Home  |  News  |  Reviews  | About Search :  HardWare.fr 



  Processors

  Motherboards

  Graphics Cards

  Multimedia

  Storage

  Imaging

  Monitors

  Miscellaneous
Advertise on BeHardware.com
Review index:
Fujifilm FinePix F710, 3 or 6 megapixels ?
by Vincent Alzieu
Published on September 22, 2004

Introduction
The FinePix F710 has three million photosites but 6 million sensors. Half of them are used to build a basic picture. So is it a 3 or 6 megapixel?


To answer this question we tested this camera in two modes- with and without interpolation- before comparing the results with the following standard digital cameras:

  • Kodak DX7630, 6 megapixel,
  • Sony DSC-W1, 5 megapixel,
  • Nikon Coolpix 4100, 4 megapixel (one of the winners of our last comparative test),
  • Canon PowerShot S1-IS.



    Super CCD versus CCD and CMOS

    Sensor resolution is expressed in megapixels, meaning the number of megapixels captured. Each pixel is represented by three photodiodes, each photodiode is contained in a photosite. A photodiode works in one color, having a red, green or blue filter like LCD screen pixels.

    Photodiodes come in the form of a single light sensitive silicon part. CCDs then transfer captured information by moving the electrical charge from one photosite to another. CMOS sensors, or more precisely their photosites, convert light information directly into electrical tension and send it to the processor. To be more accurate, there are actually more than three photodiodes per pixel. The human eye is more sensitive to green, so to have a more accurate restitution of color gradations, two green photosites per pixel are required. This technique makes it possible to measure green luminance and the chrominance of the two other colors. So we are actually speaking of “quadruplets” rather than “triplets” for pixels.

    The photosite problem

    If we go back to the above sentence, “Each pixel is represented by three photodiodes, and each photodiode is contained in a photosite”, and we add the fact that there is twice as many green photosites than red and blue, you may notice the inconsistency. It means that unlike LCD panels, each pixel isn’t physically made of three colored sub pixels, but is interpolated from one photosite by the processor.
    If we take a 4 megapixels sensor this means 4 million photosites of 1 million red + 1 blue + 2 million green. If we look at a LCD screen, it means that each pixel is made of 3 photosites (1red, one green and one blue). We need to add one green photosite (this is the Bayer filter principle) to improve color accuracy. So finally a 4-magapixels sensor only captures pictures with a physical size of one megapixel only!
    This is the reason why manufacturers insist on the fact that the number of megapixel isn’t the only component of a picture. There is also the lens and above all the sensor. This processor is in charge of obtaining a 4 megapixel photo from 4 million photosites. Its main job is first to transform a red photosite, a base for a dot, into pixel and add an average of the adjacent green and blue pixels. The sum of colors will determine the color of the final dot.

    We get back to the sensor

    CCD sensors are usually used in compact cameras and reflexes other than the Canon. Quality is from good to excellent, depending mainly on the physical size of the photosites. The smaller the sensor the less light the photosite receives and the less sensitive the camera. Logically, the camera will have more noise in the image.
    The quality of CMOS sensors ranges from excellent to awful. The worst comes from low end models where pictures have a lot of noise and color quality is unpredictable. The bad reputation of the CMOS sensor comes from these types of cameras. On the other hand there are the Canon numerical reflexes (and recently Nikon) which, of course, aren’t affected by these problems. The difference between CMOS and CCD sensors is a softer picture.


    Then there is the Super CCD, invented by FujiFilm and used exclusively by this company since 2000. Each photosite includes not one but two silicon parts. The bigger silicon part captures high lights and the smaller one low lights. Here we have the choice. If we choose a resolution in function of the number of photosites (3 megapixels for the F710) only the “high light” silicones are used. If we double the resolution (6 megapixels mode for the FP710), the processor interpolates the picture from 3 to 6 megapixels by using information sent by the “low light” silicon photodiode.

    FujiFilm introduced their camera as a 6-megapixel, but the fact that this camera interpolates the picture from 3 to 6 million dots would put this camera in the 3-megapixel category. The easiest way to establish which category this camera belongs to is to compare the two modes to each other and then compare the better mode with standard 3 to 6 megapixel cameras.


  • Page index
    1 | 2 | 3 | 4 | 5 | 6 | 7 | 8
    Next page >>
    Ergonomic aspects and the F710’s speed  




    Copyright © 1997- Hardware.fr SARL. All rights reserved.
    Read our privacy guidelines.