Good point. I think that "sounds more pleasing" is the most common goal. It's what the general public wants and is usually how "better" is defined.
In audio enthusiast circles, there are other goals.
In audio enthusiast circles, there are other goals.
To the general public, I'd guess that MP3 is better because you can get more MP3 on your ipod.
However, the following observation should be made in this context: better measurements sometimes correlate 1 to 1 with more pleasing sound. This has been demonstrated for distortion above a certain threshold and frequency response, just to mention two. But not for some other parameters, such as phase shift for example. A 24dB/octave crossover can sound very well, in spite of a 360 degree phase shift in the crossover region.
So, in order to make something sound more pleasing, it is sometimes warrented to optimize a number well defined technical parameters, making the contraption 'better' in an objective sense. And some technical parameters, such as phase shift, can be safely kept below what would be possible to achieve, in a trade off for things that make it sound better. For example, in order to stay away from driver break up, I gladly have some less phase linearity by using steeper filters to prevent that.
Now, it is my honest opinion that LP does not give anything in return for the technical deficiencies it has as compared to CD, but in a well implemented set-up the can both sound great. So, it is in the end probably all below the perception threshold, and other factors come into play.
Some people seem to think that digital has some ultra-realism & leading edge attack, sometimes lapsing into digital edginess. They assume this to be a characteristic of the extra detail that digital has. This is missing to a large extent in vinyl replay & there is a smoothness to the sound.
In fact, I'm convinced for a number of reasons, that this ultra-realism is in fact the distortion of jitter & when jitter is reduced the sound becomes much smoother & more analogue-like without any loss of detail but with no ultra-realism or edginess.
I liken it to MSG in food - it gives a kind of kick to the taste but hides the subtleties & dynamics of the real taste (& ultimately is headache inducing). Some probably prefer food with MSG?
In fact, I'm convinced for a number of reasons, that this ultra-realism is in fact the distortion of jitter & when jitter is reduced the sound becomes much smoother & more analogue-like without any loss of detail but with no ultra-realism or edginess.
I liken it to MSG in food - it gives a kind of kick to the taste but hides the subtleties & dynamics of the real taste (& ultimately is headache inducing). Some probably prefer food with MSG?
Last edited:
Most digital systems seem to have near 'blameless' specs for noise and distortion. Can jitter be a problem if the specs are so good? Is there something especially offensive about jitter noise/distortion that makes it a problem even at extremely low levels?
Last edited:
Easy.
As far as I know, the specs are for individual parts, not end-to-end. So, if you stick a bad transport on a great DAC, you'll have problems, but good specs. Eg. stick an NEC external CD ROM onto a Benchmark DAC -- it will suck.
I know of no maker that specifies THD/Noise comparing the bits to the waveform.
As far as I know, the specs are for individual parts, not end-to-end. So, if you stick a bad transport on a great DAC, you'll have problems, but good specs. Eg. stick an NEC external CD ROM onto a Benchmark DAC -- it will suck.
I know of no maker that specifies THD/Noise comparing the bits to the waveform.
There are quite a few people involved in digital audio that think jitter is a big problem.
In a way that cuts to the heart of this thread - how can something sound bad if it measures well. Two options: we're measuring the wrong things or it doesn't sound bad, it's an illusion.
The opposite is also stated - how can something sound good & measure bad. Again - wrong measurements or illusion.
A possible approach to answering this might well be pyshoacoustics. Make our measurements more focussed on what's important to the known model of the ear - that way we might actually be able to correlate what could sound good from a set of measurements! Let's face it, measurements of linearity haven't really gotten us very far as to what something might sound like?
I was curious about was how jitter shows up in the noise and distortion measurements. Presumably it is different from quantisation distortion because it scales with amplitude..? i.e. it's proportionally just as bad on a loud signal as a quiet one..?
A fascinating article on the development of a system for recovering a low jitter clock from a USB audio stream.
http://www.eetimes.com/design/audio...l-memoir-of-engineering-heartache-and-triumph
http://www.eetimes.com/design/audio...l-memoir-of-engineering-heartache-and-triumph
The problem with discussing jitter is that it needs to be qualified what type of jitter is present & the spectrum analysis of that jitter. Using a single value is as meaningless as using a single THD value.
Yes, they should be saying "I" think it sounds better, or it comes across as a fact, not an opinion, which it is. And again a vote should settle it.
Theres a lot of people out there who think MP3s and over compressed (dynamics) music "sounds more pleasing" (or better). Now if they become the majority are you going to agree with them?
It can be seen in spectrum analysis as sidebands around a tone. There are other ways to measure it, but s.a. is handy.
It's not something that shows up in normal THD specs or frequency response. Maybe in IMD, I'm not sure. Werner probably knows.
Why would I? Except maybe to acknowledge that it is what they like. There is already a lot of that around, without even dipping into MP3s. Most people listen only to amplified music. It's what they are used to and what they like.
But don't get too discouraged about the MP3 crowd. There have been a number of studies cited right here on this forum that indicate they can still hear and prefer a better format. But, it would seem, they don't really care. 😉
I'd like to mention that one thing which the vinyl nay-sayers fail to mention is that since the advent of cd, vinyl playback systems as well as the software have improved considerably over what they were in the sixties, seventies, and eighties. All of this "I pulled out all my old Rolling Stones records and stuck an ADC cartridge on them and it wasn't as good as digital" is nonsense. Recently produced vinyl by the well-known record re-issuers has wider dynamic range, better bass, and is dead quiet compared to floppy 47th stamper consumer stuff from the old days. Moving coil cartridges have also improved a great deal and make it possible to play lps that have been mastered closer to, if not over the edge of what was once considered practical.
John
John
Last edited:
It is a fact when I hear something. Not seen the McGurk effect on YouTube? Its a fact that people hear different sounds when presented with the same audio stimulus, depending on what visual stimulus they receive. Its not the person's opinion as to whether they hear one or another sound.
Voting is quite irrelevant.
- Status
- Not open for further replies.