From: Patent Tactics, George Brock-Nannestad
I am just latching on to Don's mail - I absolutely agree, but I am not in
agreement with a lot of what was written while I was away for a few days.
First of all, I have to admit that it is unlikely that any pre-amp should
tolerate a spark at the input - Tom Fine's complaint. In such a case, a
circuit consisting of properly biased germanium diodes or transistors would
reduce the max input to + or - 200 mV. The effect of feedback loop saturation
is a very real one, and those amps with the best stationary signal
performance are frequently the worst to display this on very fast input.
Headroom and short feedback loops are what we need.
I would have liked the discussion on analog vs. digital to have been a bit
more informed, but on the other hand one cannot request any hardware and
software user to have studied and understood a book like John Watkinson's
'The Art of Digital Audio' in order to permit them to write on the ARSCLIST.
For this reason we must be thankful for the many who have taken time to
elucidate a point; in this patchwork manner the general knowledge will
I am going to make a few notes that some will find boring, and I hope some
will find faulty, because that will provoke them to express themselves as
clearly as possible to correct me. The general problem is that sometimes
something that is as clear as water to some is very murky to others, and the
discussion does not take place in the same context - we all carry our own
around with us.
Now, most of what we are dealing with is bandwidth. Bandwidth and time
resolution are linked in the world of resistors, capacitors, and inductors,
but not so where our hearing is concerned. For solid proof of this go to
Milind Kunchur's webpage - they have been mentioned before on this list.
The so-called Nyquist criterion states that if you want to be able to
reconstruct a signal from a sampled representation of it, you have to decide
the highest frequency you want represented and then you have to sample at a
rate that is at least twice the highest frequency. When you then reconstruct
your signal by providing voltages at the sample rate according to the table
of values that represents the audio signal, you will have your full bandwidth
and dynamic range back, provided the resolution or bit-depth has been
sufficient. It has been claimed that this stepwise presented waveform makes
the signal unlistenable, but that is not the case, because you invariably
smooth it by filtering, so nothing above your defined highest frequency gets
out in the analog domain again. You will note that you are in full control:
define the maximum frequency and define the resolution. If they are not
sufficient to your purpose, go higher.
In order to have a sampled representation of our analog sound signal (which
is itself an electrical representation of a pressure variation around
atmospheric pressure) we have to measure it at regular intervals defined by
the sample rate and hold the value for as long it takes to obtain the tabular
value - the value that goes into the digital word stream.
Now, we have a problem with pure sampling of a waveform: if it has a
frequency that is more than half of the sampling rate, that too will be
sampled, but in this case under-sampled, which means that the result appears
to be at a completely different and inharmonic and jarring frequency, an
alias. Once we have adopted a sampling frequency we must simply ensure that
no signal above half the sampling rate is available for sampling.
This is done by filtering, so-called anti-alias filtering. With 44.1 kHz
sampling rate, no signal above 22.05 kHz is permissible. On the other hand,
we do want our 20 kHz bandwidth - this is what a young, pre-earbud ear can
mostly hear. So, our filter has to go from full transmission at 20kHz to zero
transmission at 22.05 kHz *). No problem, our telephone engineers have done
this kind of exercise for 90 years. However, such sharp cut-off filters come
with some frightful time delay distortion (phase to some), very audible down
to 2 kHz. That was the situation for about 10 years in CD audio, until
someone came up with the idea to correct the time response in the digital
domain by means of a digital filter. The currency to pay for this is total
delay time, but for something recorded a year ago, some microseconds do not
In other words, many of our problems come from the filter. If we increase the
sampling rate we can make use of our more frequent samples in two ways: we
can do a gentler filtering that does not have the delay effect at low
frequencies, or we can increase our highest frequency. If we do the latter,
we shall be able to increase the time resolution of our digital
representation. With 20 kHz the maximum slope of the waveform is only one
quarter of that of an 80 kHz bandwidth - reachable by a 192 kHz sampling
frequency and a gentler anti-aliasing filter.
Dependent on where you are on the scale of knowing electronics at graduate
level, you will miss a discussion of one-bit sampling and also of dithering.
I am not out to write a book. John Watkinson did. There is one thing that I
would love to see in the market: an interactive program where you could take
a number of input waveforms and vary the parameters bit depth, bandwidth,
sampling rate, anti-alias filter type and parameters, all to experiment on,
to see what things look like when you are in control. On the mock-ups based
on this principle I have seen, one of the most impressive is "oops, where has
it gone", when a frequency or a delay between two signals suddenly disappear.
Or when two frequency response peaks close by suddenly merge into one.
Understanding WHY they have gone seems to me to be very important.
A couple of other items came up while we were at 78rpm reproduction:
- the reason why we need an elevated bandwidth for recordings on rough
surfaces is because that is where the noise signal is. If we are to deal
properly with the noise it has to be represented correctly. And 80 kHz is a
marked improvement and will also permit us to look into low-level artefacts
at supersonic frequencies. I would refer you to my paper "What are the
Sources of the Noises We Remove", Proceedings of AES 20th Int. Conf.
'Archiving, Restoration, and New Methods of Recording', Eds. Z. Vajda, H.
Pichler, Budapest 5-7 October 2001, pp. 175-182. Unless someone else makes
the claim, I may well have been the first to state that in restoration work
using preservation copying, the bandwidth is not determined by the intended
signal but by the noise. This was at the IASA Conference in Bogensee near
Berlin in 1994.
- one might consider that the pickup would be encountering a formidable
vertical wall when it met a square wave. However, the recording is usually a
velocity recording, that is the square waveshape is differentiated (1st
derivative), which means it turns into a triangular wave. The problem is that
with a high bandwidth, the corners of this triangular wave (where the square
shifts from constant positive to constant negative and vice versa) are very
sharp, and even that may be difficult to trace. George Alexandrovich made a
7" test record with a "square wave" - there are virtually no wiggles at the
corners of the triangles. It is for testing pickups, but I have not dared to
use it on anything but the ELP.
- for the same reason a tick, modelled by a steep rise followed by steep
fall becomes two ticks, one positive and one negative.
- even if you remove these two peaks and interpolate or draw a waveform
connecting the ends, the low frequency excitation of the whole
cartridge-tonearm system (stylus, cantilever, bearing, cartridge mass,
tonearm mass and resonances) still remains as a low-level thud - a tail.
CEDAR started back in 1988 with a program developed by Peter Rayner while
still at the British Library to remove not only the ticks, but also the tail.
Well, for those of you who have not yet dropped off -
*) Actually it is not zero, but determined by the noise floor in the desired
digital representation or bit depth. Every bit added gives 6dB more headroom.
> On 11/04/2012, Tom Fine wrote:
> > As for the idea that 192kHz is somehow inferior to 96k, that sounds
> > like audiophoolery or outdated ideas based on older hardware. I would
> > say it's unecessary overkill to sample at that high a rate, for the
> > reasons you stated, but the playback results shouldn't sound inferior
> > to 96k unless something is being done wrong by the equipment or the
> > user.
> The main advantage of high sampling rates is that it is easier for
> software to distinguish clicks (impulses) from music.
> The waveform of a muted trumpet (as played by Buck Clayton, for
> instance) is very similar to a rapid sequence of clicks, so software
> operating at 44.1 KHz cannot automatically tell the difference.
> But the trumpet signal will fade out at the high frequency limit of the
> recording equipment, while clicks will have higher frequency components.
> Don Cox
> [log in to unmask]