Atmasphere,
I suspect it would be hard to set up any meaningful comparison of an amp alone in that an amp cannot produce music alone without speakers. The results could vary widely depending on speakers used, personal preferences, and all the usual culprits that result in different strokes for different folks.
My gut feel is that as you say, its all in the implementation and there are many ways to make tasty soup. |
That is true. However I have found the if given the option, people prefer a system that it not bright and harsh, given otherwise that there is no lack of detail, bandwidth and with no tonal aberration.
IOW, a proper stereo should lack loudness cues, such that you can approach the same volumes in your room that the real live music could. A typical orchestra can do 115db with ease, yet many audiophiles will not turn up the volume past 95 db because its 'too damn loud' or they get 'turn that @#$% down!' from their S.O. This mostly due to artificial loudness cues which are totally coming from distorted odd ordered harmonics, and by that I mean only 100ths of a percent.
Its been my contention that negative feedback is a major culprit, literally violating one of the most fundamental rules of human hearing: how we detect the volume of a sound. |
Hard to disagree with Ralph's last post in my opinion. |
"Its been my contention that negative feedback is a major culprit, literally violating one of the most fundamental rules of human hearing: how we detect the volume of a sound."
Maybe. Maybe not. I've got a Music Reference RM-200II, and there are some compact discs (never records, it seems) in which I can only turn up so loud before my ears bleed, and others I can turn up without any irritation.
Then you factor in the speakers, and their ability to generate clean sound at the levels you are talking about.
Seems to me the only way to make that determination is to take two identical amps, one with negative feedback and one without, and play them with the same source through the same speakers. I've not heard of anyone who has done that exercise, much less anyone who has heard it. |
Is the 115dB at a mid-hall seat, or is that the level on the stage? Apparent sound levels in a concert hall are far more influenced by the indirect sound level than in the home environment. This is due to the enormous difference in volume between a concert hall and a normal audiophile listening room (say 200-300 cubic meters vs. 20,000-30,000 cubic meters). In a home environment you listen to predominately direct sound that has nor been tonally shaped (highs rolled off) and homogenized by a concert hall volume of space. Hence it sounds louder than the exact same measured SPL in a concert hall. It really has nothing to do with amplifier design. |
The Bel Canto ref1000m2 Icepower Class D amps I am using on my larger full range OHM Walsh omnis seem to be able to go as loud as desired without an edge on most good recordings.
These amps are high damping factor which matches well to the OHMs, so I suspect there is NF applied, but not sure. If loudness cues are more present as a result, I'm not sure that it is significant enough to matter.
I also use Triangle and Dynaudio monitors on teh same system. Monitors in general will not fare as well with orchestral music at high volumes (at least not without a sub of some sort to offload much of the work), so I can say that the speakers are definitely a factor as well. |
Ralph Atmosphere said, " A typical orchestra can do 115db with ease, yet many audiophiles will not turn up the volume past 95 db because its 'too damn loud' or they get 'turn that @#$% down!' from their S.O. This mostly due to artificial loudness cues which are totally coming from distorted odd ordered harmonics, and by that I mean only 100ths of a percent."
I ask do you listen to your system at this level?
Bob |
"These amps are high damping factor which matches well to the OHMs, so I suspect there is NF applied, but not sure"
Mapman - Not only NFB is applied but it is mulitiple (two) called Mulivariable Enhanced Cascade Control. It is in the Karsten Nielsen doctorate work available on the web (Icepower). Feedback doesn't have to be deep since class D amp has inherently low output impedance even without feedback. Speaker is always connected to +Vs and GND and only direction changes (very low resistance Mosfet Bridge). Output impedance increases with frequency but DF is still about 1000 at 1kHz. |
Jamesgarvin I had a pair of Manley ref 440/200 monos that had adjustable slopes and feedback to custom taylor the sound;there were noticable differences but I did not consider it annoying just different. |
A few years ago, we had a regular and good contributor (be nice if he came back) Ar t an amp designer, suggested that in his typical ss designs he would try to keep NF to a minimum, but in his switching amps though there was much more NF, it didn't seem to be an issue. I wonder if it was due to the NF in the switching amps occurring at a faster rate? |
Unsound - Feedback in Icepower amps is called Multivariable because it consists of two feedbacks controling voltage and time.
I suspect that feedback in class D is less evil for few reasons:
Amp cannot become unstable (oscillate) since it is already oscillating.
Response can be very fast limited only by Mosfet's max current and resistance (sort of Hysteretic converter)
Voltage feedback is shallow because duty cycle is more linear than transistor characteristic(less feedback required) and also because "time feedback" already corrected most of nonlinearities. I suspect that voltage feedback is helping to deal with load regulation. Early class D amps (Tact) had no voltage feedback at all and were sensitive to load conditions.
Momentary saturation of output stage (charge trapped at the junction) that happens in bad cases of class AB is irrelevant since time (duty cycle) and not the voltage is analog quantity.
|
When I first heard the IcePower amps in my setup, the difference in the nature of teh sound from anything I had heard prior was most startling. It was most apparent in that major volume adjustments seemed to have comparably little effect on volume, though the SPLs were clearly increasing as expected.
My observation of how the IcePower amp sounds as volume is increased seems consistent to me with what Ralph describes as the benefit of not using feedback and loudness cues. Most really good systems I have heard over the years tend to behave this way more or less I believe.
At the same time, Icepower applies negative fedback clearly. So I have to conclude that the IcePower amp has achieved a lot of teh benefits that Ralph attributes to no feedback in its specific implementation.
If so, another reason why Icepower/Class D can be considered a major innovation in home audio along with the other advantages. |
Kijanki, I could be wrong about this, and perhaps it doesn't matter, but, I don't think Ar_t 's amps were ice amps.
Have the TacT amps changed since then? |
Acoustat6, yes, sometimes I do. Most music does not demand it but some does: The Verdi Requiem, on the Soria box set (RCA) is a good example of vinyl being put the limits, and stereos too. If you play the quiet spots at the correct volume, the peaks will be at 115db.
So- to answer another question- 115 db peaks occur front row center, equivalent to where the microphones are usually placed. When you are in the orchestra, you don't get the same volume exposure that the audience does unless you are in the percussion section.
Some of the class D amps do not have feedback as well as some transistor amps as I have mentioned. Some of the class D processes allow for time manipulation, which is a fancy way of saying that they have a way around this issue. I don't doubt that this may be part of why class D amps are challenging the traditional transistor art.
However the artifacts I am talking about also occur at much lower levels. BTW, it is odd ordered harmonics that allow SETs to have the 'amazing dynamics' that are so often associated with them despite their low power levels. When an SET is at low power, it makes no odd orders at all, but when you push it over about 50% of full power or so, they begin to show up. Since it is the transients that take all the power, that is where the loudness cues are now occurring too. As a result, an SET with low power might seem to play so loud that 85 or 90db seems like a lot. This is the effect of artificial enhancement of the natural loudness cues.
As a result I am careful about using the word 'dynamics' because so often when I hear audiophiles use the term they are really talking about 'distortion'.
If you can't be in the room with musical peaks at 100 db (if its uncomfortable or sounds too loud), then its very likely that the odd orders are being enhanced. |
Unsound - I remember Ar_t posting on the subject of switching amps. I don't know about TacT now. At the time I bough my Icepower based amp I read opinions on Icepower and Hypex. Both configurations (Full Bridge vs Half Bridge) have good implementation for instance Hypex based Channel Island amps.
Mapman - I also observed that listening at low volumes with Icepower is very good. I know that part of the problem is our hearing but I enjoy very good and clean sound (with very articulate bass) at low levels without any loudness correction. I cannot explain it. Do you experience the same? |
"Mapman - I also observed that listening at low volumes with Icepower is very good. I know that part of the problem is our hearing but I enjoy very good and clean sound (with very articulate bass) at low levels without any loudness correction. I cannot explain it. Do you experience the same? "
Yes, bass is very articulate at all volumes. That and the loudness behavior I described above are perhaps the two biggest differences I hear on my system with icepower compared to prior SS amps.
The loudness perceived compared to actual SPL was most dumbfounding at first especially in that teh IcePower amp power spec is 4X what I had before (500w/ch into 8 ohm compared to 120w/ch prior). Its almost like the additional power is utilized to flush out the music, kind of like blowing up a ballon, as the volume goes up yet the perceived loudness does not increase so much.
Kijanki, one other thing, with my particular speakers, the high current and efective doubling of power into 4 ohms helps produce a natural balance in the bass which also helps at low volume whereas with some amps, the bass is relatively leaner. Toss in teh articulation in the bass provided at all levels then and you have quite a winning combo that not only delivers a detailed low end, but also allows higher frequency details to come through to boot. Very satisfying to listen to! Articulate, detailed and "musical". Very hard for me find fault at present, which is a very nice place to be! |
The loudness perceived compared to actual SPL was most dumbfounding at first especially in that teh IcePower amp power spec is 4X what I had before (500w/ch into 8 ohm compared to 120w/ch prior). Its almost like the additional power is utilized to flush out the music, kind of like blowing up a ballon, as the volume goes up yet the perceived loudness does not increase so much. This sounds like there is less odd-ordered harmonic generation. |
Atmosphere said "06-25-10: Atmasphere
Acoustat6, yes, sometimes I do. Most music does not demand it but some does: The Verdi Requiem, on the Soria box set (RCA) is a good example of vinyl being put the limits, and stereos too. If you play the quiet spots at the correct volume, the peaks will be at 115db."
Hi Ralph, If you listen to the Verdi Requiem at peaks to 115db (which is an impressive level for any system) then, and since we all know that rock is supposed to be even louder than an orchestra, how loud would you listen to Cream live or other "loud" rock?
Because you said, "IOW, a proper stereo should lack loudness cues, such that you can approach the same volumes in your room that the real live music could."
Then just how loud do you listen to rock on your system to approach what a live rock concert does. Its gotta be louder than a recording of Verdis Requiem!
Bob |
Acoustat6, that's true. My system can play so loud that prolonged exposure can result in damage. But playing rock as loud as it is often performed has a bugaboo: most rock recordings are anything but live! Quite often the guitar amps have only 15 watts in the studio, so who is to know how loud such a recording is to be played.
I have a white-label Vertigo press of Black Sabbath's 2nd LP (Paranoid), which is an amazing recording and one that can bring most systems to their knees in a heartbeat. You play it loud, but even that one is hard to tell how loud it should be played.
I play in a rock band, and recently we did a memorial show where we were the only band on the bill that was not metal. The club we were playing in was a metal club. It was on that night that I discovered that metal bands don't play all that loud. The most powerful guitar amp we saw that night only made 25 watts. They rely on the PA.
So- how loud is that supposed to be? Rock is always tough because there is no good answer for it. |
Atmasphere said "My system can play so loud that prolonged exposure can result in damage."
I assume you mean damage to your ears, not your system. I agree 115db is too loud as we do not want to damage our ears. Also I think that for audiophiles in general that level is very difficult to obtain. IMHO it is an unrealistic goal.
But you say, you play the Black Sabbath loud, so...
The Black Sabbath Lp you said you play loudly then is it louder than the Verdi or is the Verdi LP the LP that you play loudest? In other words, is the Verdi LP a test lp for "you want to hear loud?" then this is the loudest we go?
I am always on the lookout for LPs that streatch my systems limits, it sounds as if thats a great test LP for a systems ability to play loud. Though I dont get anywhere near 115db!
Bob |
If you get the original Vertigo press there is a lot of bass and bass impact! If you try to play that anywhere near 105 or 110 db, which is justified by the material, most systems just can't do it- too much energy required. The trick it to be able to play it that loud without it **sounding** loud, IOW the only sense of volume should come from the LP itself, not the system. |
Hi Atmasphere, Thanks for the advice. I have several "system busters" Lps. Certainly Daphos is high on the list along with Carmen suite on RCA and a few others that have very big dynamics.
I am curious as to your system, what does it consist of? There must be some nice OTL amps, but what else do you use? I dont believe I have ever saw references to your system speakers etc. Could you post a few pictures of your system here on the 'Gon. Although I would expect that it is constantly changing. Is this system in your house or at work?
Thanks,
Bob |
Hi Bob, the system I have at home uses a pair of Classic Audio Reproduction T-3s with dual 15" woofers and field coil drivers in the midrange and tweeters. They are about 97 db.
When we first got them I noticed that phenomena of loudness cues: we were playing them a lot louder than the previous speakers, which were 89 db. Now of course you can imagine that one might do that with a more efficient speaker because it is possible, but really the limiting factor with the less efficient speakers was that the loudness cues made it uncomfortable to turn it up the volume any higher. The CALs in the system allowed the amps to make less power/less loudness cues so we were playing them louder very naturally and effortlessly. |
"Now of course you can imagine that one might do that with a more efficient speaker because it is possible, but really the limiting factor with the less efficient speakers was that the loudness cues made it uncomfortable to turn it up the volume any higher."
Hmm, I'm under the impression amp clipping generally becomes more of an issue at higher volumes with less efficient speakers?
I'm assuming that you run these with one of your tube amps, which would minimize loudness cues when distorting? All amps distort more at higher levels, correct? The question is in what way and to what extent the distortion is audible. For a given amp, this would occur sooner at lower volume with a less efficient speaker, all other factors aside, right? |
Right. We were careful to not allow the amplifiers to clip during these tests. However it really did not matter which amp we used or in what combination. The higher efficiency of the speakers meant that amps were not working as hard and so were more relaxed.
That was 12 years ago. We've made good progress reducing distortion since then, so the amps have continued to get more relaxed. |
06-25-10: Jamesgarvin
... "Seems to me the only way to make that determination is to take two identical amps, one with negative feedback and one without, and play them with the same source through the same speakers. I've not heard of anyone who has done that exercise, much less anyone who has heard it."
Hi Jamesgarvin;
Interesting. However, they could not be identical amps, even supposing that they only differed regarding feedback, but I assume you meant it ideally.
Practically speaking, we can, of course, choose one of each type from a number of respected designers and begin there.
When you referred to playing them with the same source, I'd like to reference "source" as recorded source.
Using acoustic instruments as an example:
If we are present in the recording space during the performance and we agree that the sonics are excellent, we'll have to verify that all components: mics, mic pre-amps, console or summing equipment, power-amps,
are non-feedback - as well, the same devices with feedback, We'd have to start there.
If this could be accomplished, we'd still have to consider
Class A, A/B, D, discrete/chips, SS/tube, etc.
We still haven't discussed what reference speaker system.
Considering all of the above, its amazing the amount of effort that drives the talented designers and enthusiastic lovers of audio and music.
For those concerned, having a very satisfying audio system is natural treasure.
I'm glad to see those efforts displayed here.
|
