Why is the pixel size and sensor size of a digital camera important?
Does increasing pixel count increase noise?
Big pixels vs small pixels
The purpose of this literature review is to provide links to a few articles on internet that discuss the topic of why the pixel size and / or sensor size of a digital camera are important considerations for buyers of digital cameras. As with any literature review, it is important for readers of this review to go to the source of each quotation so they can evaluate for themselves the various issues that are being discussed.
There is considerable debate amongst photographers about whether increasing the pixel count on the sensor of a digital camera will lead to increased noise at theimage level, particularly when images are recorded in low light and require a relatively high ISO setting. However, there appears to be little doubt that increasing the pixel count leads to increased noise at the pixel level.
The sensor type and size are also important factors, as explained, for example, here and here. The following is a brief literature review of some of the material on internet that deals with these issues. This page expands on the information given here.
This page has the following sections:
It is interesting to read some material published by Sony when, in 2012, the 20.2 megapixel Sony RX100 came on the market.
Here is an extract from a posting made in 2012 on the DPR Sony Alpha SLR / SLT Talk Forum:
"I think that when most cautious people are investigating whether or not to purchase a new camera that has a higher pixel count than the one they currently use, they want to ensure that the high ISO performance of the new camera is at least as good as that of the lower megapixel camera they are currently using. I don't think this is being unrealistic!
"Sony is very sensitive to the fact that many photographers appreciate the fact that increasing the pixel count increases noise and that they want reassurance that the new higher megapixel sensors have adequately dealt with this problem. Consider, for example, this statement from Sony about the high pixel count of the new Sony RX100:
Question to Sony:
"Why did you want a 20.2-megapixel sensor? I heard that this incredibly high pixel count would negatively impact noise levels, thereby decreasing image quality at high ISO settings.
Answer from Sony:
Ueda (Image Quality Design)
"It’s true that increasing pixel count increases noise. But since we manufacture our own sensors, we can easily tweak sensor specs to suit specific needs. This allowed us to craft a totally new sensor that delivers superbly detailed images with low noise. For high-sensitivity shooting we managed to reduce noise levels below those of existing Cyber-shots by combining technologies from Cyber-shot and α Series. As a result, we can shoot at up to ISO 6400 for normal photos and up to ISO 25600 when using Multi Frame NR."
This matter was also discussed further in this DPR forum posting:
"Also in the above statement, Kaimi (Product Design) said this:
"Some compact cameras keep noise down with lower pixel counts. But there are users who really want super-fine images, which require a high pixel count. But if the higher pixel count leads to unacceptable noise levels that the noise reduction system can’t reduce without affecting image quality, it is essentially self-defeating. Our new sensor achieves both super-fine images and low noise so even distant landscapes shot at telephoto focal lengths show great detail. People are going to love this compact camera for travel photography."
In December 2013, a contributor to a DPR forum agreed with the views expressed in the above statements and said that:
"Increased pixel count in same size sensor increases noise.. but you can improve the design of the sensor (that includes pixel level improvements and support electronics improvements to reduce that noise)."
A similar view about "lost sensitivity due to increased pixel counts" is taken in this article about cellphone cameras where it is mentioned that :
"Advancements in technology such as microlenses and back illumination, along with more advanced noise reduction algorithms, have gone some way toward making up for the lost sensitivity due to increased pixel counts in such a small sensor, but not far enough: today's cellphone cameras are still horribly noisy due to their small light-gathering area."
Yasuhiro Ueda: Statement from Sony's Image Sensor Business Division
On 30 August 2011, Yasuhiro Ueda of Sony's Image Sensor Business Division, Semi Conductor Business Group, issued a semiconductor briefing "Image Sensor Business". As part of this briefing, a diagram was produced titled "Misconception that reduced pixel size causes deterioration in characteristics". This diagram has sometimes been reproduced on forums to support the view that reducing the pixel size does not necessarily lead to a noticeable deterioration in image quality when the image is viewed at the image level. This might imply that the above statements from Sony that "it's true that increasing pixel count increases noise" and "some compact cameras keep noise down with lower pixel counts" may only be relevant if images are viewed at the pixel level.
Nevertheless, although debatable, it's quite possible that the above statements from both Ueda and Kaimi are referring to noise at the image level. The question put to Sony (as quoted above) seems to be referring to a concern about noise that might be visible at the image level, so it's logical that Sony's answers to the question might also be referring to noise that could be visible at the image level.
Note that one of the sensors referred to in the above semiconductor briefing is a back-illuminated 16.41mp CMOS sensor that Sony developed for use in cellphones. The image size is 4672 pixels x 3512 pixels and the pixel pitch is just 1.12 microns (with a sensor width of only about 5.23mm). Note also this section of this page which discusses why Apple increased the sensor size and the pixel size of the Apple iPhone 5S.
Under the heading "Effects of Sensor Size" this Wikipedia article distinguishes as follows between images captured in bright light and those taken using higher ISO settings:
"In the case of images bright enough to be in the shot noise limited regime, when the image is scaled to the same size on screen, or printed at the same size, the pixel count makes little difference to perceptible noise levels – the noise depends primarily on sensor area, not how this area is divided into pixels. For images at lower signal levels (higher ISO settings), where read noise (noise floor) is significant, more pixels within a given sensor area will make the image noisier if the per pixel read noise is the same."
Here is an extract from an article by Sony engineers about the Exmor CMOS imaging sensor developed by Sony for use in digital cameras:
"To achieve higher pixel counts, pixels must be made smaller. However, sensitivity is generally proportional to pixel area and will decline if pixel size is reduced. By combining its knowledge of the technology developed for CCDs, Sony has been able to compensate for the sensitivity loss caused by reductions in pixel size. The future goal is to combine higher speeds with higher resolutions while compensating for reductions in sensitivity resulting from reductions in pixel size. This will be achieved using device process technology, circuit technology and image processing technology."
Click here to read this article. Note the statement above that "sensitivity is generally proportional to pixel area and will decline if pixel size is reduced". This is referred to in the article under the heading "future challenges". As explained here "reduced sensitivity means less signal, which means lower signal to noise ratio (SNR) so you look at ways to reduce noise...Smaller pixels less signal per pixel.. when you do that you need to reduce the base noise that does not get reduced with the smaller size."
On 4 November 2013, the Digital Photography Review (DPR) website published a very informative interview with Kimio Maki of Sony titled "Every six months I want to do something new". This interview discusses several interesting topics, including the emergence of smartphones for digital photography and discussions about the rationale underlying several of Sony's cameras, such as the Sony RX100, Sony RX1, Sony A7 and Sony A7R. Kimio Maki also explains why Sony chose a small 1-inch sensor for the Sony RX10 camera, rather than a larger APS-C sized sensor.
In this DPR posting, titled "Big pixels gather more light than small pixels", it mentions that:
"If people do a reasonably conscientious internet search, they will find a lot of material that confirms what could be regarded as a “general rule” about the role of small pixels and large pixels, as nicely set out by Bob Atkins":
"So what you get with sensors is a tradeoff of resolution for noise (and the related dynamic range). Big pixels give you lower resolution, lower noise and higher dynamic range. Small pixels give you higher resolution, higher noise and lower dynamic range. The real question is what constitutes "High Enough" resolution and "Low Enough" noise. Can you have both? "
This useful article was published on September 13 2013 and it explains why the 1/3.2" sensor size on the iPhone 5 (4.54mm x 3.42mm) was changed to a 1/3.0" sensor size on the iPhone 5S (4.89mm x 3.67mm). Although Apple increased the sensor size of the iPhone 5S (when compared with the iPhone 5), the megapixels of both phones has stayed the same (8 megapixels). This means that the pixel size of the iPhone 5S (1.5 microns) is larger than that of the iPhone 5 (1.4 microns). This reverses the trend of reducing the pixel size so that the camera could have a greater megapixel count.
Article on Photography Stack Exchange: "Why and how do 'bigger pixels' make a better picture"?
This article was prompted by the iPhone 5S announcement in September 2013 when Philip Schiller, Apple's senior vice president of Worldwide Marketing, said that: " 'bigger pixels' is the key to a better picture". In answer to the question "what does this mean", it is mentioned in the article that:
"This is referring to the size of the photodetectors, which represent the pixels, on the camera's sensor. In short, the more light reaching these photodetectors, the more accurately they can measure light levels, thus reducing noise. To simplify things, noise is usually a factor in low-light photography."
The article also says that:
"The desire to cram more and more pixels into camera sensors has harmed the sensitivity of the sensors, particularly small sensors in cellphones and compact cameras, resulting in more noise at the same light levels."
Note that this article also has some useful comments from readers of the site and it is well worth reading the whole article.
This is a Youtube Video, which is described as follows:
"The Yale School of Engineering & Applied Science presents the 2011 Victor M. Tyler Distinguished Lectureship in Engineering with Eric Fossum, Professor of Engineering in Dartmouth's Thayer School and a consultant to Samsung Electronics' Semiconductor R&D Center. Dr. Fossum, who earned a PhD in electrical engineering from Yale in 1984, is one of the world's leading solid-state image sensor device physicists, best known for inventing the CMOS image sensor. His "camera-on-a-chip" technology is used in nearly all camera phones and webcams, digital-still cameras, high-speed motion capture cameras, automotive cameras, dental x-ray cameras, and swallowable pill cameras."
This video runs for about 1 hour and 7 minutes. At 43 minutes 22 seconds, Dr. Fossum showed a slide headed:
"Smaller pixels, more pixels". It is mentioned on this slide that (with smaller pixels) the performance of the pixel is usually degraded which results in a reduced SNR under low light and a smaller full-well, so worse dynamic range. Dr Fossum mentioned that when pixels are smaller, they collect less light basically.
In some postings to the "Bird Photographers" forum in 2012, Dr. Roger Clark produced some very interesting practical comparisons of cameras with pixel pitches of 4.3, 5.7, and 6.4 microns. He concluded that the 6.4 micron pixel images have the highest signal-to-noise ratio, but the least detail. This seems to conform with Bob Atkin's “general rule” stated above. This thread contains an excellent discussion about which camera to choose, one with "big pixels or little pixels?"
It is also very informative to read these articles on Dr. Roger Clark's website:
Note Dr. Clark's conclusion about the low light and good light implications "when choosing between cameras with the same sized sensor but differing pixel counts".
Dr Clark explains why large pixels alone do not improve high ISO noise performance.
Dr Clark explains why ISO is a relative gain, varying by camera and has nothing to do with sensitivity or true exposure.
This article discusses, for example, what the DxOMark sensor results mean and it gives the author's views on why large sensors outperform smaller ones. Peter van den Hamer says that "the good news is that more MPixels do not increase image noise - despite a widespread belief to the contrary". He discusses the reasons for this and mentions that you often automatically scale down to a lower resolution when you print or view the results.
In December 2013, contributors to a DPR forum thread discussed several of the issues relating to pixel size and noise, including this analysis by Peter van den Hamer. One contributor provided very informative technical evidence in support of his view that:
Similar postings were published here and here. This contributor also provided a posting titled: "Sony A57 vs A77: Proof that pixel count affects noise".
In June 2014, a DPR thread about small pixel vs large pixel sensor performance was started on the "Photographic Science and Technology" forum. The initiator of the thread asked for an updated discussion of a 2009 thread published on "Photography-on-the.net" that was titled "Small pixel sensors do not have worse performance". One contributor asked whether the 36mp full frame Sony A7R vs the 12mp full frame Sony A7S is a real world case study of large vs small pixels (see also this DxOMark sensor score comparison for these two cameras).
In August 2014, a DPR thread was started on the "Photographic Science and Technology" forum titled "Performance in low light - what is the limit?". You will find it interesting to read this reply from Dr Eric Fossum.
In September 2014, a DPR thread was started on the "Photographic Science and Technology" forum titled "Are large pixels a gimmick?". The original question in this thread asked "Can ISO and DR characteristics of a large-pixel sensor be achieved by down-sampling from a higher resolution (smaller pixel) sensor? For example think Sony A7S vs A7R." Note that the Sony A7S is a 12.2 megapixel full frame camera while the Sony A7R is a 36.4 megapixel full frame camera. This contributor provided reasons in support of the view that large pixels are not a gimmick!
The full frame Sony A7S is a 12.2 megapixel full frame mirrorless camera that has 4K video capability. The specifications for the Sony A7S are published here. The announcement of the Sony A7S was made on April 6, 2014 and can be seen here. The announcement mentioned that:
"The innovative α7S camera features a newly developed, 12.2 effective megapixel 35mm Exmor® CMOS sensor paired with a powerful BIONZ X image processor, allowing it to shoot at a sensitivity range of ISO 50 – 409600 with unprecedented dynamic range and low noise.
A comparison of the pixel size of the Sony A7S with that of the 24mp APS-C Sony A77II can be seen here. The pixel pitch of the Sony A77II is about 3.9 compared with about 8.4 for the Sony A7S. This relatively large pixel size of the Sony A7S has caused considerable interest amongst photographers, particularly those who are interested in photographing in very low light.
Before the introduction of the full frame 12.2 megapixel Sony A7S, Sony had marketed the 24.3 megapixel full frame Sony A7 and the 36.4 megapixel full frame Sony A7R. The fact that Sony then introduced a third E-Mount full frame camera with substantially larger pixels than those of the A7 and A7R, would seem to indicate that, because the sensor size of the A7, A7R and A7S is identical, the larger pixel size of the A7S has a lot to do with the aim of improving the low-light ability of the A7S when compared with that of the A7 and A7R.
For example, in a review published (on the petapixel.com site) on 30 July 2014 titled "Sony a7S Astrophotography Review", Ian Norman explained that:
"The A7s has larger pixels than pretty much any consumer level full-frame digital camera. The larger pixel size means that each pixel can collect physically more light. The more light per pixel, the better the signal to noise ratio for that pixel and so that pixel will more accurately detect the incoming light than a smaller pixel would. This means that, all other things being equal, the A7s should be capable of the best per pixel signal-to-noise ratio of any production camera. This means that it should arguably be the best camera for astrophotography yet."
It is recommended that you read the full article and see the excellent images of the "Milky Way" that were captured by the Sony A7S.
Distinguished photographer and teacher Reed Hoffmann explains why:
The “Cambridge in Colour” tutorials acknowledge that the amount of background noise also depends on the sensor manufacturing process, but they nevertheless seem to support the “general rule” as stated by Bob Atkins:
“Larger pixels receive a greater flux of photons during a given exposure time (at the same f-stop), so their light signal is much stronger. For a given amount of background noise, this produces a higher signal to noise ratio — and thus a smoother looking photo. This is not always the case however, because the amount of background noise also depends on sensor manufacturing process and how efficiently the camera extracts tonal information from each pixel (without introducing additional noise). In general though, the above trend holds true. ”
From the section titled "Myths and common misunderstandings: Larger sensor systems have less noise because they have larger pixels / higher ISO's result in more noise",Joseph confirms the view that, the smaller the pixel, the less light it can gather, as summarised below:
"Thus, the sensor with more pixels will record more detail, but will appear more noisy at 100% view, because each pixel, individually, collects less light. However, that additional detail afforded by the smaller pixels can be traded for less noise by the application of NR (noise reduction)."
James also has other sections of his paper that deal in considerable detail with noise, sensor size, high ISO's and pixel size.
Emil Martinec supports the view that the “light collection efficiency per unit area is essentially independent of pixel size”. However, with regard to low light (high ISO) applications, Martinec concludes that:
"There is an advantage to big pixels in low light (high ISO) applications, where read noise is an important detractor fom image quality, and big pixels currently have lower read noise than aggregations of small pixels of equal area."
Click here to go to an article titled "Advantages and disadvantages of cropping images to gain extra reach".
Click here to see an index of camera comparisons showing the mathematical relationships between image size, pixel size, pixel density, and reach etc.
Click here to go to the full explanatory article about the crop factor and “telephoto advantage” of an APS-C camera.
Note that Appendix 2 of the above article includes the following sections:
The following supplementary notes are designed to give you further information about how to compare the cameras listed in the above index:
Click here to see a comparison of two “theoretical” cameras, which permits the reconciliations between the percentages shown for pixel density, and pixel area, to be exactly equal.
Click here to see examples of the outstanding resolution of the full frame Sony A99 and Sony A900.
to go to the home page of Rob’s Photography New