Personal View site logo
Cinema gear deals, direct from factories - Gear deals and Gear deals section. Also check Cameras, lenses, software, gear deals.
You support is vital for us. To keep this place ad free and independent, select one of the options below.
Donations are going to community support costs, hosting, etc. Your support allows to improve and expand this site.
Camera Resolution for Video: MTF - Modular Transfer Function
  • Hi, I currently develop a tool which should display an MTF curve in real time from monitor screen. This should be the best method to analyse perceived sharpness or resolution. I used several cameras like GH2, OMD EM 5 mark II, iPhone 6s... As soon as it comes to video I can hardly see any difference in resolution according MTF50 value. At 50% it's always 2pixels difference between black and white and hence the optimum which could be achieved.

    Something here I miss or I do wrong. The discussion about resolution compared to different cameras goes on and on but I cannot track differences down by the MTF method.

    Below you can see just the tool which I called evaluSCOPE as well as a 400% crop of the sine wave test pattern grapped by 1080p with an old FD 35mm lens at F2.0 using the Olympus OMD EM 5 mark II:

    image

    MTF Test.jpg
    1186 x 1342 - 370K
  • 4 Replies sorted by
  • I think before any video mentioning you need to research how MTF measurement is done properly.

    In photo and lenses measurements.

    Also, your MTF50 results can originate simple from fact that all this cameras have good and high resolution sensors, good resizing. So MTF50 will be always quite good.

    You can make check on old camcorders, like SD1 :-)

  • Why not video. It's low resolution compared to image capture. Assume you have a m43 sensor with 17.3 mm width and 1920 pixel than its a maximum resolution of 55.5 line pairs per mm. Bottle neck shouldn't be the lenses in most cases.

  • Yep, most time it is not lenses.

    But anyway, it needs vey good read, as every measurement tool requires full understanding that are you doing.

    http://www.imatest.com/docs/sharpness/

  • Yep, results confirm with most examples. In this case only difference that MTF is displayed linear at evaluSCOPE and logarithmic at Imatest.

    Just in this case (bar pattern: lens only) I read maximum at z=336 and minimum at z=342 which gives a resolution of 6 pixel. Depending on width of film/sensor and other crop factors you could calculate it in lp/mm.

    image

    MFT Imatest.jpg
    1163 x 957 - 176K