Yes, I know, Jan, I complemented you then for that, because I had always been wondering whatever happened to the forgotten half of the dynamic duo.
As stated in the inteview, Otala spoke English better, so he got to present the paper, when, as I understood it, it was in fact Lohstroh who did most of the initial designing.
I enjoyed that text tremendously. 11 out of 10 for it.
That's another description of the sound that always my goal. Except, it has nothing to do with valves, SS, or even opamps - it's merely what correct sound comes across like. Talking about volumes and tonality changes, that 3D quality should always be there, from the tiniest whisper of a sound, to room shaking intensity. Dejan's comment about not needing the tonality headroom doesn't compute for me, because IME that is a key requirement for audio that has the live quality - music like high energy rock just doesn't work, doesn't hold together unless the tonality headroom is there; again, the goal is for the system to run at the point of clipping, with full integrity to the sound, and then be able to drop the volume steadily down to a whisper without the tonality of the whole subjectively changing. Very, very rarely I come across other systems with this capability - they are signposts of the way to go, IMO ...
Just for fun, and as an example of the process, I kicked off my PC setup this morning, from dead cold, with a high energy rock compilation, Foo Fighters, Cold Chisel, that type of thing. At deafening volumes, someone coming into the room would have to shout, and I still wouldn't have understood what they were saying. At first, a congested dense mess, the drums were a disaster - now, the strong conditioning has steadily freed up, stabilised the important elements, and big acoustic sound is building, the drums are sitting in big spaces with all the echo making sense; all the artificial reverb hangs together, the cymbals are sounding pretty decent - I have, depending on the track, at times quite a vast acoustic hanging over the little speakers on the other side of the room, which then fills the rest of the listening space.
Edit: For the Aussie crowd, just went through "Will I Ever See Your Face Again" - and, no, I don't do the "extra" lyrics,
- I like this track for testing because the rhythm(?) guitar riff in this has a fantastic driving energy to it, something a system has to get right ...
Edit: For the Aussie crowd, just went through "Will I Ever See Your Face Again" - and, no, I don't do the "extra" lyrics,
- I like this track for testing because the rhythm(?) guitar riff in this has a fantastic driving energy to it, something a system has to get right ...
Last edited:
Ahhh, me ol' buddy's back - good to see you again!
Seriously, it was a worthwhile exercise, because I just switched to Ella Fitzgerald Gold, all the detail was there, tape hiss, plenty of acoustic, but the late 50's mellow tone was not quite in place, the voices had an "edge". Turned out to be still a bit of interference picked up on one of the mains spurs, which once identified was quite easy to null out.
Seriously, it was a worthwhile exercise, because I just switched to Ella Fitzgerald Gold, all the detail was there, tape hiss, plenty of acoustic, but the late 50's mellow tone was not quite in place, the voices had an "edge". Turned out to be still a bit of interference picked up on one of the mains spurs, which once identified was quite easy to null out.
The parameters in your list there are not sufficient to synthetically replicate the sound of an instrument or the human voice.
Likewise, in the same fashion, they are nowhere near sufficient to measure the unique sound of an instrument or the human voice.
The unique sound we can recognize of up to thousands of different instruments is not very much within the perception of low / high distortion, low / high noise, flutter, wow, jitter or frequency response.
The sonic memory recall of thousands of instruments is immediate and stays with us forever.
I've seen your list on the internet before, written by someone called Ethan Winer.
It's not a complete list. A limited list is limited vision.
It doesn't even have crosstalk in it. Is that science?
The word science comes from the latin word scientia, which means knowingness.
I know that list is very limited.
What is so special about them?
There isn't even an academic paper which has shown the difference between mp3 and flac.
It's just random never-do-wells on the internet which can show that.
I'm sorry you misunderstood my note. What I was saying was that if you do not meet those minimums all the other audiophilia that get agonized over in these fora is in question. They are known aural sensitivities. Comparing two amplifiers with different frequency response and hearing a difference proves nothing more than that the difference in frequency response is audible. Same for level, distortion etc. It can be pretty sobering how much a small response difference will color the perceived sound.
"The parameters in your list there are not sufficient to synthetically replicate the sound of an instrument or the human voice." There is something like 2.5 billion people on the planet that can recognize a voice and understand the words over a 3.5KHz band with relatively high distortion and poor SNR (telephone). Many have had no problem recognizing an instrument or a specific performance even reproduced using an acoustic phonograph. I think you need a more specific description of what you are trying to say. I would submit that auditory clues alone won't accomplish your goal.
"The parameters in your list there are not sufficient to synthetically replicate the sound of an instrument or the human voice." There is something like 2.5 billion people on the planet that can recognize a voice and understand the words over a 3.5KHz band with relatively high distortion and poor SNR (telephone). Many have had no problem recognizing an instrument or a specific performance even reproduced using an acoustic phonograph. I think you need a more specific description of what you are trying to say. I would submit that auditory clues alone won't accomplish your goal.
For amusement, and possibly to give a few people a fright, there are these: Welcome to TheLoudest.co.uk - 2000w 6x9" Speakers.
They may actually deliver quite decent sound, would be an interesting experiment ...
They may actually deliver quite decent sound, would be an interesting experiment ...
And that's just what Lohstroh said in the interview. I understood the whole tone as him saying it was a joint effort.
Frank, I made the comment exclusively for the setup I use, i.e. my speakers and the Marantz combo. Because of the small room, I don't really need more, but put me in a bigger room where the speakers need to try harder and I will need more power.
And I will know soon enough, because by the end of August, I will be moving into a bigger room.
Lastly, I beleive maintaining tonality is not linearily dependent on the power available, I think it's the design that does it or doesn't. As an example, my toy, the Toshiba/Aurex SB-45 integrated amp, rated at 40 WPC, also pulls off the same trick until it goes very near to its absolute limit.
I see, I didn't follow correctly, I'm sorry.
Perhaps I was just thinking of the Ethan Winer paper, which I looked up now and it says......
"only four parameters are needed to define everything that affects audio quality: Noise, frequency response, distortion, and time-based errors."
Yes, the frequency response needs to be identical of course, before we start comparing DAC's, amplifiers, capacitors or cables, for sure!
We can rarely equalize speakers or transducers, so it's important to 'listen beyond' the frequency response in that case to get an evaluation.
I've been listening beyond FR for a while, at least I try to, some other people only listen to FR. I've even seen speaker designers say that FR is all that counts. A few of my thoughts on speaker measurement here
http://www.diyaudio.com/forums/full-range/258682-cone-preferences-paper-alum-titanium-poly-etc-3.html#post3996034
If there is a difference between cables, it's not in the FR. Alright, sometimes cables sound different and that is due to FR, but that is not what we are looking for right? I think we are on the same page.
Interesting answer.
Yes, despite the high THD, high SNR and very limited bandwidth we can still recognize a voice on a phone quite easily.
We have evolved to recognize voices, single out voices in a crowd known as the cocktail party effect http://en.wikipedia.org/wiki/Cocktail_party_effect, recognize animals in the forest, birds in the sky, recognize sounds like different metals, judge sound distances and recognize musical instruments and their quality with high skill, since music is hard-wired into us......
"Studies have shown that the human brain has an implicit musical ability"
http://en.wikipedia.org/wiki/Cognitive_neuroscience_of_music
We are not hard-wired with certain instruments or certain voices, we hear them and then we remember them. If you were to hear a violin a decade later in a blind test, you'd still recognize it is a violin, even different kinds of violins if you've been exposed to that. No visual clue is necessary, only the raw data.
No clues or hints, you simply strike any note like middle C at the same pitch and volume on various violins, pianos, synthesized instruments et cetera.
Thus, sonic memory is not "a few seconds" which I see people write over and over. I just looked up the Ethan Winer article, yup it's in there
"human auditory perception and memory is so short."
He continues to write
"With A/B testing - it is mandatory that the switch be performed very quickly."
Is it? Why?
A quick switch is only necessary for volume and FR difference, perhaps pitch.
A few people have perfect pitch, they can say "that is A#" or even hear a half note difference. Normal people need to hear a few pitches near each other to know it, like a scale.
That's volume, FR and pitch. Few humans have perfect pitch or perfect volume, perhaps since it's not very interesting or useful, I'm not sure.
Instead what we have is perfect tonal colour, we also have an excellent and long-lasting sonic memory for music, rhythm, completely nonsense patterns, nonsense waveforms and nonsense noise.
An example is, you only need to hear something like half a second of a piece of music or less and you can instantly say which song that is, at times.
I remember competitions on the radio, when they would play some very short nonsense noise, then listeners had to call in and say which song it was to win a prize.
Now, is it the case that these human skills transcend into audio equipment?
Let's start with speakers / transducers.
Yes.
They are voices or instruments as well, so to speak.
Their task is to reproduce but they always have a slightly distinct sound, in my experience, rather than transparency.
How do we assess the tonal colour of a speaker driver?
- Time character
- Spectral character
I.e.
- How does the attack and decay look.
- How does the distortion spectrum look.
Then there is waveform analysis such as sawtooth and square-wave to help with the analysis of the distortion spectrum and transient handling.
That is all I know about the tonal colour side. Perhaps volume shifting efficiency as well? Which you could call tremolo.
There are more differences of course like dynamic, electrostatic and isodynamic drivers, I am not sure where these differences are accurately measured.
Then there is acoustic space like the size and distance of the drivers.
Moving forward. Does audio equipment such as the DAC and amplifier have tonal character?
If it does, I suppose it's in the time and spectral character as well then.
In a paper I linked yesterday, ceramic capacitors had horrible EMR, electro-mechanical resonance. I am not sure where EMR and ESR fit into this picture.
I am not crazy, I am just trying to cover all corners, to know about all corners.
I am quite sure that for example NE5532 and AD8610 sound different and the difference is pretty easy to hear directly, just like I can hear an instrument directly.
I can only assume the difference is within time and spectral.
I can only assume 0.1% is too strict, or the difference is time.
Unless we move into the realm of "Leaded solder" versus "Silver / Tin solder" and how that can cause a sonic difference? If it does and escapes measurement then perhaps there are secret Silver op-amps.
Kthx for reading or any comments.
Last edited:
Kastor, I think you misunderstood Demian's list.
It was never supposed to a wholesome compilation, it was just a reminder of a few elements.
I asked Demian for help. So I sent him an initial version of what I wanted. As sent to him, it had already fulfilled a lot of criteria points, but Demian sent me a list of 19 points he thought I should address differently, or in greater depth, if you like.
Some were obvious, others made me wonder why didn't I think of that, but others required me to do serious research for my own benefit, which I am still doing, over a month later, as I was less to much less familiar with some of them.
THIS list was what Demian was referring to, not any absolute list one needs to address if he wants something above average.
It was never supposed to a wholesome compilation, it was just a reminder of a few elements.
I asked Demian for help. So I sent him an initial version of what I wanted. As sent to him, it had already fulfilled a lot of criteria points, but Demian sent me a list of 19 points he thought I should address differently, or in greater depth, if you like.
Some were obvious, others made me wonder why didn't I think of that, but others required me to do serious research for my own benefit, which I am still doing, over a month later, as I was less to much less familiar with some of them.
THIS list was what Demian was referring to, not any absolute list one needs to address if he wants something above average.
DVV, no offense intended, and I don't want in any way to question your perceptions - they are YOUR perception. But although this post was a reaction to the TIM/NFB exchange, it is of course irrelevant for that debate. You (or anybody's) personal preferences have nothing to do with whether NFB causes TIM or not.
Jan
It doesn't even need stereo or audiophile recording. 1929 Bing Crosby I was listening to the other day had it. Gilbert Briggs said something absolutely correct. Stereo verses Mono is not the issue. Two ears is the issue and we prefer two speakers. He begged people to upgrade speakers and then the amps later when funds allowed a high standard. This was early 1960's, most Stereo amps were abysmal. Good Mono is better than OK Stereo. You can hardly stop yourself hearing fake Stereo when the system works well. The small differences in the room or pick up cause that.
JLH built the 10 watt amp as two Williamson amps was too much to house. JLH stated he could not detect a difference between his new amps and the Williamson. Job done I would say. Williamson was 19 when he designed the KT 66 amp, 66 when he died. He went to the robotics industry and design robots for motorcar assembly ( from memory so forgive if the exact age etc). The Williamson was not the worlds first 0.1 % distortion amp . Leak predated it by 2 years in the 1945 TL12 ( not 12+ although that is a nice amp ). Hafler I am sure doing something?
For what it is worth distortion in perfect decay can have the 1/3 harmonic at 0.3% and still be considered hi fi. 0.1 % was taking that result 20 dB below audibility (1% THD in exponential decay of distortion harmonics). It you say that was the 1950's and not valid now remember the Quad ESL will still out perform most speakers on distortion and transparency if you take an hour adjusting them for the sweet spot. With two Neumann mics nothing really has improved when a direct feed. Some people added Kelly ribbon tweeters as to get above 15 kHz. The real interest came when the Quad and KLH made it possible to ask these questions. I think class AB needs lower distortion as by whatever name TID goes by a good AB amp needs to be better than a class A to sound convincing. With a little thought that would seem reasonable as the dam thing is trying it's best to unlatch rather than play music. Class D upsets that thought a bit. I would argue the class D makes it possible to ignore similar problems as they are shifted very high in frequency. The Quad 303 distortion is not much above - 80 dB and is also nearly in exponential decay. That is one very near perfect result as it obeys two sets of rules. LTP amps usually have very low second harmonic with noticeably higher third. That may not be as good. One can do the Naim trick and have different resistors on the long tail pair to put it back into harmonic balance. Naim has slightly raised 4th.
The 0.1% THD had to be revised in later years. If the 5th or 7th was at 0.1% it would not be a nice amp (the feedback loosing power higher up). The assumption had been that it was not going to happen when valve designs. Many Japanese amps were like that. When criticized they said no Japan domestic market amp was made that way. It was USA techno spec market. They had thought we were mad yet made what we requested. Douglas Self said that as far as he could see TID and even IM distortion is shown by simple evaluation of distortion. I think he has a point.
JLH explained below.
http://6moons.com/industryfeatures/zen/plh.pdf
Quad 405 . Because it is unlatched in the class B/C dumpers it might be easier to control.
Two neat quick pulses rather than two mixed AB pulses. One simple trick might be to put diodes in the dumper bases and hold them off until a higher voltage. Make the class A section beefier to cope (anybody care to calculate it). As said before some MOS FET's might avoid that. I dare say more feedback to suit 16 db worse dumper distortion if wanting something to boast about and FET's. The 405 is mostly a feedback amp with HF feed-forward assistance.
Two neat quick pulses rather than two mixed AB pulses. One simple trick might be to put diodes in the dumper bases and hold them off until a higher voltage. Make the class A section beefier to cope (anybody care to calculate it). As said before some MOS FET's might avoid that. I dare say more feedback to suit 16 db worse dumper distortion if wanting something to boast about and FET's. The 405 is mostly a feedback amp with HF feed-forward assistance.
Spot on, IMO. A 40W amp capable of using its full range cleanly will be very, very impressive; it's no "trick", it is merely a well designed bit of kit - back in the 80's I did the rounds of listening to every "impressive" amp I could find, in the hope that at least one would be capable of using its full range of power. Only a local effort, by Greg Ball, stood out at the time as remaining competent at higher sound levels ...
And none taken, Jan. It's really about two things:
1. I very clearly said that excessive NFB >>> may <<<, not "must" or "will", possibly cause TIM related phenomena, which is, I believe, quoting one of Otala's original conclusions from 1973, and
2. I don't think anyone here would argue that designing for wide open loop bandwidth and low global NFB is just the same as designing for 60+ dB of GNFB and damn the open loop bandwidth all the same. It is therefore reasonable that the end result will also be different, if only by a little.
All I am saying is that to me, that "just a little" seems to be the difference between whether I like it or not more often than not. This seems to be the inevitable conclusion of my 6:1 ratio of preference. It's not a question of right and wrong, it's simply a personal preference.
And it certainly doesn't make either Otala or Cordell right or wrong.
Last edited:
The catch here Frank is that the said Toshiba/Aurex integrated amp was one of their cheapest and smallest offerings, yet it sent quite a few more famous and way more expensive amps back home in tears.
On the other hand, it may be good for the money, but of course, it's hardly perfect.
The next step up from it would be my H/K 6550 integrated amp after I replaced it volume pot. Now, that's one damn serious piece of kit, whose only fault is its 50/70W into 8/4 Ohm rating. But the quality of sound, and in particular the spatial aspect, has to be heard to be believed. It's mine, but I still need to remind myself every now and then of just how good it is. But that's another story.
- Status
- Not open for further replies.