John Power wrote:
I don't see 8K. Last year's model produces too good a product to warrant an additional 3K. I think they have ridden that 8K horse as long as they could have...
The 8K was mearely a reflection of the fact that there is no alternative ( 16MP FF ). The status is still unchanged.
Am I the only one who thinks that the announcements are not over yet? I mean, sheesh, the show has not even started! They might just be giving the 1DIII a little time in the spotlight before they announce more stuff...
Supposedly they have leased the largest amount of space at PMA ever (by a large margin). The 1DIII announcement is not, to my eyes, worth THAT much interest. I think there might be something else coming.
flatdraft wrote:
i, for one, will be looking forward to the $8000 tax writeoff that my new 1Ds mkIII will be providing me this year.
John Power wrote:
I don't see 8K. Last year's model produces too good a product to warrant an additional 3K. I think they have ridden that 8K horse as long as they could have...
i'm predicting it'll fall between $7199 and $7999 but realistically banking on it being $8k. i foresee canon making the 1DsIII improved enough to charge full price.
either way, my accountant is going to ask the same question he always does... "you paid how much for a digital camera?
Am I the only one who thinks that the announcements are not over yet? I mean, sheesh, the show has not even started! They might just be giving the 1DIII a little time in the spotlight before they announce more stuff...
Evidently, at the UK press event announcing the 1DM3, Canon confirmed that the present equipment announced will be all there is for PMA. Check Keith's Northlight Images page.
My finger has been hovering over the 1dsM2 buy button on B&H for the past month. All these darn threads about the 1dsM3 have been stopping it from landing. Stop teasing me already canon!
D. von Briesen wrote:
Evidently, at the UK press event announcing the 1DM3, Canon confirmed that the present equipment announced will be all there is for PMA. Check Keith's Northlight Images page.
I have my doubts about that confirmation and whether it really came from someone at Canon with the authority to make that statement. I don't believe they would make that statement at all as I don't understand why they would. Someone may have said it but it may have just been an opinion.
- The pixel pitch is the same for both the 1DIII and 1DsII sensor.
- Both sensors are designed and made by Canon, on Canon designed and built lithography gear.
- Both cameras will have the same Canon know-how for Bayer to pixel interpolation, and since Canon didn't trumpet any major breakthrough in this regard for the 1DIII, that mechanism will be more or less the same as well.
Canon DID trumpet a number of major breaks-through:
(i) wider micro-lenses
(ii) 14 bits
(iii) better noise/processing (Digic III)
(iiii) ...just kidding
Noise is like grain in that it hurts resolution. The 1D3 has less of it (note the 6400 iso setting). Therefore the 1D3 is likely to resolve more than a 10 megapixel crop from the 1Ds2.
To me as a non-technician it looks like they took a 2592 x 3888 piece out of 1Ds II sensor, because this has advantages in production ?
Then they improved the microlenses, increased the photodiode area within the pixel area to that of the 1DII ( and 5D ? ) and increased the bit-depth to 14
Wouldn’t it be relatively easy for them to implement these improvements in a FF sensor with the same pixel size ?
So can we expect a 16.7 mp FF sensor with these improvements in the next 1Ds or 5D ?
I think I heard the same thing happen at Best Buy regarding the 30d being leaked, maybe it was the 20d, either way, it proved to be true. (I talked to my Canon rep afterward and he seemed quit pissed about it, but wouldn't confirm it). The 18 month cycle was also complete though, so....maybe I'll stop there over my lunch hour and try the same approach and see what happens.
I'll go by my local Best Buy tonight and see what they say. They may also have one at Circuit City... if they both say the same thing, then something's up.
Ok, so everyone is excited about the new 1D3 and it looks like there are some nice new features. I'm still a FF fan and have been for decades; neither the 1.3 or 1.6 crop have ever been my cup of tea as my primary interest.
From Canon's 1D3 white paper, we know that the 1D3 and 1Ds2 both have identical pixel sizes (7.2 x 7.2), but they don't share the same density or diode size. The pixel sizes of the 1D2N and 5d are both 8.2 x 8.2, the 30D is 6.4 x 6.4, and the XTi is 5.7 x 5.7.
here's the text from page 10 of the white paper.
Although the pixel size of the 1D Mark III is 1 micron smaller than the pixel size of the EOS-1D Mark II and Mark II N, the photodiode size of both sensors is the same, thanks to the optimized photodiode construction and more sophisticated processing of the 1D Mark III.
By optimizing the gap between the on-chip microlenses and improving the fill factor
(photodiode area divided by total pixel size) of each pixel, light-gathering efficiency has been improved. Also, by optimizing the photodiode's structure, light-reception efficiency has been improved. Additionally, the second-generation, on-chip noise removal circuit and the noise reduction technology explained below combine to remove noise effectively.
This has enabled the EOS-1D Mark III to be the first EOS DSLR camera to have ISO 3200 as part of its standard ISO speed range. Dynamic range at low ISO speeds is about the same with the 1D Mark III as it is with the EOS-1D Mark II and Mark II N despite the 1D Mark III’s increased resolution.
So I'm wondering what many others may be as well, and that's where does this lead us? Does that mean that assuming Canon maintains a Bayer sensor, is the optimal photodiode size 7.2, or the 5ds 8.2, or even the 20d/30d's 6.4 or the XTi's 5.7? Realistically, the XTi seems like a dead end, and Canon simply crammed more pixels in to be competitive and I'm fine ignoring it as not relevant to this discussion.
Everyone loves the low noise shots from the 5d at 8.2 (better than the 1Ds2's performance at 7.2), and now Canon says " Dynamic range at low ISO speeds is about the
same with the 1D Mark III as it is with the EOS-1D Mark II and Mark II N despite the
1D Mark III’s increased resolution". Them are weasel words.
So what's going on? Where is FF headed? Are we headed for a 1Ds with 20% more resolution (16.7x.1.20= 20MP) or is the change going to be smaller? Where will the 5d go? Will it too see a 20% increase, or since it shared the same density as the 1D2n, will it see a larger increase with smaller pixels/diodes ala the 30d (I don't think so)?
Since this is my primary interest, any constructive thoughts would be appreciated. Yes, I know we're just guessing, but.....
Edited by jamesf99 on Feb 26, 2007 at 10:50 AM GMT
BeeMan458 wrote:
Canon states in their white paper somewhere that there's supposed to be an extra stop in the highlight shoulder of the 1D3.
It was explained on another blog, that (as you probably already know) DR is a function of the noise floor (db) as opposed to bit depth.
Nope, didn't know that about the "noise floor" so thanks for pointing it out. I was one of the slow ones that thought DR could be increased by increased bit depth. As I think about it, I guess it was rather careless of me to think that. If it doesn't mean you get "blacker blacks" or "whiter whites" then the range isn't really increased, it's just the discernment of individual values that becomes easier, but.....
OK - back to this. We already get the blackest blacks and whitest whites, so why isn't this DR? DR must be a function of what can be discerned, what can be captured and held as a distinct value on the B-to-W continuum. The reason we have high contrast in slide film is because the range isn't that great; the subtle value changes are sometimes lost in place of the larger jumps. If you're now getting those more subtle changes, isn't that capturing a more complete or increased range?
Can you point out/direct me to the other thread?
The rumor mill has Canon cranking out a 22MP sensor for the 1Ds3. I don't know if Canon is on board with this number or not. Time will tell.
yeah, I've been hearing that for over a year. There was talk about it before PMA 06 and what did we get? We got the 20d Mk1.5, sometimes called the 30d.
Me? I don't know nothing as I'm just a parrot of what I read and even then I sometimes slur my lines.
jamesf99 wrote:
OK - back to this. We already get the blackest blacks and whitest whites, so why isn't this DR? DR must be a function of what can be discerned, what can be captured and held as a distinct value on the B-to-W continuum. The reason we have high contrast in slide film is because the range isn't that great; the subtle value changes are sometimes lost in place of the larger jumps. If you're now getting those more subtle changes, isn't that capturing a more complete or increased range?
Uh, not really. But the increased bit depth can increase usable dynamic range.
At the dark end of the scale, you're limited first by quantum physics and second by the size of the steps in the A/D converter. If I understand this correctly, even at 14 bits the A/D steps are larger than the quantum uncertainty. So adding bit depth means smaller A/D steps, which means more shadow detail and less posterization. Unfortunately it doesn't mean less noise. But the smaller steps could make the noise less visible.
I think you're on the right track for DR, but not following through in the thinking. DR, as I understand it, is how much the camera can capture before the shadows come out as completely black and the highlights clip. If more is recorded then you have a greater dynamic range. When the 5D first came out there were a lot of people, myself included, who were pleasantly surprised at how much shadow detail was captured, whereas on my 20D I would probably get more noise and fewer details in the shadows on the same shot captured with the same settings.
So, dynaic range is more about comparing the actual scene to what is captured.
The bit depth will help reducing the posterization. But DR is basically S/N of the sensor.
By the way, DR should never be a issue with the current Canon cameras. People are usually complaining about the blown highlights. If the BHL is centered around the light source ( like Sun or light bulb or direct reflections ). Forget it. No DR can handle that.
When the large area is blown out - like Clouds, or white papers or cloth, it is a exposure problem.
i don't know what we're in store for with the 1Ds3 but i'd much rather see a 16 or 18 megapixel sensor with 16-bit capture and 12 stops of DR than i would a 22 or 24 megapixel sensor with 14-bit capture and 10 stops of DR. hopefully canon puts quality over quantity and realizes that more is not always better.
i don't know what we're in store for with the 1Ds3 but i'd much rather see a 16 or 18 megapixel sensor with 16-bit capture and 12 stops of DR than i would a 22 or 24 megapixel sensor with 14-bit capture and 10 stops of DR. hopefully canon puts quality over quantity and realizes that more is not always better.
I totally agree with this statement. Unfortunately, the megapix is only stat that many use to compare cameras. Personally, I don't care how many pixels the camera has. I want a larger sensor that can produce more dynamic range, which I hope is the way Canon is also thinking.
