OFFICIAL: CANON EOS 5D Mark III - User Thread - Part 3

[spinworkxroy]

Correction, not if but that its i misquoted and misunderstood u. There seems to be quite a few of bad press in clubsnap on 5d3 that it's on a tipping point. We are all 5d3 users and like u, I have some wishes too. Like in body wireless flash control which I think 60d abd 7d has. FYI, I upgraded from a 500d and difference is overwhelming for me.

Hopefully I would be able to share some 5d3 pics soon! Happy shooting!

No worries..i agree, the wireless flash thing would've been good..I've been enjoying it on my 60D and when i found out i couldn't do it on the 5d3 because there's no internal flash to trigger anything…i could't do off camera flash anymore I went to buy a Phottix wireless trigger thing…only to realise there was no wireless TTL function (my fault for not researching)…so i'm back to square one… i gave up eventually..am not going to pay big bucks for a pocketwizard to have wireless TTL flash when i don't actually use flashes at all…oh…and buying the latest 600EX flash isn't going to help either You're still going to need to buy the Canon wireless flash trigger to fire the 600EX off camera…it may have a built in radio..but not the 5D3 so it can control other flashes…our camera body just can't control it…basically…more $$ need to spend..
i guess we would all love to be able to have wifi and stuff…evenif it were a cheap attachement…the lower end models have such bells and whistles but the higher end ones don't…then again…a lambo seats 2 while a sedan sits 5 so yea…the higher the model, the less "features" you're gonna have…

 

[spinworkxroy]

wonder what kind of a controlled test he did. I used the shadow area in his link to do some shadow recovery and exposure pushing myself, and i could not get anything to that level of noise in what he did.

Thanks for sharing man…i had no idea…I'm not sure what tests he did too…but he's a Canon user so he can't be doing it on purpose to screw Canon. He must be doing some super shadow recovery and some super software that manages to make things worse.

 

[raydio]

wonder what kind of a controlled test he did. I used the shadow area in his link to do some shadow recovery and exposure pushing myself, and i could not get anything to that level of noise in what he did.

Great test! The tester in Fred Miranda may have been doing something adversely to bring out the worst in the 5D3, intentionally or otherwise. Your test proves that there is another story...

 

[linkk]

agree with the post above, this means one thing don't easily believe in what other people claim in the internet and worst convincing and misinforming others of what you just read :dunno:

 

[aaSKaa]

On the other hand, we also have to accept that 5D3 does have its limitation.

 

[NewbieInCS]

The test done by Fred Miranda for the dynamic range is quite accurate. I did similar test before. Under expose, then pull up by two stops. Plus the effect of peripheral illumination, the corner and edge probably got pulled up by 5-6stops because my lens had 3-5stops of fall off at the corner that I was trying to pull. Can clearly see the banding and noise there even just viewing on my 23" lcd screen.

 

[ricleo]

The test done by Fred Miranda for the dynamic range is quite accurate. I did similar test before. Under expose, then pull up by two stops. Plus the effect of peripheral illumination, the corner and edge probably got pulled up by 5-6stops because my lens had 3-5stops of fall off at the corner that I was trying to pull. Can clearly see the banding and noise there even just viewing on my 23" lcd screen.

well, his test images are based on a 3rd party lens, so there will not be any peripheral illumination, and his test crops were in the centre/centre lower frame and not near corners where illumination will take place.

i used his other crop to do the test using the shadow recovery tool in PS3, and note. only 1% recovery was needed to bring it up to levels brighter then his super noisy 5D3 recovery. D800 recovered image placed for comparison.

 

[ricleo]

The test done by Fred Miranda for the dynamic range is quite accurate. I did similar test before. Under expose, then pull up by two stops. Plus the effect of peripheral illumination, the corner and edge probably got pulled up by 5-6stops because my lens had 3-5stops of fall off at the corner that I was trying to pull. Can clearly see the banding and noise there even just viewing on my 23" lcd screen.

now is it fair to judge a camera's performance based on pushing/recovering shadow areas by 5-6 stops? I believe the D800 sensor is very good and excellent indeed, but to say the 5D3 is lousy in dynamic range based on post processing pushing, that seems like another matter, and fact is i didn't need to go to any extreme 5-6 stops push to recover that extremely dark shadow area in the fred minranda sample.

we also have to note that the 5D3 sample was upsized to 36MP to match the D800, a pretty good job by the 5D3, based on my recovery using his samples i shld say

 

[NewbieInCS]

My point was that at 5 stop with base iso, I was already seeing banding and noise in a normal small view. So I don't think it is unfair to say that there will be more noise if you pull two stop of the shadow, sample it to 36mp, then take out a 100% crop to compare to 100% crop from d800 with higher dynamic range.

 

[ricleo]

My point was that at 5 stop with base iso, I was already seeing banding and noise in a normal small view. So I don't think it is unfair to say that there will be more noise if you pull two stop of the shadow, sample it to 36mp, then take out a 100% crop to compare to 100% crop from d800 with higher dynamic range.

I think it's fair to say the D800 has a great sensor with DR. Pull back exposure at 2 stops is fine too, and it u see what I did for the 5D3 using the fred mindara samples, I don't think the 5D3 is that bad. So not sure what methods he uses to pull back shadow detail.

But you do realise pulling back 5 whole stops from an underexposed image is frankly quite extreme. Even if the d800 sensor allows that, me as a photographer will not shoot and accept a image that was 5 stops under exposed and go back home to attempt o rescue it. I wld take another shot properly exposed.

 

[aaSKaa]

Canon NR, too much smoothening, sharpnest gone.

 

[guseliz]

I wonder why the sunlight in the bottom right image is different than the original image before processing :dunno:
How did he achieve such image after processing?

wonder what kind of a controlled test he did. I used the shadow area in his link to do some shadow recovery and exposure pushing myself, and i could not get anything to that level of noise in what he did.

 

[spinworkxroy]

The main problem here i believe is the up sampling of the 22mp to a 36mp image size..that's like 1.5x more..
if it's 22MP, then leave it as 22mp..test each camera for what it is and there's no need to "standardize" anything is what i say...you buy a camera for what it is and not for it to pretend to be something else...upsampling is never a good idea...just like how up sampling a DVD movie into a 1080p doesn't make it a full HD movie..it just pretends to be.
but yes...you shouldn't have to take a photo and crop the hell out of it to get an image you want and you shouldn't have to take a photo and increase the shadows by 5stops in the first place..
If you need to do any of the 2, i think it's time to train yourself to be a better photographer in the first place hehe..

 

[doodah]

i used his other crop to do the test using the shadow recovery tool in PS3, and note. only 1% recovery was needed to bring it up to levels brighter then his super noisy 5D3 recovery. D800 recovered image placed for comparison.

Wah... This is eye-opening! :bigeyes:

I have done similar tests in the past when some internet site compared my paltry 450D to some Nikon camera and claimed the 450D had terrible shadow noise. I rented the Nikon camera and did a side-by-side testing. No matter what I did, I simply could not cough out the same kind of noise shown on the site.

Anyway, why not post your results on FM forum and ask Fred Miranda himself what he did? Fred Miranda is a rather impartial reviewer, so I doubt he'll lie about this.

PS: It just occurred to me: did you turn on chroma noise reduction? That will help to remove the chroma noise. Of course, if one can turn on chroma noise removal without removal significant details, that may very well be the solution.

 

[ricleo]

Wah... This is eye-opening! :bigeyes:

I have done similar tests in the past when some internet site compared my paltry 450D to some Nikon camera and claimed the 450D had terrible shadow noise. I rented the Nikon camera and did a side-by-side testing. No matter what I did, I simply could not cough out the same kind of noise shown on the site.

Anyway, why not post your results on FM forum and ask Fred Miranda himself what he did? Fred Miranda is a rather impartial reviewer, so I doubt he'll lie about this.

PS: It just occurred to me: did you turn on chroma noise reduction? That will help to remove the chroma noise. Of course, if one can turn on chroma noise removal without removal significant details, that may very well be the solution.

Actually I didn't do any post processing or NR apart from the shadow highlight haha.... Perhaps his version of ACR does not handle 5D3 files well.

 

[ricleo]

Canon NR, too much smoothening, sharpnest gone.

It's not NR, it's upsizing the 22mp to 36mp.

 

[dodgethis]

Actually I didn't do any post processing or NR apart from the shadow highlight haha.... Perhaps his version of ACR does not handle 5D3 files well.

There's only one version of ACR that can handle the MkIII's CR2 anyway.

 

[Anthony Lee]

The test was done buy a guy named "theSuede" who happens to be a technical expert on digital sensors, and I have read his contribution for years. If any of you guys want to question him, make sure you do your homework first. My advise is, just let it be and enjoy your 5D3 as it's a damn good camera.

D800 spanked the 5DIII in every DXO category! - FM Forums

 

[linkk]

i hope this thread will not turn into 5d3 vs Nikon D800 yet again sigh.

 

[dodgethis]

I'm curious, those who are making the comparisons, have they gotten their MkIIIs yet? Or they just feel compelled to point every single error on this thread to convince themselves to not buy the camera?