I unfortunately got a very poor and cheap 18-55mm lens on my Canon 60D. In addition to poor quality of my photos (poor in comparison with my 50mm f/1.8 Lens which makes great pictures) the noise level of my pictures is really high. Should I upgrade my camera or get a better lens? since I'm interested in "night sky photography" this level of noise is really bothering.

For example you're taking photo of a special scene with two different lenses with the same ISO, Aperture and Shutter speed. Is the noise level of a photo taken with Zeiss Lens the same as Canon Lens?


In short, no. See What is noise in a digital photograph? for a fairly comprehensive overview of what does.

The main aspect of a lens which might cause increased noise is if you are shooting at a reduced aperture and not compensating with a longer exposure — you'll have to increase the ISO, and that amplification will make more apparent noise. But if you're shooting at the same aperture, give or take the details of different transmission (which is usually not a big deal), there will be no difference.

I suppose some forms of noise may be more or less masked by a lack of lens sharpness, but that's not going to be useful in any way.

  • Right. But what I exactly ment is that for example you're taking photo of a special scene with two different lenses with the same ISO, Aperture and Shutter speed. Do they exactly make the same level of noise? for instance a 50mm f/1.2 Zeiss vs Canon
    – SERAJ
    Nov 22 '14 at 15:25
  • 1
    Noise isn't a property of a lens, but the lens will play a large factor in the amount of light you have, which will affect what your ISO is later in the process, which can have a profound affect on noise. At otherwise identical settings, a faster lens can result in less noise. Edit: Faster should suggest that a wider aperture is being used, which may not be available on a cheaper lens. Nov 22 '14 at 15:33
  • 4
    If you use the same settings with different lenses and the light remains exactly the same, the noise level will be exactly the same. Nov 22 '14 at 15:34

I agree with @mattdm 's link to "What is noise..." as fairly comprehensive, but not complete today. It would be conventional to say that the glass and metal of a lens does not add noise to the image. A simple definition of noise might be the random disturbance of an image. In other words, if you took the same picture again and again, what might change from one to the next image is noise. Of similarly, noise is the unpredictable degradation of the image. Distortion and aberrations caused by the glass are not conventionally considered noise as they are predictable or repeatable.

But a contemporary lens today is not only glass, metal and mechanics. They contain electronics, focus servo motors and image stabilization components. These electronics can surely create electrical noise internal to the camera and effect the sensor data readout. A poor motor driver could create noise on the camera power bus and effect the image readout. A malfunctioning image stabilization system could make the image worse.

So I would offer that it is possible that two lenses could introduce different amounts of noise into the camera system.


Yes.. and no!!!! But not because if the f stop or other things you can set in your camera.. people above have been minimizing the effect of the light transmission (t stop) of a specific lens. Sometimes the t-stop of a lens is a couple stops smaller than the f stop of the lens. Although the t-stop does not affect the bokeh, it does affect the exposure. A picture taken with two different lenses and all of the settings the same will be darker on the lens with a lower t stop. You can check the T stops for a bunch of different lenses on dxomark.

So if the exposure, factoring in the T stop, is exactly the same then the pictures will be the same and so will the noise. But since one lense may have a smaller t stop then another lense one picture may be darker than the other and thus will seem to have more noise because the shadow areas will just have more apparent noise in the loss of detail


I would have said "no" a week ago, however I have change my opinion. Yes, a cheap lens can cause noise, at least in video. I was shooting with a RED in the mountains, and my good mid range lens was destroyed. I used the 18-55mm as back up. I was switching between the kit 18-55mm to my 70-200mm f/4 L-series. Same ISO, same settings and there is a huge noise issue with the scenes shot with the 18-55. I was expecting poor performance out of the kit lens, but not noise. (Which is how I found my way here.)

I am willing to bet it does poorly with chromatic aberration combined with lack of resolving power that creates the noise with this particular lens.


On the field I found out that lens quality absolutely affect the ISO reduction. It all comes down to quality of a glass and how effective it absorbs the light. Had a shooting with my kit Nikon lens and more expensive Tamron lens — same aperture, same ISO, same shutter speed. Tamron blew the kit Nikon lens out of proportion. My photos were 2 exposures darker with kit lens, and noise levels were more noticeable.

There are restrictions of the camera sensor of course, but size of a lens and quality of a glass does affect the noise levels. Nothing dramatic, but difference is there once you crop your photos. Not to mention the color quality and contrast.

  • 2
    You're misunderstanding something here (or your camera or lens is malfunctioning). Assuming that your kit lens was capable of opening up to the same aperture as the Tamron, they will be effectively the same brightness - that's what "exposure" means, and exposure is set by shutter speed, aperture and ISO (and incoming light, but I'm assuming that's not changing).
    – Philip Kendall
    Oct 3 '18 at 12:22
  • 1
    This can't be right. There may be a difference in transmission, but that difference will be a fraction of a stop. The Nikon kit lens does not have a two-stop light loss.
    – mattdm
    Oct 3 '18 at 12:47
  • 1
    Is there any chance you're setting the aperture to f/2.8 and then shooting at the max focal length of your kit lens? That would give you a two stop "difference" because the kit lens can't go faster than f/5.6 at that focal length.
    – Philip Kendall
    Oct 3 '18 at 14:46
  • I am profession photographer, so assuming that I am misunderstanding something here is pure speculation and domineering approach. My both lenses are working fine but Tamron 17-55 totallyabsorbs more light and works better. I shoot professional models with kit lens aswell, and photos are perfect, but Tamron is always better in low light enviroment when it comes to noise. I set both of my lenses on f5.6 with same shutter, ISO, and i had speedlight flash as light source and they perform differently. I also had studio shootings and with better lens I can increase ISO more and get same results.
    – Marko Beg
    Oct 4 '18 at 10:07

I found something called T-number which affects the light transmittance of Lenses, and varies form lens to lens. This fact Eventually may make you change the camera settings that is influential on the final result and photo quality for sure. more info: What is T-number / T-stop?

  • This is covered in my earlier answer.
    – mattdm
    Jul 29 '16 at 11:27
  • See the question, What is T-number / T-stop? It's important in videography, but almost never matters in practice for photography.
    – scottbb
    Jul 29 '16 at 13:18
  • Also, by itself, this answer has nothing to do with noise, which is what the original question was about.
    – scottbb
    Jul 29 '16 at 13:25
  • @scottbb How do you say that lens structure which is a physical concept is just influential in videography?
    – SERAJ
    Jul 30 '16 at 12:30
  • I never said anything of the sort. I said that it, referring to the notion of transmission stops, "is important in videography". I said nothing about physical structure of the lens being "just influential". T-stop does not describe the physical structure of a lens; it merely quantifies the total amount of light reduction through a lens from all sources (internal reflection; absorption; and blockage by the iris). This matters to videography. It almost never matters in practice in photography.
    – scottbb
    Jul 30 '16 at 13:00

At first it does not, but in practice it does. Cheap lenses have a physical F4, a aperture is the opening relative to the focal lenght, right. So F2 for a 50mm lens is 50mm divided by 2 equal 2.5mm aperture. so a Canon 50mm L and a 50mm 1.4 and a 50mm 1.8 all have a aperture of 2.5mm when shot at F2 and they produce same results. This is the theory at least.

But the 50mm 1.8 dows not have the same quality glass as the 50mm L. It has some cheap plastic elements which absorb much of the spectrum so at same ISO, F2 and 1/125 the 50mm 1.8 has just a nutella grey muddy image where the 50mm L has a lot more semitones and shades and color glare expecially in the shadows. Old timers call this shadow reading characteristic of a lens.

On an overcast day the difference is small. but enter an Orthodox church, dark and poorly lit , with just some light coming from a narrow window, and the differnece between the USM versus L glass is huge. And with the same settings the L shows you much more of the original scene and the garbage 50mm 1.8 has a lot of dark islands in the image. Try to open those up in post processing and noise level increses too. But color shades yuo will never recover.

The L lenses are like half of what a Leica lens does. now of course is up to the photographer to understand and try to overcome this and i am surprised that a lot of people only see better contrast and better constructuon in an L lens when the difference is in the colors and in between colors and their brightness. I am not talking about primary colors here but of those in between colors. you can't bring those in pp.

now a lot of people are not passionate photographers and are only after making a quick buck and have cheap 40% srgb laptop monitors so those will wonder "what is he talking about, I can slide vibrance all the way right and get crazy colors" ... lets try it this way. buy a cheap $30 tessar m42 and go do some roses and alternate the 50mm 1.8 with the tesar, both shot at F4 lets say. and then try to make the mutted dead dusty country side looking colors of the 1.8 look like those already found in the shots with the tesasar.

Yuo can try with the 50mm 1.4 too and have the same struggle. for paint colors this does not matter but when it comes to natural colors, flowers, water, different shades of the sky early in winter, and portrait, you will not make then look the same if you pp a whole day. you cannot regain what was previously absorb by the poor lens. All you can do is apply coection to what you have recorded but that you can do to the L glass too so the advantage still remains.

Nikon also marks this quality glass, with a gold ring and same as canon, has a cheap junk line 50mm 1.4 at around $350 and a 58mm 1.4 gold ring at $1500. and I see a lot of people wondering why. it's because of the glass. the glass in the L or nikon gold ring is quality optical glass where the rest is something, maybe plastic, maybe resin, maybe just lower quality glass.

  • 3
    There a lot of inaccurate claims here. To pick a couple... 50 divided by 2 is 25, not 2.5. Also, while some of the lenses mentioned use plastic in the construction of the body, none of them use plastic or resin lens elements.
    – mattdm
    Mar 18 '18 at 14:10
  • 3
    And the part about losing midtones is basically nonsense as well. Light just does not work that way.
    – mattdm
    Mar 18 '18 at 14:14
  • Also, if your Nutella is grey, you should definitely check the expiration date.
    – mattdm
    Mar 18 '18 at 16:33
  • "L" is not a type of autofocus, so the comparison of "L" vs USM is not accurate (since many L lenses use USM). Also, I wouldn't call every f/4 lens cheap. The Canon 70-200 f/4 is stellar, and so too is the 300 f/4.
    – OnBreak.
    Mar 19 '18 at 15:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.