Pye Jirsa Explains Why These Are the Two Lenses You Should Own First

When we think of "need to own" lenses, sometimes we ignore the utility for the aesthetics. Other times, when looking at new equipment it really comes down to the usability of a lens and whether it should be taking up space in your bag. If you’re trying to make those decisions now, Pye Jirsa may be able to help.

The utility of equipment is different for everyone, and in the case of wedding photographer and owner of SLR Lounge, Pye Jirsa, it really is about what lenses can reap the highest benefit to not only the style of his work but to the clients he works with every weekend. Jirsa has gone through his catalog of 2018 images and noticed a striking realization, that 65 percent of his images are taken with the Canon EF 24-70mm f/2.8L II USM. If you’ve ever watched or listened to Jirsa previously, you’ll likely remember his personal disdain for the 24-70mm lens. Even with this acknowledgement, the versatility of lens and focal length cannot be overlooked.

One of my favorite lenses ever, the Canon EF 70-200mm f/2.8L IS III USM, is the second lens that Jirsa believes should be on your list of must own equipment for any event or portrait photographer. The complementary zoom range and fast aperture make this lens a must have for photojournalist and portrait photographers alike. Coming in second in his catalog with 25 percent of all his images being taken with the 70-200mm, it shows with its ability to separate the subject from the background while eliminating distracting elements from your subjects. Though on the heavy side, the telephoto zoom is a lens that deserves to take up space in every photographer's bag. 

Jirsa promises to reveal a separate video to round out the primes he uses as well as reveal how those lenses break down into the other 10 percent of his catalog. I’ve never been a fan of the 24-70mm lens for event work, but have routinely used the Canon EF 24-105mm f/4L IS II USM lens for events when I knew that lens changes would be too slow. It’s not the fastest lens nor the sharpest and probably the lens I most loathe in my bag (even though I use the Canon EF 8-15mm f/4L Fisheye USM much, much less). The focal length starting at 24mm and reaching out to a portrait length lens is just incredibly versatile for most event work and keeps you from missing a moment.

So, how do you feel about the 24-70mm lens (or the 24-105mm focal length for that matter) and do you think you share the same usage amounts in your own work as Jirsa? Is there a different lens that you own that you don’t love but can’t live without?

Log in or register to post comments

57 Comments

Pye Jirsa's picture

In addition to the recommendation, the video and article also gives techniques and best practices on shooting creatively with these lenses. Hopefully, that part is less "duh" =)

Another great video, Pye. I share your sentiments about the 24-70. While it is a workhorse and it doesn't have the magic of lenses like the 24/1.4 and 50/1.4 Planar etc. it always delivers the goods. I started photography a long time ago and I always learn something from your videos. Thanks!

That's why Olympus put out a 12-100 Pro lens, FF equivalent of 24-200, with image stabilization, constant f4 aperture, weight 1.24 lbs. OK, not f2.8 but pretty damn versatile with superb IQ.

Pye Jirsa's picture

That would indeed be a fun lens.

I fought purchasing the M Zuiko 12-100 for quite a while, the whole time thinking I definitely didn’t want an f/4 aperture, but I have to say that it’s amazing. I’ve been using it for quite a few commercial jobs and more recently for events. And it’s completely weather sealed!
Thanks Pye, for always putting out insightful videos and being a great ambassador in the photo industry!

Rk K's picture

Except that it's equivalent to an f8 24-200, and you need loads of light to use it. M43 is best with small, fast primes.

This f/8 rubbish is only relevant to DoF, and not to exposure. It IS an f/4 lens, and only needs as much light as any other f/4 lens.

Also, f/8 gives more a good amount of DoF for portraiture, events, etc. One does not buy an f/2.8 or f/1.4 lens for bokeh, but for low-light work. Unless you are shooting a performing arts event in a dark theatre, the f/4 will do just fine.

Rk K's picture

Take that somewhere else, people here understand how cameras work. Or where do you think the two stop noise advantage for ff comes from?

The “noise advantage” is a myth. It only occurs when comparing “same image”, and when sensor technology is assumed to be equivalent. “same image” dictates same DoF, which dictates same aperture diameter, which results in different F-number.

This has NOTHING to do with the same exposure. A 50mm lens at f/4 gives the same exposure as a 100mm lens at f/4, regardless of what sensor size they are connected. This is why hand-held light-meters do not care what focal length or sensor size one has. It only cares what exposure index one has set.

TL;Dr →

Aperture diameter (and distance to subject) is all which affects DoF. F-number (and exposure time & light level) is all that affects photon noise. With the same scene, (light value), and same exposure time, and the same F-number, one will produce the same exposure, and same photos noise, regardless of sensor size, or focal length.

Rk K's picture

How is it possible to talk so much yet understand so little?! It is truly remarkable. A simple look at the dpreview studio shot comparison will instantly disprove what you just said. They use the same aperture and ff is clearly two stops cleaner at the same iso. This is because you don't understand what iso and exposure is, and how they both relate to each other and to sensor size. When you compare the same image (same fov and dof), ff will be at 4x the iso and the noise will be the same. But then your f4 lens also becomes an f8 on ff.

You do realise that you just agreed with me! …So you must have missed my point.

First thing, the exposure index does not alter the amount of light which hits the sensor. At Tv=¹/100, and Av=f/4.0, the amount of light entering the lens is the same, whether the Sv is set at ISO 100/21° or 200/24°. The exposure index does not change the light exposure; it is adjusted for the light exposure.

Second, an f/4 lens remains an f/4 lens regardless of what sensor size is used, just like a 50mm lens remains a 50mm lens regardless of sensor size. To get the same image, a 50mm lens at f/4.0 on an MFT must be replaced by a 100mm at f/8.0, on an F-type, true. Same image also means same exposure time (for the same motion blur), and therefore a different EI for the same image brightness.

Nevertheless, the MFT and the F-type gets the same amount of light in the lens, and on the sensor, when they are both set to f/4.0 at ¹/100 second exposure, regardless of the EI.

Don't believe me, check your light meter.

Rk K's picture

I didn't agree with you, you just don't understand. Just as you don't understand exposure. You need to realise that exposure is /unit area, that's what the light meter measures, that's what f-stops refer to as well. But the amount of light exiting (not entering) the lens is the light intensity (exposure) x the area (image circle). This is where the sensor size comes in and this is why FF is 2 stops cleaner at the same iso or has the same noise at 4x the iso. It's basic stuff, with loads of articles about it, I don't know why you're still clinging to the wrong conclusions. You don't change the shutter speed to get the same image, you change the iso!

Let´s clarify somethings here. Your first claim was that a 12-100mm f/4.0 lens on an MFT (Olympus
M.Zuiko ED 12-100mm f4.0 IS PRO) is only equivalent to a 24-200mm f/8.0 lens on an F-type camera.

I claimed that that “equivalency” is only in regard to the DoF, and not the exposure. (I also claimed that the super-thin DoF of a 24-200mm f/4 was not necessarily optimum, but that was an aside).

You then implied that your first statement must be true, as the «two stop noise advantage» proves it.

I claimed that the noise advantage was only a myth, only occurring when comparing “same image,” a thing which does not happen in real photography. That is to say, Tony with his D850 will pull out his 70-200mm f/2.8, set it to 100mm & f/5.6, stand on the rock, focus 5m away, and shoot at ISO 400/27⁰ with a ¹/250 second exposure time, while Joe with his OM-D EM-1X will pull out his 40-150mm f/2.8, set it to 50mm & f/5.6, stand on the rock, focus at 5m away, and shoot at ISO 400/27⁰ with a ¹/250 second exposure time. They will get similar images, (with Tony having ever so slightly more out-of-focus background/foreground), with precisely the same exposure, FoV, photon noise, etc.

Nevertheless, if Tony decides to shoot at 70mm, or 200mm, the exposure will be the same if he kept his aperture at f/5.6. Further, if Joe wanted the same DoF as Tony, he would have had to set his aperture to f/2.8, (and reduce his EI to ISO 100/21⁰), but that is irrelevant to exposure.

You then claim that a look at images by DPReview will disprove the mathematical science. No it will not. The science is immutable. You said they use, «…the same aperture and ff is clearly two stops cleaner at the same iso.» You need to explain what you mean by aperture. Aperture is a diameter with the dimension of length. F-Number is a ratio with no dimension. What you stated cannot disprove anything. Again, the science is immutable.

Do it this way; take a gray card, in steady light. Photograph it from the same distance with an F-type camera using a 200mm, 100mm, & 50mm lens, all at f/5.6. make two copies of each photograph. With the copies, crop them 50% then enlarge the crops by 200% (to get back to the original size). Of the six images, which has the most noise, and which has the least? You will find that the percentage of noise in all of them will be precisely the same, (due to the Poisson distribution), but the size of the noise will be larger in the ones you enlarged.

I then said that you agreed with me, (in that, to capture the “same image” Joe would have had to use an aperture of f/2.8 in the example above to get the same DoF), and explained that changing the EI, —what you call the ISO— does NOT change the amount of light hitting the sensor, but adjusts for the amount of light hitting the sensor, and that this is only dependent on the light value, Lv, the exposure time, Tv, and the F-number (or focal ratio).

You then claimed that exposure is unit area, and is what the light meter measures. Herein is your mistake, and why you fail to understand. “It is good when the world agrees,”™ so let us look at what ISO says about all this.*

“The exposures produced by a DSC [Digital Still Camera] are determined by the exposure time, the lens aperture, [F-number or focal ratio], the lens transmittance, the lens illumination falloff, the flare light present at the sensor, and the level and spectral distribution of the scene radiances incident on the camera lens.”
Notice, no mention of sensor size, image circles, nor exposure index, (EI). Exposure is only a matter of the exposure time, the focal ratio, and the light value.

“As the exposure is increased, the [EI] gain can be decreased, and, therefore, the image noise can normally be reduced to an acceptable level.”
In otherwords, [photon] noise is a by-product of low exposure, nothing else.

“exposure index → EI
numerical value that is inversely proportional to the exposure provided to an image sensor to obtain an image”
In other words, the EI is set according to the Lv. It does not affect the Lv, but is adjusted to the Lv. Higher exposure, lower EI. Lower exposure, higher EI. (N.B., do not conflate light value, Lv, with exposure value, EV. They are not the same thing. A light meter measures the Lv, and gives you an EV, based on the EI one sets in the light meter).

Also to clarify something else you are mistaken about, a hand-held light meter cannot measure the light exiting the lens, and is in no way “image circle” nor “sensor size” dependent. What it measures has no correlation to surface area of the sensor in any way, shape, or form. They are not format specific. Wherever you got the information that «…exposure is /unit area, that's what the light meter measures,» and «…the amount of light exiting …the lens is the light intensity (exposure) x the area (image circle),» was an erroneous source. A hand-held light meter cannot possibly know what image circle or image size you have.

TL;DR →

Light value, Lv, which is what light meters read, do not care about image circles nor sensor size. Exposure is only dependent on Lv, Tv (exposure time), and Av, (focal ratio). It is not dependent on sensor size nor EI (Exposure Index). EI does not alter exposure, but is set according to exposure.

*[ From ISO 12232:2019 “Determination of exposure index, ISO speed ratings, standard output sensitivity, and recommended exposure index.” ]

P.s., I never claimed that one changes exposure time —what you called shutter speed— to get the same image. We actually agreed on that last statement. Maybe you ought to re-read what I wrote.
P.p.s., I also made some typos. Will be correcting those. ;-)

Rk K's picture

You keep saying the same incorrect thing with more and more words that you don't fully understand. It doesn't make you any less wrong.

Exposure is /unit area, that's why you get same exposure values on any camera, with any sensor size. Noise and dof however, are not measured /unit area and cannot be determined just based on exposure. You need to take the sensor size into account (and a couple of other things for dof).

Open up the dpreview test scene, select two cameras with different sensor sizes and observe how FF has a two stop noise advantage at the same iso (and exposure) over mft.

I cannot make it any simpler than this for you. Noise is not determined just by the iso value! A FF sensor will receive the same light intensity (and photon noise if you insist) /unit area, this is why you get the same exposure, but it has 4x the area! That's why you get 4x less noise. How dim can you possibly be?!

Your example is also comletely wrong. Tony will not only get much shallower dof, but also a much cleaner, higher quality image.

The science is immutable. But you have to grasp it first.

You are conflating things again.

«Exposure is /unit area, that's why you get same exposure values on any camera, with any sensor size.»
I ask you again, How on Earth, (or any other planet for that matter), can a HAND-HELD light-meter, incident or reflective, possibly know what camera/sensor I have?!? How can my strobe meter possibly know what image circle my lens is casting?!? THEY DO NOT, CANNOT, and NEED NOT know! It is totally irrelevant, and whatever source you got that information from is lying. If I am wrong, please explain how a hand-held light meter can know these things.

Light meters measure Lux-seconds, and give an Lv reading, which is used, in conjunction with the EI value you provide, to return an EV value, which is used to set your Tv-Av combination. BTW, EV is equal to Lv when Sv is ISO 100/21⁰.

«You need to take the sensor size into account (and a couple of other things for dof).»
Er,… NO! Only two things, and two things only, determine DoF; Aperture diameter, and distance to subject.

Those DoF calculators which ask for stupid things like CoC size, focal length, Sensor size, enlargement size, etc., are wasting time. All the values they ask for are recalculated from those two already. If one really wants to complicate things, I can also argue that DoF is dependent on enlargement size and viewing distance, but we can remove those fom the equation by assuming perspective correct viewing distance, (which many of those DoF calculators already assume, or they assume a 10×8 inch print viewed from 10 inches away).

«…has a two stop noise advantage….»
And what noise advantage does the Pentax K-3 II have over the *istD? Thee is no clear advantage when all the factors of noise is not taken into account. Photon noise is only, and will always be only dependent on Lv and nothing more. You want to compare apples with oranges. Compare the Sony α7 to the Sony α7R III and tell me about the noise comparison. It is not from sensor size difference. Do the experiment I asked, then come back and explain why simply cropping the sensor, (cropping the image 50% then enlarging 200%), did not produce more noise. All a MFT sensor is, is an F-type sensor with less usable space. If an MFT has more noise based solely from its size, then cropping an F-type to 50% ought to produce the same amount of noise.

Take an image with a 50mm lens at f/2.8. Now take the same image with a 100mm lens at f/5.6, adjusting only the EI. now enlarge the 50mm image to match the 100mm image. Which has more noise? “Oh, but they are at different resolutions,” you say. What´s your point? It is only sensor size which matters, right? Wrong.

Many things can affect noise, but Photon noise is only affected by Lv.

«Tony will not only get much shallower dof, but also a much cleaner, higher quality image.»
Forget that Tony has a 45Mpx camera, and Joe has a 20Mpx camera. Not the point. Stick to the point. Give Tony a Nikon D5, or a Canon EOS 6D. I mean, really?!? What if Tony had the full-frame Nikon Df with 16Mpx or the Sony α7S with a mere 12Mpx? Will he still have a better, cleaner image because he has full-frame? The amount of Photon noise is only dependent on light level. Other things affect “cleaner, higher quality” than sensor size. Put a Pentax K-1 II on a tripod and use PixelShift. No CFA Transform approximation of any kind necessary, and the 45Mpx Nikon couldn´t come close, …but that is not the point. The point would be that they would have the “same image,” by Joe adjusting his aperture diameter to match Tony. It is basic geometry. “Which of these triangles are similar?”

The science is immutable. You do not grasp it. If you did, I would not have to explain that a hand-held light meter gives a camera agnostic reading, or that photon noise is attributed to the Poisson distribution, or that an exposure setting of ¹/250 seconds at f/5.6 at ISO 100/21⁰ is the same for all cameras, even a 10×8 inch view camera with a 400mm lens.

Rk K's picture

When did I say that exposure depends on sensor size?! I said the exact opposite, several times. This is precisely because exposure IS /unit area, and does not depend on sensor size. That's why light meters work.

The amount of information (light) captured, and consequently the noise, however does depend on sensor size. I don't write these meaningless word salads that you do, so please go back and try and understand what I'm saying.

Image noise is determined by the technology of the sensor and the amount of light hitting it. Not the intensity of the light! The amount. Obviously a larger sensor will gather more light at the same intensity. It is not measurably influenced by resolution.

When you compare the output of cameras you scale the images to the same output format (be it print, web or whatever else you need). Anything else is meaningless.

I did do these tests, as I use both and EM5II and various FF cameras. It was just as obvious as the dpreview comparison tool, which you continue to ignore because it experimentally disproves your house of cards of hypothesis. What you're saying isn't science. It's mad rambling.

Also, the word you're looking for is entrance pupil, not aperture.

https://www.dpreview.com/reviews/image-comparison?attr18=daylight&attr13...

«The amount of information (light) captured, and consequently the noise, however does depend on sensor size.»
Fine. Let us go back to THIS statement. I will ignore everything else.

You seem to have contradictions. Exposure is not relative to sensor size, but noise is, despite that photon noise is totally dependent on the Poisson distribution, regardless of what area it is on. I don´t actually get this, besides the fact that it is wrong on so many levels.

I have no idea where you got this, but the Poisson distribution tells us that if you have so much Lux-seconds of light, hitting a 1m², or a 2m², or a 4m², or an 8m² surface, that there will be a certain amount of photon noise. That is, if a room is being flooded with a certain amount of light, it does not matter what size surface it is hitting, it will have a certain amount of photon noise, and it does not matter what camera captures that image, it will have the same amount of photon noise. The more light which hits the object, the less noise. The greater the exposure, the less noise.

THERE IS NO CORRELATION between photon noise, and sensor size, (or total light gathered). Let us look at total light.

Here is a 3m×2m board. I frame it in my K-3 and a 75mm lens set to f/5.6 and exposure time of ¹/250 seconds at ISO 400/27⁰. The upper third is a white, 5 stops above middle gray, the lower two-thirds is middle gray. If I expose for the middle gray section, I will have little noise anywhere in my image. If I were to paint the bottom third a black, 5 stops below middle gray, and not change my exposure, then the bottom third will now have much noise. If I paint the middle third the same white as the upper third, the bottom still has the same amount of noise, but I have more total light. If I paint the left third and right third with white, the lower-central ninth still has the same amount of noise, despite being a much smaller area, and there being more total light. If I painted all the white areas middle gray, the lower-central ninth will still have the same amount of noise, despite there being less total light. If I painted it all black, then it will all have the same amount of noise, despite having much less total light.

The total amount of light does not change the amount of noise. If I were to use a 100mm lens, shooting only one-fourth of the board,it will have the same amount of noise. If I were to take Joe´s OM-D EM-1X, using the same settings, (Tv, Av, Sv), but a 50mm lens, I will have the same amount of noise. If I took Tony´s D850 with a 100mm lens, I will have the same amount of noise.

If I now either add another light source, or open my aperture to f/2.8, or increase my exposure time to ¹/60 seconds, then my noise will decrease.

Noise is NOT a function of total light. Exposure is NOT per unit area, Exposure is lux-seconds, and, although lux is lumens/m², that square meter is the lens pupil, (dependent on aperture diameter), not the sensor area. Now lumens is luminous flux, and that depends on focal length, but F-number is dependent on lens diameter and focal length, so now you have a dependence on F-number, and focal length and aperture diameter is no longer relevant! (That is why the focal ratio became a thing). Focal ratio, (and exposure time), affects the amount of light. Not focal length, not aperture diameter. (But we agree on that).

So there is nothing to link photon noise to total light captured, and, regarding total information captured across the entire sensor, that has no correlation to anything except your histogram.

P.s., As for your link, explain why the K-1 has less noise than the Sony or Nikon. Same size sensor, same settings. Why a difference? Nothing is proven here. There are too many variables at play, and we have no idea about the source of the noise. Compare the OM-D EM-1 to the OM-D EM-1 II. Same sensor size, same Mpx, different noise. Nothing is proven. The empirical data fails.

Rk K's picture

My link clearly and undeniably proves my point. All the FF cameras are very close to each other, and differences can be explained by the transmission of the lens, differences in sensor tech and raw processing, temperature, etc. The Em1 on the other hand is drastically worse. By exactly two stops, as predicted. You just refuse to see because it topples the house of cards that is your understanding of the topic.

"Photon noise is totally dependent on the Poisson distribution". You keep repeating this, like a mindless parrot, and while it's true, you don't actually understand what it means. It's related to the pixel noise (/photon if you will), and thus it's the noise /unit area! You increase the area and the image noise decreases, because you're not enlarging the /unit area noise as much. Simple! If you crop/zoom into an image the same thing happens. It appears more and more noisy.

Maybe you should do the experiment you keep proposing, to prove just how wrong you are. I know I did.

You also keep talking about the size of the subject but that's just nonsense. This is about the size of the sensor (and consequently that of the entrance pupil).

You keep explaining something that is obvious, but also completely irrelevant to the discussion.

You're also wrong about exposure. Again. It's true that it can be expressed as lux/s, and obviously lux is photons /unit area. But that area is the sensor. The sensor doesn't know or care what lens you put on it. You don't even need any lens. Exposure will be determined by the light hitting the sensor. Otherwise, how do you explain that longer lenses have larger entrance pupils at the same f-stop. You can't! Idiot...

«My link clearly and undeniably proves my point.»
It doesn´t, because it has no control protocol. You are claiming that the sensor size is responsible for the noise, but two MFT sensors, both 20Mpx, have noticeable different noise level. One cannot blame it on sensor size, nor on resolution, so why such a difference? Additionally, out of the three F-type sensors you had, 45Mpx, 42Mpx, & 36Mpx, the 36Mpx had the least noise.

Therefore, these two observations, (MFT vs MFT, and F-type vs F-type vs F-type), prove that the difference in noise CANNOT be SOLELY attributed to the sensor size. we are clearly seeing noise difference, without seeing sensor size differences. Therefore, the connection of noise to sensor size cannot be established. To disprove your premise, see this link:
https://www.dpreview.com/reviews/image-comparison?attr18=daylight&attr13...
the K-3 II clearly has a bigger sensor than the OM-D E-M1 II, yet has more noise, thus proving that smaller sensors have less noise, right? Empirical data!

But looking at the OM-D E-M1 vs the OM-D E-M1 II, the latter has so much more noise, and there is no size difference, therefore, we can conclude that size is NOT linked to noise. Empirical data!

How on Earth does one set of empirical data prove that larger sensors have more noise, and prove that noise is not linked to sensor size, and that smaller sensors have more noise? It doesn´t! What it proves is that noise has several variables, and your “proof” is no proof. Is does NOT definitively prove a link to noise and sensor size. Science is immutable, but the empirical data does not support your hypothesis. You just refuse to see it because it topples the house of cards that is your understanding of the topic.

Now, if one [*cough*, *cough* Northup *cough*] plotted all F-type sensors on a graph, and plotted all MFT sensors on a graph, the x-axis showing date of release, and the y-axis showing S/N ratio, then perhaps some link can possibly be gleaned, but then one still has to create graphs for sensor technology, fabrication resolution, (that is, 42nm, 32nm, 23nm, technology, etc.), sensor resolution, sensor pixel density, etc, etc. Without a control protocol, the graph data is pointless. (No pun intended).

«It's related to the pixel noise…»
No it is not. Pixels do not have noise. Pixels have one value. It is related to pixel array. Some pixels in the array will have a value slightly higher or slightly lower than the average. That deviation is noise. Yes, pixel arrays have a surface area, but… we are not speaking of the area of the entire pixel array, but the part of the array which displays the part of the image where there is noise. The Poisson distribution is a distribution, but it is a distribution over the scene, not the sensor.

«You increase the area and the image noise decreases…»

I have a 12m×8m bright white billboard with a 30cm×24cm matte black rectangle in the middle. If I have a 36×24mm sensor, and a 100mm lens at f/8.0, and 90% of the sensor is within 3% of saturation, (that is, 97% or more saturated), and one small 3.6×2.4mm rectangle of it in the center of the sensor is about 0.2% saturated, that 1% of the 36×24mm sensor will have noise, with a certain signal to noise ratio. while 99% of it will not. Now cut off 6mm from each side, and 4mm from the top & bottom, to turn it into a 24×16mm sensor. That 3.6×2.4mm block in the center still has the same S/N ratio). Reducing the sensor size did not change the amount of noise. Replace the 100mm lens with a 75mm lens at f/8.0, and that 3.6×2.4mm block becomes a 2.4×1.6mm block with the same S/N ratio. Changing lens does not change the noise. I can put on a 50mm lens at f/8.0, and cut the sensor down to 18mm×13.5mm, the 1.8mm×12mm block in the middle will still have the same S/N ratio, because of the Poisson distribution of light reflected off the billboard, not the sensor.

If I then paint a 3m×2m rectangle in the middle of the board matte black, so that the entire sensor has about 0.2% saturation, then the total light exiting the lens has changed, but the S/N ratio in that center square has not. Noise is NOT changed by total light leaving the lens. The science is immutable.

«…because you're not enlarging the /unit area noise as much.»
If I enlarge the sensor from 18mm×13.5mm to 36mm×24mm, and repaint the entire billboard bright white except for the original 30cm×24cm block, and replace the 50mm lens with a 100mm lens a f/8.0, now my area of noise has indeed increased four fold, and my noise is “spread out” over a larger area. True. But my signal is also spread out over a larger area, and so I am back to the same S/N ratio.

Sensor size does NOT alter photon noise.

«If you crop/zoom into an image the same thing happens. It appears more and more noisy.»
Er, no. It has the same s/n ratio. It is not more noisy at all. The noise is more visible, but it is not in the least bit more noisy. If one measured the noise, one will see that it is the same. If I shrink an out-of-focus picture small enough, it appears in focus. It does not mean that I took an in-focus image.

«Maybe you should do the experiment you keep proposing…. I know I did.»
Really?!? Hah, ha, ha, ha, ha, ha, ha, ha, ha, ha, ha, ha! That is a good one. I did that so long ago to illustrate something else, I doubt you did it. The science is immutable.

«This is about the size of the sensor (and consequently that of the entrance pupil).»
BOABDL!!!
AH, ha, ha, ha, ha, ha, ha, ha, ha, ha, ha, ha, ha! That is even funnier! How on Earth does sensor size affect pupil size? Do you mean that for the “same image,» we need a different pupil size? Again, re-read what I posted. A 12-100mm f/4 lens is indeed an f/4 lens, when it comes to exposure. It is ONLY a 24-200mm f/8.0 equivalent when referring to same image capture, but it is indeed an f/4 lens, because, at 12mm, it has an pupil diameter of 3mm, and at 100mm, it has a pupil diameter of 25mm. No matter how small the sensor is, the pupil diameter of the lens does not change, anymore than its focal length changes.

«You're also wrong about exposure.»
I quoted ISO. Are they also wrong?

«…lux/s….»
Nope! Lux×seconds, not lux÷seconds.

«…lux is photons /unit area.»
Er, wrong again. lux is lumen per square meter, and lumen is luminous flux, or light (photon) per second in a unit solid angle.

To re-iterate, when photon hit the matte black block, (of a certain area), they are scattered in every direction. Some of it enters through a certain sized pupil (which has a square area, 122.6mm² in our example of a 100mm lens at f/8.0). That is then scattered over a certain square area of the sensor. (8.4mm² in our example). How much gets through the lens, is determined by the pupil diameter, (a surface area, lumen per square meter), and how much it is scattered over the sensor is determined by the focal length. taking both the focal length and the pupil diameter into consideration, we eliminate those consideration using the focal ratio, or F-number, to get 8.0. We now no longer care about the scattering by the lens, nor the area of the pupil. We have simplified the equation. The amount of light forming that part of the image is the same, with the same S/N ratio.

Again, that is the ONLY reason why we use F-numbers; so that we can have lens, & film size, agnostic metering.

«But that area is the sensor.»
Er, no. That area is the area in the scene where we are seeing noise. We don´t care about the parts of the sensor where there is no noise.

«Exposure will be determined by the light hitting the sensor.»
By light hitting the sensor, where? In my highlights? In my shadows? In my mid-tones? Since most in-camera light meters meter for middle gray, it had better be for the mid-tones. But what about my HAND-HELD light meter? (Yes, it needs no lens, it does not come with lenses, it cannot accept lenses). HOW ON EARTH does my HAND-HELD light meter, meter the light hitting my sensor? It DOES NOT! It meters the light hitting the SCENE! The scene is what is metered, and the scene is what determines PHOTON NOISE!

«how do you explain that longer lenses have larger entrance pupils at the same f-stop.»
BY DEFINITION! The F-number, N, is the ratio of the focal length to the pupil diameter, such that, the focal length, 𝘧, divided by the pupil diameter, D, is given as
N=𝘧/D,
or,
D=𝘧/N

You know, I´ve actually mentioned this before. Don´t know why you thought it would be so hard for me.

«Idiot...»
…And this is where our conversation ends. I do not have conversations with those whose only hope is to resort to insults, name-calling, or vulgar language.

Good day, sir.

Rk K's picture

Man the longer your word salads the less sens you make. You can't even interpret basic data and go against undeniable evidence with your wild theories.

I clearly said that sensor tech also influences noise. And you don't even know the specs of the cameras I (or even you) chose! The Em1 has an old 16mp sensor, that's why it's more noisy. (still, it's not a big difference) The k3 is an old apsc camera, that's why it's more noisy.

Also, I had 45mp, 36mp and 25mp in the FF cameras I chose. The 36mp was the best, but they are also all very close. Far, far better than any mft camera!

What a joke you are!

Now, go back and try and understand my comment, (I doubt you're capable, the average child is more intelligent), because the rest of your comment is completely off topic comes from your continuous misunderstanding and overcomplication of a simple topic. I mean, you don't even know what pixel level noise is, but you're talking about quantum physics you have zero grasp of!

It has nothing to do with the size or shape of the subject! How could it?!

I said, “Good day, sir.”

Rk K's picture

What, you ran out of misinformation?!

Note to self: there are those who do not understand science and the scientific process, including simple geometry, & logic, and there are those who do not understand that you do not respond to those who throw insults, (or use vulgar language, although not in this case). It always seem to happen.

yanpekar's picture

I keep wondering why people give advice on what lenses other people "should own first" without knowing the needs of other people, what they shoot, and what lenses are the most optimal for the type of photography they do.

Pye Jirsa's picture

I responded to Michael similarly. From the video/article, these are the first two lenses we recommend for aspiring/pro wedding/event/portrait photographers. Photographers that shoot a blend of photojournalism and portraiture.

yanpekar's picture

Thank you for your reply. It makes sense. The title should reflect this, as otherwise it gives an impression that this combo is recommended to anyone without taking into account specific needs / type of photography people do. If you recommend these lenses for portraiture than you may consider not limiting the choice to just these two, as many prefer to use prime lens for portraits, which in many cases are more beneficial than the zoom lenses you recommended. The video may leave some people confused.

Michael Comeau's picture

It seems silly to blindly recommend this $3500+ combo to everyone.

Pye Jirsa's picture

Not blind. If you watch the video or read the article, these two lenses are recommended for aspiring/pro wedding/event/portrait photographers. That is photographers that shoot a blend of photojournalism and portraiture.

More comments