Great. Just do it level-matched and without peeking. The stuff that the hucksters whine about.
no one would listen to a poor sounding system, instinctively you would back down on the volume control, that is what i would do...
A good way to assess what an amp sounds like when driven hard is to load it up (3 or 4 ohms) using a resistor and then to put a low z attenuator across the load and plug in a pair of headphones.
I don't buy your previously espoused philosophy of listening to systems flat out fas42. I like big amps because they sound effortless and not because they can go loud.
I don't buy your previously espoused philosophy of listening to systems flat out fas42. I like big amps because they sound effortless and not because they can go loud.
No, you're not deliberately thrashing the system, you're listening loudly enough for the distortion characteristics to become apparent - simple approach, put on solo piano, and start turning up the volume so that it approaches realistic levels; typically, it will betray various shortcomings as you do so, sounding less and less like the real thing as the level increases. These behaviours then can guide you as where the problems lie.
@ bonsai: did I say "loud", as in hifi loud? The key word you used was "effortless", which is a true measure of a system's capabilities. IME, a lowly amp can do that, sound "effortless" IF the whole system is engineered well. "Effortless", in fact, is a synonym for "Able to go loud without high levels of disturbing distortion" - something that few systems are capable of ...
Last edited:
but this is the perceptual psychology key point - your brain, way below conscious level is correlating everything you've heard, read, experienced in the purchase, handling, installation, use of the device brand, model you see to construct a narrative about its qualities, place in your world view
it has been found again and again that the sighted judgment is hugely influenced by your internal state, history, "expectations" with the particular brand, model...
it is uncomfortable to make discriminations on just the audio alone - you can't go back and forth with elements of your internal mental narrative seeing affirmation, reinforcement of you mental model and the sensory input
but a good Blinding protocol is required to make discriminations that are independent of your history with the part
anecdotes that you can make discriminations that "prove" to be independent, apparently "go against expectation" are not logical refutations of the necessary conditions for unbiased sensory discrimination - we all do hear real differences quite often too
the 1 dB level difference being identified as a audible change by an expert listener isn't a counter example - it is within the expectations of Psychoacoustics
Last edited:
I find 'peeking' extreamly useful in learning to characterize via listening as to what something sounds like. And, after some listening and knowing what it is you are listening to, you get trained to hear it without peeking.
For example.... how hard is it to know what pure second (2H) sounds like. Or a specific frequency? You measure it and see it and listen to it. A DIY'er can adjust harmonic levels and harmonic order with test equipment to learn what each sounds like. This is all peeking. But it is also training to discriminate as to what you are hearing. Do the same with freq and you learn to recognize which freqs are missing or emphasised in the music system's spectrum. You peek at the data of amp and speakers and then listen and learn... memorize it for later use.
Do this peeking at data and correlating it to what you hear and eventually you will be able to make accurate judgments just by listening. Sometimes going back to data to be sure or update your memory. Some might be amazed at the subtleties one can hear with training using the 'peeking' method/process.
THx-RNMarsh
For example.... how hard is it to know what pure second (2H) sounds like. Or a specific frequency? You measure it and see it and listen to it. A DIY'er can adjust harmonic levels and harmonic order with test equipment to learn what each sounds like. This is all peeking. But it is also training to discriminate as to what you are hearing. Do the same with freq and you learn to recognize which freqs are missing or emphasised in the music system's spectrum. You peek at the data of amp and speakers and then listen and learn... memorize it for later use.
Do this peeking at data and correlating it to what you hear and eventually you will be able to make accurate judgments just by listening. Sometimes going back to data to be sure or update your memory. Some might be amazed at the subtleties one can hear with training using the 'peeking' method/process.
THx-RNMarsh
Last edited:
yes opponents throw up canards, strawmen, deliberate frabrications
to discredit or ridicule what they don't want to do or have to acknowledge
in "ears only" listening testing training is permitted, encouraged - it is only the trials you are scoring that need Blinding
you can "know everything" about the experiment: signals, sources, devices under test, resistor manufacturer, lot code...
you can arrange multiple trials with "focus" questions, drawing attention to potential audible features
you can practice to reduce "test anxiety", you can take as long as you like to listen
you just aren't allowed any method of knowing which is which during the trials except the audio waves reaching your ears
don't peek, do level match if you want any respect from those who have read Psychoacoustics, Sensory experiment Design
to discredit or ridicule what they don't want to do or have to acknowledge
in "ears only" listening testing training is permitted, encouraged - it is only the trials you are scoring that need Blinding
you can "know everything" about the experiment: signals, sources, devices under test, resistor manufacturer, lot code...
you can arrange multiple trials with "focus" questions, drawing attention to potential audible features
you can practice to reduce "test anxiety", you can take as long as you like to listen
you just aren't allowed any method of knowing which is which during the trials except the audio waves reaching your ears
don't peek, do level match if you want any respect from those who have read Psychoacoustics, Sensory experiment Design
Last edited:
Within its limitations, yes,Is this with the PC speakers?
. It has a miserable power supply, I have done almost nothing to address that, so any recording with sustained high average level content will expose its weakness there - but in other areas that matter, for me, it does pretty nicely. For example, I can run most classical pieces at absolute maximum volume, and it comes across very competently - but the wrong sort of organ recital will catch it out badly; rail sag will cripple the sound.At some stage I will have to do a major push to assemble a system with more heft - there's enough gear lying around to easily go to 120dB levels competently, including in the bass. Ah well, when the motivation kicks in ...
Yes, it's a long term process, learning to recognise various misbehaviour traits - at times I find it amazing how people can listen to obviously flawed sound, and nod their heads in approval - they haven't learnt to distinguish what's on the recording, from what the playback system is doing to the sound.
When you're in the throes of tweaking a system you have to shortcut your way through to getting acceptable sound; the combinations will kill all chances of getting somewhere in reasonable time, if everything is tested "correctly" each step of the way - this is where experience is a most powerful tool ...
Last edited:
Its a very interesting question indeed. But it seems to me there isn't a 'happy medium' here.
Higher value resistors are indeed noisy, but aren't we concerned more with SNR than absolute noise? Assuming thermal noise is the dominant factor, then that's proportional to the square root of the resistance. Hence for better SNR, going higher for the resistive load gives an advantage provided the signal scales in proportion to the resistance.
Bottom line is - higher value resistors give you lower current so power supplies should be as high voltage as possible, signals should be as large as possible. Since bipolar opamps are eventually going to be limited by current noise, JFETs are the way to go. Here's one suitable opamp - http://www.ti.com/lit/gpn/OPA552?keyMatch=opa552&tisearch=Search-EN
60V total supplies, nice low thermal resistance package to keep the input bias currents low (low temperature is needed for JFETs to have the lowest bias currents).
Thank you for the reply and recommendation.
I have never even heard of the OPA551/552 op-amps before.
60 volts sounds like a LOT of voltage going to a small IC, though.
So, you're suggesting using these IC's with higher voltage supplies and higher value resistors(perhaps 10k, or so) in the feedback loop?
I'm partial to the sound of bipolar op-amps, but I'm certainly open to other options.
OK, way way way OT:
Interesting piece. Also interesting that feral/cultured does not entirely break down by region. I found the comments of Teddy Hall (Stellenbosch) particularly interesting. The "wild" fermentation ultimately results in Saccharomyces strains dominating, as expected. Also Jeremy Seysses from Burgundy, who said that in their Provence vineyards they needed to use cultured Saccharomyces strains to complete the fermentation. One might have thought that the warmer temperatures would have reduced the need for that.
If I grew grapes and made wine in a "traditional" area, where a lot of grapes are grown and have been grown for a long time, I would probably rely on the feral yeast and supplement as needed in bad years. OTOH, if I grew grapes and made wine in a place like Nova Scotia's Annapolis Valley, where fruit orchards are everywhere, I would probably use cultured yeast in order to reduce the contributions of other wild yeasts more associated with apples etc. (Of course, I would probably also try a few low-volume fermentations with feral yeast just to see how it would turn out.)
Finally, the cultured yeast strains are S. Cerevisiae and a few others, all found in abundance growing wild in vineyards. It's not like they have been concocted in a lab via gene splicing. In fact, one thing I objected to in that article was from Olivier Humbrecht from Alsace, who said "They all have added killer genes". What BS. S. Cerevisiae is present in vineyards, breweries, and bakeries, and has been for millennia, and has been cherished by brewers and bakers for millennia, precisely because it grows at a fast rate and destroys competing organisms (including pathogens and spoilage organisms like bacteria). So even if your fruit is damaged (juice exposed to air) and some "bad" organisms are present the fermenting yeast will kill them off and take over. Humbrecht makes it sound like cultured yeast has been genetically engineered, but it has been selected via a very old mechanism (see Darwin, C.).
Interesting piece. Also interesting that feral/cultured does not entirely break down by region. I found the comments of Teddy Hall (Stellenbosch) particularly interesting. The "wild" fermentation ultimately results in Saccharomyces strains dominating, as expected. Also Jeremy Seysses from Burgundy, who said that in their Provence vineyards they needed to use cultured Saccharomyces strains to complete the fermentation. One might have thought that the warmer temperatures would have reduced the need for that.
If I grew grapes and made wine in a "traditional" area, where a lot of grapes are grown and have been grown for a long time, I would probably rely on the feral yeast and supplement as needed in bad years. OTOH, if I grew grapes and made wine in a place like Nova Scotia's Annapolis Valley, where fruit orchards are everywhere, I would probably use cultured yeast in order to reduce the contributions of other wild yeasts more associated with apples etc. (Of course, I would probably also try a few low-volume fermentations with feral yeast just to see how it would turn out.)
Finally, the cultured yeast strains are S. Cerevisiae and a few others, all found in abundance growing wild in vineyards. It's not like they have been concocted in a lab via gene splicing. In fact, one thing I objected to in that article was from Olivier Humbrecht from Alsace, who said "They all have added killer genes". What BS. S. Cerevisiae is present in vineyards, breweries, and bakeries, and has been for millennia, and has been cherished by brewers and bakers for millennia, precisely because it grows at a fast rate and destroys competing organisms (including pathogens and spoilage organisms like bacteria). So even if your fruit is damaged (juice exposed to air) and some "bad" organisms are present the fermenting yeast will kill them off and take over. Humbrecht makes it sound like cultured yeast has been genetically engineered, but it has been selected via a very old mechanism (see Darwin, C.).
You're not alone, nor had I before I figured out that higher signal voltages were better, without limit. Which has been quite a recent discovery for me
There are ICs which go considerably higher. Linear Technology for example has one (here - LTC6091 - Dual 140V, Rail-to-Rail Output, Picoamp Input Current Op Amp - Linear Technology) which operates up to 140V.
For ultimate SQ yeah I believe its the way to go, but 10k is still too low. If you have a 60V signal swing that's 3mA each way, too high a current. Think in terms of 100's of k for feedback resistors.
Me too, I'm though gradually coming round to the view that opamp sound is really in the main power supply sound. And high voltage power supplies (because they're lower current) are easier to get right than low voltage.
- Status
- Not open for further replies.