What makes a strobe or flash pop?

[richardsperry]

What makes the sound?

 

Why is there sound coming from a glass tube when it fires?

[lornesunley]

<p>electric arc ... makes noise ... just like the thunder from lightning</p>

[richardsperry]

But it's a vacuum.

 

Lightning occurs in air, thunder is the sound waves in air.

 

Sound does not travel in a vacuum.

[richard_cochran]

<p>It's not a vacuum. The tube is filled with gas at pressure that's lower than atmospheric pressure, but it's still got some pressure.</p>

<p>Anyway, the gas in the tube very rapidly heats up tremendously. That heating increases the pressure. A sudden change in pressure is pretty much the definition of a sound.</p>

[lornesunley]

<p>Handy Wikipedia reference to how a flash tube is constructed http://en.wikipedia.org/wiki/Flashtube</p>

[Palouse]

<p>Xenon gas--set off by the electrical charge from the capacitor.</p>

[jcoffin]

<p>The flash tube doesn't contain a vacuum -- if it were, you wouldn't get the flash. In an ordinary flash, the tube is filled with Xenon gas (or at least that's the primary element). You have a plate at each end charged to ~300-500 volts. Then you have an electrode wrapped around the outside of (part of) the flash tube. When you want the flash to fire, you feed ~10,000-15,000 volts onto that electrode. That ionizes the Xenon, making it conductive. The ~300-500 volt charge then flows through the ionized Xenon, causing the flash.</p>

<p>Although a flash tube converts the energy to light more efficiently than an incandescent bulb, there's still a fair amount of heat released. A decent sized battery operated flash is about 10 watt seconds, but that 10 watt seconds is released in roughly 1/1000th of a second, so during the discharge you have ~10,000 watts of power flowing. Especially given the relatively small size of the tube, even a small percentage of that being converted to heat causes the gas in the tube to heat up quite abruptly. That makes the gas expand quickly, which makes the glass of the tube vibrate, in turn vibrating the air around the tube, which we hear as a sound.</p>

[tom_mann1]

<p>The previous responses are correct, but there are also several other sources of the sound that comes from a flashgun that has just been triggered:</p>

<p>a) Whenever you discharge a capacitor, the attractive forces between the plates go from a large value (when the capacitor is fully charged) to a much smaller value (when the capacitor is discharged). They are held apart by materials which serve as a dielectric and mechanical spacer. When the forces suddenly drop, the plates try to spring back to their non-charged separation. Depending on the exact construction of the cap, this can make a significant sound.</p>

<p>b) Whenever you send a large current pulse through a wire, it generates a rapidly changing pulsed magnetic field. The B field from one part of the wire can interact with the field produced by the wires in another part of the circuit (...say, the ground return...). Just like two magnets, depending on their orientation, the two pieces of wire may either momentarily attract or repel each other. This leads to the wires moving slightly, but in a very short period of time, so very rapidly. For circuits of this size, this usually leads to a weak "click" sound.</p>

<p>c) The glass or quartz envelope of the flash tube gets a burst of UV energy deposited in it by radiation from the plasma. This leads to it quickly (ie, on the time scale of ms) expanding radially because of the coefficient of thermal expansion of the material of the envelope. Any rapidly moving surface in air (ie, the exterior of the flash tube) produces sound.</p>

<p>d) I have seen claims that the air around the flashtube is rapidly heated by the pulse of broadband optical energy from the plasma. I'm not so sure of this mechanism as air has almost no optical absorption in the portion of the spectrum passed by the envelope of the flash tube, but perhaps particles or gaseous impurities in the air do have absorption in the visible and near UV and so could also contribute to sound production.</p>

<p>Which of the above mechanisms of sound production is the most significant depends dramatically on details of construction of the flash unit.</p>

<p>Tom M</p>

[richardsperry]

Thanks for the replies

[frank_skomial]

<p>There were flashes discharging at high speed with operating voltages of 2000 VDC.<br>

Some tubes instead of polarizing electrode wrapped outside, have a narrow strip of conductive electricity paint at the back of the tube.</p>

[rodeo_joe1]

<p>The "pop" is exactly what any sound is, a pressure shockwave in air. This is mostly caused by air, dust and water vapour in proximity to the flashtube being rapidly heated by the infrared component of the light. The loudness of the pop might well change depending on the humidity and amount of pollution in the air. The tube itself will move a little as well, but most of the noise comes from air expanding.</p>

<p>Consider a 400 Watt/seconds studio flash that discharges in, say, 1/500th of a second. That's an average power dissipation of ~ 200 Kilowatts during the discharge time! Now try to imagine the amount of heat put out by 200 Kilowatts of tungsten lighting - pretty hot, yes? Even allowing for the fact that a Xenon tube is more light efficient than tungsten, there's still a huge heat output during that 1/500th of a second. So it's hardly surprising that the surrounding air gets a kick up the butt and then squeals about it.</p>

<p>BTW, if you want to hear an even louder pop from your flash, stick a black plastic bag just in front of it. The heatwave rapidly expands the plastic and hits it like a drum-skin. It can sound quite frightening, but all that's happened is that the plastic has moved and pushed some air about like a loudspeaker cone.</p>

<p>PS. Sound propagates through two main modes: Adiabatic and Isothermal. Adiabatic propagation travels as a thermal wave in air, so it should be no surprise that a thermal shock in air causes a sound.</p>

[Matt Laur]

<p>All of this makes for a good reminder about the amount of energy being released, and good reason to think very carefully about where you position naked-tube flash units around models' heads, especially things like overhead hair/rim lighting. When a tube shatters, it's pretty spectacular.</p>

[richardsperry]

Joe,

 

So the sound is not coming from the discharge inside the tube, as another poster stated. But from the air outside the tube

that's heated.

 

I'm confused now.

[lornesunley]

<p>The sound comes from both ... basically the temperature change from the electric arc causes a pressure wave (sound) the gas inside the tube transmits sound to the tube glass and that sound radiates, and the heat from the discharge also causes a pressure wave in the air surrounding the tube ...</p>

 

[tom_mann1]

<p>Rodeo, I wouldn't be so sure about your comment, <em>"... This is mostly caused by air, dust and water vapor in proximity to the flashtube being rapidly heated by the infrared component of the light. ..."</em></p>

<p>The IR transmission of pyrex can be seen here at http://infrared.als.lbl.gov/images/pyrexcurve.GIF . As you can see, no IR gets through pyrex above 3.7 microns. However, pure dry air has zero absorption from about 190 nm in the UV all the way through the far IR, so we can be 100% sure that absorption by pure dry air is not the cause.</p>

<p>The humidity in normal air does indeed show IR absorption, but it is almost entirely in a set of very narrow absorption lines around 3 microns. In contrast to the broadband absorption exhibited by your black plastic bag example, water's narrow absorption lines intercept only a small fraction of the total black body IR emission from the flash. In addition, the fundamental O-H absorption band of water around 3 microns is right where the IR transmission of the glass is dropping rapidly.</p>

<p>Dust particles certainly do have broad band absorption in the IR, but, at normal levels, they occupy such a small volumetric fraction of the air, they also don't intercept much of the IR emission from the lamp.</p>

<p>The bottom line is like I said earlier: The relative contribution of all of these different potential sources of sound can vary tremendously from unit to unit and with environmental conditions, so unless you want to do some tests, it can be hard to guess which mechanism (of those mentioned in this thread) is the largest contributor to the overall sound produced by the unit.</p>

<p>Cheers,</p>

<p>Tom M</p>

<p> </p>

[rodeo_joe1]

<p>Tom, just stick the palm of your hand in front of a flashtube while it's fired to see how little(!) heat is given off. A thin bit of Pyrex ain't gonna stop nothin'. And where does any energy absorbed by the Pyrex go? And how can you say that dry air can't be heated by the agitation of its molecules when faced with a <em>huge</em> and concentrated discharge of energy over quite a wide spectrum? Also look at how adiabatic sound propagation works, through the exchange of pressure and heat waves. Pressure and heat are almost interchangeable at low frequencies (like a "pop") as far as gases are concerned.</p>

<p>But blaming the contraction or expansion of the capacitor plates is reaching a bit don't you think?</p>

[tom_mann1]

<p>I'm sorry to have to say this, Rodeo, but you obviously don't understand even the most basic physics of the arguments I just made:</p>

<p>a) Of course the palm of my hand would get hot because it has broad band absorption throughout the visible and IR. In fact, it's absorption coefficient is so high that it will absorb the most of the optical energy incident on it in at most a few mm of depth. This is exactly the opposite of air which has weak and very narrow band absorption features.</p>

<p>b) <em>"A thin bit of Pyrex ain't gonna stop nothin" - </em>Did you even bother to read carefully what I said, and then look at the graph from one of the national labs that I cited? I said that in the IR, beyond about 3 microns, no light gets through the couple of mm of pyrex they used as a test sample. At shorter wavelengths, ie, throughout the shorter wave IR and into the visible, it's as you say, there is very little absorption of light in the pyrex, but at the wavelengths that are transmitted through pyrex, dry clean air is completely transparent, meaning it doesn't absorb energy, so it can't get hot from that mechanism. </p>

<p>c) <em>"And how can you say that dry air can't be heated by the agitation of its molecules when faced with a huge and concentrated discharge of energy over quite a wide spectrum? "</em> - Real easy. I just said it again and told you why in (b), above. You can have all the energy in the world being transmitted through something, and if that something doesn't absorb, it doesn't absorb, ie won't get hot. </p>

<p>d) <em>"Also look at how adiabatic sound propagation works, through the exchange of pressure and heat waves. Pressure and heat are almost interchangeable at low frequencies (like a "pop") as far as gases are concerned."</em> - These statements are irrelevant to the issue at hand. These statements deal with the propagation of sound, once a source of it has been established. We are trying to establish that source. </p>

<p>e) <em>"But blaming the contraction or expansion of the capacitor plates is reaching a bit don't you think?"</em> - No, it's not. </p>

<p>Acoustic noise from caps is a well known phenomena that you obviously aren't personally familiar with so you think it can't happen. I've been using and designing huge capacitor banks for physics experiments since the 1970's, and under the right conditions, one can easily hear this sound. If you don't believe me, perhaps you might believe a 2011 article from Electronic Design News on exactly this topic, stating that the noise, even from small caps can be so audible as to be annoying: http://www.edn.com/article/512775-Reduce_acoustic_noise_from_capacitors.php .</p>

<p>My bottom line is that there are many possible sources of acoustic noise from these sorts of flash units, and without doing careful tests (ie, a bit more than sticking your hand or a plastic bag in front of them), it is very difficult to say which source contributes the most to the sound you hear. </p>

<p>Tom M</p>

[Matt Laur]

<p>I accidentally ran an experiment along these lines just the other day. This is strictly anecdotal, but I consider myself to be a keen observer of equipment behavior, including the sounds made by such devices.<br /><br />I was setting up a Zeus pack-and-head rig, planning to place the head inside a large Lastolite brolly-style modifier. The intention was to use every bit the 1200WS the unit could dish out (that thing is ... very energetic). When used that way, I put the head inside the modifer naked - no reflector, etc.<br /><br />Not needing to go through the motions of mounting a reflector, I didn't pay attention, and left the black plastic protector cup in place over the flash tube. La la la, let's trigger the strobe and see how we're doing, Matt! Full power. I'm used to hearing that unit go <strong>POW</strong>! Instead I got ... <em>pop</em>. Um, and some smoke. <br /><br />That single discharge cooked off some of the black plastic on the inside of the protective cover, without, of course, touching the bulb. But what was significant was that I was discharging the unit by pressing the test button on the pack, where the very beefy capaciters are, 12 feet away from the head. Though the pack <em>does</em> emit a very minor audible noise at discharge (masked a bit by its cooling fan, I suspect), that sounds is a tiny fraction of the audible punch that eminates from the flash tube itself.<br /><br />This was especially clear to me because the now-slightly-singed protective cover (man, that smelled bad! but no harm done to the light or anything else) also substantially muffled the usual pop from the head. Which is to say, most of the noise is forward of the head's reflective face, right there where the flash tube is doing business.<br /><br />I'm disinclined to assign the classic pop to substantial movement of the quartz envelope itself (as in, it physically expanding enough, size-wise) to serve as the "drum" that, directly, moves enough air to create the pop. The clue for me, here, is the <em>quality</em> of the sound. I've discharged enough guns, lit enough flammable things, triggered enough flash powder, used enough muzzleloaders, idiotically played with enough hydrogen-filled balloons, and eaten enough beans to have a fairly acute ear for pressure waves and their many shapes. <br /><br />The thing that I note, in listening to the pop from a substantial flash discharge is the <em>wooph</em>-iness of it. Not a crack, not a smack, not the "note" I'd expect from it being the flash tube ringing, per se. This has the heat-expanding-gas feel to it, with a grace note of the electrical crackle I know is happening. Sorry I'm being so subjective and physics-for-poets about it, but I just happened to recall my recent 1200WS blunder, and it reminded me of how acutely I was listening to it for the rest of the day.</p>

[rodeo_joe1]

<p>Thanks Matt. I was about to cite the case of flash units where the tube is separated from the power pack and capacitors.<br>

Tom. Beside the above, there is plenty of additional circumstantial evidence to support the source of the POP being the proximity of the tube, and not the circuitry or capacitor(s).</p>

<p>I call witness #1 to the stand: Stereophonic hearing. This tends to locate the sound at the flash head and not at the body of the flash.</p>

<p>Witness #2: A Metz 402 or similar battery-powerpack with separate flash head. Agreed there is a slight click audible from the (ancient) power pack, but with the head (containing no capacitors) 5 feet away it's easy to hear that the characteristic low frequency pop is coming from the tube end.</p>

<p>Witness #3: Balcar and similar studio units with a separate tube and console design. Audible evidence as above and as Matt stated.</p>

<p>Witness #4: Various hotshoe guns with zoom head. The character of the sound doesn't vary much between make and model, regardless of build-quality, thickness of body plastic, etc. The sound <em>does</em> vary with the position of the zoom head; becoming deeper and louder as the tube and reflector are zoomed away from the fresnel lens. In other words as more free air is introduced in front of the flashtube. This also mitigates against the tube itself being the main source of sound.</p>

<p>Witness #5: My own studio strobes. The loudness of pop greatly increases when I put a metal snoot over the head. Now intuitively, a snoot should act as an inverse megaphone to muffle the sound, but actually seems to focus the sound wave from the front of the flash by increasing it's pressure (venturi effect). I can also feel a definite increase in the temperature of the silver reflector after a few pops - maybe irrelevent - so strike from the record.</p>

<p>Witness #6: All electronic flash apparatus taken as a whole. The characteristic discharge sound doesn't vary much regardless of the size or shape of tube or reflector used. It seems to depend solely on power and discharge time. This would seem to indicate an effect occuring in the immediate vicinity of the tube, but not originating within the tube. In short a localised heating effect. The timbre and loudness of the sound emitted also closely resemble the waveform of the light discharge.</p>

<p>In summing up I would say that adiabatic propagation of sound is totally relevent to the discussion. Heat = pressure in a gas. Heat any gas really quickly and you'll create an audible pressure wave - a sound.</p>

 

[Matt Laur]

<p>I'm sure that the entire matter is a complex combination of lots of things, varying quite a bit with everything from relative humidity to the mechanical structure supporting the flash tube and the recipe of the quartz used. I'm not discounting either RJ's or Tom's information and observations. There's a whole lot of energetic <em>stuff</em> going on when a big capaciter lets loose and a tiny lightning bolt dances around.<br /><br />I just happened to have the Zeus up again last night, and made more attention to the "clack" of the pack as it discharges. I used a cardboard tube to my ear, to help isolate spots on the pack from its fan noise. When that thing is cranked up, there's something solid but not entirely tame-sounding going on, internally. Obviously it's built to do it many thousands of times, so there can't be too much mechanical movement of dialectrics or other parts, or the damn thing wouldn't hold up very long. But it's <em>not</em> quiet inside the pack when it's dishing out the juice.<br /><br />Feeling even more curious here, I paid some more attention to the mechanical feeling at the head, as I did several test discharges. I happened to have a metal beauty dish mounted, and it sounded a lot like it was being tapped gently with a stick. The tone had a very acute attack. But on removing the dish, I was able to better hear the <em>wump</em> component of the whole audible event. There's more than one thing going on, here, to make the entirety of what we hear.</p>

[tom_mann1]

<p>Hi Guys -</p>

<p>Thanks for your detailed observations. That's exactly what is needed to figure out which of the possible physical mechanisms produce significant amounts of sound for a particular model of flash, the specific way it is set up, etc. I love your use of stereophonic hearing, the cardboard tube, etc.</p>

<p>WRT the hotshoe flashes with zoom heads sounding deeper as they are zoomed out, I still maintain that direct heating of the air is a negligible source of sound because air just doesn't absorb much light (even NIR light), particularly if it is relatively dry (as in winter), and clean of dust. If air absorbed even say as little as a few percent per meter, we wouldn't be able to see very far and landscape photographers would need to find a new hobby. ;-)</p>

<p>In the case of the zooming of the hotshoe flash, I think that what you are hearing is the lowering of the acoustic resonant frequency of the enclosed air between the Fresnel lens and the flash tube, not increased absorption of light by the larger depth of air in front of the flash tube. The change in acoustic resonant frequency is just like the variation in tone you hear if you tap on your cheek while varying the size and shape of your mouth. IMHO, the same explanation probably also explains the observations about the change of tone with the snoot.</p>

<p>IMHO, even though Joe discounted them, I think Joe's observations about the silver reflector heating up is quite relevant and may finally put us on the right path. While it is should be patently obvious that clean air can't absorb a few percent of light in a distance of a few feet, even the most highly polished metal has a reflectance at best only in the mid-90%, and often, a lot less than this. </p>

<p>In other words, every surface that receives light from the flash tube absorbs at least several percent of the light energy incident on it. The absorbed light immediately turns into a sharp jump in temperature of these surfaces. Until these surfaces cool by conduction to the cooler depths of the metal piece, this T-jump at the surface could easily heat a thin layer of air in front of these surfaces, so the air would also undergo a temperature jump on the time scale of milliseconds, and hence, expand and make a popping sound. </p>

<p>This is consistent with the "clacking" sound from Matt's beauty dish (...it's the expanding air pushing back on the metal), and, even more to the point, is consistent with the large increase in sound when you put a piece of plastic in front of the flash. </p>

<p>An interesting experiment would be to start with as close to a bare bulb setup your systems can do (eg, including no protective cover over the flash tube), and then start adding things in the vicinity of the flash tube that could absorb light, even if only a few percent, and see if the volume goes up. Obviously, the addition of such things isn't going to change the sound production by the glass / quartz envelope of the flash tube itself, the power pack, the wiring, or the nearby air. But, if the sound gets louder as you change from a bare bulb configuration to using a beauty dish or other light modifiers, I think we will have figured out our answer.</p>

<p>Cheers,</p>

<p>Tom M</p>