Oregon Oxalis; Quinault RF mushroom

Images of undisturbed subjects in their natural environment. All subject types.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

pgk
Posts: 46
Joined: Sat May 15, 2010 11:37 am

Post by pgk »

Thanks for all your comments. I've now tried stacking in Zerene (the previous stack was in Helicon) and edges are, whilst not perfect, at least substantially better and it may be that Zerene is the software which I will go with for such stacks. I think that the 'mush' I referred to is contamination - the result of oof foreground contamination on hardish edges - and Rik's explanation does make good sense. (As an aside, I wonder if the 'bokeh' of a lens has an effect on such contamination?). I've cropped both versions of the stack to really show the edge problem and will be interested to hear any further comments idc. The comparison is not absolute because the first image (lighter and lower contrast) is from HF working direct from RAW files (which seems to ignore the XMP data as far as I can tell) whilst the second is from Zerene working from 16 bit tiffs from optimised RAW files. So I think that Zerene had better base files to operate from, having said which it has produced a substantially better end result.
Image

Image
Paul

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

Post by rjlittlefield »

Paul, thanks for the crops. I prefer to call this problem a "loss-of-detail halo" -- complete loss of background detail in the vicinity of a foreground edge. It is a common artifact with Helicon. Sometimes it can be tuned out by careful choice of neighborhood sizes; sometimes not. The same sort of defect can also occur in Zerene Stacker when the DMap method is used, but the PMax method is essentially immune to it. There are other tradeoffs between DMap and PMax. Briefly, PMax is excellent at finding and preserving detail but it also accumulates more noise and is vulnerable to contrast buildup and color changes. DMap is just the opposite: it is faithful to the original images in terms of noise, contrast, and color, but sometimes misses detail. A good advanced strategy in Zerene Stacker is to stack with both DMap and PMax, then combine the best aspects of each by using Zerene's retouching tool. See the two retouching tutorials linked at http://zerenesystems.com/stacker/docs/H ... Retouching for more information about this approach.

--Rik

pgk
Posts: 46
Joined: Sat May 15, 2010 11:37 am

Post by pgk »

Many thanks Rik

I'm only just starting to try to get to grips with stacking software (running trial copies) but I shot another trial image and tried it in Zerene with substantially better results. I can see spending a lot of time trying to sort this out as I have applications for this type of image if it will work as I want. I'll post any really good images if and as I produce them in new posts.
Paul

Peter De Smidt
Posts: 233
Joined: Thu Jan 05, 2012 9:10 am
Contact:

Post by Peter De Smidt »

Those are very impressive photos!

Harold Gough
Posts: 5786
Joined: Sun Mar 09, 2008 2:17 am
Location: Reading, Berkshire, England

Post by Harold Gough »

Harold Gough wrote:Superb images. I love the wideangle close-ups, taking in middle distance and the backdrop of trees.
Having seen what Zerene Stacker can do to generate stereo pairs from a single stack of macro subjects, it seems that there is considerable potential here.

Harold
My images are a medium for sharing some of my experiences: they are not me.

F100 PHOTOG
Posts: 37
Joined: Thu Sep 14, 2006 12:08 pm
Location: Heart of Illinois

Post by F100 PHOTOG »

Hi Charles,

Fantastic images!

I'll have to say that #2 is my favorite.

Mark
Mark, from the Heart of Illinois

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

Post by rjlittlefield »

Harold Gough wrote:Having seen what Zerene Stacker can do to generate stereo pairs from a single stack of macro subjects, it seems that there is considerable potential here.
Zerene's synthetic stereo has the interesting feature that it works well exactly when you need it the most, with tiny subjects and deep stacks. The potential is more limited for larger subjects and vanishes for landscapes.

The reason is that Zerene can only reconstruct views that lie within the entrance cone of the shooting aperture. At 10X NA 0.25, the cone is wide, about 29 degrees total, so there's lots of opportunity to extract good separation. At 1:1 and f/4 nominal (f/8 effective for both camera and subject), it's only 7 degrees and the method gets more limited. Go to 1:10 at f/4 nominal (f/4.5 effective on the camera side, f/45 for the subject), and the entrance cone is down to only 1.27 degrees. That's pretty narrow for good stereo. Push it to a landscape, the cone angle goes to zero on the subject side, and the potential for synthetic stereo disappears.

--Rik

bvalente
Posts: 69
Joined: Mon Jan 18, 2010 8:13 am
Location: Los Angeles, CA
Contact:

Post by bvalente »

Rik

Might that be okay anyways (or possibly even work better) for subjects like this? The result would be a flat background and a foreground that is 3D. pretty typical stereoscopic results, I would think?


Cheers

Brian
Cheers


Brian

Brian Valente
bvalente.smugmug.com

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

Post by rjlittlefield »

Might that be okay anyways (or possibly even work better) for subjects like this? The result would be a flat background and a foreground that is 3D. pretty typical stereoscopic results, I would think?

That's a good point. When I wrote my last post, I was looking specifically at the images by pgk, at the top of the same page as Harold's comment, on page 3 of the thread. Those images show what look like far-away bushes, where the synthetic stereo effect would be minimal. But looking back to Charlie's images at the start of the thread, it does seem like there might be enough effect in the near foreground to be interesting with those.

--Rik

bvalente
Posts: 69
Joined: Mon Jan 18, 2010 8:13 am
Location: Los Angeles, CA
Contact:

Post by bvalente »

ah i see -

I think those pics you reference at the top of page three are the infinity crop from images from mid page 2, which does have near and far subjects


Brian
Cheers


Brian

Brian Valente
bvalente.smugmug.com

Harold Gough
Posts: 5786
Joined: Sun Mar 09, 2008 2:17 am
Location: Reading, Berkshire, England

Post by Harold Gough »

rjlittlefield wrote: The potential is more limited for larger subjects and vanishes for landscapes.
For those who have not tried stereo on the landscape scale, I would just point out that it doesn't work much beyond the middle distance and certainly out to the horizon. This is, of course, due to the rapidly diminishing parallax with increased distance.

This may be of general interest:

http://web.me.com/abeato/sd3dcalculator ... ntals.html

http://www.dashwood3d.com/blog/beginner ... scopic-3d/

http://nzphoto.tripod.com/stereo/macros ... indows.htm

Harold
My images are a medium for sharing some of my experiences: they are not me.

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

Post by rjlittlefield »

True. To clarify, the difference is that with Zerene's synthetic stereo the baseline is restricted to the width of the lens aperture, while for ordinary two-shot stereo the baseline can be made arbitrarily large to fit the distances involved.
bvalente wrote:I think those pics you reference at the top of page three are the infinity crop from images from mid page 2, which does have near and far subjects
Yep, and obviously I knew that two months ago because I'm the one who asked for the crops to be posted and commented on them at the time. Ah, the limits of memory...

--Rik

Post Reply Previous topicNext topic