[kh1234567890]

I just can't keep up with all this new physics, it's not what it used to be before t'Internet

You shouldn't stress over technicalities if the tool does what you want it to do

Sorry, I forgot that sarcasm is wasted on this forum. What worries me is that all this cr*p is being archived somewhere, forever. I pity the future generations !

[rparmar]

AFAICT, the only people that equated the Q with a DSLR were those that were criticizing it for not being one. This is called strawman argument.

Good thing it's not an argument I'm making then, but only one you presented. Hold on a minute -- doesn't that mean it's you presenting the straw man?

Really, I am beginning to wonder why you take every comment as a reason to criticise and be offensive. Especially when you are so darned bad at it.

[Class A]

Class A, your argument that the aperture for exposure changes with sensor format is incorrect.
....
An aperture value such as f/1.9 depends only on the size of the entrance pupil (minimum diameter of front lens element) and focal length.

You are contradicting yourself with these two statements.

First of all, neither focal length nor aperture diameter of a lens change with sensor size; they cannot, they are physical properties of the lens.

One often, however, one talks about lenses regarding their full format equivalent. For instance, the full format equivalent of a DA 70/2.4 would be a 105mm f/3.4.

The 8.5mm f/1.9 lens has its physical properties that are independent of sensor size. Some people like to convert the focal length, however, into a full format equivalent focal length (-> 47mm). That's fine but the specification "47mm f/1.9" is nonsense. This refers to a lens with an aperture diameter of 24.7mm. We both know, however that the 8.5mm f/1.9 lens only has an aperture diameter of 4.5mm. So if you say "47mm" you have to say "f/10.5" as well. Any mixing doesn't make sense. Stating that an f-ratio applies to DOF but not to exposure doesn't make sense either. Exposure is influenced by the aperture diameter which is why it would be wrong to upgrade it from 4.5mm (8.5/1.9) to 24.7mm (47/1.9).

The smaller sensor though means loss of performance at high ISO, but whether it costs 5 stops or some lesser amount I'm not sure. It certainly does seem true of APSC vs FF where APSC loses a stop in high ISO performance to FF.

Smaller sensors do not have a high ISO disadvantage. APS-C does not lose a stop of high ISO performance against FF. FF sensors have a larger area to collect light, but the need to spread the light out to this larger area reduces its intensity per sq. mm, cancelling out the larger area advantage. Please see the "Low noise benefit of FF vs APS-C equals ... zero" thread.

[jstevewhite]

In sensor technology they are. The lower the noise floor, the higher the dynamic range. Almost all of the DxOMark results basically measure noise. That is, however, completely independent of sensor size.

Smaller sensors are not intrinsically noise limited. Why would they be? Why would a sensel produce more noise just because there is less sensor around it?

True, as far as it goes; it's pixel density that people are complaining about, which generally goes up as the size of the sensor goes down. In fact it must if the MP is to be maintained. Smaller pixels mean less dynamic range, all things being equal:

Clarkvision: Digital Camera Sensor Performance Summary

is a good reference.

The lenses on the Q camera are slow like hell.
A widest aperture of f/1.9 for a 8.5mm lens may sound fast, but if you express it in full-format terms, you see that a full-frame camera can take the same pictures with a 47mm lens that has a widest aperture of f/10.5.

Of course. I have never said anything else.
With a Q camera and the 8.5mm lens, you never shoot with a focal length of 47mm.

If you want to predict the kind of images you can get with the Q standard lens, it can be useful to hypothesise about an equivalent lens on a full frame camera. That lens would be a 47/10.5.

That doesn't make sense. You're right that it isn't an f/10.5. Nor is it a 47mm. That's why I said people should not refer to the lens as a 47mm f/1.9 lens. There is an equivalent full frame lens which is a 47/10.5. For this one f/10.5 counts for DOF and exposure. You cannot separate DOF and exposure.

Here's where our problem is. If you have a full frame camera and a 47mm lens, and I have a Q with its 8.5mm lens, and a Pentax Spotmeter says our correct exposure is f1.9 @1/250 for ISO200, we will both use f1.9, 1/250th, and ISO200, and we will both get accurate exposure. Period.

DOF and exposure are unrelated. ABSOLUTELY unrelated. I can create the same exposure value with many differing DOF results. Period. f16@1/100 gives one result, while f8@1/400 gives another, and f4@1/1600, all with identical EV but different DOF.

Correct. Note however, that DOF and exposure of an 8.5mm lens at f/1.9 on the Q camera are identical to that of an 47mm at f/10.5 on a full frame camera.

No, it's not. f10.5 is NOT f1.9, no matter what format you're using. I can use the same meter to meter for my 4x5 view camera, my 35mm film camera, my APS-c DSLR, and the Q, and use the *same setting* on every one. F1.9 @1/250, ISO200. Independent of format.

Is it a wide lens?
The focal length of 8.5mm suggests that it is an ultra wide, provided you are using a full frame or APS-C sensor as a reference.

Of course you know that 8.5mm is not ultra-wide within the Q-system. You know that in order to judge its AOV in terms of the full frame "normal" reference of 50mm (or 43mm), you need to convert by the crop factor. You have to do the same to judge the lens speed. Just as "8.5mm" only sounds ultra-wide, "f/1.9" only sounds fast. The lens is slow, no arguing about that.

That's just wrong. f1.9 may not be *fast*, but it's not *slow*, either. Is a 50mm f1.9 slow? You can take pictures with the same exposure with both lenses.

[bwDraco]

You are contradicting yourself with these two statements.

First of all, neither focal length nor aperture diameter of a lens change with sensor size; they cannot, they are physical properties of the lens.

One often, however, one talks about lenses regarding their full format equivalent. For instance, the full format equivalent of a DA 70/2.4 would be a 105mm f/3.4.

The 8.5mm f/1.9 lens has its physical properties that are independent of sensor size. Some people like to convert the focal length, however, into a full format equivalent focal length (-> 47mm). That's fine but the specification "47mm f/1.9" is nonsense. This refers to a lens with an aperture diameter of 24.7mm. We both know, however that the 8.5mm f/1.9 lens only has an aperture diameter of 4.5mm. So if you say "47mm" you have to say "f/10.5" as well. Any mixing doesn't make sense. Stating that an f-ratio applies to DOF but not to exposure doesn't make sense either. Exposure is influenced by the aperture diameter which is why it would be wrong to upgrade it from 4.5mm (8.5/1.9) to 24.7mm (47/1.9).

In the first sentence, I meant 35mm equivalent, not the physical aperture. Furthermore, while this is true and makes sense in evaluating depth of field, the physical aperture ratio is always the value that determines exposure. As I said before, a smaller sensor gets less total light, but since it is spread over a smaller area, the exposure is the same.

Smaller sensors do not have a high ISO disadvantage. APS-C does not lose a stop of high ISO performance against FF. FF sensors have a larger area to collect light, but the need to spread the light out to this larger area reduces its intensity per sq. mm, cancelling out the larger area advantage. Please see the "Low noise benefit of FF vs APS-C equals ... zero" thread.

Not true. With a full-frame lens, an APS-C sensor simply crops out the area of the image circle that would otherwise be imaged by a full-frame sensor. With the same exposure settings, the amount of light that would be detected in the center of the APS-C and full-frame sensors is the same. An image sensor does not spread or concentrate light from a lens; it simply detects the amount of light coming from the lens, within the region of the image circle it covers*.

The light coming from the lens is in the form of an image circle. An image sensor does not make this circle larger or smaller and does not alter the image projected by a lens; it only reads it*. The lens and sensor are independent imaging devices. They work together to produce an image, but the sensor has no effect on the lens output.

--DragonLord

*An image sensor has microlenses that ensure that light reaching the sensor is read efficiently, and whose efficiency may vary with sensor design, but this is a matter of sensitivity and high ISO performance and has no impact on field of view or the actual amount of light reaching the sensor, just how efficiently it detects this light.

[jstevewhite]

You are contradicting yourself with these two statements.

First of all, neither focal length nor aperture diameter of a lens change with sensor size; they cannot, they are physical properties of the lens.

One often, however, one talks about lenses regarding their full format equivalent. For instance, the full format equivalent of a DA 70/2.4 would be a 105mm f/3.4.

ONLY for DOF purposes. It's still going to have the FOV of a 105mm, and your exposure is going to be f2.4, but with the DOF of a 105mm FF at f3.4.

The 8.5mm f/1.9 lens has its physical properties that are independent of sensor size. Some people like to convert the focal length, however, into a full format equivalent focal length (-> 47mm). That's fine but the specification "47mm f/1.9" is nonsense. This refers to a lens with an aperture diameter of 24.7mm. We both know, however that the 8.5mm f/1.9 lens only has an aperture diameter of 4.5mm. So if you say "47mm" you have to say "f/10.5" as well. Any mixing doesn't make sense. Stating that an f-ratio applies to DOF but not to exposure doesn't make sense either. Exposure is influenced by the aperture diameter which is why it would be wrong to upgrade it from 4.5mm (8.5/1.9) to 24.7mm (47/1.9).

Again, if you use a meter this way, you'll overexpose. Massively. Equivalence is bull anyway - it's an analogy, not a 'fact'. It's like me saying all of your FA LTD are slow glass because my 200mm f5.6 Rodenstock for 4x5" makes the FA43 a 172mm f7.6., or my 12 inch f8 Dagor makes it a 400mm f16. It's a nonsensical statement. "FAST" is about how much light it lets in; it so happens that in larger formats, fast glass also means near-zero DOF, but this factor increases beyond APS-c or 35mm too; it's *silly* to use it that way.

Smaller sensors do not have a high ISO disadvantage. APS-C does not lose a stop of high ISO performance against FF. FF sensors have a larger area to collect light, but the need to spread the light out to this larger area reduces its intensity per sq. mm, cancelling out the larger area advantage. Please see the "Low noise benefit of FF vs APS-C equals ... zero" thread.

No, "smaller sensors" don't, but smaller pixels *do*, and since putting the same number of pixels on a smaller sensor means smaller pixels, in this case, it does. I didn't read that thread - I don't need to. I've read the physics. With the same technology, larger pixels have higher DR, and lower noise than smaller pixels, period. I'm well aware that much of the magic of the D700 comes from clever noise reduction, but that doesn't negate the PHYSICS. That thread might debate the real-world relevance of the differences, but the *fact* is that, all other things being equal, smaller pixels mean less signal-to-noise ratio. Now different sensor technologies may have different sensitivities to start (iso 50/100/200), but in the same technology, it remains true that smaller pixels mean less signal-to-noise.

[Laurentiu Cristofor]

AFAICT, the only people that equated the Q with a DSLR were those that were criticizing it for not being one. This is called strawman argument.

Good thing it's not an argument I'm making then, but only one you presented. Hold on a minute -- doesn't that mean it's you presenting the straw man?

And what strawman argument is that? Certainly not the one where I criticize the Q for not being a DSLR?

Here's what you said a few pages ago:

All this proves is that no-one needs the Q -- they can just use their phone. For anyone wanting better, a larger sensor is a necessity. Also your DOF argument is flawed. A small sensor forces you to use diffraction-limited larger apertures. With a larger sensor you actually have a choice -- stop down or not.

And now:

Really, I am beginning to wonder why you take every comment as a reason to criticise and be offensive. Especially when you are so darned bad at it.

I don't criticize every statement made on this thread and I am not trying to pick on you in particular. But you should know better that criticizing a camera for not being what you want. Your earlier statements that I quoted above include at least 3 factual errors: 1) some people need compact sensor cameras or there wouldn't be a market for them if they could just use their phones 2) the Q has a larger sensor than phones do and 3) you aren't necessarily able to use small apertures on an SLR for an equivalent result to that of a compact, because the high ISO setting needed for equivalence might not be available to you. You called my arguments flawed, you implied I'm a biased Pentax fan, you ended up resorting to fallacies in your rhetoric, and now you're offended that I asked you to grow up and judge a camera against its peers. And now the thread is derailed into that blackhole provided by topics like DOF and lens equivalence. I'll have to step aside and watch the show now.

[conradj]

No, "smaller sensors" don't, but smaller pixels *do*, and since putting the same number of pixels on a smaller sensor means smaller pixels, in this case, it does. I didn't read that thread - I don't need to. I've read the physics. With the same technology, larger pixels have higher DR, and lower noise than smaller pixels, period. I'm well aware that much of the magic of the D700 comes from clever noise reduction, but that doesn't negate the PHYSICS. That thread might debate the real-world relevance of the differences, but the *fact* is that, all other things being equal, smaller pixels mean less signal-to-noise ratio. Now different sensor technologies may have different sensitivities to start (iso 50/100/200), but in the same technology, it remains true that smaller pixels mean less signal-to-noise.

Yeps. The way i've come to look at it, a lower noise floor gives you cleaner shadows with better chroma info in them, whereas a larger pixel will tend to get you more highlight headroom.

[Aristophanes]

And now the thread is derailed into that blackhole provided by topics like DOF and lens equivalence.

Derailed to that non-subjective field of optical science!

BTW--the USA Pentax website has the Q here:

Digital SLR Cameras - Official PENTAX Imaging Web Site

...but they call the header "ILC" for the Q, K-R, K-5, and 645D. They are not positioning it with compacts.

[Class A]

Smaller pixels mean less dynamic range, all things being equal:

Less dynamic range per pixel, yes, but that doesn't matter one iota for the photographically relevant dynamic range for the whole image. The article you reference, talks about pixel noise, not image noise.

Larger sensors increase dynamic range of images, pixel size is irrelevant for image noise and dynamic range. See the DxOMark article "More pixels offset noise!".

If you have a full frame camera and a 47mm lens, and I have a Q with its 8.5mm lens, and a Pentax Spotmeter says our correct exposure is f1.9 @1/250 for ISO200, we will both use f1.9, 1/250th, and ISO200, and we will both get accurate exposure.

This is an apples vs oranges comparison. Take a DOF calculator and calculate the DOF for the respective images. The DOF values will hugely differ. You need to stop down the lens on the larger sensor camera in order to get the same DOF. This means that you either have to increase the exposure time and/or the ISO value as well. Comparing images which are vastly different, doesn't make sense.

Again, if you consult a DOF calculator, you'll see that the 47mm lens on a full-frame camera need not be quicker than f/10.5 to allow you to take all the images you can take with a Q-system 8.5/1.9 lens.

You may want to consult the "Equivalence" essay by Joseph James.

DOF and exposure are unrelated.

Both are influenced by the aperture diameter.
Of course one has reciprocity for exposure, but that doesn't change the fact that -- for a given shutter speed and ISO setting -- the lens creating less DOF will also create the higher exposure.
Please check out the "Equivalence" essay.

That's just wrong. f1.9 may not be *fast*, but it's not *slow*, either. Is a 50mm f1.9 slow?

A 50mm f/1.9 is not slow, AFAIC. In order to be able to take the same images (DOF, exposure, given a fixed shutter speed and iso setting), you'd need a 9mm f/0.3 lens in the Q-system. Such lens does not exist. This is the reason why the Q-system is limiting.

"FAST" is about how much light it lets in; it so happens that in larger formats, fast glass also means near-zero DOF, but this factor increases beyond APS-c or 35mm too; it's *silly* to use it that way.

It "so happens"?

Ever wondered why increasing DOF means you get less light in? I repeatedly pointed out that fuzzy OOF areas are created by additional light. When the aperture diaphragm in a lens closes down, it reduces the amount of light coming in, reducing EV and increasing DOF at the same time. And yes, your 4x5 lenses are a lot faster than most FF/APS-C lenses. This is why you can create such shallow DOF. It is not the larger sensor/film creating the DOF. It is not the larger image circle. It is the large aperture diameter that creates the thin DOF. On a larger format, you need longer focal length to take equivalent images (same subject-camera distance, same AOV). If you keep the f-ratio constant (say f/1.9) this means the quotient "focal-length/1.9" gets larger, the larger the format. This, in turn, implies that the DOF gets thinner.

It so not "so happens that in larger formats, fast glass also means near-zero DOF".

No, "smaller sensors" don't, but smaller pixels *do*, and since putting the same number of pixels on a smaller sensor means smaller pixels, in this case, it does.

Smaller pixels on a smaller sensor means you have increased pixel pitch. Pixel pitch is irrelevant for image noise (not for pixel noise but who cares about individual pixels?). You don't need to believe me; just consult the DxOMark article "More pixels offset noise!".

I didn't read that thread - I don't need to. I've read the physics. With the same technology, larger pixels have higher DR, and lower noise than smaller pixels, period.

Again, you are referring to pixel noise which is photographically irrelevant. What counts is image noise. Do yourself a favour and read the thread + the DxOMark article.

Class A, your understanding of optics appears to be flawed.

I offered some insights into the topic. If you prefer to keep your incorrect world view, that's your perogative.

Maybe you want to consult the sources I supplied and reconsider your assessment of my understanding.

I will not further comment on incorrect statements regarding format discussions in this thread. I have provided sufficient independent sources to support my arguments and will not repeat them.

[bwDraco]

I offered some insights into the topic. If you prefer to keep your incorrect world view, that's your perogative.

Maybe you want to consult the sources I supplied and reconsider your assessment of my understanding.

I will not further comment on incorrect statements regarding format discussions in this thread. I have provided sufficient independent sources to support my arguments and will not repeat them.

Class A, I apologize about this statement, and it has been removed. I will see what other experts say so that this matter can be settled amicably and we can resume on-topic discussion. I only want to ensure that discussion on these forums is civil and polite, and I clearly do not wish to start a flame war.

--DragonLord

[Laurentiu Cristofor]

Derailed to that non-subjective field of optical science!

The field may be objective, but discussions are as subjective as any other.

[jstevewhite]

I wrote out a novel, but then pared it down to the topics relevant to this thread:

A 50mm f/1.9 is not slow, AFAIC. In order to be able to take the same images (DOF, exposure, given a fixed shutter speed and iso setting), you'd need a 9mm f/0.3 lens in the Q-system. Such lens does not exist. This is the reason why the Q-system is limiting.

THIS is the statement that's important. It's factually incorrect. Period. I'm not trying to be insulting; this is just WRONG, and I don't think it should be allowed to go unchallenged here. You would have to use a MUCH higher shutter speed with the 9mm f/0.3 lens in order to achieve the same exposure. I think f0.3 would be five stops overexposed if you used the same shutter speed.

Think about it. Have you ever used a light meter? I have. Many times. And they DON'T CARE ABOUT FORMAT. If your thesis (and Mr James) was correct, then when I metered that f1.9 ISO100 1/100 sec, the meter would have to know whether I was using a 4x5, a 6x6, a 35mm, an APS-c, or a Q. But they don't. f1.9 is f1.9, no matter what your friend says. Really. Yes, DOF will be different. It will be different between a 35mm and a MF (with equivalent FOV), and between a MF and a LF (with equivalent FOV), but exposure would remain constant across them all. That includes the Q.

Your friend Mr. James even says it: "Many make a big deal out of the fact that the same f-ratio and shutter speed result in the same exposure, and thus the claim f/2 = f/2 = f/2. " (emphasis mine)

When the aperture diaphragm in a lens closes down, it reduces the amount of light coming in, reducing EV and increasing DOF at the same time. And yes, your 4x5 lenses are a lot faster than most FF/APS-C lenses. This is why you can create such shallow DOF. It is not the larger sensor/film creating the DOF. It is not the larger image circle. It is the large aperture diameter that creates the thin DOF. On a larger format, you need longer focal length to take equivalent images (same subject-camera distance, same AOV). If you keep the f-ratio constant (say f/1.9) this means the quotient "focal-length/1.9" gets larger, the larger the format. This, in turn, implies that the DOF gets thinner.

Dude, no offense, but that's not what "fast" means. A 135mm f3.5 lens can create tiny DOF when focused close, but that doesn't make it "fast". "Fast" in general use referred to the ability to use faster shutter speeds because of large apertures; it makes absolutely no sense to suggest that they were called "fast" lenses because they made shallow DOF. The Q does not have a 'slow' lens.

I offered some insights into the topic. If you prefer to keep your incorrect world view, that's your perogative.

Maybe you want to consult the sources I supplied and reconsider your assessment of my understanding.

I will not further comment on incorrect statements regarding format discussions in this thread. I have provided sufficient independent sources to support my arguments and will not repeat them.

Suit yourself; it was not my intention to offend you, but to correct inaccurate statements. I read the sources you referenced. That doesn't change the fact - relevant to the Q - that you've made factually incorrect claims. I'm not trying to start a flame war, either, just correct a specific misstatement about the product in question - that blew up into other misstatements. The lens on the Q is NOT slow, although you're right about one thing, you won't have any razor thin DOF snaps with it.

[pinholecam]

8mm f1.9
47mm equivalent FOV with DOF of f10.5 and aperture size of f1.9 can be good in low light.
Sometimes 43mm f1.9 with DOF of f1.9 on my K5 in low light can be just too little DOF.



Don't understand all this equivalent speed of the aperture discussion.
I've got an A590 with a f2.5 lens. On a bright sunny day the sunny16 rule still applies to it. It certainly does not ask for an exposure of f2.5 (since f2.5 is already equivalent to f14) ISO100 1/100 on a bright sunny day. It asks for f16 ISO100 1/100.

[bwDraco]

Okay, so here's what Falk Lumo said when I PMed him:

There are "cameras" and "35mm equivalent cameras".

Most people are confused and fail to make the distinction clear enough.

A 35mm equivalent camera will have:

• A focal length multiplied by the crop-factor
• An f-stop value multiplied by the crop-factor
• An ISO setting multiplied by the crop-factor !! (forgotten by most people)

This means that a camera and the "35mm equivalent camera" would both expose correctly if using an external light meter.

It can be shown (in a mathematical way) that a camera and a "35mm equivalent camera"have

• identical depth of field
• identical aperture diameter in mm
• identical field of view
• identical diffraction resolution limit
• identical noise floor with a given CMOS and camera electronics technology

This means, that images taken by a camera and the "35mm equivalent camera" are indistinguishable! (which is why it is called equivalent!).

Also, the equivalent camera won't have a much different lens price or size or weight.

Where the larger sensor shows its benefit is where it allows options which aren't available with a smaller sensor:

• Otherwise impossible physical lens aperture
• Otherwise impossible resolution for a given (large) physical lens aperture
• Otherwise impossible equivalent ISO level (full frame ISO100 is ISO25 for FourThirds)
• Otherwise impossible dynamic range.

Feel free to city my PM with proper credit if you like.

Kind regards,
Falk

This is consistent with the following post:

Class A is just saying that the image you get from the 8.5mm f/1.9 ISO 100 1/250 shot taken with the Q would be identical in all respects to a shot taken with a 47mm f/10.5 ISO 3200 1/250 using a FF camera. Sensor technology differences and other issues being assumed negligible (it's theory, not practical comparison). That is the lens equivalence point he is making.

The problem is that we had all failed to consider ISO in the 35mm equivalence discussion. I hope we can get back to on-topic discussion now.

--DragonLord