Peter Loughran talkes about CROP FACTORS

Categories: GH5 camera 4 Comments


One of my readers Peter Loughran has been sending me some very interesting facts about crop factors…

Regarding the question of crop factors on the GH4.
Having used and tested these things on the the GH3 for over a year (and currently waiting for my GH4 Body to ship!) I’ve made some observations and calculations which may (if they prove correct when we get the actual camera) offer some clarification. Sorry this is a bit dry but there’s a lot of misinformation and speculation around so I thought I’d best justify my thoughts by showing the calculations used (please feel free to offer corrections if I’ve got anything wrong).
Because the GH3 (unlike the GH2 which has an over-sized constant crop factor ‘Multi Format’ sensor) actually crops its 4:3 format sensor (2x crop factor applies ONLY @ 4:3) down vertically to give 3:2 & 16:9 formats, the crop factor at 16:9 is actually ~2.18 (based on 43.27mm diagonal of 35mm still frame format of 36x24mm).

Full Frame Film/DSLR format = 36x24mm = 43.27mm diagonal (image circle diameter)

GH3 sensor (@4:3 full sensor) = 17.3×13.0mm = 21.64mm diagonal
Crop Factor = 43.27/21.64 = 2

GH3 sensor (@3:2 crop) = 17.3×11.53mm = 20.79mm diagonal
Crop Factor = 43.27/20.79 = 2.08

GH3 sensor (@16:9 crop) = 17.3×9.73mm = 19.85mm diagonal
Crop Factor = 43.27/19.85 = 2.18
(GH3 video EXIF data shows this scale factor as rounded to 2.2 but shows 35mm equivalent focal lengths correctly as multiplied by 2.18)

GH3 sensor when using the ‘Extra Tele Conversion’ 100% pixel for pixel 1920×1080 16:9 video crop mode = 7.21×4.06mm = 8.28mm diagonal
Crop Factor = 43.27/8.28 = ~5.22
(GH3 video EXIF data shows this scale factor as rounded to 5.2 but shows 35mm equivalent focal lengths correctly as multiplied by 5.22)

The GH4s new sensor appears to be same size & resolution as that in the GH3 so crop factors seem to be the same when shooting at Full HD.
Panasonics online specs. sheet for the GH4 shows exactly the same range of magnifications when using the ‘Extra Tele Conversion’ mode as those listed for the GH3:
“Extra Tele Conversion: Motion picture: 2.4x (FHD), 3.6x (HD), 4.8x (VGA)”
(Remember those magnifications are in addition to the crop factor of the sensor and selected format).
No official crop factor is so far listed, as far as I can see, for the C4K & QFHD video modes so whether the crop factor is different from that in Full HD (where it appears to use the full sensor width) remains to be confirmed.

If the GH4 is using a 100% pixel for pixel crop to obtain the 4K & QFHD images from the sensor the crop factors would be something like this:

Image sensor size 17.3 x 13.0 mm (in 4:3 aspect ratio)
Full horizontal sensor resolution is 4608 pixels over 17.3mm
C4K = 4096×2160 would imply a sensor crop of ~15.38×8.14mm = 17.4mm diagonal = Crop factor of ~2.5
4K = 3840×2160 would imply a sensor crop of ~14.42×8.11mm = 16.54mm diagonal = Crop Factor of ~2.6

Some ‘sources’ are claiming that the GH4s 4K crop factor is ~2.3 but seem to be working on the false assumption that the 16:9 (full sensor width) crop is 2x (which ONLY applies at the full 4:3 sensor aspect ratio) whereas the actual calculated 16:9 (full sensor width) crop factor is 2.18 as shown above.
Add their approximately 0.3 extra crop factor to my calculated (and EXIF confirmed) figure of 2.18 and you come to a very similar figures for the possible (100% pixel for pixel) crop factors (as calculated above) of ~2.5 to 2.6.
If that proves to be true I anticipate a lot of wailing and gnashing of teeth from those folks who long for bigger and bigger sensors.
Personally, I’d be delighted if the crop factor is a touch higher than in 1080p mode as it would give my longer lenses even more ‘pulling power’ in 4K (when shooting nature with telephoto and macro lenses) whilst still having the possible equivalent of ~18mm wide-angle coverage (more than enough for most purposes) by use of the Panasonic or Olympus 7-14mm lenses.

For all your video production needs in Scotland, get in touch with Small Video Company Ltd

4 comments on this post

  1. Peter Loughran says:

    Minor correction to my figures:
    Of course both 2160 pixel height crops in C4K and 4K should be identical at ~8.13mm (only the image width varies).
    Makes no real difference to the calculated (approximate) crop factors but here’s how they should be:

    C4K = 4096×2160 would imply a sensor crop of ~15.38×8.13mm = 17.4mm diagonal = Crop factor of ~2.5

    4K = 3840×2160 would imply a sensor crop of ~14.42×8.13mm = 16.55mm diagonal = Crop Factor of ~2.6

  2. Bo-Ming Tong says:

    I believe it is unfair to calculate the ‘crop factor’ using the diagonal this way because it would give full frame an unjustified advantage. The vertical cropping you referred to applies to a full frame camera such as a 5D mark III as well. A fair comparison would crop the full frame to 16:9 too, to 36mm x 20.25mm. It doesn’t make sense to compare different aspect ratios but if we keep the aspect ratio fixed it doesn’t matter whether we calculate the ratio of the horizontal or the diagonal.

    • Peter Loughran says:

      Sensor Envy?
      “unfair” …??? “give full frame an unjustified advantage” …???
      An advantage in what way? …it seems you are assuming that ‘bigger is always better’ (perhaps partly true in terms of low light performance at a given aperture & shutter speed) or that the (now frequently overdone …yawn) ‘shallow depth of field look’ (with its associated requirement for extremely critical focusing …even more so in 4K which will show up the tiniest error) is always a desirable side effect of bigger sensors. In situations where you have full control of lighting, focus and the ‘action’ being recorded (“Cut! …Take Two!” etc.) then such things may, or may not, be desirable but I would suggest that there are many situations where a bigger sensor may be almost unusable …try shooting fast moving news, sports or wildlife (with no chance of a retake!) using a big sensor camera and see just how many acceptable (actually in focus) shots you get.
      So, sometimes bigger is better, sometimes smaller is better …horses for courses!
      Thus crop factor figures do not imply any advantage in either direction but are merely a way of enabling us to predict, approximately, what field of view a particular focal length lens might give us on a particular camera sensor, and how that may compare to another focal length on another sensor.
      Whether you like it or not; the fact is that crop factors are, and always have always been, calculated by comparing the minimum size of the image circle required to cover a given frame size …that minimum size is equal to the diagonal of the frame regardless of the aspect ratio.
      The 3:2 aspect ratio ‘Full Frame 35mm’ format (Note: still image size, not ’35mm’ or ‘Super 35mm’ movie sizes which are different again) size of 36x24mm has been generally adopted as a known standard for comparative purposes simply because of its familiarity from decades of use by generations of photographers.
      Try examining the EXIF data from almost any make/size of camera image (preferably still image data as EXIF data for some video formats doesn’t include all the information available in still images from the same camera) and you will usually find a ‘Scale Factor To 35 mm Equivalent’ figure quoted and also a ‘Focal Length’ reading with the 35 mm equivalent figure in brackets.
      Of course it would be simpler, and easier to make comparisons, if all cameras had exactly the same aspect ratios so that the angle of coverage of the horizontal, vertical or diagonal axis could form the basis for comparison. Back in the real world, there are many different cameras with many different sized sensors shooting at many different aspect ratios …sometimes with many variations within one camera.
      For example, my new GH4 is capable of shooting stills at 4:3, 3:2, 16:9 and 1:1 aspect ratios.
      It can also shoot video at 4:3 (VGA, at two crop sizes), 16:9 (HD/FHD/4K, at a total of four different crop sizes) and ~17:9 (C4K, at one crop size) …so a total of seven possible video crop factor variations from one camera/sensor.
      The Canon 1DC shoots 1080p using the full frame sensor (vertically cropped from 3:2 to 16:9 as you say), but in 4k recording mode it’s cropped smaller to the APS-H size. You also have the option to crop the frame to a Super 35mm crop factor so it can be compatible with the newest Canon EF Cinema Zoom Lenses which are made to support that frame factor (Which is the Frame Crop on the Sony F3, Red Epic, and Alexa High End Production Cameras …I somehow doubt the owners/users of those cameras feel threatened by the ‘unfair advantage’ of the bigger sensor of the Canon 5D mark III).
      The Nikon D800/e shoots 1080p 16:9 video using a ‘Full Frame Based’ sensor readout size which, is actually smaller than the full sensor width, and also from a smaller ‘DX Based’ sensor area.
      I have shot stills with film sizes of ’10×8 inches’, ‘5×4 inches’, ‘6x7cm’, ‘6x6cm’, ’35mm’ etc. (and a little 16mm cine film) plus video with cameras with sensors ranging from ‘1/6inch’, ‘1/3inch’ etc. to the relatively big (even by professional video camera standards) sensor in the Panasonic GH3 & GH4.
      With so many variations of sizes and aspect ratios available, in both stills and video, it makes perfect sense to adopt a common familiar reference point (in this case the very familiar 36x24mm 35mm Full Frame format) on which to base the (yes, approximate) information needed to compare different camera/lens systems.

      So, relax …as the saying goes; ‘It’s not the size, it’s what you do with it that counts!’

      • Peter Loughran says:

        P.S. if you’d bothered to do the calculation of 16:9 crop factor as compared to the Canon 5d Mark 3s Full Frame 35mm 3:2 (36x24mm) (or any other Full Frame 35mm sensor) versus the Cropped Frame 35mm 16:9 (36×20.25mm), as you suggest would be ‘fairer’, then you would find the difference is less than 5% …or, to put it another way, …insignificant!
        That is unless you have the ability to visualise the difference between, say, a 50mm vs ~52mm lens (on the GH3/GH4) or a 100mm vs ~105mm lens on the Canon 5D …?
        That’s because the 35mm 3:2 (or 15:10) is not so different from the 16:9 format …obviously much closer in shape than is 4:3 (or 16:12).

        35mm Full Frame 3:2 vs Cropped Frame 16:9 compared to GH3/GH4 Cropped Frame 16:9
        (when using full sensor widths):

        35mm Full Frame 3:2 = 36mm x 24mm = diagonal of 43.27mm

        35mm Cropped Frame 16:9 = 36mm x 20.25mm = diagonal of 41.30mm

        That’s a difference of 1.97mm or about a 4.8% increase in the crop factor by using the full 35mm 3:2 (36x24mm) frame rather than the cropped 35mm 16:9 (36×20.25mm)

        GH3/GH4 sensor (@16:9 crop) = 17.3×9.73mm = 19.85mm diagonal

        GH3/GH4 16:9 crop factor of 2.18 when compared to full 35mm 3:2 frame
        (Crop Factor = 43.27/19.85 = 2.18)

        GH3/GH4 16:9 crop factor of 2.08 when compared to cropped 35mm 16:9 frame
        (Crop Factor = 41.30/19.85 = 2.08)

Post Comment

Please note: all comments are moderated by an Admin.