arthur53821
Senior Member
Go to glosary / camera system / sensor size :
http://www.dpreview.com/learn/?/Glossary/Camera_System/Sensor_Sizes_01.htm
http://www.dpreview.com/learn/?/Glossary/Camera_System/Sensor_Sizes_01.htm
-
Welcome to the new forums! Please read this first. For known issues we are working to resolve, click here.
four thirds: Please Explain
- Thread starter Todd Frederick
- Start date
David Hughes124436
Veteran Member
Snip, Snip, Snip!
Welll, to understand it you you have to look at the way things were when it started (messy) and you'll see it was a new, large and sensible standard. By making it a standard they gave it to others to use in the same way that Philips came up with an audio cassette standard and didn't mind who used it as long as they had a share and everyone was marching to the same tune and in the same direction. Seems logical to me.
It may or may not have been overtaken but you have to compare like with like. Don't bring something NASA have made into it and claim it's normal for example, 'cause it ain't.
As standards go "Four Thirds" has a lot going for it. An d the cameras take brilliant pictures.
As for "645" wasn't that the name of a new camera for the rich amateurs years ago. I seem to remember it gave the 4:3 aspect ratio but wasn't the price of a "pro" camera which would use 6 cm by 9 cm and cost a fortune. In other words it was aimed at people who wanted something bigger than 35 mm but at an affordable (comparatively) price. Then a lot of others came out. Also, it jammed a lot more pictures on a roll of 120 film (sigh).
Regards, David
Hi,
Welll, to understand it you you have to look at the way things were when it started (messy) and you'll see it was a new, large and sensible standard. By making it a standard they gave it to others to use in the same way that Philips came up with an audio cassette standard and didn't mind who used it as long as they had a share and everyone was marching to the same tune and in the same direction. Seems logical to me.
It may or may not have been overtaken but you have to compare like with like. Don't bring something NASA have made into it and claim it's normal for example, 'cause it ain't.
As standards go "Four Thirds" has a lot going for it. An d the cameras take brilliant pictures.
As for "645" wasn't that the name of a new camera for the rich amateurs years ago. I seem to remember it gave the 4:3 aspect ratio but wasn't the price of a "pro" camera which would use 6 cm by 9 cm and cost a fortune. In other words it was aimed at people who wanted something bigger than 35 mm but at an affordable (comparatively) price. Then a lot of others came out. Also, it jammed a lot more pictures on a roll of 120 film (sigh).
Regards, David
Phil Rose
Senior Member
Although that's true enough, it only helps to make David's point; how strange it would be to encounter "a [self-proclaimed] rabbit" who showed no interest about hopping and eating vegetables. Of course rabbits, unlike theologist/philosophers, are more likely to be self-evident than self-proclaimed.
Phil
KuroNeko
Forum Enthusiast
That's the part I'm missing. I've only been into DSLR's for about a year and Olympus didn't take my attention until the announce of the E-300. If I'm not mistaken the 4/3 standard was introduced back in early 2003 with the E-1? I don't know how sensors size panorama was back then, I don't think the size choose was small at that time, not for a 5 Mp sensor at least.
But as I understand, the size of the sensor is closely related to the design of the lenses, so if in a future they reach a limit in terms of resolution/noise performance and decide to size-up the sensor of future models, you won't be able to use the FT ZD lenses with those new models, will you?
KuroNeko,
if you search in this forum, you will probably find dozen of threads and thousands of posts discussing these issues. I have just two quick comments.
a) Ultimately, judge by results, not by theoretical arguments. We never assessed film camera systems by debating the chemistry of emulsions, did we?
But I cannot resist one bit of theory:
b) LENS APERTURE SIZE is the bottom line for speed/noise performance, meaning roughly the area of the lens surface over which the camera gathers light from the subject. This is determined by the aperture DIAMETER independent of focal length or format. For larger formats and longer focal lengths, same aperture diameter means higher aperture ratio. This adjustment of aperture ratio to get same aperture diameter in different formats also gives the same depth of field.
Apertuer size is the key because getting a good signal to noise ratio requires gathering enough light from the subject, and the aperture diameter measures the total rate at which light is gathered from a given subject: total photons, and therefore photons per pixel with sensors of equal pixel counts.
So the real questions of physical speed limits are
1) how large an aperture diameter can you use when you need adequate DOF?
2) how big can the maximum aperture diameters of 4/3 lenses be?
and
3) how large can the maximum aperture diameters be for a given cost?
Only (1) has an easy answer: the same for any format, so there is no inherent speed/noise advantage to a larger or smaller format once you are seeking adequate DOF: a larger format needs a higher f-stop, but can compensate with a higher ISO or longer exposure time.
For (2) and (3), Olympus has not pushed the speed limits with its entry level and mid-priced zooms, so it will be interesting to see what their promissed range of high end zoom lenses look like.
For (3), lens size, weight and cost seem roughly related to the size of the big front elements, which is roughly determined by aperture diameter and angular FOV, independent of format. So the cost/speed trade-off should not vary much between formats.
if you search in this forum, you will probably find dozen of threads and thousands of posts discussing these issues. I have just two quick comments.
a) Ultimately, judge by results, not by theoretical arguments. We never assessed film camera systems by debating the chemistry of emulsions, did we?
But I cannot resist one bit of theory:
b) LENS APERTURE SIZE is the bottom line for speed/noise performance, meaning roughly the area of the lens surface over which the camera gathers light from the subject. This is determined by the aperture DIAMETER independent of focal length or format. For larger formats and longer focal lengths, same aperture diameter means higher aperture ratio. This adjustment of aperture ratio to get same aperture diameter in different formats also gives the same depth of field.
Apertuer size is the key because getting a good signal to noise ratio requires gathering enough light from the subject, and the aperture diameter measures the total rate at which light is gathered from a given subject: total photons, and therefore photons per pixel with sensors of equal pixel counts.
So the real questions of physical speed limits are
1) how large an aperture diameter can you use when you need adequate DOF?
2) how big can the maximum aperture diameters of 4/3 lenses be?
and
3) how large can the maximum aperture diameters be for a given cost?
Only (1) has an easy answer: the same for any format, so there is no inherent speed/noise advantage to a larger or smaller format once you are seeking adequate DOF: a larger format needs a higher f-stop, but can compensate with a higher ISO or longer exposure time.
For (2) and (3), Olympus has not pushed the speed limits with its entry level and mid-priced zooms, so it will be interesting to see what their promissed range of high end zoom lenses look like.
For (3), lens size, weight and cost seem roughly related to the size of the big front elements, which is roughly determined by aperture diameter and angular FOV, independent of format. So the cost/speed trade-off should not vary much between formats.
Todd Frederick
Leading Member
I was curious about this because my C-5050 seems to have the same aspect ratio as the E-1, as does my E-10, but the E-1 sensor is much larger physically.
In the C-5050 resolution options, however, I can choose a 3:2 option which is closer to the 35mm proportions, so it would seem that even the C-5050 has a 4:3 proportion, but not the physical size of the
E-1 sensor (ditto with the E-10).
That got me thinking about this and your responses have been very helpful by way of clarification. I enjoy all of my cameras regardless. Thank you. Todd
Todd Frederick
In the C-5050 resolution options, however, I can choose a 3:2 option which is closer to the 35mm proportions, so it would seem that even the C-5050 has a 4:3 proportion, but not the physical size of the
E-1 sensor (ditto with the E-10).
That got me thinking about this and your responses have been very helpful by way of clarification. I enjoy all of my cameras regardless. Thank you. Todd
--
Todd Frederick
KuroNeko
Forum Enthusiast
I'll do it, I'm curious about that stuff. I got carried away answering the first posters that I didn't think of previous posts discussing this matter
).I wasn't implying that the system was bad picture quality wise. I'm very happy with the E-300 so far. What I was questioning is if they'll be hitting a limit with the actual size of the sensor soon, so they couldn't produce higher density sensors with acceptable noise performance, and if that'd be the case (the performance on that regard of the e-300 is what made me think about this, not theoretical arguments), and they change sensor size, will today's FT ZD lenses still usable? I wouldn't call 4/3rds a standard if they wouldn't.
Gerry Siegel
Senior Member
. I enjoy all of my cameras regardless. Thank you. Todd
And that is what counts. Aspect ratio is not square, but accomodationg to the standard page sizes, with the sole exception of the drug store print and that is of little importance to the advanced brains that visit this forum.
The size of sensors and megapixel count,for me, are only relevant as they compare to what 35mm film produce. And it appears, with a few quibbles from the peanut gallery, that the E-1 has reached that level. The model that will be announced late next month will exceed that level by far. And yet, one can hang on the E-1 and the marvelous "intelligent thinking " lenses.
Be well, prunes with your wheatena too....Gerry
And your Ovaltine
And that is what counts. Aspect ratio is not square, but accomodationg to the standard page sizes, with the sole exception of the drug store print and that is of little importance to the advanced brains that visit this forum.
The size of sensors and megapixel count,for me, are only relevant as they compare to what 35mm film produce. And it appears, with a few quibbles from the peanut gallery, that the E-1 has reached that level. The model that will be announced late next month will exceed that level by far. And yet, one can hang on the E-1 and the marvelous "intelligent thinking " lenses.
Be well, prunes with your wheatena too....Gerry
And your Ovaltine
Here is my rough summary.
If high shutter speed is not an issue, getting impressively high resoluiton and low noise levels is not a problem for 4/3" format. The big question is how the physical limits on lens and sensor speed will balance out as fas as overall shutter speed limits.
Resolution: Pixel counts for 4/3" could be pushed to 32MP and beyond simply using pixel sizes from current smaller format digicams like 8MP 2/3". (However I doubt that DSLRs will go that far, except for very expensive high end models; lens resolution has trouble keeping up with such high pixel counts, even in 35mm format.)
Noise levels and ISO: noise is fine with those smaller pixels so long as one stays to low enough ISO speed settings; about 1/2 stop slower than for "APS-C" formats, 2 stops slower than for 35mm format.
Shutter speed: Thus, lens speed becomes the issue: to get the same combination of resolution, noise levels and shutter speed, the new shorter focal length lenses for 4/3 format need to be about 1/2 stop faster than longer focal length lenses for APS-C, and two stops faster than the lenses for 35mm, which also must have twice the focal length. In each case, this means lenses of the same maximum aperture diameter for the different formats, giving the same DOF.
In general, the shorter focal length lenses used with smaller formats can have lower ("faster") minimum f-stops: for a somewhat extreme case, look at the relatively small, cheap zoom lenses for 2/3" digicams that have lower minimum f-stops than even the heaviest, most expensive SLR zoom lenses.
KuroNeko
Forum Enthusiast
Well I'm still not very knowledgeable about all the variables that play a part in the design of a camera and how they are correlated. From what I've understood from your summary, if shutter speed is important, and I think that for a DSLR that is out of question, the only chance for Olympus 4/3rds system relies on faster lenses?
So I take that the sensor size is a problem if you want to get a noise-free picture in high isos (at least as noise-free as its rivals), and to keep from selecting high isos under the same circumstances, the lens will have to be two f-stops faster than its rivals. That's sound to me like a lot, I don't think that the best Olympus glass could match the best X glass on that basis, not having into account that to get a sharp lens is more difficult (aka expensive) if it's also a fast one. am I missing something?
As I said, equal speed requires roughly equal aperture diameter. For normal to telephoto lenses at least, this roughly means the same size and weight of lenses. The 35mm format lens must have twice the focal length and so twice the minimum f-stop but the same size for the big front elements of the lens.
For example, 200mm f/2.8 in 4/3 format matches 400mm f/5.6 in 35mm format, both with 70mm aperture, and these lenses should be comparable in size and weight.
Only for wider angle lenses is it probably more difficult to get very high speed in 4/3 format, but wide angle is not the area where fast lenses are so much needed.
Anyway , comparisons to 35mm format are rather pointless, as it is fairly clear that 35mm format DSLR's will continue to be very expensive high end options only; Canon has more or less said as much. Only comparisons to APS-C formats have much relevance.
For example, 200mm f/2.8 in 4/3 format matches 400mm f/5.6 in 35mm format, both with 70mm aperture, and these lenses should be comparable in size and weight.
Only for wider angle lenses is it probably more difficult to get very high speed in 4/3 format, but wide angle is not the area where fast lenses are so much needed.
Anyway , comparisons to 35mm format are rather pointless, as it is fairly clear that 35mm format DSLR's will continue to be very expensive high end options only; Canon has more or less said as much. Only comparisons to APS-C formats have much relevance.
Canon APS sensors are 2 not 1/2 stops faster.
According to DPREVIEWS review of the E-300 the 20D's APS sensor is about 2 stops faster than the E-300. Normalized luminance noise[as shown in their graph] for the 20D at ISO1600 is about the same as the E-300 at ISO 400. They went on to say:" Visually the EOS 20D at ISO 3200 doesn't look that much noisier than the E-300 at ISO 800.The Canon EOS 300D / Digital Rebel has almost identical noise characteristics as the EOS 20D. For those who need to know the E-300 uses a Kodak CCD, the EOS 20D a Canon CMOS sensor"
KuroNeko
Forum Enthusiast
That was why I said that ZD lenses will need to be 2 f-stops faster than Canon's in order to match noise performance. All this is rising my curiosity on how well the new E model replacing the E-1 will perform in this department .
.KuroNeko
Forum Enthusiast
The question was about future prospects and limitations on 4/3 format, not performance comparisons between particular current sensors that use different technologies and different approaches to on-sensor and in-camera noise processing.
Those basic phyical limits say that the maximum ISO at which a given signal-to-noise ratio can be attained is proportional to the pixel area. Thus for equal pixel counts, the theoretical limit is about 1/2 stop faster for "EF-S" and 2/3 stops faster for DX format.
Those basic phyical limits say that the maximum ISO at which a given signal-to-noise ratio can be attained is proportional to the pixel area. Thus for equal pixel counts, the theoretical limit is about 1/2 stop faster for "EF-S" and 2/3 stops faster for DX format.
jscott
Senior Member
The 200 f/2.8 in 4/3 gives more intense illumination of a smaller image; these factors balance out so that the total amount of light received by the sensors in a given exposure time is equal, meaning that with equal pixel counts, each pixel receives the same amount of light, and so should be able to produce about the same singal to nosie ratio.
Viewfinder brightness is a bit tricky: with equal VF magnification (e.g. 0.72x at 50mm), the 200/2.8 in 4/3 would again give a smaller but brighter image. However, smaller formats like 4/3 use higher VF magnification (about 1x at 50mm), reducing the size and brightness difference. If a 4/3 body went all the way to having twice the VF magnification of 35mm, the VF images would be equal size, and also about equal brightness with these lenses.
So far, it seems that the 4/3 bodies and lenses are aiming at about the same pixel count as APS-C competitors, which is about half the pixel count of the top of the line 35mm format DSLR's, with telephoto lenses about one stop faster than 35mm options at the same FOV, and VF images of about half the area. These balance out to about the same usable shutter speed and VF image brightness. I await news of the promissed higher end 4/3 body and lenses though.
Keyboard shortcuts
- f
- Forum