Luminous Landscape Forum
Equipment & Techniques => Digital Cameras & Shooting Techniques => Topic started by: howard smith on July 21, 2005, 04:36:24 pm
-
Don, lot's of info in that question. I am an engineer (nuclear, not optical) but Michael Reichmann says, on this site, that:
"There was a query in October, 2001 on my Discussion Forum as to whether Depth of Field was calculated any differently for digital Vs. film. The answer is, no. There is no difference whosesoever. DOF doesn't care about the recording media type or size, ... ."
-
I confused depth of field and depth of focus. Sorry.
-
Here we go again...
What you're asking about is really depth of field, not depth of focus. Digital sensors are far flatter than film ever was, so depth of focus is a complete non-issue unless you're trying to calculate how sharply focused sensor dust will be at f/22.
While DOF is calculated using the same formula for film and digital cameras of all formats, (which is what Michael is referring to in the quote you cited) the variables inserted into the formula do vary depending on camera format and whether it is film or digital. Specifically, the appropriate circle of confusion value varies significantly, even between the various film formats. For digital cameras, the most appropriate circle of confusion (CoC) value is the pixel pitch of the sensor (the center-to-center distance from one pixel to the next). Since digital sensors record much more detail per unit of area than film, if you use a film-derived CoC value when calculating DOF with a digital camera, you will be unpleasantly surprised to find DOF much narrower than you anticipated. The spreadsheet available here (http://www.visual-vacations.com/Photography/dof_calculation.htm) will help you accurately predict the DOF you will get when shooting with a digital camera. Simply input the physical height of the sensor and the vertical sensor resolution in pixels, and the sheet will calculate the correct CoC value. Then input focal length and distance, and DOF for apertures from f/1 to f/64 will be displayed. You'll find that the sheet recommends using narrower apertures than you're used to using with film, but the spreadsheet tracks very accurately to real-world results. It's one of the things that necessarily accompanies using a 35mm-format camera for tasks that could only be done with medium format when shooting film.
Jonathan - thanks, I think you have come closer than others in responding to my inquiry, but, as far as I understand DOF=depth of field, and what I am asking about is more to do with the receptor of the light, being it the sensor or film. You did note that with digital sensors you do not get all the available "depth" that you would under the same circumstances with film. I am trying to say that I have seen this in action, and do not feel it is just less depth of field per se, as that is driven more by the optic's focal length and aperture used and subject distance. What I am trying to refer to is the ability to capture this range of focus at the film or sensor plane. In effect the sensor is more 2 dimensional, less depth sensitive than film. Film has an inherent thickness to the emulsion and can thus capture or record more of the zone of focus, represented by the ability to focus through the emulsion or use a smaller f stop when enlarging film sourced images and thus get a larger amount of the image in focus in the print. This when compared to digital image taken with same lens and distance etc. I am definitely not talking about DOF, but moreso what happens at the "film plane"
I guess in the end, the result is that you have to use a smaller f stop to compensate for this effect and thus compress the available zone of focus as I call it, so it can be captured on the sensor, but this gives you both a more limited choice and may push using fstops that might yield diffraction issues if pushed to use the smallest fstops.
The remainder of your explanation really has more to do with calculating the correction factor to achieve the similar DOF that could be attained by film in the same case. If you have ever used a 4x5 film camera you are probably more aware of this issue then if you have only shot 35mm.
-
Jonathan - I really think we are saying the same thing, just not agreeing on terminology. I am trying to explain a feeling that the sensor is less able to record focus depth, you say it is irrelevant, as digital sensors do not record as much depth, and are more planar, only record what is at the plane with no depth. To increase this, you stop down to increase depth of field, but in a sense you are collapsing or compressing the amount of what the image at the focal plane is able to record by the sensor. I am saying that this is an "effect" from the smaller fstop, not the cause of the need to do this. I don't know if we will ever achieve parity in this conversation, but bottom line is that you are compensating for this lack of depth to the sensors ability to record information by increasing apparent and real depth of field, but the result of this is to compress the zone of focus so the digital sensor's ability can thus record more information as being sharp. The depth of field the lens projects is a separate issue and is constant for a given scenario as long as the parameters are constant. We are both agreeing that the sensor has less ability to record depth for a given scene. :D
-
Don:
By Depth-of-Focus, do you mean sharpness? I'm trying to understand what you have experienced. I see you are NOT talking about depth-of-field. It seems you are talking about a completely different quality of the image--maybe even the printed image as opposed to the one on the plane of the medium. Could you try to explain what you see again?
Steve
-
Yes, I think I understand you, but you're still off the mark.
The out of focus parts that are still captured, are captured by neighbouring sensors rather than an emulsion layer in front or in the back of the "in-focus" layer (to the extent that such a thing exists).
Jonathan seems to be quite correct in how he explains it, I think that perhaps you need to read over his responses one or two more times.
-
Let me repeat: Sensor thickness has absolutely zero correlation to its ability to extract information from photons striking it. None. Zero. Zilch. Nada. ALL of the light striking the sensor is captured and processed and converted to bits in that very thin surface layer. There is absolutely no reason whatsoever for there to be anything below the surface layer to capture something the top layer of the sensor missed, because the top layer didn't miss anything, it got it all. There is absolutely no need whatsoever for a sensor to have significant 3-dimensionality when all you're capturing is a 2-dimensional image. Unlike a darkroom enlargement, there is no need whatsoever for light to pass through the film or sensor when exposing the image initially. Even if it isn't in perfect focus, the light will still hit the sensor and be converted to bits; all it has to do is hit the sensor anywhere and it will be recorded as part of the image. Therefore there is no reason to be concerned with depth of focus unless you're trying to calculate how far to tilt a lens to get the proper Scheimpflug focus shifting effect.
-
I understand that the sensor records or captures "all" the photons hitting it, the issue, and I will admit it is mine, is that it, of course, doesn't record bits of the image that are out of focus at the plane, but are in focus 0.1mm behind or in front of the plane, as being in "sharp focus". They are recorded as out of focus elements only.
At the risk of sounding rude, duhhh! Of course that happens, but making the light-sensitive part of the sensor thicker will not enhance the overall ability of the sensor to record a sharp image. Quite the opposite, if you made the active, light-sensitive part of the sensor 2mm thick, it would guarantee that no matter where you focused, any given image element would be out-of-focus somewhere in the sensor. And every pixel therefore would be a mix of in-focus and out-of-focus light rays. If the light rays converged 1mm below the surface of the sensor, the photons captured there would contribute to the image being in-focus, but photons captured at the surface or 2mm below the surface would contribute to the image being significantly OOF. Therefore, it would be impossible to achieve overall sharp focus no matter how you adjusted things. And that would be totally pointless and futile.
On the other hand, with a Bayer-type sensor which is optically flat and all the photon capturing and conversion takes place in the first few microns of the sensor surface, one can set the focus to coincide with the surface of the sensor and get all of the captured photons to contribute to a sharply-focused image. There is no advantage whatsoever to giving a sensor any more "depth" than absolutely necessary, as doing so will only impose an unnecessary limit to how sharply the sensor can resolve image details.
Film offers no mechanism to magically capture and record only the photons that contribute to a sharply-focused image and discard the rest. While it would be theoretically possible to stack a bunch of sensors on top of each other to make a pile 1mm deep or something and have each capture a portion of the incoming photons and combine their outputs together to get one "super image", it would be prohibitively expensive for only moderate gains in image quality.
I guess I will just live with the phenomenon and compensate by increasing f stop, which is the cure for the issue up to a point...
That is the only solution. You're pulling more detail out of every square mm of silicon than you ever could with film. One of the unavoidable side effects of that is a more stringent definition of what is "in focus". So you must stop down more to accommodate the new definition of "in focus". The bottom line is that what you're seeing has no relevance or connection whatsoever to any interaction between sensor thickness and depth of focus, it's due to decreased depth of field as a result of the digital sensor giving you a smaller circle of confusion than you're used to working with when you were shooting film. No more, no less; period, end of sentence, end of discussion.
-
BTW in traditional sense, the term depth of field is used to describe the area in front of the camera and what appears in focus, not what happens inside at the film or sensor plane. IN large format, that was called depth of focus in all the photography books I have read. So most from what I have seen here, many have misused the term totally. If you increase the depth of field by increasing the fstop as shown in my diagram above, it generally is thought of as increasing the depth of the image sharpness range from say 10 ft to infinity to 2 ft to infinity. No where is it used to describe what happens at the image plane in the camera.
I'm well aware of the difference between depth of field and depth of focus. What is incredibly frustrating is your insistence on ascribing everything focus-related to depth of focus and completely ignoring depth of field, and how the two interact. If you have a lens focused at 10 feet, light reflected from an object 20 feet away will converge in front of the film or sensor, resulting in that object appearing out-of-focus. If you want to figure out what aperture you need to render both objects in focus and what focus distance to set, you need to use the depth of field formula with an appropriate circle-of-confusion value (the sensor pixel pitch in the case of digital), not a depth-of-focus formula. The only time you'd need to think about depth-of-focus is if you were trying to calculate how much to tilt the film plane or swing the lens to bring both objects into perfect focus without stopping down. Yet you stobbornly insist on ascribing everything to depth-of-focus; you're looking at things through the wrong end of the telescope.
In fact from what you have better described it is the flatness that indeed causes the difference that I noted. The digital sensors' 2 dimensionality makes it as good as it is, but also makes it impossible to record the image sharpness in the same way as film. It is a difference, not a plus or minus per se. You may never see this, so be it. IF you never shot LF and looked at the depth I describe working in a darkroom, then you have little to no basis for commenting on what I have attempted to describe.
Your insistence on using depth of focus to describe the effect of aperture when enlarging in the darkroom is incorrect, as in that case the negative is the subject, and by adjusting the enlarger aperture you are adjusting the depth of field such that the unflat film still falls completely within the depth-of-field. Unless you're using "depth" to describe the "look" or appearance of a film image viewed on a light table or projected through an enlarger, which is something else entirely, unrelated to film flatness (or lack thereof), film thickness, depth of field, or depth of focus. Instead, it is related to the total amount of detail in the image and (in the case of digital) post-processing technique imparting a certain "look" to the image that in spite of being 2-D, takes on the appearance of being 3-D. It's achieved by making the image sufficiently clear and artifact-free that looking at a print is like looking out of a window directly at the scene, without any film grain or sharpening halos or any other perceptible artifacts of the image capture and reproduction process. And that can be achieved with either film or digital, but is not necessarily easy with either process.
I accept and trust your knowledge of digital sensor design and operation, it would be nice if you responded to me with the same respect, with out the sarcasm.
It's not sarcasm, it's intense frustration with your complete inability to grasp the concept. Or using "depth" in a way that has absolutely nothing to do with either depth of field or depth of focus. I'm not sure which.
-
I feel your frustration...:-). Truly; we look at terms differently and having coming from a 4x5 background, your use IS different from mine. Let's not try to make this a fight, I am trying to understand your perspective, and wish I felt you were doing the same???? Letting go of the jibes would help. Your tone tends to be sarcastic, whether you intend it or not, one of the blessings of a forum and the web. You have not stated whether you have done any LF work or darkroom work, it might make a difference. I am not referring to running the focus to adjust for film curvature, but focusing up and down through the emulsion itself. This is possible to some extent, and can yield different things into sharp focus. If you haven't done this, you truly don't know what I am talking about. Also using LF, during focusing on the ground glass, there is a range of what is in focus, it is not a black and white thing, you decide and control what is in focus and adjust both aperture and other controls like tilt to accomplish this. It is not, a yes or no thing.
Now I understand as you describe it, that the digital sensor is very 2 dimension like and has no depth of any siginificance. The sensor plane does not move, I understand this. Up until now, we hopefully are in some kind of understanding. You have stated, if I understand you, that the sensor would not recognize anything immediately in front of or behind its position, this too I agree on at this point, and had from the beginning. From the sensors' perspective this information doesn't exist and is why you consider it irrelevant. But from an optical perspective, there would be slightly a different image available in any of these other positions.
Using film, that has more of a thickness to its emulsion, you would also capture some of this detail at different positions within the emulsion. While this is not a huge distance, it is measurable. By using this thickness in film emulsion, it is possible to increase YOUR DOF, or my depth of sharp focus, to show this sharpness in a print. So the final print image is sharper then it would be if you used a wider aperture with less of your DOF, or have you assumed that you only use a smaller aperture to adjust for film flatness? By stopping down an enlarging lense both things really occur, not just the correction for film flatness. I can see how one would assume that since the film flatness is probably several times of magnitude larger, it would be easy to ignore the emulsion thickness issue. Do you understand what I am saying here, or am I "wrong" on this point as well, ???? If we can not agree this happens, then I will just give up further discussion, and tell you to have a nice day...:-).
The digital sensor can only record what is sharp at the plane that it is at, the photons or rays that would have converged a little in front of or behind it, are irrelevant because as you say they do not exist from its perspective. This is also why increasing your DOF works as a repair, because it compresses the focus depth at the sensor plane. More information is sharp at that point. More image is sharp. Am I getting anywhere?? Let me know, I am trying to understand this from my perspective, really.
-
I am not referring to running the focus to adjust for film curvature, but focusing up and down through the emulsion itself. This is possible to some extent, and can yield different things into sharp focus. If you haven't done this, you truly don't know what I am talking about.
That has more to do with the film not being perfectly flat than focusing on different layers in the film emulsion. It you're talking about color film, only certain colors would be in focus, as the different layers of the emulsion respond to different wavelengths of light. If focusing on a specific layer in the emulsion was a relevant factor, then you'd have a negative where objects of one color would be more sharply focused than objects of different colors. I've not shot 4x5, but I did scan film before going digital, and film flatness always had a far more significant impact on the scan than focusing on different layers in the film emulsion.
Now I understand as you describe it, that the digital sensor is very 2 dimension like and has no depth of any siginificance. The sensor plane does not move, I understand this. Up until now, we hopefully are in some kind of understanding. You have stated, if I understand you, that the sensor would not recognize anything immediately in front of or behind its position, this too I agree on at this point, and had from the beginning. From the sensors' perspective this information doesn't exist and is why you consider it irrelevant. But from an optical perspective, there would be slightly a different image available in any of these other positions.
We're in agreement so far.
Using film, that has more of a thickness to its emulsion, you would also capture some of this detail at different positions within the emulsion. While this is not a huge distance, it is measurable. By using this thickness in film emulsion, it is possible to increase YOUR DOF, or my depth of sharp focus, to show this sharpness in a print.
Here's where you're going off base. Here's why: Assuming what you say is true (which I doubt, due to the color-specificity of the emulsion layers involved) about a meaningful focus differential being achievable within the thickness of the film emulsion, your ability to extract that information optically is going to be hampered by the fact that the light from your scan or enlargement has to pass through the entire thickness of the film, not just the part that is "in focus".
To illustrate, let's conduct this mental experiment: let's load a camera with three layers of film, each of which has a perfectly clear backing, and is 1/3 the thickness of normal film. Assume perfect film flatness, and that keeping the 3 layers of film in perfectly correct registry after development is possible. Expose this "sandwich" so that focus is centered on the middle layer of film. Now you have three individual exposures with a very slight focus bracketing. At this point, you're saying that these three sheets of film have some intrinsic collective property that results in a better-quality image than any one sheet of film by itself. And that is the fatal flaw in your reasoning. Here's why:
Even if you're focusing precisely on the middle sheet of film during exposure and enlargement, the top and bottom sheets of film will still be contributing an out-of-focusness to the enlargement; they are two layers of film that were out-of-focus during exposure and are now out-of-focus again during the enlargement. That means that the projected image will be a combination of of the sharply-focused center film sheet and the double out-of-focus top and bottom sheets, with the double-out-of-focus sheets dominating the final result by a ratio of 2:1 over the contribution of the properly focused film sheet. And if you focused on the top or bottom sheet of fim during exposure and enlargement instead of the middle sheet, the situation gets even worse; instead of having 2 slightly out-of-focus sheets of film, you now have one slightly out-of-focus sheet of film and one really out-of-focus sheet of film, and you still only have one properly-focused sheet. As long as you're enlarging or scanning all three sheets of film together, there's no way to avoid this, and there's no way it can possibly improve overall image quality in any meaningful way. On the other hand, if you remove and discard the top and bottom sheets of film, that will make a significant improvement in overall image quality, as long as you manage to properly focus on the single sheet of film.
So the final print image is sharper then it would be if you used a wider aperture with less of your DOF, or have you assumed that you only use a smaller aperture to adjust for film flatness? By stopping down an enlarging lense both things really occur, not just the correction for film flatness. I can see how one would assume that since the film flatness is probably several times of magnitude larger, it would be easy to ignore the emulsion thickness issue. Do you understand what I am saying here, or am I "wrong" on this point as well, ????
I understand exactly what you are saying, I just don't buy any of it, for the reasons I've explained above in the thought experiment, as well as the following:
1. The film base is much thicker than the emulsion layer(s) for any given film stock, (95 microns/base vs 25 microns/emulsion for Agfa RSX 50 slide film in 120 format, as an example) so I don't believe the emulsion thickness is significant enough to be relevant to depth of focus or overall image quality in real-world imaging, especially given that
2. Film flatness is a problem during exposure, enlarging, and scanning, and is of much greater magnitude than overall film thickness or emulsion thickness.
3. If your premise regarding "focusing within the emulsion" had any validity, then enlarged images would render objects of one color markedly sharper than objects of another color, and one could "tune" a scan or darkroom print for best sharpness for objects of one color or another with a simple focus adjustment. And one would also have to factor in the order of color emulsion layers within the film stock being used when deciding how to focus for a given scene with objects of assorted colors.
Given that, I see no reason to conclude anything but that the thicker imaging layer of film vs a digital sensor is either irrelevant (because it's not thick enough to make a meaningful difference in focus and is completely overwhelmed by film flatness issues anyway) or is a distinct disadvantage for all the reasons I outlined in my thought experiment above. I believe that the focus effect you're seeing is primarily caused by a lack of film flatness, not "focusing within the emulsion", and that far from contributing to image quality in any meaningful way, it detracts from it.
To make a fairly bad analogy, you need to stop down some more with digital because because with digital you're only using the "middle sheet of film". You can get better resolution and clarity as a result, but you need to be more precise with your focusing and stringent with your definition of "in focus" (AKA a smaller circle of confusion) to maximize what the sensor is capable of.
-
I have to agree with Jonathan's position on this issue. There is theoretically an extremely slight focussing latitude in relation to different layers of film emulsion which are at a microsopically different distance to the lens. However, considering the numerous and much larger factors that contribute to out-of-focus effects, I would expect such minute differences, that Don thinks he is seeing, to be totally irrelevant within the larger context.
Such differences that he sees between digital and film must have another explanation.
-
The relation between depth of field and depth of focus would seem to indicate that depth of focus is typically far greater than the thickness of a film emulsion layer (approximately 7 microns for B&W films and 20-24 microns for color films), as does a bit of carefully applied logic:
In order to maintain focus within a perfectly flat 4x5 piece of Agfapan APX 100 (http://www.adorama.com/AGAP100120.html) (7 micrometer emulsion thickness) the lens would have to be perpendicular to the film plane within 0.0063 degrees. In actual practice, structural deformations of the camera body due to the weight of the lens, focus changes, breathing in the vicinity of the camera, or probably even looking at it funny would greatly exceed this, not to mention the practical impossibility of adjusting focus on a 4x5 view camera to a tolerance of 7 micrometers. That's far smaller than the texturing of the groundglass, meaning adjusting focus that precisely would have to be accomplished with a laser interferometry-controlled servo or some similar method, rather than by simply looking at the groundglass image and adjusting the focus control by hand.
A 1-degree tilt would result in a 1108 micron lens-to-subject change 2.5 inches from the center of the frame. That's 158 emulsion thicknesses for a measly 1-degree tilt, which doesn't affect the image all that much in real-world terms; it's barely noticeable in many cases. If a 158-emulsion-thickness focus adjustment is not particularly significant, then differentially focusing within the emulsion layer during exposure is impossible; all the sub-layers of the emulsion will be exposed to and record essentially identical images, meaning that even if focusing within the emulsion is possible during enlarging (a proposition I highly doubt) it is of no practical benefit because the same image is recorded at all levels of the emulsion.
At this point I think it is safe to say that claiming that differential focus within the emulsion layer during exposure is of any practical significance is the photographic equivalent of polywater (http://en.wikipedia.org/wiki/Polywater). And if it is not possible during exposure then any differential focus effects observed during scanning or enlargement can be safely ascribed to irregularities in film flatness.
-
And I am still not 100% sure that if the lenses are the same basically and DOF is DOF, why any adjustment is needed with digital compared to film. Was it the circle of confusion with film that compensated for better by having any film curvature alone. Sort of like there being two areas of imperfection that kind of complemented each other and thus resulted in a image of greater sharpness with a relatively wider aperture setting???
Depth of field, in its simplest form, boils down to this: is the sharpness and clarity of a given image element comparable to the most sharply-rendered element in the image? If the answer is yes, then that image element falls within the depth of field, and if the answer is no, then it falls outside it. There is a matter of degree involved, the transition from "in focus" to "out of focus" is smooth and gradual, not a simple binary yes/no thing. But the most important thing to keep in mind is that such comparisons are relative, and when you increase the sharpness and clarity of the sharpest and clearest element of the image, the comparison between it and any other arbitrary image element is more likely to fail and the other image element is more likely to be deemed "out of focus" than before. That is why digital images have narrower DOF than film images, even if the exact same lens, aperture, and focus distance setting are used.
Due to film grain and other factors, current digital sensors can resolve more detail per square millimeter of surface area than film. That means that your working circle of confusion is significantly smaller when shooting digital than when shooting film of the same format. If you look at any DOF calculator, you'll find that if you reduce the circle of confusion parameter, you have to reduce the physical aperture diameter by a corresponding degree to maintain the same DOF. If the circle of confusion diameter is halved, instead of f/4, you'd need to stop down to f/8 to maintain the same DOF. But the benefit you gain in exchange for this is increased overall resolution and image detail; in this instance it would be 4x overall.
Try this experiment:
Take an image file, and make two copies of it. Apply a 1-pixel Gaussian blur to the first copy, and a 3-pixel blur to the second copy. You'll find that the blurrier the image gets, the wider DOF becomes because the comparison between the most sharply focused image element and other elements is more easily judged to be equivalent the blurrier the most sharply-rendered image element becomes.
-
This is a question that relates to the ability of digital sensors to register the same depth of focus as film or not. I have know owned about 6 dslrs and about to upgrade to a 1d level camera. I think I have noted a difference in all cameras compared to film cameras and am trying to understand this. When looking at film, say in a darkroom and enlarging and making a print or projecting a slide, it is apparent that there is a "depth" to the focus within the emulsion. BY using a smaller aperture on the enlarger or running the focus on the slide differing amounts of "in focus" image can be visualized. It seems to me that digital slrs register less depth of focus then film. It does not relate to format or any given fstop that was used in shooting. I would say I have gotten my sharpest images with each camera with wideangle lenses, that would compress the depth of focus compared to longer lenses as it should be, but still doesn't seem to register the same depth as film would. I have compensated at times by using a smaller f stop then I would have had I shot on film. Are the digital sensors more 2 dimensional than film in this regard, only sensitive to the depth of focus to a smaller degree? Which might explain the phenomenon. By using a smaller f stop like 16 0r 22, this would also tend to compress the depth of focus, if I understand optics at all. Any optical engineers out there could elaborate, but please use simple terms or analogies, as I am most definitely not an engineer.
-
Here we go again...
What you're asking about is really depth of field, not depth of focus. Digital sensors are far flatter than film ever was, so depth of focus is a complete non-issue unless you're trying to calculate how sharply focused sensor dust will be at f/22.
While DOF is calculated using the same formula for film and digital cameras of all formats, (which is what Michael is referring to in the quote you cited) the variables inserted into the formula do vary depending on camera format and whether it is film or digital. Specifically, the appropriate circle of confusion value varies significantly, even between the various film formats. For digital cameras, the most appropriate circle of confusion (CoC) value is the pixel pitch of the sensor (the center-to-center distance from one pixel to the next). Since digital sensors record much more detail per unit of area than film, if you use a film-derived CoC value when calculating DOF with a digital camera, you will be unpleasantly surprised to find DOF much narrower than you anticipated. The spreadsheet available here (http://www.visual-vacations.com/Photography/dof_calculation.htm) will help you accurately predict the DOF you will get when shooting with a digital camera. Simply input the physical height of the sensor and the vertical sensor resolution in pixels, and the sheet will calculate the correct CoC value. Then input focal length and distance, and DOF for apertures from f/1 to f/64 will be displayed. You'll find that the sheet recommends using narrower apertures than you're used to using with film, but the spreadsheet tracks very accurately to real-world results. It's one of the things that necessarily accompanies using a 35mm-format camera for tasks that could only be done with medium format when shooting film.
-
This has gravitated towards Depth of Confusion.
[or was that Circle of Field?]
-
Jonathan - thanks, I think you have come closer than others in responding to my inquiry, but, as far as I understand DOF=depth of field, and what I am asking about is more to do with the receptor of the light, being it the sensor or film. You did note that with digital sensors you do not get all the available "depth" that you would under the same circumstances with film.
Unless you're using a lens with tilt/swing movements, depth of focus has absolutely zero relevance to image quality when shooting digital. Bayer-pattern sensors absorb light and do what they do in a very thin layer on their surface; there is no need to "focus through" their surface for any reason, because the light-to-bits action all happens right at the surface of the sensor. (Foveon sensors have multiple layers similar to film, but they are still orders of magnitude thinner than film.) Depth of focus is only relevant if the recording medium has significant depth, which digital sensors do not have or need. Think of it this way: with digital, the depth of focus needed to record a maximum-sharpness image is very small compared to the depth of focus needed to record a comparable film image, because the digital sensor has a much thinner and flatter light-sensitive zone.
I am trying to say that I have seen this in action, and do not feel it is just less depth of field per se, as that is driven more by the optic's focal length and aperture used and subject distance. What I am trying to refer to is the ability to capture this range of focus at the film or sensor plane. In effect the sensor is more 2 dimensional, less depth sensitive than film. Film has an inherent thickness to the emulsion and can thus capture or record more of the zone of focus, represented by the ability to focus through the emulsion or use a smaller f stop when enlarging film sourced images and thus get a larger amount of the image in focus in the print.
Again, depth of focus is completely irrelevant. The thinner recording depth of digital sensors vs film means that you can get a sharp image from a digital sensor with a much smaller depth of focus than is necessary to maintain uniform sharpness through all the emulsion layers of film, as long as the focused zone coincides with the surface of the sensor. There is no emulsion, there are no layers, only a single flat surface where all the action happens. What you're seeing is 100% related to depth of field, and 0% related to depth of focus.
This when compared to digital image taken with same lens and distance etc. I am definitely not talking about DOF, but moreso what happens at the "film plane"
You are talking about depth of field. Depth of focus is simply not a relevant factor.
-
I don't know if we will ever achieve parity in this conversation, but bottom line is that you are compensating for this lack of depth to the sensors ability to record information by increasing apparent and real depth of field, but the result of this is to compress the zone of focus so the digital sensor's ability can thus record more information as being sharp.
No, what I'm saying is that your "feeling" that a digital sensor's narrow recording depth somehow has something to do with its rendering of depth of field in the image is totally off base and irrelevant. It is NOT a "lack"; that thin, flat sensor surface does everything that the much thicker multilayer film emulsion ever could, and more.
but in a sense you are collapsing or compressing the amount of what the image at the focal plane is able to record by the sensor.
That is the essence of where you are completely wrong. Increasing the thickness of the sensor or film has no relevance whatsoever to the way DOF is rendered in a recorded image. Thinner and flatter is actually better, because you're reducing the size of the zone behind the lens that must be sharply focused to record a sharp image. It is harder to maintain sharp focus through a large range of lens-sensor distances than a small range.
What is relevant is the fact that sensor pixels are more closely spaced than the average size of grain clumps in normally exposed film, which means that a digital sensor has the ability to record more detail per surface area unit than possible with film. This reduces the size of the circle of confusion in effect when shooting digital. As a result, you find that you have to stop down more when shooting digital than when shooting film. It's all about the circle of confusion being smaller than you're used to with film, and that's a depth of field thing. It has nothing to do with "compensating" for the thickness (or lack thereof) of the sensor.
-
Jonathan - if what you say is the only relevant issue, try to explain to me, sorry if I am dense, how such a flat thin sensor can record information that is say 0.1mm [this is possibly an exaggeration in magnitude, I know] behind the actual sensor that would be captured on film emulsion. It can't, can it? but can be adjusted for by increasing depth of field. The resolution of the sensor has nothing to do with this directly, I do not think, it only records the information that is within its range. Just so I am clear, I am not saying a thicker sensor would necessarily accomplish this, the same as film, but I could see how a foveon sensor might close the gap on this issue to some extent, as long as the wavelengths further off the sensor plane are part of that layers sensitivity.
If this information is say behind the sensor, that 0.1mm, how would the sensor see it to incorporate it into the image? Would it? If not, then a thicker sensor which is capable of recording this depth would accomplish this better, no? I am really not trying to be either argumentative or overly dense, just trying to come to terms as to the cause of this apparent lack of DOF in a framework that I can understand. I have used a 4x5 for years and have done darkroom work for over 30 yrs.
-
Steve - I would be glad to try, since I seem to be failing to get my perspective across. When one focuses a large format camera for example or in a dark room, there is a range of focus in the emulsion, i.e. different parts of the image are in focus at the plane of focus at different times as you run the focus. Part of this, is of course due to the curvature how ever minimum of the film itself, but some of it is also due to the thickness of the emulsion, and the captured elements of the image being at different points within the emulsion. I do not know for sure what the distance from one part of the emulsion is from back to front, but there is a depth both to the emulsion and also to the in focus elements at each point.
Emulsion
B F B= back of emulsion, F=front of emulsion
l l
l * l *=point in sharp focus, say at F/2.8
l *l
l* l
l l
Sensor
l l
l*l
l l O
O l l O= point of sharp focus not with in the
l l sensor range, due to thinness of sensor
say at F/2.8
now by changing to a larger f stop, say f/22 this becomes
l l l l
l l*l l
l l*l l
l l*l l
l l l l
I hope this is alittle more understandable?.......
-
I still don't understand how a neighboring pixel can register something behind itself or it's neighboring pixel, can you explain this in a different way, as it makes no sense to me. This is a back to front thing, not a next pixel thing, note back to front on image above. And I have read his responses a few times......this is very frustrating, can you explain how this would be possible??? I do not think we are talking about the same thing....at this point. ???
-
I understand that the sensor records or captures "all" the photons hitting it, the issue, and I will admit it is mine, is that it, of course, doesn't record bits of the image that are out of focus at the plane, but are in focus 0.1mm behind or in front of the plane, as being in "sharp focus". They are recorded as out of focus elements only. This range is a optical property of the lense, not the sensor. It is also this point that relates to different wavelengths recording either slightly in front of, or behind the sensor plane which is the drive to make digital lenses with better convergence qualities. That is my understanding from what I have read and makes sense in an optical point of view. The image projected on to what ever media film or digital sensor is not perfectly 2 dimensional either, as far as I have read. If light was focused by the lens in such a 2 dimensional way, then no improvement would be needed in lens design.
It also seems that I am not the only one confused, if someone thinks that the next pixel over can record info from under adjacent pixels[disregarding bloom effects]. I think to understand what I am saying you have had to have worked in 4x5 or darkroom for a while and critically looked at focusing through the emulsion to experience such depth. Also in the strict sense of it, depth of field, if you use the term properly is what is seen in front of the camera as depth of the scene that is recorded in focus, and is not an black or white thing, but is a gradual process from going in and out of focus with in the range. If you have used only an 35mm slr and never looked at the ground glass or used an enlarger, you may never have experienced this depth, esp. in an af world. I guess I will just live with the phenomenon and compensate by increasing f stop, which is the cure for the issue up to a point...
-
Jonathan - first thanks for the better explanation, and also for the lack of sarcasm, "DUHH", totally uncalled for in a professional forum. At least I thought it was one, until your response. BTW - I NEVER said I thought film was better or had an advantage per se, only trying to understand the difference. BTW in traditional sense, the term depth of field is used to describe the area in front of the camera and what appears in focus, not what happens inside at the film or sensor plane. IN large format, that was called depth of focus in all the photography books I have read. So most from what I have seen here, many have misused the term totally. If you increase the depth of field by increasing the fstop as shown in my diagram above, it generally is thought of as increasing the depth of the image sharpness range from say 10 ft to infinity to 2 ft to infinity. No where is it used to describe what happens at the image plane in the camera.
In fact from what you have better described it is the flatness that indeed causes the difference that I noted. The digital sensors' 2 dimensionality makes it as good as it is, but also makes it impossible to record the image sharpness in the same way as film. It is a difference, not a plus or minus per se. You may never see this, so be it. IF you never shot LF and looked at the depth I describe working in a darkroom, then you have little to no basis for commenting on what I have attempted to describe. I accept and trust your knowledge of digital sensor design and operation, it would be nice if you responded to me with the same respect, with out the sarcasm.
From your explanation, I know understand if fact, that it is the planar nature of the sensor that does limit some of the depth, if you disagree so be it.
Lastly the discussion is OVER, when I understand the effect or when no one further responds or when a moderator closes the thread, not when you say it is. Geez, what an ego....
-
FIRST -
I have no idea about advanced optical theories and practice
SECOND -
It seems to me that if all the information in this thread is correct then what Don is saying is that there are multiple planes of focus in film. Once put on an enlarger, you can adjust the focus of the enlarger's lens and "focus through these depths". Similar to looking at a slide in a microscope. This also seems to be in my mind how one can get some more detail from a RAW digital exposure. The information has a little bit of wiggle room to still be acceptable.
THIRD -
I am under the impression that there is only on true plane of focus and therefore if you get that plane in alignment with a digital sensor you'll have the best focus one could hope for. This would support Jonathan's points. This means that the area of "sharp focus" that Don used in his diagram isn't real, just perceived.
If I am way off base on this than you can ignore the post, but I do hope that it helps clear some of the confusion for Don.
-
Don:
I liked you re-statement of the issue. It helped me understand what you are talking about.
While I've thought about this from time to time, I never gave it much attention--the notion that with a wet darkroon, you are putting the image through a second lens, with its own optical properties and focal plane.
I've only worked with 35mm film, and know what you mean about adjusting for film curvature. And if LF film is thicker than 35mm, I can see (with the help of your example) how there can be a depth of focus. I guess it also exists in 35mm film, too. I do remember developing film and having it feel thinner if I did something wrong--it's been to long to remember what error lead to "thin" negs...
And I understood your discussion of "layers" to mean layers of active elements of the emulsion--not color-sensitive layers. Is that what you were thinking? Because the chemical compounds in the emulsion have depth. The photo-sensor elements (each pixel) just count photons striking it. The film emulsion crystals change in different ways depending on how much light hits them. And those crystals are 3D. I guess the sesor also has thickness, but it is just counting photons, it's molecular structure isn't changing.
I'm curious, though--is your experience that the depth of focus is an extension of the depth of field? That is, does the depth of focus extend the depth of field on either end, or one of the ends? Given your description and graphic example I would suspect so.
But I'm also drawn back to the notion of sharpness. I wonder if the second lens (in the enlarger) acts to bring these additional elements of depth into "focus." I keep trying to think of how the depth you refer to would look on a contact print--where the neg has no lens between it and the photo paper.
But maybe that's a whole other topic and can of worms!
-
Regardless of the discussion of where the image in an emulsion actually is, I had been living with the following definitions :
Depth of field : the range of distances from the lens where the image elements are perceived as "sharp" or "in focus", while image elements at distances outside this range were considered "unsharp" or "out of focus". How these are determined/calculated requires some variables to be standardized : degree of enlargement, print size, viewing distance etc.
Depth of focus : the analogous range on the other side of the lens where the FOCUSED image is maximally sharp - as long as the film or sensor is within this range, the recording will be as sharp as possible; if the sensor or film is in the "wrong" position, the image intended to be captured (i.e. what was focused on) is not maximally sharp.
I never spent much time thinking about depth of focus because I assumed that for quite some time, lenses have been good enough to ensure it was adequate. Once this is the case, then the two parameters DoF and DoF) must be related in a predictable way - for example, if the sensor is not positioned correctly, all shots will front or back focus with respect to what is in focus on the viewfinder screen - and it needs to be sent in for calibration.
These definitions had a pleasing symmetry to them which made them easier to remember. Anyone else have a similar set of definitions?? Or are they out of date? In the long-forgotten equation containing 1/f , 1/v and 1/u, v and u refer to the distances from the lens to subject and the other to lens to film and DoF's refer to the ranges of u and v "considered sharp".
For digital, it seems like much of this discussion has left out a relevant item : the AA filter. I believe that this is substantially thicker than the sensor element on the chip is deep. All photons will go through this and some will be shifted to land up to a pixel or two away from where they would have been in its absence. Thus , there will be different amounts of the image still in focus at different depths in the AA filter, but the sensor elements will still only record a composite of those - and this controlled blurring is a good thing to prevent Moire effects.
Don, I have not experienced what describe in terms of focusing through a 4x5 onto an enlarger bed (since I only scan them) but even with my 10x loupe, I only see the transition between "all in focus" and "all not in focus" as I adjust the height of the loupe above the light table. For thick B&W emulsions, your "layers" might be detectable, but with layered colored emulsions, I think Jonathan has made a good case why you shouldn't be able to detect it as you describe (at least not without color shifts).
In your "TTY" figure, if you were to draw lines from the X's and O's to all parts of the exit pupil of the lens, you would see that the lines cross the sensor plane where adjacent or nearby sensor elements would catch them, that's why the X behind a given pixel must be captured by a nearby pixel, and points get blurred. Here the relationship of the two DoFs becomes a bit easier to discuss - the smaller the exit pupil (i.e. aperture) the more closely the rays converge on the same focus plane and the out of focus objects will "focus" either in front of or behind the AA filter, and point objects will be recorded by several pixels.
A couple of cents' worth of rambling
Andy
-
AJS and Steven - thanks, I think I understand better now, somewhat. I appreciate your patience and kindness, and simplicity of explanation or analogy. One last question, does this mean that as new lenses become more precisely made with improved convergence properties etc, that you will not have to stop the apertures down as much, or will that purely remain the only adjustment to compensate for the digital sensors' flatness. And I am still not 100% sure that if the lenses are the same basically and DOF is DOF, why any adjustment is needed with digital compared to film. Was it the circle of confusion with film that compensated for better by having any film curvature alone. Sort of like there being two areas of imperfection that kind of complemented each other and thus resulted in a image of greater sharpness with a relatively wider aperture setting???