The simplest way to explain it is to say that 0dB is the reference level for full output (unattenuated) and all the negative dB markings are relative to that reference. In other words, -20dB means 20dB (nominally) below the reference, or 20dB of attenuation. You often see a negative infinity symbol at the bottom of the volume dB scale, which of course means full attenuation. Key to getting all of this is to understand that all volume controls are attenuators, not boosters.
I was an audio operator at an FM station in the late 1980s and early 1990s. I used a walkman as a monitor instead of the sound of the console, so I had exactly the end result just like any regular listener. I learned how to use the minimum and maximum audio processor thresholds to my advantage, so I could leave the scratches and cracks noises from some records well below the music. I also used this when playing ads and the station identification so they would sound louder without sounding compressed. I really enjoyed doing this. And yes, I respected the 0dB limit at all times.
Thank you, Paul, as always. I think it's important to mention also the fact that the dB scale is logarithmic, not linear. In fact, each -10 dB is equivalent to reducing the perceived volume by half. If we consider 0 dB as 100% of the sound we can get, this is the relation between dB and volume: 0 dB = 100% perceived output level -10 dB = 50% perceived output level (approximately) -20 dB = 25% perceived output level (approximately) -30 dB = 12.5% perceived output level (approximately) and so on...
In addition, logarithms are a way of performing multiplication by adding numbers together, so for example if the signal is multiplied by 2, that adds ~6 dB, if it is multiplied by .5 , that subtracts 6 dB, and if multiplied by .25, that subtracts 12 dB. You cannot think of ratios of dB, because that's double dipping, it's an additive scale that represents things that are multiplicative. Both a source of clarity and confusion. An odd thing is preamps usually have a gain stage, the 0 dB max. is relative to this gain or the volume pot itself, if the gain is +20 then max is +20 and 0 (unity) is around the middle. Never(?) marked that way though. And yes that means "only halfway" is also about the maximum usable setting with CDs. The rest is for records and weak radio stations.
@@PetraKann Not sure what these replies are saying, but to clarify I'm multiplying voltage. Doubling that quadruples power flow. Twice the power is 3 dB.
It's like a faucet. Everyone is aware that it is not responsible for the water pressure in the pipe, but rather dampens it so that we can draw water comfortably and in measured quantities.
The faucet is actually partially responsible for the pressure you receive from the faucet itself. That's like saying a step up voltage booster turning a 12v20a signal into a 150v1.5a elevated B+ isn't responsible for the voltage the load is receiving. It's not that simple lol. The conclusion is "both and" not "either or" lol. Flow rate is more comparable to current, and pressure is more comparable to voltage. But either way, it's still a "both and" situation. Is a plate resistor or cathode resistor responsible for the voltage a vacuum tube receives across it's plates? Or is it the power supply? Or is it the transformer on the pole by the street? Or is it the hydro damn your getting 35% of your electricity from or the coal and nuclear facilities providing the remainder? This is a very tricky, almost philosophical question that will rabbit hole you back to processes of the electron itself that we don't even fully understand. Most people dont even realize that electrons flow from negative to positive not the other way around, even though the energy used for work electrically flows + to -. The electrons are physically tunneling through the crystal lattice of a conductor, at only about an inch per second, faster with higher voltages. The extreme forces they generate through field effects around the conductor are actually what pipes energy to the load, you just have to exchange electrons with the conductor to access that field of energy flow, and this is also why the insulator of a conductor actually controls the rate of energy flow. The energy that does the work is actually flowing through a field on the outside of the conductor, reverse to the flow of electrons.
Manufacturers muddy the water on this by making their preamp/receiver volume controls reference level around 10 to 12 o'clock on the volume dial. To the average consumer this gives the illusion that their gear is more powerful by being louder at a relatively low setting, say around 8 o'clock. Electronically this gives the possibility of massive clipping and catastrophic speaker damage along with making half of the range on the volume dial useless. In any case, the negative dB on the volume dial is confusing to most that are not engineers. You did a good job explaining what it actually means and made the right choice in going to a 0-100 scale as that's more relatable to the average consumer.
As a ham radio operator, in my head I kept hearing "attenuate...attenuate"...then mere seconds later Paul you said the word. The preamplifier attenuates the source, which at "line level" would drive the power amp to near its maximum output. Your explanation could not have been better.
Thanks Paul, this video is of particular interest to me because it was actually a review of an older PS Audio product line, the GCC 100 and 250, that prompted me to research and finally understand how volume actually works in most audio amps. I had been an audiofile for many years. But Like most people, I believed that turning up the volume meant turning up the juice, or power. I was fascinated to learn that the volume knob actually progressively removes or releases suppression from the full signal, which increases the power and volume level. That was interesting enough, but the real kicker was learning that the quality of sound was actually increasing as the volume went up because the power suppression that keeps the volume low is also diminishing the signal quality. For years I thought that I could actually hear more detail at higher volume levels, and I had to grin and bear the standard comment that higher volume doesn't make better sound. Granted, the difference in detail might be minimal between moderate and higher listening levels. But I was certain that I was hearing a difference, with more detail at higher volume levels. The PS Audio GCC line of amps used a "gain control cell, or gain cell" for volume control (I think). The gain control cell, if I understand it, does not control volume by suppressing the full signal, and meading the power out as the volume is turned up. Consequently, the GCC amps apparently provided superb detail at low volume levels, which most other amps with standard volume controls don't by comparison. The reviews of the GCC line, all positive, said so. But there was one disappointing aspect... none of the reviews and articles ever explained exactly how the gain control cell worked. I've also heard it called a gain cell, and I'm guessing that these two terms refer to the same mechanism? So, after this long explanation, I'm hoping that you might be able to find a little time to respond and explain how the gain control cell actually controlled volume. If I remember correctly, the GCC 100 and 250 were class D integrated amps? Does the gain cell only work with class D, or could it also be used with class AB, and A amps? Thanks again for this video, and if you have a spare minute to shed some light of how gain cells work, I'd love to finally get a little closure on that old technical question.
Negative 10 db is typically referred to as 10db "down"...from full "rated" output. (typically volts) It all started with VU meters on pro mixing consoles (in the fifties). 0db was the max value you could apply before the mag-tape became "fully saturated". The scale just became widely accepted in recording studios, then home (open reel) tape decks, later cassette decks then pre-amps and amps. (although most VU meters on amps are..."just for show")
Hello Paul. I watch your videos every day and have learned a great deal from them. Thank you for making them. From watching your videos so much, it seems to me that one of your collaborators passes you the questions they ask and you improvise a response, in a great way, to say the least. However, in this question about dB, which seemed simple, the answer didn't come out well. Sometimes improvising an answer to a question based on a very solid knowledge base is not so easy. It would be worthwhile making another video to clarify this matter properly. Again, a thousand thanks for your videos (and for your high-end products, of which I am a proud owner of 3 of them [BHK Signature Preamplifier, DirectStream DAC MK2 and the AirLens streamer]).
Hi Paul, bravo for taking this one on, it's a toughie. You just can not explain this in a short video. You do your best, but I suspect that you have lost 75% of your viewers and frustrated the remaining 25%. PS, I love your channel.
another easy way to visualize is swimming in an empty ocean without jumping up, you are practically max out at 0m (or ft or whatever unit you prefer) above sea level, but you can free dive going down (negative values)
The decibel originates from the quantification of signal loss in telegraph and telephone circuits. A logarithmic scale is used (Intensity = 10 log10[P2/P1]. Doubling the intensity of a sound produces an increase of about 3 dB. Think of it as the base 10 logarithm of a power ratio. If you halve this power ratio, then the intensity is - 3bB. You can see this in the formula above. When P2 is less than P1 (for example when you are halving the power), the P2/P1 ratio is a value less than one. When you take the logarithm of a value less than 1 but greater than zero you get a negative result. When the power ratio is greater than 1 (say you double the power ratio) then taking the logarithm gives you a positive result. Check it with your calculator. log10 (2) = 0.30103 (Doubling the power ratio) log10 (0.5) = -0.30103 (Halving the power ratio)
Great explanation. So what of the Db scale when recording? is 0 Db on the meters considered a fully saturated signal and peaks shouldn't exceed that level?
Just for a bit of clarity, there are different measures of dB. dBu, dBw and dBV. dBV is the one that is used to measure voltage, for example, between a preamp and power amp. +6dBV is approximately double the voltage that 0dBV is. It is generally regarded that people can only hear the difference in level between 1dB, so it is a really useful measurement. It is especially important when measuring the frequency response of a piece of audio equipment. It’s as simple as checking that the dB level doesn’t vary, plus or minus, by, for example, 0.1dB between 20Hz and 20kHz. It’s much easier to follow, when testing, than raw voltages.
Same thing with cable going in to your home. For signal you wanted 0-10dB. I can’t remember how many mW 0dB was but it wasn’t a a lot. Anything above or below that would fail out. As channels and internet speed is added the frequency keeps increasing. The downfall is as frequency goes up as does the attenuation. You could have great signal down low in the band but as it increased the dB would decrease at a 45 degree angle across the graph. Number of splitters and cable type and length would effect this. You had to balance out the signal and return levels. Increase or take away from the other has the opposite reaction. Only way I found to knock down the signal without increase the return was to run an extra 1-200 feet of cable in the circuit. Would be a zip tied loop under the house. This would happen to a customer that’s 50 feet or so from a node and they only have one piece of equipment ie a modem. All that noise has to go somewhere
If amplification of a signal generates distortion, does the attenuation method make it less so between the min and max volume? There is always a distortion difference between loudness levels but is it so that attenuation enables the gear to have less distortion than when the signal would be amplified instead? After all We lover the current instead of getting more. Or is it irrelevant and it does not matter which method We have in a particular piece of gear and in each case the distortion is dependant on the loudness level alone? Just my random musings 😇
Anything raisedw to the power zero is 1. So the input is multiplied by 1 to give the output..it's neither larger nor smaller. A negative (minus) power is less than 1 so input may be multiplied by a fraction to get the output so it gets decreased or attenuated. The confusion arises because dBs are powers of the number like 10²=100.. the 2 is the dB. We do this because sound intensity varies over such a wide range, we'd be talking of 1000000000000 or .0000000001 in terms of the sound intensity and that gets unwieldy. So we use the powers of the 10 base which is a smaller range of numbers.. That's the basic idea, the actuality has slightly different multipliers but this is cut down
dB is for most people a measure of the loudness/volume, not a relative scale. So in most peoples minds, we relate to dB as 60dB = normal talk, jet engine = 120dB etc. Thats why it is confusing that when in amplifiers, the "dB" is then suddenly something else, something relative to something we don't have a reference for. So, adjusting volume from 0-100 makes more sense, since none of the scales says anything about the actual volume.
that "dB" you are referring to is the sound pressure level, the dB meter shown at the amp could be "0dB" being max (output) similar to source data where 0dB means max level
The relative volume scale is primarily used for home theater. 0db is considered ‘reference’ level, which means when calibrated correctly, each speaker will be playing at 85db at ‘0’, allowing for peaks of 105db from bed layer channels and 115db from the LFE channel. This ensures that your level of loudness is how the sound engineer intended it to be and it replicates what is heard in a commercial cinema.
You're talking about something different. The question was specifically asked about volume knobs, and in that case 0dB on the knob is not 85dB of sound pressure level. 0dB on the knob is "max volume" as Paul said. ..however loud that happens to be... 85 dB, 100 dB, 130 dB, or however loud it can go.
@@Hyxtryx no, it is exactly how I stated it.. 0db on the VOLUME knob is reference level, and the negative numbers are depicting how many db from reference level the VOLUME knob is set at… That was the question !
@@AgentEagle No, it is not how you said! Show me a volume knob (physical knob, not something buried in a menu) where you can make the dB value displayed on the front of the unit go above zero. They all stop at 0dB. 0dB is "no attenuation of the signal" like Paul said. Your dB of 0dB == 85dB sound pressure is a different thing, and not what the questioner asked. If what you said is true, every amp & speaker combo would play at the exact same volume level when turned up all the way, and that is patently false.
Proof: watch the video at 0:30. The questioner asks "why does the sound get louder, the closer you get to 0dB". So in the questioner's mind, 0dB is the LOUDEST sound. That means 0dB is when their volume knob is up ALL THE WAY. It has nothing to do with 85dB. 85dB == 0dB is something different that you are talking about, and has absolutely nothing to do with the question in this video.
Maybe this will only muddy the waters, but here's one way to think of it: It's like temperature, except that temperature is linear and loudness is logarithmic --- and, technically, you can use any volume level as "zero" for comparison purposes, whereas temperature has fixed references.
Decibel is a unit of measure and input voltage can be measured in dB as well as output sound pressure but they are not the same thing. Zero dB on an amp is full power and for every -3dB volts input the power is halved. Google dBV vs dB spl for more.
Unlike Amps, Volts and Watts which are specific individual measurements, db is a ratio, a relationship between TWO measurements. e.g. -3db means nothing by itself, but if you reference 0db to 2watts. a -3db would be 1watt. A 3db increase would be 4Watts. When a change in voltage or current is doubled or halved, it has changed by 6db. When Power is double or halved, it is a 3db change. SPL is added to db (dbSPL) when sound pressure is the measurements. But it is a logarithmic scale, not linear. e.g. to double your listening dbSPL is to go from some reference like 80dbSPL (moderate concert level) would be to go to 83dbSPL, not 160dbSPL. And would require a 3db or double power increase, say 100W to 200W. But if you want to increase by 10dbSPL, from that 80dbSPL to 90dbSPL (high concert level) would require going from 100W to 1,000W! 1db is considerate a just audible difference.
Decibel is not a unit of measure. You must specify what you are measuring such as dBSPL or dBm. Wikipedia has a pretty good explanation. dB is a ratio not a measurement.
I get the concept that negative db is only relative to a “zero” db that’s actually quite loud. But the fact is that the volume produced at that “zero” db level on a sound pressure meter is actually (say) 95 db. So….I and others might need another crack at this explanation.
In the audio chain, the dB level is a measurement at a certain point in the path. The measurement can happen at source data, at any input or output, and of course the sound pressure level (SPL) at your listening position. Just like there is a voltage level at the DAC output, preamp input ... dB measurement is log scale always relative to a baseline value, e.g. 1V is defined as 0dbV, sqrt(600/1000) = ~0.775V is defined as 0dBu, and there must be a pressure value for sound defined as 0dB SPL, ... the list goes on The SPL level created by the transducer (i.e. speaker) is related but does not translate directly from the "dB value" of an earlier stage, because there are layer(s) between what you hearing as sound and what the figure is measuring Just like the rotisserie chicken temperature is related to the oven temperature (and other factors), which is related to power usage of the oven (and other factors)
@@philiptong4978 Yes but the sound pressure level at the listening position is, for the practical purposes of the listener and user of the volume knob, the only level that matters, and should reflect the level at an assumed listening position of x feet. I suspect the answer here is that calibrating a volume knob where x equals zero which equals some high decibel level, in recognition of an engineering truth internal to the device that is meaningless to the user, is very silly. Paul thus says their volume knobs go from 0 to 100 which is better. But wouldn’t it make most sense to have the volume knob go from 0 to (say) 200 db when measured at 6 meters or some other defined distance, assuming a speaker of a defined sensitivity with intervening db gradations marked accordingly? I have no idea what “40” will do to my ears, but I know in my room 95 db will be painful. And yes there are variables which make such db markings inexact- so mark each level +/- 60 db, +/- 70 db etc.
consider a property having an absolute known max limit but there is no hazardous or well defined lower limit instead of counting up from an intuitive figure of zero the system uses a concept of max being zero, counting down to represent any smaller value with negative numbers
Per Agent Eagle "The relative volume scale is primarily used for home theater. 0db is considered ‘reference’ level, which means when calibrated correctly, each speaker will be playing at 85db at ‘0’, allowing for peaks of 115db." This is what Paul should have said. Each +10 dB is perceived as twice as loud. + 3dB takes 2x the power from your amp to achieve, and +10 dB is about 8x the power. How loud is 85 dB? Well about the same as at a crowded restaurant where the collective voices are very loud, and you almost have to shout to talk to anyone else at your table. Normal room conversation at home is in the 60-65 dB range. Comfortable listening of your stereo is in the 75 dB area. Loud is 85-90 dB (and brief peaks can be more). OSHA workplace standards are don't exceed 90 dB constant for long periods of time. Rock concerts can be 100+ dB, or standing by large speakers at a wedding. Jack hammers up close around 110 dB. Speaker efficiency is measured at 2.83 v into 8 Ohms nominal (1 watt) at one meter. Most speakers are 85dB - 90dB at that level, so the excess watts in a high-powered amp are for really loud levels and/or peak levels. There are free dB apps you can put on your phone and you easily see what the sound level is wherever you choose.
You're talking about something different. The question was specifically asked about volume knobs, and in that case 0dB on the knob is not 85dB of sound pressure level. 0dB on the knob is "max volume" as Paul said. ..however loud that happens to be... 85 dB, 100 dB, 130 dB, or however loud it can go.
The dial is showing the gain margin of the amplifier. The gain margin is found in the frequency response function of the system. This is a measure of the amount that the gain can increased until the system becomes unstable. This comes from control theory and the negative sign is because the characteristic equation is in the denominator. A dB is by definition 20 log10(|g(iw)h(jw)|. thus any value will be negative because the expression for the characteristic equation is in the denominator. The use of the bode is because the frequency response of the amp can be obtain experimentally. When you increase the gain the frequency response moves up and the gain margin is reduced. When zero is reached the system is considered unstable thus its max amount that the amp can produce. To better understand this look at some youtube videos on bode diagrams hope this helps. it not a mystic once you understand the theory.
People have to understand, that db on its own is just an algorithm, the letter behind it makes it a unit. And db - only occurs on electrical db‘s like db m or db u. If you use acoustic db measurements like db SPL ,there is no minus.
dB is not a unit. It is a unit-less value. The letter you are referring to specifies what the reference 0dB value is, and is sometimes also the units for that reference value. A dB value, like -3dB is unit-less itself. It is just a ratio like 2/3 or 27/8. The units come from whatever units the reference value has, but the actual dB value itself is unit-less. > "And db - only occurs on electrical db‘s like db m or db u. If you use acoustic db measurements like db SPL ,there is no minus." There is *always* a minus. The reference value can never be 0, so any value you tell me, I can always tell you a value that is half that. Therefore, there is *always* a minus.
Ok then but, if at 0 DB an amplifier puts out its maximum power, and different amplifiers have different power levels (some are weaker, some are beefier), that means that 0 dB will translate to a different output (loudness measured in dB SPL) on different power amplifiers. And if I am not mistaken that also means that different (=rated at different wattage) power amplifiers will have different output given the same input (say -12 dB) from a preamp (because that's relative to 0 dB). But I also remember what you said about amplifier gain: two different power amplifiers can have different wattage specs and yet have the same gain and produce the same loudness at a given level of the input. This confuses me 🤔🤔
I'll know better in the mourning but I think it's due to the resistance aka the ohms that the speaker is asking from the amp. With the same GAIN 1:15 which might be 2v in and 30v out (for example) a speaker while at 4ohms will draw more watts The 3 variables which can change the equation tied to one another. Watts = Voltage * Current Less resistance means more current Power isn't that interesting
dB is a relative scale, as you understand 0dB can be defined as any value, for power measurement, it can be 1mW (= 0dBm) or 1W (= 0dBW = 30 dBm) or any other baseline value think of gain as a multiplier and wattage is the hard upper limit of power output before bad things (noise, distortion, clipping, overheat, reduced lifespan, burn out, sparks, smoke, fire, spontaneous combustion, armageddon) happen high and low is all relative a single ant can carry a lot of weight relative to its own weight (high gain) but it certainly won't be able drag you away in sleep (limited power output) a healthy person likely cannot carry multiple times of his/her own weight (low gain) but should be able to get up from bed and walk around unassisted (that is higher power than the single ant)
Great explanation! Thanks, Paul. As a product developer for my whole career, I look back on the electro-mechanical dB levels we used to measure when developing power tools. Manufacturers always recommend wearing hearing (And eye) protection with power tools due to the sound power level of either the tool or the work it is doing with your ear generally at arms length or less. You often hear references to live rock concerts and jet engines as being 100+ dB levels. I assume this kind of measurement probably relates back to voltage levels in the electronic world in some way.
yes 20log10(1)=0. In control theory the system is unstable when 1+g(jw)h(jw)=0. for this to occur the magnitude of g(jw)h(jw)must be 1 that why you have the zero. The angle is one 180 but the term is in the denominator thus it is -180 degrees. you are working with complex numbers. This scale is coming from log plots of the sinusoidal transfer function. That is why in old audio that recorded when you get pass zero dB it usually was red (unstable)
@@frederickajustagosto4765 In this case though, Unity Gain, we're talking about what the attenuator (volume control) is doing to the signal, which is this case, is nothing. Zero on any kind of level meter vs. zero on attenuator (fader, pot) are not interchangeable terms, for those following along at home.
@@carterwilliamhumphrey3373 look at the definition of a dB unity is a value of one its log10 is 0 10^0=1 that is why its zero in dB. it i using log scale. saying the same thing 1. the word unity means 1. note he says that zero is what will run amp to loudest point. people confuse log with linear. You can't have the log of zero try it in your calculator.
@@frederickajustagosto4765 We seem to be deep into the math here, and I don't know why. If you want to go into an amp circuit, a gain factor of 1 is, in fact. unity gain of the amplifier, meaning: what goes in level-wise, is what comes out. Similarly, if you put an audio fader at "0", you get the same signal level on output as input.
In the real world, 0dB is the same as 0 sound. Why not use 0dB and above? The lowest decibel a human can hear is 0 dB. In some exceptional cases, humans can hear sounds down to -15 dB. 0 dB sounds are very soft and barely perceptible by the human ear. To understand just how soft, imagine that a whisper is 30 dB and 30 dB is 1,000 times louder than a 0 dB sound. Measurements of a speaker with a microphone are never measured in - dB. Minus dB The ear hears 10 dB as twice as loud, but it is 3 dB which is actually twice as loud, even if the ear does not perceive 3 dB as twice as loud
because the PCM recording system has an absolute upper max limit, after that point everything is recorded with the same data value, imagine clipping waveform
This was not a complete explanation. A key piece of info left out is that a decibel value is a RATIO. It is always a ratio between some "reference" quantity and the quantity you're trying to describe. And it's a logarithmic scale. If it's voltages being referred to, then dropping 6dB is a halving of the voltage, and increasing by 6dB is a doubling of the voltage. If it's power, then 3dB drop is a halving of the power, and a 3dB increase is a doubling of the power. It is always a measure of the ratio between the reference value and the value you're trying to describe. What is the reference value? It's different for various things. You'll have to look that up. For sound, it's some extremely low sound pressure level at some specific distance from the sound source. For amplifier volume controls, the chosen reference level is "volume knob at full volume". Nothing else makes sense other than "volume knob at zero", but the reference value can't be zero, because anything multiplied by 0 is still 0, so you cannot have a ratio with zero as one of the values. Hence, 0dB was made to be "full volume" for amps that show the volume as dB. 0dB is also full volume for recording (as seen on VU meters), so there was already some precedent set for 0dB to mean "full volume".
What ?.0 db is .77 volts consumer level and 1.23 volts at pro levels , no ? And that is not the loudest , because of headroom ....right ? Isn't the headroom like +22 db on pro level equipment ? I think I'm now more confused than ever regarding decibels ! 😛
You, we are confused because unit is incomplete. The same happens with speaker's sensitivity, reviewers take shortcuts with the definition so the conclusion isn't accurate. Bel is a logarithmic relation between a value and it's reference. That reference must be defined in the «unit». Two analog references: 0 dBu = 0,7746 V RMS and 0 dBV = 1 V RMS. In the digital domain we have 0 dBFS that is the max level. Some useful examples: 0,3162 V RMS = -10 dBV RMS ; 1,228 V RMS = 4 dBu RMS. How do we get these numbers: 20×log(0,3162 V/1 V) = -10 dBV with means that a line tension of 0,3162 V relative to the reference V is -10 dBV, the second example 20×log(1,228 V/0,7746 V) = 4 dBu with means that a line tension of 1,228V relative to the reference «u» is 4 dBV. The full unit is defined by the combination of dB an d the reference (V, u, FS, SPL[A,B,C], etc).
the ~0.775V figure comes from result voltage of 1mW across 600 Ohm, since then sqrt(0.6)V became the baseline value of 0dBu, and +4dBu is the analogue audio level standard, if a device can handle max 24dBu then it is said to have 20dB headroom (from the nominal level +4dBu) "Consumer devices" commonly uses dBV units, 0dBV = 1V so dBu and dBV both quantify voltage with logs, but at different scale, like Celsius and Fahrenheit in log scale
There are snippets of truth all around here, but also a lot of confusion. Some is from making no distinction between the following different uses of the term decibel: (A) "Absolute" decibels, which are computed as 10 times the Base 10 logarithm of the ratio of a power level under discussion to a REFERENCE power level, which MUST ALSO be specified. [In the general sense, all dB numbers are just numeric ratios, with no units. But in the context of these kinds of dBs, since the reference signal's absolute level and unit(s) are specified, the level and unit(s) for a given signal are also known absolutely.] E.g., dBW, where the "W" refers to 1 watt, or dBm, where the "m" refers to 1 milliwatt. E.g., 1 milliwatt = 0 dBm, or -30 dBW; 2 milliwatts = +3 dBm; 10 milliwatts = +10 dBm; 400 watts = 26 dBW, or 56 dBm. The math is TEN times the log ratio because we're talking about "deci-bels", a tenth of a bel, not "bels". [Let's agree to not talk about bels any further - but if we did, the math would be just the Base 10 log of the ratio instead of ten times that]. We use logarithms because our hearing of different loudness levels correlates to a logarithmic "compression" scale of power ratios, NOT their simple linear ratios. Twice the power level is not perceived as being twice as loud - it only sounds about 30% louder. It takes TEN TIMES the power to sound twice as loud. This is behind the reason why, e.g., a 2-watt amplifier can play music fairly loudly 95% of the time with many speakers - but you might need 100 watts for the peak levels just a few decibels louder. And why a 100W amp can only play 3 dB louder than a 50W amp [and why I wish that all amps had a dBW rating as well as a wattage rating]. So we use Base 10 logarithmic ratios in the form of decibels to best correlate to the differences that we perceive between different power levels (whether it's the acoustic power of vibrating air molecules or electrical power in or out of an audio device). It was the ancient telephone industry that adopted this practice; technically, 0 dBm is 1 milliwatt into 300 ohms; but we usually gloss over that specific impedance level in our business, as we have our own common pseudo-standards, e.g., < 1000 ohms for line-level outputs, >= 10k ohms for line-level inputs. (B) A common special case of of (A) is dB SPL, where SPL - special because the reference is often not explicitly referenced, just assumed from the context. SPL refers to the specific acoustic reference "Sound Pressure Level" that most people with reasonably good hearing can just barely hear in air at 1 kHz - "the threshold of human hearing". That standard reference level is 0.00002 Pascals (Newtons/square meter - Root Mean Squared [RMS] levels, to be specific) of sound PRESSURE (NOT power, hence the math demands TWENTY times the Base 10 logarithmic ratio for these decibels - see more on this in Item C below). Sound AT THAT REFERENCE LEVEL is therefore expressed as 0 dB SPL - or often just 0 dB for short. This is the scale that we use to express levels up to, say, 40 dB [SPL] as quiet, 80 as noisy, 100 as symphony orchestra crescendo loud, 120 as painful, etc. These are sound pressure levels in air, which have NO ABSOLUTE DIRECT correlation to decibels of watts or volts or amps in electronic equipment producing such sound levels UNLESS much additional information is taken into account, such as loudspeaker efficiency (acoustic watts output for a given electrical power input), distance from the speaker, the spatial directivity pattern of the speaker, the direction of measurement relative to the speaker, and whether the measurement is performed in open 3D space or an approximation of it (such as a practical anechoic chamber, or a time-windowed measurement to simulate open space by ignoring room reflections) or actually including a room's characteristics and its sound reflections. (C) Common "throw-around" (my term) decibels, where the reference level is WHATEVER IT IS in the understood context [e.g., 10 millivolts out of a phono cartridge, or 2 volts out of a DAC, or 1 watt out of a power amp, or whatever], and the actual level is then expressed as a number of dB above (+) or below (-) THAT level. IMPORTANT DIFFERENCE: If voltages (or currents or SPL) are being compared instead of power, then the math demands that the decibel numbers be calculated as TWENTY TIMES (not ten times) the Base 10 logarithmic ratios. This is because power in watts is proportional to the voltage or current SQUARED, and the logarithm of a number squared is double the log of the original number. A 2:1 voltage ratio = 4 times the power ratio = +6 dB in power.] The subject is not terribly confusing to electrical engineers or technicians, but since decibels are all about numbers, even a minimal clear and cogent explanation does require a bit of math. I hope this helps a bit.
A preamp with a phono cartridge or microphone input DOES unequivocally amplify those signals quite a bit (but enough to drive loudspeakers). If it has a moving coil phono cartridge input, it will amplify it even MORE. MOST preamps can also amplify line-level inputs by a few decibels (perhaps 6 or so). Hence the norm of calling all such devices preamps. We could just stop there. // All preamps designed for regular human interaction also include a volume control of some type to attenuate its otherwise fixed line-level output that is designed to drive a separate amplifier that is capable of yet more amplification and much greater output power to drive low-impedance devices like loudspeakers. So-called "passive preamps" (of which there are few examples) are the truly confusing case, because regardless of what else they may do that most preamps do (such as source selection and attenuation/volume control), they by definition do NOT amplify a signal at all in the classic sense (i.e., with active devices like transistors or tubes) - only pass it as-is, or attenuate it. Some may include a transformer to boost output voltage, but that is not amplification - any such voltage boost comes at the cost of higher output impedance and less current.
Nuh ! still don't get it , but if I just stick to using whatever you want to call the knob that controls the (what I call) volume, and have it at (my idea of) a reasonable level ,it🤞👍🤞🙏 shouldn't matter if I do or don't understand all that.
This isn't a great explanation. It's simply that when you adjust the volume on your amplifier, you're not actually changing the amplifier itself - in essence the amplifier is always boosting the signal at the maximum/fixed level. Instead, you are just altering a resistor before the amplifier in order that you don't only have it playing at the maximum volume. So the negative dB are referring to the reduction before the amplifier - how much that resistor is pulling down the sound before it gets boosted by the fixed output of your amplifier. Turning the volume all the way up to "0" correctly implies that that there is now 0 db of attenuation (resistance).
the amplifier acts like a multiplier. if its gain is fixed the volume you put in is multiplied by its gain (increases). however you can change the gain on some amps that another story. but this is showing the gain margin of the amp. however I believe the manufacturer is the one that puts the scale thinking it will be the best way to sell the amp at the end that's what they do
Kudos to Paul for explaining dB without mentioning logarithms. With a background in physics logarithms make complete sense to me, but I know most people's eyes would glaze over at the mention of logarithms. Of course a more complete understanding of the dB scale does require knowledge of logarithms; e.g. it explains why voltage ratios vs. power ratios are a factor 2 apart in terms of dB.
easy to say i respect more the person who designed it. They put that in the volume to communicate the way they converse in dB. May be it better to keep learning and
The simplest way to explain it is to say that 0dB is the reference level for full output (unattenuated) and all the negative dB markings are relative to that reference. In other words, -20dB means 20dB (nominally) below the reference, or 20dB of attenuation. You often see a negative infinity symbol at the bottom of the volume dB scale, which of course means full attenuation. Key to getting all of this is to understand that all volume controls are attenuators, not boosters.
I was an audio operator at an FM station in the late 1980s and early 1990s. I used a walkman as a monitor instead of the sound of the console, so I had exactly the end result just like any regular listener. I learned how to use the minimum and maximum audio processor thresholds to my advantage, so I could leave the scratches and cracks noises from some records well below the music. I also used this when playing ads and the station identification so they would sound louder without sounding compressed. I really enjoyed doing this. And yes, I respected the 0dB limit at all times.
Thank you, Paul, as always. I think it's important to mention also the fact that the dB scale is logarithmic, not linear. In fact, each -10 dB is equivalent to reducing the perceived volume by half. If we consider 0 dB as 100% of the sound we can get, this is the relation between dB and volume:
0 dB = 100% perceived output level
-10 dB = 50% perceived output level (approximately)
-20 dB = 25% perceived output level (approximately)
-30 dB = 12.5% perceived output level (approximately)
and so on...
czcams.com/video/xSe7y9tYex8/video.html
In addition, logarithms are a way of performing multiplication by adding numbers together, so for example if the signal is multiplied by 2, that adds ~6 dB, if it is multiplied by .5 , that subtracts 6 dB, and if multiplied by .25, that subtracts 12 dB. You cannot think of ratios of dB, because that's double dipping, it's an additive scale that represents things that are multiplicative. Both a source of clarity and confusion.
An odd thing is preamps usually have a gain stage, the 0 dB max. is relative to this gain or the volume pot itself, if the gain is +20 then max is +20 and 0 (unity) is around the middle. Never(?) marked that way though. And yes that means "only halfway" is also about the maximum usable setting with CDs. The rest is for records and weak radio stations.
@@spentron1 Minor correction, multiplying by 2 adds 6 dB.
@@spentron1 ~3dB
@@PetraKann Not sure what these replies are saying, but to clarify I'm multiplying voltage. Doubling that quadruples power flow. Twice the power is 3 dB.
It's like a faucet. Everyone is aware that it is not responsible for the water pressure in the pipe, but rather dampens it so that we can draw water comfortably and in measured quantities.
The faucet is actually partially responsible for the pressure you receive from the faucet itself. That's like saying a step up voltage booster turning a 12v20a signal into a 150v1.5a elevated B+ isn't responsible for the voltage the load is receiving. It's not that simple lol. The conclusion is "both and" not "either or" lol.
Flow rate is more comparable to current, and pressure is more comparable to voltage. But either way, it's still a "both and" situation.
Is a plate resistor or cathode resistor responsible for the voltage a vacuum tube receives across it's plates? Or is it the power supply? Or is it the transformer on the pole by the street? Or is it the hydro damn your getting 35% of your electricity from or the coal and nuclear facilities providing the remainder? This is a very tricky, almost philosophical question that will rabbit hole you back to processes of the electron itself that we don't even fully understand.
Most people dont even realize that electrons flow from negative to positive not the other way around, even though the energy used for work electrically flows + to -. The electrons are physically tunneling through the crystal lattice of a conductor, at only about an inch per second, faster with higher voltages. The extreme forces they generate through field effects around the conductor are actually what pipes energy to the load, you just have to exchange electrons with the conductor to access that field of energy flow, and this is also why the insulator of a conductor actually controls the rate of energy flow. The energy that does the work is actually flowing through a field on the outside of the conductor, reverse to the flow of electrons.
That is an excellent analogy
Thank you Paul for all that you do. This question has been on my mind for many years.
Manufacturers muddy the water on this by making their preamp/receiver volume controls reference level around 10 to 12 o'clock on the volume dial. To the average consumer this gives the illusion that their gear is more powerful by being louder at a relatively low setting, say around 8 o'clock. Electronically this gives the possibility of massive clipping and catastrophic speaker damage along with making half of the range on the volume dial useless. In any case, the negative dB on the volume dial is confusing to most that are not engineers. You did a good job explaining what it actually means and made the right choice in going to a 0-100 scale as that's more relatable to the average consumer.
wow...I finally understand. I have always been confused by this....thank you
As a ham radio operator, in my head I kept hearing "attenuate...attenuate"...then mere seconds later Paul you said the word. The preamplifier attenuates the source, which at "line level" would drive the power amp to near its maximum output. Your explanation could not have been better.
Thanks for the explanation. I have also always wondered about this
Thanks Paul, this video is of particular interest to me because it was actually a review of an older PS Audio product line, the GCC 100 and 250, that prompted me to research and finally understand how volume actually works in most audio amps. I had been an audiofile for many years. But Like most people, I believed that turning up the volume meant turning up the juice, or power. I was fascinated to learn that the volume knob actually progressively removes or releases suppression from the full signal, which increases the power and volume level. That was interesting enough, but the real kicker was learning that the quality of sound was actually increasing as the volume went up because the power suppression that keeps the volume low is also diminishing the signal quality. For years I thought that I could actually hear more detail at higher volume levels, and I had to grin and bear the standard comment that higher volume doesn't make better sound. Granted, the difference in detail might be minimal between moderate and higher listening levels. But I was certain that I was hearing a difference, with more detail at higher volume levels. The PS Audio GCC line of amps used a "gain control cell, or gain cell" for volume control (I think). The gain control cell, if I understand it, does not control volume by suppressing the full signal, and meading the power out as the volume is turned up. Consequently, the GCC amps apparently provided superb detail at low volume levels, which most other amps with standard volume controls don't by comparison. The reviews of the GCC line, all positive, said so. But there was one disappointing aspect... none of the reviews and articles ever explained exactly how the gain control cell worked. I've also heard it called a gain cell, and I'm guessing that these two terms refer to the same mechanism? So, after this long explanation, I'm hoping that you might be able to find a little time to respond and explain how the gain control cell actually controlled volume. If I remember correctly, the GCC 100 and 250 were class D integrated amps? Does the gain cell only work with class D, or could it also be used with class AB, and A amps? Thanks again for this video, and if you have a spare minute to shed some light of how gain cells work, I'd love to finally get a little closure on that old technical question.
Great question and great answers. Thanks Paul
Negative 10 db is typically referred to as 10db "down"...from full "rated" output. (typically volts)
It all started with VU meters on pro mixing consoles (in the fifties).
0db was the max value you could apply before the mag-tape became "fully saturated".
The scale just became widely accepted in recording studios, then home (open reel) tape decks, later cassette decks then pre-amps and amps.
(although most VU meters on amps are..."just for show")
Excellent video Paul, thank you.
This is the video I was waiting for
Hello Paul. I watch your videos every day and have learned a great deal from them. Thank you for making them. From watching your videos so much, it seems to me that one of your collaborators passes you the questions they ask and you improvise a response, in a great way, to say the least. However, in this question about dB, which seemed simple, the answer didn't come out well. Sometimes improvising an answer to a question based on a very solid knowledge base is not so easy. It would be worthwhile making another video to clarify this matter properly. Again, a thousand thanks for your videos (and for your high-end products, of which I am a proud owner of 3 of them [BHK Signature Preamplifier, DirectStream DAC MK2 and the AirLens streamer]).
Thank you for a helpful explanation
Hi Paul, bravo for taking this one on, it's a toughie.
You just can not explain this in a short video. You do your best, but I suspect that you have lost 75% of your viewers and frustrated the remaining 25%.
PS, I love your channel.
another easy way to visualize is swimming in an empty ocean without jumping up, you are practically max out at 0m (or ft or whatever unit you prefer) above sea level, but you can free dive going down (negative values)
The decibel originates from the quantification of signal loss in telegraph and telephone circuits.
A logarithmic scale is used (Intensity = 10 log10[P2/P1]. Doubling the intensity of a sound produces an increase of about 3 dB.
Think of it as the base 10 logarithm of a power ratio.
If you halve this power ratio, then the intensity is - 3bB.
You can see this in the formula above. When P2 is less than P1 (for example when you are halving the power), the P2/P1 ratio is a value less than one. When you take the logarithm of a value less than 1 but greater than zero you get a negative result.
When the power ratio is greater than 1 (say you double the power ratio) then taking the logarithm gives you a positive result.
Check it with your calculator.
log10 (2) = 0.30103 (Doubling the power ratio)
log10 (0.5) = -0.30103 (Halving the power ratio)
3 dB up is from doubling the power, but 10 dB is from about 3x the power, and is perceived as twice as loud.
@@stevefisher8323 Power ratio
Thanks Paul Sir...
Great question 🔥
Great explanation. So what of the Db scale when recording? is 0 Db on the meters considered a fully saturated signal and peaks shouldn't exceed that level?
Just for a bit of clarity, there are different measures of dB. dBu, dBw and dBV.
dBV is the one that is used to measure voltage, for example, between a preamp and power amp. +6dBV is approximately double the voltage that 0dBV is. It is generally regarded that people can only hear the difference in level between 1dB, so it is a really useful measurement.
It is especially important when measuring the frequency response of a piece of audio equipment. It’s as simple as checking that the dB level doesn’t vary, plus or minus, by, for example, 0.1dB between 20Hz and 20kHz. It’s much easier to follow, when testing, than raw voltages.
Same thing with cable going in to your home. For signal you wanted 0-10dB. I can’t remember how many mW 0dB was but it wasn’t a a lot. Anything above or below that would fail out. As channels and internet speed is added the frequency keeps increasing. The downfall is as frequency goes up as does the attenuation. You could have great signal down low in the band but as it increased the dB would decrease at a 45 degree angle across the graph. Number of splitters and cable type and length would effect this. You had to balance out the signal and return levels. Increase or take away from the other has the opposite reaction. Only way I found to knock down the signal without increase the return was to run an extra 1-200 feet of cable in the circuit. Would be a zip tied loop under the house. This would happen to a customer that’s 50 feet or so from a node and they only have one piece of equipment ie a modem. All that noise has to go somewhere
My interpretation of what you said here is that 0db is what’s coming out of the DAC. -12db is the preamp taking 12db out of the signal?
Thank you Paul , If 50% of all the people around the globe will be like you. the world will look different - Meaning : A Better world !!!
If amplification of a signal generates distortion, does the attenuation method make it less so between the min and max volume? There is always a distortion difference between loudness levels but is it so that attenuation enables the gear to have less distortion than when the signal would be amplified instead? After all We lover the current instead of getting more. Or is it irrelevant and it does not matter which method We have in a particular piece of gear and in each case the distortion is dependant on the loudness level alone? Just my random musings 😇
Anything raisedw to the power zero is 1. So the input is multiplied by 1 to give the output..it's neither larger nor smaller. A negative (minus) power is less than 1 so input may be multiplied by a fraction to get the output so it gets decreased or attenuated.
The confusion arises because dBs are powers of the number like 10²=100.. the 2 is the dB.
We do this because sound intensity varies over such a wide range, we'd be talking of 1000000000000 or .0000000001 in terms of the sound intensity and that gets unwieldy.
So we use the powers of the 10 base which is a smaller range of numbers..
That's the basic idea, the actuality has slightly different multipliers but this is cut down
dB is for most people a measure of the loudness/volume, not a relative scale. So in most peoples minds, we relate to dB as 60dB = normal talk, jet engine = 120dB etc. Thats why it is confusing that when in amplifiers, the "dB" is then suddenly something else, something relative to something we don't have a reference for.
So, adjusting volume from 0-100 makes more sense, since none of the scales says anything about the actual volume.
that "dB" you are referring to is the sound pressure level, the dB meter shown at the amp could be "0dB" being max (output) similar to source data where 0dB means max level
The relative volume scale is primarily used for home theater. 0db is considered ‘reference’ level, which means when calibrated correctly, each speaker will be playing at 85db at ‘0’, allowing for peaks of 105db from bed layer channels and 115db from the LFE channel. This ensures that your level of loudness is how the sound engineer intended it to be and it replicates what is heard in a commercial cinema.
This is what Paul should have said, and then indicated what 85dB sounds like.
You're talking about something different. The question was specifically asked about volume knobs, and in that case 0dB on the knob is not 85dB of sound pressure level. 0dB on the knob is "max volume" as Paul said. ..however loud that happens to be... 85 dB, 100 dB, 130 dB, or however loud it can go.
@@Hyxtryx no, it is exactly how I stated it.. 0db on the VOLUME knob is reference level, and the negative numbers are depicting how many db from reference level the VOLUME knob is set at… That was the question !
@@AgentEagle No, it is not how you said! Show me a volume knob (physical knob, not something buried in a menu) where you can make the dB value displayed on the front of the unit go above zero. They all stop at 0dB. 0dB is "no attenuation of the signal" like Paul said. Your dB of 0dB == 85dB sound pressure is a different thing, and not what the questioner asked.
If what you said is true, every amp & speaker combo would play at the exact same volume level when turned up all the way, and that is patently false.
Proof: watch the video at 0:30. The questioner asks "why does the sound get louder, the closer you get to 0dB". So in the questioner's mind, 0dB is the LOUDEST sound. That means 0dB is when their volume knob is up ALL THE WAY. It has nothing to do with 85dB.
85dB == 0dB is something different that you are talking about, and has absolutely nothing to do with the question in this video.
Maybe this will only muddy the waters, but here's one way to think of it:
It's like temperature, except that temperature is linear and loudness is logarithmic ---
and, technically, you can use any volume level as "zero" for comparison purposes,
whereas temperature has fixed references.
Decibel is a unit of measure and input voltage can be measured in dB as well as output sound pressure but they are not the same thing. Zero dB on an amp is full power and for every -3dB volts input the power is halved. Google dBV vs dB spl for more.
Unlike Amps, Volts and Watts which are specific individual measurements, db is a ratio, a relationship between TWO measurements. e.g. -3db means nothing by itself, but if you reference 0db to 2watts. a -3db would be 1watt. A 3db increase would be 4Watts. When a change in voltage or current is doubled or halved, it has changed by 6db. When Power is double or halved, it is a 3db change. SPL is added to db (dbSPL) when sound pressure is the measurements.
But it is a logarithmic scale, not linear. e.g. to double your listening dbSPL is to go from some reference like 80dbSPL (moderate concert level) would be to go to 83dbSPL, not 160dbSPL. And would require a 3db or double power increase, say 100W to 200W. But if you want to increase by 10dbSPL, from that 80dbSPL to 90dbSPL (high concert level) would require going from 100W to 1,000W!
1db is considerate a just audible difference.
Yep. I am very surprised that Paul left out that key word "ratio". You can't explain what a decibel even *is* without that word!
Decibel is not a unit of measure. You must specify what you are measuring such as dBSPL or dBm. Wikipedia has a pretty good explanation. dB is a ratio not a measurement.
I get the concept that negative db is only relative to a “zero” db that’s actually quite loud. But the fact is that the volume produced at that “zero” db level on a sound pressure meter is actually (say) 95 db. So….I and others might need another crack at this explanation.
In the audio chain, the dB level is a measurement at a certain point in the path.
The measurement can happen at source data, at any input or output, and of course the sound pressure level (SPL) at your listening position. Just like there is a voltage level at the DAC output, preamp input ...
dB measurement is log scale always relative to a baseline value, e.g. 1V is defined as 0dbV, sqrt(600/1000) = ~0.775V is defined as 0dBu, and there must be a pressure value for sound defined as 0dB SPL, ... the list goes on
The SPL level created by the transducer (i.e. speaker) is related but does not translate directly from the "dB value" of an earlier stage, because there are layer(s) between what you hearing as sound and what the figure is measuring
Just like the rotisserie chicken temperature is related to the oven temperature (and other factors), which is related to power usage of the oven (and other factors)
@@philiptong4978 Yes but the sound pressure level at the listening position is, for the practical purposes of the listener and user of the volume knob, the only level that matters, and should reflect the level at an assumed listening position of x feet. I suspect the answer here is that calibrating a volume knob where x equals zero which equals some high decibel level, in recognition of an engineering truth internal to the device that is meaningless to the user, is very silly. Paul thus says their volume knobs go from 0 to 100 which is better. But wouldn’t it make most sense to have the volume knob go from 0 to (say) 200 db when measured at 6 meters or some other defined distance, assuming a speaker of a defined sensitivity with intervening db gradations marked accordingly? I have no idea what “40” will do to my ears, but I know in my room 95 db will be painful. And yes there are variables which make such db markings inexact- so mark each level +/- 60 db, +/- 70 db etc.
I’m none the wiser after this explanation!
consider a property having an absolute known max limit but there is no hazardous or well defined lower limit
instead of counting up from an intuitive figure of zero
the system uses a concept of max being zero, counting down to represent any smaller value with negative numbers
Per Agent Eagle "The relative volume scale is primarily used for home theater. 0db is considered ‘reference’ level, which means when calibrated correctly, each speaker will be playing at 85db at ‘0’, allowing for peaks of 115db." This is what Paul should have said. Each +10 dB is perceived as twice as loud. + 3dB takes 2x the power from your amp to achieve, and +10 dB is about 8x the power. How loud is 85 dB? Well about the same as at a crowded restaurant where the collective voices are very loud, and you almost have to shout to talk to anyone else at your table. Normal room conversation at home is in the 60-65 dB range. Comfortable listening of your stereo is in the 75 dB area. Loud is 85-90 dB (and brief peaks can be more). OSHA workplace standards are don't exceed 90 dB constant for long periods of time. Rock concerts can be 100+ dB, or standing by large speakers at a wedding. Jack hammers up close around 110 dB. Speaker efficiency is measured at 2.83 v into 8 Ohms nominal (1 watt) at one meter. Most speakers are 85dB - 90dB at that level, so the excess watts in a high-powered amp are for really loud levels and/or peak levels. There are free dB apps you can put on your phone and you easily see what the sound level is wherever you choose.
You're talking about something different. The question was specifically asked about volume knobs, and in that case 0dB on the knob is not 85dB of sound pressure level. 0dB on the knob is "max volume" as Paul said. ..however loud that happens to be... 85 dB, 100 dB, 130 dB, or however loud it can go.
Most receivers and preamps will let you pic which way you want. Relative vs absoute.
We measure our perreaux amplifier at one loudness level in our apartment living room area. Volume at Eviction level!
It's just a scale. Any value can be scaled , or normalised, to another range of values. It's kind of arbitrary.
The dial is showing the gain margin of the amplifier. The gain margin is found in the frequency response function of the system. This is a measure of the amount that the gain can increased until the system becomes unstable. This comes from control theory and the negative sign is because the characteristic equation is in the denominator. A dB is by definition 20 log10(|g(iw)h(jw)|. thus any value will be negative because the expression for the characteristic equation is in the denominator. The use of the bode is because the frequency response of the amp can be obtain experimentally. When you increase the gain the frequency response moves up and the gain margin is reduced. When zero is reached the system is considered unstable thus its max amount that the amp can produce. To better understand this look at some youtube videos on bode diagrams hope this helps. it not a mystic once you understand the theory.
Back in the day, the party phrase was “turn it up to 10”. Perhaps not scientific, but everybody knew what it meant, and that it was LOUD.
People have to understand, that db on its own is just an algorithm, the letter behind it makes it a unit. And db - only occurs on electrical db‘s like db m or db u. If you use acoustic db measurements like db SPL ,there is no minus.
dB is not a unit. It is a unit-less value. The letter you are referring to specifies what the reference 0dB value is, and is sometimes also the units for that reference value. A dB value, like -3dB is unit-less itself. It is just a ratio like 2/3 or 27/8. The units come from whatever units the reference value has, but the actual dB value itself is unit-less.
> "And db - only occurs on electrical db‘s like db m or db u. If you use acoustic db measurements like db SPL ,there is no minus."
There is *always* a minus. The reference value can never be 0, so any value you tell me, I can always tell you a value that is half that. Therefore, there is *always* a minus.
I guess that's why most, if not all, volume potentiometers are of the LOG value type rather than LINEAR.
human hearing perceive sound level changes in a log scale
Ok then but, if at 0 DB an amplifier puts out its maximum power, and different amplifiers have different power levels (some are weaker, some are beefier), that means that 0 dB will translate to a different output (loudness measured in dB SPL) on different power amplifiers. And if I am not mistaken that also means that different (=rated at different wattage) power amplifiers will have different output given the same input (say -12 dB) from a preamp (because that's relative to 0 dB). But I also remember what you said about amplifier gain: two different power amplifiers can have different wattage specs and yet have the same gain and produce the same loudness at a given level of the input. This confuses me 🤔🤔
I'll know better in the mourning but I think it's due to the resistance aka the ohms that the speaker is asking from the amp. With the same GAIN 1:15 which might be 2v in and 30v out (for example) a speaker while at 4ohms will draw more watts
The 3 variables which can change the equation tied to one another. Watts = Voltage * Current
Less resistance means more current
Power isn't that interesting
dB is a relative scale, as you understand 0dB can be defined as any value, for power measurement, it can be 1mW (= 0dBm) or 1W (= 0dBW = 30 dBm) or any other baseline value
think of gain as a multiplier and wattage is the hard upper limit of power output before bad things (noise, distortion, clipping, overheat, reduced lifespan, burn out, sparks, smoke, fire, spontaneous combustion, armageddon) happen
high and low is all relative
a single ant can carry a lot of weight relative to its own weight (high gain) but it certainly won't be able drag you away in sleep (limited power output)
a healthy person likely cannot carry multiple times of his/her own weight (low gain) but should be able to get up from bed and walk around unassisted (that is higher power than the single ant)
Great explanation! Thanks, Paul. As a product developer for my whole career, I look back on the electro-mechanical dB levels we used to measure when developing power tools. Manufacturers always recommend wearing hearing (And eye) protection with power tools due to the sound power level of either the tool or the work it is doing with your ear generally at arms length or less. You often hear references to live rock concerts and jet engines as being 100+ dB levels. I assume this kind of measurement probably relates back to voltage levels in the electronic world in some way.
OdB is Unity Gain. There's the term to look up, if anyone is interested.
yes 20log10(1)=0. In control theory the system is unstable when 1+g(jw)h(jw)=0. for this to occur the magnitude of g(jw)h(jw)must be 1 that why you have the zero. The angle is one 180 but the term is in the denominator thus it is -180 degrees. you are working with complex numbers. This scale is coming from log plots of the sinusoidal transfer function. That is why in old audio that recorded when you get pass zero dB it usually was red (unstable)
@@frederickajustagosto4765 In this case though, Unity Gain, we're talking about what the attenuator (volume control) is doing to the signal, which is this case, is nothing. Zero on any kind of level meter vs. zero on attenuator (fader, pot) are not interchangeable terms, for those following along at home.
@@carterwilliamhumphrey3373 look at the definition of a dB unity is a value of one its log10 is 0 10^0=1 that is why its zero in dB. it i using log scale. saying the same thing 1. the word unity means 1. note he says that zero is what will run amp to loudest point. people confuse log with linear. You can't have the log of zero try it in your calculator.
@@frederickajustagosto4765 We seem to be deep into the math here, and I don't know why. If you want to go into an amp circuit, a gain factor of 1 is, in fact. unity gain of the amplifier, meaning: what goes in level-wise, is what comes out. Similarly, if you put an audio fader at "0", you get the same signal level on output as input.
Knob is an attenuator, not an amplifier
To complicate things: 0 dB sound pressure level is absolute silence.
In the real world, 0dB is the same as 0 sound. Why not use 0dB and above?
The lowest decibel a human can hear is 0 dB. In some exceptional cases, humans can hear sounds down to -15 dB. 0 dB sounds are very soft and barely perceptible by the human ear. To understand just how soft, imagine that a whisper is 30 dB and 30 dB is 1,000 times louder than a 0 dB sound.
Measurements of a speaker with a microphone are never measured in
- dB. Minus dB
The ear hears 10 dB as twice as loud, but it is 3 dB which is actually twice as loud, even if the ear does not perceive 3 dB as twice as loud
because the PCM recording system has an absolute upper max limit, after that point everything is recorded with the same data value, imagine clipping waveform
now I'm more confused then ever
Thank you,
May be you should use some figures and diagrams to clarify your points more deeply.
All I know is I have never come close to turning my system up to 0 db
My favorite color is ham
What is that; a reddish pink. Before this came on I was reading a book of poetry; List To The Warm by Rod McKuen.
This was not a complete explanation. A key piece of info left out is that a decibel value is a RATIO. It is always a ratio between some "reference" quantity and the quantity you're trying to describe. And it's a logarithmic scale. If it's voltages being referred to, then dropping 6dB is a halving of the voltage, and increasing by 6dB is a doubling of the voltage. If it's power, then 3dB drop is a halving of the power, and a 3dB increase is a doubling of the power. It is always a measure of the ratio between the reference value and the value you're trying to describe. What is the reference value? It's different for various things. You'll have to look that up. For sound, it's some extremely low sound pressure level at some specific distance from the sound source.
For amplifier volume controls, the chosen reference level is "volume knob at full volume". Nothing else makes sense other than "volume knob at zero", but the reference value can't be zero, because anything multiplied by 0 is still 0, so you cannot have a ratio with zero as one of the values. Hence, 0dB was made to be "full volume" for amps that show the volume as dB. 0dB is also full volume for recording (as seen on VU meters), so there was already some precedent set for 0dB to mean "full volume".
How many +dB is 11? 😜
if you define 11 as 0dB, then 11=0dB
if you define 1 as 0dB, then 11 = 10×log(11/1) dB ~= +10.4dB
It would make more sense if ‘0’ db was %100 db
More is better!
What ?.0 db is .77 volts consumer level and 1.23 volts at pro levels , no ? And that is not the loudest , because of headroom ....right ? Isn't the headroom like +22 db on pro level equipment ? I think I'm now more confused than ever regarding decibels ! 😛
You, we are confused because unit is incomplete. The same happens with speaker's sensitivity, reviewers take shortcuts with the definition so the conclusion isn't accurate. Bel is a logarithmic relation between a value and it's reference. That reference must be defined in the «unit». Two analog references: 0 dBu = 0,7746 V RMS and 0 dBV = 1 V RMS. In the digital domain we have 0 dBFS that is the max level. Some useful examples: 0,3162 V RMS = -10 dBV RMS ; 1,228 V RMS = 4 dBu RMS. How do we get these numbers: 20×log(0,3162 V/1 V) = -10 dBV with means that a line tension of 0,3162 V relative to the reference V is -10 dBV, the second example 20×log(1,228 V/0,7746 V) = 4 dBu with means that a line tension of 1,228V relative to the reference «u» is 4 dBV. The full unit is defined by the combination of dB an d the reference (V, u, FS, SPL[A,B,C], etc).
the ~0.775V figure comes from result voltage of 1mW across 600 Ohm, since then sqrt(0.6)V became the baseline value of 0dBu, and +4dBu is the analogue audio level standard, if a device can handle max 24dBu then it is said to have 20dB headroom (from the nominal level +4dBu)
"Consumer devices" commonly uses dBV units, 0dBV = 1V
so dBu and dBV both quantify voltage with logs, but at different scale, like Celsius and Fahrenheit in log scale
Thanks ! @@Jorge-Fernandez-Lopez
@philiptong4978 Thanks !!
@@philiptong4978Thanks !
There are snippets of truth all around here, but also a lot of confusion. Some is from making no distinction between the following different uses of the term decibel:
(A) "Absolute" decibels, which are computed as 10 times the Base 10 logarithm of the ratio of a power level under discussion to a REFERENCE power level, which MUST ALSO be specified. [In the general sense, all dB numbers are just numeric ratios, with no units. But in the context of these kinds of dBs, since the reference signal's absolute level and unit(s) are specified, the level and unit(s) for a given signal are also known absolutely.]
E.g., dBW, where the "W" refers to 1 watt, or dBm, where the "m" refers to 1 milliwatt. E.g., 1 milliwatt = 0 dBm, or -30 dBW; 2 milliwatts = +3 dBm; 10 milliwatts = +10 dBm; 400 watts = 26 dBW, or 56 dBm. The math is TEN times the log ratio because we're talking about "deci-bels", a tenth of a bel, not "bels". [Let's agree to not talk about bels any further - but if we did, the math would be just the Base 10 log of the ratio instead of ten times that].
We use logarithms because our hearing of different loudness levels correlates to a logarithmic "compression" scale of power ratios, NOT their simple linear ratios. Twice the power level is not perceived as being twice as loud - it only sounds about 30% louder. It takes TEN TIMES the power to sound twice as loud. This is behind the reason why, e.g., a 2-watt amplifier can play music fairly loudly 95% of the time with many speakers - but you might need 100 watts for the peak levels just a few decibels louder. And why a 100W amp can only play 3 dB louder than a 50W amp [and why I wish that all amps had a dBW rating as well as a wattage rating]. So we use Base 10 logarithmic ratios in the form of decibels to best correlate to the differences that we perceive between different power levels (whether it's the acoustic power of vibrating air molecules or electrical power in or out of an audio device). It was the ancient telephone industry that adopted this practice; technically, 0 dBm is 1 milliwatt into 300 ohms; but we usually gloss over that specific impedance level in our business, as we have our own common pseudo-standards, e.g., < 1000 ohms for line-level outputs, >= 10k ohms for line-level inputs.
(B) A common special case of of (A) is dB SPL, where SPL - special because the reference is often not explicitly referenced, just assumed from the context. SPL refers to the specific acoustic reference "Sound Pressure Level" that most people with reasonably good hearing can just barely hear in air at 1 kHz - "the threshold of human hearing". That standard reference level is 0.00002 Pascals (Newtons/square meter - Root Mean Squared [RMS] levels, to be specific) of sound PRESSURE (NOT power, hence the math demands TWENTY times the Base 10 logarithmic ratio for these decibels - see more on this in Item C below). Sound AT THAT REFERENCE LEVEL is therefore expressed as 0 dB SPL - or often just 0 dB for short. This is the scale that we use to express levels up to, say, 40 dB [SPL] as quiet, 80 as noisy, 100 as symphony orchestra crescendo loud, 120 as painful, etc. These are sound pressure levels in air, which have NO ABSOLUTE DIRECT correlation to decibels of watts or volts or amps in electronic equipment producing such sound levels UNLESS much additional information is taken into account, such as loudspeaker efficiency (acoustic watts output for a given electrical power input), distance from the speaker, the spatial directivity pattern of the speaker, the direction of measurement relative to the speaker, and whether the measurement is performed in open 3D space or an approximation of it (such as a practical anechoic chamber, or a time-windowed measurement to simulate open space by ignoring room reflections) or actually including a room's characteristics and its sound reflections.
(C) Common "throw-around" (my term) decibels, where the reference level is WHATEVER IT IS in the understood context [e.g., 10 millivolts out of a phono cartridge, or 2 volts out of a DAC, or 1 watt out of a power amp, or whatever], and the actual level is then expressed as a number of dB above (+) or below (-) THAT level. IMPORTANT DIFFERENCE: If voltages (or currents or SPL) are being compared instead of power, then the math demands that the decibel numbers be calculated as TWENTY TIMES (not ten times) the Base 10 logarithmic ratios. This is because power in watts is proportional to the voltage or current SQUARED, and the logarithm of a number squared is double the log of the original number. A 2:1 voltage ratio = 4 times the power ratio = +6 dB in power.]
The subject is not terribly confusing to electrical engineers or technicians, but since decibels are all about numbers, even a minimal clear and cogent explanation does require a bit of math. I hope this helps a bit.
Then why continue the old norm of calling it a pre "amp" :)....confusing..
A preamp with a phono cartridge or microphone input DOES unequivocally amplify those signals quite a bit (but enough to drive loudspeakers). If it has a moving coil phono cartridge input, it will amplify it even MORE. MOST preamps can also amplify line-level inputs by a few decibels (perhaps 6 or so). Hence the norm of calling all such devices preamps. We could just stop there. // All preamps designed for regular human interaction also include a volume control of some type to attenuate its otherwise fixed line-level output that is designed to drive a separate amplifier that is capable of yet more amplification and much greater output power to drive low-impedance devices like loudspeakers. So-called "passive preamps" (of which there are few examples) are the truly confusing case, because regardless of what else they may do that most preamps do (such as source selection and attenuation/volume control), they by definition do NOT amplify a signal at all in the classic sense (i.e., with active devices like transistors or tubes) - only pass it as-is, or attenuate it. Some may include a transformer to boost output voltage, but that is not amplification - any such voltage boost comes at the cost of higher output impedance and less current.
Nuh ! still don't get it , but if I just stick to using
whatever you want to call the knob that controls the
(what I call) volume, and have it at (my idea of) a
reasonable level ,it🤞👍🤞🙏 shouldn't matter
if I do or don't understand all that.
This isn't a great explanation. It's simply that when you adjust the volume on your amplifier, you're not actually changing the amplifier itself - in essence the amplifier is always boosting the signal at the maximum/fixed level.
Instead, you are just altering a resistor before the amplifier in order that you don't only have it playing at the maximum volume.
So the negative dB are referring to the reduction before the amplifier - how much that resistor is pulling down the sound before it gets boosted by the fixed output of your amplifier.
Turning the volume all the way up to "0" correctly implies that that there is now 0 db of attenuation (resistance).
"Instead, you are just altering a resistor before the amplifier " - there are different types of amplifiers.
Not all control volumes work that way by attenuation with a resistance. Some adjust the gain stage.
Hence you can have a passive preamplifier, no power just passive circuitry that attenuates the volume.
the amplifier acts like a multiplier. if its gain is fixed the volume you put in is multiplied by its gain (increases). however you can change the gain on some amps that another story. but this is showing the gain margin of the amp. however I believe the manufacturer is the one that puts the scale thinking it will be the best way to sell the amp at the end that's what they do
My amp goes to 11!
No one told me there would be math.
Welcome to the digital Era
analogue electronics has math, music theory has a reasonable amount of math too
Paul just casually doxxing one of his staff at the start of the vid. Un-DeciBEL-ieveable
it should go to 11 not a hundred😂
These explanations are why most EE's go to University. Engineering ain"t easy. Good job Paul.
Kudos to Paul for explaining dB without mentioning logarithms. With a background in physics logarithms make complete sense to me, but I know most people's eyes would glaze over at the mention of logarithms.
Of course a more complete understanding of the dB scale does require knowledge of logarithms; e.g. it explains why voltage ratios vs. power ratios are a factor 2 apart in terms of dB.
Good explanation but, its still stupid to put that in a integrated or pre amp. Just give us 1-100
easy to say i respect more the person who designed it. They put that in the volume to communicate the way they converse in dB. May be it better to keep learning and
@@frederickajustagosto4765 "may be it better" to just put 1-100 and shut the fuck up