PetaPixel

Tony Northrup Makes Corrections, Replies to Critics in Part 3 of Camera Spec Debate

In this third — and final? — installment of his “camera companies are lying to you about crop factor” series that has caused such a stir, Tony Northrup hopes to silence the critics, make some corrections, and overall bring this debate to a close.

For those of you who haven’t caught up, there’s been both a part one and part two of this series, where Northrup attempts to expose camera manufacturers for allegedly misleading consumers by not properly calculating sensor size into various specs.

Screen Shot 2014-05-27 at 11.06.58 AM

Both pieces started a rather intense debate, with everyone and their brother (and sister and mom and first cousin twice removed) attempting to either debunk Northrup’s accusations or back them up.

Part three comes in at a whopping 26 minutes, so it’s not exactly a quick watch, but if you’ve been keeping up with the series and debates, you’ll definitely want to watch it to better understand how all of this has played out and decide whether or not you feel Northrup makes a valid point.

So press play and do your best to go into it with a fairly open mind… or, you know, get further entrenched in your own view… that’s also an option.

(via Tony Northrup)


 
Get the hottest photo stories delivered to your inbox.
Get a daily digest of the latest headlines:
  • http://www.flyingsuicide.net/ Oj0

    As I said, they should be sold as having an aperture which is physically f/x with the depth of field of f/y.

  • Terry_Clark

    Another moron catapulted to “stardom” by the internet. Without the world wide web this guy would never receive a moment of notice. Bottom line, who cares? Go make pictures and stop pixel peeping!

  • Rob Elliott

    That isn’t how F stop works though.. F-stop is a calculation based on the size of the opening based on the sensor/film size. It is a scaled measurement. Depth of Field is not related (directly) to the f-stop, but the plane of focus which is distinct from f-stop. The GHz on a computer processor is constant weather the processor is 32bit or 64bit. The performance of a 32bit vs a 64bit processor at the same Ghz isn’t the same though.

    Or Horsepower as a relation to speed. 625 Horsepower will take a Mclaren F1 to 200mph fairly quickly. on a Bullet train 22,000hp will only get it to 185mph.

    It isn’t dishonest to list the Horsepower.

    If he were advocating the creation of a new DOF stat it would be different. Instead he is accusing manufactures of smaller sensors of lying to consumers to trick them.. by calculating the most varied stat, to give a more standard understanding, between the wildly varying sensor sizes in compact cameras with zoom.

  • Rob Elliott

    Depth of field is caused by the Plan of Focus. This isn’t directly caused by f-stop but the physical size of the aperture. low light performance is also caused directly by the physical size of the Aperture allowing more light in on top of the normal sensitivity of the sensor and speed of shutter.

    F-stop is calculated by the size of the aperture in relation to the exposure surface (sensor/film)

    ISO is calculated as a sensitivity rating, but the noise is caused by Pixel density to sensor size ratio, more modern sensors have better noise performance with high pixel densities.

    Focal Length is also a set calculation. The actual perspective is effected by sensor size.

    Depth of Field is not a calculated measurement given on any camera.

  • Rob Elliott

    Horsepower is the engine can do the same job as 1 horse. 625 horse power.. 625 horse.

    that means you need to put 625 horse inside the F1 which is smaller running on a treadmill to create the same power. Of course you can’t fit 625 horses in the F1 body, thus to have a horse powered engine that can produce the same speed as a F1 you’d basically need 22,000 horse (the power of a bullet train that still can’t match top speed or performance)

    It isn’t really any difference.

    The Plane of Focus is determined by the physical size of the Aperture not the f-number which is calculated by the size of the Aperture in relation to the exposure surface.

    Focal Length is a set Calculation, the field of View is determined by the Focal Length on relation to the Exposure size. Which easily scales.

    It is one of the reason why there is more Kw numbers showing up on cars because that is replacing horse power (slowly) as the standard, as we haven’t been using horses to power very much in a century. Most people if you ask what Horsepower in a car means they will assume if you strap 625 horse to the front it will work the same, which isn’t the case. A Company giving the stats of a camera isn’t a lie.

    And no I don’t think Canon is implying that you will get the same… I don’t know the math in adjusting plane of focus of the top of my head. The S120 is a 5.2-26mm lens. Its maximum aperture is f1.8. Of course f1.8 on a tiny sensor doesn’t have the same plan of focus as a big camera.

    However at 20-25mm you can create a shallow Depth of field. Just like I can with a APS-C at 125mm and f9. It’s Marketing.

  • http://www.flyingsuicide.net/ Oj0

    It’s more like saying the F1 car’s 625 horsepower is equivalent to 22,000 horsepower without clarifying that it’s equivalent to 22,000 horsepower in a several ton train. It isn’t dishonest to list the actual aperture and actual focal length, but if you’re going to convert the focal distance to 35mm terms to explain the field of view you’ll get you should convert the aperture to 35mm terms to explain the depth of field to be expected. The Nokia Lumia 1020 has a camera with an aperture of f/2.2 and a 35mm equivalent or 26mm but isn’t going to get anywhere near as small a depth of field as an 26mm f/2.2 lens on a 35mm camera when distanced for the same composition.

  • David Bateman

    Yes!
    It really bothers me when people jump to use Aperture to equal depth of field. Aperture DOES NOT EQUAL depth of field.
    Depth of field will depend on focus distance to subject. So a four thirds lens with 25cm minimum focus distance can get closer to subject and have twice the field of view or tighter angle than a 135 format. So Yes you can get narrower depth of field with four thirds than you can with 135 format. I actually find this a problem, and you can see it shooting macro. Get really close to your subject, now you need to scale you aperture to f16 to keep what you might want in focus.
    Take 90min and learn the limits and advantages of your camera. If you have the money compare different cameras to see what gives you the image your looking for.
    Do not watch this guy.

  • David Bateman

    No. Because you don’t know the distance to subject! This will effect the depth of field. The company is not going to know what your doing.
    The panasonic 35-100mm f2.8 lens has minimum focus distance of 85cm. So your field of view at 100mm will look like 200mm lens on 135 format. Set your aperture to f2.8. Now a Nikon 70-200mm f2.8 VR2 lens has a minimum focus distance of 1.4m So the closest you can get to your subject is 1.4m. If you can frame it you will have narrower depth of field with the Panasonic than the Nikon 135 format.

  • GeniusUnleashed

    “F-stop is calculated by the size of the aperture in relation to the exposure surface (sensor/film)” Who taught you this? They have no idea what they are talking about. F-stop has zero to do with sensor size, it’s the relationship between focal length and lens diameter. That doesn’t change if I take my 35mm f/2 and move it from a 35mm sensor to a camera with a smaller sensor size.

    It DOES change the DOF, however as well as the viewing angle. That’s his whole point! He’s simply asking camera manufacturers to include the DOF change as well as the angle difference (that they already advertise) when they sell products. They convert the lens difference for smaller sensors because most people are familiar with 35mm standards. He’s simply asking them to tell us the DOF will change as well, that’s it. Whatever else you think he’s saying, let it go, because that’s ALL he’s saying.

  • GeniusUnleashed

    Or watch his guy and try to understand him. You don’t seem to understand him…which doesn’t make him wrong.

  • GeniusUnleashed

    No, Rob, Fstop is not calculated that way. Who told you that?

  • Rob Elliott

    The difference is that the Aperture is a set thing and the f number is something isn’t going to be understood unless they understand cameras better then many DSLR shooters.

    The Focal length conversion to 35mm is a point of reference number, and is always listed as a point of reference.

    it would be dishonest if they said a camera was a 50-300mm lens f2.8 to 5.7. When it was really a 6-24mm f2.8 to 5.7.

    The fact is the Depth of Field is not caused by the f-stop but the Focal Plane of the camera. This is where the confusion is coming from.

    f2.8 is a calculation based on Aperture and the exposure surface… the Depth of field is caused by the Focal plane of the exposure surface, which is effected by the size of the aperture and the focal plan changes caused by the elements of the attached lens.

    For example at 150mm even at f9 if you bring the subject to the front of the focusing distance of the lens while the background is far out the you will get a softer focused background. But if you take them close to the back of the focal plane the background will be in focus, all without changing the aperture.

    The smaller the sensor the focal plane is naturally much larger. This has nothing to do with the f-stop directly.

  • GeniusUnleashed

    Not sure you fully understood Tony’s premise. Question 9 is his ONLY point of making these.

  • GeniusUnleashed

    Who cares? I think everyone spending thousands of dollars on equipment cares. It must be nice to be able to throw away tons of money without a care.

  • Rob Elliott

    I’m sorry you are right it’s the focal length. It’s late and I’m not an expert.. which is why I don’t make videos. The reasoning is right.. the info is off. My apologies.

  • GeniusUnleashed

    All good. Tony is (not so simply) asking for a comparable DOF as well if they’re going to show the comparable (35mm) lens viewing angle difference.

  • Rob Elliott

    The Aperture is a factor that changes the size of the Focal Plane yes… however DOF is determined by where objects are in relation to the focal plane.

    the f-stop is not connected to focal plane but is instead a numerical value based on Aperture size in relation to the focal length of the lens.

    F-stop is not directly related to DOF which is why f2.8 doesn’t have the same DOF on every sensor size.

  • Rob Elliott

    But what he is asking for is impossible because the f-number isn’t related to that. It’s the Focal Plane which is the point I’m making. here.

    this picture is of goslings, taken at f5.6 (6.7 full frame) at 200mm (300mm full frame)

    If you had a 60mm f2.8 lens at f2.8 from the same location you would have a wider field of view, and even at f2.8 a wider focal plane when focusing to the back like this.. and as such the mother goose at the back will be in perfect focus.

    To bring the Goose in the back into the same focal plane I’d have to stop down to f9-13 maybe further, to close the aperture and widen the focal plane.

  • GeniusUnleashed

    More or less correct. Fstop is the result of the focal length/lens diameter relationship, which doesn’t change based on the camera you use if stick with the same lens.

    His premise is that, when you do (35mm) conversions, everything has to change if you change the focal length, because they are all part of the equation. So if you show the 35mm focal length equivalent, then they need to tell us the 35mm fstop equivalent, and then let us know the 35mm DOF equivalent so we can make informed buying decisions.

  • GeniusUnleashed

    That’s fine, but Tony’s talking about Apples and you’re bringing Oranges into the conversation when they aren’t relevant to his point.

    His conversation is about comparing same lenses on different sensor sizes. And then using different lenses with the same equivalent focal lengths and how the DOF changes because of the sensor size. That’s all.

  • Rob Elliott

    Ya but he’s not, he is saying you should call a 30mm f-2.8 lens a 45mm f3.5 lens equivalent. why? Because of DOF. which the F number has no relation to.

    the reason you see a 45mm equivalent is so people can easily find the lens field of view they want. if you want a 85mm field of view for your NEX.. you don’t buy a 85mm NEX lens you instead by a 57/60mm lens because it is about 85-90mm equivalent. It was caused likely because people asked a lot. people don’t ask about the f-stop. Which is not a direct cause of DOF

    What would it say…

    30mm f2… 45mm equivlient. f2.8 equivilient DOF when shot wide open with the subject near the front of the focal plane?

    DOF is a separate thing.. and if he is saying create a new DOF measurement for all camera’s that would be one thing… but saying camera manufacturers are lying because they are only converting the number that people ask to be converted… is not a lie. Particularly when you are saying Focal Plane and f-stop are directly connected..

    Manufactures are not lying to anyone when they legitimately calculate states and provide them to people.

  • GeniusUnleashed

    “people don’t ask about the f-stop.” Probably not a good idea to speak for other people, especially when you’ve already stated you’re not a professional. I ask about fstop and every professional I work with asks about fstop because 99 out of a 100 times it speaks to the level of craftsmanship and quality of the lens. That’s why L lenses are more expensive, because they have larger apertures and better quality, ie. not plastic. It also tells me, that when used on a 35mm camera, I’ll have great dof control which is the most important factor for professional photographers and pretty much anyone who takes photography seriously as an art form. When given all the information, it allows us to choose the best tool for the job, which sometimes demands APS-C sensor cameras over 35mm.

  • sus2

    Your understanding of physics is seriously flawed.

    You can ignore, perhaps, transmission and diffraction losses, but you CANNOT ignore the most fundamental law of radiated energy: the inverse square law.

    Every time you double the distance, you quarter the energy of the light.

    So, in your example, you claim that 4x more light reaches the larger sensor, but this is incorrect, because by the time it gets there, it will have travelled twice the distance, and the intensity will have reduced to 1/4 (or 1x), or exactly the goddamned same luminous flux reaching both sensors.

    That’s why the f-number is expressed as a ratio. It’s dimensionless and scales with focal length.

  • Joakim Jaldén

    “that means you need to put 625 horse inside the F1″ – no it does not, and you know that. Let’s skip the analogies. They are not really helping.

    Regarding the shallow dof argument you made. First of all, you cannot get the APS-C equivalent of 125mm on the S120. 26mm on the S120 gives you the same field of view as a 75mm on the APS-C (crop factor 1.6). 26mm with the maximum aperture of f5.7 at this focal length gives you the same dof and angle of view as 75mm on the APS-C with f16. I did an actual experiment just a few minutes ago with these setting with both the S120 and my 60D, at about 2m to the subject and 200m to the background, and “dramatic soft background” is not what comes to mind. Slightly out of focus yes, but dramatic and soft no. Even at the 1.8 on the S120 (limiting me to not more than just over 5.2mm focal length) still do not manage any better than slightly out of focus in my view. You don’t have to take my word for it though, as it is subjective. Have a look at cameralab’s review of the S120 to get a feel for what type of background blur you can get and let me know if you find this “dramatic soft”.

    Also, just so you know, I understand the physics and math of field of view and aperture. What I don’t understand is how some can be completely fine with a 35mm equivalent focal length (in terms of field of view) which is not a direct physical quantity either, and so fundamentally opposed to the 35mm equivalent aperture (in terms of dof).

    Finally, maybe Canon is not even the best example here. Take the Olympus Stylus 1 in the video. Why would they write that it has a Zuiko Digital Ultraslim 28-300mm f2.8 and superimpose it on the shadow of a real 300mm f2.8 lens if they were not implying that this is what you would get in a pocketable format? What is your theory?

  • Joakim Jaldén

    David, you don’t need to know the distance to the subject. 50mm f2.0 on a crop factor 2 camera will give you the same field of view and dof as 100mm f4.0 on a full frame, regardless of the distance to the subject. You are right though that the dof depends on the distance to the subject, but this dependence is the same for the 50mm f2.0 crop factor 2 setup as for the 100mm f4.0 setup, so you do not have to specify the distance in advance for the comparison to work. 100mm f2.0 will give you a different dof at all distances though, and this is why Tony argues that it would be better to simply say 100mm f4.0 full frame equivalent.

  • noisejammer

    @sus2:disqus
    Viewed from the same distance, light from a subject presumably has the same intensity. The reason you get 1/4 of the light is nothing to do with distance – it is that the APERTURE of a 42.5/2.8 lens is half the APERTURE of a 85/2.8 lens. Half the APERTURE = 1/4 the area. 1/4 the area => 1/4 the light.

    Can I suggest that you head off to high school for a couple of years before you comment further? Reading lessons might help. Do try to pass them.

  • David Bateman

    No that’s still wrong. I think the problem comes down to our own lazyness. A lens has specific properties that are described by the aperture and the focal length. Thats the lens. What we really need to do and Companies for that matter should do (so partial “lying” of companies) is specify the field of view for a lens in degrees. A nikon 50mm f1.2 lens on a 135 system will have x degrees (field of view) and on four thirds will have y degrees and on Nikon 1″ sensor will have Z degrees.
    The DOF will be harder to define but could be described for a reference subject x meters away from the camera to have y cm of “acceptable” focus than w cm of “acceptable” focus for four thirds and z cm of “acceptable” focus for Nikon 1″. The problems are what is acceptable focus? What should be the reference distance to subject? And a bunch of others.

  • Joakim Jaldén

    “No that’s still wrong” – In what sense is it wrong? Surely not in terms of the math? As long as the subject distance is a few times larger than the focal length the DOF is more or less the same as the DOF of an full frame (FF) camera set the FF equivalent focal length and the FF equivalent f-number. There are essentially two ways to convince you of this. Either you watch the video again and look at the examples (or even better, do the comparison yourself with your own cameras) or, if you have a “math-mind”, I have prepared and attached a proof for you which show this (as an image just because typesetting math in text is a pain).

    At the end of the day, describing the DOF in terms of an FF equivalent focal distance and FF equivalent f-numer gives a very accurate description of the DOF in terms of a FF comparison. The downside is of course that one has to have a feel for what the FF pictures will look like, but the specification of FF equivalent focal distances are already doing this for the angle of view. The big benefits over the system you are describing are however that one does not have to decide what is acceptable focus (this is somewhat subjective, although there is research on that too), one does not have to define a reference distance (the equivalence works at all medium to large distances), and it is more economical in terms of text (everything is captured by the crop factor). Note here also that acceptable focus should be determined by the final print (at the same size for both cameras), and then recalculated for the sensor, and not the other way around. Hope that helps.

  • Joakim Jaldén

    Having prepared and posted this, I realized that Wikipedia actually already have quite a nice discussion of this under the section “DOF vs. format size” in the “depth of field” article, so you may in the end be better off just reading that.

    I also made a small error in the last sentence of my attached picture. You also have to assume that he additive f term in the hyperfocal distance is neglectable, which then gives you the approximative formula used before.

  • David Bateman

    Why I say its wrong, and it is. Is because aperture strictly affects exposure. So saying its f2 on 135 format and f4 on four thirds is wrong, because its not and you will blow out your image. A light meter helps in this regard and we need to know the true aperture of the lens.
    Why this is an issue is because we expect a certain depth of field at a given aperture. But depth of field is not strictly determined by the aperture. There are many things that will effect DOF, some of which you elude to in your image above. However why your image doesn’t have an equal sign and uses an approximation sign is because of harder to quantify lens characteristics that do effect DOF. I say we expect a given DOF for an aperture and this is true. Some well documented lens will have a hard blur line, so things will be tightly in focus in the acceptable range then fall off hard to out of focus, for some lenses. Other lenses will have a large acceptable DOF range (ie like using it on four thirds but your on 135 format) for a given aperture. This has been talked about a lot and mixed in with “good” bokeh characteristics of some lenses.
    This is why I say to take 90min and learn your equipment! Or take 90min and learn about optics and trade offs using different lens elements in a lens, which will affect these properties. Or use those 90mins to read some of articles on The Online photographer (Specifically from Ctein) whom is a physicist and some times talks about these fun qualities in lenses. Don’t waste those 90min watching this guy.

  • Joakim Jaldén

    “why your image doesn’t have an equal sign and uses an approximation sign is because of harder to quantify lens characteristics that do effect DOF” – no, no, no, that is not at all why there is an approximation sign and you would know this if you would have cared to actually read the Wikipedia article. This DOF formula is based on a calculation that is based a number of axiomatic assumptions like “light travels in straight lines”. The exact (non-approximative) formula based on these premises was actually given in the attached image, and you can find the derivation and motivation for it on Wikipedia. Then there are exactly two approximative steps, one in that it is assumed that s-f is well approximated by s (subject distance dominates focal lenght, which is why the formula is a worse approximation for macro photography) and another one in that the f-term in the hyperfocal distance is neglectable (which is actually quite a good approximation). That is it. No more, no less. You could actually work out exactly how good the approximation is for particular settings if you were interested, but something tells me you are not really that interested in understanding this and that I am wasting my time arguing with you.

    If you are going to reject these idealized optics arguments as indicative of the performance of real lenses and cameras you may as well reject the notion of a focal length all together as this is also an idealized assumption (surprised?). No real lens is actually capable of focussing the light to a single focal point if you really wish to be picky at this level. Rejecting these calculations because of different bokeh on different lenses so fundamentally misses the point that it is not funny.

    And why do you even bring up the exposure red herring? I though we were discussing depth of field and not exposure and light meters. No one but you is claiming that “its f2 on 135 format and f4 on four thirds”. What was argued is that “its f4 on a 135 format and f2 on the four thirds”. Just for kicks though, make it “f4 on a 135 format with ISO 400 and f2 on the four thirds with ISO 100″ and you would actually get the same dof AND exposure at the same shutter speed, but I guess that you already knew that even though you apparently did not waste 90 minutes on these videos.

    Now I am going to calm down and spend another 90 minutes to learn my equipment as you suggested.

  • Foot of our stairs

    This is about the best explanation of this I have seen, I have seen this stuff in forums but often the realities are obfuscated by squabbling and intellectual bullies.

    However in the end it is also a bit useless as it takes 3 lengthy videos to get across I maintain that “smaller sensors are more noisy and have more depth of field” is actually a  real world truth even if it is poor technical explanation 

    It only makes sence to go beyond this if you want to hold “quality” as constant across formats or make technical explanation as to why that is aparantly the case ( as here) 
    It the real world we set the exposure to suit  the conditions we want to photograph and we buy / cary the equipment that suits the pocket and circumstance.

    So more is always better but at some point what you get is good enough, even phones take OK snaps in good light, micro four thirds will present clean images with acceptable dynamic range in even quite dull lighting,  5D or D800 will be cleaner and have more dynamic range this hasn’t changed, the knowlage  we can get some of that back if you stick an f1 lens or expose longer is useful if the subject presents that opportunity.

  • Nik Andrews

    Who cares? We look at images not pixels… When was the last time a client or a friend looked at one of your images and asked what the individual pixels looked like?