A simplistic test could guide a decision whether or not to pursue some path for further experiments or refinements. My goto example is the addition of a series resistor of around 10 ohms between a power amplifier and the speaker, to explore whether or not playing with output impedance makes any sense. On its own, the test is limited because it alters the frequency response, not just the HD. Without controlling for it, there's more than one variable being changed. But when combined with the memory of other tests like playing around with parametric EQ, I think meaningful conclusions can still be made.
6. There is no discernible difference between the two stimuli.
But that could be because of one or more the other 5 reasons. That’s the whole point of my post. Hel-loo! In other words results of tests, including controlled blind tests, rely on factors unrelated to the item(s) under test.
But that could be because of one or more the other 5 reasons. That’s the whole point of my post. Hel-loo! In other words results of tests, including controlled blind tests, rely on factors unrelated to the item(s) under test.
Last edited:
I think no meaningful conclusions can be made for several reasons that have been exhausted here many times.
I do not understand why you speculate about what could or could not be instead of doing it correct in the first place.
Assuming you are interested in finding out the facts.
Maybe you are just having the equivalent of bar talk, and that's fine too.
Jan
I do not understand why you speculate about what could or could not be instead of doing it correct in the first place.
Assuming you are interested in finding out the facts.
Maybe you are just having the equivalent of bar talk, and that's fine too.
Jan
Absolutely. In science those are known as confounding variables. It's the job of the experiment designer (i.e., the researcher and their team) to design the experiment such that the confounding variables are eliminated or at least minimized.
There are statistical methods that can be used to tease out how much variance in the dependent variable is caused by the manipulation in the experiment and how much is caused by other factors. Analysis of covariance (ANCOVA) springs to mind, but there are others.
You can always read the methods section of any scientific paper and form your own opinion on whether that particular experimental design does a good job of eliminating the confounding variables. And if you believe the researchers have missed anything obvious, I urge you to write to them with your concerns. That's how science works...
Tom
The funny thing. I learned it from a national broadcaster that have their own archestra. So yes they could know how they sounded live and on recording.
Guess they thought a violin and mic is not that much tampered with as a guitar effect box and guitar cabinet close mic.
And you need something that can fill the spectrum.
It seems the partisipants in ITU protocol tests lasts about 30 minutes. So it must be quite exhaustive.
But they try to find small differences in sound on compression, datarates, codecs, EQ and such.
So we european brodcast listeners thank them for their research and effort. Luckily not all is broadcasted at 32 kbit/s because of this
I simulated my amp designed commercially in about 1996 and post the THD at 1 watt 8 ohm. I dio never listen very loud in any case.
Should I post another at the onset of clipping. Harmonics are about -160or more dB down, did not calculate just guessed.
Should I post another at the onset of clipping. Harmonics are about -160or more dB down, did not calculate just guessed.
Here is at the onset of clipping. If I can hear distortion it is probably that of the recording equipment. Harmonics are about -142 dB below fundamental, which is below threshold of hearing.
Attachments
Last edited:
To be honest I do not know what distortion sounds like, hence my previous silly comments about the horse hair. Only distortion is speakers and there is nothing to do about that. One day when they create distortionless speakers it would be nice. Some very high end CD players and DACs may come close to this amps THD. As bad as the BBC LS3/5as are, at present they sound absolutely exceptional.
By implementing some design changes in in the schematic and simulation again I was able to drop the THD by two orders of magnitude, but the amp becomes very complex and I don't know if it could be practically measured better due to the components playing a major role. In my current amp all semies are from Hewlett Packard and Mosfets from Toshiba. All matched.
I must admit, I am no audiophile, I am just obsessed to design and develop anything to extreme and will not stop until I reach my goal. Are these things musical as they say. I don't know, it give exactly what it gets so it is hard to comment. The harmonic distribution is monotonic, but what the heck if they ar at -142 dB are they there or just a figment of your imagination.
“You can always read the methods section of any scientific paper and form your own opinion on whether that particular experimental design does a good job of eliminating the confounding variables. And if you believe the researchers have missed anything obvious, I urge you to write to them with your concerns. That's how science works... “
Most audiophiles aren’t aware of, or don’t agree on, many the variables Involved in audio testsl, including our old friend, controlled blind tests. Sufficiently broken-in speakers, cables and electronics, to name one. That’s how science works.
Most audiophiles aren’t aware of, or don’t agree on, many the variables Involved in audio testsl, including our old friend, controlled blind tests. Sufficiently broken-in speakers, cables and electronics, to name one. That’s how science works.
It is indeed fascinating how many audiophiles will throw science out the window when it doesn't confirm their beliefs, even though it was that same science that made it possible for them to listen to reproductions of music in the first place. Humans are not rational critters. That goes for audiophile humans too.
Tom
Tom
Maybe they know about the replication crisis?
EDIT: Engineers have been known to throw science out the window too if they don't agree with it. Its happened in this very forum when there is a study showing people can be affected by ultrasonics, or that DA in caps can be a real issue, or that cables can directly affect sound other than by capacitance. The paper is picked apart, reasons are found to dismiss the findings as bad science, etc. OTOH, if a paper is agreed with then it is not usually examined carefully for reasons to doubt it.
Last edited:
I'm an engineer. I know of the replication crisis. I don't throw science out because a few experiments out of many have been difficult or impossible to reproduce. See. This is where your false dichotomy breaks down.
Tom
Tom
At least you are one of those who designs and builds amplifiers.
In my humble opinion, you don't need to be an audiophile, you just need to listen to your music through your system and your amplifiers that you designed and built yourself.
And for what it's worth I do the same, I listen to my music through my system loving to feel good sensations from.
On the other hand, I believe that here we are almost all "audiophiles", just because we all love both music and audio.
Then there are few of those who only love themselves, they don't know what they're talking about and yet they talk and talk and have never designed or built an amplifier, but only their own image for profit.
They few ones pretend to know, but waits for others to answer for him since not know about almost nothing.
And they few ones created the existence of audiophiles because of the need to talk about himself.
Which in my experience is much worse than being an audiophile.