Things I (still) don't Understand :)

[Jungstar]

I'm somewhat new. Only had a few powered speakers jbl 308 and a jbl sub. Rasberry pi hat dac streamer.

I read quite a lot here on Dacs and Speakers and I'm very happy that Amir (which I thought was a japanese lady for the 1st year) has created YouTube Channel and started to debunk myths and also explain more in detail how measurements are done, very impressed. I think his measurement process deserves a video in itself. I'm a big believer in objectivism and science, yet as an economist my electronics and physics level is not rusty, it never grew beyond high school. So how do Amir measure:

Flavour of Sound? Many talk about a warm or cold speaker or a dark or bright speaker? Full body? Can we measure that?

Sound Pressure speakers might both have good values in measurements, but I would assume a larger driver area creates a larger sound pressure without distortion. But some drivers move more. Can I quantify that with watt and sensitivity?

Sinad vs Dynamic Range? what is the difference? Something about volume... And also, what's the limit for when we can actually hear noise it?

Good Low Volume Speaker? Can I see from the charts if a speaker sounds nice at low volume? 95% of the time, I just have background music.

Tubes - are they supposed to be clean? I read that tubes (amp or preamp) on purpose add distortion (or flavor) to the midrange - does that conflict with the goal of minimizing noise and focusing on that?

I'm general all the measurements are quite complicated but much appreciated. I'm happy for some debunking like cables and Psu, that Hans Beekhuyzen, almost got me.

 

[dfuller]

Flavour of Sound? Many talk about a warm or cold speaker or a dark or bright speaker? Full body? Can we measure that?

Yes - that's frequency response!

Sinad vs Dynamic Range? what is the difference? Something about volume... And also, what's the limit for when we can actually hear noise it?

Dynamic Range is basically the noise floor vs the maximum signal level before clipping. SINAD is the inverse of THD+N - it measures SIgnal vs Noise And Distortion.

Tubes - are they supposed to be clean? I read that tubes (amp or preamp) on purpose add distortion (or flavor) to the midrange - does that conflict with the goal of minimizing noise and focusing on that?

Tubes are very much not clean - they're just lower fidelity devices in general than solid state. One can design tube amps to be effectively transparent - but why bother with tubes if you're going to do that? Most tube devices these days are not designed to be super clean because there's basically no point to tubes without some sort of distortion.

Sound Pressure speakers might both have good values in measurements, but I would assume a larger driver area creates a larger sound pressure without distortion. But some drivers move more. Can I quantify that with watt and sensitivity?

Larger surface area for a woofer or mid woofer with all other things being equal equals lower distortion at a given SPL, but (in the case of mid-woofers) necessitates a lower crossover point for better directivity behavior (which can have other tradeoffs, namely higher tweeter distortion).

 

[Purité Audio]

An extremely hot curry will clean out your tubes.
Keith

 

[voodooless]

Sound Pressure speakers might both have good values in measurements, but I would assume a larger driver area creates a larger sound pressure without distortion. But some drivers move more. Can I quantify that with watt and sensitivity?

The sensitivity is usually in the measurement data. Though size of the drivers is only one of the many parameters. Same goes for distortion: a larger driver does not always distort less.

Good Low Volume Speaker?

Can I see from the charts if a speaker sounds nice at low volume? 95% of the time, I just have background music.

I guess you could look at the equal loudness curve, and find a speaker that sort of compensates for it. You can also do that with EQ, so no special speakers needed. I guess boosting lows and highs will do the trick

 

[Aerith Gainsborough]

Good Low Volume Speaker? Can I see from the charts if a speaker sounds nice at low volume? 95% of the time, I just have background music.

Generally speaking: if a speaker performs good in the measurements, it should sound good at low volume too, yes.

BUT: you need to factor in human hearing.
read this: https://en.wikipedia.org/wiki/Equal-loudness_contour

Bass and Treble response of our ears drop off a lot faster than the midrange when you lower the volume. So a nice, neutral speaker may sound perceptively worse (thin, muffled) than your typical bathtub frequency response speaker because the elevated bass and treble combat the Fletcher-Munson effect do some degree. Yet these sound horrible on higher volumes.

Naturally you can fix that via EQ but these implementations are somewhat inaccurate and temperamental.

 

[alanca3]

Sound Pressure: Generally speaking, the less a speaker moves, the lower the distortion generated. Sound pressure is quantified with dB SPL. A change in 6dB is readily apparent, while a change in 10dB sounds twice as loud. Normal conversation is around 60 dB.

The sensitivity tells us the dB-SPL output from 1 meter away when driven by 2.83 volts. This equals 1 watt into an 8 ohm load, or 2 watts into a 4 ohm load.

To increase the SPL of a loudspeaker by 3dB you must double the electrical power.


Speaker A is rated 93 dBSPL / 2.38v (1 watt)
Speaker B is rated 81 dBSPL / 2.38v (1 watt)

Speaker A only needs one watt to reach 93dB, while speaker B needs 16 watts.
(81 + 3 + 3 + 3 + 3 = 93 dbSPL)
(1 -> 2 -> 4 -> 8 -> 16 watts)

 

[Haint]

Generally speaking: if a speaker performs good in the measurements, it should sound good at low volume too, yes.

BUT: you need to factor in human hearing.
read this: https://en.wikipedia.org/wiki/Equal-loudness_contour

Bass and Treble response of our ears drop off a lot faster than the midrange when you lower the volume. So a nice, neutral speaker may sound perceptively worse (thin, muffled) than your typical bathtub frequency response speaker because the elevated bass and treble combat the Fletcher-Munson effect do some degree. Yet these sound horrible on higher volumes.

Naturally you can fix that via EQ but these implementations are somewhat inaccurate and temperamental.

@Jungstar

Audyssey's Dynamic EQ feature is a defacto Fletcher-Munson curve that scales itself with volume. The amount of boosts (adjustable with the offset value) is pretty much universally preferred in the bass (Vs. a flat target), but the treble boost is much more divisive (probably dependent on room and placement). I wish they would include an option to turn off the treble rise.

 

[dfuller]

The sensitivity is usually in the measurement data. Though size of the drivers is only one of the many parameters. Same goes for distortion: a larger driver does not always distort less.

This is an important point indeed. IIRC, the JBL 306 woofer shows better distortion behavior over parts of its range than the 308.

 

[ahofer]

@Jungstar

Audyssey's Dynamic EQ feature is a defacto Fletcher-Munson curve that scales itself with volume. The amount of boosts (adjustable with the offset value) is pretty much universally preferred in the bass (Vs. a flat target), but the treble boost is much more divisive (probably dependent on room and placement). I wish they would include an option to turn off the treble rise.

RME also has a dynamic loudness feature, as do the Buchardt speakers.

 

[Aerith Gainsborough]

@Jungstar

Audyssey's Dynamic EQ feature is a defacto Fletcher-Munson curve that scales itself with volume. The amount of boosts (adjustable with the offset value) is pretty much universally preferred in the bass (Vs. a flat target), but the treble boost is much more divisive (probably dependent on room and placement). I wish they would include an option to turn off the treble rise.

Correct, but: you still need to be able to measure SPL to set these dynamic EQ's properly during initial setup.

I used the 90dB Z-peak value as the reference point at which the Loudness compensation does nothing.

My RME can adjust both bass and treble gain, on headphones I prefer +3dB max treble instead of the factory default.

 

[levimax]

@Jungstar

Audyssey's Dynamic EQ feature is a defacto Fletcher-Munson curve that scales itself with volume. The amount of boosts (adjustable with the offset value) is pretty much universally preferred in the bass (Vs. a flat target), but the treble boost is much more divisive (probably dependent on room and placement). I wish they would include an option to turn off the treble rise.

For decades this was a solved problem with tone controls and low level listening is the most common use case for me.... too bad tone controls are out of style.

 

[rynberg]

All D&M and Anthem receivers have easily accessible tone controls; if you are talking 2-channel equipment only, that is a result of "audiophile" nuttery.

 

[Loathecliff]

Full body?

NB, that term is reserved for red wine only.

 

[Jungstar]

Yes - that's frequency response!
Dynamic Range is basically the noise floor vs the maximum signal level before clipping. SINAD is the inverse of THD+N - it measures SIgnal vs Noise And Distortion.

Tubes are very much not clean - they're just lower fidelity devices in general than solid state. One can design tube amps to be effectively transparent - but why bother with tubes if you're going to do that? Most tube devices these days are not designed to be super clean because there's basically no point to tubes without some sort of distortion.

Larger surface area for a woofer or mid-woofer with all other things being equal equals lower distortion at a given SPL, but (in the case of mid-woofers) necessitates a lower crossover point for better directivity behavior (which can have other tradeoffs, namely higher tweeter distortion).

that makes sense!!!

So..THD+N includes Distortion. The latter sounds more valid.
Why are we interested in Maximum Signal? I mean, we rarely listen at maximum volume... And is the noise recorded WHILE we play at max volume?
Would it not make sense to look at noise and distortion while listening af ie 30% of max or something more realistic? (sorry if I missed the point)

Tubes it sounds like it is not really worth measuring.

 

[voodooless]

Why are we interested in Maximum Signal? I mean, we rarely listen at maximum volume...

No, you almost never do. But a basic DAC does not have volume control. So basically it’s always at full volume . But if you DAC has a digital volume control you are basically moving the signal towards to noise floor when lowering volume. That’s why in those cases a low noise floor and high dynamic range are important factors. For the music itself it really is not very important. Also note that the recording itself also will have a noise floor, for 24 bit data, usually after 18 to 20 bits. Of course that get’s attenuated as well when you use volume control (either digital or analog).

 

[Head_Unit]

[measure]

Flavour of Sound? Many talk about a warm or cold speaker or a dark or bright speaker? Full body? Can we measure that?
@ Probably, but not well correlated to the words, because they mean different things to different people. And the frequency response is a steady-state measurement whereas what listeners hear is a time response including reflections from a room.
Sound Pressure ...I would assume a larger driver area creates a larger sound pressure without distortion. But some drivers move more. Can I quantify that with watt and sensitivity?
@ A smaller driver can theoretically produce the same sound pressure by moving more. However for comparable design budgets the smaller one will distort more.
@ Sensitivity specifications are misleading, since they don't apply well to the bass region, and they are also very often exaggerated or simply made up out of the marketing department's butt. (Trust me, I did stints in audio marketing, though I didn't stoop to this baloney).
@ I am not sure what you really mean by the question-can you elaborate more?
Sinad vs Dynamic Range? what is the difference? Something about volume... And also, what's the limit for when we can actually hear noise it?
@ Well that's a bit of a frustration here, I would like to additionally see raw noise levels without the distortion, to have some idea if the raw noise floor might be audible.
@ That's a tricky subject because even a "poor" signal to noise ratio of like 80 dB from a 100 watt (=+20 dBW) amp suggests the noise floor would be 20-80 = -60 dBW so if the speaker was pretty efficient, say 90 dB, the noise would be only 30 dB and you never hear this. BUT sometimes you CAN hear noise out of a tweeter. Seems some overlooked factor(s)
Tubes - are they supposed to be clean? I read that tubes (amp or preamp) on purpose add distortion...
@ Ah not so much on purpose as inherent in the tube designs. And of course not all are the same!
@ There may still be areas of transient distortion which we don't understand enough
@ And those harmonics may be euphoniously nice
@ Plus maybe the glow of the tubes hypnotizes us into feeling a better sound (I'm serious).