I am curious why that CMOS sensor in the high end Canon camera body has excellent low black noise compared to the images from a Fuji Pro S3/S5 with its Super CCD? I have used Fuji Pros since 2004 and have discovered that my black details in low light situations include a lot of random color noise which I do not see when reviewing images from the 1DS Mark2 body. What does the Canon body do to eliminate the random color noise in low light black detail? Is it the CMOS Sensor? Is it the body processing?
Separately, the white detail on my Fuji Pro S3 has such expanded dynamic range that I can shoot higher exposures and then lower the exposure back to normal in software and it appears to lower or mask the noise floor in the same way that Dolby Noise Reduction works for audio.
I've been a 35 mm fan for years, and have a significant investment including multiple lenses, macro, ring lights etc. For trips etc the smaller cameras seemed fine--always had small 35 mm, for instance. However, re the digital small cameras, the amount of compression is really disappointing.
Is my only choice to go back to a body/lens SLR if I want wide angle and telephoto capability AND good resolution?
No. But if you want wide/tele and really good high iso performance a SLR is where you need to be. PS camera have alway been a compromise. Small 35mm film cameras had focus issues, and lens issues too. Plus most people use iso 200-400 film which has reasonable quality even in a PS digital camera. The biggest difference is now we're blowing up the picture to 1:1 on our monitors and can see the quality defects easily. I bet if you printed your old compact photos at something like 16x20 you'd probably be unhappy with them too. That's the sort of scale we're looking at on our monitors zoomed in.
Good article, but one glaring mistake: you claimed Sony was the first to make a 12+Mpixel CMOS sensor. I think you meant the first AFTER Canon - the 5D, with its 12.8Mpixel sensor, was on the market in 2005, and the 1Ds Mark II, with a 16Mp sensor, was on the market earlier than that.
On a different front, I thought the Nikon D3 was using a Sony sensor, not a Nikon?
The 5D and 1Ds II and III are full-frame sensors. The Sony was the first consumer (read affordable) APS-C sensor. Canon now has their own 12 megapixel consumer sensor in the XSi, which we are now reviewing.
Nikon did their own designs for the D3 sensor, but they do not, to our knowledge, have the capabilities to manufacturer that sensor. Sony has manufactured sensors for them in the past and present with the D300, D60, D80, and others.
Since the new full-frame is CMOS it is likely manufactured by Sony, or possibley Samsung. Sony and Samsung (who make the Pentax 14.6 megapixel sensor) jointly own several patents on CMOS manufcaturing technology.
That should read the read "First consumer 12 megapixel APS-C sensor". As mentioned several times in the article Canon pioneered CMOS technology in a consumer DSLR with the Digital Rebel.
This article complete skips of fuji's fantastic super CCD technology. Its not really a traditional Bayar array, nor a foveon. True they don't produce a interchangeable lens SLR but they do make prosumer SLR's with their sensor. Also the very notable fuji f30 series cameras were made with this sensor. It was a true triumph in PS camera high iso usability.
Hell, then there is no excuse for it not to be included in this article. Especially when it quite clearly states there are no other sensor options at one point.
There will always be fans of one technology or another who feel slighted. I apologize for that, but Fuji is still basically a Bayer sensor with a differnt pixel shape. I have added the following to the Bayer vs. Foveon page to make Fuji fans a bit happier:
"FujiFilm produces one current DSLR with a variant of Bayer technology. It is called the Fuji S5 Pro and is basically a Nikon D200 body with a Fuji Super CCD sensor. The Fuji S5 Pro uses the Nikon lens mount. The Super CCD still uses red, blue and geen pixels in the same standard Bayer ratios. However, the shape of the pixel is hexagonal rather than the squate or rectangular pixels in other Bayer arrays. In the latest version Fuji also added smaller photosites between the normal pixels to gather "dynamic range" data.
Fuji has updated the camera body from the S3 to the S5 in the past year, but the sensor has not been updated for more than 3 years. The current Super CCD is still a 6.3 megapixel sensor, but Fuji specifes it as a 12.3 megapixel due to the addition fo the tiny "brightness" pixels. Tests indicate the true resolution is more comparable to an 8 to 10 megapixel sensor from competitors. The Fuji sensor is still basically a Bayer sensor with a different shape for pixels."
We can probably now all argue whether the Fuji Super CCD is really a Bayer variant or not. It certainly appears that way to me, and as a CCD instead of a CMOS sensor it is need of a serious update if it is to continue as a player in the DSLR market.
I'm happy. I'll won't argue that it's not a variant, but its design is different enough to warrant a mention. Even for such an old sensor, it still produces a dynamic range that exceeds notables such as the 5d, etc. I know I love my Super CCD based f30.
1> The sensor size discussion was rather muddled, with half of the information on p.4 duplicating the info on p.2
2> Yes, full-frame is really nice to aspire to, but the problem of full-frame sensor cost was not mentioned. IC lithographic machines ("steppers") cannot image a FF sensor in a single pass, making the production time cost 2-3x greater than smaller sensors (APS-H is the biggest that can be done as single pass). Worse, the large size of the sensor compared to the wafer means that yields drop appreciably as the sensor size increases. All in all, FF sensors can be 10-20x more expensive than APS-C sensors, let alone 4/3, 1/1.8", etc.
To quote Canon (2nd ref): "Regardless of future technological developments, cameras with full-frame sensors will always cost much more than cameras with smaller sensors."
steppers used to not be able to make 35 mm sensors in one pass. as canon and nikon are two of the bigger stepper manufacturers, guess what has likely happened? and both of them are smart enough to not announce that they've made the larger stepper, so as to make more profit than the other.
Page 2 emphasized the huge size difference between compact point-and-shoot-sensors and DSLR sensors as classes of sensors. Page 3 then explored how all the DSLR sensor sizes evolved within that class, and page 4 detailed the current variations in sensor size within the DSLR class. I appreciate your comments, but can you provide some specifics on what needs to be changed?
The full-frame sensor does indeed cost more, which was mentioned a number of times in the article. However, prices are still dropping for the full-frame sensor as evidenced by the fact that the street price of the Canon 5D is now around $2000.
In the above Table we can see the range of sensor sizes now available. However, the majority of Compact or Point and Shoot cameras today still generally use a 1/2.3" to 1/2.5" sensor. A few top-of-the-line compact cameras, like the Canon G9, feature a 1/1.8" to 1/1.7" sensor; and the very best compact cameras have sensors around the 1/3" to 1/2" range. In contrast, the sensor sizes for today's DSLR cameras are in the range of 4/3" and APS C. A few top pro cameras now sport 35mm-size sensors and are referred to as full frame.
Page 3
Why All these Different Sensor Sizes? 35mm first appeared on the scene in the 1930s, using an image frame of around 24x16mm. 35mm still film simply took 35mm motion picture film and spooled it into a light-tight canister, turned the spool direction by 90 degrees, and used double the frame size. (In fact, some early 35mm still cameras were referred to as "double-frame" cameras.) By the 1960s, with point-and-shoot and developments in SLR technology, 35mm had become king of the film formats. Even as film manufacturers tried to introduce other film formats, 35mm continued to grow and prosper.
...
... Actually, the image size that was always shot was the 30.2mm x 16.7mm, and the other sizes were just standard crops. The 24mm by 16mm APS C format is approximately the same size as the 1/2 frame 35mm championed by Olympus in the film era.
...
More recently, Canon has championed the full-frame sensor in their pro cameras and in their pro/amateur 5D model. Larger sensor manufacturing cost has dropped as digital sensors have evolved and it now appears likely the APS-H (1.3X lens multiplier) will eventually drop from the Canon line. Since no lenses depend on that image circle all current Canon full-frame lenses, such as those used with the APS-H cameras, will remain usable on the full-frame sensor pro models that will replace them.
The developing push for full-frame at the top of the current DSLR market is a move to a sensor that is a bit more than double the size of today's APS C sensors. And there is a significant problem with this -- the full-frame sensors are made using different process techniques to even APS H sized sensors, greatly increasing their relative costs. In short, the IC lithographic machines ("steppers") that make the sensors cannot image a full frame sensor in a single pass, making the production time cost 2-3x greater than smaller sensors (APS-H is the biggest that can be done as single pass). Worse, the large size of the sensor compared to the wafer means that yields drop appreciably as the sensor size increases. According to Canon USA, full frame sensors can be 10-20x more expensive than APS-C sensors.
It is not surprising then that dSLRs mostly use APS C and 4/3 sensors.
Nikon, Pentax, and Minolta all had significant success in the 35mm film market, so all three had a vested interest in preserving their 35mm lens mount and keeping their current 35mm system users happy. However, none of these three had the resources to develop and manufacture their own sensors, so they partnered with sensor manufacturers to produce digital SLR cameras. Of course, these manufacturers were mostly making APS C, 4/3, and smaller sensors. In recent years that partner has been Sony, so all three manufacturers have basically adopted a 23.6x15.7mm sensor with a 1.5X lens multiplier.
...
I am not an expert but it seems to me the most important point has been missed in this article: depth of field. Larger sensors allow for greater depth of field with the same lens. There is a lot of discussion of "image quality" (resolution, shutter speed, sensitivity) but depth of field alters what the image is, which is more important.
In the interest of keeping the article somewhat basic and a digestible size we did not include a discussion of the depth of field variations with compact sensors compared to DSLR sensors.
This IS an important concern for creative use of depth of field since the DOF is much greater with the smaller sensor than with the DSLR sensor. Not everyone would agree that a shallower depth of field on a DSLR is more desirable. It certainly might be in a portrait, but not in a sweeping landscape shot. The desired DOF depends on what you are shooting.
I do agree with you there that it is desirable to have deeper depth of field (particularly on macro) but I think the difference is that on an SLR you have the choice - you can shoot wide open and have shallow depth of field or you can stop down (and normally have to use up the ISO advantage) to get depth of field. On a camera with a small sensor you are generally stuck with the deep depth of field whether you want it or not.
I hope nobody is using F/1.2 when shooting a portrait. Unless you want only a nose to be in focus. With something THAT shallow, you either have to REALLY step back and use distance to your advantage to get what you want in Focus. Professional Wedding photographers will not TOUCH anything above f/5.6 as you don't want to get ONLY the bride in focus while the groom is OOF.
People, remember there are multiple factors that affect DOF. Aperture is definitely one of them, distance, and Focal Length.
Obviously you want something very shallow with great bokeh, then you would get close to your subject, open up your aperture, and have a telephoto focal length.
The thing is, on a FF sensor, you "see more" on the same focal length as a cropped sensor. Well put wesley, a 50mm is a 50mm. For those of us with cropped sensors, slap a 50mm on your camera. (or zoom to 50mm). Look at your subject with your naked eye, then look through the view finder. SAME SIZE AND SHAPE! Nothing gets larger! Supposedly a telephoto range (greater than normal) should have some zoom effets shouldn't it? I mean APS-C sensors are either 1.5 or 1.6 smaller than a full frame, that means the normal focal length should be around 27mm? So 50mm should be 75mm for nikons and 80mm for canons! That means you should see some zoom in 50mm! but nope! you don't! in fact, the only time you'll see the zoom is AFTER you fire the shutter and it displays the image to fill that 2.5" or 3.0" inch lcd you have! (or even 2.0") So where does the FF excel? you place 2 cameras exactly side by side of each other, same focal length, same focus point. Same lens. Through the Viewfinder, you'll see the nice Bokeh of your subject on BOTH cameras, exactly the same. Difference? the FF has more frame coverage! Thus, having better "DOF" (or lack thereof) because of the more frame coverage! On this note, let's all by Hasselblad's Medium FF and get some nice digital backs! ;)
Hmm ... In my understanding the difference in FF to APS-C is CROP.
A 50mm lens on APS-C has the same DOF, same perspective characteristic as on a FF camera, but the sensor is smaller, so you basicaly get a CROP of the original FF image. There is only a perceived zoom effect as the image is a smaller area.
For Nikon you get 1.5x smaller image area, but everything else is the same as on FF.
It has to do with distance from the subject as well. Perspective is the ratio of the distance to the subject to the distance to the background.
So, to get the same size on sensor, with the same lens length, you must either move closer, or further, changing the perspective, and the depth of field.
DOF is also a function of the lens. Some ultra-serious photographers are willing to pay $2500 for a Canon 85mm 1.2L over a 85mm 1.8 (which is just as sharp and which only retails for $300), because the 1.2L has a softer, more pleasing DOF effect.
Looking to buy a camera pretty soon here.
What do you folks think about the G9?
Is there a particular brand I should favor?
Also, are there any cameras available that do "tricky" stuff, like Infrared, Macro, or that neat effect when cities look miniature? (Can't find an example)
I read an article about a guy a few years back that would setup a scene, then he would walk around selectively lighting certain aspects in trippy ways to make a final composite image.
Anyone know what this is called, or where I can learn more about it?
Is "fisheye" what you are looking for for the miniature cities?
and the selective lighting is often called "Painting with light"
plenty of compact cameras can do OK macro if you can get really close to your subject, but as this is often hard for reasons of lighting and/or scaring off the subject, an SLR with a real macro lens is still beneficial.
When everybody was still on 35mm film, it was impossible to tell whether an SLR shot an image or a good point and shoot/rangefinder style camera. Image noise was a function of the film, and within a certain range, SLRs and point and shoots could reproduce the same sharpness and depth of field. All of that is pretty much impossible these days with the split in sensor sizes.
My biggest question is: if camera companies could make a pocketable 35mm/APS sized camera 10 years ago, why can't they make the same pocketable size with a digital sensor? An APS-C sized sensor in a Canon G9-style body shouldn't be impossible, and while it would be expensive, it would have image quality worth the price.
I have a feeling that the companies producing SLRs are just trying to protect their high-margin lens business - a company that's not vested in the SLR side of things could carve out a very solid niche by making a point and shoot with an SLR sensor.
And yes, I know the Sigma DP1 exists, but without a zoom lens and max aperture of F4.0, $800 is a lot to ask...
A lot of it has to do with consumers expecting zoom lenses with cameras. The larger the sensor the larger the lens has to be, especially with a zoom lens. I don't think it's possible to have a decent zoom lens on a pocketable camera with a large sensor (with current technology). This is why the DP1 has these limitations. This is also why a camera such as the Sony R1 has such a large lens compared to other cameras with smaller sensors.
However, I'd be quite happy with a fast prime lens on a well-designed, pocketable, large-sensor camera, even if it did have a bunch of downsides to it.
If the DP1 was more refined and had a much faster lens, it would be an awesome camera. Unfortunately, it's rather crude compared to cameras from other leading brands.
"My biggest question is: if camera companies could make a pocketable 35mm/APS sized camera 10 years ago, why can't they make the same pocketable size with a digital sensor? An APS-C sized sensor in a Canon G9-style body shouldn't be impossible, and while it would be expensive, it would have image quality worth the price.
I have a feeling that the companies producing SLRs are just trying to protect their high-margin lens business - a company that's not vested in the SLR side of things could carve out a very solid niche by making a point and shoot with an SLR sensor. "
I don't think that's the case at all, film and digital cameras have to be a lot more different than you perhaps appreciate. The Sigma DP1 has had a prolonged and difficult development, while I think that's partially because Sigma are less experienced it's not entirely the story. One of the main problems I believe is the way light hits the sensor, with a digital sensor to work optimally you need to have the light hitting the sensor as straight as possible. Film was more flexible in this sense which made it easer to produce small cameras using comparatively large pieces of film. I believe this was one of the difficulties in producing the DP1 as the lens is very close to the sensor which meant the light is hitting the sensor at very high angles. This is easier on an SLR as the sensor is a reasonable distance back from the rear of the lens (Olympus particularly have made a goal of telecentric lenses) although even the 35mm full frame cameras can suffer from poor corner sharpness.
Also with film you didn't have any low pass or IR filter in front of the film itself which is normally part of a digital camera, I believe the DP1 (as well as the Leica M8) don't actually have an AA filter at all which causes its own problems.
Hopefully other companies will have a go seeing the interest that the DP1 is generating, I think Panasonic who produce a 4/3 sensor and are having little success in the SLR market are in a prime condition to do so. However it's a more difficult task than it seems.
"I have a feeling that the companies producing SLRs are just trying to protect their high-margin lens business "
when Kodak got out of doing DSLR bodies (even if they were basically Nikons with a different sensor) they said it was because the profits weren't in the bodies but in the lenses - which of course they didn't make ...
Pentax is supposed to be getting out of compacts because they're becoming a commodity item & hence price-sensitive/low margin.
I could imagine someone like Leica doing what you want though as they seem able to command a significant premium.
Leica's currently tied to Panasonic sensors, and relatively speaking, Panasonic hasn't figured out how to make a sensor that has low noise above ISO 200 yet, regardless of size (otherwise I'd seriously contemplate the Lumix LX-2).
Also, while I'm willing to pay a premium for a nice fixed lens camera (I'll consider up to $700, maybe even $1000 for a camera with the right handling, image quality, and features), the market for such an expensive "small" camera is probably pretty small (unless professionals are all clamoring for a pocketable camera as a backup in the field).
Lastly, the Leica name generally commands an even healthier premium than I'm willing (or financially able) to consider...
The sensor is only half the story. The lens is the other half. This is NOT marketing hype. The difference between consumer kit lenses and pro series glass is easily discernible (even to non pixel peepers). Does this justify the 8X difference in price? I don't know.
However, I'm pretty sure that it's not possible to construct a high quality P&S with a great lens that is "pocketable," regardless of the sensor size. Good glass tends to be heavy. Even the old Leica and Contax rangefinders were not pocket cameras. They were compact, but I doubt anyone considers them pocket cameras (unless you're talking about cargo or jacket pockets, and you trust having a $6000 camera hanging out of your shirt pocket). When you add zoom functions to the equation (which is a must nowadays), you looking at some serious IQ trade-offs.
That said, I would be the first in line to purchase a good P&S camera that produces high quality professional grade images.
Honestly, I'll be happy if someone can fit an under 2" thick camera with an adequate (say, 28-112mm equivalence, f2.8 - f4.0) lens and a large/larger sensor. Maybe not even an APS-C sensor - a 4/3" design is already significantly larger than a 1/1.8" sized unit. Olympus, are you listening?
The Olympus E-420, even with its pancake 28mm equivalence lens, is too thick to put into a jacket pocket (or crammed into some looser fitting jeans).
I rarely ever take dedicated trips to take photos - so packing a backpack half full of photographic equipment to take a "nice" shot is something I'm loathe to do. But I do go cycling, hiking, and other normal touristy activities where I'd like to be able to bring a camera along...
There was the Sony R1 - APS-C sensor and 24-120 Zeiss-branded lens. Not exactly pocketable, even not having to deal with the registration distances required by an SLR with the swinging mirror. Maybe if you were willing to do something with folded optics you could fit a zoom in a smaller body, but who knows how much of an image compromise that would be.
My XT with 28 1.8 does fit in the cargo pocket of one pair of pants I own, but I would rather just have it in a small shoulder bag.
"The Foveon sensor falls between 4/4 and the Canon 1.6 in size and has a 1.7X lens multiplier."
I assume this should read '4/3'
Obvious question is what about Fuji? While I realise they re-use Nikon bodies and lenses, the discussion about Bater and Foveon makes no mention of Fuji's sensor. While it is more conventional than the Foveon, it's not just a bog standard bayer sensor and while Fuji are a minor player, so are Sigma.
For those who are not familiar with Fuji's design, they use two photodiodes at every point one larger and one smaller with the two combined to produce a single pixel in the output image. The idea is that these pair of pixels can capture more extended dynamic range than a standard bayer sensor of the same size. The latest version of this sensor is in the Fuji S5 Pro however it's limited to just 6MP although I can't remember if they still produce 12MP files from this. The S5 itself is basically a Fuji version of the very good Nikon D200 body although Nikon have of course moved on with the very impressive D300.
Yes, it should have read 4/3 and the reference is corrected. The Fuji S3 was one of my favorite Nikon bodies and the dynamic range was certainly impressive in the studio.
We did not mean to slight Fuji, but as one Nikon-mount body with a sensor that hasn't been updated in several years (the S3 and S5 sensors are the same as I understand it) we decided not to include the Fuji since there have been no sensor updates in quite a while.
"We did not mean to slight Fuji, but as one Nikon-mount body with a sensor that hasn't been updated in several years (the S3 and S5 sensors are the same as I understand it) we decided not to include the Fuji since there have been no sensor updates in quite a while. "
I can't say I really agree there, given the article more focuses on concept in parts I think the S5 sensor is still relevent as it's something slightly different to the bayer sensor. The article refers to only Bayer and Foveon which implies there is nothing else, I think for completeness even if you don't go into any detail it's still worth mentioning Fuji are doing something else.
Also the S3 and S5 sensor are not the same, while they have the same amount of pixels it appears there's been some slight improvements although clearly not much. To be far to Fuji though, the Foveon sensor hasn't really come on much either - it's gone up very slightly in resolution with some small other changes but that's it pretty much. If the Foveon sensor had been just mentioned in passing I could definitely understand leaving Fuji with a similar mention but generally Fuji and Sigma are considered in the same boat as doing something a bit different although arguably with the Nikon body and mount Fuji have had more success.
I never saw an advantage to their designs. I can't see what purpose having a smaller photo site on the sensor would do. It just has more noise, and less dynamic range than the larger sensor. I've read their papers on the subject, and they don't seem to have made a good case for it. Somehow, I think they understand that now.
"I never saw an advantage to their designs. I can't see what purpose having a smaller photo site on the sensor would do. It just has more noise, and less dynamic range than the larger sensor. I've read their papers on the subject, and they don't seem to have made a good case for it. Somehow, I think they understand that now."
Are you referring to Fuji? If so, your information is incorrect - the last measurement I saw put the S5's sensor at slightly more dynamic range than the 35mm sensor in the Nikon D3. Their real problem at the moment seems to be resolution as well as having the older D200 based body.
I think Sigma (and Foveon) would be better off if Sigma could license a major player's mount, like Kodak did with the SLR/n and SLR/c. There are plenty of people who would like to have the sensor for the situations where it excels, but have no interest in a whole Sigma SA mount setup.
I've been reading Anandtech for over 8 years now. I was a bit skeptical about your doing articles on digital cameras; however, this article put my reservations to rest. Very nicely done.
I see why it makes sense for Anandtech to write about digital cameras. The nexus between computer tech and camera tech are very obvious if you've been following the trends. I'm a professional portrait photographer. In the last few years, it has become virtually impossible to remain in business unless you have a firm grasp on the latest camera tech as well as computer tech. I spend as much time on my computer as I do behind the camera. The camera has become a computer accessory (or vice versa). Many old timers who are hanging on to film are slowly being forced out. I'm somewhat of an old timer myself because I learned about photography when it was cool to have a darkroom, but I embraced digital equipment early.
Your graphs make it very easy to explain to people why their P&S (piece of s***) cameras are simply not adequate for serious portraiture. The MP count is marketing crap. As you demonstrated, it's the physical size of the sensor that matters.
Looking forward to part II of the series. Also, do you have any solid info on the introduction of 5D Mk II?
Mostly. If you're willing to do some post processing yourself, the higher MP will enable you to decrease the IQ gap between the higher end DSLR and the ho-hum consumer camera. For this reason, since I wanted something compact and didn't really need a lot of optical zoom, I chose the Canon SD1100IS. 8MP, and while you begin to get noise at ISO400, more at ISO800, and tons at ISO1600, using a non-linear digital filter should correct most of that without blurring the image much (if at all).
I wish I did have definitive info about the 5D Mark II launch, but the best info I have is this fall at Photokina. Rumors pop up every month that the new 5D will be here in a few weeks - and the last rumor was a definite April 22, which is tomorrow. Rumors are just that - rumors.
the thing is, I've read the reviews and saw the comparisons between the 5D and the D200 (I believe that was the last CCD sensor that Nikon used). The 5D has way better per-pixel sharpness than the D200. Perhaps the technology of the CMOS made it so it finally "looks" better than the CCD. Or at least on the Canon side of things. I'm not biased towards any camera. I've own an XTi, 5D, Panasonic DMC-L1, and now a D300. I'm hoping to own the next 5D MKII (or whatever they call it).
Even the D2X used a CMOS sensor, and I thought Nikon made the D300 sensor, as well as the D3 sensor (FF). What I also find "funny" is that Sony doesn't have a FF sensor yet, but nikon does, thus leaving me to believe that Nikon came up with their own FF sensor.
Nice article Wesley, finaly one Anandtech photo related article I enjoyed reading.
On note on the growing megapixel count. Sooner more than later, DSLR in APS-C (and later full frame ones) will hit the same technology wall P&S are facing (too small photosite).
This is one factor that makes me a happy film shooter :-) I know that my limit is the scanner up to around 10MP and I have less flexibility in shooting conditions (either 2 bodies or limited by ISO and film type), in every other situation I am equal or better off.
I am waiting for an affordable Nikon full frame body and then I will make my switch to digital (but that is yet years to come).
A sensor technology improves APS-C will be able to go to high enough pixel counts and have very good quality. For most consumers some where around 12 MP is enough, if the sensors improve to where this can hit 3200 ISO with low noise 99 percent of the market would be satisfied.
The full frame bodies will replace the medium formats of the film world. There have been some really good medium format cameras that take much better images than 35mm but very few are sold. I think the article is correct in saying the full-frame DSLR will be the PRO cameras and priced as such.
The number ofpixels is directly related to the print size.
If you go by the oft quoted 300 dpi on the final print for maximum quality, you will need a sensor with 2400 x 3600 resolution for a full frame 8 x 12 print, or 8.64 MP. For an 8 x 10, it would be 7.2 MP (4/3 sensor).
For a larger 11 x 17, it would be 3300 x 5100, or 16.83 MP, or 13.86 for the 11 x 14 4/3 sensor print.
You can figure the rest of the sizes my multiplying the inch size of the print edges L x W by 300 to come up with a number.
But the truth is that 240 dpi is going to be good enough fot most prints, and youcan do the numbers that way.
When you know the numbers, you can figure out what size sensor you will need for the highest quality work.
But for most people, even 180 dpi will be enough for their prints. Going to that gives more flexibility, as the sensor pixel count is much smaller.
It's better to get a camera with a longer OPTICAL zoom rather than to go for the biggest number of sensor pixels.
It's all a matter of print size. I am an amateur/hobbyist, and anything past A4 is large for me. I have printed reasonable quality 8x12 prints from a consumer slide film and home film scanner.
I can hang them up on the wall in a nice frame and they will serve their purpose :-)
Idealy I'd need a 12-14MP full frame Nikon with goot ISO/Noise characteristics (f.e. D3) and it will satisfy my needs for years to come.
especially as I was wondering the other day what smaller process technology would do for digital sensors?
Normally of course smaller process means cooler & more importantly smaller>cheaper to make but of course with sensors you are talking a fixed size so not any cheaper.
But would the potential increased precision of circuitry mean any better image quality?
& a little nitpicking:
Sony didn't buy Konica Minolta (which continues in business) but certain assets from the Photo Imaging division of KM. As to what that actually means there is little hard knowledge outside the 2 companies as it seems that KM still retain some IP & indeed a shareholding in some of the production facilities.
Also, current thinking seems to be that the "A900" may not be called that but something else to differentiate it from the APS-C models.
The smaller process technology will have no positive effect on the sensors themselves, though it will for the associated electronics integrated on the die.
The same problems he mentioned about smaller sensing sites will remain. The smaller the sensor, the poorer the performance viz a viz larger sensors.
He did mention that the photo division was the purchase, not the entire company (unless he changed the article after your post).
I changed the wording on the Sony purchase a bit to better reflect that Sony bought the Minolta camera assets of KM and not the company. Thanks for pointing this out.
Canon's 1DS MK2 was a 16.7MP CMOS sensor also, and of course, the MK3 is a 21MP CMOS sensor...
Where i'm lost is if CCD's are so much better (in IQ) why dont' they stick with CCD's? why the move to CMOS besides the lower cost and the battery life that it saves?
Why are professional level DSLR's (From canon and Nikon) are both CMOS when we know that CCD is the way to go for better IQ?
It USED to be true that CCD's were better. Not so any longer. The best CMOS sensors are better than the CCD's they replace.
The desire to go CMOS is obvious to the manufacturers of the sensors.
CCD sensor technology is a completely different manufacturing process from that of CMOS, which the entire industry uses for everything else (almost).
Moving to that allows CMOS sensors to not only be able to integrate other electronics on the sensor chip, resulting in simplicity, price advantages, and the ability to more favorably utilize their process lines, but that higher quality you're concerned about.
Putting functions on the same chip improves the quality of the signals.
And, by the way, an error in the article: Canon was not the first to make, or use, a CMOS sensor. They were the first to come out with a high quality sensor. I believe that it was Vivitar that used the first one, though I forgot the name of the manufacturer.
Anther omission is that there are trilinear sensors used in camera backs such as the Betterlight scanning backs. So there are three different major technologies in use.
And not all of the negatives of the Foveon chip was mentioned.
No, CCD is definitely still the IQ king. CCD still has a far better SNR than any CMOS technology.
As a digital camera design engineer, I deal with image sensors every day. The major reasons why CMOS sensors are attractive are all cost related. Not only are CMOS sensors themselves cheaper, but they lend themselves to integration with other electronics better and they are MUCH nicer to design with. A typical CMOS sensor will require ground, 3.3V, and maybe some other standard voltages (1.8V, 2.5V, etc.), whereas a typical CCD will require ground and anything from 8 to 12 other DC voltage rails. For instance, I am working on a camera that requires -15V, -9V, -6V, -4V, -1.5V, ground, 2V, 3.5V, 11 V, 15V, 24.5V, and has a clock signal that must run up to 40V. While making the voltages and driving clocks at them is fairly straightforward, it requires a lot more components than a CMOS sensor design would. More components directly equals higher cost. Also, as the article mentions, more functions can be integrated onto CMOS sensors than CCDs.
The other advantage of CMOS vs. CCD is in random-access readout. If you want to read a small region of interest on a CCD, you either have to read out the whole frame and digitally ignore the parts you don't care about (no increase in read speed) or the chip has to support charge dumping, where portions of the image can be dropped without reading them out. CMOS makes it much easier to read small portions of the image, and thus things like live view are simply done.
Lastly, I'd like to mention an issue with the Foveon sensor that the article didn't mention. While the foveon technology presents itself as having three discrete pixels stacked on top of each other, the reality is much more ambiguous. Foveon relies on the ability of different wavelengths of light (colors) to penetrate to different levels in silicon, however, far from being discrete, easily separated regions, the depth of capture of different wavelengths in silicon tends to be very blurry and ill defined. This results in significant color mixing in the foveon design. They manage to pull out the images they do through the use of extensive processing. That makes the technology pretty unappealing to design with, thus the dominance of Bayer sensors. Honestly, if I needed to do a camera with true RGB per pixel, I would use a 3-CCD design where a full sensor is dedicated to each color rather than using anything like a foveon. It would be more expensive, but far better quality.
Thank you for clarifying several points from the Design Engineer perspective. I appreciate your insights into the CCD vs. CMOS issues.
Sometimes it is difficult for people to wrap their heads around the idea that a technology (CMOS) is not the low noise champion, but that it is winning nonetheless because of other attributes such as lower cost, manufacturing efficiency, lower cost, integration advantages, lower cost, lower power consumption, and lower cost. Your comments put that reality into perspective.
That's all very interesting, but unless you are designing a very high IQ, special purpose (read, very expensive) device, that's simply not an assumption that can be made..
CCD's have numerous problems. High power requirements, which lead to higher temperatures, which leads to higher noise levels, requires cooling for the best results, and so on.
There is no inherent IQ advantage to CCD's. The longer development time led to an early start, and all the advantages accrued from that. But that lead shrunk.
Also, when talking about cost/performance, we must realize that it is very important to not lose sight of the fact that performance must be compared at reasonable cost levels. NASA can afford to spend a million for a sensor, which they do, but it's irrelevant to everyone else.
As for the Foveon chip, yes, that is one of the problems I was talking about, and the biggest one.
Some enthusiastic reviews and articles have taken Foveon's word that they undergo little processing compared to Bayer chips. That's only true in the de-matrixing area, and so they don't need a the aliasing filter. But the color mixing problems are just as serious, and I've found, in using the camera, that color quality is more variable than with my 5D. Often noticeably poorer as well.
As stated on the Canon CMOS Technology page, it says that although CCD sensors achieve high IQ, they are have slow data-reading speed, which means that it wouldn't be capable to capture at a fast FPS mode.
CMOS is capable of high data-reading speed, but due to crosstalk (between pixels) the IQ suffers as a result. However, due to developments in CMOS technology, we can assume that IQ has improved closer to CCD.
My opinion is that the cost, speed, battery life and potential for improvements is what compelled many to adopt CMOS instead of CCD.
I've recently bought Canon 40D that uses CMOS (right?) and I'm a bit disappointed in the image quality over my Rebel XTi (dead). There is definitely much less noise and it's incredibly fast, but I just can't get images to be as sharp as with the XTi. In particular with 100mm Canon macro lens. I don't know if I got a defective body or the CMOS sensor is indeed softer than CCD in the XTi. I'm considering returning the 40D and fixing my XTi instead.
Canon typically tunes the JPEGS from the entry-level bodies to be more "punchy" (more saturation, more sharpening, more contrast) than on higher-level bodies, to provide results closer to the average P&S output.
If RAW, might just be processing variations between RAW profiles for different bodies. Have you tried some sharpening in photoshop or similar to compare images?
Also possible the body is missing focus, but before jumping to that conclusion I'd try the image adjustments first.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
72 Comments
Back to Article
gheinonen - Tuesday, October 14, 2008 - link
I am curious why that CMOS sensor in the high end Canon camera body has excellent low black noise compared to the images from a Fuji Pro S3/S5 with its Super CCD? I have used Fuji Pros since 2004 and have discovered that my black details in low light situations include a lot of random color noise which I do not see when reviewing images from the 1DS Mark2 body. What does the Canon body do to eliminate the random color noise in low light black detail? Is it the CMOS Sensor? Is it the body processing?Separately, the white detail on my Fuji Pro S3 has such expanded dynamic range that I can shoot higher exposures and then lower the exposure back to normal in software and it appears to lower or mask the noise floor in the same way that Dolby Noise Reduction works for audio.
bonedaddy - Wednesday, April 30, 2008 - link
I've been a 35 mm fan for years, and have a significant investment including multiple lenses, macro, ring lights etc. For trips etc the smaller cameras seemed fine--always had small 35 mm, for instance. However, re the digital small cameras, the amount of compression is really disappointing.Is my only choice to go back to a body/lens SLR if I want wide angle and telephoto capability AND good resolution?
Midwayman - Thursday, May 1, 2008 - link
No. But if you want wide/tele and really good high iso performance a SLR is where you need to be. PS camera have alway been a compromise. Small 35mm film cameras had focus issues, and lens issues too. Plus most people use iso 200-400 film which has reasonable quality even in a PS digital camera. The biggest difference is now we're blowing up the picture to 1:1 on our monitors and can see the quality defects easily. I bet if you printed your old compact photos at something like 16x20 you'd probably be unhappy with them too. That's the sort of scale we're looking at on our monitors zoomed in.CyniCat - Thursday, April 24, 2008 - link
Good article, but one glaring mistake: you claimed Sony was the first to make a 12+Mpixel CMOS sensor. I think you meant the first AFTER Canon - the 5D, with its 12.8Mpixel sensor, was on the market in 2005, and the 1Ds Mark II, with a 16Mp sensor, was on the market earlier than that.On a different front, I thought the Nikon D3 was using a Sony sensor, not a Nikon?
Wesley Fink - Friday, April 25, 2008 - link
The 5D and 1Ds II and III are full-frame sensors. The Sony was the first consumer (read affordable) APS-C sensor. Canon now has their own 12 megapixel consumer sensor in the XSi, which we are now reviewing.Nikon did their own designs for the D3 sensor, but they do not, to our knowledge, have the capabilities to manufacturer that sensor. Sony has manufactured sensors for them in the past and present with the D300, D60, D80, and others.
Since the new full-frame is CMOS it is likely manufactured by Sony, or possibley Samsung. Sony and Samsung (who make the Pentax 14.6 megapixel sensor) jointly own several patents on CMOS manufcaturing technology.
Wesley Fink - Friday, April 25, 2008 - link
That should read the read "First consumer 12 megapixel APS-C sensor". As mentioned several times in the article Canon pioneered CMOS technology in a consumer DSLR with the Digital Rebel.Midwayman - Wednesday, April 23, 2008 - link
This article complete skips of fuji's fantastic super CCD technology. Its not really a traditional Bayar array, nor a foveon. True they don't produce a interchangeable lens SLR but they do make prosumer SLR's with their sensor. Also the very notable fuji f30 series cameras were made with this sensor. It was a true triumph in PS camera high iso usability.ElFenix - Thursday, April 24, 2008 - link
fuji makes interchangeable lens SLRs with Nikon F mounts.Midwayman - Friday, April 25, 2008 - link
Hell, then there is no excuse for it not to be included in this article. Especially when it quite clearly states there are no other sensor options at one point.Wesley Fink - Friday, April 25, 2008 - link
There will always be fans of one technology or another who feel slighted. I apologize for that, but Fuji is still basically a Bayer sensor with a differnt pixel shape. I have added the following to the Bayer vs. Foveon page to make Fuji fans a bit happier:"FujiFilm produces one current DSLR with a variant of Bayer technology. It is called the Fuji S5 Pro and is basically a Nikon D200 body with a Fuji Super CCD sensor. The Fuji S5 Pro uses the Nikon lens mount. The Super CCD still uses red, blue and geen pixels in the same standard Bayer ratios. However, the shape of the pixel is hexagonal rather than the squate or rectangular pixels in other Bayer arrays. In the latest version Fuji also added smaller photosites between the normal pixels to gather "dynamic range" data.
Fuji has updated the camera body from the S3 to the S5 in the past year, but the sensor has not been updated for more than 3 years. The current Super CCD is still a 6.3 megapixel sensor, but Fuji specifes it as a 12.3 megapixel due to the addition fo the tiny "brightness" pixels. Tests indicate the true resolution is more comparable to an 8 to 10 megapixel sensor from competitors. The Fuji sensor is still basically a Bayer sensor with a different shape for pixels."
We can probably now all argue whether the Fuji Super CCD is really a Bayer variant or not. It certainly appears that way to me, and as a CCD instead of a CMOS sensor it is need of a serious update if it is to continue as a player in the DSLR market.
Midwayman - Friday, April 25, 2008 - link
I'm happy. I'll won't argue that it's not a variant, but its design is different enough to warrant a mention. Even for such an old sensor, it still produces a dynamic range that exceeds notables such as the 5d, etc. I know I love my Super CCD based f30.http://www.imaging-resource.com/PRODS/S3/S3PA6.HTM">http://www.imaging-resource.com/PRODS/S3/S3PA6.HTM
Has a reasonably good discussion on the super ccd for those interested.
haplo602 - Thursday, April 24, 2008 - link
http://www.fujifilm.com/products/digital_cameras/s...">http://www.fujifilm.com/products/digital_cameras/s...ElFenix - Wednesday, April 23, 2008 - link
http://gizmodo.com/383170/giz-explains-digital-cam...">http://gizmodo.com/383170/giz-explains-digital-cam...Ajax9000 - Monday, April 21, 2008 - link
Two criticisms of the article.1> The sensor size discussion was rather muddled, with half of the information on p.4 duplicating the info on p.2
2> Yes, full-frame is really nice to aspire to, but the problem of full-frame sensor cost was not mentioned. IC lithographic machines ("steppers") cannot image a FF sensor in a single pass, making the production time cost 2-3x greater than smaller sensors (APS-H is the biggest that can be done as single pass). Worse, the large size of the sensor compared to the wafer means that yields drop appreciably as the sensor size increases. All in all, FF sensors can be 10-20x more expensive than APS-C sensors, let alone 4/3, 1/1.8", etc.
See for example: p.68 in http://www.canon.com/technology/pdf/tech2008e.pdf">http://www.canon.com/technology/pdf/tech2008e.pdf and pp.11&12 in http://www.robgalbraith.com/public_files/Canon_Ful...">http://www.robgalbraith.com/public_files/Canon_Ful...
To quote Canon (2nd ref): "Regardless of future technological developments, cameras with full-frame sensors will always cost much more than cameras with smaller sensors."
ElFenix - Wednesday, April 23, 2008 - link
steppers used to not be able to make 35 mm sensors in one pass. as canon and nikon are two of the bigger stepper manufacturers, guess what has likely happened? and both of them are smart enough to not announce that they've made the larger stepper, so as to make more profit than the other.Wesley Fink - Tuesday, April 22, 2008 - link
Page 2 emphasized the huge size difference between compact point-and-shoot-sensors and DSLR sensors as classes of sensors. Page 3 then explored how all the DSLR sensor sizes evolved within that class, and page 4 detailed the current variations in sensor size within the DSLR class. I appreciate your comments, but can you provide some specifics on what needs to be changed?The full-frame sensor does indeed cost more, which was mentioned a number of times in the article. However, prices are still dropping for the full-frame sensor as evidenced by the fact that the street price of the Canon 5D is now around $2000.
Ajax9000 - Wednesday, April 23, 2008 - link
Some suggested edits:Page 1
[Delete http://images.anandtech.com/reviews/cameras/2008/s...">http://images.anandtech.com/reviews/cameras/2008/s... ]
Page 2
Sensors Today
Table: Sensor Comparison
Type Width (mm) Height (mm) Area (mm^2) Percent Full Frame Crop Factor (Lens Multiplier)
1/3.6" 4 3 12.00 1.39 8.49
1/3.2" 4.54 3.42 15.53 1.80 7.46
1/3" 4.8 3.6 17.28 2.00 7.07
1/2.7" 5.37 4.03 21.64 2.50 6.32
1/2.5" 5.76 4.29 24.71 2.86 5.91
1/2" 6.4 4.8 30.72 3.56 5.30
1/1.8" 7.18 5.32 38.20 4.42 4.76
1/1.7" 7.6 5.7 43.32 5.01 4.47
2/3" 8.8 6.6 58.08 6.72 3.86
1" 12.8 9.6 122.88 14.22 2.65
4/3" 18 13.5 243.00 28.13 1.89
"APS C" 23.7 15.7 372.09 43.07 1.52
APS C 25.1 16.7 419.17 48.52 1.44
APS H 30.2 16.7 504.34 58.37 1.31
35mm film 36 24 864 100 1.00
In the above Table we can see the range of sensor sizes now available. However, the majority of Compact or Point and Shoot cameras today still generally use a 1/2.3" to 1/2.5" sensor. A few top-of-the-line compact cameras, like the Canon G9, feature a 1/1.8" to 1/1.7" sensor; and the very best compact cameras have sensors around the 1/3" to 1/2" range. In contrast, the sensor sizes for today's DSLR cameras are in the range of 4/3" and APS C. A few top pro cameras now sport 35mm-size sensors and are referred to as full frame.
The APS C to 4/3 sensors of the bulk of today's digital SLR cameras are huge by comparison to those found in mainstream Compact or Point and Shoot cameras. To see the difference in the relative size of P&S sensors and DSLR sensors, look at the graphic below.
http://images.anandtech.com/reviews/cameras/2008/s...">http://images.anandtech.com/reviews/cameras/2008/s...
[Delete http://images.anandtech.com/reviews/cameras/2008/s...">http://images.anandtech.com/reviews/cameras/2008/s... ]
Why Does Sensor Size Matter?
...
Page 3
Why All these Different Sensor Sizes?
35mm first appeared on the scene in the 1930s, using an image frame of around 24x16mm. 35mm still film simply took 35mm motion picture film and spooled it into a light-tight canister, turned the spool direction by 90 degrees, and used double the frame size. (In fact, some early 35mm still cameras were referred to as "double-frame" cameras.) By the 1960s, with point-and-shoot and developments in SLR technology, 35mm had become king of the film formats. Even as film manufacturers tried to introduce other film formats, 35mm continued to grow and prosper.
...
... Actually, the image size that was always shot was the 30.2mm x 16.7mm, and the other sizes were just standard crops. The 24mm by 16mm APS C format is approximately the same size as the 1/2 frame 35mm championed by Olympus in the film era.
Page 4
Sensor Size and Multipliers
[Delete http://images.anandtech.com/reviews/cameras/2008/s...">http://images.anandtech.com/reviews/cameras/2008/s... ]
...
More recently, Canon has championed the full-frame sensor in their pro cameras and in their pro/amateur 5D model. Larger sensor manufacturing cost has dropped as digital sensors have evolved and it now appears likely the APS-H (1.3X lens multiplier) will eventually drop from the Canon line. Since no lenses depend on that image circle all current Canon full-frame lenses, such as those used with the APS-H cameras, will remain usable on the full-frame sensor pro models that will replace them.
The developing push for full-frame at the top of the current DSLR market is a move to a sensor that is a bit more than double the size of today's APS C sensors. And there is a significant problem with this -- the full-frame sensors are made using different process techniques to even APS H sized sensors, greatly increasing their relative costs. In short, the IC lithographic machines ("steppers") that make the sensors cannot image a full frame sensor in a single pass, making the production time cost 2-3x greater than smaller sensors (APS-H is the biggest that can be done as single pass). Worse, the large size of the sensor compared to the wafer means that yields drop appreciably as the sensor size increases. According to Canon USA, full frame sensors can be 10-20x more expensive than APS-C sensors.
(
As a comparison to computer components, most CPUs are less than 300mm^2. The very large and expensive G80 and R600 GPUs are about 480mm^2 and 420mm^2 respectively. Unlike CPUs and GPUs that can increase performance through a die-shrink, APS, 4/3, and full frame sensors have a fixed die size -- so Moores Law doesn't apply.
http://pc.watch.impress.co.jp/docs/2007/0423/kaiga...">http://pc.watch.impress.co.jp/docs/2007/0423/kaiga...
http://pc.watch.impress.co.jp/docs/2007/0705/kaiga...">http://pc.watch.impress.co.jp/docs/2007/0705/kaiga...
)
It is not surprising then that dSLRs mostly use APS C and 4/3 sensors.
Nikon, Pentax, and Minolta all had significant success in the 35mm film market, so all three had a vested interest in preserving their 35mm lens mount and keeping their current 35mm system users happy. However, none of these three had the resources to develop and manufacture their own sensors, so they partnered with sensor manufacturers to produce digital SLR cameras. Of course, these manufacturers were mostly making APS C, 4/3, and smaller sensors. In recent years that partner has been Sony, so all three manufacturers have basically adopted a 23.6x15.7mm sensor with a 1.5X lens multiplier.
...
CSMR - Monday, April 21, 2008 - link
I am not an expert but it seems to me the most important point has been missed in this article: depth of field. Larger sensors allow for greater depth of field with the same lens. There is a lot of discussion of "image quality" (resolution, shutter speed, sensitivity) but depth of field alters what the image is, which is more important.Wesley Fink - Monday, April 21, 2008 - link
In the interest of keeping the article somewhat basic and a digestible size we did not include a discussion of the depth of field variations with compact sensors compared to DSLR sensors.This IS an important concern for creative use of depth of field since the DOF is much greater with the smaller sensor than with the DSLR sensor. Not everyone would agree that a shallower depth of field on a DSLR is more desirable. It certainly might be in a portrait, but not in a sweeping landscape shot. The desired DOF depends on what you are shooting.
melgross - Tuesday, April 22, 2008 - link
This is a very complex subject for many people. It's not just the sensor size that determines depth of field.But, we also have perspective to worry about.
Johnmcl7 - Monday, April 21, 2008 - link
I do agree with you there that it is desirable to have deeper depth of field (particularly on macro) but I think the difference is that on an SLR you have the choice - you can shoot wide open and have shallow depth of field or you can stop down (and normally have to use up the ISO advantage) to get depth of field. On a camera with a small sensor you are generally stuck with the deep depth of field whether you want it or not.John
finbarqs - Monday, April 21, 2008 - link
I hope nobody is using F/1.2 when shooting a portrait. Unless you want only a nose to be in focus. With something THAT shallow, you either have to REALLY step back and use distance to your advantage to get what you want in Focus. Professional Wedding photographers will not TOUCH anything above f/5.6 as you don't want to get ONLY the bride in focus while the groom is OOF.People, remember there are multiple factors that affect DOF. Aperture is definitely one of them, distance, and Focal Length.
Obviously you want something very shallow with great bokeh, then you would get close to your subject, open up your aperture, and have a telephoto focal length.
The thing is, on a FF sensor, you "see more" on the same focal length as a cropped sensor. Well put wesley, a 50mm is a 50mm. For those of us with cropped sensors, slap a 50mm on your camera. (or zoom to 50mm). Look at your subject with your naked eye, then look through the view finder. SAME SIZE AND SHAPE! Nothing gets larger! Supposedly a telephoto range (greater than normal) should have some zoom effets shouldn't it? I mean APS-C sensors are either 1.5 or 1.6 smaller than a full frame, that means the normal focal length should be around 27mm? So 50mm should be 75mm for nikons and 80mm for canons! That means you should see some zoom in 50mm! but nope! you don't! in fact, the only time you'll see the zoom is AFTER you fire the shutter and it displays the image to fill that 2.5" or 3.0" inch lcd you have! (or even 2.0") So where does the FF excel? you place 2 cameras exactly side by side of each other, same focal length, same focus point. Same lens. Through the Viewfinder, you'll see the nice Bokeh of your subject on BOTH cameras, exactly the same. Difference? the FF has more frame coverage! Thus, having better "DOF" (or lack thereof) because of the more frame coverage! On this note, let's all by Hasselblad's Medium FF and get some nice digital backs! ;)
haplo602 - Tuesday, April 22, 2008 - link
Hmm ... In my understanding the difference in FF to APS-C is CROP.A 50mm lens on APS-C has the same DOF, same perspective characteristic as on a FF camera, but the sensor is smaller, so you basicaly get a CROP of the original FF image. There is only a perceived zoom effect as the image is a smaller area.
For Nikon you get 1.5x smaller image area, but everything else is the same as on FF.
melgross - Tuesday, April 22, 2008 - link
It has to do with distance from the subject as well. Perspective is the ratio of the distance to the subject to the distance to the background.So, to get the same size on sensor, with the same lens length, you must either move closer, or further, changing the perspective, and the depth of field.
finbarqs - Monday, April 21, 2008 - link
Medium Format* sorry, typo :)Johnmcl7 - Monday, April 21, 2008 - link
Other way round - a larger sensor gives you *less* depth of field which as you quite rightly point out is a strong advantage of a larger sensor.John
pinto4402 - Monday, April 21, 2008 - link
DOF is also a function of the lens. Some ultra-serious photographers are willing to pay $2500 for a Canon 85mm 1.2L over a 85mm 1.8 (which is just as sharp and which only retails for $300), because the 1.2L has a softer, more pleasing DOF effect.TETRONG - Monday, April 21, 2008 - link
Cool article.Looking to buy a camera pretty soon here.
What do you folks think about the G9?
Is there a particular brand I should favor?
Also, are there any cameras available that do "tricky" stuff, like Infrared, Macro, or that neat effect when cities look miniature? (Can't find an example)
I read an article about a guy a few years back that would setup a scene, then he would walk around selectively lighting certain aspects in trippy ways to make a final composite image.
Anyone know what this is called, or where I can learn more about it?
Much obliged!
strikeback03 - Tuesday, April 22, 2008 - link
Is "fisheye" what you are looking for for the miniature cities?and the selective lighting is often called "Painting with light"
plenty of compact cameras can do OK macro if you can get really close to your subject, but as this is often hard for reasons of lighting and/or scaring off the subject, an SLR with a real macro lens is still beneficial.
pinto4402 - Monday, April 21, 2008 - link
There was an article at http://www.popphoto.com/popularphotographyfeatures...">http://www.popphoto.com/popularphotogra.../4923/th...about infrared photography. It also discusses a company that is modifying certain digital cameras for infrared use.
bjacobson - Tuesday, April 22, 2008 - link
Infrared is largely not done due to issues with privacy.TETRONG - Tuesday, April 22, 2008 - link
No worriesThis company is making anti-infrared underwear
http://www.cramer.co.jp/products/guardshorts.html">http://www.cramer.co.jp/products/guardshorts.html
aliasfox - Monday, April 21, 2008 - link
When everybody was still on 35mm film, it was impossible to tell whether an SLR shot an image or a good point and shoot/rangefinder style camera. Image noise was a function of the film, and within a certain range, SLRs and point and shoots could reproduce the same sharpness and depth of field. All of that is pretty much impossible these days with the split in sensor sizes.My biggest question is: if camera companies could make a pocketable 35mm/APS sized camera 10 years ago, why can't they make the same pocketable size with a digital sensor? An APS-C sized sensor in a Canon G9-style body shouldn't be impossible, and while it would be expensive, it would have image quality worth the price.
I have a feeling that the companies producing SLRs are just trying to protect their high-margin lens business - a company that's not vested in the SLR side of things could carve out a very solid niche by making a point and shoot with an SLR sensor.
And yes, I know the Sigma DP1 exists, but without a zoom lens and max aperture of F4.0, $800 is a lot to ask...
idealego - Wednesday, April 23, 2008 - link
A lot of it has to do with consumers expecting zoom lenses with cameras. The larger the sensor the larger the lens has to be, especially with a zoom lens. I don't think it's possible to have a decent zoom lens on a pocketable camera with a large sensor (with current technology). This is why the DP1 has these limitations. This is also why a camera such as the Sony R1 has such a large lens compared to other cameras with smaller sensors.However, I'd be quite happy with a fast prime lens on a well-designed, pocketable, large-sensor camera, even if it did have a bunch of downsides to it.
If the DP1 was more refined and had a much faster lens, it would be an awesome camera. Unfortunately, it's rather crude compared to cameras from other leading brands.
Johnmcl7 - Monday, April 21, 2008 - link
"My biggest question is: if camera companies could make a pocketable 35mm/APS sized camera 10 years ago, why can't they make the same pocketable size with a digital sensor? An APS-C sized sensor in a Canon G9-style body shouldn't be impossible, and while it would be expensive, it would have image quality worth the price.I have a feeling that the companies producing SLRs are just trying to protect their high-margin lens business - a company that's not vested in the SLR side of things could carve out a very solid niche by making a point and shoot with an SLR sensor. "
I don't think that's the case at all, film and digital cameras have to be a lot more different than you perhaps appreciate. The Sigma DP1 has had a prolonged and difficult development, while I think that's partially because Sigma are less experienced it's not entirely the story. One of the main problems I believe is the way light hits the sensor, with a digital sensor to work optimally you need to have the light hitting the sensor as straight as possible. Film was more flexible in this sense which made it easer to produce small cameras using comparatively large pieces of film. I believe this was one of the difficulties in producing the DP1 as the lens is very close to the sensor which meant the light is hitting the sensor at very high angles. This is easier on an SLR as the sensor is a reasonable distance back from the rear of the lens (Olympus particularly have made a goal of telecentric lenses) although even the 35mm full frame cameras can suffer from poor corner sharpness.
Also with film you didn't have any low pass or IR filter in front of the film itself which is normally part of a digital camera, I believe the DP1 (as well as the Leica M8) don't actually have an AA filter at all which causes its own problems.
Hopefully other companies will have a go seeing the interest that the DP1 is generating, I think Panasonic who produce a 4/3 sensor and are having little success in the SLR market are in a prime condition to do so. However it's a more difficult task than it seems.
John
Heidfirst - Monday, April 21, 2008 - link
"I have a feeling that the companies producing SLRs are just trying to protect their high-margin lens business "when Kodak got out of doing DSLR bodies (even if they were basically Nikons with a different sensor) they said it was because the profits weren't in the bodies but in the lenses - which of course they didn't make ...
Pentax is supposed to be getting out of compacts because they're becoming a commodity item & hence price-sensitive/low margin.
I could imagine someone like Leica doing what you want though as they seem able to command a significant premium.
aliasfox - Monday, April 21, 2008 - link
Leica's currently tied to Panasonic sensors, and relatively speaking, Panasonic hasn't figured out how to make a sensor that has low noise above ISO 200 yet, regardless of size (otherwise I'd seriously contemplate the Lumix LX-2).Also, while I'm willing to pay a premium for a nice fixed lens camera (I'll consider up to $700, maybe even $1000 for a camera with the right handling, image quality, and features), the market for such an expensive "small" camera is probably pretty small (unless professionals are all clamoring for a pocketable camera as a backup in the field).
Lastly, the Leica name generally commands an even healthier premium than I'm willing (or financially able) to consider...
pinto4402 - Monday, April 21, 2008 - link
The sensor is only half the story. The lens is the other half. This is NOT marketing hype. The difference between consumer kit lenses and pro series glass is easily discernible (even to non pixel peepers). Does this justify the 8X difference in price? I don't know.However, I'm pretty sure that it's not possible to construct a high quality P&S with a great lens that is "pocketable," regardless of the sensor size. Good glass tends to be heavy. Even the old Leica and Contax rangefinders were not pocket cameras. They were compact, but I doubt anyone considers them pocket cameras (unless you're talking about cargo or jacket pockets, and you trust having a $6000 camera hanging out of your shirt pocket). When you add zoom functions to the equation (which is a must nowadays), you looking at some serious IQ trade-offs.
That said, I would be the first in line to purchase a good P&S camera that produces high quality professional grade images.
aliasfox - Monday, April 21, 2008 - link
Honestly, I'll be happy if someone can fit an under 2" thick camera with an adequate (say, 28-112mm equivalence, f2.8 - f4.0) lens and a large/larger sensor. Maybe not even an APS-C sensor - a 4/3" design is already significantly larger than a 1/1.8" sized unit. Olympus, are you listening?The Olympus E-420, even with its pancake 28mm equivalence lens, is too thick to put into a jacket pocket (or crammed into some looser fitting jeans).
I rarely ever take dedicated trips to take photos - so packing a backpack half full of photographic equipment to take a "nice" shot is something I'm loathe to do. But I do go cycling, hiking, and other normal touristy activities where I'd like to be able to bring a camera along...
strikeback03 - Tuesday, April 22, 2008 - link
There was the Sony R1 - APS-C sensor and 24-120 Zeiss-branded lens. Not exactly pocketable, even not having to deal with the registration distances required by an SLR with the swinging mirror. Maybe if you were willing to do something with folded optics you could fit a zoom in a smaller body, but who knows how much of an image compromise that would be.My XT with 28 1.8 does fit in the cargo pocket of one pair of pants I own, but I would rather just have it in a small shoulder bag.
Johnmcl7 - Monday, April 21, 2008 - link
On page 4:"The Foveon sensor falls between 4/4 and the Canon 1.6 in size and has a 1.7X lens multiplier."
I assume this should read '4/3'
Obvious question is what about Fuji? While I realise they re-use Nikon bodies and lenses, the discussion about Bater and Foveon makes no mention of Fuji's sensor. While it is more conventional than the Foveon, it's not just a bog standard bayer sensor and while Fuji are a minor player, so are Sigma.
For those who are not familiar with Fuji's design, they use two photodiodes at every point one larger and one smaller with the two combined to produce a single pixel in the output image. The idea is that these pair of pixels can capture more extended dynamic range than a standard bayer sensor of the same size. The latest version of this sensor is in the Fuji S5 Pro however it's limited to just 6MP although I can't remember if they still produce 12MP files from this. The S5 itself is basically a Fuji version of the very good Nikon D200 body although Nikon have of course moved on with the very impressive D300.
John
Wesley Fink - Monday, April 21, 2008 - link
Yes, it should have read 4/3 and the reference is corrected. The Fuji S3 was one of my favorite Nikon bodies and the dynamic range was certainly impressive in the studio.We did not mean to slight Fuji, but as one Nikon-mount body with a sensor that hasn't been updated in several years (the S3 and S5 sensors are the same as I understand it) we decided not to include the Fuji since there have been no sensor updates in quite a while.
Johnmcl7 - Monday, April 21, 2008 - link
"We did not mean to slight Fuji, but as one Nikon-mount body with a sensor that hasn't been updated in several years (the S3 and S5 sensors are the same as I understand it) we decided not to include the Fuji since there have been no sensor updates in quite a while. "I can't say I really agree there, given the article more focuses on concept in parts I think the S5 sensor is still relevent as it's something slightly different to the bayer sensor. The article refers to only Bayer and Foveon which implies there is nothing else, I think for completeness even if you don't go into any detail it's still worth mentioning Fuji are doing something else.
Also the S3 and S5 sensor are not the same, while they have the same amount of pixels it appears there's been some slight improvements although clearly not much. To be far to Fuji though, the Foveon sensor hasn't really come on much either - it's gone up very slightly in resolution with some small other changes but that's it pretty much. If the Foveon sensor had been just mentioned in passing I could definitely understand leaving Fuji with a similar mention but generally Fuji and Sigma are considered in the same boat as doing something a bit different although arguably with the Nikon body and mount Fuji have had more success.
John
melgross - Tuesday, April 22, 2008 - link
I never saw an advantage to their designs. I can't see what purpose having a smaller photo site on the sensor would do. It just has more noise, and less dynamic range than the larger sensor. I've read their papers on the subject, and they don't seem to have made a good case for it. Somehow, I think they understand that now.Johnmcl7 - Thursday, April 24, 2008 - link
"I never saw an advantage to their designs. I can't see what purpose having a smaller photo site on the sensor would do. It just has more noise, and less dynamic range than the larger sensor. I've read their papers on the subject, and they don't seem to have made a good case for it. Somehow, I think they understand that now."Are you referring to Fuji? If so, your information is incorrect - the last measurement I saw put the S5's sensor at slightly more dynamic range than the 35mm sensor in the Nikon D3. Their real problem at the moment seems to be resolution as well as having the older D200 based body.
John
strikeback03 - Tuesday, April 22, 2008 - link
I think Sigma (and Foveon) would be better off if Sigma could license a major player's mount, like Kodak did with the SLR/n and SLR/c. There are plenty of people who would like to have the sensor for the situations where it excels, but have no interest in a whole Sigma SA mount setup.pinto4402 - Monday, April 21, 2008 - link
I've been reading Anandtech for over 8 years now. I was a bit skeptical about your doing articles on digital cameras; however, this article put my reservations to rest. Very nicely done.I see why it makes sense for Anandtech to write about digital cameras. The nexus between computer tech and camera tech are very obvious if you've been following the trends. I'm a professional portrait photographer. In the last few years, it has become virtually impossible to remain in business unless you have a firm grasp on the latest camera tech as well as computer tech. I spend as much time on my computer as I do behind the camera. The camera has become a computer accessory (or vice versa). Many old timers who are hanging on to film are slowly being forced out. I'm somewhat of an old timer myself because I learned about photography when it was cool to have a darkroom, but I embraced digital equipment early.
Your graphs make it very easy to explain to people why their P&S (piece of s***) cameras are simply not adequate for serious portraiture. The MP count is marketing crap. As you demonstrated, it's the physical size of the sensor that matters.
Looking forward to part II of the series. Also, do you have any solid info on the introduction of 5D Mk II?
bjacobson - Tuesday, April 22, 2008 - link
"The MP count is marketing crap."Mostly. If you're willing to do some post processing yourself, the higher MP will enable you to decrease the IQ gap between the higher end DSLR and the ho-hum consumer camera. For this reason, since I wanted something compact and didn't really need a lot of optical zoom, I chose the Canon SD1100IS. 8MP, and while you begin to get noise at ISO400, more at ISO800, and tons at ISO1600, using a non-linear digital filter should correct most of that without blurring the image much (if at all).
strikeback03 - Wednesday, April 23, 2008 - link
Problem is that the camera has already blurred away lots of your detail at ISO 400 and up.http://www.dpreview.com/reviews/canonsd1100is/page...">http://www.dpreview.com/reviews/canonsd1100is/page...
Too bad there is no option to reduce/turn off the in-camera NR, for those of us who own a better program for it already.
Wesley Fink - Monday, April 21, 2008 - link
I wish I did have definitive info about the 5D Mark II launch, but the best info I have is this fall at Photokina. Rumors pop up every month that the new 5D will be here in a few weeks - and the last rumor was a definite April 22, which is tomorrow. Rumors are just that - rumors.andrewln - Monday, April 21, 2008 - link
This was followed in about 6 months by the introduction of the Samsung 14.6MP CMOS sensor in the Pentax D20.should be Pentax K20D
Wesley Fink - Monday, April 21, 2008 - link
Typo corrected. Unfortunately spell check can't catch model numbers that are misstated.araczynski - Monday, April 21, 2008 - link
now i know what pc tech illiterate people feel like :)its a good thing i don't care about photography, interesting read none the less.
finbarqs - Monday, April 21, 2008 - link
the thing is, I've read the reviews and saw the comparisons between the 5D and the D200 (I believe that was the last CCD sensor that Nikon used). The 5D has way better per-pixel sharpness than the D200. Perhaps the technology of the CMOS made it so it finally "looks" better than the CCD. Or at least on the Canon side of things. I'm not biased towards any camera. I've own an XTi, 5D, Panasonic DMC-L1, and now a D300. I'm hoping to own the next 5D MKII (or whatever they call it).Even the D2X used a CMOS sensor, and I thought Nikon made the D300 sensor, as well as the D3 sensor (FF). What I also find "funny" is that Sony doesn't have a FF sensor yet, but nikon does, thus leaving me to believe that Nikon came up with their own FF sensor.
melgross - Monday, April 21, 2008 - link
It's Nikon's own design, though I don't remember who makes it.haplo602 - Monday, April 21, 2008 - link
Nice article Wesley, finaly one Anandtech photo related article I enjoyed reading.On note on the growing megapixel count. Sooner more than later, DSLR in APS-C (and later full frame ones) will hit the same technology wall P&S are facing (too small photosite).
This is one factor that makes me a happy film shooter :-) I know that my limit is the scanner up to around 10MP and I have less flexibility in shooting conditions (either 2 bodies or limited by ISO and film type), in every other situation I am equal or better off.
I am waiting for an affordable Nikon full frame body and then I will make my switch to digital (but that is yet years to come).
wally626 - Monday, April 21, 2008 - link
A sensor technology improves APS-C will be able to go to high enough pixel counts and have very good quality. For most consumers some where around 12 MP is enough, if the sensors improve to where this can hit 3200 ISO with low noise 99 percent of the market would be satisfied.The full frame bodies will replace the medium formats of the film world. There have been some really good medium format cameras that take much better images than 35mm but very few are sold. I think the article is correct in saying the full-frame DSLR will be the PRO cameras and priced as such.
melgross - Monday, April 21, 2008 - link
The number ofpixels is directly related to the print size.If you go by the oft quoted 300 dpi on the final print for maximum quality, you will need a sensor with 2400 x 3600 resolution for a full frame 8 x 12 print, or 8.64 MP. For an 8 x 10, it would be 7.2 MP (4/3 sensor).
For a larger 11 x 17, it would be 3300 x 5100, or 16.83 MP, or 13.86 for the 11 x 14 4/3 sensor print.
You can figure the rest of the sizes my multiplying the inch size of the print edges L x W by 300 to come up with a number.
But the truth is that 240 dpi is going to be good enough fot most prints, and youcan do the numbers that way.
When you know the numbers, you can figure out what size sensor you will need for the highest quality work.
But for most people, even 180 dpi will be enough for their prints. Going to that gives more flexibility, as the sensor pixel count is much smaller.
It's better to get a camera with a longer OPTICAL zoom rather than to go for the biggest number of sensor pixels.
haplo602 - Tuesday, April 22, 2008 - link
^ this ^It's all a matter of print size. I am an amateur/hobbyist, and anything past A4 is large for me. I have printed reasonable quality 8x12 prints from a consumer slide film and home film scanner.
I can hang them up on the wall in a nice frame and they will serve their purpose :-)
Idealy I'd need a 12-14MP full frame Nikon with goot ISO/Noise characteristics (f.e. D3) and it will satisfy my needs for years to come.
Heidfirst - Monday, April 21, 2008 - link
especially as I was wondering the other day what smaller process technology would do for digital sensors?Normally of course smaller process means cooler & more importantly smaller>cheaper to make but of course with sensors you are talking a fixed size so not any cheaper.
But would the potential increased precision of circuitry mean any better image quality?
& a little nitpicking:
Sony didn't buy Konica Minolta (which continues in business) but certain assets from the Photo Imaging division of KM. As to what that actually means there is little hard knowledge outside the 2 companies as it seems that KM still retain some IP & indeed a shareholding in some of the production facilities.
Also, current thinking seems to be that the "A900" may not be called that but something else to differentiate it from the APS-C models.
melgross - Monday, April 21, 2008 - link
The smaller process technology will have no positive effect on the sensors themselves, though it will for the associated electronics integrated on the die.The same problems he mentioned about smaller sensing sites will remain. The smaller the sensor, the poorer the performance viz a viz larger sensors.
He did mention that the photo division was the purchase, not the entire company (unless he changed the article after your post).
Wesley Fink - Monday, April 21, 2008 - link
I changed the wording on the Sony purchase a bit to better reflect that Sony bought the Minolta camera assets of KM and not the company. Thanks for pointing this out.finbarqs - Monday, April 21, 2008 - link
Canon's 1DS MK2 was a 16.7MP CMOS sensor also, and of course, the MK3 is a 21MP CMOS sensor...Where i'm lost is if CCD's are so much better (in IQ) why dont' they stick with CCD's? why the move to CMOS besides the lower cost and the battery life that it saves?
Why are professional level DSLR's (From canon and Nikon) are both CMOS when we know that CCD is the way to go for better IQ?
melgross - Monday, April 21, 2008 - link
It USED to be true that CCD's were better. Not so any longer. The best CMOS sensors are better than the CCD's they replace.The desire to go CMOS is obvious to the manufacturers of the sensors.
CCD sensor technology is a completely different manufacturing process from that of CMOS, which the entire industry uses for everything else (almost).
Moving to that allows CMOS sensors to not only be able to integrate other electronics on the sensor chip, resulting in simplicity, price advantages, and the ability to more favorably utilize their process lines, but that higher quality you're concerned about.
Putting functions on the same chip improves the quality of the signals.
And, by the way, an error in the article: Canon was not the first to make, or use, a CMOS sensor. They were the first to come out with a high quality sensor. I believe that it was Vivitar that used the first one, though I forgot the name of the manufacturer.
Anther omission is that there are trilinear sensors used in camera backs such as the Betterlight scanning backs. So there are three different major technologies in use.
And not all of the negatives of the Foveon chip was mentioned.
s12033722 - Monday, April 21, 2008 - link
No, CCD is definitely still the IQ king. CCD still has a far better SNR than any CMOS technology.As a digital camera design engineer, I deal with image sensors every day. The major reasons why CMOS sensors are attractive are all cost related. Not only are CMOS sensors themselves cheaper, but they lend themselves to integration with other electronics better and they are MUCH nicer to design with. A typical CMOS sensor will require ground, 3.3V, and maybe some other standard voltages (1.8V, 2.5V, etc.), whereas a typical CCD will require ground and anything from 8 to 12 other DC voltage rails. For instance, I am working on a camera that requires -15V, -9V, -6V, -4V, -1.5V, ground, 2V, 3.5V, 11 V, 15V, 24.5V, and has a clock signal that must run up to 40V. While making the voltages and driving clocks at them is fairly straightforward, it requires a lot more components than a CMOS sensor design would. More components directly equals higher cost. Also, as the article mentions, more functions can be integrated onto CMOS sensors than CCDs.
The other advantage of CMOS vs. CCD is in random-access readout. If you want to read a small region of interest on a CCD, you either have to read out the whole frame and digitally ignore the parts you don't care about (no increase in read speed) or the chip has to support charge dumping, where portions of the image can be dropped without reading them out. CMOS makes it much easier to read small portions of the image, and thus things like live view are simply done.
Lastly, I'd like to mention an issue with the Foveon sensor that the article didn't mention. While the foveon technology presents itself as having three discrete pixels stacked on top of each other, the reality is much more ambiguous. Foveon relies on the ability of different wavelengths of light (colors) to penetrate to different levels in silicon, however, far from being discrete, easily separated regions, the depth of capture of different wavelengths in silicon tends to be very blurry and ill defined. This results in significant color mixing in the foveon design. They manage to pull out the images they do through the use of extensive processing. That makes the technology pretty unappealing to design with, thus the dominance of Bayer sensors. Honestly, if I needed to do a camera with true RGB per pixel, I would use a 3-CCD design where a full sensor is dedicated to each color rather than using anything like a foveon. It would be more expensive, but far better quality.
Wesley Fink - Tuesday, April 22, 2008 - link
Thank you for clarifying several points from the Design Engineer perspective. I appreciate your insights into the CCD vs. CMOS issues.Sometimes it is difficult for people to wrap their heads around the idea that a technology (CMOS) is not the low noise champion, but that it is winning nonetheless because of other attributes such as lower cost, manufacturing efficiency, lower cost, integration advantages, lower cost, lower power consumption, and lower cost. Your comments put that reality into perspective.
melgross - Monday, April 21, 2008 - link
That's all very interesting, but unless you are designing a very high IQ, special purpose (read, very expensive) device, that's simply not an assumption that can be made..CCD's have numerous problems. High power requirements, which lead to higher temperatures, which leads to higher noise levels, requires cooling for the best results, and so on.
There is no inherent IQ advantage to CCD's. The longer development time led to an early start, and all the advantages accrued from that. But that lead shrunk.
Also, when talking about cost/performance, we must realize that it is very important to not lose sight of the fact that performance must be compared at reasonable cost levels. NASA can afford to spend a million for a sensor, which they do, but it's irrelevant to everyone else.
As for the Foveon chip, yes, that is one of the problems I was talking about, and the biggest one.
Some enthusiastic reviews and articles have taken Foveon's word that they undergo little processing compared to Bayer chips. That's only true in the de-matrixing area, and so they don't need a the aliasing filter. But the color mixing problems are just as serious, and I've found, in using the camera, that color quality is more variable than with my 5D. Often noticeably poorer as well.
7thSerapHim - Monday, April 21, 2008 - link
As stated on the Canon CMOS Technology page, it says that although CCD sensors achieve high IQ, they are have slow data-reading speed, which means that it wouldn't be capable to capture at a fast FPS mode.CMOS is capable of high data-reading speed, but due to crosstalk (between pixels) the IQ suffers as a result. However, due to developments in CMOS technology, we can assume that IQ has improved closer to CCD.
My opinion is that the cost, speed, battery life and potential for improvements is what compelled many to adopt CMOS instead of CCD.
Zak - Monday, April 21, 2008 - link
I've recently bought Canon 40D that uses CMOS (right?) and I'm a bit disappointed in the image quality over my Rebel XTi (dead). There is definitely much less noise and it's incredibly fast, but I just can't get images to be as sharp as with the XTi. In particular with 100mm Canon macro lens. I don't know if I got a defective body or the CMOS sensor is indeed softer than CCD in the XTi. I'm considering returning the 40D and fixing my XTi instead.Z.
strikeback03 - Tuesday, April 22, 2008 - link
Canon typically tunes the JPEGS from the entry-level bodies to be more "punchy" (more saturation, more sharpening, more contrast) than on higher-level bodies, to provide results closer to the average P&S output.If RAW, might just be processing variations between RAW profiles for different bodies. Have you tried some sharpening in photoshop or similar to compare images?
Also possible the body is missing focus, but before jumping to that conclusion I'd try the image adjustments first.
teng029 - Tuesday, April 22, 2008 - link
the rebel xti uses a cmos sensor.http://www.usa.canon.com/consumer/controller?act=M...">http://www.usa.canon.com/consumer/contr...&fca...
melgross - Monday, April 21, 2008 - link
Since when did the XTi use a CCD?