I used to hear them, too, especially (very faintly) in fan noise, especially after I happened across all of the material on the web about EVP (electronic voice phenomena), especially right after my wife died, in 1999. But eventually I found out that I could make her or them say basically whatever I wanted to hear, which to me meant that my brain was able to do things with sound that I was previously unaware it could do, and I had only been talking to myself. What a letdown, though.
Or maybe you meant that you mostly listen to recordings of artists who have since passed away. In that case (cue Gilda), "Never mind.".
Last edited:
The "culprits" are expectation bias, sighted bias and the brain's ability to fool ourselves if we let it. When the main obsession becomes the audio components instead of the music, then begins the downward plunge into cables, power cords and magic rocks.
I can make myself "hear" the phone ringing at work even when it isn't sometimes. The mind is a scary thing.
These may be the culprits for some, but I'm still talking about playback distortion, and the reduction of that to inaudible levels. Everyone knows what obvious distortion sounds like, and beyond that there are lower levels of distortion which are less and less obvious, they become the "sound" of certain components, or combinations thereof. If all one does is to fool around, swapping bit and pieces of ancillary gear, then all that's happening is the changing of the flavour of today's distortion -- eliminating distortion is a quite different and much more difficult, irksome matter ...
No it doesn't.
I know that perfectly well, thank you. (I refer you to, for example, this post http://www.diyaudio.com/forums/pc-based/212982-bare-bones-framework-audio-dsp-pc-2.html#post3077993).
What I am saying is that a standard frequency domain test may produce one number for THD, say. I cannot convert that into a time domain result. The frequency domain test may even produce magnitudes for the harmonics. I cannot convert those into a time domain result (no phase, you see). Even if I have all the information to do the transformation into a time domain waveform, the test result may be meaningless: the standard frequency domain test probed the amp with a 1kHz sine wave, and maybe I'm interested in the amp's 'thump' at power on with no signal applied. If I record the amp's output with no signal applied in the time domain and by the magic of the DFT, convert it to a frequency domain representation, will the result tell me what I need to know? It will certainly be the mathematical equivalent, but not as informative (in this case) as the time domain representation.
In other words, my motive for doing the test, my choice of test signal, my choice of whether I choose to look at the time domain result or the frequency domain result, is what decides whether the test is "a frequency domain test" or "time domain test".
Consider feeding two minutes of music into an amp and recording the output. If I process the whole lot with a large DFT and look at the output, will the resulting spectrum tell me whether the amp's gain dipped by 0.02dB after each drum beat (for example)? Even if I break it into small windows and look at a spectrogram? No. But if I interpret the signal in the time domain, I can deduce such things: the process might be along the lines of finding what I have to do in order to match the output with the input. In this case, I would find that the best match occurred when I dipped the gain according to a certain envelope in this region. Other 'pathologies' would be revealed by having to apply other modifications in order to match the input to the output.
Last edited:
"to me meant that my brain was able to do things with sound that I was previously unaware it could do" "and I had only been talking to myself. What a letdown"
"The "culprits" are expectation bias, sighted bias and the brain's ability to fool ourselves if we let it"
It's what I said earlier. You can be influenced and your brain can form something that isn't there but this will wear off and eventually, reality sets in..
If you Spend enough time listening to audio components including speaker cables, the actual sonic charateristics of each piece becomes clear..
There's the cartoon.
Frequency domain does indeed include phase. Unless your amp has a frequency dependent delay line, it will be minimum phase and the phase is trivially extracted from the amplitude response. And even if it does have a frequency dependent delay, this shows up in basic measurement (for the record, in 40 years of testing amplifiers, I've never seen one that wasn't minimum phase except for units with an explicit delay built in, such as subwoofer amps).
Frequency and time are perfectly interconvertable and carry the same information. It may be more convenient to look at one or the other, depending on what you're looking for. That's why engineers have both available to them.
Sy, be careful here because the Fourier transform assumes a linear system. It is not directly applicable to a nonlinear system.
If one has sine wave data from the nonlinear system then one can determine the phases of the harmonics by using the Fourier Series, but this only works for a single frequency.
We're talking electronics here, which as a practical matter are an excellent approximation to an LTI. To the extent that they aren't, these also are screamingly obvious in basic measurements.
Yes, after spending enough time listening to audio components in blind AB testing it is clear which components make zero difference in perceived sound. All of these components have measurable differences that correlate directly to their physical characteristics.
I can't see much difference between those statements and "I don't suffer from expectation bias..."
All you've done is add "...when I've been listening for long enough". It's still just anecdotal.
Ah if only those that claim to be able to hear differences in audio components would do such tests but they won't. Might upset themselves when they find out what they "heard" is all in their head and all the money they wasted on the "cable of the month" could have put Jr through college
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.