We are not going to worry about 128kbps, I'm talking about using the best encoder, at the highest rate possible, with all the settings optimised for that piece of music - and compare that to the original. For example, use Lame, do some experiments to see which settings give the best subjective result - we are not even talking about using AAC, which nominally outperforms MP3 at particular rates.
I think what Antony Michaelson has to say here below about amplifier power and clipping is significant, especially to me, since I believe one of the most important factors governing perceived amplifier sound quality is that amplifier's clipping recovery behaviour.
http://www.carltonaudiovisual.com.au/?q=node/view/1020
Regarding distortion, I think that ideally it would be low in level and simple in character, with a simple relationship to the signal. Preferably even order, harmonically, too.
http://www.carltonaudiovisual.com.au/?q=node/view/1020
Regarding distortion, I think that ideally it would be low in level and simple in character, with a simple relationship to the signal. Preferably even order, harmonically, too.
And with waking up with your tube gear:
a) It's still next to you, so no, it wasn't just a bad dream;
b) It's still next to you, not repossesed, so you must have paid the wild price;
c) It has the upper hand in the matter of replacing rare and expensive tubes;
<First cuppa coffee>
d) Bearing in mind points (a) and (c) above, it's likely that said negotiations will not end today, and
e) Perhaps you should buy the latest whizz bang tube sockets to get rid of that microphony.
As a free extra, add the wife's joy at hearing you need to replace your tubes - again. However, this freebe might turn out to be very expensive if your wife decides that if you can afford new tubes, then you should be able to also afford a new bag, or dress she needs.
Frank, please enjoy the difference between 128kbps mp3 and original
https://www.dropbox.com/s/p47jax6kp52baeo/mp3-orig.wav?dl=0
Russ, no doubt there is sense in what Anthony Michaelson wrote. But, ...
He assumes a "healthy loud level". This is highly debatable regarding the actual power used. Some people have a more, and some a less sensitive hearing, so what might be great for one may be too loud for the other. I would have liked to have seen some specific figures, such as, for example, 96 dB SPL but at 3 meters, or some such.
He assumes speaker sensitivity of 87 dB(2.83V/1m. While there is no doubt such speakers exist and are sold new today, it seems to me that most have moved on to 90 dB or above. Mine, for example, do 92 dB, and when reclaculated, it seems his power level can be freduced to 158W in my case. Not to mention even higher sensitivity speakers, such as say JBL 4312, with its 95 dB sensitivity, reducing power requirements to half of what I need.
So, while I agree his arguments are valid, they have been given in too broad and unclear terms. With these things, one really should be more precise to be right.
But this should give pause to our tube friends - how effcient do their speakers need to be not to cause compession and/or limiting? The answer is simple - as much as possible above the 100 dB/2.83V/1m level, which is why in the mid 70ies some manufacturers (e.g. Altec Lansing) offered efficiences of 102 dB.
As things stand, with average efficiency speakers, in addition to wild distortion figures, they will also have compression and possibly even clipping.
The rest of us need to do some calculating to work out the maximum required power to avoid compression and limiting, and especially clipping, since solid state amps tend to clip violently and even possibly lethally regarding the speakers.
Last edited:
Dejan
My experiments with diy tube audio gear tells me that they are the perfect guinea pig for your anatomical table. They can take a lot of abuse. You may turn them red from tip to toe, they give you plenty of time to understand that you do something awfully wrong to them, they don’t smoke/explode on your face in a flip of a second and they talk to you again afterwards.
My endeavour with tube radios from 50ies-70ies has taught me that tubes –against urban myths- are very sturdy, long lasting (excluding magic eyes) creatures.
In audio gear for home use, you won’t have to think of replacing them due to aging.
I would say that they will perform flawlessly(*) for 50-60 years if the heating in-rush current is controlled and the proper on-off sequence btn HT and heaters is maintained.
Tube gear doesn’t have to cost us diyers an arm and a leg.
You may not know it but women are unconsciously attracted to tubes (consequently to tube-men too
)And since the important issue of clipping has emerged, tubes behave exemplary there
George
(*) just remove and reinsert them in their sockets once a year or two (contacts issue)
Mmmm, not necessarily. Tube amps nearly always show blocking distortion following clipping unless they're direct coupled or are designed specifically to avoid this (e.g., Crystal Palace, Red Light District).
Hi,
Exactly.
I'm still running the same valves in my preamp since 1986. When measured they're still 100% within rating. IOW, as new.
The occasional replacement of a 6080/6AS7G which is used in the output stage of my OTL is the sole exception. But then that's to be expected.
@George: I use a contact enhancing product on the pins of the smaller valves so the metal can't oxidize and good contact is guaranteed. ( I keep forgetting the name of this stuff, I'll see if I can find it)
Cheers,
Exactly.
I'm still running the same valves in my preamp since 1986. When measured they're still 100% within rating. IOW, as new.
The occasional replacement of a 6080/6AS7G which is used in the output stage of my OTL is the sole exception. But then that's to be expected.
@George: I use a contact enhancing product on the pins of the smaller valves so the metal can't oxidize and good contact is guaranteed. ( I keep forgetting the name of this stuff, I'll see if I can find it)
Cheers,
George, tube clipping is similar to dynamic compression
Lamm Industries ML3 Signature monoblock power amplifier Measurements | Stereophile.com
Lamm Industries ML3 Signature monoblock power amplifier Measurements | Stereophile.com
Last edited:
Thanks. Unfortunately, I don't think the levels were adjusted precisely enough in/after the compression, because you can hear too much of the waveform of the choral piece - better nulling would have turned the difference file more into meaningless chirps and whistles - peak level was only -17.5dB down.
Average rms level of both files is within 0.1dB order. The nature of lossy compression (both dynamic and frequency) does not allow for precise amplitude matching. There is no exact point for matching. Nevertheless, you can hear what lossy compression subtracts. The difference is a garbage of the worst kind.
George, tube clipping is similar to dynamic compression
Lamm Industries ML3 Signature monoblock power amplifier Measurements | Stereophile.com
What do you expect for the price of a house?
The issue isn't that it's tube, the issue is that it's heavily designed and lightly engineered.
That’s it.
Thank you for the proper term of the effect.
Acoustically-wise is very benign. A casual listener doesn’t understand the onset of clipping.
Raising the volume further, the weaker waveforms become louder-still clean- and the building-up structure of harmonics from the softly clipped higher waveforms makes the sound even louder and ‘fuller”.
Non fading memories: Years ago together with a friend, we were testing the perceived loudness between a PP class A tubed 12W and a SS class AB 60W, by simultaneously monitoring one channel from each amp output through a two ch oscilloscope. Winner was the 12W up to the clipping point of the SS, above which, it (the SS) became unlistenable.
SY
The few power amps I have built, all were capacitor coupled . And still I had good impressions.
After 2-3 years that my eyesight won’t allow me work with small components anymore, I will turn to all tube constructions again. I promise to investigate the issue in detail
George
Last edited:
George, the compression with the amp shown starts somewhere at 10W. I may easily have a solid state amplifier which starts clipping at 400W (the Lamm will never reach it) and has very very low distortion at any level. Even poor distortion of solid state amplifiers at low levels is a myth, well engineered SS amplifier has no problems like this. My experience is such that many people do not like precise reproduction of the recorded music and they ask for additives.
Its a big no no to compare wave shapes, don't do it!
Lossy codecs ALWAYS change the wave shapes.
What matters if you can HEAR a difference or not.
One other important thing:
Clipping causes huge amounts of high frequency energies, lossy codecs have difficulty encoding these signals. And its with these kind of samples (unnatural high frequency energies) that people can positively ABX higher bitrate codecs.
A nice test for people to take:
CD vs mp3 do you hear the difference?
I got 100% score with 48kbit/sec, I'm a master at this stuff....
What matters if you can HEAR a difference or not.
A nice test for people to take:
CD vs mp3 do you hear the difference?
I got 100% score with 48kbit/sec, I'm a master at this stuff....
Yes, it matters if I can hear the difference. Tell me, why do they use such horrible music samples for listening tests?
- Status
- Not open for further replies.