does lens choice effect noise?

[almo]

Well... That pretty much is the question.

[cozmocha]

You can have less noise with a 2.8 then a 5.6 because you wont have to have a high iso.

Not sure what differences you are looking at in the lenses?

[almo]

You can have less noise with a 2.8 then a 5.6 because you wont have to have a high iso.

Not sure what differences you are looking at in the lenses?

I guess what I am saying is cheap lenses compared to L grade lenses. I have seen the slightest bit of decrease in noise level when inspecting images shot at similar values on my 30D from non L lenses to L lenses. I can't say the same about my current camera, since I have never used anything but L's with it.

Thanks!

[drg]

Yes.

Color noise can be introduced at different levels even with a top of the line lens under best/worst conditions.

Chromatic Aberration (CA) is a type of noise in how it is corrected in image processing (in-camera or post processing).

Color bloom can be accentuated by a slower lens or a fast lens can produce an edge effect or halo if shot wide open with the light hitting the lens in the wrong way. A common variation of this is 'flare' which is filtered to a great extent by modern coatings.

High end lens offerings like the Canon luxury 'L' units excel because Canon makes a point of being sure that they perform similarly (very similarly) from one unit and model to another.

[DGK*CRONE]

Good question. I was wondering the same.

If I am shooting in a High ISO and there is noise more noise than I would like, would switching to a better lens decrease that noise?

[almo]

Good question. I was wondering the same.

If I am shooting in a High ISO and there is noise more noise than I would like, would switching to a better lens decrease that noise?

Good point. I await the answer.

[almo]

Yes.

Color noise can be introduced at different levels even with a top of the line lens under best/worst conditions.

Chromatic Aberration (CA) is a type of noise in how it is corrected in image processing (in-camera or post processing).

Color bloom can be accentuated by a slower lens or a fast lens can produce an edge effect or halo if shot wide open with the light hitting the lens in the wrong way. A common variation of this is 'flare' which is filtered to a great extent by modern coatings.

High end lens offerings like the Canon luxury 'L' units excel because Canon makes a point of being sure that they perform similarly (very similarly) from one unit and model to another.

Thanks Carl. That is great info!

[drg]

Good question. I was wondering the same.

If I am shooting in a High ISO and there is noise more noise than I would like, would switching to a better lens decrease that noise?

Slightly different problem. HIGH ISO produces noise in the camera from amplification of the sensor signal. A very very fast lens can reduce that because of a different exposure setting that can lower the ISO used.

When using HIGH ISO underexposure is a killer! It will just push the noise up into the visible picture worse because of signal to noise ratio (you old audiophiles know what I'm talking about). Slight overexposure will help.

Some camera sensors at high iso setting benefit from longer exposure times or in other words the shutter speed should be decreased (example from 1/50th to 1/10th of a second). This of course usually means a tripod at lower ISO. It can introduce other problems but you can shoot in those silly ranges of 6400 or greater if you remember to overexpose and let the sensor fill up with lots of light to push the amplified noise way down in the signal.

[Franglais]

..
If I am shooting in a High ISO and there is noise more noise than I would like, would switching to a better lens decrease that noise?

If you put the 18-55 f2.8 pro lens on your camera instead of the 18-55 f3.5-5.6 kit lens and you changed nothing else then you would see zero difference in noise.

If you did the shot at 1600 ISO f5.6 on the kit lens then did the shot at 400 ISO f2.8 on the pro lens then you see a lot less noise in the shot done on the pro lens.

Noise is dependent on the ISO you choose. The ISO you choose (in low light) is dependent on the maximum aperture of your lens. A lens at f2.8 lets through 4 times more light then a lens at f5.6 so you can reduce the ISO by 4 times.

[Loupey]

Noise is a function of the imaging sensor design. Since the sensor cannot be changed within a given body, the appearance of noise is solely dependent on the amplification (ISO) of the signal.

In other words, noise at ISO 1600 is not dependent on the lens.

Noise can be more noticeable against different backgrounds and if a scene is underexposed as others have mentioned (because the noise is further amplified during post processing).

But, I do wonder if the effect of poor bokeh from lesser lenses can give the illusion of more noise. I don't have directly comparable lenses to test that thought unfortunately.

[DGK*CRONE]

Thanks everyone! Everyones response really helped.

[drg]

If you put the 18-55 f2.8 pro lens on your camera instead of the 18-55 f3.5-5.6 kit lens and you changed nothing else then you would see zero difference in noise.

If you did the shot at 1600 ISO f5.6 on the kit lens then did the shot at 400 ISO f2.8 on the pro lens then you see a lot less noise in the shot done on the pro lens.

Noise is dependent on the ISO you choose. The ISO you choose (in low light) is dependent on the maximum aperture of your lens. A lens at f2.8 lets through 4 times more light then a lens at f5.6 so you can reduce the ISO by 4 times.

This isn't entirely the whole story.

The two lenses expensive or 'kit' will produce a different and measurable color range/gamut in the image. The less expensive lens is far more likely to produce an aberration under the same conditions as the 'pro' lens unless they have identical coatings in which case the pro lens will be a poorer quality image. If not there would be little need for the more expensive lens as speed of the lens in alone isn't the only factor with its ability to produce a quality image. Some of the great images of photographic history were shot with lenses that started at about f/8 or even greater.

What you say? The larger primary element in the pro lens needed to produce the lower f-number requires a different quantity of flare and color correction coating to produce at least as good a quality of image as the cheaper lens. The nature of the materials used for these coatings also aid in shifting the chromatic problems either out of the visible range or 'compressing' them in a non-objectionable visual range.

Uncoated lenses from the black and white era, lens that were optimized for a particular type of film, and lens coating that are damage or have deteriorated demonstrate how much effect coating have when mounted on any DSLR. You'll see from those examples that a lens can produce gross problems in that area we call noise. Examples of color optimized lens offerings include the magic Kodak and Nikon lenses for the original Kodachrome.

Panavision made lenses for motion picture film for specific emulsion numbers and they need a lot of post work to get them to look right with a digital imaging system.

I have uncoated Leitz lenses that are optical clear as a possible but don't do well with any color film or sensor. I also have a wonderful Hasselblad lens from the olden days that hasn't ever worked very well on a digital back because of some internal flaw that doesn't show on film but creates micro flares and hot spots on the digital sensors.

My point is that there are several image issues that are noise in how we correct them. Color noise is the most noticeable in the average image. The sensor records that noise and may make it significantly worse. De-mosaicing, the cameras IR filter, and signal amplification can alone create problems but the reason we spend a ton of money on those top of the line pro lenses is to get the best light falling on that sensor to start!

The rest of this comment I'll respond to another similar post about ISO levels and noise.

[Skyman]

This isn't entirely the whole story.

The two lenses expensive or 'kit' will produce a different and measurable color range/gamut in the image. The less expensive lens is far more likely to produce an aberration under the same conditions as the 'pro' lens unless they have identical coatings in which case the pro lens will be a poorer quality image. If not there would be little need for the more expensive lens as speed of the lens in alone isn't the only factor with its ability to produce a quality image. Some of the great images of photographic history were shot with lenses that started at about f/8 or even greater.

What you say? The larger primary element in the pro lens needed to produce the lower f-number requires a different quantity of flare and color correction coating to produce at least as good a quality of image as the cheaper lens. The nature of the materials used for these coatings also aid in shifting the chromatic problems either out of the visible range or 'compressing' them in a non-objectionable visual range.

Uncoated lenses from the black and white era, lens that were optimized for a particular type of film, and lens coating that are damage or have deteriorated demonstrate how much effect coating have when mounted on any DSLR. You'll see from those examples that a lens can produce gross problems in that area we call noise. Examples of color optimized lens offerings include the magic Kodak and Nikon lenses for the original Kodachrome.

Panavision made lenses for motion picture film for specific emulsion numbers and they need a lot of post work to get them to look right with a digital imaging system.

I have uncoated Leitz lenses that are optical clear as a possible but don't do well with any color film or sensor. I also have a wonderful Hasselblad lens from the olden days that hasn't ever worked very well on a digital back because of some internal flaw that doesn't show on film but creates micro flares and hot spots on the digital sensors.

My point is that there are several image issues that are noise in how we correct them. Color noise is the most noticeable in the average image. The sensor records that noise and may make it significantly worse. De-mosaicing, the cameras IR filter, and signal amplification can alone create problems but the reason we spend a ton of money on those top of the line pro lenses is to get the best light falling on that sensor to start!

The rest of this comment I'll respond to another similar post about ISO levels and noise.

This is great info, but It doesn't answer the question. The term noise is most often used to refer to unwanted information generated by a digital camera's sensor. This means that a lens can't effect noise except when the lenses performance changes the settings of the camera in such a way that the sensors performance is degraded.

Effects that are introduced from optical properties of a lens are most commonly refered to as aberations.

It is confusing in that the end result might be similar however to correct or adjust for a fault you need to identify where it is coming from.

[drg]

Noise is a function of the imaging sensor design. Since the sensor cannot be changed within a given body, the appearance of noise is solely dependent on the amplification (ISO) of the signal.

In other words, noise at ISO 1600 is not dependent on the lens.

Noise can be more noticeable against different backgrounds and if a scene is underexposed as others have mentioned (because the noise is further amplified during post processing).

But, I do wonder if the effect of poor bokeh from lesser lenses can give the illusion of more noise. I don't have directly comparable lenses to test that thought unfortunately.

There are different kinds of noise! The two most commonly thought of are luminance and color. They are not all from merely using a higher ISO value.

Luminance (brightness) noise may be unexpectedly a problem at low ISO numbers such as in an image with a lot of dynamic range when examining either the highlights or lowlights depending upon the range optimized for in the exposure setting.

There can be noise created by internal scattering in the lens before the light ever reaches the sensor! The light bounces back and forth between lens elements in the lens. Tele-extenders can cause this with a mounting ring that is more reflective than the lens in its internal light path.

A real world analog is similar to glare in bright sunlight. Or a dirty car windshield or camera lens. This will make it harder to see because of the extra light that is randomly scattered that lowers the ability to perceive a clear image. Noise. Noise reduction/analysis software will fix a lot of this but not all regardless of its source.

Sensors are the largest culprit with ISO related noise due to simply straight amplification, but that is not the only factor with Color Noise. Bloom occurs with some sensors that over saturate due to design choices when gathering light under extreme conditions too. How the lens is designed to transmit light across the spectrum (again those blasted coatings and optical material choices that the manufacturers are so secretive about) may not be proportional to the 'real' world. Some will produce dramatic and visible shifts in color. A sensor can then over or under expose and we have a visible problem that can be confused with sensor noise. Then the amplification makes it really bad. This too can occur at any ISO range.

Any lens can generate specific problems. Special cases include:

If the lens has not been optimized for digital sensors with a straight light path from the rear element to the sensor surface, the light will scatter across the face of the sensor and produce color noise because the relative intensity of the light is 'off' from what the Bayer type sensor is set up to decode. Film didn't care, sensors do and this noise generation comes entirely from the lens.

ISO generation in a digital camera isn't linear. A simple experiment is to shoot any scene with relatively unchanging light at several wide ranging ISO settings available in the digital camera of choice and the look at the Histogram for those images. It doesn't just shift from left to right. It also will change in relative intensity.

You may have to look at RAW data in third party tool to see this as I have seen a few of the newer camera models that are obviously playing with everything in-camera to let the user know what the image 'can be' rather than what is, particularly with 'help modes' enabled. These shifts can cause color clipping that a lens may multiply. The camera may 'know' about that sensor issue, but not about a particular lens and thus not be able to 'fix' the resulting image wide problem or 'noise'. It can look just like bloom in many cases.

Where a lens can create added or distinct problems magnified with higher ISO shifts is with falloff at the corners where it will cause a shift in the dynamic range of the image and the noise will be worse in those image portions affected by the light falloff. The sensor is thus underexposing in portions of the image and the combination of the sensor and the lens magnify each others problem. The lens actually creates the condition that produces the most noise.

This alteration in the histogram demonstrates one problem with lens design. The light at moment isn't changing but the way the camera sensor responds does. The lens isn't dynamic yet in consumer cameras so there is an optimum range established. Exceed that range and the lens will introduce problems. Misuse the lens (light directly on the front element will cause flare and is not always a spot or a bloom), shoot with a damaged or dirty lens and you can produce a type of noise in a digital image. I am not just talking spots either! Then there are the chromatic aberration issues (both color shifts are lens based) that sensors just faithfully record. There is an edge effect with sensors that I believe is frequently confused with CA but that is the more dreaded purple fringe as opposed the cyan/red or yellow/blue combos associated with the prismatic effects from mechanical optic mismatches or again, the lens produces noise.

NOTE:[ Electrostatic driven internal camera elements with polarization technology has been demo'd at M.I.T., R.H.I.O.T., CalTech and a few other of that kind of school. We can hope! ]

There is a whole set of problems associated with other facets of lens design that create other image problems, but many of them magnify noise in a digital camera. One that is very interesting is a computer designed lens that I owned that had a laser cut plastic element that is one of the best designed film zoom lenses I've ever owned. Yet on a digital camera it produces the weirdest color flares due to the discrete nature of each pixel in the sensor that captures edge effects from the micro prisms on the lens surface.

Lens problems will create noise. Sensors will create noise. They will multiply each others problems. It starts with a lens, good or bad, and the sensor records some portion of it all!

[drg]

This is great info, but It doesn't answer the question. The term noise is most often used to refer to unwanted information generated by a digital camera's sensor. This means that a lens can't effect noise except when the lenses performance changes the settings of the camera in such a way that the sensors performance is degraded.

Effects that are introduced from optical properties of a lens are most commonly refered to as aberations.

It is confusing in that the end result might be similar however to correct or adjust for a fault you need to identify where it is coming from.

See my second post. Bottom line is that it can be difficult to tell where the 'noise' starts.

The original questions was whether a lens choice effects noise. Yes, it does.

[DGK*CRONE]

If you put the 18-55 f2.8 pro lens on your camera instead of the 18-55 f3.5-5.6 kit lens and you changed nothing else then you would see zero difference in noise.

If you did the shot at 1600 ISO f5.6 on the kit lens then did the shot at 400 ISO f2.8 on the pro lens then you see a lot less noise in the shot done on the pro lens.

Noise is dependent on the ISO you choose. The ISO you choose (in low light) is dependent on the maximum aperture of your lens. A lens at f2.8 lets through 4 times more light then a lens at f5.6 so you can reduce the ISO by 4 times.

This answered my question and really helped me understand the difference a lens and camera body better.

THANK YOU.