Hrow wrote:
Congrats to FMers. According to my son who is a student there, we are now being studied by researchers from the University of Pennsylvania who are looking at the impact of Internet rumors on product marketing. Needless to say, the basic premise of the study isn't one that is all that favorable and boils down to how easy is it to get relatively smart people to believe relatively stupid things without having one shred of evidence pointing in any direction whatsoever. Don't worry, FM isn't alone in being targeted and hopefully we will get higher marks than some of the other sites being looked at....Show more →
Henry, what have you done
Didn't your son tell you you shouldn't let the subjects of observation know that they are being observed? That might affect the study in a dramatic way!
The phenomenon isn't new, I'd guess. Call it Hype, or worse, Mass Hysteria. It's just a lot faster and has a lot more reach, thanks to the world wide web. Anyway, if you can, let us know of the findings, sounds like an interesting subject.
Hrow wrote:
Congrats to FMers. According to my son who is a student there, we are now being studied by researchers from the University of Pennsylvania who are looking at the impact of Internet rumors on product marketing. Needless to say, the basic premise of the study isn't one that is all that favorable and boils down to how easy is it to get relatively smart people to believe relatively stupid things without having one shred of evidence pointing in any direction whatsoever. Don't worry, FM isn't alone in being targeted and hopefully we will get higher marks than some of the other sites being looked at....Show more →
"FF is expensive to make, so it is not that easy to just drop the price of a 5D to that of a 30D."
FF doesn't have to match the current price of a 30D in order to be successful. It simply has to provide a cost/value that makes in competitive among buyers who might otherwise have purchased a 30D. There is room between the price of a 30D and some higher price for a FF body that costs less that the list price of the current 5D.
On a related topic, clearly a 5D at the current list price would not compete with a new 12 MP camera like the one Puts suggests. But there is quite a bit of room between a $4500 "pro" body and a sub-$1000 400D. At least enough room for a less expensive 5D-like body and perhaps room for a more capable 5D-like body as well?
"Congrats to FMers. According to my son who is a student there, we are now being studied by researchers from the University of Pennsylvania who are looking at the impact of Internet rumors on product marketing..."
And congratulations to you for invalidating their study - unless the the purpose of the study is to measure differences between a discussion thread that knows it is being studied and a control group that does not know. (Researchers cannot assume that study subjects behave "normally" when they know they are being observed.)
If it is a real study (as oppposed to a "class project") your son may well have some difficult and awkward (and, I hope for him and you, not career changing) explaining to do about how he compromised his research.
Dan
Edited by danmitchell on Feb 11, 2007 at 09:01 AM GMT
Didn't your son tell you you shouldn't let the subjects of observation know that they are being observed? That might affect the study in a dramatic way!
The phenomenon isn't new, I'd guess. Call it Hype, or worse, Mass Hysteria. It's just a lot faster and has a lot more reach, thanks to the world wide web. Anyway, if you can, let us know of the findings, sounds like an interesting subject.
You are right, mass hysteria is nothing new and often its value is under appreciated. For example, had not the fine folks in Salem been so diligent, the USA might be a nation of witches today.
"And yes, my “understanding” is that Canon have tried very hard to make these two cameras have everything that could be keeping a top end film photographer from moving to digital…. You can expect a range of accessories / grips etc to match..."
The few people who shoot high end 35mm do so because of computer skills (no, not everyone of them, I'm making a point) I don't see how this camera addresses that. There's also a high correlation between age and film.
So if Canon is taking this position, I guess they're referring to camera size. 1V high end is cool. Better than EOS-3, I think.
But I do understand a milestone hear. There are no major functional/ergonomic areas left to conquer from the 35mm film era. But that's more of a Canon thing than a photographer thing, IMHO.
Returning to Erwin Puts comments and the strange remarks about lack of the 1D memory buffer that appear to be some botched translation. Some have speculated that this is a continuous jpeg mode where, assuming memory cards are fast enough, pictures will be written to your card until it is full without speed or burst length limitations from buffer size/speed.
Why jpeg mode? Due to lower than full res jpegs? What about RAW mode? Any reason that would be slower than full res jpeg mode? It would seem to require LESS processing.
jcbenner wrote:
Returning to Erwin Puts comments and the strange remarks about lack of the 1D memory buffer that appear to be some botched translation. Some have speculated that this is a continuous jpeg mode where, assuming memory cards are fast enough, pictures will be written to your card until it is full without speed or burst length limitations from buffer size/speed.
Why jpeg mode? Due to lower than full res jpegs? What about RAW mode? Any reason that would be slower than full res jpeg mode? It would seem to require LESS processing.
File size.
If we assume that they are saying taht the JPEGs at 8MP can be stored without using a buffer, then the highest quality JPEG is around5MB, RAW 8MB, and RAW + JPEG 11MB, using figures form the 1D11
If we assume that it is the full 12MB that are stored as JPEG without using the buffer, taking the figures for the 5D we get JPEG 3.5MB, RAW 11.5MB and RAW + JPEG 13.5MB
If my memory serves the reason for the larger 1DII JPEG at 8MB is that the 1DII has some sort of super high definition JPEG available.
If I weren't too lazy I would check on the flash card to see how many MB/sec they will handle - something like a Sandisk extreme should do best.
Not that it will matter a hoot, as plainly it will not affect the user whether the buffer is being used for any reasonable burst, unless you want to kill the shutter in one session! - How long does it take at 8fps to do 150,000 clicks?
Regards,
jcbenner wrote:
Returning to Erwin Puts comments and the strange remarks about lack of the 1D memory buffer that appear to be some botched translation. Some have speculated that this is a continuous jpeg mode where, assuming memory cards are fast enough, pictures will be written to your card until it is full without speed or burst length limitations from buffer size/speed.
Why jpeg mode? Due to lower than full res jpegs? What about RAW mode? Any reason that would be slower than full res jpeg mode? It would seem to require LESS processing.
That's likely down to simple throughput numbers. An auto-crop 1.3x image (from a 12 mpixel full frame sensor) will be 8 mpixel. At JPG fine, that's about 3 megabyte per image. Give or take some. (My 6.3 mpixel 10D averages 2.2 megabyte over thousands of JPG fine shots.)
It's easy to determine approximate throughput numbers. If you take the write speeds of the fastest performing CF cards out there, Sandisk's Extreme IV pushes up to 40 MB/sec, you can see that 8 FPS continuous is in the cards. (Pun intended. )
Okay, now try to write 8 RAW files to card in one second. That's not 3 MB per image, but around 8 MB per image. Current high-speed CF cards can't handle 64 MB per second quite yet...
Tentacle wrote:
That's likely down to simple throughput numbers. An auto-crop 1.3x image (from a 12 mpixel full frame sensor) will be 8 mpixel. At JPG fine, that's about 3 megabyte per image. Give or take some. (My 6.3 mpixel 10D averages 2.2 megabyte over thousands of JPG fine shots.)
It's easy to determine approximate throughput numbers. If you take the write speeds of the fastest performing CF cards out there, Sandisk's Extreme IV pushes up to 40 MB/sec, you can see that 8 FPS continuous is in the cards. (Pun intended. )
Okay, now try to write 8 RAW files to card in one second. That's not 3 MB per image, but around 8 MB per image. Current high-speed CF cards can't handle 64 MB per second quite yet......Show more →
No problem storing the around 4MB images for 12MP either without using the buffer, on the extreme 1V
Regards,
DaveMart wrote:
No problem storing the around 4MB images for 12MP either without using the buffer, on the extreme 1V.
That depends on the framerate. If SanDisk claims "up to 40 MB/sec" then I'll build in a little margin myself. The tiny letters on the SanDisk site warn that the 40 MB/sec number is based on their own testing. It's the combination DIGIC III / High-speed CF that needs to deliver, not some best-scenario fine-tuned cardreader in a test set-up.
It may be a stretch, but 32 MB/sec might be possible. If it doesn't, then we'll have to wait for the next gen CF cards and see if Extreme V and the likes will enable sustained full frame 8 FPS bursting.
dcmiller wrote:
Maybe it writes jpg/large/fine continuously only to SD. That way it's a Canon controller.
DIGIC III improves AF and AE calculations and probably bring fps up to 10
What if the sensor is not bayer? Hmmmm?
10fps, sure - that's within the shutter and mirror speeds of the 1v - and should just about go on to the Extreme 1V in 8MP mode according to Tentacle's calculations - but I doubt it is not bayer - not a whisper of that in these latest leaks.
Regards,
dcmiller wrote:
Maybe it writes jpg/large/fine continuously only to SD. That way it's a Canon controller.
DIGIC III improves AF and AE calculations and probably bring fps up to 10
What if the sensor is not bayer? Hmmmm?
Frame rate, in terms of maximum output, is mostly bottlenecked by the number of output channels of the sensor and the sample rate of the pre-processing engine, which is a discrete chip that sits between the sensor and DIGIC. At least, that's the case with DIGIC II.
Regarding Bayer vs Foveon output is tricky. If we're talking RAW numbers, then it's easy: Foveon has a higher number of photo sites, compared to a bayer equivalent sensor.
It's a different situation if we'd try to compare JPG outputs of equivalent sensors. The size of a compressed image isn't determined by resolution, but by information content. But now we're drifting into Information Theory. If that holds true, then a foveon JPG image should have about the same size as it's Bayer equivalent size image.
The best speed data I know of for flash memory is from Hans-Jürgen Reggel , and
in his testing, the Extreme IV is able to write at speeds of about 26MB/second. His testing shows that he can write to a Extreme III card at around 17MB/second. So it may be possible to get close to 6 fps JPEG from a 12MP camera written to todays fastest CF cards. However I wouldn't say that would be a easy task.
If canon can write to flash memory at anywhere near 26MB/second, then we all should be very happy and or praise them a great deal because.... according to Rob Galbraith, the fastest any camera he has tested can go is about 9.4MB/second, that being the Nikon D80, Canon's fastest writing camera to date goes at around 8MB/second.. that being the 1Ds MIII
