Pages: 1 2 [3] 4   Go Down

Author Topic: New Leica Digicam  (Read 20478 times)

leonvick

  • Full Member
  • ***
  • Offline Offline
  • Posts: 108
    • http://www.plateaulight.com
New Leica Digicam
« Reply #40 on: December 08, 2003, 12:25:01 am »

I can't believe this: five (5)  pages of supertech mumbo-jumbo hyperbole about DOF without a single reference to the formula or to the almighty critical role of the "circle of confusion"! The latter is essentially the size of the circle created on the focal plane (film or sensor) by its intersection with the cone of light created behind the lens, by the lens, from a point of light in front of the lens. From a subject point that is precisely in focus, that cone of light converges exactly on the focal plane. From subject points that are out of focus, those cones of light converge in front of or behind the focal plane. The farther a point is out of focus, the farther the point of convergence from the focal plane and the larger the circle projected upon the focal plane. Depth of field is now the the distance between the nearest and farthest points, at a given aperture and lens extension that project acceptably small circles of confusion on the focal plane.

Acceptable, in this case, might be construed as a subjective judgement (dependent upon a viewer's visual acuity and viewing distance) but convention has defined it to be (at the focal plane)  0.00058 or 1/1720 of the focal length assuming that a viewer's distance is that at which a "proper perspective" is obtained. Thus, the acceptable subtended angular size of the circle of confusion of any image is dependent upon both the image size and the viewing distance.

Returning to the camera, the nearest point sharply defined in front of the focal plane is

(d^2*tanC)/(t+d*tanC)

where d=the distance focused upon, C=the angular size of the circle of confusion at 1/1720 of the focal length and t=the effective diameter of the lens (=the focal length divided by the f/number). Likewise, the farthest point sharply defined in front of the focal plane is

(d^2*tanC)/(t-d*tanC).

DOF=the distance between these two points.*

* These formulas are from the Van Nostrand's Scientific Encyclopedia, D. Van Nostrand Company, Inc., New York, 1968. Other sources may be more detailed but these will do for educational purposes, at the very least.

Note that no mention is made here of the size of the focal plane. That's because this is accounted for by the assumption that any real lens used in the calculation will have a sufficient FOV, behind the lens, to cover the film placed at the focal plane. So plug in your various film, focus distances and sensor sizes using any consistent number system and have at it! All the answers are here!  
Logged
Leon
Wherever I go, there I am.

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
New Leica Digicam
« Reply #41 on: December 08, 2003, 01:43:49 pm »

Quote
My interest in this thread began with a comment from BJL who suggested the smaller format digicam has greater potential than we sometimes imagine. Here is a quote that caught my attention.
Indeed, we have wandered a very long way from that initial comment of mine, which was about low light capability, not primarily about depth of field. And even further from the original subject of this thread, the Leica Digilux 2.

   I propose starting a new thread for anyone still discussing this. In order to attract Ray in particular, it will include arguments as to when and why we CAN expect some of the new, smaller digital formats to continue to offer less in the way of large apertures at a given angular field of view than 35mm format, and hence less latitude for extremely shallow depth of field and for fast, low light photography, even though 35mm currently does better than any larger format in these respects.
Logged

victoraberdeen

  • Full Member
  • ***
  • Offline Offline
  • Posts: 205
    • http://www.abovo-media.com
New Leica Digicam
« Reply #42 on: December 09, 2003, 06:52:45 pm »

BJL, What are you talking about, you’re on some mission but what is it!

You can get the photo with any camera, Leica, Rollieflex, MPP exposing a 30th at f5.6 and get a publishable photograph, if the focal length of the lens is the same you’ll even have approximately the same depth of field. The light making the image is the same regardless of image size. Ergonomics is another story!

If, however you want the subject to have the same percentage size in the image, you’ll need to use different lenses and for the same depth of field range a different aperture. If this is what you have been banging on about, it is the wrong reason to choose your format. The selection of format should be based on the purpose of the produced photograph and what determines acceptable and your own standards of image quality.
Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
New Leica Digicam
« Reply #43 on: December 11, 2003, 09:59:03 am »

Quote
One should also bear in mind that, by coincidence, 35mm is also very close to the length of the image so the term does not seem totally absurd even if one were not aware that it actually referred to the total width of the film.

2/3rds on the other hand seems to have no connection whatsoever with any dimension of the 2/3rd's format
Ray, that is quite a stretch: a format is designed with a frame 18x24 mm, diagonal 30mm, with the film width increased to 35mm by then need for sprocket holes on movie film; Leica doubles up the frame to 24x36, and by pure coincidence, 35mm becomes a "reasonable description"!

   The dimensions like 2/3" (based on old vidicon tubes for TV, from long before silicon chips) do have a clear connection to sensor dimensions that is more than a coincidence; they are approximately proportional to sensor size, so one can quickly compare relative sizes, for instance noting that a 4/3" sensor has about twice the linear dimensions and four times the area of a 2/3" sensor.
   The relation to actual dimensions is not so horrible either, for the few of us who actually care about such internal details; the height, width, and diagonal are about .4, .5, and .6 of that dimension; multiply by 10, 13 and 16 to throw in the metric conversion too.

   About manufacturers wanting to hide the small sensor size, I would note on the contrary that in the mass market, small is cool. Many digicams are heavily marketed on the basis of being small, and Fuji boasts about efforts to make sensors smaller. Why would or should most customers be either surprised or bothered by the fact that their small camera contains a small CCD (if they even know what a CCD is, or that the camera contains one), so long as they are happy with its performance? This is on a par with thinking that the average casual photographer is bothered by getting more depth of field. Things are very different in the mainsteam than up here on Mount Olympus, or Mount Canon.
Logged

victoraberdeen

  • Full Member
  • ***
  • Offline Offline
  • Posts: 205
    • http://www.abovo-media.com
New Leica Digicam
« Reply #44 on: December 12, 2003, 05:34:10 pm »

I agree on the LCD, are they doing the hood again?
Logged

victoraberdeen

  • Full Member
  • ***
  • Offline Offline
  • Posts: 205
    • http://www.abovo-media.com
New Leica Digicam
« Reply #45 on: December 10, 2003, 02:19:02 am »

Yes, got me there, I was looking at another digicam... 2/3rds a silly term really - for them that still use inches
Logged

BruceK

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 87
New Leica Digicam
« Reply #46 on: December 10, 2003, 03:25:46 pm »

Quote
To what important photographic dimension does "35mm" refer to?

The width of the film.

    Bruce
Logged

victoraberdeen

  • Full Member
  • ***
  • Offline Offline
  • Posts: 205
    • http://www.abovo-media.com
New Leica Digicam
« Reply #47 on: December 11, 2003, 07:38:25 pm »

DPReview just posted images of the Leica take a close look at the front element it looks most odd. Anyone seen this before?

Hi res images
Logged

BryanHansel

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 369
    • www.bryanhansel.com
New Leica Digicam
« Reply #48 on: December 01, 2003, 04:55:49 pm »

I saw this earlier today also.  I was so excited until my stomach flipped when I read electronic viewfinder.  At least it is 100% coverage.  Leica listed this camera on their web site at:


http://www.leica-camera.com/digitalekamera...x2/index_e.html


I wish it listed how fast from power on until the first shot, and how fast of a shutter delay it had from squeezing the shutter release.
Logged
Bryan Hansel
[url=http://www.paddling

pmkierst

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 78
New Leica Digicam
« Reply #49 on: December 05, 2003, 12:03:13 am »

Quote
The point is that when people skip all the math and skip all the theorizing, they end up concluding something stupid like DOF depends on sensor size when it does not.

In your example, the improved DOF you see is due to the smaller sensor using a much smaller physical aperture.  It is not caused by shorter focal length!

It does not hold true in the real world that DOF effectively depends on sensor size.  It depends on magnification and aperture.  To fix your example, if I shot the scene with 35mm f/8 and 1/250, then shot the same scene with with the 4x5 and appropriate lens at f/28 and 1/60, the DOF for the 4x5 would be marginally better despite the fact that the aperture was identical.

You need to understand that holding the f-number constant across formats is meaningless and bogus.  Once you figure that out, you'll realize the DOF claim is wrong.

Small sensor cameras require less light to expose them so they get by with smaller physical apertures and deliver correspondingly good DOF.  If it were a 5050 you were shooting in the above example, f/8 would be the smallest aperture available.  Why?  It's diffraction limited at f/8!  Shooting a 5050 at f/8 is like shooting your FF 35mm SLR at f/32 all the time (except that your 35mm would provide better DOF).

Now let's talk macro.  With macro you are typically trying to maximize DOF (or at least make sure it's enough).  If you are suggesting that a 5050 provides better DOF for a macro composition that FF 35mm you are sorely mistaken.  At higher magnifications especially, 35mm will trounce the 5050.  Do the math or better yet, take the pictures and you will see.  Do that experiment, Paul, it might make you blue in the face.
OK, "F-Stop" numbers are dependent on focal length and physical aperture (and it is a non-obvious equation depending on exactly where the actual physical device is located). So saying F/8 for one lens and F/8 for another (different focal length) lens already accounts for the difference.

Why is it so stupid? You seem to be quite passionate about this, yet you miss the fundemental point: To get the same DoF for the same Field of View (FOV) for the same speed sensor, you get to use a higher shutter speed on a small format camera. Or, if you have a specific shutter speed in mind, you get more DoF on the smaller format camera.

Note that a 35mm is diffraction limited somewhere around F/18. And probably has about the same DoF for the same FOV.

And yes, I have shot macro on a 4x5. Let me tell you, it is not pleasant. Recommended method (by reputable sources such as "View Camera Technique") to get more DoF? Use a shorter focal length and crop; i.e. use a smaller format.

And yes, if the lenses were equal quality and equally corrected for macro, I would say a small sensor will out-perform a larger sensor.

And the point of all this is usually trying to get shallow DoF, where a small sensor camera can be very troublesome because F/2 on those short focal lengths still gives a lot of DoF.  And yes it is because the physical aperture is smaller, but who cares? Same outcome either way. Getting all caught up in physical aperture sizes really doesn't change the outcome; too much DoF.

And I have done the math.

Since neither of us will move from our positions, my end of the discussion is finished. I agree to disagree.
Logged
Paul K.

Craig Jones

  • Guest
New Leica Digicam
« Reply #50 on: December 05, 2003, 04:32:27 am »

Quote
Dear me! I don't quite know what to say, Craig!  :D  What sort of nonsense are you trying to perpetrate!  ???

The F stop is one of the great concepts of Photography precisely because it is[/i] meaningful across formats.
Meaningful only for exposure purposes!  It is not the same as aperture!
Quote
A 300mm lens at f8 will always require the same shutter speed in the same lighting conditions whether that lens is attached to a 35mm camera, a tiny digicam or a 12"x16" field camera, and it will always produce the same DoF from the same shooting position.
No one cares about DOF when you've made no attempt to hold composition constant.  If what you are saying is that f/8 delivers constant DOF regardless of camera, format, or focal length so long as the camera doesn't move (or the photographer doesn't) that's obviously nuts.  Why don't you talk about different formats with matching perspectives since that's what everyone will ultimately be interested in.  We take pictures of subjects, not just pieces dependent on what our formats allow us to have.
Quote
I've already mentioned in my response to Victor that 'only' changing the sensor size does not affect DoF. But you must realize that nobody designs and manufactures camera lenses without regard to the sensor size, film size, ie. format of camera the lens will be used with. However, it's true that lenses designed for a larger format can be and sometimes are used for a smaller format, the obvious example being the recent crop of DSLRs such as the D60, 10D and 300D.

So let's look at what happens when we take identical shots from the same position with a 50mm lens on a 1Ds and 10D. If the lens is the same and the F stop is the same and the distance to the central subject is the same . . . it follows that the exposure will be the same and the DoF will be the same. That's the beauty of the F stop system.
Actually it is the result of absolutely nothing changing other than the crop.  It has nothing to do with the "f-stop system" and anyone can see that.  The optics and shooting geometry are unchanged so it's no surprise that the image rendered on the film plane is identical.  All that's different is how much is recorded.

Quote
But something will be different, won't it? The 10D shot will be severely cropped. In oder to get the same angle of view with the 10D I'll have to step back a few paces. Consequently, the DoF will now be greater on the 10D shot because I've increased the distance between camera and subject. Alternatively, I could just whack on a 28mm lens to get the same field of view from the same position. That would also have the effect of increasing DoF for the 10D shot (at the same aperture).
In the first case you've reduced magnfication.  In the second case you've also changed the physical aperture assuming you meant "same f-stop".  The actual focal length would be 31mm, so the second case is just theoretical.  If you adjusted the f-stop to account for the change in focal length (same physical aperture) then DOF would not increase.

Quote
Now you could argue that the increased DoF that the smaller format 10D seems to exhibit has nothing to do with its smaller sensor size and that the real reason for the increased DoF is the fact that you either (a) had to step back to get a wider FoV or ( had to use a wider angle lens.
I would argue that in the first case but not the second.  Focal length is not a determiner of DOF, aperture is.
Quote
However, there are reasons behind reasons. If the 10D did not have a smaller size sensor than the 1Ds, there would be no reason to step back or to fit a wider angle lens. In this example (which extends across all formats), the size of the sensor could be viewed as the more 'primary' cause of the increased DoF.
Except that now there would be no difference in DOF because the original persectives match.  There would no longer be a reason to step back or change the lens.  The only reason one could argue that the theoretical FF 10D would have better DOF in this case is that one could claim it had a larger COC due to its lower resolution.  Clearly, two full frame 35mm SLRs with identical lenses, exposure settings, and perspectives deliver identical DOF when it's measured the same way.
Quote
Even from a purely mathematical point of view, there is a connection between sensor size, F stop and aperture if we're talking about standards lenses for the different formats. F stop is a relationship between the focal length of a lens and the physical dimensions of the aperture (F=FL/D). The focal length of a standard lens (as opposed to a telephoto or wide angle lens) is given by the diameter of the image circle the lens throws on the sensor. The size of the sensor or format bears a very direct relationship to the image circle. All 35mm lenses for example are designed to have an image circle a bit larger than the diagonal of the 36x24mm frame (about 43mm). The diagonal of an 8"x10" plate for a field camera is about 12.8"= 320mm. It's no accident that the standard lens for an 8x10 field camera is just that . . . 320mm.

Get my point?  :)

Well, no I don't get your point.  Different format sizes require different focal lengths for matching perspectives.  Is that your point?  I think we all know that.  You are, however, mistaken about DOF and f-stops.  You must change the focal length when you change format to match the perspective.  When you do, you must change the f-stop so that the physical aperture remains the same.  Then the DOF and perspective will match.  The exposure will be off, but that's due to the change in exposure area of the sensor.  Assuming like ISO, you will then have to adjust shutter speed.  If you choose to leave the f-stop the same, DOF will change but it is due to the change in physical aperture, not the sensor size.
Logged

Craig Jones

  • Guest
New Leica Digicam
« Reply #51 on: December 06, 2003, 06:08:51 pm »

Quote
Yes I did. You tried to say that F stop and DoF considerations do not apply across formats. When I challenged you on this point, you then wrote that it only applied to exposures and it was nuts to pretend that a constant DoF across formats was relevant because people do not take photos of bits and pieces but whole compositions (or words to that effect).
I did not.  I said that fixing f-stops while varying focal lengths across formats was invalid if you were trying to compare DOF of the two formats.  In fact, what I said was that the resultant decrease in physical aperture (due to fixing f-stop) was the cause of the improved DOF, not the smaller sensor size itself.  You did not challenge this statement.    The rest of your quote just proves that you didn't understand what I was saying at all.

Quote
What is nuts or irrelevant, or what constitutes a bit or a piece or a whole composition, is a matter of opinion; a matter of taste and a matter of esthetics. I'm trying to cut through your obfuscation with some facts.

That a 300mm lens at a given f stop will have the same size aperture and the same DoF (from the same shooting position)[/I]irrespective of the format the lens was designed for is surely of some academic interest to those who didn't already know it, and is surely of more than academic interest to owners of APS format DSLRs who didn't already know it.
I'm not sure how this ties into the discussion, but it's true that if you fix the focal length and f-stop, DOF will be the same regardless of format given identical CoC.

Quote
You seem to be looking for direct causes.
Of course I am!  What directly controls DOF is all that matters in this discussion.  Sensor size does not determine DOF.  The sensors themselves only record the image rendered on them!  DOF is in the optics (though sensor size effects magnification for a desired perspective).

Quote
Are you trying to say that only direct causes are valid? A lot of things that happen are a consequence of a consequence of a consequence etc. Exponents of Chaos Theory will have us believe that a butterfly flapping its wings in South America might just tip the balance and cause a storm in Australia (I wish it would. It's still pretty dry over here.) This is an extreme example and I don't take it seriously. That reducing sensor size might (and certainly will if you keep the f stop and FoV the same) result in an increase in DoF has a much more direct causality.
Look who's obfuscating now!

Quote
By the way, it's difficult to hide a phrase like "same f stop" in such a short sentence. I don't know how I managed that.

If the educated photographic community were to support the notion that smaller formats offered inherently better DOF only because it's indirectly connected or observed to be true in typical usage, then such incorrect assumptions would lead to the belief that small formats are superior for macro (untrue) or small formats are unable to limit DOF (untrue).  How specific examples of a format perforrn is irrelevant to the inherent capabilities of that format.
Logged

Craig Jones

  • Guest
New Leica Digicam
« Reply #52 on: December 07, 2003, 10:11:16 am »

Alright Ray, I'll enlighten you assuming it's possible.
Quote
Quote
I did not.  I said that fixing f-stops while varying focal lengths across formats was invalid if you were trying to compare DOF of the two formats........  
This reminds me a bit of the sorts of arguments I used to have with my ex wife. "But you said .... "No I didn't, I said...." and so on. I wish I'd had a tape recorder running continuously.  :)
We do have a recorder running. but you chose to paraphrase (incorrectly) instead of quoting since it suited your purpose.

Quote
Fortunately on this forum we have a record of everything written. This is exactly what you wrote:-

Quote
You need to understand that holding the f-number constant across formats is meaningless and bogus. Once you figure that out, you'll realize the DOF claim is wrong.


Thank's for removing that sentence from its context, Ray. That statement was most clearly about DOF, not exposure, and was entirely correct.  Interesting that you can find a quote now since it suits you.

Quote
This statement is clearly, definitely and demonstratively wrong.

Holding the f-number constant across formats is neither meaningless nor bogus. The physics that determines that DoF will remain the same for a lens of a given focal length and given f stop irrespective of camera format, is the same physics that determines that an increase in DoF will result if the focal length is shortened and the f stop kept constant. Every photographer should know this.
Please Ray.  You have no intention of holding focal length constant (or anything else for that matter).  The thrust of your endless arguments is that you reserve the right to vary every parameter as it suites you, then claim that small sensors have a DOF advantage.  Why don't you settle down and pick a single, unchanging, meaningful test so that progress can be made?  Are you afraid of the consequences?

As I've said many times, you cannot change sensor size and focal length (leaving f-stop constant) then claim that the DOF change is due to the sensor size.  It is due to the change in physical aperture and change in the acceptable parameters of the shot which may not be obvious to some people.  If you want to discuss DOF, then talk aperture.  If you want to talk exposure, then talk f-stop.  The two are not the same, and your dogged commitment to keeping them confused is just an attempt to skirt the issue.

Quote
It is also sometimes of practical significance to LF photographers who use a shift mechanism for architecture shots. As I'm sure you already know, to use tilt and shift requires that the image circle is significantly larger than the film format. In order to achieve this larger image circle, LF photographers will sometimes use a lens designed for a larger format (say 8x10) with  a smaller format body (say 4x5). They have to be aware that in doing this, DoF does not change for the same f stop. This is not meaningless or bogus.

Now you might argue that this is obvious, and I would agree that it is obvious to those who have thought about it and already know it. But it's not necessarily obvious to those who have not thought about it.

In any case, I wasn't aware this discussion was about the relative transparency of various statements. None of my statements in this thread are incorrect (as far as I know. If any are, please enlighten me  :)  )

As I recall (and there's a accurate record to back me up), it was you who first accused me of trying to "perpetrate nonsense" then accused me of deliberately obfuscating the issues.  Now all you offer is specious comments about chaos theory and specialized lenses, as if any of that is necessary or valuable to the argument.  You are attempting to shift the argument to a discussion lenses of a fixed focal length even though I've never mentioned it and it has nothing to do with the original argument.  It is something you can get right, though.  Thanks for the clarification.

If you aren't enlightened at this point, I'd suggest you take on BJL's proposal.  It is thoughtful and reasonable.  It is curious that you believe that your artistry is forced to change because you vision of a shot is hopelessly tied to the format.  I believe in reality that the conclusion is inescapable once you agree to play the game.
Logged

victoraberdeen

  • Full Member
  • ***
  • Offline Offline
  • Posts: 205
    • http://www.abovo-media.com
New Leica Digicam
« Reply #53 on: December 08, 2003, 02:20:16 am »

Thank you Leon, some sanity at last.  If you review the article notes on technical aspects of photography I posted earlier, I you will find your reference to Circle of Confusion. Of course we may both be missing the point, it maybe just about size
Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
New Leica Digicam
« Reply #54 on: December 09, 2003, 02:59:40 pm »

Quote
... Enter a nice gentleman like yourself who hands me a Leica Digicam; do I now use F2 at 1/250th? Almost certainly not. I might well stick with the same f stop and shutter speed, thus lifting those constraints, achieving greater DoF and a more appropriate shutter speed for a hand held shot with the shorter lens.
Thank you Ray; your example shows that in this situation, the smaller format probably gives you more flexibility; in this case by making it easier to achieve more depth of field in a limited light situation. This corroborates my comments about smaller formats having an advantage in low light situations. You are almost saying that with the larger format, you might like to get more DoF than F/5.6 gives you, but are forced to a less satisfactory compromise by the limitations of that equipment (superable if you had sufficiently fast film or sensor.)

   And as you say, the struggle to achieve more DoF is probably far more common than the struggle to achieve less, so to the extent that this is true, smaller format cameras might have some advantage (but at the cost of lower image quality in other ways).

P. S. Ray, I just read your further comment on this after replying to your earlier post; maybe we are now all agreed that smaller formats are likely to be better off, or at least not worse off, in typical low light situations?
Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
New Leica Digicam
« Reply #55 on: December 10, 2003, 03:47:46 pm »

Quote
Quote
To what important photographic dimension does "35mm" refer to?

The width of the film.

    Bruce
Right of course: the total film roll width, including space for sprocket holes and maybe the movie sound-track. For obscurity and irrelevance to digital photography, this is on a par with the "diameter of a vidicon tube" explanation for sensor dimensions; which makes me wonder why I so often read people addressing scorn at the latter and none at the former. Since this happens a lot in the US too, it is not just the use of inches!

P. S. On the subject of strange dimensions, there is also Canon saying things like "The EOS 300D features an APS-C sized sensor". Actually APS-C is 25.5x17mm, not the 300D's 22.7x15.1mm, so going from the vague APS to a name that refers to specific, wrong, dimensions puzzles me.
Logged

victoraberdeen

  • Full Member
  • ***
  • Offline Offline
  • Posts: 205
    • http://www.abovo-media.com
New Leica Digicam
« Reply #56 on: December 11, 2003, 07:30:50 pm »

This 2/3" is because electrical engineers did it that way! It is odd 11mm is almost half an inch at 25.4mm where as 2/3 would be 16.93mm.

This summs it up ->  DPReview - Making (some) sense out of sensor sizes
Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
New Leica Digicam
« Reply #57 on: December 01, 2003, 05:16:39 pm »

OK, I will play the devils's advocate, and suggest that the Leica Digilux 2 (aka the Pansonic something-or-other) might at least be a succesful "concept camera" for a certain photographic niche. Like a concept car, it will not necessarily sell in large quantities though, or at all.
   Apart from the appeal to some of us of a compact digital camera that allows one to control focus, aperture and focal length with rotating wheels on the lens barrel, the big debate will be over sensor and pixel size, and the EVF. I will leave the latter to another post.

   The key here is noting that smaller formats can typically be used at lower f-stop's, and hence at lower ISO settings, off-setting their weakness when compared at equal ISO, and that bigger sensors can only provide their potentially greater image quality when provided with enough light to avoid underexposure.
Larger sensors can give higher image quality when provided with enough light, but an "information theoretic" approach suggests that with a given, limited amount of light, any sensor big enough to take in all the available light without highlight blow-outs, combined with a fast enough lens, can give "image detail" as good as a larger sensor, which has to be unerexposed with that same light levels. Below a certain moderate level of total available light, even 2/3" format can roughly match any larger format. However the lens speed of the Digilux 2 is not enough to fully take advantage of this: I dream of a future f/1 zoom lens in this very small format.

   My measure of "detail" is roughly resolution (pixel count) times tonal detail at each location (S/N ratio). If the same total amount of light is gathered (same exposure duration, same aperture diameter and hence f-stop in proportion to focal length and sensor size), sensors of different sizes with the same number of photosites will each get the same amount of light per site, but the larger sites will have more thermal noise, so worse S/N ratio. If the larger sensor instead has more photosites of the same size as the smaller sensor, each photosite gets less light, so again worse S/N ratio, balanced against higher resolution. Downsampling to the resolution of the smaller sensor will return one roughly to the situation of both sensors having the same photosite count; the smaller one still wins on noise.
   For the Digilux 2 used at its maximum of f/2 with a shutter speed of 1/100, the sensor can handle all available light at about EV 9 or below (6 stops or more below bright sunlight), rising to EV 10 if 1/200 is needed and so on, so for a larger sensor to have an advantage in light this low or with shutter speeds at least this fast needed, it can only do it by going to larger aperture diameters and hence less DOF; as far as DOF, the equivalent f-stop is about f/5.6 with APS DSLR's. Allow an APS DSLR a zoom lens going down to f/2.8, and the threshold where the Digilux 2 is at no disadvantage goes down to EV 7 at 1/100, EV 8 at 1/200 and so on. Optimal f-stop for image quality seems to be smaller; around f/11 in APS format, and using that and the Digilux 2 at the equivalent f/4 pushes the break-even to about EV 11, just four stops below bright sunlight.

   As to the "concept camera" idea, the potential is that (a) the maximum usable light level per pixel might be increased by several stops by ideal like latest Fuji's SuperCCD design, which would increase by several stops the light level needed before a larger format has any advantage ( in the very small 2/3" format, a f/1 lens should be no harder to make than f/4 for comparable FOV in 35mm format, and that would add two stops of high shutter speed potential © the resolution can be increased; Sony has already announced 8MP in 2/3" format, and that is more resolution than the overwhelming majority of prints need (being significantly more than given by any APS format DSLR!)

   Finally, do not be mislead by the rather common claim that more, smaller pixels on a sensor of the same size must give worse noise in an image: smaller pixels in the same quantity on a smaller sensor do that, but increasing pixel count in proportion to pixel size roughly balances out the noise level seen in a print of a given size, and also balances any apparent loss in fineness of tonal gradations and range, all through the extra "dithering" or "averaging" effect of printing more, smaller pixels. A print can have far wider dynamic range than the individual pixels: consider half tone printing where the many tiny printed pixels are pure black or white, or consider the pieces of black metalic silver on a traditional print as "1-bit pixels".
   In summary, I think that there is a reason why sensor makers are moving almost uniformly towards more, smaller pixels despite the horror often expressed on the internet at this trend, and I think this is the explanation.
Logged

Craig Jones

  • Guest
New Leica Digicam
« Reply #58 on: December 04, 2003, 05:35:24 pm »

Quote
Quote
Quote
However, if you want to maintain the same angle of view from the same shooting position, DoF increases, for any given f stop, as the sensor size decreases.
That's because when you adjusted the focal length, you changed the physical aperture to keep the f-number constant.  DOF didn't increase because of the smaller sensor, it increased because you closed the physical aperture!
Um yeah, true. What is your point?

I don't get why people argue this over and over and over again. Skip all the math, skip all the theorizing. Try this:

Take a small sensor camera. Make it a P&S, a 35mm, what ever. Select a 3D scene. Select an exposure; lets say 1/250 @ F8 @ ISO 100. Take picture.

Now take a 4x5. Insert ISO 100 Film (or a digital back if you are quite well off). Select focal length for the same composition from the same position. Select the same exposure. Do not use movements. Take picture.

Print both at the same size; lets say 8x10. Visuallly inspect. Do they have the same DoF? I can assure you it is not even in the same ballpark. The DoF in the 4x5 will be very very very thin. The P&S DoF will be very good.

Now argue all you want on why, but it still holds in the real world: DoF EFFECTIVELY depends on the sensor size. You can blame it on focal length if you want, you can point at this variable or that changed, argue to you are blue in the face, but you will still have more DoF for your desired composition on the small sensor camera.
The point is that when people skip all the math and skip all the theorizing, they end up concluding something stupid like DOF depends on sensor size when it does not.

In your example, the improved DOF you see is due to the smaller sensor using a much smaller physical aperture.  It is not caused by shorter focal length!

It does not hold true in the real world that DOF effectively depends on sensor size.  It depends on magnification and aperture.  To fix your example, if I shot the scene with 35mm f/8 and 1/250, then shot the same scene with with the 4x5 and appropriate lens at f/28 and 1/60, the DOF for the 4x5 would be marginally better despite the fact that the aperture was identical.

You need to understand that holding the f-number constant across formats is meaningless and bogus.  Once you figure that out, you'll realize the DOF claim is wrong.

Small sensor cameras require less light to expose them so they get by with smaller physical apertures and deliver correspondingly good DOF.  If it were a 5050 you were shooting in the above example, f/8 would be the smallest aperture available.  Why?  It's diffraction limited at f/8!  Shooting a 5050 at f/8 is like shooting your FF 35mm SLR at f/32 all the time (except that your 35mm would provide better DOF).

Now let's talk macro.  With macro you are typically trying to maximize DOF (or at least make sure it's enough).  If you are suggesting that a 5050 provides better DOF for a macro composition that FF 35mm you are sorely mistaken.  At higher magnifications especially, 35mm will trounce the 5050.  Do the math or better yet, take the pictures and you will see.  Do that experiment, Paul, it might make you blue in the face.
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
New Leica Digicam
« Reply #59 on: December 05, 2003, 08:44:27 am »

Craig,
You're not telling me anything I don't know. I don't understand where you got the idea that I think F stop is synonymous with aperture diameter. What you are doing is engaging in obfuscation, perhaps because you don't like the method by which I've explained things; perhaps just to show off your knowledge.

Nothing you've written so far provides any evidence or reason why what I've written on this subject is wrong. What you seem to have done is draw an incorrect inference from some of my statements and are now arguing against that incorrect inference. I can't be responsible for people misunderstanding what I say. All I can do is try to be as clear as possible.

So let's go back to what I actually wrote.

Quote
However, if you want to maintain the same angle of view from the same shooting position, DoF increases, for any given f stop, as the sensor size decreases.

This statement is absolutely true. I can't fault it. I have not said that sensor size determines DoF. If you think that's what I meant, then it's you who has made an incorrect inference. DoF is never determined by one factor, not f stop alone, not physical aperture alone not focal length alone not sensor size alone. But sensor size is[/i] going to determine the focal length of the lens required for a certain composition shot from a certain position.

Perhaps I could have provided a fuller explanation. Isn't this always the case? I could have tacked on the obvious explanation that the reason DoF increases, at a given f stop, as sensor size decreases, is that to maintain the same field of view it's necessary to use an increasingly shorter focal length lens. There's a relationship between sensor size, lens focal length and Field of View. There's also a relationship between focal length and physical aperture size. It's called F stop.

I really don't know what you are arguing about.
Logged
Pages: 1 2 [3] 4   Go Up