It's well known in the industry that distortion figures may not be directly related to percieved sound quality. This is what George is trying to point out. So there must be some other aspect improved that relates with percieved sound quality. It's too bad that the method to find a relation between sound quality and measured data is not openly available. I personally only use single tone harmonics testing in the process of diagnosis, not the final judgement.
Related comments here...http://www.diyaudio.com/forums/anal...optocoupler-volume-control-3.html#post2386926
onesdx, you have not answered my question yet...
onesdx, you have not answered my question yet...
This is just... untrue ! And George himself claims just the contrary... even for bad reasons !
You're definitely not in agreement with what most of the professional audio engineers know for a long time ago now... Please, would you explain what means exactly your signature : "Hear the real thing!"
Last edited:
Only Nelson, your just riding on his coat tails, he showed us this over a year ago, and designed a buffer for the Lightspeed so it can drive his low impedance amps, don't take his credit, it's not professional at all.
"![/QUOTE]I think you gave a wrong explanation of a subjective fact. Me and other think that the subjective improvement of the series-shunt setup is almost only due to the odd harmonics and, moreover, we know where are coming from, these odd harmonics..."![/QUOTE]
Only you belive this, read what owners of high end active pre's have found, over 300 posts
AudiogoN Forums: Lightspeed Attenuator - Best Preamp Ever?
Cheers George
You're kidding, aren't you ? Because ten or even thousand of people prefer a distorted sound is enough to proof that your explanation of "contact issue" is true ?...
Well, I think we must stop this discussion here since religion and science aren't on the same level...
Before stopping this vain hope of mutual understanding, here you'll find a visual evidence that our senses are sometimes fooled... What do you think of the gray level of squares A and B ?...
Attachments
Things have now stopped screeching in the night . . .
Sounds to me Dave like the Sowther is ringing it's brains out with the load (7K-12K) load that the Lightspeed presenting . . .
And, as it's sounding to me now, you were spot on. I dug out an old (Ben Duncan's 1988 HFNRR design) op-amp I/V stage from where it's lain under the bed these last 10 years and swapped it for the Sowthers. It's working a treat.
Many thanks!
Sounds to me Dave like the Sowther is ringing it's brains out with the load (7K-12K) load that the Lightspeed presenting . . .
And, as it's sounding to me now, you were spot on. I dug out an old (Ben Duncan's 1988 HFNRR design) op-amp I/V stage from where it's lain under the bed these last 10 years and swapped it for the Sowthers. It's working a treat.
Many thanks!
Now, I call this trying to involved unrelated data to prove a point. Shame. So if you wish to stop, nobody is going to beg you to to continue.
I have asked you about the specific resistance values of your test setup, you have never provided a valid answer. Seems like you are not quite sure what you did.
Last edited:
Distortion measurements is only one means of troubleshooting a circuit under the assumption that linear behavior should be a goal. However, even though harmonics exist through this kind of measurement, it may create a perception of tone in an instrument played. This does not mean it will sound bad, it just sounds different. Since, for example, different violins are going to sound different anyway, so it does not matter if the reproduction has a different harmonic content because you don't know what the original one sounds like anyway. Therefore, the reproduction can still sound real. This is why I have the signature to "Hear the reall thing", when you are able to percieve certain characteristics in an instrument that cannot be properly analyzed through data, you know what to listen for. So from what I can see from the data you look at, it's probably going to take about 3 to 4 years to really start to get the understanding involved with linking data and listening perception.
For the record, I currently do not use LDRs, and have only had one listening session with those connected to an amplifier I have as a demo to a third person. The LDRs beat the pants off the volume control I had in there. At that time, I was not aware how the input impedance varied with frequency. Therefore, I would encourage you to do input impedance measurements on your LDRs. Distortion measurements will not tell you what you are looking for.
Last edited:
So you won't have to look it up, here I am quoting myself.
I think I had it right back then. Starting with post 3677, I think ondesx's points can be summarized thus:
-LDRs have greater distortion than a simple resistor or stepped attenuator
-The lowest measured distortion device is what we must listen to
-"We", known to this forum as "ondesx" have developed a superior product to any LDR based attenuator
-ondesx's attenuator's actual configuration/schematic must be kept secret, although he has no product to sell
-ondesx is better and has a better understanding of the science/electronics/measurements/etc. than the rest of us (side note is that he used another member of "we" to obtain those measurements)
-If you disagree with him, you are following a religion, not a science
I asked specifically if he had ever built or listened to a Lightspeed or clone. What system(s) had he used to listen to his or any other attenuator versus the subject of this thread. I asked what deficits he heard in the LS. I asked about his testing methodology to ensure temperature wasn't causing his perceived variations. Finally, I wondered if he would be willing to demonstrate his product in comparison to an LDR attenuator.
The answers were predictable. No reply to system used, whether he'd ever listened to or built an LDR attenuator. Vague replies about a "standard" test method. No reply as to the sonic flaws heard in the LS. Naturally, he lives in a part of the world far away from any one else who might own an LDR based attenuator.
His insinuations continue about those of us who use human ears to evaluate various products whose purpose is to provide us with the most realistic sounding reproduction of music possible as perceived through our human ears.
I may be wrong, but I think that onedesx has never listened to an LDR setup. I also think that in the near future, the not-for-sale Jackspeed will soon become a commercial product and its supporting literature will tout its superiority over the LS (and probably everything else ever made).
Sorry, but for me, this argument with ondesx is pointless.
Stuart
Heh, lot's of companies are trying to pump things out without really managing comparisons with older products. I bought a new cell phone thinking it had more features than my old one, only to discover the phone is not as smart as my old smart phone. Product support is even more dumb, they did not know more than the phone owner.
Just one answer because all other words in your post are bla bla bla...
You have the picts in one of the above posts of my own MKII Balanced LS... Please see #3703. You deserve some glasses don't you ?
Last edited:
Again I think you're unable to read the measurement values and that you ignore the advice of George in terms of optimal resistance values...
Then, to be clear and exhaustive: 7K is the resistance value recommended by George himself for several reasons (maintain a low noise, not overload the source, etc.), then for 16x attenuation, i. e. -24 dB you deal with Rseries = 6.558K and Rshunt = 442 ohms. The sum, as you'll easily verify it, equals 7K...
For the MkI the Rseries is 15K fixed value, then for 24 dB of attenuation you'll get 946 ohms.
That's all folks !
All things can perfectly be compared and the results matches the expected behavior as soon as one knows what is going on with these LDRs !
Is that clear for you now?
No.
I was asking about your other test configurations which you claim lower distortion. I have no question about what George has done. He has passion in what he does. I see information in other peoples data, but yours are just 1KHz test signals, not very informative. Pity.
Last edited:
First, I'd like to know what other volume control configurations you have already actually tested. The resistance values? Simple shunt-series type configurations do not have linear gain. You know, when we look at data, we want to make sure what we are comparing.
First, I'd like to know what other volume control configurations you have already actually tested. The resistance values? Simple shunt-series type configurations do not have linear gain. You know, when we look at data, we want to make sure what we are comparing.
I'm sorry but I'm not sure to understand your request. The plotted curves are for the MKI i. e. for a series fixed resistance (this is the definition of the MKI) and the series-shunt LDR setup i. e. the MkII. The way both work is obviously well known in this thread. Then, what do you mean by "what other volume control configurations you have already actually tested ?" perhaps another valiue than 15K for the fixed series or a different total of the 7K (=series+shunt values) for the MkII ? Or perhaps are you asking for my last attenuator not using LDRs ?...
Secondly, what do you mean by "Simple shunt-series type configurations do not have linear gain" ?
Both setups are compared for THE SAME attenuation value (-24dB for the published curves) and the most favorable "total resistance" for the MkII. Then, the derived data are quite well comparable. I did the measurements for all the attenuation values allowed by these setups. I also published, in one post above, a couple of curves at 0dB (i. e. no attenuation). Obviously the differences are much more visible. These data point out, as expected, that :
- first, the MkII suffers for greater distortion than the MkI
- secondly, the MkI almost generates only even harmonics
- thirdly, the MkII almost generates only odd harmonics.
Depending of the attenuation level the number and the levels of these harmonics are different. But the rules remain. The reason of these results is related to the Fourier Transform of the output signal.
Last edited:
Please calm yourself George! ... You knew all that. Where is the problem? If you like the sound of the LS and many people do agree, all is fine.
All I did was answer questions and recall that we could not just rely on what we hear, like we can't only trust on our eyes as demonstrated by the drawing with squares A and B of the same gray level, even if everyone will agree obviously to declare that each square is of a different level... Even with this information, the picture remains realistic !...
All I did was answer questions and recall that we could not just rely on what we hear, like we can't only trust on our eyes as demonstrated by the drawing with squares A and B of the same gray level, even if everyone will agree obviously to declare that each square is of a different level... Even with this information, the picture remains realistic !...
Last edited: