Pan said:
Clearly you have never looked at real music on an oscilloscope. Voices? String instruments? Organs? You should have stopped with flute, at least you had that one semi-right, although apparently only by accident.
A sound card?
Riiiight... No wonder you don't believe anything else matters.
You definitely need to get out and listen to more stuff. High feedback, low feedback, tube, solid state, digital, analog...it's all useful for getting a feel for what can and cannot be done. While you're at it, it wouldn't hurt to attend a number of concerts where people are playing unamplified music.
Until you show some evidence that you've done your homework there's not much point in continuing.
Grey
janneman said:
Here is the free download of Audio Diffmaker software.
http://www.libinst.com/Audio DiffMaker.htm
From what I heard about it, I can really recommend it 😎
A good audio tool.
Grey, you say that measurements are no indicator of sound quality and yet you say that amps with very low measured distortions have low sound quality.
From this I would be able to predict which amp you would reject simply by using measurements. So in your particular case, measurements are a very good indicator of sound quality. 😉
From this I would be able to predict which amp you would reject simply by using measurements. So in your particular case, measurements are a very good indicator of sound quality. 😉
GRollins said:
I have a scope but it's true, I have never loked at music waveforms on it. OTOH no music reach up to the MHz range so a scope is not necessary.
I do look at waveforms in the software I use for recording music though. I use Reaper and Audacity right now but used Cubase earlier. The presentation is identical as on a scope only you have much higher dynamic range using the soundcard/AD instead of the scope. The soundcard only has 60-70kHz bandwith against the MHz BW of the scope but I think 60-70kHz is enough for music.
The flute pic you saw from me is recorded with a pair of Sennheiser MKH8040 mic's into a Lynx Aurora and that's the soundcard I commonly use for listening on my stereo. It sounds fenomenal and is a mastering grade converter of the type that is used for the best CD's in your collection. It's audibly transparent at the highest sampling speed (192kS/s) and have extremly good measured performance.
No, the flute graph is recorded by me so obvisouly not by accident.
The Lynx Aurora outperforms most any CD player you can find out there. It is transparent at high speeds and close to transparent at 44.1kS/s so that is evidently not a limitation. Of course I believe that things matters... why do you put words in my mouth??
I find your behaviour disturbing!
I play, write, and record music myself with friends. We play in home studios and in various live venues. I recorded my friend the other day playing a grand piano in a hall.
I play acoustic and electric guitars, Irish wooden flute, tin whistle, banjo, harmonica and various percussive instruments. Having top notch recording gear such as matched pairs of Sennheiser MKH8040, MKH8020, Earthworks QTC1 and other mic's and pro-master grade AD-DA converters and mic preamps I think I have a relatively good situation in listening to live music and bring the recorded material with me home.
I often make sure to hear various forms of performances such as concerts, opera, street musicians and so on.
You really are embarassing!
/Peter
back on track
I see this thread is maybe diverting from the original article. Upon further reflection I also see there's no easy solution. With that in mind, I ask Mr Pass if he can recommend any good recordings of unaccompanied sackbutt.
I see this thread is maybe diverting from the original article. Upon further reflection I also see there's no easy solution. With that in mind, I ask Mr Pass if he can recommend any good recordings of unaccompanied sackbutt.
A good way to check for distortion of significane is to insert a piece of equiment in a complete chain and see if there's a difference in sound with the DUT in and out of the chain.
Let's start out with a high performance rig (subjectively and objectively) and some critical hard to handle music and signals.
We listen to the rig as is, and then we insert the DUT, if the sound does not change we must consider the DUT being transparent.
If the DUT is designed with feedback loops and even global feedback loops it would be a strong indication that some people are, shall we say slightly confused about the matter..
One example; a CDP (CD player) is feeding an amp, now let's stick a AD and DA between the CDP and amp and adjust levels to 0.05dB.
Or; insert a power amp driving a complex nonlinear dummy load between the preamp and power amp of the main chain.
The main amp is fed via a attenuator after the DUT.
Preamps can of course be tested this way as well.
In all three scenarios above it is perfectly possible to end up without a detectable difference in sound. This has been done with electronics using feedback circuits and IC opamps.
/Peter
Let's start out with a high performance rig (subjectively and objectively) and some critical hard to handle music and signals.
We listen to the rig as is, and then we insert the DUT, if the sound does not change we must consider the DUT being transparent.
If the DUT is designed with feedback loops and even global feedback loops it would be a strong indication that some people are, shall we say slightly confused about the matter..
One example; a CDP (CD player) is feeding an amp, now let's stick a AD and DA between the CDP and amp and adjust levels to 0.05dB.
Or; insert a power amp driving a complex nonlinear dummy load between the preamp and power amp of the main chain.
The main amp is fed via a attenuator after the DUT.
Preamps can of course be tested this way as well.
In all three scenarios above it is perfectly possible to end up without a detectable difference in sound. This has been done with electronics using feedback circuits and IC opamps.
/Peter
It's been done. An eye-opener.
http://www.bostonaudiosociety.org/bas_speaker/abx_testing2.htm
Jan Didden
http://www.bostonaudiosociety.org/bas_speaker/abx_testing2.htm
Jan Didden
Evass....
since no one has recommended any recording let me throw out some of my favorites. Tocata en fugue in D minor, performed by E.Power Biggs(actually anything by him)on Telarc (I think).
Joe Satriani,Surfing with the Alien(album). Last but not least The Dereck Trucks Band, Songlines(OMG) this will astound you.
As music is as a personal taste thing I guess you cold call mine eclectic, take it with a grain of salt as you would all of the prior opinions. We are all looking for the ultimate system(within some budget grand or small) . If you build it yourself it has more personal satisfaction and you can make changes, Kind of like the optometrist(is this better or this). Only though your own eyes or ears can you make these judgements. Listen and enjoy...
Regards, Elwood
since no one has recommended any recording let me throw out some of my favorites. Tocata en fugue in D minor, performed by E.Power Biggs(actually anything by him)on Telarc (I think).
Joe Satriani,Surfing with the Alien(album). Last but not least The Dereck Trucks Band, Songlines(OMG) this will astound you.
As music is as a personal taste thing I guess you cold call mine eclectic, take it with a grain of salt as you would all of the prior opinions. We are all looking for the ultimate system(within some budget grand or small) . If you build it yourself it has more personal satisfaction and you can make changes, Kind of like the optometrist(is this better or this). Only though your own eyes or ears can you make these judgements. Listen and enjoy...
Regards, Elwood
I'd just like to sidestep all the insults directed towards people who accept DBT and measurement for what they are, i.e. science, and draw discussion back to the origin of this thread, NP's article.
Is the article seen as offering solace to those who believe that whatever they 'hear' during uncontrolled listening is indisputable audio truth, caused exclusively by changes in the sound waves in the air, and all that remains is for poor old inadequate science to catch up with them via new scientific discoveries?
What say the author?
Is the article seen as offering solace to those who believe that whatever they 'hear' during uncontrolled listening is indisputable audio truth, caused exclusively by changes in the sound waves in the air, and all that remains is for poor old inadequate science to catch up with them via new scientific discoveries?
What say the author?
No solace is on the menu. I simply think that it's interesting to look at
stuff that is not getting much consideration when people speculate
about the audibility of distortion.
😎
stuff that is not getting much consideration when people speculate
about the audibility of distortion.
😎
The simplistic suggestions, like compare output and input and see what's different, have been done. It's not that they weren't useful back in 1940, it's that they've taken us pretty much as far as they can. That's why I feel servo subs are worth exploring--when passive subwoofers regularly exhibit 5% or more of THD and lackluster bandwidth, that's an application that's ripe for negative feedback. Once you get THD below some arbitrary percentage (.1% is often put forward as the limit of perception) it's time to look to other testing techniques. The direction to go is to use music itself, as that's where things fall short.
--There almost has to be a dynamic element to the problem. Does the signal fail to meet the peak voltage/current value expected? That might mean a circuitry shortcoming, but honestly my first thought would be to look at the power supply design. If, indeed, the power supply fails to deliver sufficient current on time, then you would expect power amps to be more prone to this than preamps. Not surprisingly, you do find a loose trend in that direction.
Current limiting, in the conventional sense, means the circuit in question cannot supply enough current to drive the load to the intended voltage swing. What I envision is something along the lines of dynamic current limiting. If you wait long enough, the current actually is available, but sufficient current to swing a sharp, sudden peak is slow to arrive due to, for instance, ESR in the power supply caps, or a voltage regulator that is slow to respond due to a cap at the base/Gate of the pass device (put there with the intention of reducing noise) that takes too long to fall.
I heard a really expensive system once that sounded bloody marvelous...until someone did something dynamic, whereupon everything fell apart. Although I have never seen schematics for the equipment in question, I suspect it had voltage regulators that were unable to respond quickly. The solution in this case would be a redesign of the regulator, obviously.
The slow delivery problem can be approached in several ways. One is film bypass caps, or in the case of line level equipment, all film caps. It's impractical to use all film caps in most power amps. Another possibility is an oversized power transformer, which would give you lower DC resistance on the secondary, which reduces the recharge time when the power supply caps are depleted, and reduces the Zout of the power supply as a whole. A more ideal voltage source.
--Is it something like thermal distortion or perhaps some other poorly understood mechanism? To date, there's been very little work done on the thermal thing to find out just how bad the problem is..or isn't, as the case may be. One problem is that you're either caught trying to dynamically control the temperature of the device as a whole as I described above, or you're left trying to reduce the thermal lag of the device itself.
Oh, boy! Now there's a can of worms! Do you take a stock device and grind down the back plate or epoxy case in an effort to reduce the hysteresis effect of the casing? That's going to go over real well in the field. Suppose you've got a lightning-struck piece of equipment that needs a new transistor. You can't just solder in a new device and expect it to work properly--you've got to special order the thing from the company who made the equipment, because the erstwhile TO-220 case has had its back plate ground and polished to 1/4 of its original thickness. That's not going to go over well...especially if the company is out of business (not impossible, given the economic collapse) or has run out of the part due to cessation of production (2SK389, anyone?).
--Is it a parts problem? Resistors come the closest to their ideal model. Capacitors and semiconductors are nowhere near what they should be. A good film cap is within shouting distance of the ideal, but it's expensive and bulky and could still use improvement. Semiconductors are just flat-out ridiculous. What should be a straight line is badly curved. The curves obey, more or less, idealized mathematical laws which can be used to predict distortion (enter the simulation program), but even then there are imperfections due to things like the Early effect, so you have to add ninety-nine band-aids to get the model to come anywhere close to reality. I'd like to know if there are dynamic problems involved that aren't completely described.
Does "Resistor Man" (a nod to Horowitz & Hill and other texts that use the same general analogy) have trouble responding fast enough? At first blush, you'd tend to think no. After all, you can readily design a circuit that responds out into the MHz region, right? But bear in mind that radio frequency circuits generally have narrow bandwidths. Very narrow bandwidths. In fact, RF designers knock themselves out trying to build circuits that, ideally, have infinitely narrow bandwidth. Hmmm... Okay, so what if you were to try for wide bandwidth? It turns out that there really aren't that many applications that require low noise, wide bandwidth, good dynamic response, and low distortion. Astronomers, for instance, want low noise electronics. So far, so good. For at least some applications, they want wide bandwidth. Excellent! Now we're getting somewhere. But they aren't all that worried about dynamic range. With the exception of certain short-lived phenomena like supernovae, you're not going to see really huge amplitude variations in the signals. Okay, so maybe this isn't working so well after all. Although it wouldn't hurt to pick your friendly local astronomer's brains for low noise techniques...
(Beware, they cool circuits with liquid nitrogen when they're really serious about low noise. I stopped at liquid water...)
In short, audio has nearly unique requirements. We don't know as much about how circuits respond under the conditions we need to meet. Computers need ultra-fast response, but don't give a flip about dynamic range because everything is defined to, say, 0-3 volts. Radio transmitters must by law be held within certain frequency limits, lest they cause problems with adjacent stations. Toaster ovens only care about maintaining a set temperature...and they don't even do that all that accurately. We use electronics every day for a million things, but the controls for an elevator just don't give us a lot to work with in the audio sense. Ditto for stop lights, electric winches, and variable speed drills. We're nearly on our own because there's not big money at stake. Or lives. Or national security. Or much of anything, for that matter. So the innovations have to come from the companies that produce the gear or from motivated hobby-builders. And don't dump on the DIYers until you're read up on the history of some of the audio companies. Some really cool things have been developed by guys working solo in their garages.
There are other things I could throw in here, but you get the idea. There's plenty of room for discussion without assuming the position that THD is all there is and copping an attitude. If you really, truly feel that THD is sufficient, then go build a circuit with opamps and use 80 or 100dB of feedback per chip and be happy.
If you are open to the possibility that there's more to life than infinite feedback around an opamp, then contribute something useful and quit acting like a troll.
Grey
--There almost has to be a dynamic element to the problem. Does the signal fail to meet the peak voltage/current value expected? That might mean a circuitry shortcoming, but honestly my first thought would be to look at the power supply design. If, indeed, the power supply fails to deliver sufficient current on time, then you would expect power amps to be more prone to this than preamps. Not surprisingly, you do find a loose trend in that direction.
Current limiting, in the conventional sense, means the circuit in question cannot supply enough current to drive the load to the intended voltage swing. What I envision is something along the lines of dynamic current limiting. If you wait long enough, the current actually is available, but sufficient current to swing a sharp, sudden peak is slow to arrive due to, for instance, ESR in the power supply caps, or a voltage regulator that is slow to respond due to a cap at the base/Gate of the pass device (put there with the intention of reducing noise) that takes too long to fall.
I heard a really expensive system once that sounded bloody marvelous...until someone did something dynamic, whereupon everything fell apart. Although I have never seen schematics for the equipment in question, I suspect it had voltage regulators that were unable to respond quickly. The solution in this case would be a redesign of the regulator, obviously.
The slow delivery problem can be approached in several ways. One is film bypass caps, or in the case of line level equipment, all film caps. It's impractical to use all film caps in most power amps. Another possibility is an oversized power transformer, which would give you lower DC resistance on the secondary, which reduces the recharge time when the power supply caps are depleted, and reduces the Zout of the power supply as a whole. A more ideal voltage source.
--Is it something like thermal distortion or perhaps some other poorly understood mechanism? To date, there's been very little work done on the thermal thing to find out just how bad the problem is..or isn't, as the case may be. One problem is that you're either caught trying to dynamically control the temperature of the device as a whole as I described above, or you're left trying to reduce the thermal lag of the device itself.
Oh, boy! Now there's a can of worms! Do you take a stock device and grind down the back plate or epoxy case in an effort to reduce the hysteresis effect of the casing? That's going to go over real well in the field. Suppose you've got a lightning-struck piece of equipment that needs a new transistor. You can't just solder in a new device and expect it to work properly--you've got to special order the thing from the company who made the equipment, because the erstwhile TO-220 case has had its back plate ground and polished to 1/4 of its original thickness. That's not going to go over well...especially if the company is out of business (not impossible, given the economic collapse) or has run out of the part due to cessation of production (2SK389, anyone?).
--Is it a parts problem? Resistors come the closest to their ideal model. Capacitors and semiconductors are nowhere near what they should be. A good film cap is within shouting distance of the ideal, but it's expensive and bulky and could still use improvement. Semiconductors are just flat-out ridiculous. What should be a straight line is badly curved. The curves obey, more or less, idealized mathematical laws which can be used to predict distortion (enter the simulation program), but even then there are imperfections due to things like the Early effect, so you have to add ninety-nine band-aids to get the model to come anywhere close to reality. I'd like to know if there are dynamic problems involved that aren't completely described.
Does "Resistor Man" (a nod to Horowitz & Hill and other texts that use the same general analogy) have trouble responding fast enough? At first blush, you'd tend to think no. After all, you can readily design a circuit that responds out into the MHz region, right? But bear in mind that radio frequency circuits generally have narrow bandwidths. Very narrow bandwidths. In fact, RF designers knock themselves out trying to build circuits that, ideally, have infinitely narrow bandwidth. Hmmm... Okay, so what if you were to try for wide bandwidth? It turns out that there really aren't that many applications that require low noise, wide bandwidth, good dynamic response, and low distortion. Astronomers, for instance, want low noise electronics. So far, so good. For at least some applications, they want wide bandwidth. Excellent! Now we're getting somewhere. But they aren't all that worried about dynamic range. With the exception of certain short-lived phenomena like supernovae, you're not going to see really huge amplitude variations in the signals. Okay, so maybe this isn't working so well after all. Although it wouldn't hurt to pick your friendly local astronomer's brains for low noise techniques...
(Beware, they cool circuits with liquid nitrogen when they're really serious about low noise. I stopped at liquid water...)
In short, audio has nearly unique requirements. We don't know as much about how circuits respond under the conditions we need to meet. Computers need ultra-fast response, but don't give a flip about dynamic range because everything is defined to, say, 0-3 volts. Radio transmitters must by law be held within certain frequency limits, lest they cause problems with adjacent stations. Toaster ovens only care about maintaining a set temperature...and they don't even do that all that accurately. We use electronics every day for a million things, but the controls for an elevator just don't give us a lot to work with in the audio sense. Ditto for stop lights, electric winches, and variable speed drills. We're nearly on our own because there's not big money at stake. Or lives. Or national security. Or much of anything, for that matter. So the innovations have to come from the companies that produce the gear or from motivated hobby-builders. And don't dump on the DIYers until you're read up on the history of some of the audio companies. Some really cool things have been developed by guys working solo in their garages.
There are other things I could throw in here, but you get the idea. There's plenty of room for discussion without assuming the position that THD is all there is and copping an attitude. If you really, truly feel that THD is sufficient, then go build a circuit with opamps and use 80 or 100dB of feedback per chip and be happy.
If you are open to the possibility that there's more to life than infinite feedback around an opamp, then contribute something useful and quit acting like a troll.
Grey
grollins,
I agree that THD is one measurement and like any measurement it has limitations. It should not be over interpreted. Nevertheless, it can still provide information about what the amplifier is doing when distorting which can be helpful in modifying the design to improve THD. The simpler "distortion" measurement i proposed (which i did not intend to claim as my own, I just like to take things down to first principles and work up from there thinking we might find a common ground somewhere) may be able to quantify the difference between the input and output but it really does not tell you much about how the amplifier is distorting. To get a better idea of what is going on you need higher order analysis of which THD measurement is one. The trick is to find the correct measurement which captures the difference "you" are hearing. Based on your posts I think you know all this already but i did not want assume it.
I believe that Nelson has at least in part, if not fully, bridged the "apparent disconnect between subjective experience and simple measurements of distortion". Do you agree? (this is my attempt to drag the thread back to directly discussing the article). If not, what measurement would capture the differences you are talking about?
Now, i have a direct question about the article for anyone who feels like helping out. Why does feedback create more complex distortion? Where do those 5th, 6th, and 7th harmonics come from? I assume feedback adds it's own non linearity? Does this nonlinearity arise from the delay?
I agree that THD is one measurement and like any measurement it has limitations. It should not be over interpreted. Nevertheless, it can still provide information about what the amplifier is doing when distorting which can be helpful in modifying the design to improve THD. The simpler "distortion" measurement i proposed (which i did not intend to claim as my own, I just like to take things down to first principles and work up from there thinking we might find a common ground somewhere) may be able to quantify the difference between the input and output but it really does not tell you much about how the amplifier is distorting. To get a better idea of what is going on you need higher order analysis of which THD measurement is one. The trick is to find the correct measurement which captures the difference "you" are hearing. Based on your posts I think you know all this already but i did not want assume it.
I believe that Nelson has at least in part, if not fully, bridged the "apparent disconnect between subjective experience and simple measurements of distortion". Do you agree? (this is my attempt to drag the thread back to directly discussing the article). If not, what measurement would capture the differences you are talking about?
Now, i have a direct question about the article for anyone who feels like helping out. Why does feedback create more complex distortion? Where do those 5th, 6th, and 7th harmonics come from? I assume feedback adds it's own non linearity? Does this nonlinearity arise from the delay?
GRollins said:
The small tube power amps we have built use all filn caps and it makes a BIG difference. These modest little amps have slain some giants.
dave
okapi said:
It does not rise from the delay but directly from the mathematics of the feedback. Assume you have an amp that has a curved transfer function (input to output; if it was a straight line it would have no distortion). Very often that is a cubic curve, something like: Vout = 20* Vin + 0.5* Vin^3. So, the output is 20 times the input (undistorted) plus a little bit of the 3rd harmonic (the distortion). OK?
So far so good. Now, feed that Vout back to the input and again run it through the 20*Vin+0.5*Vin^3 formula. You get 3rd harmonics of the main signal again, plus 3rd harmonics of the 3rd harmonic which gives you a 9th harmonic. Etc. It's quite involved mathematicvally to get it all straight because no amp is always exactly only 3rd harmonic. So, it's easier to measure, see the curve in Nelson's article.
In the article it is also clear that if you make the fb high enough, eventually all harmonics disappear under the audibility horizon. But there is always a tradeoff for the designer: do I put more effort in the linearity of the amp before fb, or do I put more effort in high open loop gain and let the fb take care of things? That tradeoff is part of the art and is why there are so many different amps that all can sound quite enjoyable.
Jan Didden
I have my weak moments when I loose self control and all of a sudden sounds comes out of my mouth. Normally though I restrict myself to instruments and let one of my friends who are a trained singer do the vocal job!! 🙂
I DO care about the people around me. *lol*
/Peter
I DO care about the people around me. *lol*
/Peter
planet10 said:
That's why I said 'most.' My main tube amps use just under 1000uF of polypropylene caps (Solen) bypassed with polystyrene (MIT) and are astounding. The topology isn't all that remarkable, although I did arrange for more drive for the output stage than usual.
The amp puts out something like 130W (4 x 6550 per channel) at about .1% distortion if I recall correctly, but absolutely stomps most other amps. The cost of film caps makes it impractical for commercial outfits to build power supplies that way, but there's no reason you can't do it as an individual.
One of the first things I noticed when I turned on my Aleph 2s was that the image from the Alephs was a foot or two forward of the image from the tube amps. Very peculiar. I traded a few e-mails with Nelson over the matter which ended with him concluding that it was 'a mystery.'
Some aspects of imaging are reasonably well understood. Some are more subtle. For example, a lot of people assume it's all down to phase relationships. While it's true that the predominant part is due to phase, there are other things at play...
POP QUIZ:
You hear a live band at some distance. What happens to the frequency response of the band's sound?
...tick...tick...tick...
Okay, time's up. Hand in your answers.
Oooops! Too many said nothing happens.
That's not true. High frequencies attenuate faster over distance than low frequencies. Don't believe me? Go listen for yourself. The cymbals become mushy and get lost entirely, but the bass drum thud comes through clearly even quite some distance away.
So what's that mean in terms of hi-fi?
Think it through. If a circuit uses feedback, and that feedback produces high frequency byproducts, then the comparative prominence of the higher frequencies brings the image forward. You with me so far? But...now the imaging cues are screwed up. Let's assume all the phase information is presented perfectly. That gives you one impression. But the high frequencies are slightly skewed, telling you that the image should actually be slightly forward of the phase image. As a result, your impression of the total image is a little blurred and less distinct.
For the record, my tube amps and the Aleph 2s I built (don't confuse them with commercial product) have very similar specs. But the image is notably different. Why? I offer the above as a rough guess. If my memory hasn't developed too many holes, the Alephs use around 20dB feedback compared to my tube amps' 10dB or so. Is that the cause of the difference in the imaging? I'll guarantee that it's not the entire explanation but I'd be willing to bet it's a contributing factor.
Jan alludes to higher feedback killing the higher harmonics generated by the feedback itself. Indeed, there are graphs that show low feedback circuits having a certain amount of distortion, mid-level feedback circuits having increased distortion, and higher feedback amps beginning to slowly win the battle, leading to a curve that tapers off slowly towards reduced distortion once you begin using 60-80-100dB of feedback. But then you get other oddities, like that too-silent thing I mentioned above. I don't know how to measure that, but the digital people eventually figured it out. Surely there's some way to get a handle on a similar phenomenon in high feedback amps.
(Jan, do you have that graph--or a link to it--handy? This is one of those picture's worth a thousand words things.)
Let's take another perspective. If, as I suggest, high feedback circuits kill low level information, then you're going to get a different sort of screwed up image. Assume the phase information is intact (I don't think it is, but let's not complicate this). You're going to get one reasonably clear image from that signal. Now, if the intra-note silence is as I say, you're going to get a different set of cues that will vary depending on the recording venue! If it's a bright hall, the loss of high frequency information will make the image seem farther away. If it's a comparatively dead hall, there's less high frequency information to be lost and you're dealing more-or-less with just the loss of the trailing, incoherent echo signal, which makes it seem closer. Either way, you're getting mixed signals, so to speak, and your brain has trouble localizing the image. It's a little blurred and indistinct.
Which turns out to be fairly well correlated in the listening room, although I do not claim that this is the whole story, by any means. Note that I skipped right past the phase question. Throw that in and the waters get murky, indeed. It's not just one question, it's a whole bag-full. A lot of people seek simplistic answers to complex problems. Don't believe me? Look at the socio-political history of the United States over the last eight years. No matter how much some people hate it, there are times when you just have to sit down and think...seeking a more 'nuanced' approach to the whole thing.
Feedback is one such question, though I do not claim that the fate of the free world hangs in the balance.
Grey
