Well 32000 is within reach for high speed B&W films when pushed. I believe Kodak Tmax 3200 is happy being pushed 3 or 4 stops, which will take it to ISO 50,000 or so. I've pushed it 2 stops and got good results. It'll be contrasty as hell, but that's half the fun.
Kodak only sell Tmax 3200 in 35mm canisters, as the metal canister is required to stop the film being fogged by cosmic rays - that'll give you an idea how sensitive it is!
Cesare
I do a fair bit of DSP development work, and used to work full time doing that sort of thing. Fan noise is really difficult to deal with when you are trying to hear some odd clicks/pops in some DSP algorithm you are developing.
My solution was to put the machines in the neighbouring room, with the cables fed through a hole in the wall. It might sound a little extreme, but it took 1/2 hour with a jigsaw.
Well, my original 802.11b airport is still going strong after a fair number of years. I bought it with an iMac G3/450, so i'd guess that was 1999 or somewhere around there.
It's handled dialup access, bridging a wired ethernet to wireless machines, and now hangs off my adsl switch/modem as a wireless bridge. It really was a revolutionary product.
An Anton Piller order is very serious, and is only handed out in extreme situations. It is used when you want to protect your IP (Intellectual Property), and when you can demonstrate that directly approaching the offending company/individual will most likely cause them to destroy the evidence.
For example, say you are a brand owner, and you find there is a company setup in a factory somewhere making counterfeit goods, if you approached then, they would disappear overnight, only to continue from a different location, with all evidence gone. An Anton Piller order allows you to 'bust' the operation, and seize basically whatever you want in order to stop the damage to your brand (typically the manufacture of counterfeit goods).
I shoot a Canon D30, which is a 3.2mp dSLR. I get prints of comparible quality to 35mm film. I tend to do 6*9 inch prints, or mount 10*15 inch prints.
The digital processing needed to get a 10*15 inch print (at 300dpi) from a 3.2mp image isn't difficult. I usually apply an unsharp mask to the original, then do a bicubic interpolation to the required resolution. A little extra sharpening if you want, and this gives a very acceptable print.
Which 5mp camera have you used which led to these problems? Why are you so sure this is a problem with 5mp, and not a problem with the particular model of camera that you were (are) using?
All PowerPC Macs have come with ethernet, so that's since around 1992 or so. The current range all have gig ethernet, including powerbooks.
You are right about the graphic cards as being a joke though - especially as apple have used OpenGL acceleration in their GUI (Quartz Extreme). The apple UI is really improved by a decent amount of video ram. You'd have to factor a 256Mb card, especially if you are running multiple screens.
Another point which should be pointed out is that these Apple cards are dual DVI. For whatever reason noone seems to think that the average PC user wants dual DVI...
Cesare
Granted, bitmaps don't scale very well in certain situations, and vectors do, but this doesn't mean that you should use vectors for everything. Apple sensibly use a large bitmap size which is scaled down for most display purposes, and scaling bitmaps down is pretty good. The application icons are something like 256*256, so you rarely want to scale them up.
Let's face it, if you scale an icon to 4*4 pixels, it's not going to look good no matter if it started life as a vector or a bitmap:-)
And while we're on the subject, and someone please hire a graphic artist at microsoft and get some decent icon's put together? Windows icons are really really poor, and make the UI look cheap. It's not going to break the bank to get an artist in for 6 months...
BTW, about your laptop thing - why can't you set a larger font size in XP at the moment? I'm not sure I understand.
Except your mailbox overflows when you are away on holiday for a couple of weeks because of the volume of spam.
Cesare
Re:So isn't the megapixels rating a farse?
on
70 Megapixel Webcam
·
· Score: 3, Informative
Well, technically speaking most high end SLRs can't actually take the whole scene in one go at fast shutter speeds.
Shutters on cameras are made of two curtains - the first one is normally closed, the second one normally open. When you depress the shutter release, the first curtain opens to start the exposure, and the second curtain closes to end the exposure. At fast shutter speeds (1/125 on my manual pentax, 1/200 on my Canon D30) the second curtain starts to close before the first curtain has finished opening. In effect then, the film or sensor is exposed to a slit of light between the opening and closing curtains.
So, it's maybe worth remembering that when you say 'you can expose the whole scene in one go' with an array sensor, it isn't always as simple as that!
You're point is very valid about interpolation. Some cameras are advertised with 'effective' resolutions. Saying that, even a real 3mp camera is interpolating 9mp of colours from 3mp of colour information (since each cell is only receiving one colour pixel).
OSX wasn't designed for PPC - it is layered on the Mach/FreeBSD, which is a mature UNIX, and the Apple guys owe alot to the Next guys, who were working on 680x0 processors.
Windows XP? Designed? Well i'll let that pass, but given that it is derived from the NT source which was released on many different architectures in it's time (including PPC BTW) it's far from x86 specific.
Also the PPC isn't very RISC, it's quite a large instruction set. And as for x86 processors being CISC, they gave up on that a long time ago., and all the processors you are likely to see are CISC cores with a x86 translator.
I'm sorry, but we are talking about a company here.
It's pointless to anthropomorphise companies and suggest that they have in any way behaved in an underhand manner - they are just a name.
You want to really understand what has motivated the company staff to rip off your government, and it will probably come down to the greed of one or two people. The guys involved were probably on a performance related bonus scheme, so had an incentive to increase their sales. Lot's of the people working on this project were probably unaware of the markup, unaware of what the requirements were, they were just doing what they had been told to do.
Well, I don't think this is what is going on. 1024 bits of resolution per cell? I don't think so.
Most high end photographic CCDs are 12 bit, and I guess 16 bit will be available if you pay enough money. Remember, the CCD charge must be sampled, and you just don't get 1024 bit samplers. 24 bit samplers usually produce 18 or 20 bits max or signal, and the rest is just noise. I'm sure these guys have stuff an order of magnitude better than that, but that would be another 3 bits...
I'm guessing the 1 Gig figure is for all 60 CCDs, so there is actually 1024/60 bits per cell, which works out around 17, so this is probably an approximate figure, and the CCD is read at 16 bit resolution.
I was wondering if he is actually able to edit a 2Gb file in Photoshop.
I wonder what limits it has for file editing. I seem to remember it having a pixel limit of 64000 although that may be rubbish (or removed in an earlier Photoshop release).
Also, the article states that he corrects the colours 'from memory'. Seems odd that he doesn't take a 35mm shot at the same time to use for colour balance.
Yeah, the virus scanner people generate a fair bit of traffic themselves. It does go to show though that there is a fine line between providing a useful service to the less well informed, and spam.
Putting it another way, i'd prefer if the average joe who doesn't know much about this stuff at least was informed that their machine was infected with a mail sending virus. If only a few of them acted on it...
Well, come to the UK and sit stationary on the M25 some time if you want to explore the 'building more lanes increases traffic flow' idea:-)
The only problem with your idea is that if anything, we've been moving away from parellel to serial forms of communication between devices in the last few years. Most modern protocols are serial, replacing the older parallel standards (think USB, Firewire, Serial IDE etc). The reason for this is the difficulty in implementing parallel busses as the speed increases. You basically get crosstalk problems between your lines.
Patents are filed with broad claims and specific claims. Other companies get to comment on the claims. I had a quick look at the site and it wasn't clear whether this has been granted, but the filing date is Oct 28 2002 so prior art must obviously predate the filing date.
It looks to me like the broadest claims cover 'multimedia player' so will cover video/audio players. Who knows of the claims which ones Apple expects to succeed with, and with which they are trying it on.
Same here in the UK. My wife works for a firm of patent agents and the advice is always 'don't discuss this with anyone before filing'. You can't patent what you have disclosed to other parties (unless under an NDA).
There are plenty of other ways that Apple could protect its IP in the UK. They'll have design rights for example. It might be that in the US there is some confusion about what is and what is not patentable, and for large companies the advice is then 'file first, ask questions later'. Even if it isn't granted, it will rattle through the system for a number of years giving Apple some room against competitors.
I find it interesting to note that Linux is deemed insecure when it is likely that the attacks weren't on the kernel, but on processes running on the system (e.g a buggy/badly configured Apache server).
Shouldn't therefore the comment be about the relative security of Apache vs IIS etc? How come Linux gets it in the neck when there is an Apache vulnerability?
As for the low OSX figures, this is probably more related to the number of deployments out there, but that is a different matter.
I'd also be interested to see the Linux figures broken down by platform. I'd suspect that most attacks will be buffer overruns on x86 machines. By running Linux on a different processor (say sparc) you probably avoid these attacks because the exploits haven't been coded for this processor.
Your colour vision is slower than your peripheral black/white only vision. Your peripheral vision is probably faster to help you detect movement - i'm sure someone here will tell you why (or at least have an opinion, it is/. after all).
If you don't look directly at a monitor such that it is still within your vision the flickering of slower frame rates is very obvious. For example, if you read a book placed on your desk in front of your monitor, you'll probably find the monitor distracting (since you see the flicker).
This story is complete nonsense. It has so little going for it i'm surprised it even made it onto/.
Let's get this right. The Russians can't write a bit of control software for a pipeline? I'm not sure they would even have bothered. They had plenty of people to throw at problems like this. They would have posted people around the place to monitor pressures etc. Dull work, but that was the Soviet way of doing things.
If they had a gas explosion, most likely cause was someone screwed up, not a piece of control software.
And as for 'we can make software pass quality checks but then fail in use' well of course you can. I write code like that for a living!
I went from a IIcx (16Mhz 68020 with a 68881 FPU) to a 7500 (100Mhz PPC 601) and got something like a 100x speedup when running some of my floating point based code.
So no, the 601 didn't suck. It was about as fast as you could get in any PC at the time.
I think you are remembering running emulated 680x0 code on the PPC which of course takes a big hit. A wonder of software engineering, but if I remember the emulated machine didn't do FPU correctly, so any code which had FPU optimised code paths didn't get run in the emulator, which caused quite a slowdown.
The 68040 was quite a change from the earlier processors (I think the '040 was the first with a pipelined architecture, and it included FPU on board), and so it had a considerable speedup.
The 68000 was a real CISC processor with all the trappings. I remember those days well, but don't miss it;-)
I'm quite aware that the XP isn't an Athlon 64. I'm also quite aware of the prices of processors (and how much I paid for them when they were released). I'm not entirely sure how the price of a processor affects it's performance though - maybe you can help me out with that one.
Also, I can compare the performance of processors at the same frequency. I can even compare the performance of processors at completely different frequencies as well. That's what i'm doing. It's called benchmarking.
My question remains, anyone willing to run a test harness on an Athlon 64 for me? I'd be interested to see how my code runs on it.
Slashdot Mirror
Well 32000 is within reach for high speed B&W films when pushed. I believe Kodak Tmax 3200 is happy being pushed 3 or 4 stops, which will take it to ISO 50,000 or so. I've pushed it 2 stops and got good results. It'll be contrasty as hell, but that's half the fun. Kodak only sell Tmax 3200 in 35mm canisters, as the metal canister is required to stop the film being fogged by cosmic rays - that'll give you an idea how sensitive it is! Cesare
I do a fair bit of DSP development work, and used to work full time doing that sort of thing. Fan noise is really difficult to deal with when you are trying to hear some odd clicks/pops in some DSP algorithm you are developing.
My solution was to put the machines in the neighbouring room, with the cables fed through a hole in the wall. It might sound a little extreme, but it took 1/2 hour with a jigsaw.
Cesare
Well, my original 802.11b airport is still going strong after a fair number of years. I bought it with an iMac G3/450, so i'd guess that was 1999 or somewhere around there.
It's handled dialup access, bridging a wired ethernet to wireless machines, and now hangs off my adsl switch/modem as a wireless bridge. It really was a revolutionary product.
An Anton Piller order is very serious, and is only handed out in extreme situations. It is used when you want to protect your IP (Intellectual Property), and when you can demonstrate that directly approaching the offending company/individual will most likely cause them to destroy the evidence.
For example, say you are a brand owner, and you find there is a company setup in a factory somewhere making counterfeit goods, if you approached then, they would disappear overnight, only to continue from a different location, with all evidence gone. An Anton Piller order allows you to 'bust' the operation, and seize basically whatever you want in order to stop the damage to your brand (typically the manufacture of counterfeit goods).
Cesare
I shoot a Canon D30, which is a 3.2mp dSLR. I get prints of comparible quality to 35mm film. I tend to do 6*9 inch prints, or mount 10*15 inch prints.
The digital processing needed to get a 10*15 inch print (at 300dpi) from a 3.2mp image isn't difficult. I usually apply an unsharp mask to the original, then do a bicubic interpolation to the required resolution. A little extra sharpening if you want, and this gives a very acceptable print.
Which 5mp camera have you used which led to these problems? Why are you so sure this is a problem with 5mp, and not a problem with the particular model of camera that you were (are) using?
All PowerPC Macs have come with ethernet, so that's since around 1992 or so. The current range all have gig ethernet, including powerbooks. You are right about the graphic cards as being a joke though - especially as apple have used OpenGL acceleration in their GUI (Quartz Extreme). The apple UI is really improved by a decent amount of video ram. You'd have to factor a 256Mb card, especially if you are running multiple screens. Another point which should be pointed out is that these Apple cards are dual DVI. For whatever reason noone seems to think that the average PC user wants dual DVI... Cesare
Granted, bitmaps don't scale very well in certain situations, and vectors do, but this doesn't mean that you should use vectors for everything. Apple sensibly use a large bitmap size which is scaled down for most display purposes, and scaling bitmaps down is pretty good. The application icons are something like 256*256, so you rarely want to scale them up.
:-)
Let's face it, if you scale an icon to 4*4 pixels, it's not going to look good no matter if it started life as a vector or a bitmap
And while we're on the subject, and someone please hire a graphic artist at microsoft and get some decent icon's put together? Windows icons are really really poor, and make the UI look cheap. It's not going to break the bank to get an artist in for 6 months...
BTW, about your laptop thing - why can't you set a larger font size in XP at the moment? I'm not sure I understand.
Cesar
Except your mailbox overflows when you are away on holiday for a couple of weeks because of the volume of spam. Cesare
Well, technically speaking most high end SLRs can't actually take the whole scene in one go at fast shutter speeds.
Shutters on cameras are made of two curtains - the first one is normally closed, the second one normally open. When you depress the shutter release, the first curtain opens to start the exposure, and the second curtain closes to end the exposure. At fast shutter speeds (1/125 on my manual pentax, 1/200 on my Canon D30) the second curtain starts to close before the first curtain has finished opening. In effect then, the film or sensor is exposed to a slit of light between the opening and closing curtains.
So, it's maybe worth remembering that when you say 'you can expose the whole scene in one go' with an array sensor, it isn't always as simple as that!
You're point is very valid about interpolation. Some cameras are advertised with 'effective' resolutions. Saying that, even a real 3mp camera is interpolating 9mp of colours from 3mp of colour information (since each cell is only receiving one colour pixel).
Bzzz, wrong.
OSX wasn't designed for PPC - it is layered on the Mach/FreeBSD, which is a mature UNIX, and the Apple guys owe alot to the Next guys, who were working on 680x0 processors.
Windows XP? Designed? Well i'll let that pass, but given that it is derived from the NT source which was released on many different architectures in it's time (including PPC BTW) it's far from x86 specific.
Also the PPC isn't very RISC, it's quite a large instruction set. And as for x86 processors being CISC, they gave up on that a long time ago., and all the processors you are likely to see are CISC cores with a x86 translator.
Apart from that, a very well argued case.
I'm sorry, but we are talking about a company here.
It's pointless to anthropomorphise companies and suggest that they have in any way behaved in an underhand manner - they are just a name.
You want to really understand what has motivated the company staff to rip off your government, and it will probably come down to the greed of one or two people. The guys involved were probably on a performance related bonus scheme, so had an incentive to increase their sales. Lot's of the people working on this project were probably unaware of the markup, unaware of what the requirements were, they were just doing what they had been told to do.
Let there be light....
Cesare
Well, I don't think this is what is going on. 1024 bits of resolution per cell? I don't think so.
Most high end photographic CCDs are 12 bit, and I guess 16 bit will be available if you pay enough money. Remember, the CCD charge must be sampled, and you just don't get 1024 bit samplers. 24 bit samplers usually produce 18 or 20 bits max or signal, and the rest is just noise. I'm sure these guys have stuff an order of magnitude better than that, but that would be another 3 bits...
I'm guessing the 1 Gig figure is for all 60 CCDs, so there is actually 1024/60 bits per cell, which works out around 17, so this is probably an approximate figure, and the CCD is read at 16 bit resolution.
I was wondering if he is actually able to edit a 2Gb file in Photoshop.
I wonder what limits it has for file editing. I seem to remember it having a pixel limit of 64000 although that may be rubbish (or removed in an earlier Photoshop release).
Also, the article states that he corrects the colours 'from memory'. Seems odd that he doesn't take a 35mm shot at the same time to use for colour balance.
Yeah, the virus scanner people generate a fair bit of traffic themselves. It does go to show though that there is a fine line between providing a useful service to the less well informed, and spam.
Putting it another way, i'd prefer if the average joe who doesn't know much about this stuff at least was informed that their machine was infected with a mail sending virus. If only a few of them acted on it...
Because once a solution becomes commercial, the spammers get hold of it and work out how to modify their spam so that it gets through.
Well, come to the UK and sit stationary on the M25 some time if you want to explore the 'building more lanes increases traffic flow' idea :-)
The only problem with your idea is that if anything, we've been moving away from parellel to serial forms of communication between devices in the last few years. Most modern protocols are serial, replacing the older parallel standards (think USB, Firewire, Serial IDE etc). The reason for this is the difficulty in implementing parallel busses as the speed increases. You basically get crosstalk problems between your lines.
Patents are filed with broad claims and specific claims. Other companies get to comment on the claims. I had a quick look at the site and it wasn't clear whether this has been granted, but the filing date is Oct 28 2002 so prior art must obviously predate the filing date.
It looks to me like the broadest claims cover 'multimedia player' so will cover video/audio players. Who knows of the claims which ones Apple expects to succeed with, and with which they are trying it on.
Same here in the UK. My wife works for a firm of patent agents and the advice is always 'don't discuss this with anyone before filing'. You can't patent what you have disclosed to other parties (unless under an NDA).
There are plenty of other ways that Apple could protect its IP in the UK. They'll have design rights for example. It might be that in the US there is some confusion about what is and what is not patentable, and for large companies the advice is then 'file first, ask questions later'. Even if it isn't granted, it will rattle through the system for a number of years giving Apple some room against competitors.
I find it interesting to note that Linux is deemed insecure when it is likely that the attacks weren't on the kernel, but on processes running on the system (e.g a buggy/badly configured Apache server).
Shouldn't therefore the comment be about the relative security of Apache vs IIS etc? How come Linux gets it in the neck when there is an Apache vulnerability?
As for the low OSX figures, this is probably more related to the number of deployments out there, but that is a different matter.
I'd also be interested to see the Linux figures broken down by platform. I'd suspect that most attacks will be buffer overruns on x86 machines. By running Linux on a different processor (say sparc) you probably avoid these attacks because the exploits haven't been coded for this processor.
Your colour vision is slower than your peripheral black/white only vision. Your peripheral vision is probably faster to help you detect movement - i'm sure someone here will tell you why (or at least have an opinion, it is /. after all).
If you don't look directly at a monitor such that it is still within your vision the flickering of slower frame rates is very obvious. For example, if you read a book placed on your desk in front of your monitor, you'll probably find the monitor distracting (since you see the flicker).
So, no caffeine required!
This story is complete nonsense. It has so little going for it i'm surprised it even made it onto /.
Let's get this right. The Russians can't write a bit of control software for a pipeline? I'm not sure they would even have bothered. They had plenty of people to throw at problems like this. They would have posted people around the place to monitor pressures etc. Dull work, but that was the Soviet way of doing things.
If they had a gas explosion, most likely cause was someone screwed up, not a piece of control software.
And as for 'we can make software pass quality checks but then fail in use' well of course you can. I write code like that for a living!
I went from a IIcx (16Mhz 68020 with a 68881 FPU) to a 7500 (100Mhz PPC 601) and got something like a 100x speedup when running some of my floating point based code.
;-)
So no, the 601 didn't suck. It was about as fast as you could get in any PC at the time.
I think you are remembering running emulated 680x0 code on the PPC which of course takes a big hit. A wonder of software engineering, but if I remember the emulated machine didn't do FPU correctly, so any code which had FPU optimised code paths didn't get run in the emulator, which caused quite a slowdown.
The 68040 was quite a change from the earlier processors (I think the '040 was the first with a pipelined architecture, and it included FPU on board), and so it had a considerable speedup.
The 68000 was a real CISC processor with all the trappings. I remember those days well, but don't miss it
Ok :-)
Still looking for an Athlon 64 to try. None of my friends have (yet) taken the plunge.
I'm quite aware that the XP isn't an Athlon 64. I'm also quite aware of the prices of processors (and how much I paid for them when they were released). I'm not entirely sure how the price of a processor affects it's performance though - maybe you can help me out with that one.
Also, I can compare the performance of processors at the same frequency. I can even compare the performance of processors at completely different frequencies as well. That's what i'm doing. It's called benchmarking.
My question remains, anyone willing to run a test harness on an Athlon 64 for me? I'd be interested to see how my code runs on it.