It is interesting how much the frequency responce of a microphone preamplifier must be reduced in order to use for the human voice.
If this is not done the frequency responce from singer vocal chords to listeners brain is not flat. The frequency responce from mic to loudspeaker is only part of the chain [see Hollywood reference].
If a typical studio directional microphone is used at less than a few inches distance the proximity bass increase may begin at as high as 1KHz. It then roughly increases by 6 dB per octave as frequency goes down looking like half of a bathtub curve. Such curves can be readily observed on any directional microphone data sheet in reference to small distance from source
To compensate for this the preamplifier needs to do the inverse. This means the frequency responce of the microphone preamplifier may need to start dropping at 1KHz (-3dB point) with a slope of 6dB over octave below 1KHz. It depends on distance. 3in source to mic may need 470Hz turnover frequency etc.
And this is only part of the story. Additionaly there are the Fletcher Munson equal loudness curves and the voice effort curves.
The first describe how our mind hears, how its frequency response curve changes by intensity.
The second describe how the frequency spectrum of voice changes by how loud we sing.
So after compensating for the proximity effect the bass and treble frequencies may still sound exaggerated to our brain. As an example we may need to gently high cut starting at 10KHz.
This needed process has similarity with the playback de emphasis used to compensate emphasis while disc recording. And it has beneficial results. By reducing treble to get the the flat perception described above, the hiss noise goes down too. And any higher order harmonic distortion of an electron tube, JFET, or bipolar junction transistor decreases too.
One way to implement the high cut is the effective gentle first order low cut filter. The old trick of a C series with R filter across the plate load, the resistor being variable. It can be seen as treble control in schematics of EMI, Emmerson, Dancette tube record changers etc.
This process has another beneficial effect. Reducing high order harmonics means rounding of the extreme clipped peaks that happen all the time when using music signals. Rounding makes them more acceptable and the higher average to peak signal ratio means higher loudness. Similar process to FM radio preemphasis de emphasis.
The above may produce a dramatic increase to signal to noise ratio and loudness.
So the best preamplifier for voice applications that need 3dB points at 1KHz and 10 KHz means bandwidth restricted by 6-7 octaves (from 20 Hz to 1 KHz is 6 octaves and 1 octave form 10KHz to 20KHz) to reproduce correct voice, ie flat frequency responce to brain.
This facilitates extremely the design of input and output transformer.
Very high step up rations may be used. If state of the Art Nanocrystalinne cores such as Magnetec Nanoperm are used which have x100000 relarive to air magnetic permeability, then very few turns are needed.
So they can be wound anywhere by the user to fit the exact application.
The concept of thinking from singer actor vocal chords to listener brain is restated and emphasized. This is knowledge shared by the pioneer sound engineers of Hollywood sharing their 1930's findings on the Journal of SMPE. This concept may explain why some vintage microphones, recordings and preamplifiers shine. Sometimes what we do not do is important.
To return to the topic, lower bandwidth means the gain can be increased in audio transformers.
If we use in the chain as preamplifier front end such as the Pleiades series V one tube preamplifiers, noise further reduces.
The Pleiades preamplifier is open source. Schematics are freely available. And it can also be freely manufactured as all other Pleiades electronic concepts.
It is battery powered, only a few volts are needed. This is made possible by using a high resistor from anode to grid to discharge the grid from the induced negative charge by the nearby cathode boiling electrons. The secondary of the input transfor must be capacitance coupled to the control grid so,as not to influence the grid potential.
The cathode of the electron tube is deliberately underheated to further reduce the cathode steamlike electron emission traffic congestion. Low heater voltage apparently for similar if not more reasons is used on the legendary Neumsnn U47 microphone.
By doing all the above we are left with a dramatic increase in signal to noise ratio. This means even a deliberate faint voice can be amplified in a clean way and without thermal noise hiss. The low and high frequency roll offs that have to be introduced are added by the outside passive components. Electron tubes and other devices are inherently flat from DC to MHz. The surrounding components may be used to advantage to reproduce flat frequency responce to the listener brain.
The turnover frequency depends on application. For example which microphone is used and what distance. If the singing reproduced intensity is different from the actual intensity of the acoustic sound emitted by the singer in the studio.
The point is that of looking at the chain as a whole system (from vocal chords to brain) and taking this into account in the design of the microphone preamplifier.
An attempt is the Pleiades amplifier. By using only one EF183 (triode connected), 2 transformers, an anode to grid resistor, one capacitor and a battery. With it we can possibly obtain the best sounding and quietest from noise signal in the world.
It can be perfected.
The very few components needed between vocal chords and brain make the most natural sound.
The Pleuades pre preamplifier is small, has low power consumption and is portable.
The Pleiade K117 version implemented with a K117 or 2SK117 JFET fits inside the female microphone connector and the other side of the audio cable gets phantom power when connected for example to iPad GarageBand or any mobile device that can power electrets.
The fact that dynamic microphones have their broad band reasonance in the mid band (resistance control of reasonance), greatly synergies with the above concepts.
Ribbon (reasonance at low frequency) as they are mass controlled and condenser and crystal microphones which have the disphragm reasonance at high frequency as they are stiffness controlled are excellent microphones too and can sound very natural and low noise too using the above concept.
Reference: Sound picture recording and reproducing characteristics - Loye, Morgan, Journal of motion pictures sound engineers
If this is not done the frequency responce from singer vocal chords to listeners brain is not flat. The frequency responce from mic to loudspeaker is only part of the chain [see Hollywood reference].
If a typical studio directional microphone is used at less than a few inches distance the proximity bass increase may begin at as high as 1KHz. It then roughly increases by 6 dB per octave as frequency goes down looking like half of a bathtub curve. Such curves can be readily observed on any directional microphone data sheet in reference to small distance from source
To compensate for this the preamplifier needs to do the inverse. This means the frequency responce of the microphone preamplifier may need to start dropping at 1KHz (-3dB point) with a slope of 6dB over octave below 1KHz. It depends on distance. 3in source to mic may need 470Hz turnover frequency etc.
And this is only part of the story. Additionaly there are the Fletcher Munson equal loudness curves and the voice effort curves.
The first describe how our mind hears, how its frequency response curve changes by intensity.
The second describe how the frequency spectrum of voice changes by how loud we sing.
So after compensating for the proximity effect the bass and treble frequencies may still sound exaggerated to our brain. As an example we may need to gently high cut starting at 10KHz.
This needed process has similarity with the playback de emphasis used to compensate emphasis while disc recording. And it has beneficial results. By reducing treble to get the the flat perception described above, the hiss noise goes down too. And any higher order harmonic distortion of an electron tube, JFET, or bipolar junction transistor decreases too.
One way to implement the high cut is the effective gentle first order low cut filter. The old trick of a C series with R filter across the plate load, the resistor being variable. It can be seen as treble control in schematics of EMI, Emmerson, Dancette tube record changers etc.
This process has another beneficial effect. Reducing high order harmonics means rounding of the extreme clipped peaks that happen all the time when using music signals. Rounding makes them more acceptable and the higher average to peak signal ratio means higher loudness. Similar process to FM radio preemphasis de emphasis.
The above may produce a dramatic increase to signal to noise ratio and loudness.
So the best preamplifier for voice applications that need 3dB points at 1KHz and 10 KHz means bandwidth restricted by 6-7 octaves (from 20 Hz to 1 KHz is 6 octaves and 1 octave form 10KHz to 20KHz) to reproduce correct voice, ie flat frequency responce to brain.
This facilitates extremely the design of input and output transformer.
Very high step up rations may be used. If state of the Art Nanocrystalinne cores such as Magnetec Nanoperm are used which have x100000 relarive to air magnetic permeability, then very few turns are needed.
So they can be wound anywhere by the user to fit the exact application.
The concept of thinking from singer actor vocal chords to listener brain is restated and emphasized. This is knowledge shared by the pioneer sound engineers of Hollywood sharing their 1930's findings on the Journal of SMPE. This concept may explain why some vintage microphones, recordings and preamplifiers shine. Sometimes what we do not do is important.
To return to the topic, lower bandwidth means the gain can be increased in audio transformers.
If we use in the chain as preamplifier front end such as the Pleiades series V one tube preamplifiers, noise further reduces.
The Pleiades preamplifier is open source. Schematics are freely available. And it can also be freely manufactured as all other Pleiades electronic concepts.
It is battery powered, only a few volts are needed. This is made possible by using a high resistor from anode to grid to discharge the grid from the induced negative charge by the nearby cathode boiling electrons. The secondary of the input transfor must be capacitance coupled to the control grid so,as not to influence the grid potential.
The cathode of the electron tube is deliberately underheated to further reduce the cathode steamlike electron emission traffic congestion. Low heater voltage apparently for similar if not more reasons is used on the legendary Neumsnn U47 microphone.
By doing all the above we are left with a dramatic increase in signal to noise ratio. This means even a deliberate faint voice can be amplified in a clean way and without thermal noise hiss. The low and high frequency roll offs that have to be introduced are added by the outside passive components. Electron tubes and other devices are inherently flat from DC to MHz. The surrounding components may be used to advantage to reproduce flat frequency responce to the listener brain.
The turnover frequency depends on application. For example which microphone is used and what distance. If the singing reproduced intensity is different from the actual intensity of the acoustic sound emitted by the singer in the studio.
The point is that of looking at the chain as a whole system (from vocal chords to brain) and taking this into account in the design of the microphone preamplifier.
An attempt is the Pleiades amplifier. By using only one EF183 (triode connected), 2 transformers, an anode to grid resistor, one capacitor and a battery. With it we can possibly obtain the best sounding and quietest from noise signal in the world.
It can be perfected.
The very few components needed between vocal chords and brain make the most natural sound.
The Pleuades pre preamplifier is small, has low power consumption and is portable.
The Pleiade K117 version implemented with a K117 or 2SK117 JFET fits inside the female microphone connector and the other side of the audio cable gets phantom power when connected for example to iPad GarageBand or any mobile device that can power electrets.
The fact that dynamic microphones have their broad band reasonance in the mid band (resistance control of reasonance), greatly synergies with the above concepts.
Ribbon (reasonance at low frequency) as they are mass controlled and condenser and crystal microphones which have the disphragm reasonance at high frequency as they are stiffness controlled are excellent microphones too and can sound very natural and low noise too using the above concept.
Reference: Sound picture recording and reproducing characteristics - Loye, Morgan, Journal of motion pictures sound engineers
No comments:
Post a Comment