Jesse Evans
Offline
• • • •
Upload & Sell: On
|
 p.1 #7 · R5 Dynamic Range Test and Comparison | |
ketang wrote:
Hi everyone,
I'm a dual Canon and Sony shooter who has owned the 5D I, III, and IV and Sony A7R II and III. About 2 years ago I switched over completely to Sony while waiting for a Canon full-frame mirrorless camera that offered what the 5D series did without the DSLR drawbacks and with the R5 I have taken the plunge. I made a comparison below between the Canon and the Sony A7R III as part of a process to figure out how to ETTR with the R5. On the Sony I can shoot in HLG and I find that it is nearly perfect in showing me where the highlights will blow while still opening up the shadows so that I can see what the final image could look like. For the R5 I first compared the lower contrast JPEG picture profiles with the HDR PQ setting and am finding that the latter is more helpful for RAW ETTR, but more on that later.
For this comparison, I took bracketed shots on each camera controlling for a set of variables and chose the maximum exposures where I could still rescue all the highlights. Some details:
Shot in RAW (Sony with compressed RAW because that's how I shoot)
FE and RF 24-105mm at ISO 100, 29-31mm, f/5.6 to match framing to the extent possible
Sturdy tripod, remote shutter, live view magnified manual focus on same spot
Files processed in ACR. Canon files first converted to DNG with Adobe DNG Converter (ACR support not yet available).
+100 Shadows, -100 Highlights, NR off
You will see I have two Sony versions and one Canon. There is one Sony image and one Canon marked "Exposed for Highlights" that resulted in 1/5s for the Sony and 1/3s for the Canon. This is how I would shoot in real life. The differences in shutter speed could be the result of ISO 100 having differing actual sensitivity, the T-stop of the lenses, etc. I also chose a Sony image shot at the same shutter speed as the Canon and marked this as "Exposure Matched to Canon." I then pushed all the files a bit to better show the midtones and deep shadows. For the "Exposed for Highlights" versions I pushed an identical +0.25 stops. For the "Exposure Matched to Canon" file from Sony I had to push +0.45 stops to match the brightness (on this image the highlights are clipped). In other words, there is a .2 EV difference in push between the two approaches. Below is a small preview of the canvas where I have placed 100% crops from around the frame, followed by a link to the full-size image:
https://photos.smugmug.com/photos/i-mwJ75F8/0/002c7457/X4/i-mwJ75F8-X4.jpg
Full-Size (26MP, 4829x5398)
Some observations:
For ETTR, I am finding that the best approach with the Canon is to set the exposure where the VF and LCD show the highlights at their max before clipping, then increasing the exposure by 2/3 of a stop
When exposure is matched for the highlights, the Sony shows more noise in the shadows.
The Sony shows more detail in the shadow areas, both the "mid" shadows and darkest shadows. This is probably due to some combination of the lack of an AA filter, less baked-in NR, and lens differences (I think the Canon lens is softer). The second row of images shows this difference in crispness the most.
The third row of images shows the deepest shadows. My eyes see that while the Canon shadows look brighter, they look like there has been some NR applied and both macro and micro contrast look lower.
PhotonstoPhotos has posted the DR results from the R5 and confirmed that NR is being applied to RAW images from the base ISO up to 640, which may be a first for Canon in a long time. I usually prefer to handle NR myself and from that perspective there is an advantage to the Sony approach. However, I have to say that the clean look of the Canon has its benefits as the Sony would require more chroma NR and if you aren't careful that can throw off the colors across the image. The Canon may be more forgiving for maintaining the right colors in the shadows at the expense of some detail, and I find that valuable. All in all, I want to stress again that the usable dynamic range from each camera is highly useful and the bigger differences are with ergonomics and handling, AF tracking, IBIS, features and exposure tools, lens options, etc. Each camera has its own advantages across these categories but both are highly capable.
Hope you found this helpful....Show more →
Is it possible to share a drive or dropbox link with the raw files you used for some additional analysis?
|
