Do it by the text book every time, get the same results every time.
Instead, thinking laterally can produce new ideas that move the world forward. The Lightspeed is one of these ideas. Simple and elegant.
You can analyse the theory all you want but it works very well in practice.
Those concerned with channel balance should consider adding a balance control instead of trying to match LDRs to rediculous levels of precision.
Regards
Paul
Instead, thinking laterally can produce new ideas that move the world forward. The Lightspeed is one of these ideas. Simple and elegant.
You can analyse the theory all you want but it works very well in practice.
Those concerned with channel balance should consider adding a balance control instead of trying to match LDRs to rediculous levels of precision.
Regards
Paul
I ran a test of 460 and tested at 5 points. Was following Georges comment from several months back that this was best, so why not try it. I got no more matches at 5 points than I do at 4 points and I figured I will stick with 4. It has served me well for a few years now. That said though, it would be great to be able to match at 20 points but way to time consuming. On that little tiny board I have posted schematics of I added a balance circuit (left/right) and it is because while I go for very very close matching it is not always perfect, well never perfect. In my own system when I fire it up for the first time with a new set of LDRs I dial the image into the center with my balance control at normal listening level. I dont find a need to mess with it further even after several months of listening at different volumes. I know for sure that there is a bit of difference between the Left series LDR and the Right series LDR but the image still stays in the center so its not off so much that it needs me to continue messing with balance pot everytime. Bravo for George matching at 5 points. I dont find it necessary and dont have complaining customers but more is always better.
MarkusA, Maximus and SoongSC have that DIY persistence and 'glass half full' optimism that is so commendable and such a great part of this website. I think we sometimes look at audio circuits in a way to "2 dimensional" way. Get out the check list, measure everything, does it meet the checklist? Yes, well it must sound good then. This makes sense only to a point. Its part of the process but if it measures well and sounds poorly then something is amiss. We can and should, I think, measure with our ears, but there also has to be that 3rd or 4th dimension that we arent measuring and dont know to measure that is effecting what we hear. George measured some of it years ago with that fancy HP instrument he was able to use and it gave him some insight to better sound. Its one of those dimensions we dont normally measure. Maybe its because people are not usually able to use such high precision equipment and yet are able to come up with wonderful sounding equipment without those measurements, but perhaps the percentage of great sounding equipment would rise if everyone was measuring what George measured/saw/heard and if more people looked for something new. Maximus, MarkusA and SoongSC all may not have come right out and said it or even meant it but I think I can take from their recent posts that they would agree there is something we arent measuring but that we are hearing and simply because its not measured doesnt mean its not there.
Its just a resistor right? So we can put signal through it and measure voltage, current, resistance, impedance, inductance, distortion, etc etc but obviously it doesnt SOUND like other resistors. So there is something we are not measuring that we are hearing. Talk about measuring all you want but if you cant point out incredible sound on your spec sheets then the spec sheets are not telling the whole story.
Well my mind is obviously all over the place today. Happens a lot
but keeps life interesting. Hope you all are having a great time with your families whether you do or dont celebrate the upcoming holidays. We have been having a great time and hope you all are as well.
Uriah
MarkusA, Maximus and SoongSC have that DIY persistence and 'glass half full' optimism that is so commendable and such a great part of this website. I think we sometimes look at audio circuits in a way to "2 dimensional" way. Get out the check list, measure everything, does it meet the checklist? Yes, well it must sound good then. This makes sense only to a point. Its part of the process but if it measures well and sounds poorly then something is amiss. We can and should, I think, measure with our ears, but there also has to be that 3rd or 4th dimension that we arent measuring and dont know to measure that is effecting what we hear. George measured some of it years ago with that fancy HP instrument he was able to use and it gave him some insight to better sound. Its one of those dimensions we dont normally measure. Maybe its because people are not usually able to use such high precision equipment and yet are able to come up with wonderful sounding equipment without those measurements, but perhaps the percentage of great sounding equipment would rise if everyone was measuring what George measured/saw/heard and if more people looked for something new. Maximus, MarkusA and SoongSC all may not have come right out and said it or even meant it but I think I can take from their recent posts that they would agree there is something we arent measuring but that we are hearing and simply because its not measured doesnt mean its not there.
Its just a resistor right? So we can put signal through it and measure voltage, current, resistance, impedance, inductance, distortion, etc etc but obviously it doesnt SOUND like other resistors. So there is something we are not measuring that we are hearing. Talk about measuring all you want but if you cant point out incredible sound on your spec sheets then the spec sheets are not telling the whole story.
Well my mind is obviously all over the place today. Happens a lot
but keeps life interesting. Hope you all are having a great time with your families whether you do or dont celebrate the upcoming holidays. We have been having a great time and hope you all are as well.Uriah
There are two different ways to "match" these components. One is to buy a bunch of them and then, with a great deal of patience, to measure 5 points until two LDRs are similar enough to consider they are "matched". But I recommend to retry the same measures of these 2 components a day or two after this first measure... Some interesting surprises will arise...
The other method is to determine, whatever the LDRs, the requested value of current for a given resistance. Taking, of course, the same patience for the measures... When a control is performed one or two days after, one will find a better match between the measures than with the previous method.
The other advantage of this second method is that almost all the LDRs can be used without buying bunches !...
One more is that, obviously, if some LDR must be replaced, any new one will do and we don't have to find a new matched set !...
Again, this "5 points method" unfortunately wouldn't work for balanced use... The differential approach of the signal in balanced mode requires a much better "matching" all over the usable range. Whereas in single-ended mode a mismatch can be compensated by a balance control between channels, the same "mismatch" is much more, difficult or perhaps not possible to correct.
I have no good suggestion for "monitoring" the LDRs. But, their "calibration" from time to time is of course perfectly possible and I did it several times !...
The improvement of this method is the "automatic calibration", i. e. the automatic measure of all the LDRs. This will take from one hour to a day long if a very good matching is required. But who cares, it can be done when one is out for the day or the week-end !...
Last edited:
@udailey
Yes, though we are 'measuring' that something with devices appropriate to its format, namely ears and minds.
Why does 'measure' have such a visual connotation.?
I 'hear' you
If all goes as planned, you might see a thesis on this subject.
Product may take a while after that; if it happens, I'd like George to have the first opportunity of evaluating whether it fits into his operations or not. Got my fingers crossed.
Last edited:
If all goes as planned, you might see a thesis on this subject.
Either I haven't explained it correctly or you don't understand English, but my comments and my explanations aren't any dark theory, or a "to do" list...
This attenuator is already a real functional device. Not a project...
Anyway, I worked a lot on this subject for two years, and I have enough material to write a book on the subject... But who cares ?
Onesex, I don't understand, you have monotonously tried to denigrate the Lightspeed Attenuator since a few pages back for it's so called unacceptable distortions, why do you persist in this thread. The only way your going to be happy is by including it into an active stage and then putting a s--- load of feedback around it to satisfy the measurement side of yourself.
Cheers George
Cheers George
Now you said it : you don't understand... I haven't "denigrate" the LS, I said that the distortion values aren't compatible with the development of a high-end attenuator. This is a totally different sentence. And this is why I choose a different way to achieve this goal...
Anyway, I acquired enough skill in the improvement of this device to help those who want (as previously said in post #3778). Remember the warning of the moderator : "If you're not interested in engaging other members who have technical criticisms, you're under no obligation to do so." Again, I understand you're happy with your simple LS MkII. And those who already are like you, don't have question or intention to change anything. It's OK for me...
But, some other, like I did it in the past, are probably ready to improve this design. Why would you prohibit to all these guys to benefit from some developments of the community ? Isn't a contradiction with the reason of being of this forum ?...
I think what George is saying is that you can also open a new thread to talk about your ideas like George has done. Note that most forum members just say directly what the distortion is caused by. But further discussion on what the improvement is can be discussed in a different thread. If you have normal measurements to show how much better it is, then fine, post the product measurements. Up to now, nobody has even heard the product you are talking about.
Many Chinese based forums communicate as you do, but when invited to any audiophile gathering to demonstrate, they become very defensive when they cannot demonstrate audibly agreeable improvement.
I'm sure there are probably a few people that participate in this forum whom live not too far from where you are. Maybe another's listening impression might help.
Many Chinese based forums communicate as you do, but when invited to any audiophile gathering to demonstrate, they become very defensive when they cannot demonstrate audibly agreeable improvement.
I'm sure there are probably a few people that participate in this forum whom live not too far from where you are. Maybe another's listening impression might help.
Last edited:
All this worrying about attempting (close to) perfect channel balance is a waste of time. Too many variables exist for the "perfection" to matter.
Your ears are not perfectly matched
Your room most probably has different sonic absorption/reflection properties from side to side (depending on who else is in the room, at least)
Your phono cartridge (if you use records) doesn't have perfect channel match
Your tube equipment almost certainly has different gain from channel to channel.
Your speakers will have some channel to channel differences.
You won't always be seated to within 0.1% of the center of the speakers
Your source material won't be that closely balanced
Simplicity is the ultimate sophistication. There are three reasonable alternatives:
Live with minor imbalances (George's setup)
Add a balance control (as Uriah does)
Build with two gain controls. Set one for volume then set the other to balance, it's really quite easy.
Want to improve it? Make a better LDR!
Stuart
Your ears are not perfectly matched
Your room most probably has different sonic absorption/reflection properties from side to side (depending on who else is in the room, at least)
Your phono cartridge (if you use records) doesn't have perfect channel match
Your tube equipment almost certainly has different gain from channel to channel.
Your speakers will have some channel to channel differences.
You won't always be seated to within 0.1% of the center of the speakers
Your source material won't be that closely balanced
Simplicity is the ultimate sophistication. There are three reasonable alternatives:
Live with minor imbalances (George's setup)
Add a balance control (as Uriah does)
Build with two gain controls. Set one for volume then set the other to balance, it's really quite easy.
Want to improve it? Make a better LDR!
Stuart
It sounds to me like Ondesx is proposing a system of control that reduces distortions by perfectly matching current to that required to produce a given resistance using microprocessor controlled current sources to feed each LDR with the currents chosen from a look up table to determine the current needed for each level of attenuation. That sounds like it would be more accurate and the concept is fairly simple to those with programming capabilities. Not me, but my son thinks he could do it rather quickly.
The net sonic result is probably significantly up the diminishing returns curve, compared to the Lightspeed. You'd also have to concern yourself with shielding the digital portion from the analog. I don't know if that would be a significant issue or not.
George has been kind enough to share his work with us and many people seem to like the performance a lot, even though there may be areas of that may be improved. Yes, it likes a high impedance load. George stated that. It's fairly easy to add a buffer if needed. Better tracking may produce better distortion measurements, but is it audible? If it is to you, enjoy. Don't denigrate a design that many find quite satisfactory because it doesn't meet YOUR needs.
Ondesx, you seem to have a viable idea. However, the level of criticism of the Lightspeed and personal attacks against George are completely unwarranted. Many of us like to go for the simplest solution that works well. This thread is about an idea that performs very well with a simple implementation. If you want to discuss a way to potentially take the performance up a notch, why not start your own thread titled something like "A More Accurate Lightspeed" the way Zen Mod did with his "poor Serbian man's Lightspeed" thread?
Your continued participation in this thread is fairly disruptive and the tone of many of your posts makes it difficult to focus on the merits of your concept or George's. I (and probably others) would be interested to see your ideas in another less confrontational environment. Maybe you don't mean it that way, but much of what you write come off as harsh. I hope that you will start a new thread to discuss your ideas.
The net sonic result is probably significantly up the diminishing returns curve, compared to the Lightspeed. You'd also have to concern yourself with shielding the digital portion from the analog. I don't know if that would be a significant issue or not.
George has been kind enough to share his work with us and many people seem to like the performance a lot, even though there may be areas of that may be improved. Yes, it likes a high impedance load. George stated that. It's fairly easy to add a buffer if needed. Better tracking may produce better distortion measurements, but is it audible? If it is to you, enjoy. Don't denigrate a design that many find quite satisfactory because it doesn't meet YOUR needs.
Ondesx, you seem to have a viable idea. However, the level of criticism of the Lightspeed and personal attacks against George are completely unwarranted. Many of us like to go for the simplest solution that works well. This thread is about an idea that performs very well with a simple implementation. If you want to discuss a way to potentially take the performance up a notch, why not start your own thread titled something like "A More Accurate Lightspeed" the way Zen Mod did with his "poor Serbian man's Lightspeed" thread?
Your continued participation in this thread is fairly disruptive and the tone of many of your posts makes it difficult to focus on the merits of your concept or George's. I (and probably others) would be interested to see your ideas in another less confrontational environment. Maybe you don't mean it that way, but much of what you write come off as harsh. I hope that you will start a new thread to discuss your ideas.
OK Bob. It is exactly what I done, but the distortion won't lower by a correct matching... Unfortunately !
It seems that you or somebody near you, can do it. Nice. It's enough for me...
Happy Christmas !
Last edited: