off starting at around 8Khz, but 7.5 managed to preserve low frequencies better than the 15ips with a slightly more ‘solid’ bottom end and therefore was widely used in rock recordings in the 60’s and 70’s. (You can look this up, but it has to do with the relationship of the wavelength of the signal and the width of head gap…the slower the tape speed the easier for a ‘given’ head gap to record and read the lower, or longer, wavelengths.)
When switching between the 2 speeds you should expect to get a very much improved high frequency response with 15ips when compared to 7.5ips but perhaps a somewhat less tight low end. Note that 15ips will also provide less THD (Total Harmonic Distortion) than 7.5ips. There is also a shift in the frequency of the noise by an octave between 7.5ips and 15ips, with the noise at 15ips sounding an octave higher than 7.5ips. This difference has always been argued, with some preferring the noise signature of 7.5ips and others preferring the noise signature of 15ips. As time progressed and 30ips became popular many used 30ips because as the noise shifted upwards yet another octave from 15ips, it moved further away from the fundamental musical frequencies and thus became less obtrusive. Experiment and form your on opinions on these issue of speed vs. noise vs. frequency response.
At the time of the 351’s popularity there were a number of pre-emphasis curves in use worldwide. Because of the inherent limitations in analogue tape recording these curves generally applied high frequency pre-emphasis equalization during recording and then applied the reverse curve upon playback. The net result of this was to maintain high frequency response while reducing tape noise.
The most popular curve in Europe at this time was CCIR. In America the curve was NAB. Ampex also had their own short lived curve know as Ampex Master Equalization (AME). The machine at Olympic, as best we can determine, was a NAB machine although Ampex would provide CCIR machines on special order. Additionally, NAB was the curve used for the vast majority of all the American pop recordings done on this machine. The NAB curve provided the most accurate sonic signature of this legendary machine, so NAB was the pre-emphasis curve chosen by Waves for the Kramer MPX tape model.
Waves modeled and have provided you with two Bias settings. “Normal Bias” (this is the Ampex recommendation for Bias adjustment…straight out of the original Operators Manual for the 351). “Normal Bias” was recommended in the first years of production by Ampex to try and reach the best recording levels with the minimum amount of distortion and the maximum frequency response. This setting produced a fairly low noise level (noise is around 60db lower then the peak signal), and around a 2 to 3db of high frequency loss with a moderate amount high frequency distortion.
During the 60’s…after a number of years of working with these tapes, many Professional Studio Engineers and Technicians discovered that by over biasing (increasing the amplitude of bias signal) by only a small amount they could improve the high frequency response and at the same time lower the noise level. This was called “Over Bias” and each tape type, Studio, Recording Engineer and Technician had their own way of calibrating the bias to achieve their preferred sonic qualities.
For MPX we modeled the -3db over bias, which was agreed by most engineers to be the point where you got the best high frequency response, least amount of distortion and a best Signal to Noise ratio on 3M Scotch 207 tape. (Actually -.7db at 700hz for 15ips but set at -3db at 15Khz for accuracy … you adjusted nominal Bias to accomplish the peak level of the signal and then continued beyond the peak until the level began dropping the desired amount, therefore the term ‘over bias’.) When you switch between “Normal Bias” to “Over Bias” mode, you should expect to hear a bit less noise, clearer high frequencies (reduced distortion), and a bit more overall dynamic range (clarity … once again the result of less overall THD).
Defined: The power emitted by the head responsible for transferring program information on to tape is measured as a quantity of lines of flux or fluxivity once on the tape. This is measured in NanoWebers per meter (commonly abbreviated as nWb/m). When picking an operating level for tape (Flux), the general rule is that the higher the operating level the further away you are from the noise floor but the closer you are to the point of distortion
Flux is the amount of density of magnetic particles on the tape per meter (nWb/m), the higher the flux level – the higher the recording level on the tape. In order to reach higher recording levels you need to have tapes that can manage higher flux level (many modern tapes can reach very high recording levels with minimum distortion with the added benefit of reduced noise or ‘tape hiss’).
Ampex initially set a standard recoding level that they called “Ampex 0” (or Ampex Operating Level) for recordings made on their machines in 1950. This recording level was at a flux level of 185 nWb/m. All early alignment tapes were Ampex 0 or 185 nWb/m signals and were the standard for all recording. As tapes were developed to handle greater and greater flux (or higher recording) levels the industry continued to rate tape flux levels based on the original Ampex 0. For example a tape that was designed to record signals at 250nWb/m was said to be recording at +3db (over Ampex 0 of 185nWb/m). As point of reference, since the Flux Control on Kramer MPX is calibrated in nWb/m here is a quick reference guide for comparison (Source Quantegy):
-2db = 150 nWb/m
0db = 185 nWb/m (Ampex Standard Operating Level)
+3db = 370 nWb/m
+5db = 320 nWb/m
+6db = 370 nWb/m
+9db = 520 nWb/m
Science aside, early on many engineers discovered that by pushing or abusing the recording levels on tapes it created some very musical and frequently desirable side-effects, especially in Rock and Roll recording. Tape has a very unique way of going into saturation or overload. Whereas Digital is basically ‘go or no-go’ with over modulation producing clipping, as the levels are raised on analogue tape a number of things increase simultaneously and fairly gradually: Third Harmonic Distortion (THD), Inter-modulation Distortion (IM), Modulation Distortion, a mix of other aberrations of distortion, many of which are still not fully understood, and ultimately when pushed hard, tape has a tendency to ‘Saturate’ which is a form of compression, unique to analogue tape. Many engineers, even when digital was available, still preferred to record certain instruments on analogue tape (especially Rock and Roll drums) pushing the levels on the tape well beyond their ideal operating conditions to gain this saturation or compression.
Modelling this behaviour of changing sonic behaviour across a wide range of flux levels was perhaps one of Waves biggest challenges. Ultimately Waves was able to create a continuous flux control that ranges from -2db below the standard Ampex 0 well into unknown territories, invaluable for a wide variety of special effects.
Since 3M Scotch 207 was rated at between 185 nWb/m or 250 nWb/m (because opinions vary about the ‘ideal’ level for Scotch 207, we have chosen the more conservative level of 185 nWb/m for the MPX default, although we modelled a wide range of flux levels to be able to accurately create the variable flux control)), you will find that this plug will be especially sensitive to settings above the recommended level of the tape sampled. When reaching higher flux levels you will notice that the low frequencies and very high frequencies become more and more distorted, as the noise level goes down. Additionally there is another layer of distortion since these increased operating levels also stress the valve input and output stages giving you the bonus of an additional level of overall “tube distortion”.
Kramer MPX is the first tape modelling plug-in that has actually managed to create a continuously variable flux control, letting the user sonically understand the changes between different recording levels. Add to this that this is also being done on a fully valve electronics tape machine and you truly have a one of a kind plug-in…fully unique in the industry.
To simplify the MPX Flux control just think of it as ‘Tape Drive’ or ‘Saturation Control’.
As you increase the Flux level on the tape the output of the MPX plug remains constant (at Unity Gain)
unless you chose to adjust the output gain control.
Although the Noise on the MPX is defaulted to off, it is strongly suggested that you take a listen to it. If this can be said…you may find it to be some of the nicest noise you will ever hear since it is a combination of the tape hiss generated by analogue tape recording, overlaid by the thermal valve noise of the Ampex 351 input and output electronics.
Wow and flutter in analogue tape recordings results from the turning speed inaccuracies (FM), and flutter effect (AM) caused from changes in the physical alignment of the tape on the actual recording head, coupled with the ‘slip and slide’ of the tape going through the transport mechanism and a variety of mechanical ‘frictions’.
In a perfect world, there would not have been Wow and Flutter. Many ‘in the day’ considered Wow and Flutter the same as surface noise on a vinyl disc…just something negative one had to live with since that was ‘state of the art’. But like noise, there are many that will feel that any model is not complete without all the flaws as well. For that reason we have provided you a manual Wow control on the GUI of MPX. It is defaulted to the Wow and Flutter we modelled on our sample machine. You can increase it for a somewhat more enhanced effect (although it would never be too obvious [unless your machine was broken]…as it was always a subtlety of the analogue tape process), or if you chose, you can move to a more idealized world and turn the Wow off. Having or not having Wow and Flutter and or Noise has nothing to do with the primary advantages of the Analogue Tape Recording Sonics so use them or not at your own discretion. You will always have the advantages of the ‘Ampex Tape Sound’ with them or without them.
To sweeten the Kramer MPX package we added a variable delay control (0ms-500ms) that routes the tape playback of the plug-in back to the input of MPX. This creates a very basic feedback Tape Delay effect across the entire signal (the direct signal is always included in the mix). A Low Pass (LP) filter was added to the delayed feedback path to allow you to filter out any unwanted high frequencies. This delay/feedback feature is intended to be very basic. It does not offer a wide control section, but we think it contributes additional value to the MPX plug-in. Used carefully, this function has the capacity to create some very lovely tape delay sounds. It also needs to be pointed out that the delay is only effected by the Delay Time control in the Delay Section. It is NOT affected by the running speed of the transport. If you need to ask us why we included this bonus function, the answer is quite simple…because we could!
What to expect from the MPX
The Default Set-up of MPX, without any adjustments, will provide identical results to having made your recording on 3M Scotch 207 at 15ips using a NAB pre-emphasis curve with an over-bias of -3db at 15Khz and at a recording flux level of 185 nwB/m. This will yield the sound of the basic industry set-up for this machine at the height of it’s era. The only item remaining for you to decide will be whether or not to have tape and valve amplifier noise, and if so, exactly to what degree.
Of course you also have other options with regard to Bias, Wow, Speed and most dramatically,
a continuously variable Flux control starting at the ultra conservative setting of -2db below the classic Ampex Zero
all the way to some rather massive, heretofore unattainable in a plug-in, extreme analogue tape saturation effects.
Meter Transfer Switch:
On the face of the meter you will find a Meter Transfer switch that transfers the meter from reading the Input to the MPX plug or the Output of the MPX plug. This purely a meter transfer switch and has no effect on the monitored sound.
Mode Switch or Input/Reproduce:
In place of the classic Record light you will find an illuminated yellow or amber lamp. The label above it will be in the default mode of Repro, which means the output of the MPX is after the model of recording on tape. In the Input mode you will only be hearing the sound of the valve input and output electronic stages (or directly through the machine without any tape running). This is not a pure Bypass and this sound is, in and of itself, valuable for many applications. Other than the audio monitor transfer and the light being on for Reproduce and off for Input, when in Reproduce the transport reels will also be turning to insure you that you are hearing the result of recording on analogue tape.
What Should You Expect to Hear:
The sound of analogue magnetic tape recording may be new or un-familiar to some born in the Digital Age. The goal of the design of all analogue tape recorders was the same…to provide a transparent, colourless method of media storage. The Ampex 351 recorder was quite successful, in it’s day, at accomplishing this goal. But because of the state of the art at the time of it’s development it still had a wide variety of limitations: Tape Saturation, Tape Noise, Harmonic Distortion, Modulation Distortion, Phase Shift and limitations in frequency response to name but a few.
Although many think that analogue tape will improve the sonic quality of their signal, by today’s standards and measurements it does quite the opposite. Measured by modern technology it lowers the overall resolution of a signal. In fact the signal to noise ratio of an analogue tape recording is not good by today’s standards. It fails to accurately reproduce both high and low frequencies. It’s THD measurements are not good by current standards (more then 1% THD) and yet listeners still find its sound pleasing.
So just why is this sound so desireable? Well for a number of reasons, but firstly, contrary to measurements and the theoretical loss of high frequency response, because of the non-linearity of the NAB pre-emphasis curve, coupled with the Third Harmonic Distortion created by the analogue recording process….the ultimate subjective result is slight increase in the quantity and clarity of the higher frequencies.
Even as Digital Recording has come of age, and with hindsight 50/50, what many considered to be a limitation of analogue tape recording has, in fact, become desirable. Digital has been criticised by many as being clinical and cold sounding, while analogue technology has been touted as sounding warm, clear and musical. Therein lies the big difference in what you should hear in the MPX plug. Granted if you want to play with the settings (and please feel free to do so) you can create some very dramatic tape effects of Saturation, Noise, etc.. But by using the MPX with its default settings it will provide you with a extremely accurate model of, not only analogue tape recording, but of analogue tape recording on what many considered to be the premier valve analogue tape recorder…the Ampex 350 transport coupled with the Ampex 351 valve record and reproduce electronics.
A description of the sound of Kramer MPX is that it is best ‘felt’ as a warm, sweet and clear musical sound. Start there as your basis and then experiment with the controls to discover all the additional sonic qualities available to you and choose those that best suit your tastes and your style of music.
In Practical Application:
It is not possible to create a model with this level of detail and flexibility and at the same time keep it’s resource demands low. Running MPX does require a fair amount system resources as you might expect. If you want to used MPX on individual tracks…please do so, as there are no ‘rules’ in our art form. But MPX may perform best on sub-mixes of Drums and Percussion, Strings, Guitars, Vocals and anything that requires analogue ‘warmth and clarity’…and don’t forget what MPX could contribute to digitally sampled instruments and digital synthesizers. Many will also find that across an entire mix and especially in mastering, MPX may prove to be invaluable. MPX is the result of almost two years of failures and triumphs and contains the heart and soul of many contributors. Our sincerest hope is that you will enjoy MPX and that it will become an invaluable component in your sonic toolkit.
Kramer MPX used in conjunction with the Kramer HLS and the Kramer PIE finally completes the magic of the original Kramer Olympic Studio Recording Chain.