fredmiranda.com
Login

Moderated by: Fred Miranda
  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Canon Forum | Join Upload & Sell

       2       end
  

Archive 2009 · Direct ISO comparison 7D and 20D

  
 
Drayken
Offline
• • •
Upload & Sell: Off
p.1 #1 · Direct ISO comparison 7D and 20D


Ok, not sure I did it correctly, but I am at the end of my return period for the 7D, so I was doing some last minute tests. I used the same lens, 17-55 IS, and shot at ISO 400, 800, 1600. In RAW, converted to DNG in Lightroom 2. No noise reduction. (it "may" have had 20 on luminance, but if it did, it did on all the files) I did a 100% crop. So obviously the 7D pics are larger. I was not sure how to downsize it to make it even. Anyone wanting to do it, feel free. I cannot host the original DNG files, but can email them if someone wants them.

What I see, is that the 7D resolves alot more, it looks much sharper. But is there more noise Below are the 1600 samples, and a link to all of them.

http://www.pbase.com/docphotography/sample

http://www.pbase.com/docphotography/image/119510099/original.jpg

http://www.pbase.com/docphotography/image/119510101/original.jpg



Nov 18, 2009 at 05:07 PM
thedigitalbean
Offline
• • • • •
Upload & Sell: On
p.1 #2 · Direct ISO comparison 7D and 20D


Don't use Lightroom 2.5, use DPP as its a converter that fully supports both cameras. LR's support in 2.5 is very preliminary.

Next a better comparison would be uprez the 20D file to the same as the 7D and then to compare at 100%.



Nov 18, 2009 at 05:09 PM
Drayken
Offline
• • •
Upload & Sell: Off
p.1 #3 · Direct ISO comparison 7D and 20D


How would I downsize the 7D correctly? I took a 7D file, sized it to the same pixels and res, but then do I use bicubic, or what? And is that all I need to do? If so, When I did it, The 7D still looked like more noise. It all cleans up very well, though. Like I said, I am just running things through since I got one more day before I am "stuck" with the 7D. And YES, I have taken real photos. I haven't been pixel peeping the entire time, lol.


Nov 18, 2009 at 05:27 PM
cameron12x
Offline
• • • •
Upload & Sell: On
p.1 #4 · Direct ISO comparison 7D and 20D


Some of us would love it if you could upload some RAW files from both the 20D and the 7D for comparison.


Nov 18, 2009 at 06:12 PM
dehowie
Offline
• • •
Upload & Sell: Off
p.1 #5 · Direct ISO comparison 7D and 20D


Even better get two prints done at 18x12 and see which looks better...
Start comparing them how you will use them.
And the winner is..7D



Nov 18, 2009 at 06:27 PM
Drayken
Offline
• • •
Upload & Sell: Off
p.1 #6 · Direct ISO comparison 7D and 20D


dehowie wrote:
Even better get two prints done at 18x12 and see which looks better...
Start comparing them how you will use them.
And the winner is..7D


I would assume you are right, and plan to print them tomorrow. Did not have time tonight.

I can email the originals, but cannot host.



Nov 18, 2009 at 06:55 PM
Gochugogi
Offline
• • • • • •
Upload & Sell: Off
p.1 #7 · Direct ISO comparison 7D and 20D


I agree, the only comparison that matters is one that uses the image the way you might: two optimized 12x18 prints viewed side by side or, if you never print, huge desktop wallpapers (e.g., 2560 x 1600 pixel wallpaper if you own a 30 inch display). Everything else is sorta abstract...


Nov 18, 2009 at 07:08 PM
Drayken
Offline
• • •
Upload & Sell: Off
p.1 #8 · Direct ISO comparison 7D and 20D


Gochugogi wrote:
I agree, the only comparison that matters is one that uses the image the way you might: two optimized 12x18 prints viewed side by side or, if you never print, huge desktop wallpapers (e.g., 2560 x 1600 pixel wallpaper if you own a 30 inch display). Everything else is sorta abstract...


The thing is, I can make a web sized jpg, and it will still have more noise than the 20D file. It cleans perfectly well, but it just doesn't look to me, like the other pics I see posted at times. I have seen 100% crops posted, and they are perfect. The poster "says" no processing has been done, so I wonder how their pics have no noise, yet mine do. Sigh, today is the 14th day of purchase, D-day.



Nov 19, 2009 at 03:25 AM
Ian.Dobinson
Offline
• • • • • •
Upload & Sell: Off
p.1 #9 · Direct ISO comparison 7D and 20D


Drayken wrote:
The thing is, I can make a web sized jpg, and it will still have more noise than the 20D file. It cleans perfectly well, but it just doesn't look to me, like the other pics I see posted at times. I have seen 100% crops posted, and they are perfect. The poster "says" no processing has been done, so I wonder how their pics have no noise, yet mine do. Sigh, today is the 14th day of purchase, D-day.



Expossure? if you expose to the right you are less likely to get noise or at least less likely to induce more noise when you proccess and are tempted to brighten the image (which you probably wouldn't do if you expossed to the right



Nov 19, 2009 at 04:03 AM
Ian.Dobinson
Offline
• • • • • •
Upload & Sell: Off
p.1 #10 · Direct ISO comparison 7D and 20D


also 100% crops on such large files really bear no relation to what you normally see.
How much worse would the 8mp file look increased to the same 18mp of the 7D?
Also dir you try the 7D in 10mp RAW? it may give you a better test against an 8mp 20D



Nov 19, 2009 at 04:05 AM
cameron12x
Offline
• • • •
Upload & Sell: On
p.1 #11 · Direct ISO comparison 7D and 20D


Drayken wrote:
The thing is, I can make a web sized jpg, and it will still have more noise than the 20D file. It cleans perfectly well, but it just doesn't look to me, like the other pics I see posted at times. I have seen 100% crops posted, and they are perfect. The poster "says" no processing has been done, so I wonder how their pics have no noise, yet mine do. Sigh, today is the 14th day of purchase, D-day.

You may want to download a "trial" version of Adobe Photoshop CS4 (or Elements, or Lightroom 2.6)) and download Adobe Camera Raw (5.6).

I believe all of them use ACR 5.6 to convert from RAW to JPG, and it appears that v5.6 may have cleaned up much of the problem.

See this thread.



Nov 19, 2009 at 04:52 AM
Gary Gray
Offline
• •
Upload & Sell: Off
p.1 #12 · Direct ISO comparison 7D and 20D


Oh boy, another computer monitor high ISO noise comparison between a 8 megapixel 6 year old camera and a 18 megapixel 6 week old camera.

Science is having it's finest moment here.

Geeks gone wild!



Nov 19, 2009 at 10:49 AM
Gochugogi
Offline
• • • • • •
Upload & Sell: Off
p.1 #13 · Direct ISO comparison 7D and 20D


Gary Gray wrote:
Geeks gone wild!


I think I bought that DVD...



Nov 19, 2009 at 12:00 PM
Drayken
Offline
• • •
Upload & Sell: Off
p.1 #14 · Direct ISO comparison 7D and 20D


Gochugogi wrote:
Maybe they used JPEGS at default (in-camera NR). Or used DPP and didn't realize the default settings use the same NR as the in-camera JPEGs. Or exposed to the right and pulled the exposure down a stop before converting RAW to TIFF. It only makes sense to me to process a file at the optimal settings that result in a beautiful rendering of the image. It's easy to make it look ugly if you want to. For example, below is a pixel level view of an ISO 1600 image (17-55 [email protected]) I shot with my 7D (full image below it).
...Show more

You said you overexposed? Your exif shows you shot Exposure Bias 0. I used +1/3. Is this what your talking about?



Nov 19, 2009 at 04:18 PM
Gochugogi
Offline
• • • • • •
Upload & Sell: Off
p.1 #15 · Direct ISO comparison 7D and 20D


Drayken wrote:
You said you overexposed? Your exif shows you shot Exposure Bias 0. I used +1/3. Is this what your talking about?


I'm not sure what you're trying to say. I can say I was in charge of the exposure, not the camera. The meter is merely a tool to help you estimate exposure. It can't think and doesn't know a black wall from a white one.

According to the camera histogram, I exposed to the right: no shadows and the mids were all pushed to the right. You don't necessarily need to use EC to expose to the right. How can this be so? The door was extremely dark toned and not a normal distribution of tones. Hence, the meter rendered dark tones as mid-tones (i.e., it overexposed). I knew the HLs were minimal so I let it stay--a perfect "expose to the right" setting. I had to pull it down one stop prior to conversion for the correct tonality in the final image.

On the other hand, had this scene been a landscape with a normal distribution of tones, I would have needed to add +EC or used substitute metering. What's substitute metering?: meter an area you want rendered as medium, use AEL and bang away. If I meter a mid toned area of grass or clear blue sky I'm get a perfect exposure for an average landscape. If I meter dark earth I get about 1 stop of overexposure. A black rock, 1.5-2 stops of overexpose. Dried grass, about 1 stop of underexposure. White sand or snow, 1.5-2 stops of underexposure. That's the basic idea.



Nov 19, 2009 at 07:12 PM
Drayken
Offline
• • •
Upload & Sell: Off
p.1 #16 · Direct ISO comparison 7D and 20D


Well, my father came over, and we shot at ISO 400,640,800,1250,1600,3200. We used a tripod, and used AV mode. Set +1/3 exp bias. Results:

-1st, the 5.6 RAW did make things slightly better, across the board.

-2nd, every shot we compared, my father's was slightly exposed brighter, and slightly less noise. (both being "slightly") So, I was going back and forth, wondering why. Then finally, I noticed our shutter speeds. His was ALWAYS one step slower than mine. Example: ISO 800 = 1/20 vs 1/25 and ISO 1600 = 1/30 vs 1/50. So, I assume even this slight of a difference in shutter speed, would affect exposure and noise. I am not sure why they differed. Maybe how we were standing, or maybe just the difference in the cameras themselves. I sure wish I would have set them in manual mode, but since all the shots differed by one step, I assume that is why I saw more noise.

-3rd, I took both camera ISO 800 files, resized to 360dpi and printed. I know that I never cared for printing my 20D ISO 800 files without using noiseware. I did not sharpen, nor use noiseware, and I will be honest, I would have accepted them "as is". They were very sharp, and noise was not as noticeable. So good in fact, that I then ran through noiseware, then nik sharpener, and reprinted. Hardly any difference at all. That is how sharp they were to begin with.

I decided to keep the camera. I cannot see my father and I both, having poor noise. It is just a huge jump from our 20D and 30D cameras. I have to get used to seeing them on the monitor at larger sizes, while I edit.



Nov 19, 2009 at 07:18 PM
Drayken
Offline
• • •
Upload & Sell: Off
p.1 #17 · Direct ISO comparison 7D and 20D


Gochugogi wrote:
I'm not sure what you're trying to say. I can say I was in charge of the exposure, not the camera. The meter is merely a tool to help you estimate exposure. It can't think and doesn't know a black wall from a white one.

According to the camera histogram, I exposed to the right: no shadows and the mids were all pushed to the right. You don't necessarily need to use EC to expose to the right. How can this be so? The door was extremely dark toned and not a normal distribution of tones. Hence, the meter rendered
...Show more

I always want to learn how to take better photos, so I am trying to understand all you are saying. First things first. Did you shoot in manual? If so, then I think I understand. I shot in AV, so if I am not mistaken, I cannot expose to the right, unless I choose +EC, correct? And thank you very much for your explaining and help.



Nov 19, 2009 at 07:24 PM
Drayken
Offline
• • •
Upload & Sell: Off
p.1 #18 · Direct ISO comparison 7D and 20D


Gary Gray wrote:
Oh boy, another computer monitor high ISO noise comparison between a 8 megapixel 6 year old camera and a 18 megapixel 6 week old camera.

Science is having it's finest moment here.

Geeks gone wild!


With all due respect, I would have agreed with you 100% of the time, until Canon released the MKIII, and I remember all the complaining and all the people saying how it was user error and such. Well, ever since then, I guess some paranoia has set in, lol. I tend to check, and recheck things. I kept my last camera for quite some time, and plan to do the same with this one, so I was just making sure.



Nov 19, 2009 at 07:30 PM
Gochugogi
Offline
• • • • • •
Upload & Sell: Off
p.1 #19 · Direct ISO comparison 7D and 20D


Actually I was in Av mode. Shooting to the right has nothing to do with exposure mode and everything to do with your histogram: the tones representing the image are pushed up against the right side of the histogram. That's it. In a scene with a very restricted dynamic range like my door, the shadows were placed just to the left of center and the mids near the right side. Why? the meter is to stupid to know the door is supposed to be a medium brown rather than light brown. No user intervention needed.

I think you're assuming the meter knows the difference between a white wall and a gray wall and therefore always automatically gives you the right exposure. Unfortunately it does not. Go to a wedding and fill the VF with a portrait of the bride in white dress. Shoot at metered value and see what happens. The image is underexposed because the white dress was expected by the meter to be an average distribution of tones, not just light tones. So it made that white dress gray. Now fill the VF with the groom in black tux. He's way overexposed--meter thought he should be gray too--and you just "exposed to the right" in P mode. Of course if you expose too far to the right the whites can't be recovered when you drag the exposure down before RAW conversion.



Nov 19, 2009 at 07:46 PM
Drayken
Offline
• • •
Upload & Sell: Off
p.1 #20 · Direct ISO comparison 7D and 20D


Gochugogi wrote:
Actually I was in Av mode. Shooting to the right has nothing to do with exposure mode and everything to do with your histogram: the tones representing the image are pushed up against the right side of the histogram. That's it. In a scene with a very restricted dynamic range like my door, the shadows were placed just to the left of center and the mids near the right side. Why? the meter is to stupid to know the door is supposed to be a medium brown rather than light brown. No user intervention needed.

I think you're assuming the meter
...Show more

Ok, so are you saying that since your particular picture was of a darker subject, the camera over exposed some? You did not need to do anything? When I read your first post, you made it sound as if you did that on purpose. If I were to "choose" to shoot to the right, then I would use the +EC, correct?



Nov 19, 2009 at 09:30 PM
       2       end




FM Forums | Canon Forum | Join Upload & Sell

       2       end
    
 

Welcome back
Log in to your account