SKAN wrote:
This video from Tony Northup explains it better:
Crop Factor: Why you multiply the aperture by the crop factor when comparing lenses
https://www.youtube.com/watch?v=f5zN6NVx-hYI would also like to quote one viewer's comment
[gamerguy00 2 years ago]
So in short, just so people don't misunderstand this - you're actually talking about the the effective DoF, not the actual amount of light coming through the lens. Logically, an f/2.8 on a full-frame is still a f/2.8 on a m43-camera, however the effective depth of field is equivalent of a f/5.6. This has to do with the size of the sensor, as you very well explained. Great explanation, just some bits and pieces that could've been misunderstood.
This video from Tony Northup explains it better: b... (
show quote)
I am not sure what the difference is between "effective DoF" and actual DoF. This is how things that should be relatively straight forward quickly go off the rails.
Do actually think that it is fair to compare a camera released in 2008 (5D Mk II) with one released in 2014 (Olympus OM-D E-M10)? A fairer comparison would have been a full frame camera that was closer., generationally, to the Oly, like a 5D Mk III (2012), or better yet, a Nikon or a Sony, both of which have good low light/high ISO performance. But that is the subject of another thread.
Actually TN's explanation is as clear as mud - at 3:23 he says that smaller sensors
"introduce a lot more noise because they are gathering a less total amount of light" which is where his confusion starts. The sensor does not gather anything - the lens does the gathering and projects it onto the sensor. But he seems to contradict this with his next statement - at 3:29 he says
"you get the same amount of light per square inch of the sensor" which is correct, but then he flips again when he says
". . . but each pixel will be getting much less light because there is simply less sensor size." Looks like TN is confusing sensor size and pixel size. If he is comparing sensors of the same resolution - the pixels in the smaller sensor cameras will be smaller, and that is the reason why they gather less light. For cameras where the pixel size are the same, you'll find that noise will be comparable when viewed at 100%. (See explanation and examples from DPReview, link below).
At 4:00 he makes his case - to multiply the the aperture on the full frame camera by the crop factor of the smaller sensor camera, in this case the Olympus being a crop of 2x, he uses a focal length on the full frame camera that is 2x the one used on the Oly, and suggests that the aperture on the full frame camera should also be multiplied by the same crop factor, to make a fair comparison. This makes sense and is supported by DoF calculators.
He then goes into a vortex of confusion when at 4:22 he says
"a lot of people think that the Olympus 45mm F1.8 provides the same image quality as the full frame Canon 85mm F1.8 lens. Because they have a 2x crop factor, they multiply 2 times 45mm and they think the lens becomes 90mm F1.8 - it doesn't - it becomes a 90mm F3.6 lens" The truth is that his comparison is valid only for depth of field, and his notion that there is a difference in image quality BECAUSE of the difference in DoF is baseless. If there is a difference in optical quality it is directly related to lens design parameters. If there is a noise difference it's entirely possible that the older camera (by 6 years) will not perform as well as a newer one. If there is a difference in how much detail is recorded, then there is no question, the Canon, with it's 21 mp sensor wins that round. And if you compare prints of comparable images magnified to the same size, the prints from the full frame camera, being magnified only half as much as the ones from the Oly, will show fewer lens flaws, noise and less focus/blur softness if present. An incorrect assumption and corresponding confusing statements at best.
At 5:27 he goes on to state
"people think that smaller sensors have worse image quality. But that's not really the case, at least it's not that simple. If you were to shoot them with exactly the same settings, in most cases the smaller sensor would have a lower level of quality" so far the science suggests he is correct. However, he offers the following as an explanation
". . . because each pixel is gathering less light, and that just introduces more and more noise. The more light light you give your pixels the cleaner your image is going to be. The way you can adjust for this is by using faster lenses on smaller sensor. In the example I just did, shooting with a full frame camera at F5.6 required me to increase the ISO two stops. I was able to shoot two stops lower on my Micro 4/3, and that gave me better image quality because the M4/3 camera was shooting at a lower ISO. You can get similar quality between larger sensors and smaller sensors if you use lower ISOs." So he is trying to make an equivalency between the terms image quality, noise and DoF. His explanation is that smaller sensors have pixels that gather less light. This is not necessarily true, especially if the pixels are the same physical size. The OM-D-EM10 has a pixel size of 13.99 µm², and the pixel size of the 5D Mk II is 41.09 µm². There is no question that the 6 year difference in release date will factor heavily, but even with that taken into consideration, direct comparisons between the two cameras show comparable results at low ISOs, and poorer results with the Oly at higher results (see second link below for a review of the Oly, along with a raw image noise comparator). As the ISO is increased, the noise more quickly increases to the point that it starts to impact the detail recorded in the smaller sensor camera.
His Oly is a 16 mp camera, his Canon is a 21 mp camera. The 33% higher pixel count in the Canon, together with the fact that it is a much larger sensor, completely trumps the image quality of the Oly, when you make prints of the same images, with all of the settings comparable. The 5D Mk II will still have more detail and less noise at 3200 ISO than the Oly will have at 800 ISO. That is a more realistic portrayal.
If you look closely, he is saying that he could shoot the Oly with an ISO two stops lower, but his video shows something different - at 5:58 he shows the Canon is set to ISO 3200, then shows the Oly (6:04) at ISO 200. Then goes on to say that the two stop difference gave him better image quality. Ok. maybe this was a mistake but it should have been corrected if such was the case. His spoken words don't match the text overlay. He later uses another side by side text overlay comparison and he seems to correct his mistake. But you can see where it is confusing.
Ah, but at 6:58 he makes the following statement -
"when you apply the crop factor to the aperture you will end up using a lower ISO on your smaller sensor, but that will improve the picture quality and give you image quality results that are closer to the full frame equivalents."Well that says it all, doesn't it. Improving picture quality to the point that the smaller sensor will be "closer" to full frame results. He
admits that full frame is better right there. Even when you increase the ISO by two stops on the full frame camera. And then he goes on to issue a plea to the mfgrs to be more "honest" about focal length equivalence. Hmm . . .
Some "accurate" background info can be found here:
https://www.dpreview.com/articles/5365920428/the-effect-of-pixel-and-sensor-sizes-on-noisehttps://www.dpreview.com/reviews/olympus-om-d-e-m10(Use the comparator to look at raw test images for both the Canon 5D Mk II and the Oly EM 10)
Yeah, mostly word salad with some accurate facts mingled in.