i totally agree with you on the 1st point, but for the second point.. i don't think we're talking about the technique/skill of the driver here as what's important is the car itself. Whether the car has good handling is another matter, though, of course.
i for one wouldn't think that the 16-18bit grayscale is so important, plus in the case of MRI and CT image the resolution is much less than the resolution of a typical scanner (note that that's the resolution of the MRI/CT image per se, not the resolution of the film, which is actually quite good.)
HRCT is still the same story, although it mentions 'high resolution'.
Re:Depends on what you're looking for!
on
Film to X-rays?
·
· Score: 1
in the case of a lung mass, or a bone tumor, then if i was the doc i would probably take a PET CT and get a printed film. I couldn't think of a center using just X-Ray +/- contrast for a lung mass or a bone tumor.
btw i don't think we need the 3600x3600 dpi resolution to diagnose/monitor some arthritis, as far as i'm concerned, i'm happy with 1024x1024 image of CXR/AXR on monitor.
I know. As I said, interpolation, blah blah blah. My point here is that whether you interpolate it into a sine wave or leave it as a triangle/square, you are not *accurately* representing the original waveform, and therefore the sound will not be identical.
you are correct if, and only if the absolute highest frequency component is higher than the sampling frequency divided by 2. (I hope to use this wording to clarify a bit, the term 'nyquist frequency' has many meaning depend on which textbook you have read)
any signal, which, the highest frequency component is f, when sampled above frequency 2f will be accurately represented provided that there is infinite amplitude resolution. do we have anything disagreeing here?
and again, there is no, not even a single commonly used codec can accurate represent anything with a freqeuncy component of infinity.
Hardly. But I *can* hear the difference in things like mp3 vs uncompressed audio, so I know audiophiles are not entirely full of shit.
Ah, the joy of clarification - i think we're just in some miscommunication. Audiophiles are those people who would believe audio cables being directional, or that the design of power supply has anything to affect the quality of sound, or 2+2=5, for extremely large values of 2.
I think we're in agreement here. But again, find me some commercially successful music whose absolute highest component is 11KHz. I don't think you will. 11-22KHz is only one more octave, but the harmonics in that octave are very important for the timbre of the sound.
and i bet no more than 0.01% of all people can hear anything above 22kHz at -20dB with respect to peak CD volume; and any commercial record i've seen do not have anything above -50dB for signals beyond 22kHz.
Yes, i'd agree that the timbre of an instrument is related to its harmonics, but what i'd recommend you to try is to place a low pass filter starting at 18kHz, and see if you noticed anything in a reallife record. I once played around with some audiophiles with one single digital recording with and without 20kHz low pass and i've yet to see anybody could reveal this in a properly carried out ABX test.
I know. As I said, interpolation, blah blah blah. My point here is that whether you interpolate it into a sine wave or leave it as a triangle/square, you are not *accurately* representing the original waveform, and therefore the sound will not be identical.
i would be glad to tell you there is nothing such as a perfect triangle, square wave or so, be it naturally generated or electronically generated. anything that could be generated could be represented, though not everything generated could be heard; those that could be heard is provided in CD, and CD provided just that. codecs like musepack, or AAC provide even better compression on audio because it further take out the inaudible components of audio.
this is probably a consequence due to a very commonly used, pirated version of windows which unattendedly installs a version of windows xp in traditional chinese without asking for the hostname.
this version of install-disc is commonly shared among hong kong and taiwan bittorrent communities.
"I studied electroacoustic music and digital recording at university" and "My speciality is electronics" and then you don't even understand that when we're talking about nyquist frequency we're talking about the frequency component, after FFT, DCT, MDCT or [insert your favourite algorithm here], which are sine waves.
in fact, when human are hearing, in our inner ear different frequency compoennt of the sound wave stimulate different region, effectively doing a transform from the time domain to frequency domain, and as the cutoff of human hearing is around 20kHz (in fact, it's much lower in older people who are in their 10s or 20s, by 20kHz we're talkin' about kids)
> You may replace "violin" with the higher-pitched instrument of your choice =).
I'm sorry, but most instrument does not have fundamental frequencies above 20kHz and harmonics above 20kHz are usually too insignificant in terms of energy for them to be heard in reality. feel free to actually do an ABX test on a 24/192 record with and without a 20kHz low pass filter at 24dB/octave.
Yes, if the *absolute highest harmonic* is 11KHz
i thought by highest component i'm talking about an even more absolute measurement of frequency (note harmonics are also complex waveforms which can be broken into their own frequency components).
Here, I'll illustrate with a sawtooth wave, sampling frequency 44KHz:
I'm sorry, but there is nothing such as sawtooth in any natural instrument. if you want to say sawtooth in synth, you're free to do so, but the synth are usually giving you something your kind samples are doing anyway... sawtooth wave has a frequency component of *infinity* and thus cannot be sampled accurately by any commonly used technique.
The smoothness of these waves is (as I said) because of the interpolation my system is using. Without that, the closer I got to the Nyquist frequency, the more my wave would look like a triangle or square wave, depending on what data points were discarded and the shape of the original wave.
by the PCM standard, it will NEVER look like a triangle or square wave which both has a frequency component of infinity; PCM must pass through a low pass filter in order to accurately represent a waveform below nyquist frequency. i'm sorry, but by what you've done, i smell audiophile.
p.s. i'm an audio enginneer; and yes, i'm extremely picky
i think it's you that need to read up about digital audio and nyquist frequency.
Nyquist theorem If input signal has _maximum frequency_ (bandwidth) f, sampling frequency must be at least 2f
First, you are definitely wrong about 'for an example......four times a second', as when we are considering the nyquist frequency you should look at the frequency domain to see the highest harmonics; and a signal with highest frequency component of 11kHz is always faithfully sampled with 44kHz sampling (by that i mean if the bit resolution is high enough you can reconstruct the signal perfectly, yes. perfectly).
and for your information, a violin actually has frequency ranging from the low end of some hundred Hz to around 8kHz. 8kHz is often quoted in designing a violin to be a soft limit when you are tweaking parameters to get the violin sound good. for violin recording a low pass put at 16kHz usually result in transparent recording.
if you have a 11kHz component in a waveform and you sample that at 44kHz, you are sampling it 4 point per cycle, which is ADEQUATE to represent the original waveform.
and nothing could be reduced to a square wave, it is also not that higher frequency would result in sine wave, it would undergo aliasing and result in a very low frequency waveform.
though, as violin goes, it actually have harmonics that go up to around 30kHz when played sul ponticello, or up to 50kHz when playing double-stop; the reason why CD is said to be relatively transparent is because no Homo sapien can feel those frequency as far as i know from any properly controlled trials, some audiophiles claims they do, so i would wonder if they are in the Canis genus or not.
it does. there is selection pressure given everytime when tumor cells grow. e.g. immune system, chemo, radiation etc. so if you tried to treat cancer with drugs, chances are, if you do not eliminate them totally, what happen would be tumor cells expressing exporters of drugs.
in short,
(1) they reproduce (2) they are susceptible to changes
for spinal cord injury (brown-sequard syndrome (hemisection) or whatsoever), I don't think embryonic stem cells are going to do any good in this field unless we have a very careful computer controlled 'neuron repair' or 'neuron association' process which automagically associate neurons with what they should associate.
Right, neurons do have their plasticity in brain and even if connection goes wrong it can still work. But then do anybody here know what regenerates fastest when stem cells are put into spinal cord hemisections? C-fibres! those unmyelinated fibres which confers PAIN and temperature sense in the spinothalamic tract.
depends on how you think human experiements has been done in china but most of the result is basically they get something back, like 3 moving muscle in arm (that's not a hell lot, come on) and they also get some really painful life and become dependent on analgesics. Depends on how you think they may or may not be a good thing.
yes, it is. but i gotta say that they have some pretty good connection to HKIX (hong kong internet exchange) and overseas bandwidth is something like 20Mbps.
But nevermind, most Hong Kong people use bandwidth for local transfer anyway - that is - bittorrent of movies, apps, stuff like that.
can somebody tell me how can we eat to have chronic lower respiratory diseases (e.g. COPD, etc.) or how can we have septicemia by eating stuff? I would be glad to see a recipe.
there is nothing unnatural about anything created by human beings. we are a part of this world, and anything we do is just like what a herd, or thousands of herds of leopard, panda (as if they can do anything), cows, sheeps or tiny little mice do.
just to say a 'for-example'. those who do not even consider to benefit from the environment (reads: going to hospital when they have rectal bleeding) are doomed to extinction of their own genes.
nobody knows how intelligence is inherited, nor would there be any significant experiment be done in the future, assuming that the ethical committee (tm) does not change their crappy mind. (on that, i suppose if there is such thing as an ethical committee in the past years there won't be vaccinia vaccine (pardon for redundancy) at all.) -- (1) you will not be able to carry out human experiment, (2) you will be able to carry out animal experiment, but nobody cares the result, and (3) as all of us know, the sample size for identical twins studies are way too small for any significance citation.
but then while intelligence may or may not be highly inheritable, what gives rise to a human is highly inheritable, and identical twins tends to have similar intelligence despite being raised in different background.
and the problem isn't that the 'feebleminded' are outbreeding the able ones, but the fact that the community is trying hard to work against natural selection (reads: social security)...
Slashdot Mirror
i totally agree with you on the 1st point, but for the second point.. i don't think we're talking about the technique/skill of the driver here as what's important is the car itself. Whether the car has good handling is another matter, though, of course.
come on, this is worse than somethingawful or goatse...
> Women are pretty.
a significant portion of the female population in the world is not pretty
the cheapest way of getting a copy is to have it copied, by the radiology department. i think it cost less than anything else which give you a film.
i for one wouldn't think that the 16-18bit grayscale is so important, plus in the case of MRI and CT image the resolution is much less than the resolution of a typical scanner (note that that's the resolution of the MRI/CT image per se, not the resolution of the film, which is actually quite good.)
HRCT is still the same story, although it mentions 'high resolution'.
in the case of a lung mass, or a bone tumor, then if i was the doc i would probably take a PET CT and get a printed film. I couldn't think of a center using just X-Ray +/- contrast for a lung mass or a bone tumor.
btw i don't think we need the 3600x3600 dpi resolution to diagnose/monitor some arthritis, as far as i'm concerned, i'm happy with 1024x1024 image of CXR/AXR on monitor.
you are correct if, and only if the absolute highest frequency component is higher than the sampling frequency divided by 2. (I hope to use this wording to clarify a bit, the term 'nyquist frequency' has many meaning depend on which textbook you have read) any signal, which, the highest frequency component is f, when sampled above frequency 2f will be accurately represented provided that there is infinite amplitude resolution. do we have anything disagreeing here? and again, there is no, not even a single commonly used codec can accurate represent anything with a freqeuncy component of infinity.
Hardly. But I *can* hear the difference in things like mp3 vs uncompressed audio, so I know audiophiles are not entirely full of shit.
Ah, the joy of clarification - i think we're just in some miscommunication. Audiophiles are those people who would believe audio cables being directional, or that the design of power supply has anything to affect the quality of sound, or 2+2=5, for extremely large values of 2.
I think we're in agreement here. But again, find me some commercially successful music whose absolute highest component is 11KHz. I don't think you will. 11-22KHz is only one more octave, but the harmonics in that octave are very important for the timbre of the sound.
and i bet no more than 0.01% of all people can hear anything above 22kHz at -20dB with respect to peak CD volume; and any commercial record i've seen do not have anything above -50dB for signals beyond 22kHz. Yes, i'd agree that the timbre of an instrument is related to its harmonics, but what i'd recommend you to try is to place a low pass filter starting at 18kHz, and see if you noticed anything in a reallife record. I once played around with some audiophiles with one single digital recording with and without 20kHz low pass and i've yet to see anybody could reveal this in a properly carried out ABX test.
I know. As I said, interpolation, blah blah blah. My point here is that whether you interpolate it into a sine wave or leave it as a triangle/square, you are not *accurately* representing the original waveform, and therefore the sound will not be identical.
i would be glad to tell you there is nothing such as a perfect triangle, square wave or so, be it naturally generated or electronically generated. anything that could be generated could be represented, though not everything generated could be heard; those that could be heard is provided in CD, and CD provided just that. codecs like musepack, or AAC provide even better compression on audio because it further take out the inaudible components of audio.
this is probably a consequence due to a very commonly used, pirated version of windows which unattendedly installs a version of windows xp in traditional chinese without asking for the hostname.
this version of install-disc is commonly shared among hong kong and taiwan bittorrent communities.
"I studied electroacoustic music and digital recording at university" and "My speciality is electronics" and then you don't even understand that when we're talking about nyquist frequency we're talking about the frequency component, after FFT, DCT, MDCT or [insert your favourite algorithm here], which are sine waves.
in fact, when human are hearing, in our inner ear different frequency compoennt of the sound wave stimulate different region, effectively doing a transform from the time domain to frequency domain, and as the cutoff of human hearing is around 20kHz (in fact, it's much lower in older people who are in their 10s or 20s, by 20kHz we're talkin' about kids)
> You may replace "violin" with the higher-pitched instrument of your choice =).
I'm sorry, but most instrument does not have fundamental frequencies above 20kHz and harmonics above 20kHz are usually too insignificant in terms of energy for them to be heard in reality. feel free to actually do an ABX test on a 24/192 record with and without a 20kHz low pass filter at 24dB/octave.
Yes, if the *absolute highest harmonic* is 11KHz
i thought by highest component i'm talking about an even more absolute measurement of frequency (note harmonics are also complex waveforms which can be broken into their own frequency components).
Here, I'll illustrate with a sawtooth wave, sampling frequency 44KHz:
I'm sorry, but there is nothing such as sawtooth in any natural instrument. if you want to say sawtooth in synth, you're free to do so, but the synth are usually giving you something your kind samples are doing anyway... sawtooth wave has a frequency component of *infinity* and thus cannot be sampled accurately by any commonly used technique.
The smoothness of these waves is (as I said) because of the interpolation my system is using. Without that, the closer I got to the Nyquist frequency, the more my wave would look like a triangle or square wave, depending on what data points were discarded and the shape of the original wave.
by the PCM standard, it will NEVER look like a triangle or square wave which both has a frequency component of infinity; PCM must pass through a low pass filter in order to accurately represent a waveform below nyquist frequency. i'm sorry, but by what you've done, i smell audiophile.
p.s. i'm an audio enginneer; and yes, i'm extremely picky
i think it's you that need to read up about digital audio and nyquist frequency.
...four times a second', as when we are considering the nyquist frequency you should look at the frequency domain to see the highest harmonics; and a signal with highest frequency component of 11kHz is always faithfully sampled with 44kHz sampling (by that i mean if the bit resolution is high enough you can reconstruct the signal perfectly, yes. perfectly).
Nyquist theorem
If input signal has _maximum frequency_ (bandwidth) f, sampling frequency must be at least 2f
First, you are definitely wrong about 'for an example...
and for your information, a violin actually has frequency ranging from the low end of some hundred Hz to around 8kHz. 8kHz is often quoted in designing a violin to be a soft limit when you are tweaking parameters to get the violin sound good. for violin recording a low pass put at 16kHz usually result in transparent recording.
if you have a 11kHz component in a waveform and you sample that at 44kHz, you are sampling it 4 point per cycle, which is ADEQUATE to represent the original waveform.
and nothing could be reduced to a square wave, it is also not that higher frequency would result in sine wave, it would undergo aliasing and result in a very low frequency waveform.
though, as violin goes, it actually have harmonics that go up to around 30kHz when played sul ponticello, or up to 50kHz when playing double-stop; the reason why CD is said to be relatively transparent is because no Homo sapien can feel those frequency as far as i know from any properly controlled trials, some audiophiles claims they do, so i would wonder if they are in the Canis genus or not.
you're right about the sampling part, though.
I do not think much of musicians playing electric guitar has any substantial knowledge of how the MOSFETs (or tubes) in their nice amp works.
IMHO, they are smart if they know their music well, and they can produce decent music.
it does. there is selection pressure given everytime when tumor cells grow. e.g. immune system, chemo, radiation etc. so if you tried to treat cancer with drugs, chances are, if you do not eliminate them totally, what happen would be tumor cells expressing exporters of drugs.
in short,
(1) they reproduce
(2) they are susceptible to changes
==> (3) they evolve
well, not really.
for spinal cord injury (brown-sequard syndrome (hemisection) or whatsoever), I don't think embryonic stem cells are going to do any good in this field unless we have a very careful computer controlled 'neuron repair' or 'neuron association' process which automagically associate neurons with what they should associate.
Right, neurons do have their plasticity in brain and even if connection goes wrong it can still work. But then do anybody here know what regenerates fastest when stem cells are put into spinal cord hemisections? C-fibres! those unmyelinated fibres which confers PAIN and temperature sense in the spinothalamic tract.
depends on how you think human experiements has been done in china but most of the result is basically they get something back, like 3 moving muscle in arm (that's not a hell lot, come on) and they also get some really painful life and become dependent on analgesics. Depends on how you think they may or may not be a good thing.
ketamine-bp
more accurately, in china, the saying is "if its belly face downwards, its edible"
well, they are actually doing what we call Fiber to the home (as opposed to fiber to the building "FTTB").
only the last few meters (as opposed to last mile) is Cat5(e).
yes, it is. but i gotta say that they have some pretty good connection to HKIX (hong kong internet exchange) and overseas bandwidth is something like 20Mbps.
But nevermind, most Hong Kong people use bandwidth for local transfer anyway - that is - bittorrent of movies, apps, stuff like that.
can somebody tell me how can we eat to have chronic lower respiratory diseases (e.g. COPD, etc.) or how can we have septicemia by eating stuff? I would be glad to see a recipe.
there is nothing unnatural about anything created by human beings. we are a part of this world, and anything we do is just like what a herd, or thousands of herds of leopard, panda (as if they can do anything), cows, sheeps or tiny little mice do.
just to say a 'for-example'. those who do not even consider to benefit from the environment (reads: going to hospital when they have rectal bleeding) are doomed to extinction of their own genes.
I've got karma to burn, so:
nobody knows how intelligence is inherited, nor would there be any significant experiment be done in the future, assuming that the ethical committee (tm) does not change their crappy mind. (on that, i suppose if there is such thing as an ethical committee in the past years there won't be vaccinia vaccine (pardon for redundancy) at all.) -- (1) you will not be able to carry out human experiment, (2) you will be able to carry out animal experiment, but nobody cares the result, and (3) as all of us know, the sample size for identical twins studies are way too small for any significance citation.
but then while intelligence may or may not be highly inheritable, what gives rise to a human is highly inheritable, and identical twins tends to have similar intelligence despite being raised in different background.
and the problem isn't that the 'feebleminded' are outbreeding the able ones, but the fact that the community is trying hard to work against natural selection (reads: social security)...
ketamine
the court hearing is said to be held in 1st april?
i don't know if they're doing that, but if you are using adblock, just
Tools|Adblock|Preferences...
select [x] Hide ads
use adblock
o m/*
block:
http://www.clicksor.com/*
http://z1.adserver.c
http://www.paypopup.com/*
well i think you should note that these service cost the taxpayers to setup. defeated government service = taxmoney wasted, period.
government and corporation going together, surprise, surprise.
well, to many, a ratio is not something presented in dB, so perhaps you could as well accept his ignorance (sp?).
btw, i suppose with spread-spectrum (aka spam) techniques one can transmit information (reads: advertisement) even when in a negative SNR time.