Longer focal length or shorter for good IQ for low light portrait photography?

Suman Vajjala

Well-known member
Messages
150
Reaction score
9
Hi,

I have a question. Consider two lenses with same f-ratio but one has a larger focal length than the other (say 200 mm vs 50mm). Which lens would give better image quality when shooting portraits in low light considering the same composition, same lighting conditions and same f-ratio?

Thanks and regards

Suman
 
... I thought initially (now I know it is incorrect) that longer focal length, faster lenses will capture more light than a shorter one (keeping the same composition etc).
Light gathered is the product (multiplication) of the entrance pupil area and the solid angle. So, it's easy to think of it this way ...

Holding f# constant if you double the focal length you double the diameter of the entrance pupil (and get 4x the area).
But you also halve (1/2) the angle of view (and as it works out the solid angle is a quarter (1/4)).

So the 4x and the 1/4x cancel out.
Thanks, Bill. May I know why one uses solid angle instead of area?
They are both used; it's solid angle times entrance pupil area.
 
... I thought initially (now I know it is incorrect) that longer focal length, faster lenses will capture more light than a shorter one (keeping the same composition etc).
Light gathered is the product (multiplication) of the entrance pupil area and the solid angle. So, it's easy to think of it this way ...

Holding f# constant if you double the focal length you double the diameter of the entrance pupil (and get 4x the area).
But you also halve (1/2) the angle of view (and as it works out the solid angle is a quarter (1/4)).

So the 4x and the 1/4x cancel out.
Thanks, Bill. May I know why one uses solid angle instead of area?
The f-number, as the quotient of the focal length and aperture diameter (thus related to aperture area), is an approximation that breaks down with low f-numbers (starting to become significant at f/2, a minor issue by f/1.4, and quickly diverging for lower f-numbers, still).
No, it is not an approximation; and in any case this has nothing to do with his question.

What point do you think you're trying to make?
 
... I thought initially (now I know it is incorrect) that longer focal length, faster lenses will capture more light than a shorter one (keeping the same composition etc).
Light gathered is the product (multiplication) of the entrance pupil area and the solid angle. So, it's easy to think of it this way ...

Holding f# constant if you double the focal length you double the diameter of the entrance pupil (and get 4x the area).
But you also halve (1/2) the angle of view (and as it works out the solid angle is a quarter (1/4)).

So the 4x and the 1/4x cancel out.
Thanks, Bill. May I know why one uses solid angle instead of area?
The f-number, as the quotient of the focal length and aperture diameter (thus related to aperture area), is an approximation that breaks down with low f-numbers (starting to become significant at f/2, a minor issue by f/1.4, and quickly diverging for lower f-numbers, still).
No, it is not an approximation; and in any case this has nothing to do with his question.

What point do you think you're trying to make?
DOH! I totally messed up -- my apologies! I got it in my mind (for whatever reason) that he was talking about why use the angle of rays defined by the entrance pupil rather than the f-number. Total brain-fart!
 
Why same f-stop?
It was to even out the comparison. I thought initially (now I know it is incorrect) that longer focal length, faster lenses will capture more light than a shorter one (keeping the same composition etc).

Regards

Suman
The point of my remark is that shorter lenses are typically faster (say, 50/1.4 vs. 200/2.8), so one could exploit this, with all the implication about the DOF.
I understand your point now. In the same example, am I right in supposing that the out of focus background will be smoother for 200/2.8 than 50/1.4?

Regards

Suman
In principal yes, with a deeper DOF, but it is not that simple because a different part of the background will be in the frame, and if the background is too close, the 50/1.4 will make it smoother because it has shallower DOF.

You can play with this:


 
Neither. There are good and bad lenses in all focal lengths. The choice of focal length depends entirely on the shooting situation and your preferences.
Hi,

My question is more from a theoretical standpoint. If all other factors are equal then which lens produces best IQ? Or, is IQ independent of focal length?
If by image quality you mean the technical stuff that they test on lens reviews, there is what Great Bustard said below, that if we are talking in full frame terms, 50mm lenses are usually sharper than 200mm lenses. Also known as 'nifty fifties' 50mm lenses are the Goldilocks of lens design and often the sharpest lenses with the least aberrations, distortion etc. are 50 or 85mm lenses. They will also be cheaper than a 200mm lens of corresponding quality.

But I can't imagine a situation where it would make sense to choose a portrait lens of a particular focal length because it's optically better than one with another FL. As GB also points out, with similar framing, the perspective and depth of field will be different with different FLs. Since we are probably more sensitive to/critical of minute differences in human faces than any other subject, that becomes the most important aspect. Only with a flat object such as a painting or a wall I would choose my optically best prime regardless of FL provided there's ample light and I can get far enough from the subject since perspective doesn't matter in this case.

If you want to do a general comparison of lenses of different FLs in terms of IQ you can read lens reviews. After going through a sufficient number you should have a fair idea of the correlations.

You mention low light so you are clearly after a fast lens. Longer lenses get bulkier and more expensive for the same maximum aperture and IQ if they even exist. They are also harder to hold steady and small movements produce more blur so you have to use a higher shutter speed which gives you a lower exposure and a more noisy image so shorter lenses have a clear advantage here.
If the image magnification (basically, framing) and aperture are the same, the depth of field will be the same. Perspective will be different because that's not determined by focal length. It's determined by camera position.
 
Why same f-stop?
It was to even out the comparison. I thought initially (now I know it is incorrect) that longer focal length, faster lenses will capture more light than a shorter one (keeping the same composition etc).

Regards

Suman
The point of my remark is that shorter lenses are typically faster (say, 50/1.4 vs. 200/2.8), so one could exploit this, with all the implication about the DOF.
I understand your point now. In the same example, am I right in supposing that the out of focus background will be smoother for 200/2.8 than 50/1.4?

Regards

Suman
In principal yes, with a deeper DOF, but it is not that simple because a different part of the background will be in the frame, and if the background is too close, the 50/1.4 will make it smoother because it has shallower DOF.

You can play with this:

https://dofsimulator.net/en/?x=EfQA4QQRoAAIJEwkAAADgAA

https://dofsimulator.net/en/?x=EH0AcQEEYAAIJEwkAAADgAA
Thank you very much for the links. I will try them out.

Regards

Suman
 
To answer why we refer to solid angle:

The answer is that we actually refer to the product of the solid angle and an area. For example we take the solid angle of acceptance and the entrance pupil area (The aperture area as seen from the front). This product is called the ettendue. It is the same number for the product of exit solid angle and exit pupil. It is a conserved number.

Relating to this thread on the effect of using a long lens at longer distance versus using a shorter one closer: If the subject distances scale with focal length, and the f/# stays the same so does the ettendue. It will be the same for both lenses.
 
To answer why we refer to solid angle:

The answer is that we actually refer to the product of the solid angle and an area. For example we take the solid angle of acceptance and the entrance pupil area (The aperture area as seen from the front). This product is called the ettendue. It is the same number for the product of exit solid angle and exit pupil. It is a conserved number.

Relating to this thread on the effect of using a long lens at longer distance versus using a shorter one closer: If the subject distances scale with focal length, and the f/# stays the same so does the ettendue. It will be the same for both lenses.
Thank you, for the explanation. I have come across solid angle but never used. I think, it will take some time for me to digest this information :)

Thank you & Best regards

Suman
 
To answer why we refer to solid angle:

The answer is that we actually refer to the product of the solid angle and an area. For example we take the solid angle of acceptance and the entrance pupil area (The aperture area as seen from the front). This product is called the ettendue. It is the same number for the product of exit solid angle and exit pupil. It is a conserved number.

Relating to this thread on the effect of using a long lens at longer distance versus using a shorter one closer: If the subject distances scale with focal length, and the f/# stays the same so does the ettendue. It will be the same for both lenses.
Thank you, for the explanation. I have come across solid angle but never used. I think, it will take some time for me to digest this information :)
Sure. Solid angle is basically the 3D version of your regular 2D angle.
 
To answer why we refer to solid angle:

The answer is that we actually refer to the product of the solid angle and an area. For example we take the solid angle of acceptance and the entrance pupil area (The aperture area as seen from the front). This product is called the etendue. It is the same number for the product of exit solid angle and exit pupil. It is a conserved number.
Yes, and to risk confusion, got always conserved in practice.
In other words not everything is always delivered from the exit pupil to the sensor (film).
For example, vignetting.
Relating to this thread on the effect of using a long lens at longer distance versus using a shorter one closer: If the subject distances scale with focal length, and the f/# stays the same so does the etendue. It will be the same for both lenses.
 
The point of my remark is that shorter lenses are typically faster (say, 50/1.4 vs. 200/2.8), so one could exploit this, with all the implication about the DOF.
I understand your point now. In the same example, am I right in supposing that the out of focus background will be smoother for 200/2.8 than 50/1.4?
There is a simple approximation for background blur for an infinitely distant background: and by "infinity", that is any background about ten or more times as far as the distance from the camera to the subject. So if you are five feet from your subject, "infinity" is approximately fifty or more feet beyond your subject.

Background blur = subject magnification x focal length / f-stop value.

Subject magnification is equal to the size of the object on your sensor divided by the size of the subject in real life. More magnification, such as getting closer to your subject, will increase background blur, so half the subject distance means twice the blur. Larger sensors are good for giving more background blur, as you can get closer to the subject with the same focal length, increasing the subject magnification..

Doubling the focal length, or halving the distance to the subject, or halving the f/stop will all double the background blur.

Focal length / f-stop is the measure of the lens' "entrance pupil diameter" and a larger diameter will give you more background blur on a distant background. So for your two lenses:

200 mm / 2.8 = 71.4 mm

50 mm / 1.4 = 35.7 mm

So the longer lens will give you double the background blur, but only for a suitably distant background.

For close backgrounds, you'll have to go back to the depth of field calculators, and eventually for a really close background, the lens with smallest f/stop will give you the maximum background blur.
 
The point of my remark is that shorter lenses are typically faster (say, 50/1.4 vs. 200/2.8), so one could exploit this, with all the implication about the DOF.
I understand your point now. In the same example, am I right in supposing that the out of focus background will be smoother for 200/2.8 than 50/1.4?
There is a simple approximation for background blur for an infinitely distant background: and by "infinity", that is any background about ten or more times as far as the distance from the camera to the subject. So if you are five feet from your subject, "infinity" is approximately fifty or more feet beyond your subject.

Background blur = subject magnification x focal length / f-stop value.

Subject magnification is equal to the size of the object on your sensor divided by the size of the subject in real life. More magnification, such as getting closer to your subject, will increase background blur, so half the subject distance means twice the blur. Larger sensors are good for giving more background blur, as you can get closer to the subject with the same focal length, increasing the subject magnification..

Doubling the focal length, or halving the distance to the subject, or halving the f/stop will all double the background blur.

Focal length / f-stop is the measure of the lens' "entrance pupil diameter" and a larger diameter will give you more background blur on a distant background. So for your two lenses:

200 mm / 2.8 = 71.4 mm

50 mm / 1.4 = 35.7 mm

So the longer lens will give you double the background blur, but only for a suitably distant background.

For close backgrounds, you'll have to go back to the depth of field calculators, and eventually for a really close background, the lens with smallest f/stop will give you the maximum background blur.
Hello Mark. Thank you for the technical explanation regarding background blur.

Regards

Suman
 
I understand your point now. In the same example, am I right in supposing that the out of focus background will be smoother for 200/2.8 than 50/1.4?
The quantity of background blur is distinguished from the quality of background blur, and smoothness is quality measure, and not a quantity of blur. In my previous reply, I was only considering the quantity of blur and not its quality at all.

The "smoothness" of a blurred background is a function of optical design. Some lenses are specifically designed to have smooth out-of-focus regions behind the point of focus, which is usually called good bokeh. To guarantee good bokeh, you need a lens that is designed to delver good bokeh. Lenses that are valued for portraiture usually have smooth bokeh, while many inexpensive, general-purpose lenses have poor bokeh where the background blur is jittery or distracting.

This article describes bokeh in detail:

 
If the image magnification (basically, framing) and aperture are the same, the depth of field will be the same.
Good point.
 
If the image magnification (basically, framing) and aperture are the same, the depth of field will be the same. Perspective will be different because that's not determined by focal length. It's determined by camera position.
 
Hello everyone. I have performed the tests today to see if the exposure remains the same when using two different lenses with same f-stop value and same composition.

All tests are performed with same ISO and f/4 f-stop value. Photos are attached here which are straight out of camera. The first photo was shot at 24 mm, the second one at 50mm and the last at 105 mm. The camera metering was set to spot focus and care was taken to ensure same area for metering for the three shots (using diagonals & grid lines). Please note that the subject is the white card (also for metering).

Conclusions: Shutter speed remains more or less the same. Also, the exposure is consistent for the three images. (Please note that this is not performed with perfection as the framing, angle etc are different.) This is as pointed by several forum members here. Thanks to everyone for the help and support. I have learned a lot of things from this discussion.

Image 1: 24mm
Image 1: 24mm

Image 2: 50mm
Image 2: 50mm

Image 3: 105 mm
Image 3: 105 mm

Best regards

Suman
 
Last edited:
Hello everyone. I have performed the tests today to see if the exposure remains the same when using two different lenses with same f-stop value and same composition.

All tests are performed with same ISO and f/4 f-stop value. Photos are attached here which are straight out of camera. The first photo was shot at 24 mm, the second one at 50mm and the last at 105 mm. The camera metering was set to spot focus and care was taken to ensure same area for metering for the three shots (using diagonals & grid lines). Please note that the subject is the white card (also for metering).

Conclusions: Shutter speed remains more or less the same. Also, the exposure is consistent for the three images. (Please note that this is not performed with perfection as the framing, angle etc are different.) This is as pointed by several forum members here. Thanks to everyone for the help and support. I have learned a lot of things from this discussion.

Image 1: 24mm
Image 1: 24mm

Image 3: 105 mm
Image 3: 105 mm
A well done test! Just one thing: if not spot metering, then, of course, the camera may meter differently since the scene luminance as a whole may change. For example, had you used evaluative metering, the top scene would have metered lower than the bottom scene, because the top scene, as a whole, is brighter than the lower scene.

Also worth noting is that difference in lens transmission and/or vignetting, whether for different lenses or for different focal lengths within the zoom range, may also affect the metering -- see this post for more specifics.
 
Hello everyone. I have performed the tests today to see if the exposure remains the same when using two different lenses with same f-stop value and same composition.

All tests are performed with same ISO and f/4 f-stop value. Photos are attached here which are straight out of camera. The first photo was shot at 24 mm, the second one at 50mm and the last at 105 mm. The camera metering was set to spot focus and care was taken to ensure same area for metering for the three shots (using diagonals & grid lines). Please note that the subject is the white card (also for metering).

Conclusions: Shutter speed remains more or less the same. Also, the exposure is consistent for the three images. (Please note that this is not performed with perfection as the framing, angle etc are different.) This is as pointed by several forum members here. Thanks to everyone for the help and support. I have learned a lot of things from this discussion.

Image 1: 24mm
Image 1: 24mm

Image 3: 105 mm
Image 3: 105 mm
A well done test! Just one thing: if not spot metering, then, of course, the camera may meter differently since the scene luminance as a whole may change. For example, had you used evaluative metering, the top scene would have metered lower than the bottom scene, because the top scene, as a whole, is brighter than the lower scene.

Also worth noting is that difference in lens transmission and/or vignetting, whether for different lenses or for different focal lengths within the zoom range, may also affect the metering -- see this post for more specifics.
Thank you. I had deliberately put the subject in the center as I thought vignetting may influence the results. Also, I chose spot metering as I wanted to control the exposure for the subject. As spot metering was available only for a small centeral portion of the frame (~2.5 or 5%, I don't remember which is correct), I was forced to keep the subject in the center. I am glad that I did these things as I now know from your post that my guesses were correct!

Best Regards

Suman
 
Last edited:

Keyboard shortcuts

Back
Top