• Welcome to the all-new HUG. All content has been converted from the old system, and over the next days we will re-style HUG in a more familiar way.

Psychoacoustics - what is it and why does it matter?

Simon

New member
Risk/benefit and approval

Risk/benefit and approval

How overbearing does it have to be to prevent unplanned outcomes? fluoxetine, hpv vaccine, thalidomide?
Perhaps this is going a bit off topic, but I’m sure Alan will tell us if he thinks it is!

The judgement is essentially one of risk versus benefit; the problem is that this is an extremely complex field and the information that one needs to gather in order to come to a judgement of risk versus benefit is often not easily obtained. The “some quarters” I was referring to above are clinicians, not drugs companies! They are concerned that, particularly for certain very serious conditions, the processes that are mandated for the treatment approval process are so “heavy” that the development of effective new treatments is held back.

If one is developing a drug to treat morning sickness (thalidomide was an example of one such) then since it may be used by a large number of vulnerable (but not seriously ill) people, one must be very careful to monitor for adverse effects. If one is developing a drug to treat acute myeloid leukaemia, then one is dealing with very seriously ill patients who have a high probability of rapid death. The scales of the risk/benefit equation are different for these different cases. The argument that some make is that the treatment approval process has been so conditioned by considerations relevant to the former type of situation that patients in the latter type of situation suffer.

BTW, I’m an academic statistician with no vested interest in drug development.
 
B

Bromo33333

Guest
Are subtleties always processed?

Are subtleties always processed?

Again, A-B testing is designed to highlight the sonic difference between A and B at the instant of switch-over and this it does extremely well. Are you saying that if I presented you with a slide show of two pictures of family members that you were familiar with but one picture had a colour cast that you wouldn't be able to recognise it on switch-over? Of course you would: the instantaneous comparison would jolt your brain into appreciating the differences. It's exactly the same with A-B audio testing - shown on a video clip here just recently.

Conversely, if the slide show image A transitioned (faded) progressively into the B image over a period of minutes as you watched, do you think you would suddenly say at any one point 'wow, that family member seems to have green skin'? I don't think so. Your visual/audio memory is not designed to make long-range comparisons; there is no evolutionary benefit for that behaviour. Our senses are designed to produce snap conclusions, side by side comparisons, better/worse, yes/no, pass/fail, go/no-go, danger/safe, attractive/unattractive, loud/soft, bitter/sweet, hot/cold ....
My point isn't to refute A/B testing - just point out that there is enough research going on senses/brain/detection that we are likely going to be justified, and sometimes surprised once conclusions are drawn.

I have run informal A/B testing to choose what I felt was the most accurate sound for a familiar piece of music (such a informal test put Harbeth on my short list, BTW). I always find it interesting to hear that our empirical methods may not always tell the whole story (or in this case, they might) - it was relevatory to contemplate that possibly details aren't always noticed with something familiar because you are working on playback of your memory and not processing all the information of your senses. I suppose *if correct* it would be a question of threshold in which one would notice a difference - to speculate - perhaps the amount of change, and the type of change. Green skin on a relative might be an easy trigger, but shirt color may not. But the work of my friend is mostly visual which is connected to the brain entirely differently than hearing.

To bring it back to the evolutionary argument - is that subtleties aren't always processed simply to keep the brain lightly loaded enough to deal with threats. To wit - perhaps the exact shade of color of your relative's shirt and the type of necklace worn, earrings, and the type of ribbon on the hat may not be noticed by your brain while under a stressful event - because it has no value - the fact that the relative is in danger, say, and the details surrounding that may be processed. Perhaps, and I am speculating here, the brain is as adept in throwing information out as picking what to process. I do not know how this would figure in an A-B test, but if the speculation were proven true, it might mean that in an A-B test where you are trying to bring details out, they have to cross a certain threshold before you might notice them, and things like state of mind might influence how much and what you may process.

I know this sort of test shouldn't be very controversial - and the controversy is very much overblown. My intent is not to fan flames, or take a position - just point out that it is an exciting time in audio - and we may start understanding very precisely the nature and processing of sound by the brain (and how the brain works, too)!
 
B

Bromo33333

Guest
A-B empiricism

A-B empiricism

I actually prefer the A-B test - I believe that it is the best empirical solution and tool we have to date - and probably the best way to evaluate two speakers, say, than anything else.

What I am (unsuccessfully, I fear) trying to say is that as we learn how the brain works better and how the senses tie in, and how we perceive as a result - we may eventually have some scientific light shone on the whole process that we didn't have before. It could be controversial for the mere fact that perception may be influenced by matters well outside the controls of the test. Or not. We frankly, don't have a "theory of mind" let alone a "theory of perception" that links all the facts together sufficiently well to say for sure. Until then we have the A-B test. After that, we may have a few more things to norm it against or a few alterations to it in general. But not now.

I will leave it that those vehemently opposed to the A-B it may be scoundrels, rubes or idiots - but I will leave the door open to the slight possibility that they aren't and we don't yet understand what's going on enough to declare so with certainty.

But as an engineer, I feel we do have a good empirical test in the A-B, but I look forward to the day when we have something better, if such a thing is possible.
 

A.S.

Administrator
Staff member
A-B cannot rank absolutes ...

A-B cannot rank absolutes ...

I actually prefer the A-B test - I believe that it is the best empirical solution and tool we have to date - and probably the best way to evaluate two speakers, say, than anything else. ....
It's vitally important to keep very much in mind that the A-B test is not designed to rank absolutes, as in 'A is absolutely the best speaker in the world...'. That would be a ludicrous misrepresentation of the modest capabilities of a properly constructed A-B test. All we are concerned about is just appreciating the sonic differences (or not) between (speaker) A and B presented in front of us at this moment, on this day, in this room, with these electronics, with this brain connected to these ears, knowing what we know, in an emotional and health state common to both A and B. That's all. If we were then to take this a stage further and through a series of comparisons compare A v B v C v ...... Z we could maybe develop a ranking table of some sort. But that's an academic exercise and of no use to me, or you.

The practical use of A-B testing is to replicate the sort of situation the designer finds himself in, or the consumer at a dealers where he cannot decide on A or B. There are speakers on the market which, to my ears, are lamentably colored and it really is a surprise that they receive top-ranking in the media. They really shouldn't be on the market*. Some of the best advertised are simply shocking, human voice sounding utterly unnatural. They add nothing to the accumulated knowledge and craft of two generations of speaker engineering and probably perform less well than speakers of equivalent size/price point made in the 1960s. The reason that these contemporary designs are marketable is because of soppy, casual appraisals amongst those we trust to properly critique audio on our collective behalf. Were A-B comparisons made using instantaneous changeovers then should there be any redeeming features of a design (there must be something marketable) we could be rightly directed to that plus point and the weaknesses also highlighted against a reference.

I do not design perfect speakers. I don't have the technology nor the skills nor the tools. But it's usually a shock to experience what the consumer (apparently) thinks is natural sound when to my ears, it is wildly short of that performance even in the middle frequencies on speech where hifi loudspeakers are idling and not working hard at all. And the latest trend of hugely cranked-up HF (typically 50-100% more output from the tweeter than would be considered natural a generation ago) doesn't seem to trouble any critic judging from what I read in the print media. An A-B comparison against a correctly balanced speaker would reveal that boost in the first minute of change-over even to a critic with abnormal or aged-related hearing acuity.

Hence, decade by decade hifi speakers become irritatingly brighter in the high frequencies. It's just not natural. You have to ask yourself what sort of reference sound a modern critic considers 'right'. Clearly this one issue has a massive influence on whether they like or not the product they are required to evaluate.

* Surely the most basic of all requirements of a high fidelity loudspeaker is that it reproduces the human voice without obvious coloration or change in tonal character. A loudspeaker that is incapable of doing so cannot be considered a high fidelity loudspeaker.
 

Pharos

Member
The mother of reference sounds

The mother of reference sounds

Alan, you cite the inadequacies of reproduction of some speakers on human voice; "Some of the best advertised are simply shocking, human voice sounding utterly unnatural."

I could not agree more, and we had a conversation about this aspect some months ago. I think that this one task is probably the hardest to achieve, and your post reminded me of something that I would like your help with.

I do not have the necessary equipment or expertise to try to produce a reference recording of speech, and with regret the 'Mother' of all sources with which we are both familiar, in my opinion is paying scant regard to producing such.

Do you know of any reference recordings of speech which have been thoroughly vetted for elimination of proximity effects and eigentones? For example, well recorded voices in the open air, voices in a semi live room, perhaps of conversation, and providing all the necessary acoustic stereo clues?

I would love to make such and have a design in mind, but surely extremely high quality mics. would be needed and much measurement to check the results.
 

A.S.

Administrator
Staff member
Speech - and using it as a test

Speech - and using it as a test

...I do not have the necessary equipment or expertise to try to produce a reference recording of speech, and with regret the 'Mother' of all sources with which we are both familiar, in my opinion is paying scant regard to producing such.

Do you know of any reference recordings of speech which have been thoroughly vetted for elimination of proximity effects and eigentones? For example, well recorded voices in the open air, voices in a semi live room, perhaps of conversation, and providing all the necessary acoustic stereo clues?

I would love to make such and have a design in mind, but surely extremely high quality mics. would be needed and much measurement to check the results.
You make some interesting points here. My view would be as follows:

1. I absolutely agree that carefully recorded human speech - as you say, the Mother of all reference sounds - replayed over loudspeaker systems can rapidly highlight the inadequacies of the loudspeaker. That's because whilst few ordinary listeners know how musical instruments really sound 'in the raw' everyone knows how fellow humans sound. And since most of the problems that bedevil loudspeakers fall in the speech band, playing speech is like taking a zoom lens to the sonic picture being painted by the speaker. It's blindingly obvious to even the untrained listener that there are characteristics in most loudspeakers which instantly alert the listener to the fact that he is hearing a reproduced voice rather than the real thing, so speech is a really excellent analytical tool. And cheap and quick to use (no royalties etc.).

2. I'd caution you about making a speech recording indoors, in stereo. That's sure to add layer upon layer of confounding variables. All you need is a mono recording, with as little of the environment around the voice as possible. An anechoic chamber would be a good solution; outside in the garden another (but there would be an inevitable ground reflection to consider). There is already a speech recording here on HUG made in the (now demolished) BBC chamber. I also have other library sources.

3. You don't need an expensive microphone per se, you need a microphone with a flat frequency response. That rules out all the standard microphones you'd see in the normal recording studio: they just are not flat enough or sufficiently coloration free. That could mean as simple as a Panasonic omindirectional electret capsule (cost about $4) which with a tiny bit of frequency response correction at the extremes of the response is amazingly flat, and used in the DIY type sound checking microphones. It's a little hissy which actually is a good thing because that hiss will give you valuable information about the tweeter performance on playback.

Here is a DIY microphone that should be capable of making a good quality speech recording; total build costs about $15.
 

Pharos

Member
Tish-boom and bandwidth

Tish-boom and bandwidth

I think that my referance to 'Mother' was misunderstood, though I agree that human perception of speech is far more the norm than musical instruments, and has been a part of our evolution on which our survival has crucially depended. (Perception of subtle intonations and nuances).

I was actually referring to what was a much loved organisation which could well be described as setting 'The Reference' standard for broadcast speech quality forty years ago. It's now very much more diverse output must have affected its ability to concentrate on sound quality.

I think the increase in top which many speakers are exhibiting is very much a correlate of what is known in Hi-Fi circles as 'The Hi-Fi Sound' perhaps epitomised by a 'tish-boom' charcteristic. ('If there is a prominence of top then it must be going high, and if a prominence of bass then it must be going low - so it must have a wide bandwidth.')

I have just watched a film about the blowing up of a bridge at Remagen, mainly out of interest in the sound quality, and it was dated '69. The speech was entirely natural, sounding as in the open air, and this prompted me to consider that the film may have had LF curtailment. But other sounds indicated a very full bandwidth and deep bass, though not as excessive as is often the case with modern films. So it can be done.

I have used the electret condenser mics. which you kindly illustrated and recommended for speech recording, but I had wondered about their quality, and whether or not it was good enough. (Surely the tiny aluminium diaphragm will give a metallic colouration?) Could I use my wall mounted PZM?

I used a Xsed pair of the former capsules in my WMD6C cassette recorder 20 years ago and thought that the sound metallic, and was now thinking along the lines of Calrecs being necessary. Although valid, it also seems to me that the ground reflection is very much less significant than many other colorations, those from other reflections and proximity effects.

What a waste to demolish the BBC's acoustic chamber, they are I presume very costly to create.

Your caution about an indoors recording of speech in stereo counters what I had thought were some of the main aims of stereo.
I had thought that one would try to capture the early reflections in a semi live room of for example two people talking, and with that carefully done recording, attempt to replay it in another room in the hope that the captured acoustic would give a 'false acoustic' in the listening room. I appreciate that the replayed sound would be subject to processing by the acoustic of the listening room, but although a compromise, with the optimum amount of damping in the listening room perhaps it can be done.

One reads reviews about how certain systems extend depth to behind the rear walls, and also width beyond that of the listening room, and that is what I had hoped to achieve maximally.
 

A.S.

Administrator
Staff member
The 'reality-subset'

The 'reality-subset'

....Your caution about an indoors recording of speech in stereo counters what I had thought were some of the main aims of stereo.

I had thought that one would try to capture the early reflections in a semi live room of for example two people talking, and with that carefully done recording, attempt to replay it in another room in the hope that the captured acoustic would give a 'false acoustic' in the listening room. I appreciate that the replayed sound would be subject to processing by the acoustic of the listening room, but although a compromise, with the optimum amount of damping in the listening room perhaps it can be done.....
I caution you about overloading yourself with compounding variables: to make serious analytical progress in audio you really have to reduce complexity. Trying to record a stereo acoustic to replay over stereo speakers in another (stereo) acoustic is way too complex a decoding task. Try it by all means but I think you will have the greatest difficulty teasing apart what you hear and drawing any valid conclusions.

Stereophonic sound is a grand illusion. Any sense of depth, height and width is entirely phantom, in the brain of the hearer. It is a sonic conjuring trick that we enjoy. The sonic image in our brain is no more real than the image of ghosts or fairies at the bottom of the garden. It's wholly personal and whilst seemingly real to you, may be totally invisible to others.

The quote from one of the last BBC audio designers, David Mathers in his report "On the design of loudspeakers for broadcast monitoring" (BBC RD 1988/14) I've extracted really says it all.

The Panasonic microphone capsule is widely used for measuring loudspeakers, and designing LS systems. I do not believe that it could be applied to that if it was significantly colored. I'm certain that it would be good enough for your recordings.

As for sound quality in broadcasting these days, the rot started when producers disliked the look of big, quality microphones in-shot after the clip-on lapel mini-mic appeared. I understand that the poor-old embattled BBC has so many management issues these days that sound quality is not and has not been a priority for many years. Neither is training. Sky seem to have captured the moral and technical high ground there. However, behind the scenes we have a major project underway to put sound quality back onto the agenda and to positively influence the next generation of sound professionals to craft sound with more care. More on that over the next few months.

>
 
Attachments only viewable to members

Pluto

New member
Small is best in microphones

Small is best in microphones

I have used the electret condenser mics. which you kindly illustrated and recommended for speech recording, but I had wondered about their quality, and whether or not it was good enough. (Surely the tiny aluminium diaphragm will give a metallic colouration?)
Mics have a great advantage over loudspeakers despite the fact that their roles are more or less exact opposites within the transducer world.

A modern mic. capsule has near-zero mass. In general, audio authorities will tell you that capacitor mics. are the best sub-type amongst the core microphone technologies, which are moving coil, ribbon and capacitor (also known as condenser or electrostatic) mics. The fundamental reason for this is that the moving diaphragm has to be no larger (or heavier) than is required to produce a measurable change in capacitance in the presence of audio. This in turn means that any colourations due to the mechanical properties of the capsule are likely to be at far higher frequencies than those in which we are interested.

The pick up pattern of a minute capsule suspended in free space with one face exposed to air movement (i.e. sound) is omnidirectional. Such a capsule will not become directional until the wavelength of the incoming sound starts approaching the size of the capsule and its enclosure and this means many tens of kHz. Fundamentally, if you desire a truly omnidirectional microphone, the smaller the better.

So, low mass and small size are your friends if you desire a low colouration omnidirectional microphone. It is ironic that the measures we have to employ to make a microphone discriminate in favour of sound arriving from one direction or another are probably the biggest single cause of microphone "characteristics". Certainly the very flattest response microphones tend to be omnidirectional and small.
 

Pharos

Member
Deception, adaptation and the merry-go-round

Deception, adaptation and the merry-go-round

I was of course aware that the sound reproduction process is based on creating an illusion, and remember well going one evening to a demo of quadraphonic sound in the early seventies, one other member of the auditioning team being a concert musician. She said; "No Hi-Fi sounds like the real thing", and she is right. There is always to an extent a suspension of disbelief, and one phenomenon I experience underlines that.

If whilst listening to FM, one changes stations, one will, depending on the nature of the differences of programme, experience a massive quasi 'shock' at each change. It takes several seconds to habituate to the new sound and identify the instruments in the music on each station, and I believe this to be because of internal 'software' adaptation and acclimatisation.

I think this process of adaptation is what Alan is trying to avoid by using only 30 sec bursts in DBT. If we were anywhere near producing a good deception in sound reproduction we would only experience the change as one of source stimulus, as we do for example when exiting a tube station and entering the street.

In my years of trying to improve sound equipment I have made some relaitvely slight improvements on numerous occasions, and for a while they have resulted in a sense of joy and delight in which I want to listen again to all of my music. Gradually this change becomes accepted as the new norm, and the quality of reproduction becomes more familiar.

In my opinion this reflects that the change is heard as a better deception, and then gradually with time the brain runs its 'software' and 'learns' the new sound, finally resolving it into another deception, which although better, is flawed.

This process involves a great deal of time, work and expense, and produces only temporary illusions of 'getting to the truth', and is I believe, often what keeps audiophiles on the 'merry-go-round' of repeatedly changing equipment.

It is also interesting that the cat notices, and is even fooled and very upset for a time by the differences resulting from a change, but habituates after a while, no longer taking any notice of it.

I had assumed that the cheaper end of the mic. market were poor in performance, and that for speaker measurement purposes really expensive ones were necessary. (I have a pair of AKG 451EBs.)
 
Last edited by a moderator:

A.S.

Administrator
Staff member
The vital measuring microphone ....

The vital measuring microphone ....

....I had assumed that the cheaper end of the mic. market were poor in performance, and that for speaker measurement purposes really expensive ones were necessary. (I have a pair of AKG 451EBs.)
The 451 would be a very poor choice as a reference microphone. As you can see from the article I have already linked to a characteristic of AKG condenser microphones is a rising top. That is a no-no in a reference microphone where you need, above all other considerations including that of low noise, a flat on and off axis response. Otherwise you are unable to be sure whether any lumps, bumps and general trends in the measured response are the speaker, the microphone on axis or the undue emphasis of the microphone to reflections from the environment around the loudspeaker.

In the attached data sheet, which is generic, nice and smooth, all ripples painted out and no manufacturing tolerance indicated and may or may not represent your actual 451, I have marked the two features that would debar it use as a reference: the dramatically rising top above 5kHz or so, and the ever-increasing sensitivity to off axis sounds in the upper audio region. Conversely, a $2 Panasonic capsule is much flatter (I know, I have made comparisons against a truly flat B&K reference) but as already stated, the Panasonic would be much hissier. It really couldn't be used to make a hi-fi low-noise recording - it's inherent noise would be irritating to listen to.

If you do attempt to make stereo recordings indoors with your 451s, as I said some posts ago, you will be overwhelmed with difficulties determining what is what: the room, the speakers, the mic, the mic positioning etc. etc..

If you absolutely must have total confidence in the frequency response you measure, then you have to use a B&K 4191 or similar precision omnidirectional mic, which truly has an almost straight line response. But omnis indoors would give you negligible stereo. If you thumb through back copies of recent hi-fi magazines and look at the published frequency response curves there are many questions that we, the public, need answers to regarding speaker frequency curves made by the reviewers ...:

  • What type of measuring microphone was used by the reviewer?
  • Where exactly was it placed relative to the speaker?
  • Does it have a Calibration Certificate, and was that issued by an international authority or is it a DIY job?
  • Is the mic acceptably flat as a reference tool?
In addition we need to know ...

If the speaker response shows significant deviation from flat at the extremes of the audio band ....

  1. Is that truly a characteristic of the speaker, or is it the mic, the room, the amplifier, the cables or is it deliberate, the designer's intention?
  2. Does a boost compensate for the designer's hearing limitations at EHF or for his interpretation of his aged customer's hearing acuity?
  3. Does the designer wear a hat, glasses or is the sound wave into his ear impeded in any way that could disturb his hearing acuity?

Many questions. The variable we have most chance of eliminating is that of a really good, flat microphone. We can take a view on the other factors.

>
 
Attachments only viewable to members

Pluto

New member
Caution! The danger of using a "studio grade" directional microphone

Caution! The danger of using a "studio grade" directional microphone

I had assumed that the cheaper end of the mic. market were poor in performance, and that for speaker measurement purposes really expensive ones were necessary. (I have a pair of AKG 451EBs.)
Nature, unlike commercial interests, never gives you something for nothing.

The ultimate problem with tiny electret microphones (which bears no relationship to cost or even frequency response) is that the smaller the moving component, the lower the output of the capsule which requires the microphone's built-in head amplifier to run at higher gain, thus worsening the noise performance. So, in the most general terms, small microphones will tend to exhibit poorer noise performance than bigger ones.

You imply above that you use AKG 451s for speaker measurement. I hope this isn't really so! The cardioid 451 is notorious for its edginess from about 2kHz upwards. These mics., unlike the omnidirectional case where sound is only allowed to impinge upon one face, obtain their polar characteristics by allowing a controlled amount of the sound to impinge upon the rear face of the capsule as well as the front, (look for holes cut into the mic body just behind the front grille) causing partial acoustic cancellation. The idea behind this somewhat messy process is that almost no sound from directly in front of the mic is allowed to reach the rear of the capsule[SUP]1[/SUP] while sound originating from behind the microphone will cause nearly complete cancellation[SUP]2[/SUP]. This is why these mics have secondary apertures which allow some sound to impinge upon the rear of the capsule.

This acoustic cancellation process is a comb-filtering nightmare and it's surprising that directional mics using this method can sound as good as they do, but do not use such a mic. for any kind of reference or measurement purpose!

[SUP]1[/SUP] Or at least, arrives at the rear face of the capsule out of phase, thus augmenting such signals
[SUP]2[/SUP] As it arrives at the rear face of the capsule in phase, thus cancelling such signals
 
Attachments only viewable to members

A.S.

Administrator
Staff member
What can you get away with?

What can you get away with?

Here is the comparison I made last summer between a $2 Panasonic capsule mic and two top-line B&K reference capsules costing about $1200 each. The 4190 is good to 20kHz and the 4091 to 40kHz. Note the vertical scale. The measurement was made by rigidly clamping either mic/capsule in a DIY vice so I could be as sure as possible that the capsules were at the same point in space. Only the tweeter was connected and driven in the speaker. Yes, there will be local rflections from the vice, but they will be commmon and somewhat smoothed by the 1/12th octave measurement analysis.

What do you think? Would the $2 mic be good enough or is there still that slight worry that a cheap as chips capsule may mislead you, if not today in a year's time?!

>
 
Attachments only viewable to members

Pharos

Member
Variations between microphones

Variations between microphones

That from my perspective is really very good - especially if that close agreement is followed down to the bottom end.

In my position I can I think regard the Panasonic as a reference; surely most speakers, even expensive ones deviate way more than that from flat.

Is that quality of performance obtained at the other end of the spectrum?

Addendum.
From the three graphs it seems that there is a great agreement in the curve deviations from flat, and, given the differing heritages of the Panasonic and B&K mics, these very probably are due to objectively present factors in the measuring environment - reflections interference etc., and on which they agree, which therefore particularly validates the Panasonic mic..
 
B

Bromo33333

Guest
Quick and easy v. precision measurement

Quick and easy v. precision measurement

Test equipment has to be both precise and reliable. I have used Hewlett-Packard test gear (now Agilent) for the precision, but also the reliability of the gear. You cannot predict when something drifts - and it is exceedingly hard to track down precisely when something would produce measurement error - and it is generally worth it to buy something that is reliable as well if the measurement is a critical one.

So for me, I might have the $2 capsule as a "quick and easy" measurement device in an engineering environment (not production) if there was a solid reference available that was used before a series of measurements were done, and only the calibrated reference was used for shippable things. But that's getting into ISO stuff.
 

Pharos

Member
Panasonic total frequency response

Panasonic total frequency response

My buying 451EBs was not for measurement purposes, but for the recording of music, for sampling sounds out in the open to later be processed in a sampler and 'instruments' made form them.

The listed typical uses for the 'B' largely validate my choice I think, (vocals and classical guitar), but I had even wondered recently about trying to get hold of a STC 4038 for pure vocal recording.

Is it possible for anyone to confirm the response of the Panasonic in the mid and bass regions please?
 

A.S.

Administrator
Staff member
Mic correction and 4038

Mic correction and 4038

....Is it possible for anyone to confirm the response of the Panasonic in the mid and bass regions please?
A) the correction (normalising) file supplied with a calibrated Panasonic capsule embedded in a semi-pro measurement microphone applies no correction data below about 1kHz or so B) from this we can assume that the deviation from flat compared with a B&K reference omni is very small up to about 1kHz.

If you are thinking about using the Panasonic for music/speech recording, I say again: forget it. Whilst they may well be flatter than your 451s, they will be much hissier. Pluto has explained why: a tiny capsule has a stiff diaphragm that hardly moves, hence a small voltage is produced from an impinging sound wave, and the ratio between self-noise from the mic's electronics and useful amplification of the sound wave is narrowed = more hiss.

We've covered the 4038 ribbon mic before, and used it in speech recordings (here somewhere?) and it also has low sensitivity, is a little 'chesty' and needs some HF shelf lift to sound correct. Nice, low coloration sound but the fig-8 polar pattern is an issue when used indoors.

The Behringer 8000 may use the Panasonic capsule. It can be purchased calibrated from here. The calibration service shows typical before-cal frequency response as this.

Personally, I would caution over-reliance on a $2 capsule regardless of how much correction-EQ is applied to flatten its response. There will always be doubts. How about temperature stability? Or drift (already mentioned)? If you absolutely must be able to separate the performance of the mic from that of the loudspeaker you really have to bite the bullet and spend money on a B&K omni. Or perhaps the more affordable DPA omni. What appears to be this mic (or a variant of it) is sold as a reference mic by one audio test manufacturer. We have one and it's well made, needs 24V phantom powering. I've never used it for recording though.

I don't know what more useful info I can give you.
 

Pharos

Member
Mic use ....

Mic use ....

I have two separate interests; sound reproduction and music creation, and the mic, usages are very relevantly separated. Thank you for all the info..
 

Pluto

New member
Tinkering or going pro?

Tinkering or going pro?

I had even wondered recently about trying to get hold of a STC 4038 for pure vocal recording
A pair of 4038s in decent condition will not be cheap. About £1500 a pair, new. Yes, they are still made in small quantities! Nature, as I said earlier, never gives you something for nothing and in this particular case the quid pro quo is that, in most instances, the figure-8 polar pattern is something of a liability. Don't get me wrong here - there are venues where a coincident pair of figure-8 ribbons in front of a soloist or small ensemble will produce a divine result, but this is unlikely to be the technique of choice most of the time. The other issue with 4038s is that they have incredibly low output by modern standards and a great many modern mic. amps may not have enough gain available under all conditions. One more thing - ever thought about the mechanics of assembling a coincident pair of two microphones that each contain such powerful magnets that you cannot stop them embracing each other?

If you are seriously interested in small-scale recording, you will be far better off with a pair of mics. that offer a switchable polar pattern. But the real play, in my view, is to buy one of these. Cheap, they ain't. But in terms of the results you can obtain because of the flexibility they offer an investment in this direction is, arguably, far more useful than tinkering with odd 'bits and pieces' of microphones here, there and everywhere.
 
Top