5D3suffer blackdots(EDIT:my mistake,just poorly tuned sharpening algo4incamJPGS

[LetTheRightLensIn]

EDIT: I see the older version of the thread got brought back to life so I re-titled this one to not be misleading. I see the new thread got attached to the end of the original thread. Anyway on the new thread it was determined that is much more likely a poor tuning of the in cam jpg sharpening and NR algorithms than anything to do with the black dot issue of the 5D2.

Sorry my mistake. It's probably not a big deal at all. it wont' affect RAW at all. Hopefully they will fine the jpg sharpening/NR before release (these samples were taken months ago apparently). Still not sure why Canon always uses awful sample shots, each camera release seems to lower the bar for the samples they provide.






Wow, after the big deal about the black dots with the 5D2, they post a bunch of 5D3 samples where bright objects have black haloes. The issues looks worse than it ever did with the 5D2.

Hopefully they can fix it again. A bit surprised they have the same sensor issue again.

 

[bycostello]

Re: 5D3 suffers from the black dots problem of the early 5D2

shocking there is especially surrounding all the hype...

 

[thenickdude]

Re: 5D3 suffers from the black dots problem of the early 5D2

Can you post a link to one of the photos which has this problem?

 

[Mt Spokane Photography]

Re: 5D3 suffers from the black dots problem of the early 5D2

A google of 5D MK III black dots has not broght up a single post. Why not provide a link to a valid image?

 

[HurtinMinorKey]

Re: 5D3 suffers from the black dots problem of the early 5D2

This image ?

http://web.canon.jp/imaging/eosd/samples/eos5dmk3/downloads/02.jpg

I see noise in the second pic (it's taken at 6400ISO, so it's what you'd expect).

Okay, I'm seeing some halos around the stars in the first pic. Is this what you are referring to? Could this just be a JPEG artifact?

<a href="http://s1241.photobucket.com/albums/gg507/HurtinMinorKey/?action=view&amp;current=blackHalo.jpg" target="_blank"><img src="http://i1241.photobucket.com/albums/gg507/HurtinMinorKey/blackHalo.jpg" border="0" alt="Photobucket"></a>

 

[HurtinMinorKey]

Re: 5D3 suffers from the black dots problem of the early 5D2

Here it is at 400% magnification.

 

[LetTheRightLensIn]

Re: 5D3 suffers from the black dots problem of the early 5D2

This image ?

http://web.canon.jp/imaging/eosd/samples/eos5dmk3/downloads/02.jpg

I see noise in the second pic (it's taken at 6400ISO, so it's what you'd expect).

Okay, I'm seeing some halos around the stars in the first pic. Is this what you are referring to? Could this just be a JPEG artifact?

<a href="http://s1241.photobucket.com/albums/gg507/HurtinMinorKey/?action=view&amp;current=blackHalo.jpg" target="_blank"><img src="http://i1241.photobucket.com/albums/gg507/HurtinMinorKey/blackHalo.jpg" border="0" alt="Photobucket"></a>

Yeah most of the brighter stars are surround by a black halo. And even on the snowboarder pic it looks like some is outlines in dark blue and black halo, which really brings it into a territory where it never showed at all on the 5D2. On the 5D2 it would mostly show at high ISO when shooting things like Christmas lights and you'd get a black blob to one side of every specular highlight or bright point. Or if you shot night city scapes, each bright window would have a black stripe running along it. Luckily they were able to fix it in firmware. Hopefully they can this time too. I'm kind of shocked it has it again after all that was made out of it when the 5D2 got released with the issue.

It actually looks worse to me than the original 5D2 black dot issue.

Maybe it is just jpg artifacts but it really looks like the 5D2 issue, only a bit worse.

And I am not the only one to notice this. I see that many people on DPR and FM and other sites independently noticed it. In fact CR seems to be the only place where it isn't getting much attention.

Hopefully they can fix it again in firmware and hopefully it doesn't rob anything in the process of the fix.

 

[Loswr]

Re: 5D3 suffers from the black dots problem of the early 5D2

In fact CR seems to be the only place where it isn't getting much attention.

Until now...

 

[pedro]

Re: 5D3 suffers from the black dots problem of the early 5D2

isn't that a pre-production unit?

 

[HurtinMinorKey]

Re: 5D3 suffers from the black dots problem of the early 5D2

It has been suggested that this is an artifact of sharpening. I'm gonna test this once I get home.

 

[well_dunno]

Re: 5D3 suffers from the black dots problem of the early 5D2

As long as it is not suggested they are alien spaceships, we are safe...

 

[Curmudgeon]

Re: 5D3 suffers from the black dots problem of the early 5D2

Deeply disappointed. I was hoping for pink dots this time.

 

[rumorzmonger]

Re: 5D3 suffers from the black dots problem of the early 5D2

At 400% magnification, everything in the image is an interpolation artifact... :

You need to look at the image at 100% in order to see what is actually coming out of the camera.

 

[DavidRiesenberg]

Re: 5D3 suffers from the black dots problem of the early 5D2

According to the EXIF, that image was taken back on October 9th. So we're talking about a camera that will be at least six months old on release date. That means it is probably 5 months older than the production model.

 

[rocketdesigner]

Re: 5D3 suffers from the black dots problem of the early 5D2

At 400% magnification, everything in the image is an interpolation artifact... :

You need to look at the image at 100% in order to see what is actually coming out of the camera.

This a 4x magnification and someone started another gripe thread .... ?

I've heard of pixel-peeping, but this more like anal-pixel-peeping.

 

[DJL329]

Re: 5D3 suffers from the black dots problem of the early 5D2

When I first read the subject line, I thought this was a joke, but now that I've seen the full image, I'm sure of it.

 

[LetTheRightLensIn]

5D3 black outlines - something weird with in cam jpg sharpening+NR (or maybe

EDIT:
maybe it is something different after all since some of it seems to be merely very dark shades and not solid 0,0,0 which it was on the 5D2 (or maybe the same issue but just partially fixed?)

Anyway hopefully they will have better tuned the algorithms on the release firmware if it's just the new jpg processing engine causing it and hopefully it's not some issue that will be there even in the RAW files.

Whatever the case it's pretty sloppy to have so many mangled and ugly samples online to represent the new camera. The fall mtn HDR scene is just scary.

-----------------------


This was a very real problem on the initial 5D2s which Canon themselves admitted and eventually fixed. And these 5D3 samples hint at the issue being far worse on the 5D3 so what gives?

Shouldn't it be made known to Canon rather than hidden away until it's too late?

And I am not to the only to point this out.

And don't laugh it, make a nice print your nighttime city scape and when eery window it outline in black rings maybe you won't be laughing to much any more.

If Canon has said it's been fixed for the release models or there has been some explanations then how about stating that instead of just locking the thread for no apparent reason???

 

[chito]

Re: Why was the 5D3 black outlines thread locked???

wow, that's a valid point, i didn't know about the 5D2.. was it a hardware or software issue?

 

[YellowJersey]

Re: Why was the 5D3 black outlines thread locked???

Good point.

 

[DJL329]

Re: Why was the 5D3 black outlines thread locked???

With the 5D Mark II, the black dots were clearly visible on a 100% crop. In the sample image in the 5D Mark III black dot thread, the dots are barely visible at 400%.

wow, that's a valid point, i didn't know about the 5D2.. was it a hardware or software issue?

The problem on the 5D Mark II was corrected with a firmware update.