The battle of megapixels. Thoughts from Radozhiva

Add your thoughts in the question about megapixels in the comments.

The battle of megapixels. Thoughts from Radozhiva

The battle of megapixels. Thoughts from Radozhiva

Nowadays there are quite a few cameras with matrices with a very large number of pixels on board. Many amateur photographers are sure that the more pixels the better, but there are nuances and subtleties.

Some disadvantages of cameras with a large number of pixels:

  1. The more megapixels there are on the camera's matrix, the smaller these pixels are. In general, the smaller the pixel on the camera's matrix, the less light it can absorb, and the stronger the noise in the picture can appear. In general, a small pixel size (large number of megapixels) results in more noise. Fortunately, scientific progress does not stand still, and each new generation of cameras simultaneously increases the number of pixels and tries to reduce the level of noise produced. But there are certain limitations that are very difficult to go beyond.
  2. Cameras with a lot of pixels produce huge files, especially when shooting in RAW format. For example, files from a 36-megapixel monster Nikon D800 weigh about 70MB. The process of copying the shot material from a small photo shoot to a computer with an HDD with a hard drive can take tens of minutes.
  3. Processing images consisting of a large number of megapixels requires very powerful computers.
  4. On cameras with more megapixels, it is more difficult to achieve 'per pixel' sharpness. For super sharp images, you need use good resolution lenses... These lenses are usually very expensive. Therefore, it is customary to say that a matrix with a large number of pixels is more difficult to "feed" with ordinary lenses. Often times, closing the aperture well is enough to achieve a super sharp image.
  5. A matrix with a large number of pixels has a lower diffraction threshold. This means that on cameras with a large number of pixels, it is more difficult to achieve a sharp image with closed apertures. This is a serious problem with such cameras. More details about this can read here.
  6. When using matrices with a large number of megapixels, it is more difficult take off due to pixel-wise movement. On such cameras, you have to reduce the shutter speed to avoid grease.
  7. In fact, modern digital cameras do not have real pixels, in the usual sense of the word, you can read about a kind of `` wrapping '' pixels in the section about Bayer pattern.

Advantages of cameras with a large number of pixels:

  1. The possibility of a strong crop of the picture without much loss in image quality. This is very convenient for post-crop. But at the same time, in order to take advantage of this advantage, it is necessary that the original image be very well detailed, which in turn imposes restrictions on the optics.
  2. Ability to print on materials of a very large format, for example, on sheets of A1 format and more. This is a very strong point of such cameras. Sometimes this is the only serious argument in favor of cameras with a large number of MPs.
  3. In most cases, multi-pixel cameras have very wide dynamic range and large color depth, It is very important for accurate color reproduction and post photo processing.
  4. Cameras with high density sensors may not use an AA filter. At the same time, the sharpness of the image can significantly increase, but the possibility of obtaining moire in photographs also increases.

My experience:

As a photographer, personally for myself once and for all concluded that megapixels play a much smaller role than many people think about it. I very rarely need strong framing, I print canvas with wedding photos from my 12 MP camera without any problems, I have a huge number of printed A4 glossy photos from 6 megapixel cameras. Larger pixels allow the use of non-top-end lenses, and batch processing of photos takes much less time :).

On the topic of megapixels, I also advise you to look in the sections'Pixels and Subpixels", "Gigapixels'and'Megapixel Math'.

Material prepared Arkady Shapoval. Training/Consultations | Youtube | Facebook | Instagram | Twitter | Telegram

Add a comment: Alexey

 

 

Comments: 172, on the topic: The battle of megapixels. Thoughts from Radozhiva

  • Sasha

    Diffraction is always present at any aperture, only on closed apertures it is better visible, since diffraction is created by rays passing along the edge of the aperture blades regardless of the size of the “hole”, with a closed aperture the number of non-deflected rays is comparable to the number of rays deflected from the trajectory and diffraction is more noticeable.
    When filming it was believed that the finer the grain of the film, the sharper the frame, the finest-grain films were used for large image enlargements, it never occurred to anyone to enlarge the frame to the extent that each grain could be seen separately and talk about some kind of grain sharpness. I think if a frame with a large number of pixels, say 24 pixels, is reduced to a size of 6 pixels, then with a closed aperture, say 13 (which is still valid up to 6 pixels without the influence of diffraction), both frames will be sharper with 24 pixels.
    Now monitors are produced with different resolutions. On monitors with small dots at 100% magnification, the photos will also look sharper.
    I think the number of pixels as matrices and the resolution of the monitors will not stop.

    • juris

      for a large hole, what is the diffraction? what are you talking about ?? ... there was no need to skip physics !!!

      • Michael

        There is diffraction at any hole - this is just physics) just the size of the Euler circle is incomparable with the size of the fixing cell (pixel, grain, etc.)

  • Samuel

    you said that you print canvases. is how much how many cm?

    • Arkady Shapoval

      I periodically print canvas for clients, A1 format, approximately 594 x 841

  • Jury

    shooting on canon 400d. With this device you can get beautiful pictures. you need a little head and hands. Well, let pimples verb up to the foam. Truly, I look at canon 70d. Ff and canon 80d require a good lens.

    • Lynx

  • Max

    I’ve been shooting for 7 years on Canon 1d Mark 2, I want to buy 1D Mark 2, it suits 90% of tasks. Optics L. I tried to shoot at 5dm2 and tripyatak, a completely different picture, I do not like it.
    My wife Nikon d40, a gorgeous picture, but 3 focus points is hell (when I took it after my unit (45 focus points), I almost cried)). But the picture at d40 is gorgeous
    Now I am faced with a problem, but it is inherent in optics (70-200 L f4), without a stub, for 200 micro-shifts, because of this, photo banks often reject photos. You can solve the problem with a monopod or glass with a stub.

    • Karen

      “But 3 focus points is hell” is strange, but I use one central focus ... ..

      • Oleg

        Yeah, from the category of “what prevents a bad dancer” ...

        • Karen

          Probably ))). As soon as it became possible to choose the number of focusing points, I immediately switched to one central one, and since then I have been photographing it, pointing it at my eyes and shooting. First of all, we pay attention to the eyes, and if they turned out to be sharp, then the portrait was a success. I also read somewhere that the most accurate setting for sharpness is through the center point.

          • Ivan

            Colleagues, have you tried 45 points? This is the purest buzz. I also did not understand why and focused one by one, but after trying it, I understood. 3, 9, 11, 19 - weren't lying around, to be honest.
            Now I grieve that I have to be content with 9th, I also use one in the center of 95% of the shots.
            You can live and shoot with one central one - I don't argue, and it doesn't really bother you, but 45 pieces are completely different.
            It's like shopping dumplings (which you can live for a week, probably) to compare with food in a restaurant.

            • Karen

              Can the camera set maximum focus on each individual point? Of course not, she sets the average focus for all using her algorithm. So why do I need such a focus, if I can choose the desired point myself and focus on it ... .. eyes.

              • Karen

                And keep in mind that this is not the photo itself, but a 3 megapixel clipping of a 10 megapixel photo taken at a shutter speed of 1/15 sec. 40mm lens on crop with hand. It was still necessary to set the sensitivity to 400 (quite a working number for Knon 40D.) And the shutter speed to 1/30 sec, because there is a small stir. And the original photo looks like this.

              • Ivan

                What does “maximum focus for each individual point” mean? The camera focuses and illuminates those points that are caught in the grip. To select another (others) - turn the control wheel, everything is simple and fast.
                What did you want to demonstrate with a photograph? I do not understand.
                You can choose from the available points on any camera and focus on it.

              • Karen

                I see no point in having the camera choose the focus point for me, i.e. if it matches my desire, then it's good, if not, then you need to turn the wheel. Why not select the focusing point yourself from the very beginning and not aim using this point? And why the central point, because I read somewhere that the central point is the most accurate. With this photo, I wanted to demonstrate the accuracy of aiming one central one on the eye.

  • Zhorik

    They were modest that you are Max, I hoped that you had a Hassel ... oh, disappointed with your fantasy)

  • Alex

    Are you saying that if I shoot not at 24 megapixels, but at 13 ... by reducing this in the camera settings, I will get a better picture?

    • Vladimir

      Image quality is a relative concept, what exactly is meant, sharpness - color rendition - noise. Sharpness can be subjective and will increase, but not a fact, but the noise in theory should decrease, since the camera software - with decreasing resolution, processes neighboring pixels as one, as if averaging their values.

      • Karen

        And in my opinion, choosing a lower resolution, we simply turn off the pixels on the periphery. This is clearly visible on compact cameras when switching from 4: 3 to 3: 2. For example, Panasonic LX1. With a 16: 9 format, the maximum resolution is 3840x2160, and when switching to a 3: 2 format, the vertical resolution remains the same - 2160, but horizontally decreases to 3248, and when switching to a 4: 3 format, the resolution decreases to 2880 with the same 2160 vertically. Those. the horizontal side pixels are being turned off. Therefore, when switching to a lower resolution, the digital noise does not decrease.

        • zengarden

          When decreasing the size at the same aspect ratio, no "pixel off" occurs, everything is used, but with intra-camera interpolation. The quality of the picture will not be better than if you resize in the editor; only the file sizes are smaller.
          If you change 16: 9 to 4: 3 or 3: 2 - yes, the picture will be cropped by the camera to the desired aspect ratio.

          • Karen

            There is such a format, sRAW, with it 4 nearby pixels are combined into one, i.e. if your camera has 16 megapixels, sRaw is shot at 4 megapixels, and although the number of pixels decreases by 4 times, the size of the photo (in Canon) is reduced by only 1,5-2 times. Some notice a decrease in digital noise. On the example of Panasonic, the number of vertical pixels remains unchanged, only the number of horizontal ones changes. As a result of interpolation, the photo quality should deteriorate, but this does not happen, and the camera takes the same quality, only the shooting angle decreases.

  • Romanov

    Those who did not hold the D800 in their hands will always be afraid of it, and adhere to the myth (it's not about the pixels)
    After picking up the D800, he no longer wanted to pick up his D3 and D700, they seemed to me just junk. Of course, there are difficulties in a large number of pixels, increase your level, and don’t

    • Vladimir

      I support. It was also d800. Then I bought a sony a99, with a matrix stub. I sold it, feeling a shortage of 36 megapixels, although I liked the Sony, a very understandable camera, even a child can easily remove it. Now I'm saving up for nikon z7.

  • Romanov

    In the very near future, cameras of 150 and 200 megapixels are waiting for us, progress is growing. Stock up on RAM) In about 15 years we will recall 36 mega pixels with a laugh

    • Alexey Shch

      Many are very happy with 6mp cameras from 2005) Even on A4 prints.
      I speak as the owner of cameras with a much larger number of megapixels than 36))

      • Boris

        Of course it will suit 6mp for many, the print quality remains at the level not only of 2005, but of the last century :)) Now try to print on the Noritsu QSS 3801 (640dpi) photo lab, how then do you look at 6mp? Over36mp of your camera will come in handy)

        • Paskuda

          I live in Nizhny Novgorod. There are a couple of normal photo labs in the city, the rest all print pictures on inkjet shitty printers. Tp-shkam will print pictures of his children from the phone, but I see the stripes of an inkjet printer. Or laser. I don't understand them. Therefore, if I print photos in rare cases, then I go to the photo machine. One here was closed a couple of years ago, broke. They say it's expensive to repair and maintain. And I looked where we have on Noritsu (I learned from your comment) they print, so I didn’t find it in Nizhny Novgorod. One only in Dzerzhinsk.

  • Serge

    Good day. I read your public and appreciate, probably already five years. I am writing for the first time) The question may seem silly, but I would like to hear the answer here. Actually, what dpi do dslr cameras have in rav files? Does dpi change with camera compression in jpeg? Well, why do you need 300 dpi, it’s clear what to print, but why exactly 300? Thanks. Happy holiday to all

    • Boris

      Let me answer, the topic of high-quality photo printing is very partial to me. To begin, I’ll write that I think about 300 dpi. A long time ago someone decided that a person can separately see in the picture only up to 6 pairs of lines / mm., Which corresponds to 300 dpi. This quality was taken as a standard and most photolabs can not give out more detail, even if there is a file with a size of 100 megapixels. You can still agree that 300 dpi is enough for large print formats, from A3, A2 and more. Such prints are more convenient to view from a greater distance and they will look clear. And the most common formats 9x13, 10X15, 13X18, 15x21 and 20x30cm are examined point-blank, just like a smartphone or tablet monitor. Here it becomes noticeable that 300 dpi is not enough. It’s not just that smartphones now have displays with a pixel density of 500 dpi and higher. Therefore, for really high-quality printing, you need a photo lab with the same dpi density as a smartphone or more.
      Is there a connection between 300 dpi and the rav-file, then only in max. the size of the finished image at which sharpness is not yet reduced. Same thing with jpeg. And in the camera, jpeg really shrinks, when compared with the same frame, converted from a rabbit. The size in pixels of the ravah that the jpeg of the camera may be the same, but the sharpness on the ravah is higher. Therefore, jpeg needs to be done independently from a rab on a computer, if maximum quality is required. For myself, I concluded this: for a photolab, you need from 600 dpi, for a camera, you need from 100mp or higher to print quality up to 30x40 cm or photo books. I completely agree with the reader Romanov about future cameras at 150-200 mp.
      Well, I wrote, as I understand it myself, maybe Arkady or other readers will explain it even differently.

      • MISHA

        But is there a lab with 500dpi? For film Kodak hectare 100 and 300dpi probably a lot!?

        • Michael

          To scan film and 2000 dpi is not enough. Talk about the fingerprint. Although from the standpoint of my eye the above is written nonsense. I do not see more than 300 dpi

    • Michael

      dpi RAW no. As many as you ask, so much will be. There is a resolution of the image is measured in pixels (dots), there is a resolution of the printing equipment is measured in dots per inch (DPI). The final image size in centimeters (meters, millimeters) will depend on the ratio of image resolution to print resolution. The concept of dpi has no relation to data formats (RAW JPEG). Namely, 300 dpi is accepted in print, because the human eye cannot distinguish higher resolution, i.e. typing in higher resolution simply does not make sense.

  • anonym

    Who writes what, the discussion is not right at all, there are two types of photographers, those that work specifically as photographers in advertising studios, and so on, and there are creative photographers who care about the plot, so that they have 6 pixels and 24 they care about the plot and plot, it is clear that in the new cameras there is almost no noise, on the other hand, the old cameras give a picture more pleasantly, and this artistic noise and hair can be torn by beginners only

    • Peter Sh.

      What a horror. So creative photographers cannot be professionals?

      • Major Mahoney

        It is impossible to optimize the task by two or more parameters. This is how I was taught at the Soviet institute. In the people it sounds like you chase two hares, you won't catch a single one. I think so…

  • anonym

    Counted a little https://ibb.co/casiym

    • Alexey

      That is, for printing with a resolution of 300dpi of a picture measuring 29 by 19 cm., For viewing from a normal distance, an 8 megapixel camera matrix is ​​enough.

      • serge

        It turns out - yes. And Arkady wrote about this more than once.

      • Valentine

        And for printing 15 x 20, 4,3 megapixels is enough. Both large and for money is good. On a Full HD monitor, such a 4,3 megapixel image can be viewed with a twofold increase. That is, even the 6 megapixel matrix has the ability to crop without damage to the print. Given that the density of 300x300 is very good, and often, at a density of 200x200 people cannot see the difference, then, in fact, the final resolution may be even lower. This can be used not only to save space, but also for technical improvement of the final frame.

        • serge

          https://ibb.co/eXC5w6 5 megapixels soap dish Olympus.

          • Valentine

            No comments

          • Valery A.

            Well, a soap dish, well, legs, but what's the catch?

            • Michael

              Probably in the artistic blur frame

            • serge

              We are discussing megapixels, not the artistry of photography. Specially selected a photo without faces (or else they will file a lawsuit :)). For comparison. According to the detail on the screen and when printing regular size photos, the difference between 5 megapixels and 20 is very difficult to notice. One huge plus of the big megapixels I see is that you can cut off the floor of the frame and you get a completely normal photo in detail.

              • Michael

                You are right, there will be no difference. Just an example, well, completely soapy, even for viewing at a 2 megapixel screen and a joke no one understood)))

        • anonym

          Yes, no joke, just a 6 megapixel photo for comparison on the screen.

  • Ksenia

    Arkady, so tell me, an ignorant amateur of high-quality photography, will I jump over to the D3100 with the Nikon D800 or will there be enough potential for the D700 (I want to take the secondary)? But sometimes I photograph on a commercial basis, that is, the requirements for the picture and quality and the device itself are above average.
    I read articles about both, and remained in some confusion. :)

    • Igor

      If you need a good picture, take Nikon D4s. You will not regret. The hurt of this unit is enough for many years.

    • Peter Sh.

      Ksenia, you understand what’s the matter. Ask which prof. the camera is more suitable for you, the same as asking which concert piano will sound prettier. The best thing you can answer is one of them is new, and the other is white.

      No one except you can know this. Good luck!))

      • Ksenia

        Thank you, but the fact is that I read that a large number of pixels sometimes complicates the work, that is, it is more difficult to make a sharp frame, and that the d800 is not friendly with inexpensive glasses .. but the d700 is not at all modern, Probably, doesn’t it make sense to buy an outdated device, even for a start?

        • Valentine

          Perhaps both would not make sense to buy. Judging by your text: 1 - you do not shoot very much, therefore the capabilities and resource of these cameras are excessive. 2 - you are going to save on optics, etc. - irrational budget allocation, not the best results in the output. What doesn't suit the D3100, what exactly do you want to see in the new device, what are the flash optics, how much do you want to spend on a new camera and optics?

          • Ksenia

            Valentine, I shoot quite often and would like to grow in it. Very often I have to shoot baptisms, I suppose that the D3100 is much inferior in ISO to the above devices. Well, the quality of the picture is also a very important factor (as for printing in large format and photo books). As for the lenses - yes, my budget does not yet allow using the manufacturer's optics. At the moment, Nikkor has 50 1.8, Sigma 17-70, Helios 44m4 and a flash from Sigma in the arsenal.
            Well, I think, in the region of 50 tr to spend on a new camera. :)

            • Michael

              Download RAW from D800 and look, compare with your 3100. It is unlikely that you will see a difference in noise. Spend 50 on a box and 17-70 will fall on a shelf (or shooting in crop mode, which makes no sense). Will you use 1.5 fix? Very irrational

            • Peter Sh.

              Ksenia, in this case, the D610 / D750 is more suitable for you.
              Also need at least Nikon 50mm f / 1.4.
              But even with the lenses that you have, the image quality in dim light will be incomparably better than that of the D3100. Both in noise and in color. If the D800 is better, then slightly. However, it will be more difficult to handle and handle.

              • Paskuda

                As far as I understand the D800 is a more professional full frame than the D750. And you advise the opposite.

            • 1Dl3

              Ksenia, given your budget, it is irrational to take the D800 “because I want”.
              Well, you buy it in a so-so state with a range of 50 to 100 thousand frames, but you have no money for normal optics ...
              All the same, the D800 will require decent glasses, but you do not have them. Not a single corresponding matrix of this camera.
              In general, the abundance of D800 on Avito should hint at something.
              People just rushed to get rid of them. They don’t get rid of good cameras ..
              D800 is a potentially problematic camera that will torment you with its dull green look at the concept of "White Balance"

              • 1Dl3

                And also make you know what SHEVELENKA.
                And the ISO workers there are worse than on the D5100 .. Surprise surprise?
                Take the NEW D750 on credit with a whale 24-120 / 4G. If you know how to take pictures and shoot often, then you will quickly beat it off, because this is a working machine, not a stand-studio camera ala D800 / D800e / D810 / D3x for shooting with the right expensive light.

              • Vladimir

                Well, you and the troll, mixed d3x with d800. Stomp out of here

              • Paskuda

                But why do you all advise an almost amateur full frame in the face of the D750, when the D800 is more professional. Does the D800 shoot better than the D750? And that the D800 is problematic, not like a Kalashnikov assault rifle?

              • Dmitry Kostin

                Damn, I hope this coronavirus is resolved and I finally get a D3x from Japan. A friend bought a Z6, immediately sold it (after 2,5 months - an anti-record), now he has Z6_mk2. D3x sits on a shelf and gathers dust. For one thing and compare it with 1Ds_mk3. I really hope that it will have the same color profile as the D3s, I really liked it.

              • Dmitry Kostin

                2 Paskuda:
                It is not the camera that shoots, but the photographer.
                Now the Z5/Z6/Z6mk2 will be more interesting and promising - these cameras, like the D780, and the Canon EOS R have fallen in price quite well in a couple of months.
                D810 also fell in price.
                If you look from the point of view of the reliability of used cameras, repairmen advise the eight hundredth series and D4 / D4s, etc.

  • Я

    Well, canvases are not an indicator even once. The structure of the canvas fabric creates the illusion of detail, sharpness, well, by analogy with the slight “noise” of a soap photo, in order to create the illusion of fine detail. But on photo paper 12 megapixels when printing at least 60x80 nothing at all, plotter printing or photo-chemical printing is not at all important. Moreover, even when preparing for publication on the WEB, multipixel images when reduced look much more detailed than low-pixel ones ... so the opinion that “why overpay for many pixels, is enough for me” is actually profane, if the task is worth getting a commercially viable product, not taking pictures on vacation.

  • Igor

    I definitely bought this canon 3Dsr 5 years ago and realized that I spend more time on processing, longer and the files aren’t lifted. But with 1D in good light and the Kodak matrix, it’s more pleasant for the eye than 5Dsr. pumping money.

  • Valentine

    Sony announced on July 23, 2018 a new sensor for smartphones with 48 effective megapixels (8000 x 6000). The description and pictures hint that with such a sensor, you can see small details like in a telescope. Earlier, at the end of 2017, for adult cameras (medium format), Sony announced 100 and 150 megapixel sensors. In general, the battle of megapixels is in full swing on all fronts.

  • anonym

    Lord! Got a question. It is known that a camera with a large physical pixel has some advantages. (Noise, diffusion circle, etc.)
    And if you set a small image size in the camera menu, will it be the same? What then happens? Can small pixels-neighbors merge into one large? I understand that this is not the same thing, but is it possible to achieve a similar effect in this way?
    I apologize if there is already material on this topic. I haven’t met yet. Thanks.

    • Past the crocodile

      No, they cannot).
      The decrease is due to essentially intracameral dill.

      • Past the crocodile

        In general, the user with the nickname 1ds is right, but in general.
        Under new cameras for 20mp new glasses. Under the top bz from the boot and Nikon, only expensive news.
        And nothing else.
        Once Arkady was already counting the film on mp. It seems in the review of a film niche. It turned out roughly 24-26 as I recall.
        From experience, for example, Nikonovskaya film 80-200 Ed 2.8, the second mod plays on all about 2-3-4d where up to 16. At 750k, for example at 2.8, these are already tears. Like a medium-sized boot whale for the 40th late 90 - early 00.
        Takumara, for example, film is good, ideally straight up to 5 from the stump with her 16. On k-1 with her 37, well, that's still clownery. There are discrete currents and then not all, and already not as on 5ke - there is nothing to sprinkle with.
        Karoch for manuals from Nikko recommend d4, d3. The first one is direct good, the second one is generally ## smiling. 12 straight exactly))).

  • Michael

    Questions about “fat” and “thin” pixels. If we assume that the quality is transferred to 100d from L to M, and then to S, then how will this affect the density or pixel size. And will the lenses for the film give an acceptable sharpness and others. Or is “quality” something else? Of all the experts said above, this remained with the card ...

    • Novel

      Are you talking about LMS in JPEG or RAW? In JPEG, this is a normal downsize. Everything is more interesting in RAW.

      Arkady has written about debayerization. Everything is removed from the matrix in this form:

      RGGR
      GBGB
      RGGR
      GBGB

      It is a square of 16 “pixels”, each “pixel” has 14 bits of brightness and is covered by a filter that allows a certain part of the spectrum to pass through. Here, in fact, in this 4x4 square there will be four measurements of red, 8 measurements of green and 4 measurements of blue, which, due to debayering and interpolation, will turn into 16 ordinary RGB pixels.

      Here is a regular RAW - this is a sample from the matrix before debayering, reaped by the Loseless method (without loss).
      JPEG, M-JPEG and L-JPEG are the result of processing a normal RAW file using the camera - debayering and basic color correction, possibly adaptive to the scene. This is where all lovers of “color” and “skin tone” break their spears, and in fact the camera's ability to process RAW according to its guided rules and taking into account user settings - color temperature, saturation, brightness, sharpness. LMS-jpeg is downsize RAW processing by one third and one half.

      M-RAW and S-RAW, from what I remember, do not contain a sample from the matrix, but already the result of debayering. And not in RGB format, but in YCbCr (or YUV) format - that is, one luminance pixel and two color-difference pixels. And if in the usual RAW of 16 pixels of the matrix (each of which encodes only one color channel) we get 16 ordinary (bloated) three-channel pixels, then in M-RAW it will already be 9 pixels, and in S-RAW - 4. Debayering occurs according to different principles with subsequent conversion to another color space. RAW converters understand this format and allow you to do all the same actions with it as with ordinary RAW, but everything happens according to different algorithms and does not allow extracting some extreme things that can be done in ordinary RAW.

      • Novel

        Not quite correctly conveyed the thought. The drawing is simpler:

        RAW

        RGRGRG
        GBGBGB
        RGRGRG
        GBGBGB

        can be converted to 16 RGB pixels, some of it is "composed" by interpolating adjacent pixels using complex debayering algorithms.

        M-RAW

        YUV, YUV, YUV
        YUV, YUV, YUV
        YUV, YUV, YUV

        16 measurements of ordinary RAW turned into 9 pixels in YUV format, the debayerization has already passed according to other algorithms. The conversion to RGB is trivial, the result will be 9 regular RGB pixels, the image is 1.33 times smaller on each side.

        S-RAW

        YUV, YUV
        YUV, YUV

        16 measurements of ordinary RAW turned into 4 pixels in YUV format, then everything is like M-RAW, the image is less than 2 times on each side.

    • Novel

      Answering your question directly, shoot in the usual RAW, it has more options for image processing, you can sharpen in Lightroom / Adobe camera raw at the conversion stage, additionally with noise cleaning, and then additionally sharpen in Photoshop.

      The most effective method of sharpening in Photoshop is to create a copy of the layer, use the High-pass filter on the copy, change the blending mode to Overlay, change the opacity of the layer if necessary, if oversharp. We select the radius in High-pass depending on the size of the image and the subtlety of the details we need. In principle, you can gently drag the slider up from the minimum until a color appears on the gray filter mask.

      • Michael

        Yeah…! Difficult for a schoolboy with an academician ...
        I'll ask quite simply with the hope of a simple answer. Why you can shoot an old ef at 40d, but not at 100d. And why do I need raw, I just need a sharp ipeg!

        • Novel

          When shooting at 100d JPEG, compressed to the same size as 40D should be at least as good.

          • Michael

            Already good! And then - what is a “fat” and thin pixel? It is clear that the pixel density on the 20d matrix, reduced to the FF size, will be greater than that of 6d. But are they the same fat? Or is it necessary to talk about other matters here?

            • First name

              you have an LCD TV, the picture is pseudo-numbers from it so-so, if you move 3m away, the picture gets better, and if there’s fire even 10m away)

              the camera simply downsizes the frame size depending on the M / L preset, you can just as well open Photoshop and put even less, say 1280 * 960, and in general the picture will become a gun, and it doesn't matter what fat there is a pixel, everything will be more beautiful on the wrapper )

              • Michael

                Yes, I have a simple question: is the pixel size the same or different for different matrices? Well, what is the question and the answer, please!

              • Roman

                The higher the matrix resolution, the smaller the pixel size, of course. This is if matrices of the same size are compared - DX / crop / APS-C or FX / full frame.

  • Michael

    Those. pixel size depends on density and matrix size. So? Because they cover the entire area of ​​the matrix. This is understandable. On the FF, the pixel is “bold”, maybe there are less of them than on the cropped one? With this sorted out! Now the next question: why is bold better than thin? And also interestingly, with an increase in the density on the FF, the pixel size loses weight and approaches the old cropped ones? So what's the progress ?!

  • Stanislav Ts.

    Arkady, good afternoon. In the article you wrote that the advantage of many pixel cameras in a wider DD. The question is: where does this come from? Shouldn't a thicker-pixel matrix have a higher DD due to a thicker pixel and its larger, so to speak, photon capacity?

    I looked at DxoMark FF cameras with 12 megapixels (Nikon D3 / 3s, D4, D5, Sony A7s) - all of them have low DD (at least in the tests) compared to 24-36 megapixels. Sonya was especially surprised: A7 (24 Mp) has DD as 14.2 feet, Sony A7S - 13.2. I don’t think Sony’s engineers have specially created a matrix with a reduced DD for the A7S.

    Nikon D5 also surprised: DD has the same 12.3 stops as D3s. I doubt very much that Nikon's engineers will pump up and put a poor matrix in the top camera.

    The third point: I pulled shadows and overexposures with Pentax K-5 II (crop, DD 14.1 stop by DxO) and with D700 (DD 12.2 stop). Feels like the picture from the D700 stretches much better both in highlights and shadows, although in theory it should be the other way around.

    The feeling that DD is tied to the pixel size in addition to the matrix quality itself, but the dependence on the pixel size seems to be the opposite.

    • Dima

      It’s probably no longer relevant, but I’ll put in my 5 cents. Only photo wankers believe in Dxo. Their Nikon D5100 has 13.6 stops, and the Canon 5dm3 has 11.7 stops. This is already funny because it’s a full frame, and it’s not the most ancient, and in Africa it’s a full frame. Cameras of approximately the same period. But in reality everything is different. Where the d5100 would have asked for bracketing, the 5dm3 calmly managed with one frame and there were no failures in either the light or the shadows. The RAVs with the 5dm3 are very elastic, but a little worse than the d800. Everything fit well and there was no need to fiddle around with the sliders too much. But in terms of noise, the D800 is really noticeably better, that’s true.

      I'm more interested in the question, if a larger pixel catches more light, therefore on ancient cameras where there were 6-10 or on a full frame 12, for example, should there be, all other things being equal, faster shutter speeds, especially when there is a lack of light?

    • Dima

      It’s probably already irrelevant, but I’ll say it, I’m more interested in if larger pixels catch more light; it turns out that matrices with large pixels, all other things being equal, can allow shooting with shorter shutter speeds to obtain the same exposure in the frame?

      Dxo can't be trusted. Their d5100 has 13.6 stops and 5dm3 11.7. In practice, it’s the other way around) It’s hard to compete against a full frame no matter how you look at it.

      • Dmitry

        A larger pixel catches more light, which has a positive effect on the signal-to-noise ratio. Such matrices can work at higher ISOs than matrices with small pixels. But this is true for matrices of the same generation. For example, 6MP SSD Nikon matrices from the mid-late 800s are not capable of going above ISO3200. And modern small-pixel crop matrices work quite well 7000 with good quality of the output signal. Signal processing also has a big impact. More advanced and heavier algorithms allow you to pull out a picture by a stop or two, all other things being equal. For example, developers from the old Nikon D5000, when developed with DXO pureraw, make it possible to shoot even large children’s portraits at ISO up to 6400, and everything else at ISO 12800 and push-ups up to 800 in post. Using the stock in-camera algorithm, I tried not to shoot more than ISO 1000-XNUMX.

        • Dima

          I didn’t receive an answer to the question) You have to compare it yourself. A large pixel that catches more light, in theory, requires less exposure time. For example, take the Nikon D700 and compare it with the D800. All other things being equal, will the shutter speed be shorter on the d700? The conversation is not about ancient SSD matrices. And relatively modern.

          Don’t pay attention to the above, the first message didn’t seem to be sent, I changed it and then two appeared. The point is clear.

          DXo pure Not a bad thing, but I didn’t understand a couple of things, how to suppress noise and increase sharpness not before opening the level, but at the final stage? Topaz Sharpen suited me much better, you can use it either independently or as a plug-in, the settings are much more flexible, even minor mistakes and blurs are corrected with a bang.

  • Yuriy

    People who want to look for pixel contrast really need pixels and a technical lens.

Add a comment

Copyright © Radojuva.com. Blog author - Photographer in Kiev Arkady Shapoval. 2009-2023

English-version of this article https://radojuva.com/en/2011/12/battle-of-megapixels/?replytocom=206088

Versión en español de este artículo https://radojuva.com/es/2011/12/battle-of-megapixels/?replytocom=206088