The Science Behind Broadcasting

Probably the most important part of any radio broadcast or recording, is the ability to capture a voice.  So we know that when we speak, vibrations in the voice box create a pressure wave in the air, this travels through to cause vibrations in your ear, and they are then converted into electrical signals, and fed to the brain where they're decoded.

But, all the kit that we use for radio is electric, so we do need a device that converts that pressure wave into an electric current and this is what a microphone does.  To find out how microphones do that, we asked our own Dave Ansell if he could build one from scratch, and to do that, he looked back into the history books...

Dave -   The first kind of microphone which was practically used was developed by Edison in the 1880s to work with the telephone, and that was based on having a little pot full of carbon granules.  You apply a voltage across those carbon granules, and you get a current out.  Now if you shake that using the sound wave which is coming from your voice, that will change resistance of those carbon granules, so that will change the current going through it and you have an electrical signal.  But the problem is, the resulting electrical signal is horrible and it gives a very, very bad rendition of the sound it was producing.  But it was very cheap and you didn't need to amplify it, so it's very useful in the early days.

Ben -   How did people take those ideas and build on them?

Dave -   The next big improvement in microphones was done by a guy called Wente based in Bell labs in 1916 and I'm going to build a model of that now.  This is based on the principle of capacitance.  I've got a piece of tin foil and if you apply a voltage of that tin foil, some current will flow into it, a tiny, tiny amount, a little tiny bit, and it might become negatively charged.  Now, if you apply the opposite voltage to the second piece of tin foil then that will become positively charged and for a certain voltage, a certain amount of charge will flow from one plate to the other, so we say the system has a certain capacitance.

Ben -   But if you allow those two bits of tin foil to touch then presumably, the current will just flow from one to the other and instead of having a way of storing energy, a capacitor, what you've got is just a very poor circuit.

Dave -   That's right, so we don't want to do that.  So, I'm going to separate these two pieces of tin foil with a piece of cling film which is a really good insulator.

Ben -   So what we now have is a very basic model of the standard electrical component which is a capacitor, and that is capable of storing a small amount of charge, but that still doesn't look to me like it can collect sound from the environment and turn it into an electrical signal.

Dave -   The way it does this is a property of capacitors.  This is because if you got the two plates, a positive plate and negative plate very close to each other, they stabilize each other.  The positive plate attracts the electrons in the negative plate and so, the voltage you need to push electrons in gets lower, so you can get more charge on for the same voltage, and we say has a higher capacitance.  The closer those two plates are together, the higher the capacitance is.  So, if the plates move backwards and forwards, the capacitance is going to change up and down.

Ben -   So, in order to turn this into a microphone, we need to find a way of changing the distance between the two.  Is that as simple as shouting at it?  Essentially, we put a vibration in it from our voice that changes the distance between the two plates, and that gives us a model of that sound wave in the changing electrical current?

Dave -   That's what I intend to do, yes.  Shouting is going to be involved.  As the sound waves hit it, they're going to move the plates together and apart.  The capacitance is going to change, so the amount of charge on those plates is going to change, you get a current flowing in and out, which it should be able to detect as an electrical signal.

Ben -   So, we're wiring up our 2 pieces of tin foil and a piece of cling film now.  I would be amazed if it does work.  Dave, over to you, what's next?

Dave -   Well I've now got this connected to our recorder so we should be able to listen to it later.  I'll give it a go.  I might speak quite loudly because my experience is this is not the most sensitive microphone in the world, but let's give it a go.

Ben -   Okay, hold on.  I had better stand back if Dave is going to be shouting.  Dave's voice is loud at the best of times.  Off you go then, Dave.

Dave -   So I'm now shouting very loudly at this capacitor based microphone!

Ben -   Let's have a listen and see what it sounded like.

This week, scientists have uncovered residues of 7,000-year-old cheese.  Chemist Professor Richard Evershed from the University of Bristol is one of the authors of a paper in the journal Nature this week which describes how he and his team have made this rather milky discovery.

Chris -   So, how did this story begin?

Richard -   Well, it began truly about 30 years ago when one of our co-authors, Peter Bogucki, was putting together evidence from sites he'd been excavating with colleagues in Poland where they've been uncovering these perforated pot shapes, ceramic pot shapes and he propounded a theory based upon those pot shapes that they were basically cheese strainers and started to develop the idea that these early farmers, which is the Linearbandkeramik culture, were dairy farmers.

Chris -   When you say they were perforated pots, we're talking clay pots with fairly substantial holes drilled in them, rather like a top of a pepper pot?

Richard -   Very numerous holes, if you look at the images that are in the article.  They're all over the vessels. They look like a sort of small colander.

Chris -   Obviously one can say they look like the kind of thing that modern-day people use to strain cheese, but just because they look like it doesn't mean they definitely were being used to strain cheese.  Is there anything else they could've been used for?

Richard -   That's right and in fact, we've had an interest in identifying dairy products from the archaeological record for a number of years.  These vessels were one of the major sources of sort of typological evidence, but I must admit, I've been rather sceptical and I'd often said in my lectures, but surely, this could've been used for straining anything.

Chris -   So, how did you approach dealing with this and trying to find out what they really were used for?  How did you discover cheese?

Richard -   This is where the two studies collide.  Our research, going back about 12 years now has been focusing on looking at the origins of dairy by looking at the pottery vessels, usually cooking vessels.  We've developed a methodology, a chemical method which allows us to identify milk fat.  It's based upon the fact that animal fats, which are the most common type of organic residue you find in archaeological pottery when you study pottery from Europe, can be identified through their fatty acid signatures.  Based upon studies that we've done, studying many modern animals, we've developed a very robust proxy which allows us to separate ruminant, carcass fats from dairy fats, pigs, horses can all be separated.  So we can quite constantly identify different fats.  So if we go into archaeological pottery, recover the residual fats in there, we can identify these fats quite unambiguously as to whether they're dairy fats or carcass fats.

Chris -   These pieces of pottery have been in the ground in this site in Poland, in the centre of Poland for 7,000 plus years.

Richard -   It's remarkable, isn't it?  Yes.

Chris -   And you can get fats out of them, that were from the cheese or the dairy products these people were working on?

Richard -   Absolutely.  It never ceases to amaze me. It's due to the fact that these fats are preserved deep within the fabric of the pottery and I imagine that they go into sort of molecular size pores in the alumina silicate matrix where they're protected from say, bacteria getting in there to degrade them.  And in the right sort of soil types, they're protected from leeching from under the pottery by ground water.

Chris -   So, you've got specimens of the pots, you took small samples which enables you to isolate these fats.  What do those fats tell us then?  They say that there was dairy material in these pots?

Richard -   That's right and just to clarify, we're not only focusing on these cheese strainer vessels.  We've looked at a large number of other vessel types from the site and quite specifically, what it shows is that these perforated vessels, these putative cheese strainers are the ones that contain dairy fats.  And so, if you then start to put this together this as like a forensic exercise, we've got dairy fats associated with perforated vessels and you start to ask yourself the question:  What other milk processing activity could you reasonably have been doing which required you to strain some sort of milk product?  Well, it's cheese making.  There is no other milk processing activity that uses a straining activity.

Chris -   Why do you think these people wanted to make cheese?

Richard -   Well, there's two motivations.  One, you would be producing  - if you had a substantial dairy herd as the animal bone assemblages from these sites seem to suggest - you would have surplus of milk at certain times of year, in the summer time for instance.  So, the production of cheese would've been a way of producing a non-perishable product that you could then preserve throughout the year so you could just take full advantage of the nutritional properties of milk in the winter time.  But there is another critical factor and this is where Peter Bogucki comes back into the story with his paper from the 1980s.  It's related to this lactose intolerance phenomenon.  It's very likely, almost certain that these early farmers were lactose intolerant, so they would, potentially, have got quite poorly, had they been drinking substantial amounts of whole milk.  So, it looks like this straining activity was perhaps the recognition of this and they had found a way of getting around the lactose intolerance problem.  What you're doing when you separate the curds from the whey, the whey contains the lactose.  So you're actually producing a lactose reduced product when you produce cheese and thus, lactose intolerant farmers would've been able to exploit the nutritional qualities of milk.

For more on cheesemaking, check out the
Naked Scientists Scrapbook:

Porcupine quills penetrate skin better than a hypodermic needle because of tiny backwards-facing barbs at their tips, which also render the quills very difficult to pull back out.  This trick could now inspire better medical equipment such as needles and tissue adhesives.

North American Porcupines carry around 30,000 spines and are known to be an effective defence mechanism - they penetrate tissues easily and are often difficult to remove.  Unlike those found on similar animals such as the African porcupine or our own hedgehog, North American Porcupine quills have a series of microscopic barbs on their tips.  These barbs intrigued researchers, including Dr Jeffrey Karp from the Brigham and Women's Hospital in Boston, who then undertook a series of experiments to see what effect these barbs had.

Publishing their results in the journal PNAS, the researchers describe a number of experiments, measuring the force required to penetrate a sample of pig skin, and then to pull the quill back out.  They repeated their tests with quills where the barbs had been sanded smooth, and with other items of a similar diameter, including a hypodermic needle.

Unsurprisingly, having backwards-facing barbs means more effort is needed to remove a porcupine quill than a smooth quill - almost four times as much force was required as with a barbless quill.

The surprising result, and for the researchers the most interesting, was that the presence of barbs reduced the force needed to penetrate into the muscle.  While an 18 gauge hypodermic needle takes around 0.59 Newtons of force to pierce into pig flesh (bought from a local butcher), the barbed quill required only 0.33N to achieve the same task.  Samples taken from the skin and observed under a microscope also showed a much smoother puncture site, suggesting that the tissue had absorbed less energy from the quill, and therefore taken less damage.

Computer modelling suggests that the barbs concentrate force onto a smaller area, much like the jagged edge of a serrated knife.   This reduces the need for the tissue to deform around the entire circumference of the quill, resulting in easier penetration and less damage.

The authors argue that mimicking the barbed structure could allow creation of microneedles that penetrate easily but resist buckling, and enhance a range of other biomedical applications.

Ben -   If you're not currently listening online, then you could be using one of three different ways to receive your radio.  We've already heard about AM and FM, but now, there's also DAB - that's Digital Audio Broadcast.  The BBC first started broadcasting digital radio back in 1995 and now, most of their output can be found on a DAB radio.  But, if FM was working, then why go digital?  Still with us is John Adamson, BBC Production Editor for London and the Southeast, and as we said earlier, 'the Digital Doctor'.  So John, why did we want to go digital?

John -   I suppose it was a combination of quality and quantity.  So, it enhances listeners choice and offers better quality, particularly in marginal signal conditions.  

If we go back to AM just very briefly - medium wave and long wave, there's another radio station every 9 kilohertz apart.  So, the audio bandwidth is only 4.5 kilohertz, barely better than telephone quality, not very good at all.  Also, a finite number of frequencies that cannot be re-used very often because their transmissions spill over a very wide area, they can't be very well targeted.  In the case of long wave, there's only 15 channels for the whole of Europe, so you can imagine it's quite difficult to manage that frequency of the spectrum effectively.  Medium wave has rather more, but it's still four and a half kilohertz audio bandwidth to avoid interference to adjacent stations.  Aerials are rather large.  Not a lot of choice there really.  Also, medium wave has a problem of after dark interference; Strange noises, the background continental interference, whistles, whines, and buzzes in the background.  FM made things a lot better of course because we got stereo as we said earlier and more immunity to electrical interference.  Also, because it was on the VHF band, around about 3 metres wavelength, aerials were a sensible size, so you could target your audience quite effectively.  So, in the case of BBC Radio Cambridgeshire, they've got a transmitter at Madingley that's targeted just for the environment around there. 

Why do we go to DAB?  Well, FM still had a problem in marginal areas where there's something called multipath interference.  I don't know if you've noticed this as you're driving along or using a portable radio, but sometimes you get all sorts hisses and burbles, and squeaks and things.  That's multipath interference and that's a function of analogue radio - that if you get a wanted signal and a reflected signal which inherently must be delayed with respect to the original, it will interfere.  

If you think of analogue television, which was shut down just last year, you may remember seeing ghosting on the picture.  That is the visual equivalent of the interference you would get on audio.  

So, the reflected signal always causes a problem to the main.  DAB offered something rather new and that was, that because of the modulation system used and the coding systems used, you could have a number of transmitters, all working on the same frequency. 

So, for example, the BBC national digital multiplex operates on one frequency right around the country.  All the transmitters work together and although some of them will be delayed with respect to others, if there are in a certain window of timing some delay, they will add, and the listener won't get any interference. 

High quality audio is possible as well.  Thanks to digital audio encoding, mpeg compression, that allowed us to squeeze more audio into a given bit of spectrum.  If you were using linear digital audio, you would need around about 1.5 megabits or 1.5 megahertz worth of space for one stereo audio channel.  That's not very efficient.  But with mpeg, this allowed us to reduce that to say, 256 kilobits, and that was alleged to give around about CD quality.  So, in the early days, what we were thinking of with DAB was, we'd have more channels, but not many more, but high quality, and a more rugged, reliable, transmission system

Ben -   Lets come back to the audio compression in a minute, because I think we need to point out that that's different from the type of compression we were talking about before.  When we were talking about amplitude modulation and frequency modulation, that's essentially building a model of the sound of that pressure wave in the modulation.  But with digital radio, you're not doing that.  It's not a direct model of the sound.  It's differently encoded.

John -   Yes, it's all digitally sampled with a lot of ones and zeros effectively, and all gets assembled back in the decoder again.  So, if you were to listen to DAB radio spectrum on an analogue set, all you'd hear is a lot of random noise.

Ben -   And how does the signal change with distance?  I have a digital radio at home, and it seems to be, ironically, digital.  It's either perfect and sounds beautiful or the signal is dreadful.  Whereas with an FM radio, I can tune it in, and it sort of seems to fade out with distance, the quality fades out.  Is digital always as good as it's going to be, or not present?

John -   The theory is that you either get a perfect signal or you're getting no signal at all, but in reality, there's that grey area where the signal isn't quite strong enough for your set decoder to make sense of.  So, what you get is this characteristic sound which has been described as "bubbling mud".  The difficulty with that is that it's a really irritating sound.  With FM, you just get more and more noise.  The signal gets weaker or you might get these burbles that I mentioned to you earlier, and squeak sounds with multipath distortion.  But with the digital burbling mud sound, it's really quite difficult to follow the programme.  So, it's not perfect, that is true.  Because of the modulation and coding system that is used, it's generally better for mobile and portable reception.  So if you're driving along in your car, it should be more rugged than an FM equivalent.  And of course, you're on the same frequency, so you're not having to retune every few miles as you find a different transmitter.

John - It's all to do with the amount of compression and the type of compression that is put into the signal. As I said earlier, audio processors nowadays are very clever devices, multiband devices, so they chop up the audio spectrum into different chunks, and they process them separately.

You can put an amount of basic compression in, and you can add clipping. Clipping in moderation is fine because it adds intelligibility. In fact, old telephone systems used to have lots and lots of clipping on it, which made it nice and easy for people to understand what was being said, even on a really crackly horrible line.

Chris - This is where, if you got a wave, if you're literally just chopping the top off the wave, isn't it, to make it a bit quieter.

John - That's right. Exactly that. Now, that in moderation is fine, but it can be quite tiring after quite a short period of time. Normally, around about 5 to 10 minutes, some people start to notice the clipping effects in the audio processing that's used.

Unfortunately, a lot of broadcasters, when they put the audio compression in, do it for an instantaneous reaction to listeners who perhaps are just tuning down their dial, they find their radio station, and they think, "Wow! That's loud! That's very impressive" and are going to hang on in there for a bit.

That's all well and good, but very quickly, this fatiguing effect can take over, and it affects some people more than others. That's the strange thing, you might find one person that doesn't notice the fatiguing effects. But other people find it really, really distracting and disturbing, and want to tune away very quickly.