Ok, started to disassemble to .dll-file. Sorry to say, but if Canon EOS-1DX is the awaited camera, it seems like it's still reading only 12bits for color depth from the raw sensor data.
Function called (in the same library):
_psParseGetMainImageFromCR2@12
Notice: @12
But beware, my assembler knowledge is a bit outdated and the same is true for my build in parser
Anyway, I have to get some sleep now ... any volunteer to continue in the meantime ?
chaosman wrote:
Ok, started to disassemble to .dll-file. Sorry to say, but if Canon EOS-1DX is the awaited camera, it seems like it's still reading only 12bits for color depth from the raw sensor data.
Function called (in the same library):
_psParseGetMainImageFromCR2@12
Notice: @12
But beware, my assembler knowledge is a bit outdated and the same is true for my build in parser
Anyway, I have to get some sleep now ... any volunteer to continue in the meantime ?
I think the name 1Dx may be just some sort of internal Canon name for the 1DII series - see this old post: http://forums.dpreview.com/forums/read.asp?forum=1019&message=5783965
That makes some sense as all the other cameras are old models.
Regards,
DaveMart
The reason I ask is that this info is very consistent with my source (who visited an old school friend who was a EOS marketing / product development 'insider' in Canon Japan). He (on a mission from me to find out) seemed very confident, though (being back in early July) the dates were just all EOS cameras replaced 'by February / March 2007'. If your source is different and saying (basically) the same this helps (me at least) get more confident in the rumour. (As confident as you can be anyway!)
In which case I can seriously worry about whether to buy the 40D or 7D (and go full frame)... The fun of it!
philllie1 wrote:
and what about your much promoted 3d
Phil
Don't be too quick to dismiss the 3D. The thing that might change is it's FF. Everything in those 3D specs could be delivered by Canon if they wish. However, 2007 is going to be enormous IMO. I don't expect anything for photokina that hasn't been announced unless they drop a bombshell on opening day.
Tentacle wrote:
We've been here before. What's a 10D over a 300D? Magnesium body, a bit faster, better AF, more functions. What's a 20/30D over a 350D? Magnesium body, more fps, better AF, more functionality. Ok, and the 30D has spotmetering over the 20D. Leaves you a pretty good idea on what to expect.
Noise-wise? We've been here too. 300/350/400D doesn't have ISO3200, 10/20/30D does. Each time we see slightly better SNR, extended ISO 3200 looks better each generation. Expect a similar step with the 40D. Remember that the increase in pixelcount negates some of the SNR gain.
Native ISO3200? I really really hope so, but I am not expecting it, based on what we've seen before with the 10/20/30D evolution. Remeber that DIGIC III will only bring better software based noise reduction to the table. It all depends on the sensor (and not on DIGIC) if ISO 3200 will be true or extended. ...Show more →
Expect improved (expanded AF) for the 40D. The fact the entry level AF is the same as the 30D indicates to me Canon will distinguish 40D from 400D by improving the AF along with the other usual differences. Buffer will probably go to 15 RAW, to put more distance between 400D. I would expect it will use a different sensor of 10.4MP with improved AA filter and microlens array, plus with digic III, lower power (= lower heat) deliver lower noise and of course offer ISO 3200. 40D should be a much bigger jump than 30D was IMO.
chaosman wrote:
Ok, started to disassemble to .dll-file. Sorry to say, but if Canon EOS-1DX is the awaited camera, it seems like it's still reading only 12bits for color depth from the raw sensor data.
Function called (in the same library):
_psParseGetMainImageFromCR2@12
Notice: @12
But beware, my assembler knowledge is a bit outdated and the same is true for my build in parser
Anyway, I have to get some sleep now ... any volunteer to continue in the meantime ?
The @12 indicates that the return value from the function is 12 bytes large. That means it is probably either a double or a structure/object. Either way, I don't think we can deduce image bit-depth from it.
Ha, great minds think alike - I was just contemplating hunting through the DPP files for strings.
One way to check whether the 1Dx is a new camera would be to find out how long it's been in DPP dll's - if it dates back to v1.6 then clearly it isn't new ;)
Sadly I suspect Photokina is done from a Canon perspective. Roll on PMA, and the 40D (based on the usual 1.5yr cycle for the nnD range).
Mark Shaxted wrote:
They could be internal names used by Canon software - D2000C = 20D?
Nope it was an ancient DIgital camera from Canon, before Christ... Exactly D2000. There's a special *.EXE convertor for it aside DPP Updater. at the same download page
Pixel Perfect wrote:
Expect improved (expanded AF) for the 40D. The fact the entry level AF is the same as the 30D indicates to me Canon will distinguish 40D from 400D by improving the AF along with the other usual differences.
Buffer will probably go to 15 RAW, to put more distance between 400D. I would expect it will use a different sensor of 10.4MP with improved AA filter and microlens array, plus with digic III,
IMHO nothing prevent Canon to throw 5D's AF-system into 40D body.
And they could boost the burst mode up to 6fps (with "choosable" 3fps like in 30D). DigicII had enough power to process 64Mp per second as seen from both 1D/1Ds II bodies
I might be a bit slow here but has anyone noticed the new Pentax has 9 cross sensors in its AF system. Or am I reading it wrong? this can only mean super AF system and I hope Canon will follow in their thinking.
I still feel the Dustbuster should remain where it is at the moment at entry level. I don't see the need for it and find it so easy to keep my sensor clean.
Also any one at Canon reading this Pleas follow and weather seal the body. I would love to buy a 5D but it wont improve my photos only time and practice will, I can't warrant spending NZ$5 000.
I would expect it will use a different sensor of 10.4MP with improved AA filter and microlens array, plus with digic III, lower power (= lower heat) deliver lower noise and of course offer ISO 3200. 40D should be a much bigger jump than 30D was IMO.
DIGIC III is a discrete chip, likely in BGA-package (judging from DIGIC and DIGIC II chips). That means whatever heat DIGIC III generates, which can be expected to be lower than it's predecessors, will NOT influence the SNR of the sensor.
Please note that you should see 2 things separate:
- Refinements in sensor design and production (better prevention of unwanted impurities) improve the Signal to Noise Ratio. This is usually partially offset by a bump in pixeldensity. This will determine if true ISO 3200 will be possible.
- DIGIC II vs DIGIC III: the latter has more processing power and is likely to have a smaller power draw. This will enable more sophisticated software based noise reduction, which in turn will make ISO3200 look better than before.
The new processing chip will have no say in wether true ISO 3200 is possible, or if we have to do with extended (underexposed and software pushed) ISO 3200. That depends purely on the qualities of the new CMOS sensor, not on DIGIC.
"Get ready for EOS 7D = stripped down smaller 5D $2100
40D new AF, 10.4MP digic III,
5D II new AF, 16.2MP, faster"
all three of those bodies have their attraction (although i'm wondering how you can strip a 5d any further without taking away FF), but they are all very close in range, with no 1-series body. if you took out the 7D and added a 1ds3 or 1d/1ds combination camera, it would make more sense...at least to me.
When the 400D came out and was so close to what my friend had been told in his Japan visit I asked him to stretch a friendship and find out what DIGIC III did for SLRs and what about the EOS 1 series. It’s taken a few weeks but I have a (fairly vague but still interesting) answer.
DIGIC III is a faster CPU supporting higher data transfer rates and allowing many more on the fly computations in processing an image. Some examples of how this can be used to deliver benefits in an SLR are:
• Support for faster write speeds to new generation CF cards allowing larger images at high FPS rates. This includes the potential for (?) continuous write for JPEGS up to a certain size.
• ‘Much more sophisticated and faster AF tracking and processing’, especially balancing between multiple AF points and tracking moving objects.
• More advanced noise reduction algorithms, allowing higher ISO to be used and providing smoother medium ISO images.
• More accurate exposure processing for better dynamic range preservation.
• Better image processing for generating JPEGs that are usable straight out of the camera and have higher dynamic range / preservation of detail in bright and dark areas.
Me want… Me want…
The EOS 1 Series is apparently outside my friend’s contact’s product / marketing area (Canon apparently treat the professional SLRs as their own specialist market). However he said:
• They (Canon) had been working on two different professional cameras for a while, but would probably only release one for now (? I don’t know what this means)
• Several generations of prototypes of both cameras had been in use for some time by selected pro-photographers.
• The new EOS 1 series had recently had a design upgrade which was delaying its availability.
• Canon was keen to release it as soon as possible as it it would be a flagship camera that was ‘untouchable’ in the market.
Not much new here really- but still interesting (for those that want / can afford / need a true pro camera).
Again, don’t know how real this is, but the previous info from this source was spot on… And my friend swears his Canon Japan contact is well connected…
When the 400D came out and was so close to what my friend had been told in his Japan visit I asked him to stretch a friendship and find out what DIGIC III did for SLRs and what about the EOS 1 series. It’s taken a few weeks but I have a (fairly vague but still interesting) answer.
DIGIC III is a faster CPU supporting higher data transfer rates and allowing many more on the fly computations in processing an image. Some examples of how this can be used to deliver benefits in an SLR are:
• Support for faster write speeds to new generation CF cards allowing larger images at high FPS rates. This includes the potential for (?) continuous write for JPEGS up to a certain size.
• ‘Much more sophisticated and faster AF tracking and processing’, especially balancing between multiple AF points and tracking moving objects.
• More advanced noise reduction algorithms, allowing higher ISO to be used and providing smoother medium ISO images.
• More accurate exposure processing for better dynamic range preservation.
• Better image processing for generating JPEGs that are usable straight out of the camera and have higher dynamic range / preservation of detail in bright and dark areas.
Me want… Me want…
The EOS 1 Series is apparently outside my friend’s contact’s product / marketing area (Canon apparently treat the professional SLRs as their own specialist market). However he said:
• They (Canon) had been working on two different professional cameras for a while, but would probably only release one for now (? I don’t know what this means)
• Several generations of prototypes of both cameras had been in use for some time by selected pro-photographers.
• The new EOS 1 series had recently had a design upgrade which was delaying its availability.
• Canon was keen to release it as soon as possible as it it would be a flagship camera that was ‘untouchable’ in the market.
Not much new here really- but still interesting (for those that want / can afford / need a true pro camera).
Again, don’t know how real this is, but the previous info from this source was spot on… And my friend swears his Canon Japan contact is well connected…
Geoff ...Show more →
All a bit depressing really, as there is no real increase in DR or high ISO noise performance talked about there, only fooling around with it some more in software - let's hope that they don't just end up with watercolours.
I suppose though strictly speaking you asked about DIGICIII, and a real increase in high ISO or DR would be as a result of a different sensor.
Thanks for your information, which I certainly find credible.
Regards,
DaveMart
DaveMart wrote:
All a bit depressing really, as there is no real increase in DR or high ISO noise performance talked about there, only fooling around with it some more in software - let's hope that they don't just end up with watercolours.
[...]
As much as you want it, consider this: Canon has no reason to leapfrog all existing specs. The competition doesn't warrant it. A 10 mpixel 40D will likely be a match for the much more expensive D200, and that will be enough. It will also be able to deal with the rest of the 10 mpixel bunch: the D80, Sony Alpha 100 and the Pentax K10D.
Canon will not blaze away its advantage (assuming it has the tricks ready and waiting) if there is no need to. Compare this to the x86 CPU industry, or the GPU industry. You do enough to get the performance crown back so you get the customers. Going completely overboard and "pwning" the competition 3 times over is counter productive in the long term because you will have difficulties improving on your own technology. So, you use your advantage piece-meal. The big strides are only used when there's catching up to do, not when you're in front.
Oh no? How about the 39MP Imacon (Hasselblad), and Phase One backs - that I am going to be forced to buy if Canon doesn't upgrade it's EOS 1 Camera this month? When they came out with the 1Ds Mk2 (which many think has already been made obsolete by their own 5D about a year ago), they were right up there with the best from the medium format cameras. Now they are way again behind. (And let's not even bring up the WA lenses... but at least there is a solution for that).
Come on Canon - don't force me to buy those big bulky slow expensive cameras again!
Corne wrote:
I might be a bit slow here but has anyone noticed the new Pentax has 9 cross sensors in its AF system. Or am I reading it wrong? this can only mean super AF system and I hope Canon will follow in their thinking.
This is the same AF sensor layout as the Pentax *istD, *istDS, *istDS2, K100D and Nikon D2H.
Don't know about the Nikon. But on the older Pentax models, this AF sensor is really nothing to get excited about.
Understand the AF system on the K10D is a lot better. It better be.
As much as you want it, consider this: Canon has no reason to leapfrog all existing specs. The competition doesn't warrant it. A 10 mpixel 40D will likely be a match for the much more expensive D200, and that will be enough. It will also be able to deal with the rest of the 10 mpixel bunch: the D80, Sony Alpha 100 and the Pentax K10D.
Canon will not blaze away its advantage (assuming it has the tricks ready and waiting) if there is no need to. Compare this to the x86 CPU industry, or the GPU industry. You do enough to get the performance crown back so you get the customers. Going completely overboard and "pwning" the competition 3 times over is counter productive in the long term because you will have difficulties improving on your own technology. So, you use your advantage piece-meal. The big strides are only used when there's catching up to do, not when you're in front....Show more →
That's self-contradictory.
You say that Canon doesn't need to introduce anything significantly better than its current specs because it has no competition, then you say that if they do leapfrog their current specs, they won't be able to do it again. That's fuzzy logic, a technological Maginot Line.
A. If they leapfrog their current specs, that puts them even farther ahead of the competition and insulates them better against a surprise release from someone else.
B. If they don't leapfrog their own specs, then it looks like they already have difficulty improving their own technology--look at the disappointment over the 30D.
C. There is no reason to think they " will have difficulties improving on your own technology" any more in the future than they have in the past.
D. Planning to make big strides only after you've fallen behind is how people lose races. The other guy may have more steam at the finish than you think.
Comparing the camera industry to the microprocessor industry is a misleading analogy. There are no wildcard players in the processor industry and a lot less chance of being blindsided by disruptive technology (like Minolta's introduction of marketable autofocusing, which obsoleted Canon's T90 a year before the T90 even hit the streets), nor are end-consumers the major customers of processor manufacturers--the "value-adding" computer manufactures make a difference.
I doubt Canon ever forgets the Maxuum 7000 or the possibility that something like it could happen again, any more than nations put trust in the Maginot Line concept.
?) continuous write for JPEGS up to a certain size.