Discussion in 'General Audio Discussion' started by 0Hz, May 8, 2018.
You beat me to it, lol.
Nobody beats Gordie!
Yes, I knew they occasionally switched sides. But their iconic guitars were Fender for Hendrix and Gibson for Page. My point was that Gibson has a warmer sound and Fender an edgier cutting sound.
There are effects pedals designed to do exactly that. They get it about 80% right. But I doubt you would have seen Stevie Ray Vaughn doing one of his hits playing a Gibson with a Strat effects pedal. If you want a Strat sound, get a Strat. If you want that smooth., warm Gibson sound, get a Les Paul or better yet, an ES335.
Effects pedals? I thought they rolled tubes in those to change the sound?
They do all kinds of things to affect the output. And that supports my argument. Dave said all amps should have the same waveform output. Well, if it’s a sine wave, maybe. For a complex musical signal, components and design process the signal slightly differently. That’s what makes different amps have signature sounds. I didn’t intend to go off in the weeds discussing guitars. I was just using them as an example of warm vs. crisp sounds.
Which is another way of saying different amplifiers have different amounts of distortion at different frequencies and driving different loads. The term I've used before for this is an amplifier's distortion profile.
The better the amplifier (in an ideal sense), the less the output waveform -- sine or otherwise -- deviates from the input waveform. However, that doesn't necessarily mean listeners will always prefer lower distortion, as proven by preference for (for example) relatively high distortion tube amplifiers. Listeners may prefer one distortion profile over another, independent of absolute distortion levels.
Most certainly when using the utterly useless THD metric as it doesn't acknowledge two facts:
1. Absence of harmonic profile, especially high order components for which we are highly sensitive.
2. Bears little resemblance to what occurs when amplifying correlated dynamic content with rich harmonic structures, aka music.
Exactly. That which the human perception relates to hearing music as opposed to test tones.
I agree completely. Distortion profile is a good way of describing it. You would have to have thousands of data points to accurately compare amplifiers unique complex waveform processing (distortion profile). But, alas, we are mostly left with subjective analyses that end up with terms like “warm”, “bright” or “flat”.
When Bob Carver did the Carver challenge he used the term Transfer Function to describe the sum total of any and all deviations from perfect reproduction. The T series amplifiers were named after Transfer Function.
In engineering, a transfer function (also known as system function or network function) of an electronic or control system component is a mathematical function giving the corresponding output value for each possible value of the input to the device. It is often represented as a graph, called a transfer curve or characteristic curve. The transfer function provides information which specifies the behavior of the component in a system, and is used in the mathematical analysis of systems, particularly using the block diagram technique, in electronics, control theory, project management, and construction management.
In terms of sound physics, would a simple defn be wherein electrically reproduced sound closest replicates the original source that the ear can appreciate. In this way it seems warm or most natural.
The closest reproduction of the input signal wouldn't necessarily seem warm, particularly as what's described as "warm" is often due to colouration.
This is such a complicated subject because there are so many variables. One we haven’t discussed is volume. Volume clearly has an effect on sound perception, because that’s why many unit have a loudness mode. As the volume goes down, our perception of highs and lows diminishes. So any comparison of sound would need to include the question, What was the volume?
So, getting all the help you need in this thread?
I hear you on the "colouration" aspect of an "inout signal", as this topic was discussed in great detail on a thread in "Instruments" wherein "recording sessions" were explained/shared.
From the time that the sounds are recorded, then modified by sound boards, a lot of the original results have been changed and "coloured"...much to the chagrin of some of the artists who wanted something different. However the contracts that they signed did not give them final word on the final product/result.
Then to muddy the water even further, you have the people at home with both analogue and digital gear, further changing the end sounds to suit their own personal tastes...brightening or warming them.
However, would not the nature of ANALOGUE which offers a smooth an continuous infinite number of frequencies/harmonics reflect the real world of sound. This is what I was alluding to in the way of hearing natural sound, whereas DIGITAL has only a finite set of possible values, and attempts with tiny steps to approximate the original sound. So, these two worlds may come close, but will always be worlds apart?
If we could hear infinitely high frequencies, that would be true. However, we only hear up to 20khz or so, which means (per the Nyquist-Shannon Theorem) we only need double 20khz as a sampling frequency in order to digitally record and re-play audio with equal or greater accuracy than equivalent analog circuitry. Any difference between sampled values and actual signal values are converted to noise, which is usually lower level than the noise inherent in any analog system.
In short, the "tiny steps" that notionally represent the values between digital samples are either at frequencies well beyond our hearing, or they're noise well below normal background levels, and analog systems are no better than digital (and often worse) at dealing with both.
Just to clarify, how could the analog be "worse, if smooth curved and continuous? Is this not the way that we hear in the real world.
I do realize if enough sampling is done, most ears can't discriminate the difference other than sound techs and some musicians...as they propose.
Warm speaker is a speaker with a natural roll off in highs with accurate bass. Not enough roll off to dull a performance, but just enough so when engineers use mics with elevated high ends or place mics to close to the mouth the over emphasis of the highs isn't un-natural. young listeners have been over exposed to so many un-natural recordings that they have accepted them as faithful, which they are not. Listen to some one speak they next time you have a conversation. Does their voice have the sibilance of so many recordings or live performances on TV or at a sound re-inforced concert. Go to a live opera and sit in the 10th row or more. Is that sound what you hear on a recording? Nope. I sat in the middle and on the sides of brass bands, symphonic wind bands, orchestras, lab (big) bands and a few jazz septets. I mean unless the trumpet player is blowing directly in your face, which never happens in real life but always seems to happen on recordings, a natural roll off is the only thing that can relieve fatigue. Thats why Enoch Light used RCA ribbons to take the edge off recordings so emphasized by most condenser mics. Flat speaker is acceptable if the tweeter doesn't have a resonant high end and your willing to attenuate the highs 4 to 6 db with the tone controls. The problem with Klipsch speakers is when they introduced a new crossover to protect the tweeters it introduced a 6 db peak on the 6.3 Khz 1/3 octave. I haven't measured the series III speakers, but I admit they are a bit bright, but thats what tone controls are for.
I have a spectrum analyzer and when my friends start leaning on me about their systems I drag out the unit and check their systems. Its amazing what a little tweaking can do.
That's something to think about, more than just a little.
The output of a digital recording/transmission system is always analog, as is the input. So, from the point of view of your ears, digital and analog systems are both analog. What you hear is smooth, curved, and continuous in either case.
A typical wholly-analog audio chain might be microphone to tape to tape master to record master to record to stylus to cartridge to preamplifier to amplifier to speakers. Each part of that chain adds some distortion and noise.
A typical wholly-digital audio chain might be microphone to ADC (analog to digital converter) to DAC (digital to analog converter) to amplifier to speakers. The purely digital domain between the ADC and DAC -- if uncompressed or losslessly compressed -- adds no distortion or noise. The microphone, ADC, DAC, amplifier and speakers add some distortion and noise; for the mike, amplifiers and speakers, it's the same as their equivalents in a wholly-analog system. Thus, the overall distortion and noise must be less than or equal to that of an equivalent typical wholly-analog system -- assuming a good ADC and DAC -- as we no longer have the distortion and noise adding effects of the tape, tape master, record master, record, stylus, cartridge, and preamplifier.
With good ADC and DAC circuitry, and sufficiently high sampling rate and bit depth, the ears cannot discriminate between digital and analog because there's nothing to discriminate except the potentially lower noise and distortion of the digital system. The input is analog and the output is analog in either case.
Separate names with a comma.