Yes, lots of people report something similar. Two possible explanations (both might be part of the truth):Ken Newton said:
1. Optimising one thing (low THD) can lead to a designer missing other important things or introducing new problems.
2. Some people actually prefer a little low order distortion (there is some experimental evidence to support this, as well as the anecdotal evidence of the popularity of 'tube buffers' etc. - I believe there is even some evidence that some prefer 2nd and others prefer 3rd)
My view is that low THD can mean low IMD, which must be a good thing provided other problems are not created in the process. Low order harmonics below a few percent are generally not audible, and below what the speakers will be producing anyway. So straining for extra zeroes after the decimal point is silly, but carelessly including unnecessary distortion through poor design or adjustment is equally daft.
The PCM2707 is not a current output DAC so no I/V converter is required for that one.
The PCM1704 is a current output DAC and does need an I/V converter.
Which one are you talking about?
You assume that what happend in '70's is still happening. Any low THD decent device today will have low intermodulation products too. The lessons from the deceiving ads from '70's where learned by all manufacturers that matter.
On the other hand, poor THD always sounds bad. Coloration of a musical program might be "desirable" only if is linear (like tone controls). That means maintaining the low harmonics.
Sure, there are people like to stick a sub in the car trunk and crank it up to 1000W with all the car panels rattling. And they say "it sounds goood". But that doesn't make them right.
Also, your argument that you don't have to be better than the speaker THD. So, in this thread that is about I/V stage of a DAC, you recommend DAC's with THD around 1%? So... 7 bits of data (0.8%) are enough?
Last edited:
Some people prefer Russian caviar, some people prefer French fois gras. Who's correct?
Sorry, I don't understand your point. Exactly what of all that happened in the 1970's am I supposed to be assuming?SoNic_real_one said:
Are you talking about frequency response (maintaining what is already there in the signal) or distortion (adding new harmonics)? I thought we were talking about distortion, but then you introduce the new topic of maintaining harmonics.
I think from the context of my remarks it is clear that I am talking about low order distortion. Low bit DACs don't just produce low order, but lots of high order too. So for DACs some zeroes after the decimal point are needed, except perhaps for low order distortion introduced by the analogue circuitry. The DAC chip itself needs to be as linear as possible; the following circuitry can sometimes be less accurate.
Amplifiers, DAC's, preamps, and loudspeakers all have different flavors as well. Everyone's entitled to their choice of flavors.
Why? The errors are compounding on the audio chain. It's better to keep everthing as transparent as possible.
But there is a RIGHT way to reproduce something faithfully. If you want to listen to music as close as possible of the way the original artist made it, there is only one correct "way".
If you want to make your own mix of original music with added distortions, that is not part of technical audio reproduction anymore. Call it "art" if that floats your boat.
Last edited:
Her are two examples of a different approch, that is common base/gate that also represents a very low impedance, but doesn't necesarily use high-feedback opamps.
peufeu from France
...and Pioneer ODR reference car audio system ( YES ...for CAR ... but extremly well sounding) :
Ciao
peufeu from France
...and Pioneer ODR reference car audio system ( YES ...for CAR ... but extremly well sounding) :
Ciao
Attachments
Most of us do though (purchase only high-feedback solid-state amplification components). Most being people in the world. Other than for instrument amplifiers, that is.
None of us have a Ken Newton at home to plug in to perform our 'perceived sound quality' evaluations so how much further forward your observation is taking us is open to question.
Perhaps you guys who have such faith that there are perceptible but unmeasured and unmeasurable qualities in recorded music believe that you can swamp the opinions of the realists among us here but you should take note that your credibility with the general buying public diminishes daily, you are increasingly regarded as a fanatic rump, unworthy of serious consideration. Your opinions preponderate only where you congregate. Even Stereophile thinks it may have gone too far.
Please, you defenders of common sense and sensibility, however uncouth and untutored your expressions, keep them coming. It's time we drove these hi-falutin proponents of anti-intellectualism and pundits of the ineffectual back into their boxes for a while so that we can get on with some grounded DIY.
It wasn't the only one advocating such a configuration. Jocko Homo I think was the first (now is banned here, moved on a pesonal forum).
Requires special transistors (hard to find and need matching). Per Pass measurements the average THD is 0.01%. That is -80dB or less than 14 bit resolution (and not adding the noise).
Other aproach: Marantz tried the discrete OpAmp HDAM from 90's (in CD-63-SE for example). I still have to hear one to belive it.
Mark Levinson and Musical Fidelity tried the path of discrete OpAmps too. Burson Audio sells them too. Heck, even I buid a few (bipolars, jfet) in 80's.
Again, results are nothig spectacular compared with the new crop of integrated OpAmps.
About the claim "measures bad but sounds good": http://www.benchmarkmedia.com/discu...our-ears-lie-where-does-truth-lie-examination
Last edited:
Fairly typical straw man. Let's have a show of hands here - who believes that the perceived sound differences we hear in I/V conversion are unmeasurable?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.