Human hearing beats the Fourier uncertainty principle
Lynn Olson posted this in the lengthy "Beyond the Ariel" thread, it's pretty interesting stuff, particularly as regards formats, masking, etc.
Lynn Olson posted this in the lengthy "Beyond the Ariel" thread, it's pretty interesting stuff, particularly as regards formats, masking, etc.
SY from my view point you insert the part of digital that is not far off a-d-d-a . The cd production with high speed error correct on the fly and heavy handed mixing cause more often a problem . Inserting the A-d-d-A I can believe is hard to hear above the noise floor . Last month had the good fortune to compare master tape to cd to vinyl and high def 24/192 digital at a friends Studio and on a high end system . Found the high def and cd very close to the master . However the vinyl was more enjoyable to listen . There lies the horns of the dilemma the less accurate one is more (here we go ) more musical and more involving . There may be the big problem the human interface that likes distortion being the cause of the digital discourse.
well, is it not so?
I work in the corporate offices of a company approaching 50,000. Let's just say the standard deviation around advancement by competency is shockingly wide.
So in the vinyl mastering/cutting/pressing process are there components that are deliberately included for their 'musicality' rather than their ability to deliver flat frequency response and low distortion? - because if there are I've never heard of them. And if they don't exist then it's an amazing coincidence that, although the originators of the LP system were boring old engineers looking for flat frequency response, low noise and low distortion, the inherent distortions in the process just happen to be the exact same distortions that audiophiles find pleasing to the ear to this day. Is there something magical about 33 1/3 rpm, vinyl and 12" that they found worked best in listening tests? Or was the truth about those choices rather more prosaic?
Of course a master for vinyl has to have its bass mixed to mono to avoid needle skip on playback etc. and its treble reduced to prevent cutter overheating and all that. Could these limitations in fact be the key to why people prefer the vinyl sound?
I bet it would be possible to do a pretty good digital simulation of a 'vinyl effects processor' using some linear and nonlinear DSP.
Yes, if you include 'straight wiring' the signal path I'm pretty sure going back at least as far as the Fine/Cozart Mercury classical recordings of the Fifties. Certainly by the time Sheffield was going strong and rebuilding monster tube cutters.
Highs are boosted on vinyl.
Here we have two parts the eq of the record (vinyl ) and the phono section . There may be some cross talk in the vinyl based storage that is not there in cd . 30 years ago I used a Denon 103 MC cartage that had like 43 db of channel that at one of those ortofon test show was considered very good . Channel separation on CD is much more. Bass summing is an other point. Digital is a different medium of storage than analog as such like triode vs pentode it need to be used differently . You may be on the right road adding spice ( distortion on an absolute basis ) making the music more palatable to many. Yes it can be over done like most thing and seam to be a lot in the modern history of recorded music.
I don't think anyone pretended it's about harmonic distortion.
I have done my share of tests. simulated a non-linear curve with music, artificially added harmonically-related content to a sine wave. yes, with music even 10% or more of THD is not audible. who said it was? and who said this discussion is about 10^-1000000 THD?
so, you subjectivists really think types like me are like that?
Last edited:
it did make the graphics card on my computer run 4% faster without any overclocking - which I considered to be an objective "measurement" of sorts!
being relatively skint, I do not have an oscilloscope or spectrum analyser
neither do I have any education in electronics or electrical engineering
I will see if there is some kind of spectrum analyser software I can use on my computer... although I keep forgetting to do this!
being relatively skint, I do not have an oscilloscope or spectrum analyser
neither do I have any education in electronics or electrical engineering
I will see if there is some kind of spectrum analyser software I can use on my computer... although I keep forgetting to do this!
Reading the vinyl mastering tips from the professionals is quite educational. Basically, the vinyl mastering process is tolerant of 'girl and guitar' but much less so with anything more dynamic, where artificial limiting of low and high frequencies and out-of-phase stereo is essential. 'Diameter loss' is a term I hadn't heard before, but I've always known about the consequences.
All this mangling (or self-restraint in the recording process) produces the audiophile's preferred sound. It could so easily be done in DSP, including the much sought-after 'diameter loss sound' as playback progresses.
Vinyl Mastering FAQs
Chicago Mastering Service--CD and Vinyl Mastering Facility in Chicago, IL
All this mangling (or self-restraint in the recording process) produces the audiophile's preferred sound. It could so easily be done in DSP, including the much sought-after 'diameter loss sound' as playback progresses.
Vinyl Mastering FAQs
Chicago Mastering Service--CD and Vinyl Mastering Facility in Chicago, IL
Last edited:
There's more to it than just vinyl "distortion": digital distortion is inherently more disturbing and fatiguing, even though it's much less obvious, more subtle for many people. Long term listening is the giveaway for whether the digital sound is sufficiently clean, if one starts to lose interest in what you're listening to then you hearing the classic symptoms of this type of distortion.
One can learn to pick this artifact with a bit of practice, and then it becomes obvious as soon as you hear it ...
Words, words...
Can it be measured? If so, what sort of level are we talking about?
And would it survive the LP process? It would be fascinating if something at the lower limits of measurability could still be audible buried in analogue noise and distortion orders of magnitude greater.
I'm sure it could, but is anyone sufficiently motivated to do the exercise? As you say, there's a lot of talk here, but relatively little follow up, maybe we're all too "senior", and tired, to worry about jumping through another hoop ...
It's a dynamic process, time dependent; equivalent to the slow build-up of a static charge. You could do something like capturing the spectrum of a sound segment, music, with a very pronounced high frequency spectrum - something like a short workout on splash cymbals. Keep the segment on repeat, and capture the spectra every 5 minutes, over an hours or more. Then run an analysis, comparing the spectra, how they evolve over the time frame. I believe there will be a very clear pattern, if the system has the problem to a decent degree ...
Edit: The LP process? Yes, it should be perceivable ... personally, I'm amazed at how information is "buried" in ancient, primitive mechanical recordings that have been transcribed to CD, that is "clear as a bell" when a system is working right - providing, providing (!!), that the sound hasn't been "fixed up" by one of these "No Noise" processes ..!!
It's a dynamic process, time dependent; equivalent to the slow build-up of a static charge. You could do something like capturing the spectrum of a sound segment, music, with a very pronounced high frequency spectrum - something like a short workout on splash cymbals. Keep the segment on repeat, and capture the spectra every 5 minutes, over an hours or more. Then run an analysis, comparing the spectra, how they evolve over the time frame. I believe there will be a very clear pattern, if the system has the problem to a decent degree ...
Edit: The LP process? Yes, it should be perceivable ... personally, I'm amazed at how information is "buried" in ancient, primitive mechanical recordings that have been transcribed to CD, that is "clear as a bell" when a system is working right - providing, providing (!!), that the sound hasn't been "fixed up" by one of these "No Noise" processes ..!!
Last edited:
Unity gain in the DCX, no filters. Dead easy. Tape loop did not change the gain (I checked)
I don't understand your post. Can you clarify?
It can be. You'll see a difference in the harmonic structure. Even tho very low in level, I believe those changes are a big part of what we hear. When high order harmonics abound, that can sound rough or sour, just as can oscillation.
In subjectively harsh sounding circuits (analog or digital) I tend to see high order harmonics rising as frequency rises. In more "analog" sounding circuits, low order tends to rise as frequency drops. I haven't made a rigorous study of that, just some observations on my part.
- Status
- Not open for further replies.