| View previous topic :: View next topic |
| Author |
Message |
Eddy Clerx

Joined: 30 Jul 2010 Posts: 74 Location: Netherlands
|
Posted: Sun Aug 01, 2010 11:05 am Post subject: artefacts - whats wrong |
|
|
Complete pictures looks funny but I wonder what causes these ugly artefacts.. Could anyone help
 |
|
| Back to top |
|
 |
rjlittlefield Site Admin

Joined: 01 Aug 2006 Posts: 12577 Location: Richland, Washington State, USA
|
Posted: Sun Aug 01, 2010 5:31 pm Post subject: |
|
|
What software and what settings are you using?
--Rik |
|
| Back to top |
|
 |
Eddy Clerx

Joined: 30 Jul 2010 Posts: 74 Location: Netherlands
|
Posted: Mon Aug 02, 2010 1:46 pm Post subject: |
|
|
Hi Rik,
I use helicon focus 5.1. The parameter settings were standard, radius 8, smoothing 4. A second run of r=3 and s=4 did not give significant better results.
tnx Eddy |
|
| Back to top |
|
 |
rjlittlefield Site Admin

Joined: 01 Aug 2006 Posts: 12577 Location: Richland, Washington State, USA
|
Posted: Mon Aug 02, 2010 3:18 pm Post subject: |
|
|
This is a common artifact using Helicon Focus Method B. Sometimes it is associated with uneven illumination between frames, typically caused by irregular flash output. But sometimes it appears in other cases too. The only way I know to address it within Helicon Focus is to switch to Method A. If it also appears in Method A, then I don't know any other tricks to get rid of it.
If you're in a position to consider other software, then take a look at Zerene Stacker. Zerene Stacker's PMax method is essentially invulnerable to this sort of halo and loss-of-detail artifacts, and it often gives cleaner rendering of hairs and overlapping structures. The standard disclaimer is that I wrote Zerene Stacker; the reason I did that was these artifacts we're discussing.
--Rik |
|
| Back to top |
|
 |
Eddy Clerx

Joined: 30 Jul 2010 Posts: 74 Location: Netherlands
|
Posted: Tue Aug 03, 2010 12:35 am Post subject: |
|
|
The lighting is surprisingly contstant. sb600 manual 1/8 with big softbox and a sb26 m 1/64 direct light. The sb26 delivers very constant light.
I will try Method A first and compare it with zerene. Don't worry about that disclaim. It looks very very nice Rik. As the French say, Chapeau! |
|
| Back to top |
|
 |
Eddy Clerx

Joined: 30 Jul 2010 Posts: 74 Location: Netherlands
|
Posted: Tue Aug 03, 2010 11:31 am Post subject: |
|
|
zerene does indeed a better job, the surrounding however need to be cleaned up. It reveiles all my sensor dust that I never knew it was there. Combine the 2 outputs is something I have to practice on yet.  |
|
| Back to top |
|
 |
rjlittlefield Site Admin

Joined: 01 Aug 2006 Posts: 12577 Location: Richland, Washington State, USA
|
Posted: Tue Aug 03, 2010 11:58 am Post subject: |
|
|
Yes, getting trails from sensor dust is a problem. Some people remove those using a healing brush in Photoshop. I usually do it with retouching in Zerene. In uniform background, brushing from a single source image will reduce the trails to a single spot, then brushing from a source image one or two frames away will kill the spot. With deep stacks, I often use retouching to replace PMax background with either the rearmost frame or a DMap background because those are less noisy.
--Rik |
|
| Back to top |
|
 |
scitch
Joined: 29 May 2010 Posts: 431
|
Posted: Wed Aug 11, 2010 11:57 am Post subject: |
|
|
Eddy;
Can you discuss your technique and equipment?
Thanks,
Mike |
|
| Back to top |
|
 |
Eddy Clerx

Joined: 30 Jul 2010 Posts: 74 Location: Netherlands
|
Posted: Thu Aug 12, 2010 11:32 am Post subject: |
|
|
The first stacks I shot with d700 nikon + bellows + schneider componon 50/2,8 on slide, fixed on a ply board(5cm). The steps I measured with slide ruler. This resulted in jumping pictures. Now I am the proud owner of this ..
20micronsteps 20x9cm Here I'm testing an old 28mm f3,5 Ai . works great
Wireless release. The light comes from a sb600 thru a big softbox (normally on the camera scaring away bugs!) and a sb26. Now I use the D200 for testing (shutter!wareoff)
(sorry for the poor pics )
 |
|
| Back to top |
|
 |
Eddy Clerx

Joined: 30 Jul 2010 Posts: 74 Location: Netherlands
|
Posted: Thu Aug 12, 2010 3:41 pm Post subject: |
|
|
stack shot with the f3,5 28mm nikkor old and dusty fem. Oryctes nasicornis, with this setup. and full extended single shot
 |
|
| Back to top |
|
 |
scitch
Joined: 29 May 2010 Posts: 431
|
Posted: Thu Aug 12, 2010 5:24 pm Post subject: |
|
|
Thanks for the discussion and the pictures.
Mike |
|
| Back to top |
|
 |
PaulFurman

Joined: 24 Oct 2009 Posts: 595 Location: SF, CA, USA
|
Posted: Sat Aug 14, 2010 12:30 pm Post subject: |
|
|
| In my experience, very very small focus steps reduce those kind of halos, at least using Zerene. The background halos, I handle by retouching from one layer, or whichever layer has the edge in focus which might be impractical from objects angling toward the camera. |
|
| Back to top |
|
 |
Eddy Clerx

Joined: 30 Jul 2010 Posts: 74 Location: Netherlands
|
Posted: Sat Aug 14, 2010 2:41 pm Post subject: |
|
|
| What worries me most is the soft halo's of the highlights.. should I skip this lens (28mm f3,5 H.C.) and look for something else? |
|
| Back to top |
|
 |
rjlittlefield Site Admin

Joined: 01 Aug 2006 Posts: 12577 Location: Richland, Washington State, USA
|
Posted: Sat Aug 14, 2010 3:12 pm Post subject: |
|
|
Examine your individual frames where the highlight is in focus. If you have the halos there also, then they could be coming from the lens. But if the individual frames do not have the halos, then they are being introduced in the stacking process. It is not uncommon for stacking to introduce such halos. What happens is that the software sometimes makes mistakes in choosing between a bright highlight whose edge is OOF but nonetheless high contrast, and real detail that is well focused but nonetheless lower contrast.
If the halos are being introduced during stacking, then there are a couple of routes to consider.
One of them is to manually retouch to create some combination of DMap and PMax outputs and/or individual source frames that do not have the halos.
Another is to use more diffuse lighting in which the effective light source does not have a hard edge. Eliminating the hard edge on the light source also eliminates the hard edge on the reflection, and that makes it easier for stacking software to do the right thing. To accomplish this, use a diffuser that sits relatively far from the real light source and is large enough that the edges of diffuser are not illuminated by the flash. From the images that you have here, it looks like the diffuser on your softbox is still illuminated clear to its edges.
--Rik |
|
| Back to top |
|
 |
|