Home · Register · Search · View Winners · Software · Hosting · Software · Join Upload & Sell

Moderated by: Fred Miranda
Username   Password

  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Canon Forum | Join Upload & Sell

1      
2
       3              5       6       end
  

Archive 2012 · 5D vs 5DM3 High ISO
  
 
skibum5
Offline
• • • • • •
Upload & Sell: Off
p.2 #1 · p.2 #1 · 5D vs 5DM3 High ISO


snapsy wrote:
Here is a demonstration of impulse noise on the 5D that is likely being NR by the 5DM3's DSP/firmware, like how the D3s does. Since it's in the digital domain the same should be achievable using software on the 5D's raw:

5D ISO 6400 (1600 pushed 2 EV) vs the 5DM3 ISO 12800 (1 EV delta), 100% crop


I think some RAW converters, such as ACR, largely remove all the salt and pepper stuff automatically from all cameras.



Dec 29, 2012 at 11:05 PM
snapsy
Offline
• • • •
Upload & Sell: On
p.2 #2 · p.2 #2 · 5D vs 5DM3 High ISO


skibum5 wrote:
I think some RAW converters, such as ACR, largely remove all the salt and pepper stuff automatically from all cameras.


LR uses the ACR engine and it does not remove the noise automatically, at least on the 5D as demonstrated.



Dec 29, 2012 at 11:07 PM
corndog
Offline
• • • •
Upload & Sell: On
p.2 #3 · p.2 #3 · 5D vs 5DM3 High ISO


If you really want to compare an aspect of two cameras, all other aspects should be made as close as possible, imo.


Dec 29, 2012 at 11:08 PM
Jeff Nolten
Offline
• • •
Upload & Sell: On
p.2 #4 · p.2 #4 · 5D vs 5DM3 High ISO


Hmmm, just looking at the 1 stop and 2 stop images the noise looks most similar at 2 stops. But what surprised me is the 5Dc looks sharper in both images, look at the laces and the bear's foot pad. This seems strange considering the downsampling of the 5D3 image. Is this an artifact of the image reduction?

Edited on Dec 29, 2012 at 11:10 PM · View previous versions



Dec 29, 2012 at 11:09 PM
skibum5
Offline
• • • • • •
Upload & Sell: Off
p.2 #5 · p.2 #5 · 5D vs 5DM3 High ISO


snapsy wrote:
LR uses the ACR engine and it does not remove the noise automatically, at least on the 5D as demonstrated.


All I know is I've loaded RAWs with tons of salt and pepper into ACR and it doesn't show up much and yet in some other RAW converters it is still there in spades. (and your sample says DPP)
Even your 5D sample there doesn't show as much salt n pepper as I'd expect to actually be there without any software converter filtering.




Dec 29, 2012 at 11:09 PM
snapsy
Offline
• • • •
Upload & Sell: On
p.2 #6 · p.2 #6 · 5D vs 5DM3 High ISO


Jeff Nolten wrote:
Hmmm, just looking at the 1 stop and 2 stop images the noise looks most similar at 2 stops. But what surprised me is the 5Dc looks sharper in both images, look at the laces and the bear's foot pad. This seems strange considering the downsampling of the 5D3 image. Is this an artifact of reduction?


To my eyes the difference is about 1 1/3 stops. Sharpness for the 8MP looks about the same to me. Btw, for the 5D images I took 3 images and kept the sharpest (PDAF), whereas on the 5DM3 I manually focused once in LV. All shots were at f/8



Dec 29, 2012 at 11:12 PM
snapsy
Offline
• • • •
Upload & Sell: On
p.2 #7 · p.2 #7 · 5D vs 5DM3 High ISO


skibum5 wrote:
All I know is I've loaded RAWs with tons of salt and pepper into ACR and it doesn't show up much and yet in some other RAW converters it is still there in spades. (and your sample says DPP)
Even your 5D sample there doesn't show as much salt n pepper as I'd expect to actually be there without any software converter filtering.

I only use DPP to capture the screen comp (I like it's UI better). The images were processed in LR4 and exported as 16-bit ProRGB TIFFs.



Dec 29, 2012 at 11:13 PM
Sven Jeppesen
Offline
• • • •
Upload & Sell: Off
p.2 #8 · p.2 #8 · 5D vs 5DM3 High ISO


Remember that those small pics are both downsized and PP in LR4


Dec 29, 2012 at 11:18 PM
snapsy
Offline
• • • •
Upload & Sell: On
p.2 #9 · p.2 #9 · 5D vs 5DM3 High ISO


Here is the 5D upsampled to 22MP, compared to the 5DM3 at native 22MP. It would probably fairer to the 5D to upsample both to some arbitrary resolution so that both images undergo some level of artifacting from the process.

5D ISO 1600 vs 5DM3 ISO 3200 (1-stop delta), 22MP, 100% Crop
5D ISO 1600 Full Image, upsampled to 22MP
5DM3 ISO 3200 Full Image, native 22MP



Dec 29, 2012 at 11:29 PM
mirkoc
Offline
• • •
Upload & Sell: Off
p.2 #10 · p.2 #10 · 5D vs 5DM3 High ISO


Thank you for the comparison. It would be nice to see the same comparison but with 1.5 stops underexposed and then in pp recovered pictures. Or upsampled 5D at ISO 100 comparison.


Dec 30, 2012 at 12:37 AM
 

Search in Used Dept. 



Jeff Nolten
Offline
• • •
Upload & Sell: On
p.2 #11 · p.2 #11 · 5D vs 5DM3 High ISO


Yes, thanks for posting this test Snapsy.


Dec 30, 2012 at 12:44 AM
PhilDrinkwater
Offline
• • •
Upload & Sell: Off
p.2 #12 · p.2 #12 · 5D vs 5DM3 High ISO


Like for like at 1600-3200, the difference is likely 1-2 stops. I've not heard many people quote more.

However the 5d3 will shoot much higher than 3200 and get a usable result.



Dec 30, 2012 at 01:26 AM
thedutt
Offline
• • • •
Upload & Sell: On
p.2 #13 · p.2 #13 · 5D vs 5DM3 High ISO


Thanks for posting the shots. Ive had all 5 series cameras at some point and observation is that a direct comparison is fairly complex to perform. The comparison that is the biggest difference in my book is the one that rrxjon posted about. 5DIII provides useable photos are high ISO which the previous 5Ds did not. This is both due to its ability to focus under those circumstance as well as sensor / jpeg procession. The numbers can be made to sing in a number of ways by changing the parameters around, e.g, a ISO3200 5D classic shot in day light is very useable, but the same ISO3200 shot in low dark light with f2 @ 1/50s may not be.




Dec 30, 2012 at 04:46 AM
Ghost
Offline
• • • •
Upload & Sell: Off
p.2 #14 · p.2 #14 · 5D vs 5DM3 High ISO


I can also argue that my 5dc can nail focus just of a lamp post as well as a 5d3 or 1dx.

5d3 > 5d2 > 5d.

Look I love my 5dc as much as any 5dc owner would. But fact of the matter is the 5d3 is leagues ahead of the 5dc under non-ideal conditions.



Dec 30, 2012 at 06:37 AM
Lasse Eriksson
Offline
• • • •
Upload & Sell: Off
p.2 #15 · p.2 #15 · 5D vs 5DM3 High ISO


Ghost wrote:
I can also argue that my 5dc can nail focus just of a lamp post as well as a 5d3 or 1dx.

5d3 > 5d2 > 5d.

Look I love my 5dc as much as any 5dc owner would. But fact of the matter is the 5d3 is leagues ahead of the 5dc under non-ideal conditions.


+1

If shooting high ISO in real life, then you do it because you have very little light. The dark background would make the 5D banding and noise look a lot worse. Here with good light, small aperture and white background it's a different thing.
And who is claiming there is a 3 stop difference like you write?



Dec 30, 2012 at 07:05 AM
snapsy
Offline
• • • •
Upload & Sell: On
p.2 #16 · p.2 #16 · 5D vs 5DM3 High ISO


Lasse Eriksson wrote:
If shooting high ISO in real life, then you do it because you have very little light. The dark background would make the 5D banding and noise look a lot worse. Here with good light, small aperture and white background it's a different thing.
And who is claiming there is a 3 stop difference like you write?


There are deep shadows in the scene I created. There was mention of 3 stops in the thread at http://www.fredmiranda.com/forum/topic/1176816.



Dec 30, 2012 at 12:37 PM
justruss
Offline
• • • •
Upload & Sell: Off
p.2 #17 · p.2 #17 · 5D vs 5DM3 High ISO


Not sure what the argument here is. The 5d3 improved on the 5d2 which improved on the 5D which improved on the 1ds (maybe even 1ds2/3). No single step was a revolution, but high ISO from 1ds to 5dmk3 sure looks revolutionary.

Other manufacturers have made bigger jumps. Just as it always has been.

It's pretty wonderful that every year for the past decade we've been able to say, "Shooting has never been so easy, so fun, and so technically-empowered. If there are limits they are mostly mine."



Dec 30, 2012 at 02:08 PM
kevinsullivan
Offline
• • •
Upload & Sell: On
p.2 #18 · p.2 #18 · 5D vs 5DM3 High ISO


What I appreciate about the 5D III at high ISO, in comparison to earlier Canon bodies, is not just the somewhat better noise performance but the much better color rendition. To me that's what makes the 5D III usable at high ISO for color photography.


Dec 30, 2012 at 03:14 PM
cineski
Offline
• • • •
Upload & Sell: Off
p.2 #19 · p.2 #19 · 5D vs 5DM3 High ISO


I'm one of the 3 stop crowd. The max I'd ever shoot my 5D is 1600 in the real world (wedding or commercial photography). My 5D3 I shoot often at 12,800. That's a 3 stop advantage for me in real world shooting. Does that mean the 5D3's grain at 12,800 is the same as the 5D at 1600? Heck no. But there's much more to grain and the overall image than what's being presented in this thread. On top of that, the 5D3 doesn't have to have exposure nailed at 12,800 in order to get a perfectly usable image. There's a bit of leeway with that. Even with the 5D2, the max I'd shoot is ISO 3200. Sure, 6400 is good....if you nail exposure. Just a hair off and the image is blown with horrible qualities to the blacks. Back to the 5D and 5D3, the difference between these two cameras with ISO is vast in real world shooting in my opinion. Again, I do recognize the 5D3 has more grain at 12,800 than the 5D at 1600, but the grain structure is filmic and beautiful and there's depth to the image.


Dec 30, 2012 at 03:35 PM
Lars Johnsson
Offline
• • • • • • •
Upload & Sell: Off
p.2 #20 · p.2 #20 · 5D vs 5DM3 High ISO


snapsy wrote:
There are deep shadows in the scene I created. There was mention of 3 stops in the thread at http://www.fredmiranda.com/forum/topic/1176816.


I would also say that 3 stops are closer to the truth than the "exactly 1 stop" you say it is. When you do real life shooting and printing



Dec 30, 2012 at 03:58 PM
1      
2
       3              5       6       end




FM Forums | Canon Forum | Join Upload & Sell

1      
2
       3              5       6       end
    
 

You are not logged in. Login or Register

Username   Password    Retrive password