Canon DSLR Video Compared: 1DX Vs 5D Mark III Vs T4i

Canon DLSRs are popular for video shooting, and you can pull beautiful footage from the low end EOS D series to the high end 1D series. But what is the difference in image quality in a price range from $1000-$7000? We've had a EOS 650D, 5D Mark III, and 1DX all in ou possession, so we decided to take a closer look.

We took all three cameras and shot the same scene with the same lens (Canon 24-105 f/4) and settings (f/9, 1/50, ISO 100). We use a picture style with sharpness at 0, contrast at 0, and saturation at -2. We adjusted the focal length on the crop-sensor 650D to match that of the other two cameras. Obviously this isn't completely scientific, but we did our best.

When viewed on the web at less than full-screen, the differences between the three are very faint. But when you blow 'er up a bit, the first thing that stands out is the terrible moire patterns in the T4i shot. Look at the false colour in the bricks. Clearly the 5D Mark III and the 1DX are worlds better at resolving that type of detail.

As for image sharpness, there difference is ever so slight, but the 1DX is a bit better than the 5D Mark III, and the 5D Mark III is a bit better than the 650D. You can see it best in the leaves of the trees.

For the last step we added a bit of sharpness in post (Sharpen effect in Premiere Pro CS6, value of 55) to see how the footage responded. The difference immediately became clearer at 350% crop. The 5D Mark III lost a lot of detail and displayed more noise than the 1DX, while the 650D became sludgy and gross. Ew.

But AGAIN, this is pixel-peeping! Yes, the T4i's low price is reflected in that terrible moire, but the difference in video quality on those higher end cameras is minuscule. Paying $7,599.95 for a 1DX will get you a tad bit more sharpness, and probably a bit better performance at high ISOs (though we have not tested this yet). The 1DX's real benefits are almost entirely in its beastly photo-snapping capabilities.


Comments

    IMHO compression is the real enemy of video quality once you get into 5D mk2/3 territory. Take a still photo of the scene that you video, resize it to FHD resolution and it will look miles better than a framegrab. I know that you always have to balance the competing factos of price, battery life, storage space, processing power, recording time and quality, so here's hoping the balance tips further towards quality soon :--]

      Not entirely correct. Look at the trees in particular in the close up shots. 1DX vs 5D3. Both use the same compression yet the 1DX is significantly better. Canon are clearly cooking the 5D3 output on purpose.

      Also the D800 should by all accounts have worse video because of the line skipping required due to the higher MP sensor, yet it takes better footage (clarity/sharpness) than the 5D3.

    A true test is between the EOS DSLRCamera range and the EOS-C (Cinema series).

    Video is more of a bonus feature to these cameras.

    Then there's that old issue of sensor size as well. Even with a lens dialed up to f2.8 or even f4 (as per your sample lens) you can get much better Depth of Field (therefore boke) on the full frame cameras than on the smaller sensor models and that's what sealed the deal for me when I bought the 5DMk3.

      The mk3 is purpose built for taking photos. It also does video. It works but only produces reasonable video under some circumstances. The video is like towing a caravan with a Ferrari.

      Also, the magic lantern boys have cracked the firmware. U can expect them to bridge the quality gap to the purpose built video rigs by some way. Certainly 1080p 60 is possibility, higher bit rates and even possibly uncompressed video. + all the regular toys it brings

        ^ Right on the dot man, I use ML on my MK II, they have really opened up a can of (good) worms by enabling extra cool features, they have even allow HDR video.

Join the discussion!

Trending Stories Right Now