Talk:Four Thirds system
|This is the talk page for discussing improvements to the Four Thirds system article.|
|WikiProject Photography||(Rated C-class)|
- 1 Why it's not an "Open Standard"
- 2 Name explanation
- 3 Image proportions:
- 4 Lens/body communication?
- 5 Image sensor format
- 6 Open Standard?
- 7 300 = 600?
- 8 Sensor size comparison
- 9 Smaller sensor = more noise
- 10 Requested move
- 11 Advantages section
- 12 Advantages regarding dust
- 13 Removed one of the disadvantages
- 14 Depth of field
- 15 Wide aperture wide lenses hard to design
- 16 Discussion of older mounts
Why it's not an "Open Standard"
Quoting the footnote at the bottom of http://www.four-thirds.org/en/about/benefit.html (as of 13:36 30 Jan 2008 UTC)
* Details of the Four Thirds System standard are available to camera equipment manufacturers and industry organizations on an NDA basis.
Full specifications cannot be provided to individuals or other educational/research entities.
- I had read this exact quote on the official website and I too wondered how they could call this an "open" standard. However, this statement also clearly states that any company may enter in an agreement with them to get the full specification; I guess that is what they see as "open". I find it most intriguing that they would prohibit individuals and other researchers to access the standard... Anyway, back to the "citation needed", I think what is controversial about the marked phrase is the "commonly accepted definition" statement. Maybe this could be rephrased with something such as this: "This is claimed to be an open standard; it is however accessible only to companies and under a non-disclosure agreement." (and then link the page mentioned above). AlexFicelle (talk) 17:43, 10 November 2008 (UTC)
I agree that the standard can't be classed as an "open" standard in the normal sense of the word, it is limited to certain entities (companies only), and the fact that it is available under a non-disclosure agreement makes it no different to other proprietary standards owned by a company or organisation. For the standard to be open it has to be published, open to all entities (individual and companies), available at nominal cost (i.e. cost of distribution+administration), and available without restriction on the use of the information. I can understand that the companies involved in the Four Thirds organisation feel it is very open, but this is only because their normal business practices are to have very secretive standards used for product development. (18.104.22.168 (talk) 04:43, 5 December 2009 (UTC))
Can anyone explain the connection between the name "four-thirds" and the size of the sensor? The article presently contains the text
The name of the system comes from the size of the imaging sensor. The sensor used is a "4/3 inch" type sensor. The length is based on the diameter of particular size of Vidicon tube—the actual light sensitive area being smaller. The actual length of the sensor diagonal is 22.5mm.
So more specifically, my questions are: In what way does the name come from the size? In what way is the length based on the diameter of some Vidicon tube? Note that 4/3 inch = 33.9 mm, which is much bigger (50% bigger) than 22.5 mm. That is, the sensor is nowhere near 4/3 inch in diagonal. Fg2 04:56, August 11, 2005 (UTC)
- Indeed it's not the physical dimension of the sensor. There's a good explanation of the confusing fractional definitions camera manufacturers use here: http://www.dpreview.com/learn/?/Glossary/Camera_System/Sensor_Sizes_01.htm. 22.214.171.124 08:45, 23 December 2005 (UTC)
No, it is not the dimensional size of the Four-Thirds sensor, and I think the original author who inserted the reference to old analog video camera tubes was taking a leap. Anyway, he mis-spoke. The "4:3" or "4/3" refers to the aspect-ratio of the sensor. That is, the ratio of width to height: four units wide, by three units tall (as stated in the article, 17.3mm by 13mm; note that "micro-Four Thirds" cameras have the same exact sensor dimensions, the difference is the smaller lens mount). And so it is. The confusion has arisen because the name is spoken as "four thirds", which sounds better than "four-by-three." Googlyelmo (talk) 21:00, 20 October 2013 (UTC)
I have asked this question on Olympus press meeting, and after few days, I have received e-mail from their technical division:
question: If the image circle is a part of a standard, does it mean that the sensor does not need to have 4:3 proportion? Can there be a camera constructed, by a manufacturer, meeting the 4:3 standard and have the square (1:1) image sensor put inside the same image circle?
Answer: yes just the circle is defined that the ratio of the sensor can also be different from 4/3, but 4/3 was determind as perfect measure.
What data is communicated between lens and body?
Specifically, does the lens communicate the subject distance to the body, and can the body instruct the lens to focus at a specific distance? Abu ari 09:04, 2 April 2007 (UTC)
- Using the E400, I am yet to find such features, I am sure there are no such features.
- I think the answer is "no" to both questions. But this is not the place to post questions like this, this page is for talking about the Wikipedia article itself. Try posting your question on a digital photography forum, for example: http://forums.dpreview.com/forums/forum.asp?forum=1022
- --RenniePet (talk) 23:52, 24 November 2007 (UTC)
Image sensor format
have removed this link - the four thirds system covers aspects in addition to just the sensor format - its more about the imaging circle and the lens mount than the sensor size.
For the record, I don't think the separate page is warranted - but that's probably a discussion left until after the issue of whether to delete the "full frame" page is decided. --Hmette 06:56, 7 April 2007 (UTC)
- Agree that the standard covers aspects in addition to just the sensor format... but it does cover the sensor format, not just the imaging circle. So why you'd remove the link escapes me. Expand the description by all means but don't remove useful content is my advice. Andrewa 09:55, 7 April 2007 (UTC)
"Four Thirds is not an open standard, however, as it does not meet the "allowing anyone to use" criterion commonly accepted as the definition of an open standard." This sounds like point of view as there's no citation. On wikipedia's own Open Standard page it appears that the term itself has a broad array of interpretations and there is no clear definition of the term at all. I recommend removing this line or citing a source which shares the author's criticism as a way to highlight criticism of the standard's closed nature.
Reply to the above- this is not just a point of view, as the standard is only available to companies. Wikipedia definition of Open Standard, first line "An open standard is a standard that is publicly available". Fails even this 1st generally accepted "open standard" criterion. It is only open when compared to the "trade secret" approaches camera manufacturers normally take. Some viewpoint from authors must be accepted on Wikipedia otherwise a manufacturer can claim anything is "open" in a press release and the Wikipedia entry just parrots the press release! Claims which are not substantiated by reality should be pointed out. --126.96.36.199 (talk) 04:51, 5 December 2009 (UTC)
300 = 600?
The explanation that a 300 mm lens has the same angle as a conventional 600 mm lens was bogus. The difference in angle is a direct result of the ratio of sensor size on many Olympus SLRs to a standard 35 mm frame. I suspect, but don't know, that telecentricity might make for a shorter 300 mm lens, but that doesn't change its focal length.
Groogle 23:20, 3 December 2007 (UTC)
- I've tried to restore the comparison in a more correct way. Let's talk about it here if you think I'm wrong. Thanks. --RenniePet 16:10, 4 December 2007 (UTC)
The angle-of-view given by any given lens on a Four Thirds (or micro Four Thirds) camera is NOT precisely double that of the same focal length on 35mm film or full-frame digital, though for common practical purposes it can be treated as such. The latter have a 3:2 (or 1.50x as wide as high) aspect-ratio, while FT/MFT are of course 4:3 (or 1.33x as wide as high, noticeably squarer). This difference in geometry affects the conversion factor. Angle-of-view is measured across the image plane diagonal, and will be a few percent shorter/wider than double.Googlyelmo (talk) 21:15, 20 October 2013 (UTC)
Sensor size comparison
The article says "Its area is ... around 15× larger than the 1/2.5" sensors typically used in compact digital cameras". Looking at the diagram, the area of the Four Thirds sensor is 225 mm², and the area of a 1/2.5" sensor is 25 mm². So shouldn't it say "9x larger"? Matthewk (talk) 17:01, 23 December 2007 (UTC)
Smaller sensor = more noise
I'm a fan of the Four Thirds System, but in the interests of honesty I think the statement about noise problems should say that it begins to be relevant from above ISO 400. It previously said "ISO 400 and up". Another editor changed that to "ISO 1600 and up". Wish that was true. Anyway, I've now changed it to "above 400", and added a link to a brand-new review of the Olympus E-3 in DPReview, where it says, for that model at least, "Noisier results than most competitors above ISO 400 (usable up to ISO 1600)". --RenniePet (talk) 00:24, 23 February 2008 (UTC)
The article says that the Four Thirds flange to focal plane distance is significantly shorter than a number of "competing" brands. In the case of Canon FD (hardly a competitor) the difference is actually only 3.4mm, which is not enough to allow an adapter to be made that allows infinity focus. FD lenses can be converted to fit a Four Thirds mount but it is a destructive process that prevents the lens from being used again with an FD body. A Canon FD to Micro Four Thirds adapter is, however, available. 188.8.131.52 (talk) 18:54, 24 March 2009 (UTC)
Advantages regarding dust
The smaller sensor makes the system less likely to suffer from dust problems. There's simply less sensor area where the dust could accumulate. More importantly, the small sensor rarely needs the lens to stop down too much. Unlike larger formats, the Four Thirds system can attain excellent depth of field even at f/5.6 or f/8.0. It rarely needs to be stopped down smaller than f/11 --- which is when a potential dust problem rears its ugly head. Consequently, dust is rarely visible in Four Thirds even when present. In contrast, 35mm full-frame format are often stopped down to f/16 or smaller --- thereby exposing all the dust that may be in the sensor.
- I've deleted some of the material. The writer was working from two premises. 1) That the proportion of a lens that is in use is smaller when the iris is set to a smaller f-stop. (This is true.) 2) That the amount of distortion caused by a piece of dust of a given size is less with a smaller sensor. (This is false, in fact, it's the opposite.) In the second premise the writer was confusing the *chance* of a piece of dust hitting a sensor with the *effect* of the dust being on the sensor.
- But besides that, the writer was engaged in original research speculation. If a sensor is a charged -- a "dust magnet" -- then being smaller would be a bad thing -- since all dust in a confined area would tend to be drawn to a smaller surface. The writer doesn't appear to consider that, or know anything about it.
- My concerns with small sensors is that I can't as easily see the dust, and can't as easily remove it. That a mistake I make in cleaning will mar a proportionately larger part of the surface. Piano non troppo (talk) 02:55, 30 July 2009 (UTC)
- There's really no way to compare dust problems on 4/3 and on other systems since all 4/3 cameras have Olympus' nifty dust cleaner technology. In practice they're basically dust-free, but this has nothing to do with Four Thirds itself, just a particular gizmo that Olympus sticks on the cameras. --184.108.40.206 (talk) 04:11, 30 August 2009 (UTC)
Removed one of the disadvantages
First off statement had no sources.
First statement was that one of the disadvantages stated that as of 2009 only olympus is the only manufacturer of four third system cameras. Panasonic dSLR cameras are all four third and has also started releasing Micro Four Third System Cameras (http://panasonic.net/avc/lumix/systemcamera/gms/index.html) Leica
As for saying that the orginal supporters have distanced themselves from four third, again where are the sources. All the supporting companies are still listed as supporters on the official fourthird website. --Unsungheroz83 (talk) 08:37, 4 August 2009 (UTC)
Also even if Olympus ever becomes the last four thirds, I don't see how it would be cassified as a disadvantage, this would not hinder it's performance nor it's cost and most of it's four thirds top performing lenses come from Olympus anyway.—Preceding unsigned comment added by 220.127.116.11 (talk • contribs)
- You were right to remove the paragraph. Contained mostly WP:Original research (e.g. "have seemingly distanced"). Unrelated: it seems you logged out between edits, that's a bit confusing.... Rror (talk) 12:31, 4 August 2009 (UTC)
Depth of field
Article currently reads:
- Because of the higher crop factor, an image shot at a given relative aperture and angle of view will have more depth of field on Four Thirds. This results in less control over depth of field, compared to larger formats. For instance, a 35mm full frame DSLR can match the depth of field of a Four Thirds camera by simply closing the aperture down two stops. On the other hand, it may be more difficult or impossible for a Four Thirds system to match the depth of field of a 35mm full frame camera with fast lenses.
This seems wrong. For any size sensor, for a print of a given size (which of course is a larger 'enlargement' for a smaller sensor), the depth of field at a given aperture and angle of view should be the same; isn't that a universal fact of camera optics? If someone doesn't find a WP:RS for this claim, I will delete it. --macrakis (talk) 05:35, 4 January 2010 (UTC)
This link about DOF seems to explain it well wrotniak.net "For example: you are using a Four Thirds camera (M = 2) with a zoom lens set to the focal length of 25 mm (EFL = 50 mm) at the F/2.8 aperture. The depth of field you will be getting will be identical to that obtained with a 35-mm film camera using focal length of 50 mm, providing the same image angle, closed down to F/5.6 (as 2.8×2=5.6)." - Victor Vorski 7 January 2010 —Preceding unsigned comment added by 18.104.22.168 (talk) 15:40, 6 January 2010 (UTC)
- The corresponding passage was written by me, I think (although I am not sure anymore, it has been a while). In my opinion, a specific source is unnecessary for two reasons: (1) this being common knowledge (or so I thought) to anybody in the digital photography field and (2) the DOF depedency being the result of a little bit of simple math. I can include it in the article, but it would be superfluous and disruptive in my opinion; it is also already covered in the DOF article: Depth of field#DOF vs. format size. The simple explanation for the format dependency of DOF is that f/2.0 on FT is in fact smaller on an absolute scale than f/2.0 on APS-C, which in turn is smaller than f/2.0 on 35mm (since the focal lengths corresponding to a given viewing angle are all different). --Kilessan (talk) 13:23, 7 January 2010 (UTC)
- I think I understand the source of confusion here. I looked over the DOF calculations you pointed to, and they all appear to be based on DOF on the sensor (contact print DOF). But what most people care about is DOF on the print (enlargement DOF). Yes, DOF on the sensor is inversely proportional to format size for the same distance and field of view. But smaller formats must be enlarged more for a given print size, which exactly cancels this effect. Right? --macrakis (talk) 18:39, 9 January 2010 (UTC)
- Well, no. Take for example a portrait photograph, taken with a 35mm full-frame camera at 100mm focal length with an aperture of f/2.0. For the sake of simplicity we will also assume that every picture is printed on paper of the same size, (e.g. letter). The lens has a very wide aperture (or more precise: an entrance pupil) of 50mm (such lenses are rare and expensive, but they do exist, e.g. the Nikkor 85mm f/1.8). Not regarding DOF for the moment, you can get exactly the same image if you take a Four Thirds camera at focal length 50mm, since its crop factor is about 2.0.
- However: the DOF is a function of the geometry in front of the camera and of the absolute size of the aperture. We did not change anything in front of the camera. But what about the aperture? For focal length of 50mm, an entry pupil of 50mm means f/1.0. However, f/1.0 lenses are extremely rare (and none exist for FT, at least to my knowledge), so you will have to use e.g. a f/1.4 or f/1.8 lens. Only now, the aperture actually is only 28-36mm, so it is significantly smaller, thus giving you an increased DOF. This is a very well-known fact in digital photography, being also the reason for images shot with compact cameras looking quite "flat" by comparison (since a tiny 1/2.5" sensor camera produces images with near-infinite DOF). If you have both a compact and DSLR, just try it: choose the "equivalent" focal length for both cameras, identical F-numbers, and shoot images with both devices.
- The problem of "contact" vs. "enlargement" DOF actually does not exist, since one always assumes that abberations smaller than 1/1500th of the image diagonal (and thus: sensor diagonal) are still perceived as "sharp". So the calculations actually all deal with the the enlargement case. If you are still not convinced, I can include the calculations in the article, but I really think they do not belong there. Also, it is explained quite well in the DOF article. I am effectively quoting that article (and numerous other books and calculations), and this is not a FT-specific discussion anymore. If you question that crop cameras in general (APS-C, too!) show increased DOF proportional to crop factor, I suggest to move that discussion to the DOF article. --Kilessan (talk) 21:53, 9 January 2010 (UTC)
- I finally got around to going through the derivation, and now I understand. Sorry for the confusion. I agree that the calculations do not belong in the article. On the other hand, I think the discussion of depth of field in the article is rather wordy. I think it would be enough to say that at equal f-stop and angle of view, the DOF is roughly inversely proportional to the sensor size, so Four Thirds has somewhat more depth of field than APS-C cameras and much less than point-and-shoot cameras. --macrakis (talk) 20:24, 18 January 2010 (UTC)
- The DOF phenomenon is not directly related to the size of the image sensor as much as the pixel size. A smaller sensor of the same resolution will have higher pixel density, thus smaller pixel size, and thus smaller circle of confusion, generally yielding higher DOF. A larger sensor with the same pixel density will have the same DOF, all other variables held constant.—Preceding unsigned comment added by 22.214.171.124 (talk) 17:53, 15 June 2010 (UTC)
Wide aperture wide lenses hard to design
Article currently reads:
- A telecentric optical path means more aggressive retrofocus design for wide and normal lenses, which makes wide aperture wide and normal lenses harder to design. Currently, there are only two lenses shorter than 30mm with apertures brighter than f/2: The very costly Panasonic/Leica 25mm f/1.4 and the Sigma 24mm f/1.8 Macro.
I agree that these lenses are expensive compared to 'normal' lenses in other formats--perhaps there simply isn't much demand for a fast normal lens? On the other hand, the Olympus 14mm-35mm f/2.0 zoom costs about the same as the 35mm near-equivalent Nikkor 24-70mm f/2.8, which is somewhat wider, but one stop slower. The Canon equivalent is about 25% cheaper (but still one stop slower), so I'm not sure what any of that proves about how "hard" it is to design these telecentric wide-angles.... --macrakis (talk) 05:48, 4 January 2010 (UTC)
- To 'prove' that statement will be quite impossible, yes. I am not too happy with the paragraph myself, though it does hold some truth from an optical point of view. A short focal length means a rear lens very close to the sensor. Telecentricity requirements (and the mirror) mean that this cannot usually be done, so one has to use retrofocus design. This is undisputed, I think. Retrofocal designs "tend to be" more expensive simply because they include more optical elements, which either reduces image quality or increases costs.
- A case in point, perhaps, is the ultra wide-angle 7-14mm f/4.0 zoom by Olympus. This corresponds roughly to a 10-20mm zoom on APS-C. For the latter there are two options e.g. by Sigma (one with f/4.0-4./5.6 and one with f/3.5). The Olympus/Zuiko lens is even a bit slower than the faster Sigma, while costing almost thrice as much. It still is a magnificent lens, but extremely expensive. That said, I am still not too happy with the paragraph, since it is too imprecise and open to interpretation. --Kilessan (talk) 13:46, 7 January 2010 (UTC)
- Haven't wide-angle lenses on all SLR cameras had to be retrofocus designs, regardless of telecentricity? --macrakis (talk) 14:42, 7 January 2010 (UTC)
- Good question. I would think that for any form factor there is a focal length that is "too small" and thus needs retrofocus. If this is the case indeed for all usual wide-angle lenses on APS-C and 35mm systems, I cannot say. For example, the flange-sensor distance is much larger in Nikon cameras than in Canon. However, the question concerning the cost of the lens is not only if retrofocus is employed at all, but also the amount by which the focal point has to be shifted. And of course, how many lenses are produced and sold, thus on market share. This discussion effectively underlines my problem with the passage: while it does indeed become increasingly hard to maintain certain telecentricity requirements when using very small focal lengths, this is nearly impossible to quantify in terms of lens cost. So, the statement in the article does not seem wrong to me from a technical point of view, but it does not contain much "hard" information. At this point I am in favour of removing the statement altogether. --Kilessan (talk) 15:29, 7 January 2010 (UTC)
- As far as I know you need to use retreofocus if the lens is wider then the flange focal distance. This is the case for any lens wider then about 38mm on the four-thirds standard. The Canon EF-Mount eg has about 44mm so and Nikons F-Mount 46mm, and that's where the different is for wideangle lenses you almost always need retrofocus but on 35mm the 50mm is well above the flange focal distance but for APS-C using the same mount the pictures changes already a 35mm APS-C lens needs to be built as a retrocus lens. The point from which wideness retrofocus is needed changes. --firstname.lastname@example.org 06:57, 19 January 2010 (UTC)
Discussion of older mounts
I think the paragraph about adapters for old mounts is too long, and the reference to a website on OM legacy lenses far too specialized (and non-notable in the WP sense) to bother mentioning. I would suggest instead something like:
- Because the flange-to-film distance is small, lenses from other systems can often be fitted to Four Thirds cameras with simple mechanical adapter rings.