70D vs D7100 ISO Comparison at 100%

[MichaelTheMaven]

Ill be posting a bunch of tests between these 2 cameras. Im learning a lot of really interesting things, and so far it has been an absolute slugfest. The Nikon has been surprising me in positive ways, and then the 70D comes back and crushes the D7100 in others. Ill have a full write up available soon.

Check this out. No commentary on my part, decide for yourselves. Quick and dirty test with description of what I did here:

http://www.michaelthemaven.com/?postID=2894&canon-70d-vs-nikon-d7100-iso-noise-chart

 

[mountain_drew]

Ill be posting a bunch of tests between these 2 cameras. Im learning a lot of really interesting things, and so far it has been an absolute slugfest. The Nikon has been surprising me in positive ways, and then the 70D comes back and crushes the D7100 in others. Ill have a full write up available soon.

Check this out. No commentary on my part, decide for yourselves. Quick and dirty test with description of what I did here:

http://www.michaelthemaven.com/?postID=2894&canon-70d-vs-nikon-d7100-iso-noise-chart

Thanks, that's going to be interesting. From another test, I have the impression that the canon is especially bad at low-iso with purple and blue to a lesser extent. Myabe there's something on the Canon with the blue channel.

 

[ajfotofilmagem]

Through my eyes, the 70D images look much better, and I do not care if DXO says otherwise. 8) However, the auto white balance makes strange colors randomly. I also realized that exposure varies greatly in different ISO. You used aperture and shutter speed manual or automatic? Thank you in advance.

 

[Marsu42]

Quick and dirty test with description of what I did here

It's great you're doing reviews on your blog and are the first one to occupy the search engines by carpet-bombing forums, but "quick and dirty" won't do here, and you most likely know it - esp. in this case you gotta be joking :-\ ... "I shot a color chart, hand held, in fluorescent light, with AWB on both cameras, standard ISO noise reduction turned on."

One thing it might show is that Canon has some stronger edge contrast enhancement in place which might also explain the magical sharpness increase some people noticed in res charts.

 

[MichaelTheMaven]

@Marsu42...no, not joking.

This is the only forum I have posted on, so I wouldn't call that "carpet bombing"as you say. I've found some sincere people on here have great recommendations, feedback, suggestions for future tests, etc.

The results speak for themselves, if you disagree, show us your data.

If you can't do that, well.... then you can't authoritatively say anything can you?

 

[Aglet]

Quick and dirty test with description of what I did here

It's great you're doing reviews on your blog and are the first one to occupy the search engines by carpet-bombing forums, but "quick and dirty" won't do here, and you most likely know it - esp. in this case you gotta be joking :-\ ... "I shot a color chart, hand held, in fluorescent light, with AWB on both cameras, standard ISO noise reduction turned on."

One thing it might show is that Canon has some stronger edge contrast enhancement in place which might also explain the magical sharpness increase some people noticed in res charts.

You think this might turn into another long bumpy ride? ;D

 

[ME]

Where are the shots of bbq grills, walls and dr comparison shots? Any comparisons without these is not a valid analysis. And 300% would be better. Images of color charts alone is not enough, though the 70D colors are more pleasing to me. :

 

[pedro]

Well. looks pretty well to my eyes. Is that RAW already? The candles on the posted link look great at 6400 and 12800. Looking forward to Canon's improvements in coming FF sensors.

 

[Marsu42]

This is the only forum I have posted on, so I wouldn't call that "carpet bombing"as you say.

I really have to excuse for that assumption, just seemed very probable given the quick & dirty approach - so sorry, I'll check next time.

The results speak for themselves, if you disagree, show us your data. If you can't do that, well.... then you can't authoritatively say anything can you?

Imho if it is to be worth anything, it has to be done not handheld but with no shake (tripod or similar, contrast af, timer + mlu) - only then it could count as a quick comparisons of jpeg engines and awb metering.

 

[Skulker]

If you can't do that, well.... then you can't authoritatively say anything can you?

;D ;D

I thought the first post was design to court controversy! Then this! I did literally Laugh out loud. I don't know if you mean it as a joke, but it worked for me.

 

[Loswr]

...in this case you gotta be joking :-\ ... "I shot a color chart, hand held, in fluorescent light, with AWB on both cameras, standard ISO noise reduction turned on."

+1

The results speak for themselves, if you disagree, show us your data.

The results say nothing useful - I can quite authoritatively state that your method is badly flawed, which renders your 'results' uninterpretable and no conclusions can be drawn. As Marsu42 points out, AWB and NR are major confounds to a test like this, and handheld shooting doesn't help matters.

Sorry, but the fact that those factors are so easily controllable (you mentioned that you shot under fluorescent lighting, why wouldn't you just set WB to fluorescent?) makes me question the validity of any other such 'test' you perform.

 

[Jim O]

What are we looking at
Different profiles, different contrast curves, different USM and default noise
settings

Give me the raw files and I let you know what can be seen or not

Thus spaketh the true maven of all things photographic.

 

[ajfotofilmagem]

...in this case you gotta be joking :-\ ... "I shot a color chart, hand held, in fluorescent light, with AWB on both cameras, standard ISO noise reduction turned on."

+1

The results speak for themselves, if you disagree, show us your data.

The results say nothing useful - I can quite authoritatively state that your method is badly flawed, which renders your 'results' uninterpretable and no conclusions can be drawn. As Marsu42 points out, AWB and NR are major confounds to a test like this, and handheld shooting doesn't help matters.
Sorry, but the fact that those factors are so easily controllable (you mentioned that you shot under fluorescent lighting, why wouldn't you just set WB to fluorescent?) makes me question the validity of any other such 'test' you perform.

So Neuro. That's what happens when you reaches 10,000 posts? A title of "CR GEEK" under your avatar ... : I understand that some tests are done in a hurry, but need to be repeatable in similar situations. I for one would like a quick test that showed the effectiveness of the JPEG engine and noise reduction in the camera. Since it excludes other variables with auto WB, and hand-held camera.

 

[Badger]

Good lord! No good deed goes unpunished!
Take it for what its worth, or don't.
Thanks for taking the time and effort Mike.

 

[Jim O]

Good lord! No good dead goes unpunished!
Take it for what its worth, or don't.
Thanks for taking the time and effort Mike.

Is spreading potentially inaccurate information a "good deed"? If one wants to publish something even semi-scientific, then one needs to use some semblance of the scientific method. At the very least, one should control as many variables as possible. Using the same tripod in the same location for both cameras would have been a good start. Using the same lens - perhaps a Tamron made for each line of camera, would also help.

I think there are reasonable criticisms here that, IMMHO, make any conclusions suspect. They may well be borne out with more rigorously controlled testing, but without that, they're really just pictures.

 

[Loswr]

Ok, I conclude that the D7100 has problems with color fidelity and exposure metering that occur specifically at ISO 800. Maybe it wasn't a useless test after all... :

 

[ajfotofilmagem]

Ok, I conclude that the D7100 has problems with color fidelity and exposure metering that occur specifically at ISO 800. Maybe it wasn't a useless test after all... :

That's good to know. Just never use the D7100 in ISO800 and everything is great! ;D ;D ;D ;D

 

[Badger]

I was just curious to see how these two cameras did, hand held with auto white balance at different ISOs

 

[ishdakuteb]

Give me the raw files and I let you know what can be seen or not

number of people in this forum have been asking for your raw files for more almost 2 years. have you posted one? none... then why bother to ask people to send you a raw file?

You think this might turn into another long bumpy ride? ;D

where is the link to your photography page? should you be proud to let people see your artist side lol

 

[risc32]

probably because that swatch turns orange at all iso settings except red at iso800 on the nikon. i'm not on a calibrated monitor right now, and i can't say what is more accurate, but certainly iso 800 looks much different then any of the others.