I'd say amp and speaker must cooperate properly to get the best final result. Sometimes we can pinpoint the fault, sometime we can't. Something is happening in between them.
A spk + B amp score 100
A spk + C amp score 80
I won't come to the conclusion of B amp is better than C. Instead, we should say A+B is a better combination (given all of them have no obvious intrinsic problem).
Not to mention the different tastes.
In general, if a device delivers more sounds in a system, it usually takes more 'responsibility', it can be bring higher fidelity or more problems. In the example of above, if B amp gets higher score because it delivers "more correct sounds", we might say it's better. But if B amp gets its score by hiding some bad sounds, then it gets very complicated - where does the bad sound come from? why can B amp hide it? or why does the other amp reveal it? is that bad sound really in amps? or it's in the speaker? or it's a product of the wrong combination? ...
A spk + B amp score 100
A spk + C amp score 80
I won't come to the conclusion of B amp is better than C. Instead, we should say A+B is a better combination (given all of them have no obvious intrinsic problem).
Not to mention the different tastes.
In general, if a device delivers more sounds in a system, it usually takes more 'responsibility', it can be bring higher fidelity or more problems. In the example of above, if B amp gets higher score because it delivers "more correct sounds", we might say it's better. But if B amp gets its score by hiding some bad sounds, then it gets very complicated - where does the bad sound come from? why can B amp hide it? or why does the other amp reveal it? is that bad sound really in amps? or it's in the speaker? or it's a product of the wrong combination? ...
Only tubes I've listened to recently are 6LU8s and I haven't designed with or read up much on tubes, so I can't really comment meaningfully on the tube side. Something I have noticed working with sand parts with 50+MHz GBP is the sound takes on something of a tube quality. If there's a sand solution out there with higher fidelity than an LME49713/LME49990+LME49600 I don't know what it is. Only two ICs and a few passives to solder, too---parts for a 49713+49600 channel cost about USD 20 and weigh maybe 15 grams, though with 105+dB sensitivity the 49600's optional. BUF03's pretty good on GBP and slew but neither it or Bride of Zen have the linearity I'd look for. The AD811---BUF03's replacement---is a reasonable competitor to the 49713; both parts are in a class where I'd be wanting surface mount and good layout.
There are a couple class Ds I find interesting and may try for fun. But for the most part the technology's not at the point where I'm ready to adopt it yet. Glancing at a couple datasheets even pretty good parts like the ADAU1592 yield whisper level noise at horn sensitivities.
A chord should sound scary instead of pleasing?
There's not much difference between high order products and the notes chosen for chords...
Last edited:
A chord should sound scary instead of pleasing?
Don't know where scary comes from... the ear/brain evolved to detect things that stand out, they don't have to be scary, but if further attention decides they are scary, then its time to run away (or prepare to fight)
A chord won't have notch distortion or high order distortion components (unless amplified -- class B or AB needed in the former, feedback helps in the latter)
dave
Masking is what makes 2nd & to a lesser extent 3rd not so objectionable, it doesn't do so well on higher orders.
For instance speakers can produce gobs of 2nd & 3rd, higher orders not so much. Amplifiers on the other hand can.
And when i'm talking about higher order harmonics, i'm talking about distortion products, not natural harmonics produced by an instrument...i thot that should be clear since the discussion has been about distortion.
dave
For instance speakers can produce gobs of 2nd & 3rd, higher orders not so much. Amplifiers on the other hand can.
And when i'm talking about higher order harmonics, i'm talking about distortion products, not natural harmonics produced by an instrument...i thot that should be clear since the discussion has been about distortion.
dave
Because amplifiers do, speakers don't... that is why you can hear an amplifiers numerically much smaller high order distortion products thru a loudspeaker even if the total distortion of the loudspeaker is higher than the amp's. The speaker has none of the really nasty (unnatural, amplifier) ones that aren't masked.
dave
dave
I wouldn't go as far Francec but, yeah, I'm having some trouble here too. There's nothing in the DS, Gedlee, or Rnonlin metrics which weights higher order products more heavily or treats masking differently at higher order harmonics. But I get the impression Dave's pretty locked into the idea and I'm not real keen on going around on it more.
QUOTE: "However loudspeakers do reproduce everything passed to them by the amplifier including high-order harmonics."
Except we know loudspeakers have the highest level of distortion in the whole chain of reproduction and, therefore, can't reproduce accurately the products of the amp or preceding units. What's more, if these higher order harmonics are outside of the capacity of the speakers to reproduce them, how do we even get to hear them?
Frank
Except we know loudspeakers have the highest level of distortion in the whole chain of reproduction and, therefore, can't reproduce accurately the products of the amp or preceding units. What's more, if these higher order harmonics are outside of the capacity of the speakers to reproduce them, how do we even get to hear them?
Frank
Hi cls,
Did you ever measure the response of your tube amplifier on the terminals. with the speaker load?
And connected the same speaker to a solid state amp and measured the response.
If you do you will see it is probably the reaction on the load that you like of the tube amplifier in combination with harmonics this is what your ears like. But not a correct reproduction. And sure a speaker + tube-amp combination that fits your taste.
I have a tube amp my self and plan to build one for the reasons above. For correct reproducing the recording not reliable.
This is my fundamental gripe with the nature of these threads; I don't care that anyone likes the sound of something but please don't say it is "better" unless there is some framework on which to base the premise.
As Helmuth says, you may prefer a tube amp but to claim superiority when it has measurably higher levels of distortion smacks of self-deception.
Frank
I'm aware of only minimal data supporting it, but the TAD metric incorporates the most solidly argued hypothesis I've come across about why tubes and sand can both offer good perceived performance despite significantly different distortion characteristics---namely that if a source's distortions approximate the ear's distortions the distortions get masked. Worth thinking about, though I believe a much stronger case needs to be made with regards to the ear's ability to manage harmonics from different sources for the metric to be based on solid theory. As written, the thesis assumes the ear can reliably distinguish between the harmonics of, say, E1 and E2 notes struck in the same chord as well as the harmonics a tube would generate of each notes' harmonics. It's not at all clear to me that's correct; ears can be quite discriminating but they're not infinitely descriminating. I also rather suspect a change to Rnonlin's OME (outer middle ear) weighting function instead of relying on POTS oriented post-war approximations of the ear would improve TAD's correlation with percieved quality. Understanding of the cochlear amplifier has, after all, improved a bit in the past 60 years or so. Though, in fairness, Rnonlin was published in 2003 and Cheever wrote his thesis in 2001.
My testing shows electronics is very good at producing distortions which are not harmonically related to the input signal. These distortions are very very easy to hear where harmonic distortion is not nearly so easy to hear. Speakers produce a lot of HD which changes the timbre of the music. Speakers also produce a lot of self noise which raises the noise floor causing a "window" of sound effectively located between the noise floor and the signal. In the case of the electronic non-harmonic distortion there is no signal occurring at the non-harmonically generated tone. This makes the tone very easy to affect the perceived sound. Further, these types of distortions have a tendency to increase (some say pile up) with increasing frequency. This leads to the perception of brightness or high frequency fuzz of whatever malady located in the higher bands. In bad amplifiers this effect can even appear in the bass. Speakers tend not to make non-harmonically related tones unless there is something very wrong with the speaker such as voice coil buzz, cabinet buzz, flopping spider or surround, or just bad mechanical whatever. Some drivers end up with very high Q resonators like this very expensive ceramic mid I tested was dead flat except for a 14dB peak at 2442Hz. At that frequency the driver rang like a bell. With such a bell characteristic 2442Hz was always present no matter the drive signal always coloring the sound with its presence.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.