Speaker Impedance - please confirm my interpretation

[DocG]

Devialet 120 and Devialet 200 from what I've seen

The test results for the 120 are here. And here's a quote from the article:

tested with the latest v7.1 firmware installed, this 120adH amplifier met its 120W/6ohm specification with a full 2x75W/8ohm, doubling to 2x 150W/4ohm – this power achieved with a peak level (0dbfs) digital input and the volume control set at +0.5db (not 0db). the output impedance is vanishingly low at <2mohm. distortion at this output is still a fabulously low 0.0005% through bass and midrange,a performance largely maintained over the top 40db of its dynamic range (from 75W down to just 7.7mW).the a-wtd s/n ratio is also a huge 116.8db at this level. distortion does increase with frequency, however, particularly above 10kHz, reaching 0.018% at 20kHz/1W/8ohm. the harmonics at this level are inaudible, of course. the response is flat out to +0.08db/20kHz with 44.1/48kHz media, –0.1db/40kHz with 96kHz files and –4db/90kHz with 192kHz files. as with earlier devialet amps, jitter is incredibly low – just 24psec at 10W output with 24-bit/48kHz media. PM

 

[TrevC]

Of course one could legitimately quote double the power output at 4 ohms simply by derating the 8 ohm value.

 

[NJB]

...if I want to form a shortlist of speakers that play relatively quietly, given that the amplifier is not a variable, then speakers with a lower sensitivity and a higher resistance should be a better bet. There are enough other factors involved to mean that this might not be true, but it is a starting point.

 

[hg]

...if I want to form a shortlist of speakers that play relatively quietly, given that the amplifier is not a variable, then speakers with a lower sensitivity and a higher resistance should be a better bet. There are enough other factors involved to mean that this might not be true, but it is a starting point.

The sensitivity should tell you how loud it will play but sometimes you need other data to compare sensitivities that are quoted in different ways. What is relevant when trying to decode a sensitivity of X dB:

1) The distance of the microphone from the speaker and it's orientation relative to the driver/s.

2) If the value is for a single frequency or an average over a range of frequencies.

3) If the speaker is radiating into full space (i.e. an anechoic chamber), half space (i.e. flat on a wall) or what.

4) If the input is a specified voltage or a specified power.

5) The type of input: noise or tones and how filtered.

6) If the manufacturer is making up a large number to appeal to people with valve amplifers (e.g. Audio Note, Zu)

 

[SteveR750]

...if I want to form a shortlist of speakers that play relatively quietly, given that the amplifier is not a variable, then speakers with a lower sensitivity and a higher resistance should be a better bet. There are enough other factors involved to mean that this might not be true, but it is a starting point.

The sensitivity should tell you how loud it will play but sometimes you need other data to compare sensitivities that are quoted in different ways. What is relevant when trying to decode a sensitivity of X dB:

1) The distance of the microphone from the speaker and it's orientation relative to the driver/s.

2) If the value is for a single frequency or an average over a range of frequencies.

3) If the speaker is radiating into full space (i.e. an anechoic chamber), half space (i.e. flat on a wall) or what.

4) If the input is a specified voltage or a specified power.

5) The type of input: noise or tones and how filtered.

6) If the manufacturer is making up a large number to appeal to people with valve amplifers (e.g. Audio Note, Zu)

You're right, but perhaps a little complicated to follow as a rule of thumb that the OP was after perhaps.

As a simple guide, If you want "quiet" speakers, buy ones that have a higher quoted impedance. The advantage here is that they won't demand too much from your amp. Additionally, look for low sensitivity (the standard definition is sound pressure level measured at 1W input continous sine wave input at 1m distance from the drive unit in an anechoic chamber). Bear in mind if you do want to listen to it loud, you will need an amplifier with a higher rated power (watts) output.

 

[lindsayt]

Whilst this is correct in theory, power supply limitations in budget amplifiers will mean that this is not the case in a real world system other than at very modest output levels.

OK let's take one example. My Creek CAS 4040 is a budget amplifer. Always has been. It was budget when it was new, budget 2nd hand. It produces 30 watts into 8 ohms before clipping. I'm not sure how much it can do into 4 ohms. Let's say 40 watts, to be conservative and to keep the maths simple.

Lets say we've 85 db for 2.83 volts at 1 metre efficient 4 ohm speakers. So that means we've got 2 watts producing 85 dbs at 1 metre per speaker. 20 watts would produce 95 dbs. 40 watts 98 dbs.

For the stereo pair that would be 101 dbs at 1 metre. Let's say we're listening at 3 metres range. That would give us about 90 dbs at the listening position. That's loud. Loud enough to seriously annoy any nearby neighbours. For compressed music, that's loud enough to risk damaging your hearing with a prolonged listening session.

Varous measurements I've made with a calibrated sound pressure meter indicate that the vast majority of listeners, spend the vast majority of their time listening at levels where the highest transient peaks reach less than 86 dbs.

Listening to music where the peaks are hitting 75 to 85 dbs are what I'd call generous volumes. Volumes where the music has physical impact and where it's easy to hear everything that's going on in the mix. There's no way I'd describe them as very modest output levels.

Of course, if you have insensitive hearing, 85 db peaks may well sound modest to your ears. Which is fine and not something I'd argue with. However I would urge anyone with properly working hearing to treat the recommendations of anyone with insensitive hearing with extreme caution due to the probability of them having rather different shaped Fletcher Munson curves.

 

[hg]

Lets say we've 85 db for 2.83 volts at 1 metre efficient 4 ohm speakers. So that means we've got 2 watts producing 85 dbs at 1 metre per speaker. 20 watts would produce 95 dbs. 40 watts 98 dbs.

For the stereo pair that would be 101 dbs at 1 metre.

Playing mono but not stereo where the energy might be moved to the right or left dropping us back to 98 dB.

Let's say we're listening at 3 metres range. That would give us about 90 dbs at the listening position.

That's loud. Loud enough to seriously annoy any nearby neighbours. For compressed music, that's loud enough to risk damaging your hearing with a prolonged listening session.

For stereo we have 87 dB at 3m. This is a safe level for typical home use. Damage starts at about 85 dB average, 8 hours a day, every day. Less exposure means the level can be louder without damage.

If we have compressed music with 12 dB peak to average then the average level needs to be 75 dB to avoid clipping the peaks at 87 dB. That is not loud for listening to music, it is about 10 dB below standard levels. On the quieter side of reasonable. Unfortunately uncompressed music requires a headroom of more like 20 dB in order to rarely clip peaks giving an average level of 67 dB which is too quiet to hear the music as intended.

Listening to music where the peaks are hitting 75 to 85 dbs are what I'd call generous volumes. Volumes where the music has physical impact and where it's easy to hear everything that's going on in the mix. There's no way I'd describe them as very modest output levels.

So that is an average level of 55 to 65 dB for uncompressed music. This is around the level for conversation and background music?

Of course, if you have insensitive hearing, 85 db peaks may well sound modest to your ears. Which is fine and not something I'd argue with. However I would urge anyone with properly working hearing to treat the recommendations of anyone with insensitive hearing with extreme caution due to the probability of them having rather different shaped Fletcher Munson curves.

The Fletcher Munson curves shows why listening to music at low levels gives an incorrect tonal balance with the bass and treble being reduced. As a result a person needs to listen to music at the intended level in order to hear the intended tonal balance. In a cinema this level is set to 85dB average with 105 dB peaks (and no this does not damage your hearing because you are not exposed 8 hours a day every day). The music industry as a whole is largely indifferent to quality and has never bothered to set a reference level but music engineers work close to the cinema levels. Some a bit quieter, some a bit louder (but they will damage their hearing because of the long exposure). The quietest viable level is going to be perhaps 75 dB average and 95 dB peaks because the Fletcher Munsion curves are kicking in at this level.

 

[lindsayt]

hg, have you ever measured how loud you typically listen to your hi-fi with a calibrated sound meter? And how sensitive are your speakers and how far do you sit from them?

If you're listening at levels where the sound meter drops to about 75 dbs and peaks at 87 dbs and you consider this "On the quieter side of reasonable" for listening to music at home, then it would seem that your ears are rather insensitive, or you have unusual tastes when it comes to how loud you like to listen to your hi-fi.

And on the Fletcher Munson thing. There's no way that everybody has the same shaped Fletcher Munson curve. As the hearing deteriorates, some frequencies will be affected more than others. So, what would sound tonally neutral to somebody with insensitive hearing would likely sound too bright to someone with good hearing.

And yes, most the time when I'm listening to music at home, someone can come in the room and talk to me in their normal voice and I will be able to hear them fine over the music.

And speakers tend to be less efficient in the bass than in the midrange or treble. They also tend to have their lowest impedance there. Bass drums and bass guitars are almost always mixed to be in the centre of the soundstage. So, when it comes to clipping, it's fine to consider the output of both speakers.

And I haven't even mentioned that a budget amplifier like my Creek CAS4040 will deliver about 40 watts into 4 ohms with a continuous sine wave, but can deliver more than that as brief transient peaks without clipping.

 

[hg]

If you're listening at levels where the sound meter drops to about 75 dbs and peaks at 87 dbs and you consider this "On the quieter side of reasonable" for listening to music at home, then it would seem that your ears are rather insensitive, or you have unusual tastes when it comes to how loud you like to listen to your hi-fi.

I listen at roughly the same level as a cinema, the same level as music engineers and, more importantly, the level that music sounds tonally balanced. It is the standard/normal level for listening to music because quieter levels are perceived as tonally incorrect. I would not expect anybody with a reasonable interest in music to listen at any other level although it is becoming clear to me that many of the audiophiles on this forum do listen to their inadequately sized speakers at levels well below the standard.

A while back I moved into a temporary flat where playing music at standard levels would cause problems for the neighbours. Rather than listen to music at unsatisfactory TV levels I forced myself to adapt to the unnaturalness of headphone listening as being the lesser of two evils.

At home I have an uncalibrated sound level meter and a calibrated(ish) capacitor microphone (not useable for legal noise measurements but has a calibration curve). I usually have access to properly calibrated equipment via the day job although not at the moment.

And yes, most the time when I'm listening to music at home, someone can come in the room and talk to me in their normal voice and I will be able to hear them fine over the music.

This appears to be too quiet to perceive music with the intended tonal balance. If it is your preference then no problems so long as you make no claims to it being in some way correct. It may also be normal among your peers which I would not have expected but now think might be the case. The whole notion of high fidelity really does seem to have disappeared.

 

[SteveR750]

If you're listening at levels where the sound meter drops to about 75 dbs and peaks at 87 dbs and you consider this "On the quieter side of reasonable" for listening to music at home, then it would seem that your ears are rather insensitive, or you have unusual tastes when it comes to how loud you like to listen to your hi-fi.

I listen at roughly the same level as a cinema, the same level as music engineers and, more importantly, the level that music sounds tonally balanced. It is the standard/normal level for listening to music because quieter levels are perceived as tonally incorrect. I would not expect anybody with a reasonable interest in music to listen at any other level although it is becoming clear to me that many of the audiophiles on this forum do listen to their inadequately sized speakers at levels well below the standard.

A while back I moved into a temporary flat where playing music at standard levels would cause problems for the neighbours. Rather than listen to music at unsatisfactory TV levels I forced myself to adapt to the unnaturalness of headphone listening as being the lesser of two evils.

At home I have an uncalibrated sound level meter and a calibrated(ish) capacitor microphone (not useable for legal noise measurements but has a calibration curve). I usually have access to properly calibrated equipment via the day job although not at the moment.

And yes, most the time when I'm listening to music at home, someone can come in the room and talk to me in their normal voice and I will be able to hear them fine over the music.

This appears to be too quiet to perceive music with the intended tonal balance. If it is your preference then no problems so long as you make no claims to it being in some way correct. It may also be normal among your peers which I would not have expected but now think might be the case. The whole notion of high fidelity really does seem to have disappeared.

I have the same opinion on listeing levels. One of the expctations of my "hi-fi" is that it plays instruments at a similar volume and tonality as the real thing. Try playing a drum kit in your living room, you certainly cannot talk over it.

On Sunday, I had the pleasure of an afternoon listeing to a live chamber orchestra, and there is no way you could hold a conversation over anything other the quietest passages, so it's not peculiar to modern / rock music.

 

[Covenanter]

If you're listening at levels where the sound meter drops to about 75 dbs and peaks at 87 dbs and you consider this "On the quieter side of reasonable" for listening to music at home, then it would seem that your ears are rather insensitive, or you have unusual tastes when it comes to how loud you like to listen to your hi-fi.

I listen at roughly the same level as a cinema, the same level as music engineers and, more importantly, the level that music sounds tonally balanced. It is the standard/normal level for listening to music because quieter levels are perceived as tonally incorrect. I would not expect anybody with a reasonable interest in music to listen at any other level although it is becoming clear to me that many of the audiophiles on this forum do listen to their inadequately sized speakers at levels well below the standard.

A while back I moved into a temporary flat where playing music at standard levels would cause problems for the neighbours. Rather than listen to music at unsatisfactory TV levels I forced myself to adapt to the unnaturalness of headphone listening as being the lesser of two evils.

At home I have an uncalibrated sound level meter and a calibrated(ish) capacitor microphone (not useable for legal noise measurements but has a calibration curve). I usually have access to properly calibrated equipment via the day job although not at the moment.

And yes, most the time when I'm listening to music at home, someone can come in the room and talk to me in their normal voice and I will be able to hear them fine over the music.

This appears to be too quiet to perceive music with the intended tonal balance. If it is your preference then no problems so long as you make no claims to it being in some way correct. It may also be normal among your peers which I would not have expected but now think might be the case. The whole notion of high fidelity really does seem to have disappeared.

You must live in some strange world if you think any but the smallest minority live somewhere where it would be possible to listen at concert hall levels. Your arrogant dismissal of everybody else is just that - arrogant.

Chris

 

[hg]

You must live in some strange world if you think any but the smallest minority live somewhere where it would be possible to listen at concert hall levels.

Concert levels are significantly higher than the standard level of 85dB average with peaks of 105dB. Perhaps 10 - 20 dB higher and will require serious studio speakers to meet with high fidelity. PA speakers can give the level but usually not the fidelity.

 

[pauln]

I listen at roughly the same level as a cinema, the same level as music engineers and, more importantly, the level that music sounds tonally balanced. It is the standard/normal level for listening to music because quieter levels are perceived as tonally incorrect. I would not expect anybody with a reasonable interest in music to listen at any other level although it is becoming clear to me that many of the audiophiles on this forum do listen to their inadequately sized speakers at levels well below the standard..

And what is it exactly that makes you the arbiter of listening to music? I haven't read so much arrogant bo11ocks for a long time. Does that imply that all the people that want to chill a bit or don't want to get their ears constantly blasted or maybe can't afford big speakers don't have an appreciation of, or interest in music?

And despite what people may tell you, size does matter - just not when it comes to speakers, then it's quality that counts.

 

[Vladimir]

I'm in the SPL camp. I have never gone to a music performance event to listen to finer details, timbre and imaging. I am blown away by larger-than-life SPL first, then lack of distortion, then ambiental effects. I've even stopped listening to anything else except those 3 factors in Hi-Fi. If it cant go loud, clean and portray ambience of a live event, I'm not convinced. Might as well play music on my PC monitor (which I do). I stopped listening for imaging, soundstaging, 3D, timbre etc. since I moved on from Aiwa mini systems to separates (20 years ago). I recognize tonal balance as important but not a deal-breaker like power compression at 90dB.

I had my fair share of bokshelf speakers (mini monitors). Doesn't do it for me, neighbours or no neighbours. I'd rather rock out on my headphones to SPL that maintains that larger-than-life effect for me.

This paradigm initially converted me to an objectivist. I simply stopped listening for what wasn't there and just want more of what's there. Bird in the hand sort of thing...

 

[hg]

And what is it exactly that makes you the arbiter of listening to music?

I am afraid the nonlinearity of our hearing was established long ago by science not me. In the days of high fidelity before the audiophile thing kicked off in the 70s most amplifers would have a "loudness" button which boosted the bass and treble in an attempt to compensate when listening at quieter levels than standard. But these days audiophiles seem to have no interest in high fidelity or perhaps they deny reality. It can be difficult to tell.

 

[pauln]

And what is it exactly that makes you the arbiter of listening to music?

I am afraid the nonlinearity of our hearing was established long ago by science not me. In the days of high fidelity before the audiophile thing kicked off in the 70s most amplifers would have a "loudness" button which boosted the bass and treble in an attempt to compensate when listening at quieter levels than standard. But these days audiophiles seem to have no interest in high fidelity or perhaps they deny reality. It can be difficult to tell.

I'm aware of Fletcher Munson, or equal loudness curves. As has been pointed out, many music lovers are not in the enviable position that you seem to be in where they can afford the equipment and the detached house with grounds that puts the neighbours out of earshot and have no children asleep upstairs or wife watching TV in the next room but they still want to listen to music. Should they just not bother because the tonality is compromised? Perhaps read a book instead and leave the music listening to those that can truly appreciate it?

Maybe this is the reason that so many consumer speakers and headphones have a smiley frequency response?

 

[hg]

As has been pointed out, many music lovers are not in the enviable position that you seem to be in where they can afford the equipment and the detached house with grounds that puts the neighbours out of earshot and have no children asleep upstairs or wife watching TV in the next room but they still want to listen to music. Should they just not bother because the tonality is compromised?

That is a choice for them. As mentioned above, when I could not play music through speakers at standard levels the lesser of two evils for me was to use headphones at standard levels rather than listen quietly to speakers. Before I became accustomed to listening at standard levels I would fairly happily listen to music at quieter levels (young and poor) but I was more easily distracted and the focus was not the same. Having now got used to something better going back to my old quieter speakers would not bring the same level of enjoyment. Which is a pity but that's how we seem to be made. There is a name for this effect which I cannot recall. Anyone?

 

[Vladimir]

Please remind us what are your big boy speakers for critical listening.

 

[hg]

Please remind us what are your big boy speakers for critical listening.

I haven't said but they are DIY speakers of my own design. In the context of this discussion they use two 8" woofers but the cabinets, although inevitably not small, are not big boys. Not sure I do much critical listening with it almost always being for pleasure.

 

[Vladimir]

Please remind us what are your big boy speakers for critical listening.

I haven't said but they are DIY speakers of my own design. In the context of this discussion they use two 8" woofers but the cabinets, although inevitably not small, are not big boys. Not sure I do much critical listening with it almost always being for pleasure.

I'm currious which drivers are you using for your speakers and are they sealed or ported cabinets?

 

[lindsayt]

I think my musical tastes are fairly typical of most people in this forum in that I mostly listen to studio recorded rock and pop music, with some live recordings and a wide smattering of other genres.

Here's the thing. If you want to replay rock or pop studio recordings, what actual level represents a 1 to 1 ratio of being in the room when the performance was done?

Do you go by the volume the vocalist would be singing at if they were standing in your room? In that case your listening levels would be similar to mine - about 70 dbs (with all these figures it depends if you want to recreate front row / middle of the room or rear of the room or concert venue).

Do you go by the volume of the piano? About 70 to 80 dbs.

Do you go by the volume of the drumkit? 90 to 100 dbs.

Or do you go by the volume of the band when you saw them live? Again, front row, right in front of the speaker stack, middle of the venue or rear? In which case you could be looking at 100 to 120 or more dbs.

The point being, that for this type of music, each instrument, plus the vocals will have a different natural volume. And at a concert, some instruments are amplified more than others.

For me, in general, the most important aspect is the vocals. So I like to listen at a roughly 1 to 1 ratio for the vocalist volume. If it's late at night, or I just want semi background music it will be quieter than that. From time to time I will turn my system up. When I do, the drumkit will sound more realistic, but the vocals will sound unnaturally loud, in the same way that they do at most concerts.

 

[SteveR750]

I set the volume to whatever I think is about right for an unmic'd instrument. So, a drumkit, if it's in the mix. Vocals will often have to be amplified to compete.

Guitars always should be anyway whatever's going on, and louder than anyone else.

 

[andyjm]

I think my musical tastes are fairly typical of most people in this forum in that I mostly listen to studio recorded rock and pop music, with some live recordings and a wide smattering of other genres.

Here's the thing. If you want to replay rock or pop studio recordings, what actual level represents a 1 to 1 ratio of being in the room when the performance was done?

Having sat in a few studios during recording sessions, 'being in the room' sounds nothing like the finished version, particularly the relative loudness of intruments and vocals. While I don't think that aiming for vocals at real life levels is a bad thing to go for, don't think for a second that the resulting playback sounds much like the original session.

A more (?) interesting problem is for the classical music enthusiasts trying to recreate the real life experience of an orchestra in a domestic setting. This is where you need dynamic range.