Extenders and high iso with the R5

AlanF

Desperately seeking birds
Canon Rumors Premium
Aug 16, 2012
13,130
25,154
Concerns are frequently expressed about lenses like the RF 100-500mm with f/7.1 being narrow and so the images must surely be noisy because high isos are necessary in dim light? Some claim the 400mm f/5.6 is superior in low light because it is 2/3 of a stop faster (or a 1/3rd faster than the 100-500mm at 400mm f/6.3). I've pointed out that the 100-500mm has the same size front element of the 400mm f/5.6 that although you need to crank up up the iso by 2/3 of a stop to maintain the shutter speed, the extra magnification of the longer lens compensates for any extra noise. Others are genuinely worried that a 2x extender needs an extra 2 stops of iso at f/14, and I've said don't worry as the R5 handles noise well and again the 2-fold gain in magnification compensates for 2 stops higher iso. To show this, I have today taken a series of images in cloudy overcast conditions at 20m distance of a magazine cover with the 100-500mm on the R5 from 1/100s to 1/6400s, increasing the iso from 160 to 10k, and in parallel with the RF 2x at the same shutter speeds and isos from 640 to 40k. The RAW images were processed using my standard procedure with DxO Photolab PL4. The noise reduction was with its DeepPRIME and luminosity at its standard 40 on the slider - this is mild and loses minimal detail. Lens sharpening was on but I decreased the global to -2 as the standard 0 is a little aggressive.
Here are the images in pairs - at 500mm and 1000mm with 2 extra stops of iso. The crops are 100% - 1 pixel of crop = 1 pixel of original. Draw your own conclusions!
 
  • Like
Reactions: 4 users
Concerns are frequently expressed about lenses like the RF 100-500mm with f/7.1 being narrow and so the images must surely be noisy because high isos are necessary in dim light? Some claim the 400mm f/5.6 is superior in low light because it is 2/3 of a stop faster (or a 1/3rd faster than the 100-500mm at 400mm f/6.3). I've pointed out that the 100-500mm has the same size front element of the 400mm f/5.6 that although you need to crank up up the iso by 2/3 of a stop to maintain the shutter speed, the extra magnification of the longer lens compensates for any extra noise. Others are genuinely worried that a 2x extender needs an extra 2 stops of iso at f/14, and I've said don't worry as the R5 handles noise well and again the 2-fold gain in magnification compensates for 2 stops higher iso. To show this, I have today taken a series of images in cloudy overcast conditions at 20m distance of a magazine cover with the 100-500mm on the R5 from 1/100s to 1/6400s, increasing the iso from 160 to 10k, and in parallel with the RF 2x at the same shutter speeds and isos from 640 to 40k. The RAW images were processed using my standard procedure with DxO Photolab PL4. The noise reduction was with its DeepPRIME and luminosity at its standard 40 on the slider - this is mild and loses minimal detail. Lens sharpening was on but I decreased the global to -2 as the standard 0 is a little aggressive.
Here are the images in pairs - at 500mm and 1000mm with 2 extra stops of iso. The crops are 100% - 1 pixel of crop = 1 pixel of original. Draw your own conclusions!
Thanks for doing this. Just ordered the 2x extender!
 
  • Like
Reactions: 1 user
Upvote 0
IF (and that is a big if) my on order 100-500 ever arrives, I'll sell my EF 1.4x (and EF 100-400 because I can't afford overlapping both zooms in my stable) but keep my EF 2x for the 300mm f/2.8 ii. I'll live with the 100-500 for a while before deciding on a TC not because of IQ (I really appreciate all the work Alan and others have done on testing the combos) but because of the limitations in the zoom range, length of the RF zoom with the TC attached in my bag and the hassle of placing and removing the TC on the RF lenses. I think I understand why the TC extends into the zoom lens limiting the range and extending the length of the combo with the TC attached. While not a big deal generally, even lining up the lens cap is just a bit more awkward than on the EF. Then, when I decide to move forward with the RF 1.4x, maybe there will be some used or refurbs out there.

So thanks to all of you who have posted the images and figures showing that the 100-500 with a TC is really an option. The collective knowledge on these forums is great.
 
Upvote 0
Concerns are frequently expressed about lenses like the RF 100-500mm with f/7.1 being narrow and so the images must surely be noisy because high isos are necessary in dim light? Some claim the 400mm f/5.6 is superior in low light because it is 2/3 of a stop faster (or a 1/3rd faster than the 100-500mm at 400mm f/6.3). I've pointed out that the 100-500mm has the same size front element of the 400mm f/5.6 that although you need to crank up up the iso by 2/3 of a stop to maintain the shutter speed, the extra magnification of the longer lens compensates for any extra noise. Others are genuinely worried that a 2x extender needs an extra 2 stops of iso at f/14, and I've said don't worry as the R5 handles noise well and again the 2-fold gain in magnification compensates for 2 stops higher iso. To show this, I have today taken a series of images in cloudy overcast conditions at 20m distance of a magazine cover with the 100-500mm on the R5 from 1/100s to 1/6400s, increasing the iso from 160 to 10k, and in parallel with the RF 2x at the same shutter speeds and isos from 640 to 40k. The RAW images were processed using my standard procedure with DxO Photolab PL4. The noise reduction was with its DeepPRIME and luminosity at its standard 40 on the slider - this is mild and loses minimal detail. Lens sharpening was on but I decreased the global to -2 as the standard 0 is a little aggressive.
Here are the images in pairs - at 500mm and 1000mm with 2 extra stops of iso. The crops are 100% - 1 pixel of crop = 1 pixel of original. Draw your own conclusions!
Thank you so much for yet another well researched and informative post (with examples) to help educate the forum about the burning questions we have for the 2x. Yet again assumptions about the 100-500 being too dark to be good fall apart in real world testing.
Much appreciated as always!
 
Upvote 0
Concerns are frequently expressed about lenses like the RF 100-500mm with f/7.1 being narrow and so the images must surely be noisy because high isos are necessary in dim light? Some claim the 400mm f/5.6 is superior in low light because it is 2/3 of a stop faster (or a 1/3rd faster than the 100-500mm at 400mm f/6.3). I've pointed out that the 100-500mm has the same size front element of the 400mm f/5.6 that although you need to crank up up the iso by 2/3 of a stop to maintain the shutter speed, the extra magnification of the longer lens compensates for any extra noise. Others are genuinely worried that a 2x extender needs an extra 2 stops of iso at f/14, and I've said don't worry as the R5 handles noise well and again the 2-fold gain in magnification compensates for 2 stops higher iso. To show this, I have today taken a series of images in cloudy overcast conditions at 20m distance of a magazine cover with the 100-500mm on the R5 from 1/100s to 1/6400s, increasing the iso from 160 to 10k, and in parallel with the RF 2x at the same shutter speeds and isos from 640 to 40k. The RAW images were processed using my standard procedure with DxO Photolab PL4. The noise reduction was with its DeepPRIME and luminosity at its standard 40 on the slider - this is mild and loses minimal detail. Lens sharpening was on but I decreased the global to -2 as the standard 0 is a little aggressive.
Here are the images in pairs - at 500mm and 1000mm with 2 extra stops of iso. The crops are 100% - 1 pixel of crop = 1 pixel of original. Draw your own conclusions!
This is excellent, thank you. I was on the fence about the 100-500 and extenders, now no more.
 
  • Like
Reactions: 1 user
Upvote 0
This is excellent, thank you. I was on the fence about the 100-500 and extenders, now no more.
Software that handles noise well is important. My preference is for DxO PL4 as you can use their DeepPrime or Prime noise reduction at the press of a click, and do a remarkable job without requiring any skill, retaining maximum detail.
 
  • Like
Reactions: 1 users
Upvote 0
Topaz Denoise can also do a decent job, but you have to use the new 4-way preview to compare the different algorithms, the 'winner' varies from picture to picture. And in a few cases I had to mask out damselfly eyes to preserve the detail. So it's not as easy as DxO PL4 seems to be, judging from AlanF's description.
 
  • Like
Reactions: 1 user
Upvote 0