Tom_W wrote:
Once again, the image circle projected by an EF-S lens is not large enough to accomodate in-body IS since the sensor will move out of the area covered by the lens. The same is true with a full-frame camera with EF lenses. There are only three viable alternatives - shrink the sensor size enough to allow the sensor movement to stay within the projected image circle, revamp the entire lens lineup to enlarge the projected image circle, or stick with the more effective lens-based IS system.
Fine - provide in-body IS in the cropped sensor bodies but only when they are connected to EF lenses. That solves the problem without further reducing the sensor size and thereby creates a fourth viable alternative - I'm sure some creative thinking by someone more capable than me can create even more alternatives than your three.
I'm not crying for in-body IS --- but it isn't some horrible monster that must be avoided at all costs! I do mostly telephoto work and in-body would not benefit me much if at all. But that doesn't mean I don't want to see it explored in the Canon line.
Well, if Dan Martin claims he knows for sure the 18-55IS is in development I will take his word for it. It does make sense in my opinion... although I do have to agree that in-body-IS seems like a better way to go against Sony. I can already see websites like CNet giving Sony's Alpha2 a higher score than the RebelXTi2 because of in-body-IS.
Here's another interesting forum in FM that I've been reading... it's under General Gear Talk... perhaps DSLR cameras as a whole will be obsolete in 9 years? Checkout the link:
Stunnaz wrote:
The market wants in-body IS. If you read the user reviews on the Alpha, a majority will mention in-body IS as one of the main reasons they chose the Alpha.
Of course that's how people who purchased the Alpha would answer the question. The key to remember is that a majority people purchasing DSLRs value this feature so little that they buy products without the feature.
rebelxtnewbie wrote:
18-55 IS? Why would they put IS is such a cheap quality lens. I would rather see them make a 17-85 IS MK II with beter optics and sell that as the kit.
Why put IS in a cheap quality lens?
Ask yourself why people buy the alpha to use their body IS with cheap quality lenses! It's not like the alpha owners are spending the money on any of the high end KM glass that is out there. The alpha is being used with cheap lenses for the most part.
JSeaman wrote:
Fine - provide in-body IS in the cropped sensor bodies but only when they are connected to EF lenses. That solves the problem without further reducing the sensor size and thereby creates a fourth viable alternative - I'm sure some creative thinking by someone more capable than me can create even more alternatives than your three.
Well, there's always vignetting as an alternative as well.
Seriously, it would seem futile to me for Canon to put much effort at all into creating a system of image stabilization that is inferior to the one already presented.
I'm not crying for in-body IS --- but it isn't some horrible monster that must be avoided at all costs! I do mostly telephoto work and in-body would not benefit me much if at all. But that doesn't mean I don't want to see it explored in the Canon line.
Perhaps you're not crying for it, but some on this thread are screaming for it as if it were the second coming of Jesus! And many of them seem to ignore the basic optical barriers to the technology (which is why I pointed them out).
Canon could come out with something that only works with EF lenses and cropped sensors, but that would reduce the value of some desirable lenses like the 10-22 EF-S. Alternatively, they could develop an electronic version of sensor-based IS with no physical movement at all (which might be worthwhile in the long run), but even that will still have to deal with the boundaries of the image circle as projected by current lenses.
Ask yourself why people buy the alpha to use their body IS with cheap quality lenses! It's not like the alpha owners are spending the money on any of the high end KM glass that is out there. The alpha is being used with cheap lenses for the most part.
I believe that people buy the Alpha because of the mistaken belief that in-body IS is a valuable substitute for horrible high-ISO performance.
Really, I think that putting IS into a cheap lens is a good idea. The 400D class of camera is, in general, an entry level camera. Most that buy it will probably not move to the next price tier, and will probably never buy a lens with a fast aperture. Folks here are the exception for the most part - enthusiasts & pros with a bit higher photography budget than what is typical.
Here's an interesting article that came out when the Alpha was first introduced (almost a year ago)... I hate it when they mention how much more profitable SLR cameras are compared to point-and shoots, it just reminds me how much I overpaid for my SLR gear. The article even said that Canon can still squeeze out a profit if prices of SLRs fall by as much as 40%!
rebelxtnewbie wrote:
18-55 IS? Why would they put IS is such a cheap quality lens. I would rather see them make a 17-85 IS MK II with beter optics and sell that as the kit.
Three reasons: €, $ and ¥.
For the 55-250, I'll make three predictions:
- It'll be PlasticFantastic™ like the current 55-200.
- 58mm filter diameter.
- 5,6 max aperture at 250mm.
Tom_W wrote:
I believe that people buy the Alpha because of the mistaken belief that in-body IS is a valuable substitute for horrible high-ISO performance.
Now *that* brings up a good point...It's my opinion that high ISO performance is Canon's main reason for current market share. It seems to be the *main* reaon cited by people who have dumped another brand and switched to Canon...
What IF body IS negatively affects high ISO performance on the CMOS sensors?
mirages wrote:
If you beleive that Canon does not have people working today on In-Body IS you are mistaken, and not really nearly as knowledgeable as you perceive yourself to be.
Actually, I believe they've aready *worked* on it...as in past tense, done, and over...but I don't have any inside knowledge
Tom_W wrote:
I believe that people buy the Alpha because of the mistaken belief that in-body IS is a valuable substitute for horrible high-ISO performance.
Tom_W wrote:
Well, there's always vignetting as an alternative as well.
Seriously, it would seem futile to me for Canon to put much effort at all into creating a system of image stabilization that is inferior to the one already presented.
Perhaps you're not crying for it, but some on this thread are screaming for it as if it were the second coming of Jesus! And many of them seem to ignore the basic optical barriers to the technology (which is why I pointed them out).
Canon could come out with something that only works with EF lenses and cropped sensors, but that would reduce the value of some desirable lenses like the 10-22 EF-S. Alternatively, they could develop an electronic version of sensor-based IS with no physical movement at all (which might be worthwhile in the long run, but even that will still have to deal with the boundaries of the image circle as projected by current lenses. ...Show more →
Tom_W wrote: Alternatively, they could develop an electronic version of sensor-based IS with no physical movement at all (which might be worthwhile in the long run)...
Your quoted statement directly above was EXACTLY what I was alluding to when I stated that Canon was doing research as we speak about implementation of In-Body IS in the forseeable future. To believe otherwise is pure marketing ignorance.
And, don't believe that it is not in the realm of possibility that they can overcome the image circle boundaries electronically as well. All it takes is innovation and ingenuity, which is how DSLR came about to begin with. I assure you that Canon is not limiting its development by anyones conceptual limitations here.
Canon has the deepest pockets in this arena, so no one should doubt for a minute that they are not using their formidable resources in R&D in this and other critical marketing areas. Having been there I can tell you with absolute assurance that they are, and will absolutely continue to do so. After all, we are not talking watermelon conversation, but about a LOT of money at stake here .
mirages wrote: Alternatively, they could develop an electronic version of sensor-based IS with no physical movement at all (which might be worthwhile in the long run)...
Your statement above was EXACTLY what I was alluding to when I stated that Canon was doing research as we speak about implementation of In-Body IS in the forseeable future. To believe otherwise is pure marketing ignorance.
And, don't believe that it is not in the realm of possiblity that they can overcome the image circle boundaries electronically as well. All it takes is innovation and ingenuity, which is how IS came about to begin with. I assure you that Canon is not limiting its development by anyones conceptual limitations here.
Sure it is a possiblity that Canon can overcome image circle boundaries electronically... but then you're talking about interpolating light that did not hit the sensor for the entire exposure as well as omiting light that was captured but was not wanted.
Possible, Yes. A good idea?
Digital photography is already hampered by the need to nail exposure. Too dark and we've got noise, too bright and we lose our highlights, too large of an aperture and we've got light falloff at the corners because the sensors do not catch light the same way as film does. Now you and others want to pile on to this by manipulating what light is received at the edge of the sensor based on measured camera movement?
That is a pretty insane thing to do considering that Canon, Nikon and Sigma have already solved the original problem with the use of lens based IS/VR/OS.
I understand that many people believe that Canon and Nikon both charge way too much for their lenses that both have high quality optics and IS. Please look at what Sony and Pentax are offering to go with their IS. Pentax has some great limited prime lenses, but almost all of them are totally unavailable right now. Sony has some great KM lenses available, but they will cost you more than a Canon or Nikon IS lens.
I admit that you are right... they can invest their hard earned money into body based IS. But in reality all they are doing is solving a problem they whipped a few years ago with the 75-300 IS.
nads wrote:
Sure it is a possiblity that Canon can overcome image circle boundaries electronically... but then you're talking about interpolating light that did not hit the sensor for the entire exposure as well as omiting light that was captured but was not wanted.
Possible, Yes. A good idea?
Digital photography is already hampered by the need to nail exposure. Too dark and we've got noise, too bright and we lose our highlights, too large of an aperture and we've got light falloff at the corners because the sensors do not catch light the same way as film does. Now you and others want to pile on to this by manipulating what light is received at the edge of the sensor based on measured camera movement?
That is a pretty insane thing to do considering that Canon, Nikon and Sigma have already solved the original problem with the use of lens based IS/VR/OS.
I understand that many people believe that Canon and Nikon both charge way too much for their lenses that both have high quality optics and IS. Please look at what Sony and Pentax are offering to go with their IS. Pentax has some great limited prime lenses, but almost all of them are totally unavailable right now. Sony has some great KM lenses available, but they will cost you more than a Canon or Nikon IS lens.
I admit that you are right... they can invest their hard earned money into body based IS. But in reality all they are doing is solving a problem they whipped a few years ago with the 75-300 IS. ...Show more →
nads wrote:
I admit that you are right... they can invest their hard earned money into body based IS. But in reality all they are doing is solving a problem they whipped a few years ago with the 75-300 IS.
IF, as you say, the problem was whipped a few years ago, they wouldn't be spending ANY of their hard earned money now, would they?
And, since we have both come to the conclusion that this is not the case, I do not understand the logic of your statement .
The majority of camera shoppers don't research at all. They're not like us. They're not hanging out on camera forums. They're not even reading reviews.
When I look around me, I see the opposite: people doing an extreme dosis of research before spending $800 on a dSLR. And if they weren't doing any research, chances are they'd buy a Canon or Nikon if only because they are the most well-known camera brands.
What I would like to know is the percentage of Sony A100 buyers that were KM users in the past and want to be able to use their old lenses. I know a couple who bought a Sony for that reason only (and weren't interested at all in a BIS system).
mirages wrote: nads wrote:
I admit that you are right... they can invest their hard earned money into body based IS. But in reality all they are doing is solving a problem they whipped a few years ago with the 75-300 IS.
IF, as you say, the problem was whipped a few years ago, they wouldn't be spending ANY of their hard earned money now, would they?
And, since we have both come to the conclusion that this is not the case, I do not understand the logic of your statement .
Regardless, 'Nuff Said.
Edited by mirages on Apr 21, 2007 at 01:21 PM GMT...Show more →
That was quite a leap of faith you took with my statement. My statement is that Canon CAN invest their money that way. Also, if such a hypothetical world existed where they chose to invest money in BIS, that money would simply be wasted in an effort to solve a problem that was previously solved with the 75-300 IS lens.
Let me be perfectly clear. I do not believe that Canon is working on IS in the body. I believe this because I feel the technology has severe drawbacks:
* It would destroy the market for a lenses like the 17-55IS, 70-200 f/4 IS, 70-200 2.8 IS and even 24-105 IS.
* It will not work with FF unless all lenses are redesigned with larger image circles
* It will not work with ef-s for the same reason
* It is not as effective as Lens IS at long focal lengths.
* It will deteriorate image when using most wide angle lenses
* An Xti + 18-55IS is indistinguishable from an alpha + 18-70 in the eyes of uneducated consumers who only learned that they want IS from P&S cameras.
* The competition is not truly forcing their hand... the 2 stops given by the alpha's IS are lost in the alpha's ISO performance.
* Body Based IS is a DOOMED TECHNOLOGY.
Seriously ask yourself why Canon would knowingly bring to market the 17-55 ef-s IS knowing that they would put out a body with IS within a year or two? This is a lens that can only be used on entry level products in the first place. Who out there would buy this over a Tamron 17-50 if they owned a Canon body with built in IS?
Canon has already "shown it's colors". They did so when they released an ef-s IS lens and they did so when they released an ad campaign meant to turn consumer opinion against this technology which they do not intend to bring to market.
Canon has shown it's colors by pursuing affordable FF. We know now that Full Frame and sensor-IS are mutually exclusive. We also know that sensor resolution is going to continue to improve, eventually to the point that crop sensors out resolve most lenses. At that point the only way to improve resolution given the same pixel density will be to use a larger image circle. Crop sensors are already dramatically hampered by diffraction, and this will only get worse as sensor resolution increases.
Now you can imagine a mythological electronic based version of IS all you want. But the reality is that the only way that will happen is through the advancement of processing power and the improvement of sensor technology. Unfortunately for those dreaming about this vapor ware, that kind of improvement will also yield more affordable FF sensors. The very technology required to make this imaginative version of IS work is the same technology that will make it obsolete!.
Now this is mostly a list of real world obstacles that I believe must be overcome to make sensor-IS a longterm, mainstream reality in the world of DSLR's. As I see it, the set of tools currently available from Canon is more than enough to take excellent photographs in every conceivable situation. I do not think that is rational to undermine what has been proven to work simply to give us this new technology that is in no way MEASURABLE BETTER than lens-based is.
Canon's long term goals have been based on the use of FF and lens-based IS for a long time now. Turning that strategy on it's head in order to serve the few vocal proponents of an inferior technology is not the way for a company to serve the people that really matter - it's shareholders.
Please, if you really hate that Canon is charging you so much for IS built into each lens, feel free to move to Sony. With Sony you will only lose two stops of usable ISO, but you will make those two stops back up with the use of their IS. (That is of course unless you are currently using a canon lens with IS and your subject isn't moving too fast for a shutter speed 1/4th of what you were going to use with the Canon body).
nads wrote:
Seriously ask yourself why Canon would knowingly bring to market the 17-55 ef-s IS knowing that they would put out a body with IS within a year or two? This is a lens that can only be used on entry level products in the first place. Who out there would buy this over a Tamron 17-50 if they owned a Canon body with built in IS?
Good question. Why is Canon wasting time with these new EF-S lenses that offer L quality optics (in this example possibly better), at close to L prices, but are incompatible with their entire body range, and are of substandard built quality?
nads wrote:
Digital photography is already hampered by the need to nail exposure. Too dark and we've got noise, too bright and we lose our highlights, too large of an aperture and we've got light falloff at the corners because the sensors do not catch light the same way as film does. Now you and others want to pile on to this by manipulating what light is received at the edge of the sensor based on measured camera movement?
Don't you think that it won't eventually be possible for the circuitry to vary iso
sensitivity in real time across individual elements of the sensor according to
maintain exposure within algorithmically calculated bounds ?
nads wrote:
.....
Now this is mostly a list of real world obstacles that I believe must be overcome to make sensor-IS a longterm, mainstream reality in the world of DSLR's. As I see it, the set of tools currently available from Canon is more than enough to take excellent photographs in every conceivable situation. I do not think that is rational to undermine what has been proven to work simply to give us this new technology that is in no way MEASURABLE BETTER than lens-based is.
......
Your entire post sums it up nicely - there is no rational technical reason for Canon to follow Sony's scheme when Canon offers something better already.
.