Luminous Landscape Forum
Site & Board Matters => About This Site => Topic started by: dreed on March 09, 2012, 07:44:45 pm
-
In Nick's article, he says:
"Rangefinder cameras might be early 20th century technology, but this is a very 21st century baby."
I disagree.
If I look at the top of the camera, I see shutter speed in fixed fractions or multiples of a second and exposure compensation from -2 to +2 in thirds. There is nothing more 20th century than having to shoot based on someone else's idea of how fast you can move the shutter or under/over expose a photograph.
The offered speed steps do not allow for accurate selection of a shutter speed for proper exposure unless you're in a studio and have complete control over lighting.
The concept of a "standard set of shutter speeds" needs to be thrown out or at the very least, be augmented with much more flexibility than is available today so that it is possible to get closer to obtaining a 100% correct exposure or using "expose to the right" more accurately than is possible today.
Whilst the limited number of shutter speeds may tickle the fancy of various photographers that have grown up with cameras ("ooh, look, a shutter speed dial just like how I remember they used to be...") that were limited in this way, it is really an out dated concept that no longer serves the interest of photographers looking to get the best out of a scene.
-
Help me understand, please.
You're saying that you need exposure compensation finer than 1/3 stop increments? You're saying that having shutter speeds in 1 EV increments is a problem?
Can you explain how this would limit your photography, and also how you would implement and take advantage of something better?
Michael
-
In Nick's article, he says:
"Rangefinder cameras might be early 20th century technology, but this is a very 21st century baby."
Am I the only one who cringes when he calls this a "range finder?" Over and over again.
-
Help me understand, please.
You're saying that you need exposure compensation finer than 1/3 stop increments?
You're saying that having shutter speeds in 1 EV increments is a problem?
Can you explain how this would limit your photography, and also how you would implement and take advantage of something better?
The problem as it affects me...
The problem that I've observed is that when trying to ETR accurately, even when using 1/3s, is that there's one step that's quite away from the right edge of the histogram and the next step results in highlights being blown out. There is no middle step which allows me to get closer. This of course depends on the scene and the light as sometimes the steps are small enough that I can get reasonably close but just as often I cannot.
The problem in general...
The problem for real life is that the light from the sun does not arrive through the atmosphere in 1/3 stop increments, it is a continuous scale from really bright sunlight in the middle of summer to the dark of night in the middle of winter. A correct exposure is relative as to what is required to expose 18% grey "correctly" using the available shutter speeds and aperture. Thus whilst the "correct exposure" might be 1/250 or 1/125, it is not likely to be the correct shutter speed for 18% grey but is rather the best approximation of what's required to achieve that. Whilst the use of thirds allows the camera to get closer to being correct, if the correct shutter speed was actually 1/170 then the camera will likely select 1/160. So even if I take away ETR, the fact that the sun is continuously variable rather than providing light in fixed fractions and that we're forced to use specific fractions means that it is highly unlikely that an outdoors shot is ever 100% correctly exposed.
... for those that work in studios where light can be 100% controlled, these problems do not exist (well, as long as your light meter tells you that something is 100% correctly lit rather than rounding up/down.)
How I would fix the problem...
Fixing the problem requires a completely new approach to the use of both the shutter and aperture for photography. 50 years ago, we were much more limited in scope with what could be done. Today, cameras are aware of what lens is attached and what their capabilities are so we can redefine how we take a photograph. Today cameras are computers, not just a box with buttons, levers, somewhere to put film and attach glass.
In the first instance, being able to select the exposure of a photograph in distance from the right of the histogram should be possible. Thus I should be able to say "take a photograph that provides maximal exposure" (no blown highlights) in fully automatic mode and in shutter/aperture priority modes, remove a degree of freedom or two. Now it may be that there are a few specular highlights that the photographer is uninterested in, so clearly forcing the photographer to stay at or under that limit is not enough flexibility.
With a touch screen LCD on a camera, I should be able to look at a live mode photograph and drag the histogram with my fingers (or buttons) to where I want a particular crest to be and for the camera to decide what shutter speed is required to put it there. If the camera needs to use 1/178 or 1/93 to create the required histogram then that is the camera's problem, not mine. Personally, I see this as being useful to the entire spectrum of photographers.
What I'm looking for is a file of picture data that gives me a histogram of a particular shape and the camera should be doing what it can to enable me to do that rather than limit me.
In the second instance, if I'm just after a correct 18% grey then a "correct exposure" should use whatever shutter speed (and/or aperture) is required for that rather than just one of a handful. The most important step here is to allow a full spectrum of both shutter speed and aperture, so that I can have 1/193 at f/4.85 if that's what delivers a correct exposure for 18% grey, resulting in all of the model on the beach wearing a swimsuit being in focus and correctly exposed.
The above isn't to say that being able to set aperture and shutter speeds should disappear but rather that it is no longer good enough to only allowed setting those in fixed, discrete intervals and that it is an aspect of early film photography that by itself is past its use by date.
Cameras are no longer dumb. They've got small computers in them. It's been cute pretending that they've only been able to do one thing (write sensor data to a file) but it is time to move on. There's so much more potential waiting to be tapped.
As an example of more potential, there are often parameters that I may choose to fix, such as aperture for depth of field or shutter speed to control motion blur of either the subject or camera. Now that cameras have computers in them and know what lens is attached, the camera should report the aperture with something more meaningful such as "focus range 5' to infinity" or 'focus range 3.25" to 3.26"'. Some lenses have a distance scale on them but not all and simply put, if the looking at the lens can tell me that, why can't the camera's live view display? For cameras that have live view and/or electronic view finders, it should be possible to have the camera determine what the "freeze motion" shutter speed is without needing the operator to guess if it is 1/250, 1/500 or 1/2000.
-
Meanwhile, in other news, I much agree with Mr. Devlin- this is one sweet camera. The samples I've downloaded from the Fuji website seem to have a beautiful "look" and palette all their own. The random array sensor, no AA filter, and excellent glass are quite a combo. This, plus the other features are nearly enough to make me change my D800e pre-order to the Fuji, and have $1500 left over for lenses! Anyway, not gonna happen, as I also already have a newer aps-c camera, but I'll be looking forward to reading ongoing reviews of the Fuji, and admiring its "cool factor".
-
Am I the only one who cringes when he calls this a "range finder?" Over and over again.
I admit I was puzzled when he started talking about AF and EVF. To me rangefinder means manual focus with overlapping images in the middle of an optical viewfinder. Certainly it "looks" like a rangefinder, but does it actually focus like one?
But then I decided perhaps it was more about how the viewfinder operated, not in regards to focusing but in regards to framing like a traditional rangefinder viewfinder.
-
I admit I was puzzled when he started talking about AF and EVF. To me rangefinder means manual focus with overlapping images in the middle of an optical viewfinder. Certainly it "looks" like a rangefinder, but does it actually focus like one?
But then I decided perhaps it was more about how the viewfinder operated, not in regards to focusing but in regards to framing like a traditional rangefinder viewfinder.
I agree there are similarities in look and a tiny bit in function, but there are such similarities to a point and shoot and mirrorless models as well. Range finders have been defined for a very long time and this camera, as much as I like it (and my x100) is not a range finder. A range finder is as you describe. It has a "range finder" integral to the design of the camera which works on this principal. We really must resist categorizing cameras by look and not function. There are too many who read this who will just take it as gospel that it's a range finder.
On the other hand I'm open to learn, perhaps I'm not seeing something.
Personally I don't think the viewfinder frames like a range finder in OVF/EVF or live view.. but I see where the illuminated framing lines in the OVF are "kinda sorta in a totally new technological way" like a range finder. Pretty low bar to categorize a camera if you ask me.
I think this is an exciting camera. I'll admit to not finding enough to the newly announced DSLR's to replace my current models, but the Xpro1 is different.. I've enjoyed the x100 so much I don't see myself passing up this camera.
-
beautiful woman, almost toooo much detail!!!
-
"Am I the only one who cringes when he calls this a "range finder?" Over and over again. "
It's all arguing semantics really and will probably confirm to michael exactly what he thinks about internet forums ;-) but here goes anyway...
... a "rangefinder" is presumably any device that allows you to determine the distance to something. So whether that be achieved by parallax, by sonar, radar, contrast detection or a measuring tape, it's still pretty much rangefinding. I've always been happy to think of the Contax G cameras as rangefinders even though they use an AF system rather than optical parallax methods, but that's just me. On the whole I agree it can be confusing for techies when people use very specific terms in a looser sense. So, yes, let's not call the Fuji a rangefinder, let's give it a more accurate name. How about:
"a direct view, optical tunnel viewfinder cameras, with switchable LCD information overlays and switchable alternative EVF option that focuses automatically using an electronic rangefinding system with the option of manual focus by eye using a magnified EVF view".
There you go, much better than "rangefinder" which should be reserved for Leica, Zorki and FED. Confusion all cleared up ;-)
D
-
@Dreed: I had the same initial reaction as Michael: 'this is interesting but I don't quite get what he means'. What is it about 1/98.5th of a second that's so missing from this guy's life? ;)
Your longer post really clarified it: just like us, you lament the inability to get *every last damn drop of photons* into the sensor WITHOUT burning out a single important pixel, and wonder why the hell cameras can't do this. You, like me, find full 1/3rd stop increments too crude, at least based upon the readout we get on the instantly generated in-camera histogram.
We are totally on the same page on this one.
This is why I wrote that "ETTR" mode would be one of the most important advances we could get on cameras (for our purposes landscape shooters).
The cruder answer, though, I believe is that the in-camera generated histograms are actually not a whole lot more accurate in their display of burn-out than 1/3rd stop increments. They are based on jpegs and are conservatively biased.They cause us to leave light 'on the table' with every shot anyway.
Most cameras already have continuously variable electronic shutters, the fail-point on this is colour-specific, pixel-level metering, or something like that - whatever it is that keeps anyone from building an ETTR-enabled camera.
We'll both be at the front of the line for that one!
@Klaban: yeah, I'm really interested in trying it with Leica glass too, but based on what I've seen of the manual focusing on EVF cameras, I am not optimistic that this will be useful for anything but certain landscape work.
Sean Reid has pointed out to me that, in fact, one CANNOT move the focus-point around in the OVF. What is being displayed on the overlay is the point on the screen (approximately) which will be brought up as an enlarged view on the EVF is you depress a particular button. This 'zoom-in' focus assist will be useful, but it's more limited than what I got the impression the camera could do.
I will write a longer clarification of this in my next installment. (this limitation makes sense when the impact of paralax with the OVF is taken into account)
The X-Pro1 also does not have focus-peaking. Focus peaking is a really annoying way to focus, imho, but it works. The raison-d'etre of the X-Pro 1 is that it autofocuses. I suspect the camera's use with Leica
glass will be limited to tripod-based scenics, using liveview. That said, I suspect that the image quality will be phenomenal, since one's only using the sweet-spot of the already super-sweet Leica glass, and without an AA filter. It will make us all wish the M10 will scale a sensor of this quality out to FF size (ie: 35-40MP) to really USE the Leica glass.
@Steve (etc). I'm profoundly amused by the rage provoked by my calling this a rangefinder camera. Sean Reid called me to set me straight on this already. Yes, I know it does not possess a rangefinder mechanism. I will talk about this interesting (but ultimately pointless) semantic debate in my next installment. The main reason I resist caving in and calling it something else is that RF has come to mean more than 'a camera which focuses through image triangulation'. There's only one production digital camera left that does that, the Leica. There will likely be no more. But the term persists. Manual rangefinding was invented as a way to focus back when there was nothing better. While I love RF cameras, it is not for the way they focus. It is for their size, shape, sound and view-finding way of seeing the subject.
To me, continuing to use the term "rangefinder" to describe cameras which have all of those attributes, save-and-except for the eponymous, and deeply annoying, method of focusing, is logically consistent and communicationally effective. But I remain open to the opposite view...
Cheers,
- N.
-
I was going to write approximately what Nick just wrote about the definition of rangefinder. In the debates on the Leica forum L-Camera, I've argued that some form of focus confirmation would be extremely welcome with 90 and 135mm M lenses, but the purists argue that if we had it, it really wouldn't then be a rangefinder...or it would be a degraded one. That's fine with me; I'd take a degraded rangefinder kind of experience if I could get it. I'd take an m4/3 with a 150% view if I could get it...you know, a rangefinder-like deal where you could see around the edges of a window that defines the image. That more defines a rangefinder experience to me than the split image.
Where I differ with Nick is in his expectations of IQ, either in the expected M10 or really, any other camera now. I think somebody good in Lightroom could take any top-end camera and make prints that are pretty indistinguishable from each other at anything but the largest sizes. I don't think there's gong to be any magic in the Fuji, or the M10, or the D800...I think now it's going to be a matter of sorting individual preferences in matters of handling, weight, viewfinder, etc., but IQ is going to be high all across the spectrum.
JC
-
Whilst the limited number of shutter speeds may tickle the fancy of various photographers that have grown up with cameras ("ooh, look, a shutter speed dial just like how I remember they used to be...") that were limited in this way, it is really an out dated concept that no longer serves the interest of photographers looking to get the best out of a scene.
It is called "significant figures." Your stress on your idea of perfection has nothing to do with the reality of there issue. You don't use micron scales to design houses for a good reason. Your technical-only approach to photography is not really valid.
-
This rangefinder definition is a one of those things folks like to talk about. In photography (this is not just about finding distance) is defined as a parallax method separate from the imaging optics that determine how far a subject is away from the camera. This can be MF or AF--the Contax G series and Hexar AF use such an AF system. And the system does not have to be coupled to the rangefinder system--"coupled rangefinder" is what a Leica is. The Leica actually has a combination viewfinder and rangefinder.
Unfortunately, and especially in the internet age, folks like to throw around terms without a clear understanding of their meaning. They also invent them. What normally happens is we get a common term and a technical terms--"theory" is one such term with two definitions. Calling the X-Pro1 a rangefinder is just saying because it looks like a rangefinder it is one. (At least it is not that new silly term used for the Nex 7, "rangefinder style.") Nick can use any term he likes, I will know what he is talking about regardless. Professionally, I have no problem separating what people mean and the words they choose to use. I don't really expect reviews on Luminous Landscape to use correct technical nomenclature, nor do I care.
-
@Dreed: I had the same initial reaction as Michael: 'this is interesting but I don't quite get what he means'. What is it about 1/98.5th of a second that's so missing from this guy's life? ;)
Your longer post really clarified it: just like us, you lament the inability to get *every last damn drop of photons* into the sensor WITHOUT burning out a single important pixel, and wonder why the hell cameras can't do this. You, like me, find full 1/3rd stop increments too crude, at least based upon the readout we get on the instantly generated in-camera histogram.
We are totally on the same page on this one.
This is why I wrote that "ETTR" mode would be one of the most important advances we could get on cameras (for our purposes landscape shooters).
The cruder answer, though, I believe is that the in-camera generated histograms are actually not a whole lot more accurate in their display of burn-out than 1/3rd stop increments. They are based on jpegs and are conservatively biased.They cause us to leave light 'on the table' with every shot anyway.
Most cameras already have continuously variable electronic shutters, the fail-point on this is colour-specific, pixel-level metering, or something like that - whatever it is that keeps anyone from building an ETTR-enabled camera.
Indeed and this is the most frustrating part of looking at current digital cameras - knowing that the camera hardware can do more but that it is simply limited by their software!
I wonder sometimes that if camera manufacturers could be more open with the software and allow plugins to be loaded then problems like this could be solved by 3rd party software. This would allow the camera manufacturers to deliver standard cameras that operated like people expected but allow those seeking to get more out of their camera to do so. That would be another step in recognising that the camera is now well beyond being just a light box like it was in the 20th century.
Thinking about this from a different angle, maybe it can be solved now for cameras that allow remote shutter activation via cable. The problem that we have today is that the shutter activation with those devices is currently driven by humans. Why couldn't the cabled remote allow me to dial in how long to keep the circuit closed in tenths (or some other scale that changes logarithmeticaly), thus giving us finer control over the length of time that the shutter is open? A manufacturer that approached this problem with a well designed device could have one "knob" to set shutter speed and have a set of adapters that you attached to the cable for connecting to the camera like universal power supplies do for laptops, etc. (Some part of me wonders if I should have tried to design and build such a device using VC money rather than post the idea in a public internet forum.)
It is called "significant figures." Your stress on your idea of perfection has nothing to do with the reality of there issue. You don't use micron scales to design houses for a good reason. Your technical-only approach to photography is not really valid.
I understand what you're saying but the problem is that the offered resolution in shutter speed and aperture is no longer sufficient. That is, there are fractional values between what is offered that would be of benefit. This can easily be observed in difference between the presentation of histograms of successive photos taken only 1/3 EV apart.
----------------
An addendum to this is that in all my internet searching I've failed to find a proper reference to when, where and who made the agreement on the current fractions used for shutter speed and aperture size measure. Does anyone know if this was just a gentleman's agreement between various companies at the time or is there an actual camera standard, such as an ISO or IEEE document, that defines this?
-
...but the problem is that the offered resolution in shutter speed and aperture is no longer sufficient...
In your opinion. And yet, great, well-exposed photography is being done all the time.
-
Am I the only one who cringes when he calls this a "range finder?" Over and over again.
That made me cringe too, over and over. "Rangefinder" and "RF" have a clear and accepted meaning. Redefining a word just because you want to doesn't make any sense to me. There are very few true rangefinder cameras around, so we're just going to expand that term to describe other attributes? That's a bit like expanding the term "citrus" to include bananas, or expanding the term "cotton" to include polyester. Let's please call a thing what it is, not what something else is.
Not having a rangefinder doesn't make it a bad camera. Far from it. Skipping the rangefinder mechanism is what allows autofocus and accurate framing with a wide range of lenses. Not having a rangefinder is an important selling point for this camera, because in at least one important way it is BETTER than a rangefinder: it has autofocus. :) And it has a hybrid viewfinder, which no rangefinder camera offers. Not having a rangefinder is part of the reason this camera exists, and it is one reason why many photographers will buy it, especially photographers who, because of eyesight or other reasons, would not get along with a rangefinder.
-
Language is malleable and the meaning of words change over time.
We speak of "dialing" a phone, yet when was the last time that any of us saw an actual dial on a telephone.
Michael
"When I use a word," Humpty Dumpty said, in a rather scornful tone, "it means just what I choose it to mean - neither more nor less."
"The question is," said Alice, "whether you can make words mean so many different things."
"The question is," said Humpty Dumpty, "which is to be master - that's all."
Through the Looking Glass.
-
It is called "significant figures." Your stress on your idea of perfection has nothing to do with the reality of there issue. You don't use micron scales to design houses for a good reason. Your technical-only approach to photography is not really valid.
Well said. This quest for some imaginary perfect exposure reveals a misunderstanding of exposure. The correct exposure is the one that feels right; no further precision is needed. Now I need to go adjust the temperature by a few hundredths of a degree. ;).
-
This preoccupation with precision overlooks the fact that errors are in all these systems and that the accumulated error erases any possible nuance that one believes one is getting.
This has been compounded in the digital age by the belief in absolute numerical accuracy of electronics. I truly believe that we are not going to "leave light on the table" if we do not micromanage every quantifiable step along the way. The profoundly small adjustments that can be made both in LR and PS will completely obscure these slight differences in exposure. Moreover, as MR and others have maintained, the real proof is in the print which introduces another level of variability of it's own.
I like thinking in terms of music where the notes are starting points for the artist. No two violinists will use their bows in the same way. I am not advocating sloppiness in technique but I do believe we lose sight of the art when we focus on the numbers.
-
I consider this camera to be a "rangefinder formfactor" camera as opposed an "SLR formfactor" camera so calling it a rangefinder is just fine. A bit of genericization of terms is fine with me. Otherwise we really start to go down the road of having to come up with a new category every time a tiny bit of technology gets updated.
As to the holdover of "stops" in exposure, I totally disagree with the OP. These numbers mean things. There is a very specific relationship between Apertures, Shutter-speeds and ISO's. I remember about 13 years ago I was buying a new flash meter and another photographer was in the store buying one too. He was all upset because the granularity of the digital meters was only 1/10 of a stop. His concerns were bogus and the shop owner proved it to him by asking him to try and meter the scene exactly the same way twice.
ETTR for maximum dynamic range and noise control is great with SOME cameras (Olympus and Panasonic are exceptions due to the way the pixels combine), but almost without exception when you get within 1/3 a stop of clipping in the histogram you've already clipped sensels. This is especially true with the derived colors like yellow, cyan and magenta. With Oympus/Panasonic files, increase that margin a bit more.
As a portrait photographer I am most interested in skin tones. So far, from the samples I've seen, Fuji has done very well in this regard. But some serious hard comparisions would be nice. In the world of landscape photography we really aren't overly precise in things. We always fix stuff in post. But for the majority of working pros, having a camera that nails things down in camera is critical.
-
Fuji tend to do well on skin tones (and other subjects), but they should do as they have a lot of experience in this area (it's subjective but still a valid point)
As for this model it looked very interesting, until I discovered Fuji had gone for "focus by wire", this dated relic of camera experimentation got dumped decades ago because it basically sucked v mechanical focus. That alone is a deal breaker for some.
Maybe I'm overly fussy, but whilst Fuji seemed to be tuned into "photographers" for many of their newer releases, clearly the "awful focus by wire" missed someone's brain at Fuji HQ as it did for some of their previous offerings.
-
"Am I the only one who cringes when he calls this a "range finder?" Over and over again. "
It's all arguing semantics really and will probably confirm to michael exactly what he thinks about internet forums ;-) but here goes anyway...
... a "rangefinder" is presumably any device that allows you to determine the distance to something. So whether that be achieved by parallax, by sonar, radar, contrast detection or a measuring tape, it's still pretty much rangefinding. I've always been happy to think of the Contax G cameras as rangefinders even though they use an AF system rather than optical parallax methods, but that's just me. On the whole I agree it can be confusing for techies when people use very specific terms in a looser sense. So, yes, let's not call the Fuji a rangefinder, let's give it a more accurate name. How about:
"a direct view, optical tunnel viewfinder cameras, with switchable LCD information overlays and switchable alternative EVF option that focuses automatically using an electronic rangefinding system with the option of manual focus by eye using a magnified EVF view".
There you go, much better than "rangefinder" which should be reserved for Leica, Zorki and FED. Confusion all cleared up ;-)
D
1. Well ya.. You didn't really say that did you?
2. All I know about Michael and how he feels about forums can probably be summed up by knowing this is his forum. You did it again... :o
3. A yardstick would be a "range finder" by this definition.. so no.. not really. We all know what "range finder" in the context of cameras means. Well, except the author of this article.
4. Humor is good.
-
@Steve (etc). I'm profoundly amused by the rage provoked by my calling this a rangefinder camera. Sean Reid called me to set me straight on this already. Yes, I know it does not possess a rangefinder mechanism. I will talk about this interesting (but ultimately pointless) semantic debate in my next installment. The main reason I resist caving in and calling it something else is that RF has come to mean more than 'a camera which focuses through image triangulation'. There's only one production digital camera left that does that, the Leica. There will likely be no more. But the term persists. Manual rangefinding was invented as a way to focus back when there was nothing better. While I love RF cameras, it is not for the way they focus. It is for their size, shape, sound and view-finding way of seeing the subject.
To me, continuing to use the term "rangefinder" to describe cameras which have all of those attributes, save-and-except for the eponymous, and deeply annoying, method of focusing, is logically consistent and communicationally effective. But I remain open to the opposite view...
Cheers,
- N.
1. I'd be amused by someone feeling rage over the issue too.
2. Good for Sean Reid. ::)
3. Since when and to whom? Why am I always the last to know such things..
4. Well heck, if that's the criteria we don't have to fear the demise of the range finder. Seems to be a lot more of them than I knew about just yesterday.. :D
5. That's good. I like to think I'm open minded too. I'm probably not, but I like thinking so.
It will be interesting to read how you address this in your next installment which btw I'm looking forward to. Great information and fun reading. I haven't yet decided if I'll get mine, but articles like yours help us all make such decisions. Thank you.
-
ETTR for maximum dynamic range and noise control is great with SOME cameras (Olympus and Panasonic are exceptions due to the way the pixels combine), but almost without exception when you get within 1/3 a stop of clipping in the histogram you've already clipped sensels. This is especially true with the derived colors like yellow, cyan and magenta. With Oympus/Panasonic files, increase that margin a bit more.
Are you really sure you know what you are talking about?
What is it about the "way the pixels combine" in Olympus and Panasonic "files" that makes them not "great" candidates for ETTR?
-
For those of us with digital speedometers:
-
Well said. This quest for some imaginary perfect exposure reveals a misunderstanding of exposure. The correct exposure is the one that feels right; no further precision is needed.
There isn't a quest for a perfect exposure, rather for an exposure to be correct according to our desires.
In some instances, that desire means over exposing the image to a point just before that where detail that we care about is over exposed.
In other instances, it can mean something else.
The problem is that even though we've got modern tools, they're still imposing antiquated limits and boundaries on how we achieve an exposure that feels right. In effect, for those that like to shoot raw and ETTR, the tools are not providing us with what we need to obtain an image that has been exposed how we would like it to be. The tools are capable of giving us much greater freedom than we have today in choosing parameters for the exposure that feels right to us.
As much as digital photography has advanced the art, some aspects of it appear to be well and truly welded to the past. To me, if a camera is going to be branded as a tool of the 21st century then it needs to be prepared to break those bonds.
The Lytro Light Field Camera is perhaps the first real example of a 21st century camera. It challenges a great many of the existing preconceptions of what is required to take a photograph. In comparison, everything else is pretty much a case of "lets replace film with a digital sensor."
-
"I once read about five monkeys that were placed in a room with a banana at the top of a set of stairs. As one monkey attempted to climb the stairs, all of the monkeys were sprayed with jets of cold water. A second monkey made an attempt and again the monkeys were sprayed. No more monkeys attempted to climb the stairs. One of the monkeys was then removed from the room and replaced with a new monkey. New monkey saw the banana and started to climb the stairs but to its surprise, it was attacked by the other monkeys. Another of the original monkeys was replaced and the newcomer was also attacked when he attempted to climb the stairs. The previous newcomer took part in the punishment with enthusiasm. Replacing a third original monkey with a new one, it headed for the stairs and was attacked as well. Half of the monkeys that attacked him had no idea why. After replacing the fourth and fifth original monkeys, none had ever been sprayed with cold water but all stayed the fuck away from the stairs.
Being here longer than me doesn't automatically make your adherence to a rule, or the rule itself, right. It makes you the fifth replacement monkey. The one with the weird red arse and the first to point and screech when anyone approaches the stairs. I would be the sixth monkey, at home in bed trying to come up with a viable excuse not to spend another fruitless day locked in a room with five neurotic monkeys."
http://www.27bslash6.com/timesheets.html
-
Are you really sure you know what you are talking about?
What is it about the "way the pixels combine" in Olympus and Panasonic "files" that makes them not "great" candidates for ETTR?
Not enough time at the moment to go into detail here, but Olympus/Panasonic sensors utilize two different greens. It is vital that during raw conversion all for pixels are combined properly otherwise you get artifacts and color shifts--especially in the highest exposure values.
When you look at the histogram in RGB mode, you think that all theee colors are safe, but what you are looking at is a converted image, not the sensor data. The information you are looking at is two steps removed from the sensors actual A-D conversion. First step is passing through the built in demosiac algorithm and the second pass is the breaking down of exposure values of the three primary colors.
But consider what happens with a derived color such as yellow. You are combining greens and reds. To achieve an ETTR exposure with yellow, the green sensels are usually OK even though you have blown out the red sensels. When you look at the RGB histograms the red looks fine because it is a post conversion view of the damaged image. This usually isn't a problem when the scene is made up of pure red, blue and green colors.
-
I like monkeys, I had one, once, for about a night.
I was fifteen, and bought the thing in Madras whlilst spending the day there awaiting a train connection home. It seemed quite friendly when the guy handed it over in exchange for some rupees, but after some hours perched up on the luggage rack in the compartment, its happy demeanour turned sour and all that it seemed disposed to display was teeth.
I realised that taking this thing home hadn't been my brightest idea; I also realised that I'd probably be both bitten by the monkey and that neither of us would greeted with a lot of enthusiasm at home. So, I passed ownership (what a fond concept!) on to another kid moving further down the line... I often wondered where the monkey ended up. Other than confused, I mean.
But I still like monkeys, and their behaviour re. steps is quite intelligent; note they seldom buy cameras, even though I'm sure they could easily learn how to operate them. We have much to learn from our cousins.
Rob C
-
Paracetamol, Rob, paracetamol.
Don't dare to take it, Keith; it clashes! (I assume you're not suggesting it for the monkey?)
Fake coffee works a bit, but only as long as I don't remember that it's fake.
;-(
Rob C
Bracketing etc.
I think folks might be looking at this thing the wrong way around. Think horses for courses.
If you want a camera that allows all the multiple functions of an slr, then buy an slr. If this camera is, indeed, modelled on the M concept of photography, why should it provide bracketing, which implies a tripod at the very least? Cameras of this ilk are supposedly street-wise instruments, light and convenient machines that can provide the discretion and speed that slr bodies simply can't. You want to make a statement, carry a huge slr; you want to be cool and quick and, preferrably, not noticed, carry a "rangefinder", which I am perfectly happy to consider this particular camera to be, regardless of semantic punctiliousness expressed within these pages. I am totally prepared to accept that there are basic camera types: rangefinder, slr, tlr, field, monorail and even process. Is it hard to guess where this one fits, does one even have to guess?
As for needing more than a third/quarter stop control... never in my life, and that included heaps of Kodachrome and Velvia. But then, I was always interested in the image I was making; the technical birth pangs were of no consequence at all. As someone mentioned, the moment you go into PS or whatever, all bets are off; you might as well have stayed within a third of a stop and enjoyed your shoot. Unless, of course, it is the splitting of hairs that makes your day.
I would love to have my D700 sensor inside a light, small and convenient body that doesn't cripple me after half-an-hour of walking. Yes, I am old and decrepit, but some day, with luck, you will all be, so don't knock the requirements.
There never has been and probably never will be the perfect, universal tool; for me, the closest it got was the Nikon F. But I was young.
Rob C
-
Not enough time at the moment to go into detail here
I don't think you know what you are talking about.
-
...
I think folks might be looking at this thing the wrong way around. Think horses for courses.
...
I am totally prepared to accept that there are basic camera types: rangefinder, slr, tlr, field, monorail and even process. Is it hard to guess where this one fits, does one even have to guess?
As for needing more than a third/quarter stop control... never in my life, and that included heaps of Kodachrome and Velvia. But then, I was always interested in the image I was making; the technical birth pangs were of no consequence at all.
...
As someone mentioned, the moment you go into PS or whatever, all bets are off; you might as well have stayed within a third of a stop and enjoyed your shoot. Unless, of course, it is the splitting of hairs that makes your day.
...
Why does the way in which photography is performed need to be forever wrapped up around how it was practised in the 20th century using film?
For those of us that shoot raw, every photograph requires us to go into PS or whatever and I suspect that what those of us that shoot raw would like is for the camera to make it easy to maximize the information that PS is given so that we can go back to just caring about the image.
-
As for needing more than a third/quarter stop control...
All I know is, nobody will ever see my images exactly as I intend unless they are looking over my shoulder at my monitor, or viewing the print under my lights.
Control and accuracy (as opposed to precision) are important, but at a certain point these becomes meaningless.
Paul
-
Help me understand, please.
You're saying that you need exposure compensation finer than 1/3 stop increments? You're saying that having shutter speeds in 1 EV increments is a problem?
I wouldn't claim to personally need exposure compensation finer than 1/3 stop increments or that shutter speeds in 1 EV increments have ever been a problem for me. However, I do (very humbly, being no kind of engineer) wonder whether, if there are technologically advanced civilisations on other planets that also do digital photography, they necessarily do exposure, shutter speed and sensitivity exactly the same way we do, or if we did the (thought or material) experiment of redesigning a digital camera from scratch without regard to current practices but taking maximum advantage of current technology, we would come up with what we have now. If speed, aperture and sensitivity can be made continuously variable, why not allow them to be? Of course it is useful to have some sort of equivalence between the impact on exposure of changes in each variable, but the current numerical conventions as to how changes are measured are hardly optimal for that purpose - they are just what we are used to. If perfect, or significantly improved, precision is difficult to achieve, and benefits no-one, then sure, don't bother. But if it could be done with current technology, and has practical benefits for some photographers, then why not?
-
Why does the way in which photography is performed need to be forever wrapped up around how it was practised in the 20th century using film?For those of us that shoot raw, every photograph requires us to go into PS or whatever and I suspect that what those of us that shoot raw would like is for the camera to make it easy to maximize the information that PS is given so that we can go back to just caring about the image.
IMO, because it's still exactly the same exercise: the placing of a measured quantity of light onto a surface, making the most of the two variable factors that control it to the greatest extent - brightness and duration of exposure. So far, that doesn't depend on the number of the century.
Your second point, about 'going back to just caring about the image' sort of implies that you enjoyed that condition during an earlier period - which can only mean the time when you used film... If the controls were sufficient then to allow absorption of the self by the creative aspects of the medium, why see a problem with the same (though it's actually better) situation today? Yes, I can understand perfectly well that there might be some theoretical value to 'perfect' control of parameters, but from a practical point of view, it means zilch when we are already able to be so very accurate. It's akin to thinking that if we don't breathe a precise quantity of air with each breath, and a given number of such breaths per minute that we shall suddenly die. Life is a very flexible concept and photography no less so.
Anyway, on the well-proven Ferrari/Lamborghini principle it would all end up making cameras even more expensive and prone to breaking down.
Rob C
-
IMO, because it's still exactly the same exercise: the placing of a measured quantity of light onto a surface, making the most of the two variable factors that control it to the greatest extent - brightness and duration of exposure. So far, that doesn't depend on the number of the century.
Your second point, about 'going back to just caring about the image' sort of implies that you enjoyed that condition during an earlier period - which can only mean the time when you used film... If the controls were sufficient then to allow absorption of the self by the creative aspects of the medium, why see a problem with the same (though it's actually better) situation today? Yes, I can understand perfectly well that there might be some theoretical value to 'perfect' control of parameters, but from a practical point of view, it means zilch when we are already able to be so very accurate. It's akin to thinking that if we don't breathe a precise quantity of air with each breath, and a given number of such breaths per minute that we shall suddenly die. Life is a very flexible concept and photography no less so.
Film photography is a different activity to digital photography.
They may both be photography and use similar equipment but they're not the same.
Now just how accurate is photography?
If you use thirds and your shutter speed is 1/30 then it is possible that the camera has rounded up from 1/28 or down from 1/35.
If it is rounded down from 1/35 then the 1/30 shot is over exposed by 17%.
That is huge and I don't know how anyone could consider it to be "accurate."
-
Film photography is a different activity to digital photography.
They may both be photography and use similar equipment but they're not the same.
Now just how accurate is photography?
If you use thirds and your shutter speed is 1/30 then it is possible that the camera has rounded up from 1/28 or down from 1/35.
If it is rounded down from 1/35 then the 1/30 shot is over exposed by 17%.
That is huge and I don't know how anyone could consider it to be "accurate."
With all due respect, your example is complete nonsense. If you can measure, let alone see the difference in exposure between 1/35 sec and 1/30 sec you'd better show us how, because it flies in the face of both logic and practice.
In any event, most lenses can be adjusted in 1/3rd stop increments, and most electronic shutters are continuous, with almost infinitely variable settings when adjusted by a cameras auto mode. Manually they are in full stop increments, but that's a human factors issue, not a technical limitation.
You are confusing precision with accuracy.
You're right on one thing though. Exposing for digital is different than for film. For digital, there is only one technically optimum exposure when shooting raw, at base ISO, and that's at the maximum exposure that avoids blowing non-specular highlights. Fussing over even a 1/2 stop there though is more like a fetish than concern for optimum technical quality.
Michael
-
Don't you just love the breath of fresh air in the morning? Far better than napalm!
Rob C
-
With all due respect, your example is complete nonsense. If you can measure, let alone see the difference in exposure between 1/35 sec and 1/30 sec you'd better show us how, because it flies in the face of both logic and practice.
I must admit that after I wrote that, I was thinking to myself "how can it be that film based photography worked as it did if exposure steps were so limited?" because as you say, the practice would suggest that finer grained shutter speeds are not required for our eyes.
And yes, you're right about the difference of what would be 1/6th of a stop being too difficult to discern - through simulating the before/after with LR I could see the colour change but if I turned away and looked back, it was damn near impossible to pick between the two (or at least I couldn't using the laptop that I'm currently using. Maybe if I tweaked the monitor or glued it to my nose it would be easier to notice the difference.) I suppose I should have tried this out before to test that argument.
-
I don't think you know what you are talking about.
If you are interested in dialog, I'll gladly participate. But shutting a person down with this kind of statement is a waste of a good forum. I expect it in Dpreview, not here.
-
And yes, you're right about the difference of what would be 1/6th of a stop being too difficult to discern
If you think the difference between 1/30th and 1/35th is one sixth of a stop you need a better understanding of the fundamentals before you start pedantic arguments here.
-
If you are interested in dialog, I'll gladly participate. But shutting a person down with this kind of statement is a waste of a good forum. I expect it in Dpreview, not here.
Making stuff up is a "waste of good forum".
-
Ok... I admit... I missed this thread initially and got to it when it reached its third page. However, after reading just the OP, all I can eloquently say is:
Booooooo!
-
Making stuff up is a "waste of good forum".
Yes, I have a fertile imagination. Thanks for the compliment. But in this case my fertile imagination is what caused me to question why certain things failed in digital imaging. Questions like when reds blow out, why do they shift to yellow? Why the color shifts when using highlight recovery? This was the basis for quite a bit of research and testing with color charts with and without color filters, comparing numerous Raw converters and looking at the pixel data inside the Raw files themselves.
But I'm guessing that you have done the same thing and came to another conclusion. I didn't test Nikon files so your results might be different. Canon, Panasonic, Minolta and Olympus files, yes; Nikon files, no. That was also why I was specifically mentioning Olympus and Pansonic files. These I know well. Nikon files, not so much. Whatever you know about Raw files with Nikon may not necessarily apply with other brands. The filteration on the sensors is quite different.
Feel free to show me your results that prove that I'm making this up. Sorry, but you cannot use just Lightroom to prove your point. That would be far too easy to debunk so don't even bother going there. You'll need to try harder than that.
On a related note, in your D700, what is the dynamic range of the blue channel under daylight balance? What is it under tungsten after WB adjustment? Is it better to apply WB correction or use a color correction filter?
Simple questions sometimes demand careful analysis and research. These questions can sometimes open up a Pandora's Box.
Ken
-
Ok... I admit... I missed this thread initially and got to it when it reached its third page. However, after reading just the OP, all I can eloquently say is:
Booooooo!
Fine, but I still don't see anything "21st century" in this camera - or most other digital cameras.
Speaking of 21st century, there's only another 7 years before we'll have the Esper Machine (Blade Runner) and whatever camera was used to take that photo.
-
Fine, but I still don't see anything "21st century" in this camera - or most other digital cameras.
Speaking of 21st century, there's only another 7 years before we'll have the Esper Machine (Blade Runner) and whatever camera was used to take that photo.
It would definitely be nice to have an ETTR exposure function in the camera with user programmable limits and appropriate protections to keep the colors intact at the upper limit. Technically, this is entirely possible today if the manufacturers put a programmer on it for a week.
The latest/greatest Sekonic meter actually is a fantastic tool in this regard. You create exposure profiles for your films or cameras. With a quick spot meter reading of the highlight you want to put to ETTR, a press of a button will tell you what exposure to use. It's accurate to 1/10 of a stop.
Ken
-
I've been lobbying both publicly as well as privately for in-camera ETTR exposure for years, and while several manufacturers I spoken with agree that it makes perfect sense, thus far none have the smarts to implement it. The DNG standard even has a built-in correction factor so that screen review and post processing preview could display a "normalized" image. Alas most camera makers also abjure DNG.
Sigh
-
... if we did the (thought or material) experiment of redesigning a digital camera from scratch without regard to current practices but taking maximum advantage of current technology, we would come up with what we have now. If speed, aperture and sensitivity can be made continuously variable, why not allow them to be?
In my alternate photographic universe, thee would still be settings for shutter speed, and for aperture (but maybe with the option of specifying an effective aperture diameter rather than an aperture ratio, for more direct relevance to DOF), and then a way to specify how light levels in the scene are scaled to digital output levels (maybe metered mid-tones placed at 18% of maximum level, or metered highlights placed at just below maximum level, and including exposure compensation). But there would not be an ISO speed, because that is a carryover from film that does not fit so well with the operation of an electronic light detection signal and processing device, and the one that is most remote from directly describing any feature of the composition.
Instead, gain would be the replacements, and the three parameters of aperture, exposure duration, and "tonal placement" could dictate the degree of gain needed, and do so on a continuous scale. Also, the camera could usually take care of how much of that gain to apply to the analog signal before ADC, how much to apply after ADC as digital rescaling, and how much to simply note in the raw file for subsequent application in post-processing. If the camera's calculation gives "exposure index 729", and this means that the amplifications is such that the maximum output level corresponds to the signal from a well filled to 22% of capacity, why not?
-
BJL's approach of altering the channel gain on-sensor before A-D conversion has a lot of merit. The roots of this are firmly planted in decades of audio and video processing. Still imaging is actually pretty late to this game. What BJL is suggesting is actually one of the tools at our disposal in scanning film with some scanners and software. It's nothing that need to be invented. Just applied.
The concept that stops is old school and stopless is 21st century has an inverted parallel. I'm a darkroom rat. Grew up in the darkroom, have Dektol in my DNA. The single greatest thing I ever did for the darkroom was buy an RH Designs Stopclock Professional and ZoneMaster II meter. Both operate in..... STOPS! Printing exposures are generally considered infinitely variable. By working in stops and fractions of stops, it not only simplified the process greatly, but achieved levels of accuracy and control never experienced before.
Ken
-
I've been lobbying both publicly as well as privately for in-camera ETTR exposure for years, and while several manufacturers I spoken with agree that it makes perfect sense, thus far none have the smarts to implement it. The DNG standard even has a built-in correction factor so that screen review and post processing preview could display a "normalized" image. Alas most camera makers also abjure DNG.
Sigh
What makes you think the camera isn't already doing this prior to RAW data creation?
And at the danger of repeating myself ad nauseam: suppose the sensor has a non-linear response at its lower and upper bounds, how would you suggest implementing ETTR?
-
Yes, I have a fertile imagination. Thanks for the compliment. But in this case my fertile imagination is what caused me to question why certain things failed in digital imaging. Questions like when reds blow out, why do they shift to yellow? Why the color shifts when using highlight recovery? This was the basis for quite a bit of research and testing with color charts with and without color filters, comparing numerous Raw converters and looking at the pixel data inside the Raw files themselves.
For someone who doesn't have time to directly address the issue at hand, you sure have a lot of time for other stuff and musings ...
Nothing that you have offered explains why Olympus and Panasonic cameras can't benefit from ETTR. You've made vague references to multiple kinds of green sensels and the fact that the histograms in digital cameras are based on rendered images ... but you have yet to explain your comment that I questioned about Olympus and Panasonic cameras and their unsuitability for ETTR.
I use ETTR techniques with my Panasonic GF1 regularly.
-
For someone who doesn't have time to directly address the issue at hand, you sure have a lot of time for other stuff and musings ...
Nothing that you have offered explains why Olympus and Panasonic cameras can't benefit from ETTR. You've made vague references to multiple kinds of green sensels and the fact that the histograms in digital cameras are based on rendered images ... but you have yet to explain your comment that I questioned about Olympus and Panasonic cameras and their unsuitability for ETTR.
I use ETTR techniques with my Panasonic GF1 regularly.
No, I'm not saying that ETTR can't be used with Olympus/Panasonic files, but that the point where you get color shifts in recovered highlights occurs sooner. Depending on color, this can occur about a half stop away from clipping. This also occurs with other cameras too, but to different degrees of success.
In nature photography there are few exacts. It's definitely a different world than product photography, for example. I'm not sure there are many product photographers who use ETTR. Not only does it add too much variability, but the risk of color shifts is too great.
When you get too close to the cliff with ETTR, bad things can happen. I'm sure I'm not the only one who has reverted to a bracketed shot because I pushed my histogram too far even though it was still within the lines.
Ken
Ken
-
No, I'm not saying that ETTR can't be used with Olympus/Panasonic files, but that the point where you get color shifts in recovered highlights occurs sooner.
Really? Show us.
I hope by "recovered" you don't mean to imply that these "highlights" you are "recovering" came from saturated sensels ... as that isn't ETTR ... that's called ETFTTR.
-
Really? Show us.
I hope by "recovered" you don't mean to imply that these "highlights" you are "recovering" came from saturated sensels ... as that isn't ETTR ... that's called ETFTTR.
I could. But I see where this is going. You have already come to a conclusion, and no facts will prove otherwise. This is one of those no-win situations where if you disagree with the tests, challenge the tester. This is how you have operated in the past.
What I would rather do is get people thinking about this and doing their own controlled testing to know exactly where their system hits the limit. I will say that anybody who thinks the histogram is an accurate representation of what is going on at the sensor is very ill-informed. You will require a balanced light source and a Kodak color test target. Take a series of photos of the target varying the exposure from Five stops under to five stops over. (or to the maximum limit of your lighting and exposure control). Process the files and see what really is happening at the top and bottom points of exposure. It will be an eye opener. Especially as you apply highlight and shadow recovery, but even exposure-compensation will reveal much. What happens with the derived colors (yellow, cyan, magenta) is even more challenging to the conventional wisdom.
This is just the digital equivalent to building log curves. In the B&W photography world, there are those who have done this and understand how the films actually respond. For everybody else, they just overexposed and underdeveloped the film because others said so.
Alas, few will test. The rest will ETTR and end up using the bracketed exposure not knowing why the ETTR shot didn't work right.
Ken
-
What I would rather do is get people thinking about this and doing their own controlled testing to know exactly where their system hits the limit. I will say that anybody who thinks the histogram is an accurate representation of what is going on at the sensor is very ill-informed.
Software like rawnalyze (now we have a modern version = www.rawdigger.com) to see the real raw data histogram and UniWB to get in camera histogram as close to the raw data histogram as possible were in place for years... I do not think there are readers on this forum who still think what you allege they think.
-
I've been lobbying both publicly as well as privately for in-camera ETTR exposure for years, and while several manufacturers I spoken with agree that it makes perfect sense, thus far none have the smarts to implement it. The DNG standard even has a built-in correction factor so that screen review and post processing preview could display a "normalized" image. Alas most camera makers also abjure DNG.
In your conversations with manufacturers, have they ever said or hinted at why they have not or will not adopt DNG?
-
I could. But I see where this is going. You have already come to a conclusion, and no facts will prove otherwise. This is one of those no-win situations where if you disagree with the tests, challenge the tester. This is how you have operated in the past.
Um, no. I haven't come to any conclusions ... you have ... and you should be able to explain exactly how you arrived at that conclusion since you claim to have done the testing.
So far, you haven't showed ANY evidence or facts.
You've made unsupported claims, I questioned them and you have (so far) failed to back those claims up with any facts or evidence.
The fact that the histogram isn't a perfect guide to RAW clipping is true ... but irrelevant. This is the case with all cameras today. That doesn't mean you can't increase SNR with proper ETTR ... it just means you need to understand how to properly ETTR with the knowledge that the histogram is a crude guide ... but that's the same with my Nikon as my Panasonic.
You have claimed that there is something special happening with Panasonic and Olympus cameras due to different kinds of green sensels.
Prove it.
-
In your conversations with manufacturers, have they ever said or hinted at why they have not or will not adopt DNG?
I guess it then will limit their innovations:
1) case in point = Panasonic, software optics correction and the time it took Adobe to provide a new version of DNG to accomodate that.... so should Panasonic wait for Adobe ? should Panasonic try to put the necessary information into some tag where all proprietary information from manufacturer is stored as one big binary chunk of data ? should Panasonic disclose their development to a 3rd party (Adobe) if they were trying to keep that in secret ?
2) case in point = Foveon raw files - Adobe still do not store the information natively (for how many years no proper support ?)
---
who adopted DNG really among companies w/ marketshare > 0.0x % ?
1) Pentax
2) Ricoh
3) Samsung
4) Leica (for cameras that are not rebranded Panasonic's)
---
what happened now ?
1) Samsung dropped DNG support after they stopped rebranding dSLRs developer by Pentax
2) Ricoh purchased Pentax - you have one company instead of 2 now
3) Leica still uses it (for cameras that are not rebranded Panasonic's)
---
neither are actually innovating anything just using 3rd party sensors... companies that do design/make their own sensors are not going to be bound by a standard (Samsung's move, as their do their own sensors, is a perfect example - once they were no longer using based Pentax designs, where DNG was inherited, they dropped DNG immediately)
-
What I will want to know is how is the manual focus with the M mount adapter. If I end up with this camera ( Likely! )I will to start with Fuji X to Canon FD as I have a bunch of orphaned glass I wouldn't mined trying on it. As I will have a range of glass that will take some time for them to put out and it's payed for. Thoughts. My first post thanks.
-
http://www.luminous-landscape.com/forum/index.php?topic=63617.0
Image66 ... New thread so we don't have to hijack the Fuji discussion to explore your theories.