Strange Artifact with DMAP

A forum to ask questions, post setups, and generally discuss anything having to do with photomacrography and photomicroscopy.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

billjanes1
Posts: 91
Joined: Fri Dec 30, 2016 1:59 pm
Location: Lake Forest, IL, USA

Strange Artifact with DMAP

Post by billjanes1 »

Today I was doing depth of field focus step to check for focus banding and noted a strange banding artifact with DMAP and also with Helicon using metohod 2 (depth map). After some experimenting I think I found the cause. The stack was takin under florescent illumination using magnetic ballast and likely cycling at 129 Hz (USA power). The artifact was no longer present when stacking was done under tungsten illumination. Exposure was f/5.6 @ 1/125 sec. Is this a common problem?

Image[/img]

rjlittlefield
Site Admin
Posts: 23561
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

I cannot remember a problem like this ever being reported before.

The EXIF data shows that this was shot with a Nikon D3 using 105 mm f/2.8 lens at f/5.6, with an exposure time of 1/125 second. Is that correct? If so, then I cannot think of any reason why there would be periodic defocusing of the image. There could be some variation in brightness both between and within images, but that would not produce anything that looks like this.

What do the individual source images look like?

--Rik

billjanes1
Posts: 91
Joined: Fri Dec 30, 2016 1:59 pm
Location: Lake Forest, IL, USA

Post by billjanes1 »

rjlittlefield wrote:I cannot remember a problem like this ever being reported before.

The EXIF data shows that this was shot with a Nikon D3 using 105 mm f/2.8 lens at f/5.6, with an exposure time of 1/125 second. Is that correct? If so, then I cannot think of any reason why there would be periodic defocusing of the image. There could be some variation in brightness both between and within images, but that would not produce anything that looks like this.

What do the individual source images look like?

--Rik
Rik,

The EXIF data are correct. The individual images look OK to me. The 9 image stack under fluorescent illumination can be viewed on the Adobe Cloud here. The DMAP image is shown first, followed by the individual images of the stack.

Stacking 17 images under the fluorescent lights yields less artifact as shown here.

A nine image stack under tungsten illumination is relatively free from artifact. The individual images as well as the stacked result with DMAP and Helicon Method 2 look similar.

Regards,

Bill

PS
Those not familiar with viewing images on the should know that one can view an image by clicking on it, and an enlarged view can be toggled by clicking on the 4 outwardly pointing arrows and one can go back to a smaller view by clicking on the 4 inwardly pointing arrows.

rjlittlefield
Site Admin
Posts: 23561
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Thanks for posting the source images.

The problem is that your source images are not in focus order. img_001 through img_008 step focus nicely from bottom to top of page, but then img_009 is almost an exact duplicate of img_004, focused in the area where you're seeing the artifact.

What's happening is that in that problematic area, the focus-detection algorithm is alternately detecting frames 004 and 009 as being the best focus. That results in the depth map slewing through frames 005, 006, 007, and 008, which have nothing in focus in that area. In bands where the slew happens, the result ends up showing blur.

The solution to this problem is simple: just eliminate img_009 from the stack, so that you process only frames 001-008 where focus steps neatly from one extreme to the other.

Looking at your result from the other stack, I think you have a similar problem there, but the resulting artifacts are not so obvious.

--Rik

rjlittlefield
Site Admin
Posts: 23561
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

OK, I took a detailed look at the 17-frame stack. Focus order there is quite weird. Starting in frame 001, focus moves up the target page in 002, stays almost unchanged in 003, moves down a hair in 004, then progressively up in frames 005 though 008. In img_009 it jumps down again, farther up in 010, down in 011, up again in 012, up farther in 013, down again in 014, up in 015, up in 016, and finally way down in 017.

In the 9-frame tungsten stack, focus order seems to step correctly from one extreme to the other.

So now the question becomes this: why are your fluorescent stacks out of order, while your tungsten stack is not?

--Rik

billjanes1
Posts: 91
Joined: Fri Dec 30, 2016 1:59 pm
Location: Lake Forest, IL, USA

Post by billjanes1 »

rjlittlefield wrote:OK, I took a detailed look at the 17-frame stack. Focus order there is quite weird. Starting in frame 001, focus moves up the target page in 002, stays almost unchanged in 003, moves down a hair in 004, then progressively up in frames 005 though 008. In img_009 it jumps down again, farther up in 010, down in 011, up again in 012, up farther in 013, down again in 014, up in 015, up in 016, and finally way down in 017.

In the 9-frame tungsten stack, focus order seems to step correctly from one extreme to the other.

So now the question becomes this: why are your fluorescent stacks out of order, while your tungsten stack is not?

--Rik

Rik,

Thanks for investigating this matter. The stacks were taken with Helicon Remote. Looking at the EXIF of fluorescent stacks I see that the shots are arranged in chronological order but your investigation reveals that the focus order is erratic. Perhaps the software or the camera autofocus system is malfunctioning. However, I did not change the settings of the camera or software between the stacks. I will double check my workflow.

Regards,

Bill

rjlittlefield
Site Admin
Posts: 23561
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Very interesting! I was afraid it was something like that. I suggest touching base with the support people at Helicon. Perhaps they've seen it before.

--Rik

billjanes1
Posts: 91
Joined: Fri Dec 30, 2016 1:59 pm
Location: Lake Forest, IL, USA

Post by billjanes1 »

rjlittlefield wrote:Very interesting! I was afraid it was something like that. I suggest touching base with the support people at Helicon. Perhaps they've seen it before.

--Rik
I will do that. The mis-ordering of the images was suggested by the observation that the artifact was present with DMAP but not PMAX. I would think that focus banding would affect both stacking algorithms.

It would be very nice if Zerene would produce its version of Helicon Remote. Your Stackshot interface is brilliant.

Bill

Post Reply Previous topicNext topic