I see. You think amps that measure the same all sound the same. But you think amps that measure worse sound better than amps that measure better. Are you listening to yourself?
Is there actually a difference?
Speakers sound different - that is very obvious. But I’ve never experienced a drastic change between amps. Disclaimer that I’ve never personally ABX tested any extremely high end gear.
With all these articles claiming every other budget amp is a "giant killer", I’ve been wondering if there has ever been blind tests done with amplifiers to see if human ears can consistently tell the difference. You can swear to yourself that they do sound different, but the mind is a powerful thing, and you can never be sure unless it’s a truly blind test.
One step further - even IF we actually can tell the difference and we can distinguish a certain amp 7/10 times under extreme scrutiny, is it really worth the thousands you are shelling out to get that nearly-imperceivable .01% increase in performance?
Not looking to stir up any heated debate. I’ve been in audio for several years now and have always thought about this.
With all these articles claiming every other budget amp is a "giant killer", I’ve been wondering if there has ever been blind tests done with amplifiers to see if human ears can consistently tell the difference. You can swear to yourself that they do sound different, but the mind is a powerful thing, and you can never be sure unless it’s a truly blind test.
One step further - even IF we actually can tell the difference and we can distinguish a certain amp 7/10 times under extreme scrutiny, is it really worth the thousands you are shelling out to get that nearly-imperceivable .01% increase in performance?
Not looking to stir up any heated debate. I’ve been in audio for several years now and have always thought about this.
142 responses Add your response
| Post removed |
Looking at basic specs like damping factor and THD+N doesn't tell you much of anything about an amp. I'd rather have an amp with .01% THD+N with a beautiful negative 2nd harmonic than one that's .001% and all high odd order distortion.The problem is that spec sheets generally don't tell you information like this- so it can easily appear that two amps are measuring the same when in fact they are not. I regard that as the bigger problem- the spec sheets are an attempt to make the gear look better on paper than it really is- a good example of the Emperor's New Clothes. Actually an amplifier that exhibits a 3rd harmonic as its primary distortion component will have overall inherently lower open loop (no feedback) distortion than one that makes a 2nd. |
Indeed. And you and I both know why. Well matched parts. When we talk about measurements here I'd like to think we're talking about more than the most basic facts on the side of the box. To really understand what a distortion figure means it needs to be related to an FFT. I don't think .005% is very impressive or desirable if it's 5th order. |
"sounds better" is very subjective. Totally depends on what one is listening for. If one doesn't really know what real symbols sound like (vs electronic generated), violins, pianos, vs electronic, then how would they know it is accurate? And just to be clear. Unless a component has exactly the same transfer function as another different piece, there is no way it is going to operate, let alone sound the same. The ability to drive difficult loads. How robust is the power supply? etc. You can make an amp that sounds acceptable at low levels driving easy loads. But, when you really need power (not volume), it is like trying to get large flows of water through a small diameter pipe. not going to happen. That is why power supply design to date is still extremely important. This is similar to discussions about wine. lots of people can't tell the difference between a Cabernet Sauvignon and a Merlot, let alone tell you the nuances of the wines. But, a simple test and they will. have them taste a wine, then have them smell certain items that are found in the wine, like earth, certain fruits, etc. Then ask them to taste the wine again. The majority that I've seen all of a sudden can taste these things in the wine. Why? because they now know what to taste for. Same for music reproduction. play live unamplified music, then recorded music with the same instruments. or even music they know well. See if they don't tell you the instruments sound not quiet real on their recorded music. if, the system or certain components aren't quite up to task. To me, it totally depends on who's listening and what they are listening for. enjoy |
There is no doubt in my mind that I could tell the difference between the Pass labs amp that is my main go-to amp and my previous go-to amp the Aesthetix Atlas. There are of course amplifiers that sound very similar. That would not be those two. Both very fine amplifiers and I still go back to the Atlas every now and then. So undoubtedly. The differences are not subtle. |
You eliminate yourself from having any experience with the high end, so that means you have little frame of reference to ponder amp differences. I have tons of experience and can safely say that they all sound different. If the difference means you want to listen to your system every chance you get, then an amp is worth almost any price to some. If you find yourself choosing to do other things, rather than listen? Well, that’s crap! In absolute terms there may be small differences, but the enjoyment gap could be huge. |
The ability to drive difficult loads. How robust is the power supply? etc. You can make an amp that sounds acceptable at low levels driving easy loads. But, when you really need power (not volume), it is like trying to get large flows of water through a small diameter pipe. not going to happen. That is why power supply design to date is still extremely important.The ability to drive difficult loads doesn't seem to be helpful to any amplifier, because its easy to see on the spec sheet and by listening to the amp on a hard load that its making more distortion. The problem is, such distortion is audible as brightness, harshness and a loss of detail. You can take any amplifier and just by putting it on a higher impedance/easier-to-drive load, it will immediately sound smoother with more detail. That is true whether its tube, transistor or class D. You can both hear it and see it in the specs- this is one of the few areas where the ear and the specs agree. I do agree though that robust power supply design is one of the things that set high end amps apart from the mid fi. |
I do agree with the information that some of the designers and builders have said. I have mentioned the Lafayette my 550 tube amp many times before mainly because of the power supply. I have a friend who buys used gear all the time from estate sales. From hi xx amps to old Sony maranta kenwood etc. I can hear the differences between every one of them. Not very hard. I also repair gear. And do a lot of counterpoint amps. I can hear those differences very easily. Go figure. Agree with matching. The counterpoint sounds better with matching they ipolar transistors and the fm acoustic amps sound great for ss amps because of the parts matching. Happy listening |
Indeed. And you and I both know why. Well matched parts. When we talk about measurements here I'd like to think we're talking about more than the most basic facts on the side of the box. To really understand what a distortion figure means it needs to be related to an FFT. I don't think .005% is very impressive or desirable if it's 5th order.+1 Yes- you aren't going to get much in the way of parts matching when the amp is built to a 3-figure price target. |
Asianator, Best way is bring amp to your home use device to calibrate sound levels. Have a friend hook up different amps covered so you don't know what is hooked up. Take as long as you want and decide what you like. Usually tube has good presence but loss of fine detail. Solid state sounds more tight but very detailed. I have found tube pre with solid state application works best. Yes most important is the speaker and how it works with your room. |
Usually tube has good presence but loss of fine detail. Solid state sounds more tight but very detailed.I find that tubes generally have more detail than solid state, in particular when it comes to low level detail. Its the low level detail that is the argument for tubes. Solid state by contrast is brighter (due to higher ordered harmonic distortion), and often people mistake that for detail. |
Can you explain that? The same phenomenon that give rise to various orders of harmonic distortion exist with both tubes and transistors. Well matched tubes will tend towards higher odd order harmonics just like transistors do. It seems to me that tubes as gain devices get a lot of credit they don't deserve because they're typically used in much simpler topologies that's should get the credit, but don't. |
Yes I can. The short explanation is that tubes generally are more linear. Semiconductors have a significant non-linear input capacitance which as any solid state designer knows can be tricky to deal with (in fact radios are tuned by the use of varactor diodes, which take advantage of this capacitance), and often requires the application of loop negative feedback to correct the non-linearities that result. In a nutshell, the more linear the circuit, the better it is at low level detail. It is true that The same phenomenon that give rise to various orders of harmonic distortion exist with both tubes and transistors.But it is also true that tubes have less of that phenomena than transistors, which is why the odd ordered harmonics to which you refer are at a much lower level in all tube circuits. That is the basis of the tubes/transistor debate, since that lack of higher ordered harmonics is why tubes sound smoother. It really is too bad that Sony sat on the SITs, if you will pardon the expression :) They really did offer the solid state replacement for a tube! |
Nobody builds JFET amps because nobody makes power JFETs. Nelson had SemiSouth produce power JFETs to his specs for his amps and the kits he sold, but SemiSouth went bust. That's why the vast majority of his designs use good old fashioned MOSFETs. I'm not sure input capacitance is an insurmountable issue. Tubes certainly have their technical drawbacks to be overcome. I'm not one of those who automatically cringes at feedback. |
There are so many variables that can screw up such a test. The speakers being a big issue. Some speakers are dark, dark speakers will hid variations. If you find a good full range driver, you have no crossover issues, and assuming that the speakers are articulate, these should make it easier to identify differences in equipment assuming that the high end gear is truly articulate, expensive equipment doesn't guarantee that the equipment is articulate or neutral, it only guarantees that it'll cost a lot. In my system non audiophiles can consistantly identify different gear, be it amplifiers, most cables, etc. Mind you, they don't always prefer the better (more neutral) items. Just as some audiophiles prefer dark sound some neutral, and some prefer bright equipment. I'm sure that some people couldn't tell any differences, but I haven't found anyone who didn't hear changed, when there actually was a chance. Then again, I use Teflon V-Caps, or Russian caps, Mundorf Silver Gold Supremes, and even particular resistors, so my system is tweaked to the max. Only my Yggdrasil is stock. |
Why doesn’t someone else use this approach??No-one makes SITs. They can be a bit fragile too! I'm not sure input capacitance is an insurmountable issue. I'm not one of those who automatically cringes at feedback.I'm not either, but one thing should be abundantly clear at this point- even though tubes have their weaknesses, they are a lot easier to use to build a circuit that makes less of the distortions that the ear finds particularly offensive. Funny that no-one has stepped up to the plate to fix what's wrong with transistors! If the industry had a weighting system on the harmonics produced by an amplifier, we'd probably see some change. We've known since the 1930s how much more sensitive the ear is to higher ordered harmonics than it is to lower orders, but little has been done about that in terms of circuit design- we've actually made it worse! That is why we are having this conversation right now: in a nutshell, the industry doesn't **want** to do anything about it, because it *costs too much money*. So it puts its head in the sand, hoping somehow our ears will change despite millions of years of development :) For example, if there was a range of SIT devices; signal, driver and various output devices, tubes would be gone. All done- history. But that isn't going to happen, so in all likelihood, if you have grandchildren that care about audio, they will be having this conversation decades from now. |
Just for those unaware, SITs and VFETs are just power JFETs. There's always laterals, but nobody is doing much with those. Kinda like nobody did much with the big JFETs. The solutions to the problems transistors have has been solved and pretty much nobody really cared. Is this higher order distortion really a big issue? I'm not sure it is when it's kept at or below a low noise floor. The marketplace doesn't seem to think it's an issue. |
In my opinion, it would be VERY easy to set up a double blind test where people would be unable to hear differences in various equipment. I would love to find a skeptic to work with where I pick the speakers, and the high end equipment. The speakers I would use would be 8 ohm, and over 90DB at 1 meter, not a difficult load, and very efficient. The difficulty comes into play with the DOUBLE blind part. Switching between 2 systems, level matching them, keeping the testers in the dark, and being able to immediately switch back and forth between gear so as to assist people in hearing differences would be the most difficult part. Anything truly random could have you listening to the same system after hitting the random selector switch 20 times before hearing the other system. I think a double blind start, and then allowing the listeners to swap systems at will would make far more sense. Honestly though I have strong suspicions that the naysayers have been allowed sufficient free reign to allow them to game the system. I think that the most important part of the test would be ensuring that the speakers are extremely articulate. I believe that there are many speakers out there that would easily mask differences. Other announces such as acoustically lively rooms, as in tiled floors, or excessively dead rooms, as in rooms with heavy carpet, and foamed walls could have an effect. I am not a room treatment guy, I figure music is played in imperfect rooms, but I do believe that in a double blind test too hot or too dead a room could color the results. I'd really like to be involved in one of these tests before ascribing any merit to them. I think that they could too easily be rigged otherwise. |
Double blind is meaningless to me. If you were to participate in a test of that type, I think it would be invalid because your brain would be racing trying to hear differences, and doubting itself. There are so many unknown variables about the way a brain processes an event like that. I believe that there are amps with sonic differences that are very easy to hear, and others that are more subtle, but it is best to listen to any amp over a period of weeks with all different types of the music you like so that your mind can build a profile of the gestalt, and whether it is right for you. |
The only reason double blind tests exist is so naysayers and non-believers can say, “I bet it can’t pass a blind test.” Obviously any test is fraught with all manner of complications and error so you can’t really say a single blind test has any real significance, especially if the results are negative. |
Dave_b, taking time to get a handle on a system's sound isn't the same as being able to hear differences. Before you take a hearing test, you are supposed to have had 8 hours of only low level noise, that can impact an A/B test. OTOH, if someone can't tell the difference between my amp and preamp, given a very good source, and articulate speakers, something is wrong with their hearing. To know what my system does well and does poorly would require more extensive listening, IMHO. I could be wrong, but my experience says otherwise. |
Dave_b, only that when doing an A/B particularly with consumer vs transparent articulate equipment, listeners ought to be able to identify each, though which they prefer would be a matter of taste. As for us, when we do something to our system it's often a small tweak, and it can take us an extended period in order to fully grasp its total effect. Often we can immediately hear a difference, but some components need to burn in, and even if not, it's often not immediately obvious the extent of the changes. So in one context I agree completely, but if you were referring to an A/B with consumer quality gear, and truly high end gear, then I disagree. |
Just for those unaware, SITs and VFETs are just power JFETs.SIT stands for Static Induction Transistor. VFETs are SIT devices. Power JFETs are a different beast:https://en.wikipedia.org/wiki/JFETThere's a lot less info on SITs:https://en.wikipedia.org/wiki/Static_induction_transistor But there's a pretty good thread about Nelson Pass's rework of the Sony VFET circuit in DIYAudio.com. The kit that he came up with was excellent- all sold out now... It is one of the best solid state amps I've heard. |
That's not at all in agreement with what I've been told by Nelson. When I asked about the viability of adapting the F5 to power JFETs he pointed me to an example he'd built using the Sony devices. SITs, VFETs, and JFETs are all vertical junction field effect depletion mode devices with triodic conductance curves. I've always seen them generically called JFETs unless somebody was talking about a specific brand of device. |
Interesting discussion, which I am joining late. I recently (about a year ago) purchased a pair of Magnepan 1.7i’s These speakers are not particularly easy to drive, so I knew I would need a few watts. Having blown my budget on the speakers, I went searching for a bargain amp. I found a Parasound H2200 (300+ w/pc rms) on ebay for about $400. This seemed to have plenty of power, but the speakers did not sound like they did in the dealer’s listening room. The dealer was AR dealer and he auditioned the speakers with some pretty high end tube equipment. I tried a couple of different SS pre-amps, but was not satisfied. Finally, I decided to try a tube pre-amp. After doing a bit (a large bit, actually) of research I decided to try the Transcend 6SN7 preamp (can you guess what kind of tubes it uses?). Wow!! Now I can’t say if they sound as good as they did powered by all that Class A tube equipment, but this seems to be a good pairing. The Aric Audio gear really opened up the sound of the Magnepans. Maybe someday I will try some Class A Tube power amplification, but I don’t know how to compare the numbers (How does 68w Class A amplification compare to 300w Class A/B under a difficult load?) Anyway I am satisfied for now. And my next $3000 expenditure will be for a nice (fairly) high end double boiler Cappuccino machine. Priorities. :) |
| Post removed |
| Post removed |
How the amp performs depends very much on your speakers. It it is easy to slap a pretty chassis and spend money on marketing and call it a Lamborghini of AMPs. The exponential curve of SPL vs power requirements make amp design a trade off. Also a trade off for listener is like speakers, how far to sit from them and levels one likes to listen to etc. It is not very difficult to A/B test in your home. Just have an app for SPL match after swapping the cables and listen to the same track. Again in you may not like the AMP that performs well :), due to your sound preference. |
