Canon EOS C50 Rolling Shutter and Dynamic Range Numbers are in
- Thread starter Canon Rumors
- Start date
Why did they only test latitude for RAW and not in CLOG 2?
This was such a lazy test.
Gerald Undone's review is so much better.
This was such a lazy test.
Gerald Undone's review is so much better.
Upvote
0
That's in open gate. You can't use that video metric against a camera that isn't filming open gate. By comparison, the Lumix S5II films in open gate with a rolling shutter speed of 25.4ms. Using more of the vertical height of the sensor is detrimental for rolling shutter performance.
Upvote
0
There is a typo in the article, it states 18.1 seconds, which should be 18.1 milliseconds (ms).
Upvote
0
Read the article and I have a few interesting takeaways after we've combined to roughly 30 hours on the C50.
- - - - - - - -
It’s interesting that they’re almost criticizing the camera’s rolling-shutter performance in open gate - while failing to mention that it’s actually the fastest hybrid camera rolling shutter they’ve tested in open gate. Just…interesting.
Dynamic range is another curious interpretation of the data. The results line up almost perfectly with the FX3’s dynamic range numbers, yet that context somehow never makes it into the discussion. It also beats the XF-AVC performance of the Nikon ZR, Sony FX2, Lumix S5II and S1II...oh, and the Sony BURANO - again, interesting that this camera isn't impressive. Perhaps because the sensor doesn't stand out as remarkable in any category?
The Nikon ZR "RED Raw" (*the only recording mode the camera will grade in RED color science) is measured to reproduce a maximum of 13.8 stops of total dynamic range, with a median of 10.2 stops, and a suspiciously clean noise floor in RAW. This strongly suggests noise processing is taking place in RAW and many users have noticed the digital splotchy noise. Meanwhile, the Canon is delivering over 15 stops of total dynamic range and “only” 9.98 stops at the median - without any noise reduction applied at all. Once noise reduction is added, you’ll easily exceed 12 or even 13 stops of clean dynamic range, because Canon isn’t baking noise reduction into its RAW footage in the first place. There is certainly noise on this sensor when recording in 7K, but it was incredibly easy to clean up 7K open gate and then downsample to 4K output and look truly spectacular. Normalizing the footage or in the oversampled 4K 24/30/60P is about as good as it gets.
Every head-to-head test I’ve seen of the Nikon ZR shows worse low-light performance than the C50. (The FX3 is still to this day, the king of low light.)
Gerald Undone ran the exact same tests using the exact same charts and recorded 15.1 stops of total DR with a 10.1 median at ISO 800, and 14.7 / 8.9 at ISO 6400 in RAW. In XF-AVC, the camera measures 15.1 total and 12.8 clean. Those numbers are widely regarded as good-to-great, and I’m perfectly happy with them. This article, however, leans toward being unimpressed - borderline disappointed - by performance that is better than or competitive with virtually every camera on the market. Apparently landing within 0.2–0.5 stops of the C400 means that camera is a world-beater while this one is somehow disappointing.
- - - - - - - -
Our time with the C50 is still early. I’ve got roughly 20 hours on mine, and a friend has 7–10 hours on his. I’ve already shot two large projects and multiple days at a trade show, primarily using open-gate 7K in Long-GOP (the lowest bitrate available in Open Gate), and the footage has been STELLAR. There is one clip in particular that we shot wide in 7K open gate at ISO 800 Long-GOP and we had to crop in beyond APS-C and still had 4K information and noise was inperceptable - even after the crop! I've only needed to shoot RAW video on one job ever and that was only because my camera at the time (the R3) didn't have CLOG2 like the rest of the cameras on the project. Some people exclusively shoot RAW and I would be more curious to see how the THREE options for RAW handle noise in a studio environment, sharpness and dynamic range. Because RAW LT is available in a very manageable bit rate. Maybe I'll mess around one day in my backyard and office when I have some downtime.
Which really just reinforces the obvious: you need to actually use a camera and form your own conclusions. Framing results a certain way can mislead readers, as it appears this article wants to do. The reality is, this camera absolutely holds its own against the C80, C400, R1, R5 II, R5C, S5II, S1II, FX2 and FX3.
- - - - - - - -
It’s interesting that they’re almost criticizing the camera’s rolling-shutter performance in open gate - while failing to mention that it’s actually the fastest hybrid camera rolling shutter they’ve tested in open gate. Just…interesting.
Dynamic range is another curious interpretation of the data. The results line up almost perfectly with the FX3’s dynamic range numbers, yet that context somehow never makes it into the discussion. It also beats the XF-AVC performance of the Nikon ZR, Sony FX2, Lumix S5II and S1II...oh, and the Sony BURANO - again, interesting that this camera isn't impressive. Perhaps because the sensor doesn't stand out as remarkable in any category?
The Nikon ZR "RED Raw" (*the only recording mode the camera will grade in RED color science) is measured to reproduce a maximum of 13.8 stops of total dynamic range, with a median of 10.2 stops, and a suspiciously clean noise floor in RAW. This strongly suggests noise processing is taking place in RAW and many users have noticed the digital splotchy noise. Meanwhile, the Canon is delivering over 15 stops of total dynamic range and “only” 9.98 stops at the median - without any noise reduction applied at all. Once noise reduction is added, you’ll easily exceed 12 or even 13 stops of clean dynamic range, because Canon isn’t baking noise reduction into its RAW footage in the first place. There is certainly noise on this sensor when recording in 7K, but it was incredibly easy to clean up 7K open gate and then downsample to 4K output and look truly spectacular. Normalizing the footage or in the oversampled 4K 24/30/60P is about as good as it gets.
Every head-to-head test I’ve seen of the Nikon ZR shows worse low-light performance than the C50. (The FX3 is still to this day, the king of low light.)
Gerald Undone ran the exact same tests using the exact same charts and recorded 15.1 stops of total DR with a 10.1 median at ISO 800, and 14.7 / 8.9 at ISO 6400 in RAW. In XF-AVC, the camera measures 15.1 total and 12.8 clean. Those numbers are widely regarded as good-to-great, and I’m perfectly happy with them. This article, however, leans toward being unimpressed - borderline disappointed - by performance that is better than or competitive with virtually every camera on the market. Apparently landing within 0.2–0.5 stops of the C400 means that camera is a world-beater while this one is somehow disappointing.
- - - - - - - -
Our time with the C50 is still early. I’ve got roughly 20 hours on mine, and a friend has 7–10 hours on his. I’ve already shot two large projects and multiple days at a trade show, primarily using open-gate 7K in Long-GOP (the lowest bitrate available in Open Gate), and the footage has been STELLAR. There is one clip in particular that we shot wide in 7K open gate at ISO 800 Long-GOP and we had to crop in beyond APS-C and still had 4K information and noise was inperceptable - even after the crop! I've only needed to shoot RAW video on one job ever and that was only because my camera at the time (the R3) didn't have CLOG2 like the rest of the cameras on the project. Some people exclusively shoot RAW and I would be more curious to see how the THREE options for RAW handle noise in a studio environment, sharpness and dynamic range. Because RAW LT is available in a very manageable bit rate. Maybe I'll mess around one day in my backyard and office when I have some downtime.
Which really just reinforces the obvious: you need to actually use a camera and form your own conclusions. Framing results a certain way can mislead readers, as it appears this article wants to do. The reality is, this camera absolutely holds its own against the C80, C400, R1, R5 II, R5C, S5II, S1II, FX2 and FX3.
Upvote
0
And I read your comment and just assumed you mean milliseconds and didn't even notice the typo. Nice catch!
Upvote
0
Gerald Undone is a YouTuber with zero qualifications to perform these test... But sure, take his test results and ignore CineD.
Upvote
0