What would you write into that "RAM"? There are no analog bits.
Bob said:... you must sample the
original signal at a rate equal to at least twice its bandwidth (actually,
very slightly higher, to avoid a particular degenerate case which
could occur at EXACTLY 2X the bandwidth) in order to preserve
the information in the original and avoid "aliasing."
And exactly what DO you hear when those signals run their "magic"?Radium said:Audio IS mechanical so something has to move.
Then how is it possible for a digital audio device [like an iPod] to
play audio back w/out moving a disc or tape? There is a digital chip
storing info and playback does not require moving anything. Just how
is this possible? If it works in digital, why can't it work in analog?
Ron said:Well, some types of RAM bits are stored as analog voltages
on a MOS gate capacitor. I think old CCD devices could
output some measure of the voltage per bit cell. Or you
could consider the charge digital if you could count the
number of electrons in each well.
Radium said:I'm curious to why there are no purely-analog devices which can
record, store, and playback electric audio signals [AC currents at
least 20 Hz but no more than 20,000 Hz] without having moving parts.
Most of those voice recorders that use chips [i.e. solid-state] are
digital. Analog voice recorders, OTOH, use cassettes [an example of
"moving parts"].
Radium wrote:2. Decreasing the spatial frequency of the images in the video-signal
without low-pass filtering the images or increasing their sizes. An
example of this would be making the sharp areas of an image look
duller without decreasing the "sharpness" setting [an example of low-
pass filtering] on the monitor or increasing the size of the image.
Normally, when the size of an image is decreased, its sharpness
increases [it's like compressing a lower-frequency sound wave into a
higher-frequency one]. Likewise, when the size of an image is
increased, it looks duller [like stretching a higher-frequency sound
wave into a lower-frequency one]. Low-pass filtering simply decreasing
the sharpness of an image while increasing its dull characteristics --
which is what I don't want.
That's a reasonable summary of what you don't want to do. What do you
think you might do instead?
#1 Decreases the rate at which objects in the video move without
decreasing the video's playback speed or eliminating originally-
rapidly-moving objects [such as the rapidly flapping wings]
Something has to give. If the flapping of the wings is slowed, so is the
motion of everything else.
Huh?
Says who? You're reasoning from false analogy again.
Sampled analog systems are certainly
not very common today (unless you count certain forms of
modulation as "sampling," and in fact there are some very close
parallels there), but the theory remains the same no matter which
form of encoding is used. In any event, you must sample the
original signal at a rate equal to at least twice its bandwidth (actually,
very slightly higher, to avoid a particular degenerate case which
could occur at EXACTLY 2X the bandwidth) in order to preserve
the information in the original and avoid "aliasing."
Radium said:Radium wrote:2. Decreasing the spatial frequency of the images in the video-signal
without low-pass filtering the images or increasing their sizes. An
example of this would be making the sharp areas of an image look
duller without decreasing the "sharpness" setting [an example of low-
pass filtering] on the monitor or increasing the size of the image.
Normally, when the size of an image is decreased, its sharpness
increases [it's like compressing a lower-frequency sound wave into a
higher-frequency one]. Likewise, when the size of an image is
increased, it looks duller [like stretching a higher-frequency sound
wave into a lower-frequency one]. Low-pass filtering simply decreasing
the sharpness of an image while increasing its dull characteristics --
which is what I don't want.That's a reasonable summary of what you don't want to do. What do you
think you might do instead?
The video-equivalent of changing the 'pitch' of audio recording
without changing the playback speed.
#1 Decreases the rate at which objects in the video move without
decreasing the video's playback speed or eliminating originally-
rapidly-moving objects [such as the rapidly flapping wings]Something has to give. If the flapping of the wings is slowed, so is the
motion of everything else.
The motion of 'everything else' *is* slowed. However, the playback
speed remains constant.
Repetitive or cyclical motion (such as a ball bouncing, or a wagon
wheel rotating, or a bird-flapping its wings, or an exposed model of a
piston engine operating, or a flag waving in the wind) in the movie
are slowed without lengthening the clip.
Sorry that should read "makes a still image less sharp by stretching
everything within the image without increasing the size of the image
or eliminating sharp portions of the original image"
My bad.
Anyways, this is an original picture:
http://www-dse.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/sab/report.normalimage.jpg
OK
This is how the picture looks after low-pass filtering -- YUK!:
http://www-dse.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/sab/report.lopass.jpg
I don't want low-pass filtering. I simply want all frequencies to be
downshifted similar to decreasing the pitch of audio without slowing
the playback speed. The analogy is lower the frequencies of all
components in the image w/out increasing the size of the image or
doing any low-pass filtering.
http://www-dse.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/sab/report.html
How is it false?
Radium said:Sampled analog systems are certainly
not very common today (unless you count certain forms of
modulation as "sampling," and in fact there are some very close
parallels there), but the theory remains the same no matter which
form of encoding is used. In any event, you must sample the
original signal at a rate equal to at least twice its bandwidth (actually,
very slightly higher, to avoid a particular degenerate case which
could occur at EXACTLY 2X the bandwidth) in order to preserve
the information in the original and avoid "aliasing."
Is the CCD [Charge Coupled Device] a "sampled analog system"?
The fact that it's an AC (inherently-varying) signal being recorded,
means that *something* has to move... if only some amount of
electrical charge. If the electrons don't move, the output can't vary
and all you have is a DC voltage.
And, in fact, this concept of moving electrical charges is the basis
for one type of analog signal storage and playback device which has no
moving (mechanical) parts... the CCD, or Charge Coupled Device. It
consists of a large number of charge storage devices (typically MOSFET
transistors with dielectrically-isolated gates) hooked up as a sort of
shift register or "bucket brigade". Each gate stores a charge which
is proportional to the input signal present at a given moment in time.
Several thousand times per second, a clock pulse causes each storage
cell to generate an output voltage proportional to the charge in its
storage gate, and then to "capture" onto its gate the signal being
presented by the previous gate in the chain.
In effect, the signal is propagated down the chain at a rate
proportional to the clock rate.
Why aren't these devices used more than they are? They're not very
efficient, and they're noisy. Every time the charge is copied from
one cell to the next, a bit of imprecision (noise) creeps in... so the
fidelity isn't great. And, because the device has to be able to hold
a very wide range of charges (since the charge is directly
proportional to the signal level) the storage gates have to be fairly
large.
The net result is that an audio CCD is capable of storing a
decent-quality signal for only a few tens or hundreds of milliseconds,
from input to output.
Another sort of a purely analog signal-storage device, with no moving
parts other than the electrons which convey the signal, is a simple
length of transmission line (with perhaps some amplifiers mid-way).
Put a signal in at one end, get the same signal back out the other end
some number of microseconds or milliseconds later.
Dave said:Radium said:I'm curious to why there are no purely-analog devices which can
record, store, and playback electric audio signals [AC currents at
least 20 Hz but no more than 20,000 Hz] without having moving parts.
Most of those voice recorders that use chips [i.e. solid-state] are
digital. Analog voice recorders, OTOH, use cassettes [an example of
"moving parts"].
The fact that it's an AC (inherently-varying) signal being recorded,
means that *something* has to move... if only some amount of
electrical charge. If the electrons don't move, the output can't vary
and all you have is a DC voltage.
And, in fact, this concept of moving electrical charges is the basis
for one type of analog signal storage and playback device which has no
moving (mechanical) parts... the CCD, or Charge Coupled Device. It
consists of a large number of charge storage devices (typically MOSFET
transistors with dielectrically-isolated gates) hooked up as a sort of
shift register or "bucket brigade". Each gate stores a charge which
is proportional to the input signal present at a given moment in time.
Several thousand times per second, a clock pulse causes each storage
cell to generate an output voltage proportional to the charge in its
storage gate, and then to "capture" onto its gate the signal being
presented by the previous gate in the chain.
In effect, the signal is propagated down the chain at a rate
proportional to the clock rate.
Why aren't these devices used more than they are? They're not very
efficient, and they're noisy. Every time the charge is copied from
one cell to the next, a bit of imprecision (noise) creeps in... so the
fidelity isn't great. And, because the device has to be able to hold
a very wide range of charges (since the charge is directly
proportional to the signal level) the storage gates have to be fairly
large.
The net result is that an audio CCD is capable of storing a
decent-quality signal for only a few tens or hundreds of milliseconds,
from input to output.
Another sort of a purely analog signal-storage device, with no moving
parts other than the electrons which convey the signal, is a simple
length of transmission line (with perhaps some amplifiers mid-way).
Put a signal in at one end, get the same signal back out the other end
some number of microseconds or milliseconds later.
Once again, they're not terribly efficient and are prone to be noisy.
For storage of large amounts of information, in a small space, with
high fidelity, using digital storage techniques is much more
efficient - largely because each storage cell must only store 2
different information states (0 and 1) rather than a large number of
possible levels.
charges represent the signal at a particular instant of its average over
a particular interval. (My CCD digital camera can take time exposures.)
A CCD's content may not be quantized in amount, but it is quantized in
time. In a camera, where the charges pertain to individual pixels, the
result is also quantized in space.
Radium wrote:
That's just arm-waving words. Describe the result, not as an analogy,
but as a specification. If it turns out that you can't think critically
after all, I have no time for you.
Explain how everything can slow town without increasing the time to
complete a motion. Sounds have duration and pitch. motion has no analog
of pitch in that sense. Describe the result you want, not "something
like" the result.
Tell me again how the crankshaft can take run one fifth speed without
using more time to make a turn.
Tell me again how everything in an image can be stretched to double size
without making the image twice as big.
Justify why you think that images and sounds are subject to the same
transformations.
Images have no visual equivalent of pitch. Pitch is temporal. Images are
spatial.
Is CCD a form of analog non-volatile RAM?
I wonder how a PC would perform if it used CCDs in place of digital
storage devices. Lots of errors.
Where is the "storage" in this device?
Where is the signal being stored?
Dave said:"Digital" and "subject to aliasing" are two different things.
As I believe the term "digital" is usually meant, it implies a
two-state (on/off) storage representation. It's not just that the
signal amplitude is quantized, but that the quantization uses a
power-of-two representation and storage system of some sort.
In that sense, an audio CCD uses a digital clocking structure to move
the charge along, but uses a non-digital system for representing the
signal level (a linear number of electrons). Yes, it's quantized in
time, and the electron charges themselves are quantized... but I don't
think that either of these qualifies it as "digital".
"Analog" is a very fuzzy and imprecise term, and I think that a CCD
can reasonably be called an analog system.
Even audio cassette tape is quantized in both time and amplitude, at
the level of the individual magnetic domains in the oxide or metal
particles.
"Digital" and "subject to aliasing" are two different things.
As I believe the term "digital" is usually meant, it implies a
two-state (on/off) storage representation. It's not just that the
signal amplitude is quantized, but that the quantization uses a
power-of-two representation and storage system of some sort.
Radium said:The purpose of this visual "pitch-shifting" is like a way to record/
playback/transmit/receive/store supreme-quality video while using the
least bandwidth and storage space necessary when low-pass filtering is
not an option.
Using this video frequency-shifting, a high-quality video can be
stored in an extremely slow moving video-cassette with limited amount
of tape. Due to the video-tape's extremely slow speed the temporal and
spatial frequencies of the incoming video signals must be downshifted
in order to be encoded at such slow speeds. Due to the limited length
of film in the cassette, the movie must not be made longer than what
it originally is. Due to other inadequacies in the film, the spatial-
frequency must also be decreased, but the image size must not
increase.
A 2 hour high-quality movie should be able to be stored in device with
limited high-frequency response and limited amount of storage space.
There should be absolutely no aliasing -- temporal or spatial - but at
the same time, the length of the movie should not be increased, sizes
of objects in images should not increase, image size should not
increase and no low-pass filtering should be used.
I wish I knew. This 'pitch-shifting' is a lot more confusing than I
thought. Yet I still find it so interesting. Sorry.
Tell me again how everything in an image can be stretched to double size
without making the image twice as big.
Nothing in the image has its size increased. They are simply smoothed
out.
This is similar to a graph of digital audio in Adobe Audition. You
decrease the pitch of the audio in the file by half [without changing
the tempo] and the waves in the graph will appear twice as long but
without increasing the horizontal length of the graph.
Justify why you think that images and sounds are subject to the same
transformations.
The less sample rate you have in digital audio, the lower the
frequency of the audio must be in order to prevent aliasing. There
isn't enough bandwidth to include the higher-pitches.
Similarly an imaging device with insufficient spatial bandwidth will
result in image distortion if excessively fine detail is put into the
camera.
Hence, if you want to get decent imagery in a low-bandwidth imaging
device, your best bet is to decrease the spatial frequency because
transferring it into the imaging device.
Just like if you have an 11.025-KHz-sample-rate digital audio device,
you need to make sure the pitch of the audio you are inputting into
the device does not exceed 5.5125 KHz.
Spatial frequency is how fine or dull an image is. Pitch is determined
by audio frequency. I am using the spatial frequency as an analogy.
If a digital audio device can play audio back without
any moving parts, why can't an analog audio device be designed to do
the same?
Another sort of a purely analog signal-storage device, with no moving
parts other than the electrons which convey the signal, is a simple
length of transmission line
Don said:My reading of the possible systems goes like this.
analogue - a continuous representation of the original signal
sampled - a representation of the signal at discrete time points
quantized - a sampled signal, but with the possible levels constrained
to a limited set of values
digital - a quantized signal, with the individual levels represented
by numbers
Aliasing is going to happen as soon as you move beyond the first line
of that list.
