I was involved in the Professional Audio field in the 1970s and 1980s and learned much about acoustics and the fundamentals of recording sound. In particular, the design and application of microphones was an area of special interest. Of course, the transition from analogue to digital recording has been studied with great interest, with techniques developed and applied accordingly, although, one still misses the analogue days of tape, which invoked an artistry of their own.
A Reference Book for Audio
The book, ‘Audio Technology, Music and Media’, published by Springer Applied Sciences, is a useful reference for anyone interested in the history of recorded audio. It will be of particular interest to audio engineers and producers as it also covers recording techniques and more from a practical perspective. It is highly recommended for all those interested in audio and the journey from performance to produced media.
An independent review of the book may be found here.
In the interest of promoting true stereo sound, I offer recording, engineering and mastering for orchestras, smaller ensembles, choirs and ethinic music of all kinds, usually at no cost if they are amateurs or performing for charity. There are various halls available for rent at low cost nearby. Alternatively, I will travel for special events if transport is provided. This is a valuable service as very high quality results are obtained with both high resolution files and a reference CD being provided, which the artists may use, without restriction, for whatever purpose they wish. Many have already benefited from this service. Indeed, I have recorded 24 concerts / events in the past two years alone. I can also provide high resolution files in various formats for download from the Internet.
It seems that, these days, not many understand what true stereo sound actually is, as the majority of recordings they hear have been recorded with multiple microphones being connected via a mixer into a computer recording programme or digital recording device. This can provide a nice, clean sound, but it is not the real sound. So here is an opportunity to be recorded in true stereo sound, and create a record of how your orchestra, ensemble or choir actually sounded at a given point in time. This is invaluable for archiving purposes as well as creating great sounding recordings.
Stereo sound was defined by Alan Dower Blumlein in 1929, although, was not put into practical use until the advent of the tape recorder in the late 1940s. Even then, the first recordings were monophonic with companies such as EMI producing their massive BTR 1 tape recorder which eventually became converted for stereo use. Blumlein not only invented the concept of stereo, as well as multi-channel sound recording, but also designed moving coil microphones and defined their proper use in order to record a three dimensional sound field with just two channels. These techniques have become largely forgotten, except by a few specialist sound engineers, with the advent of multi-channel mixers and tape recorders (now digital recording straight to computer or dedicated digital recorder). The multi-channel approach found favour particularly within the field of popular music, as it allowed for multiple takes or ‘passes’ to be made, gradually building up a wall of sound. The classical music fraternity soon followed suit as it allowed for microphones to be placed in front of individual instruments or groups of instruments and the sound balanced after the event. But there are problems with this approach. In particular, the complex phase relationships between instruments are destroyed and the natural reverberation that exists within and between instruments can become unbalanced and unnatural. In addition, it results, with orchestral recordings, in the sound of multiple instruments, artificially balanced together, rather than the sound of the orchestra as a whole, and as a member of the audience would hear it. The orchestra ceases to sound like an orchestra, which is a shame, because the sound of a full orchestra, or even a chamber orchestra, playing together is a beautiful sound. Consequently, the stereo techniques defined by Alan Blumlein remain valid today. A coincident, or near coincident, pair of microphones can, if positioned correctly, capture a wonderful, three dimensional sound which sounds like an orchestra should sound. Furthermore, the technique is just as valid for a small folk group, a choir or even a solo musician, as true stereo captures the sound as it was heard within the space in which it was created. I continue to use this technique, plus some special variations of my own, according to the venue concerned.
provides a valuable insight into recorded sound, as well as sound in general. Enthusiasts will find much of interest within its pages, regardless of the particular sphere in which they operate.
When preparing the final, master, files, a minimalist approach is taken with as little alteration as possible to the original sound file. Typically, it is just a question of inserting silence at the beginning and end of the file and checking the overall sound. No equalisation or alteration to the original dynamics will be undertaken unless absolutely necessary. Minor blemishes may be corrected. The resulting sound retains the freshness and immediacy of the original performance. This is in contrast to those who like to apply compression, tonal correction and various special effects, all of which serve to degrade the original quality of sound as any alteration at all will have an effect. This is typically heard as a slight ‘softening’ or ‘blurring’ of the original signal. Resampling (for example to 44.1 kHz for CD) also affects the immediacy of the original sound although, in many cases, the trade off remains beneficial due to the improved quality in the middle registers. It all depends on what is being recorded and for what purpose.
Mastering an original sound file
High Resolution Audio
The arguments for and against recording in high resolution are many. There is a tendency to now consider ‘CD’ sound as not good enough but, actually, a good recording made at 44.1kHz sampling and 16 bit can sound superb. Personally, I like to mostly record at 48kHz and 24 bit, depending on the source, as this does provide for a useful extension in dynamic range and higher frequencies. However, if you examine a spectrogram of an orchestra playing a typical piece, you might be surprised at how little sound there is at the higher frequencies. In recording solo violin, I have noticed, with the highest notes, content just short of 30kHz, which will include the fundamentals and harmonics. But this is quite rare. I have recorded quite a lot at 96kHz sampling rate, but the audible difference in sound quality between 48kHz and 96kHz can be minimal, depending on what you are recording. Instruments which are rich in both harmonics and detail will benefit from 96kHz sampling. For example, when recording a solo accordian, the higher sampling rate certainly captured the overall timbre of the instrument more faithfully. Generally, I am much more interested in capturing the dynamics of a good performance and maintaining the immediacy of sound that one hears at the podium position. Of course, the other point about high resolution files, is that you have to have something to play them back on. It is not enough that such a device can play back high resolution files, it also needs to have the inherent quality to exploit their range, as does any other component being used (amplifiers and loudspeakers for example). In addition, if you record at 96kHz or higher and then downsample to 44.1kHz to produce CDs, does the refinement of the higher sampling rate remain audible? Many maintain that it does, albeit in a diluted manner. Audiophiles will no doubt discuss this, and related points, until the cows come home. However, if you listen to some of the best classical recordings made in the early 1960s, using simple techniques, you might wonder what all the fuss is about.
Lossless audio compression
There are several popular lossless audio formats, the foremost being perhaps the open source flac format and the ape format from Monkey’s Audio, plus some platform specific algorithms. They all claim to be lossless. That is, you should not be able to tell the difference between an encoded file and an original WAV file (including at high resolutions). However, if you experiment, you will find that they all sound subtly different. Consequently, they cannot be truly lossless. In addition, if you experiment with the encoding settings within any one format, you will find that the results also sound slightly different, the faster settings always sounding better. Which also gives the lie to them being truly lossless. If they were, they would sound the same at each setting.
However, we are really splitting hairs here as, in reality, they do sound very good indeed. I have a preference for flac, encoded at the fastest setting, although I have found the ape format to be especially good for choral music and opera. The Apple and Windows proprietary formats have their own characteristics. As with all things in audio, it is good to experiment and discover your own preferences.
Its becoming popular to remaster existing recordings, supposedly rendering them in higher quality sound than the original. Sometimes, this is indeed the case, but not always. Remastering requires a combination of art and scientific understanding if it is to be successful. It is easy to loose information during the process, especially regarding the leading edges of transients, resulting in files which are slightly ‘soft’ to the discerning ear. However, if undertaken with care, then acceptable results may be obtained, but only if the mastering engineer understands his craft.
Sadly, some remastering engineers seek to ‘improve’ the original recording by adding characteristics which were never in the original, whether it be extreme equalisation or artificial stereo sound. In some cases, even instrumental lines are added which were not in the original composition. Consequently, good audio remastering should be akin to the conservation of a work of art in other mediums. The original should, if possible, be restored to its original condition. In the case of audio, the removal of tape hiss may also be entertained where present but, other than that, let’s stay faithful to the original which, after all, was intended as a faithful record of the original event.
Microphone preamplifier noise
Recording engineers like to ponder the various specifications of recording devices, especially those who are recording audio for video. Their favourite seems to be the noise floor of microphone preamplifiers and their are endless dialogues on the subject on the Internet. They seem to believe that the equipment boasting the lowest noise preamplifiers must be the best. And yet, this is simply one of a number of factors which influence the sound of a recording.
Firstly, the microphones used impart their own signature upon the recorded sound. They, in turn, have a brace of factors which make up this signature, including their inherent signal to noise ratio, their self noise, dynamic range, sensitivity and, especially, their transient response or speed. The diaphragms used in microphones behave differently according to their overall design and the materials used. Condenser microphones have internal electronics which similarly vary between models and add a little colouration of their own. Then there is the combination of microphone and microphone cable to consider. Many will argue that balanced cables should not affect the sound. However, they are a physical transmission device with electrical properties of their own and, as such, must impart a small amount of character to the sound as received at the microphone input to the mixer or recorder. So, there are many variables before we even reach those microphone preamplifiers. When we do reach them, they have several factors, other than noise, to consider. These include their own transient response, dynamic range and the overall character of sound that their electrical circuits produce. The signal will then pass through other circuitry before reaching the analogue to digital convertors. These devices play a large part in the perceived sound quality of a recording device. There are a plethora of factors inherent to their design which will add a further colouration to the sound as it travels through, finally reaching the recording media, such as, for example, an SD card. The quality and reliability of this media also have a part to play in the realised recorded sound. Every digital byte leaving the analogue to digital convertors must be imprinted cleanly and fully onto the media, with no errors or omissions.
Finally, we have our recorded sound. However, it’s journey from the microphone diaphragm to the recorded media has passed through a number of stages, each one of which has influenced the original sound wave. It seems churlish therefore to focus upon one parameter, such as microphone preamplifier noise and use this as a yardstick with which to judge the performance of recording equipment. Personally, I would be more interested in the overall musicality of the entire chain, from microphone to SD card (or computer hard disc). In addition, factors such as build quality, reliability and usability will be more important to many recording engineers than the odd dB of theoretical preamplifier noise which, in any event, should never be considered in isolation.
Following on from the discussion about microphone preamplifiers, there is the additional factor of ambient noise to consider. This should be taken into consideration when considering noise floors and audio performance in general. The level of ambient noise varies quite considerably under natural conditions. This is why recording studios go to great lengths to create a quiet environment. Under typical symphonic concert conditions, an ambient noise of around -45dB to -50dB will be encountered, much higher than the noise floor of modern recording devices. And this is when the audience is quiet. At interval times, audience movement and chatter will raise this level dramatically. For broadcast speech and video work, an ambient noise of -60dB may be considered acceptable. For recording in natural environments, especially at night, one often expects the ambient noise to drop to around -70dB or lower, but this very much depends upon location. In some areas, wildlife has its own ideas about ambient noise levels and, in some areas, it actually increases at night. For the purposes of recording live sound, within typical venues, if one can achieve an ambient noise floor of around -50db, that may be good enough to achieve a good recording. Of course, we can attempt to filter out some of this noise, especially that which occurs below 40Hz or even 80Hz. The problem is, that much of it is much higher up in the frequency range and, if removed, may affect the wanted signal. Consequently, in many cases, we have to accept that there will be a degree of ambient noise and that this will be at a higher level than the noise floor of the recording equipment being used. However, this need not detract from the enjoyment of a good live recording. After all, the term ‘record’ may be defined as recording the original event, including its natural noise signature. Under studio conditions, we remove this natural signature and create a cleaner, quieter sound. But is it any better in absolute terms? That is for the listener to decide.