Pink Carnations

[Round Tuit]

I have been inspired by Dan's work for some times now and aspire to one day create a picture that will come near the perfection that he achieves in his flower creations. In the mean time, I keep trying.
My latest attempt was taken in the early morning a few days ago using natural light from a north facing window. It is a stack of 7 images using the Pmax algorithm from Zerene.



C&C welcomed as usual

[DanK]

Andre,

thanks very much for the undeservedly big complement. I'm glad that the images I've posted have encouraged you to try more like this.

It's hard to tell at this resolution, but it looks to me as though the stacking worked very well. I don't see obvious stacking artifacts.

My one major suggestion is that your lighting is harsh, with a bunch of hot spots and possibly some clipping. In my experience, with things like flowers, we often perceive natural light as being softer than it really is. In this case, I would use a small diffuser, like this https://www.bhphotovideo.com/c/product/1220015-REG/westcott_1470_illuminator_white_translucent_20.htm l. to soften the light, and I would probably expose about half a stop darker.

For flowers like this, it's worth having Zerene stack with both algorithms and comparing the results. DMap does better at preserving colors, but it is much more vulnerable to stacking halos. Given how deep this flower is, my guess is that DMap would produce a lot of halos. If not, you might like the results better. If it does produce halos, then the only choice is to stick with PMax as you have or to touch up the DMap image from the PMax composite. I do that often, blowing the images up to 50% to get a clearer view of the halos. It often works well, but it can be tedious and time consuming.

I hope this helps. Keep them coming.

Dan

[Manfred M]

André - the same general observation that Dan has pointed out regarding the "hot" petals on the camera right side. I wonder if a more diffuse light source (i.e. a north facing window) might help control this.

When it comes to using Zerene, I will defer to Dan. While I have started to use it in some of my work, Dan has a lot more experience than most people I know.

The third comment deals with how you posed your carnations. I think there may be some lessons to learn from how Dan arranges and presents his flower shots. When I look at this image, the green leaves along the top and bottom are rather distracting. Arranging them differently so nothing more than the stem crosses the bottom of the frame and the placement of other leaves complements rather than competes with the flowers might be worth considering.

[xpatUSA]

André - the same general observation that Dan has pointed out regarding the "hot" petals on the camera right side. I wonder if a more diffuse light source (i.e. a north facing window) might help control this.

Yes, the GIMP measures about 35% blown (255/255) pixels in the flower at camera right. Curiously, there are no large areas of such, perhaps implying a lot of tiny specular reflections.

[Dave Humphries]

Hi André,

I concur with all the earlier comments.

Viewers can easily appreciate/evaluate Manfred's third point by holding a hand in front of the topmost green leaves sprouting above the blooms.

Even blown up big, I am not seeing stacking errors, although my opinion is based on what I have learnt from DanK here over the years (in terms of what to look for).

Well done,
Dave

[xpatUSA]

Yes, the GIMP measures about 35% blown (255/255) pixels in the flower at camera right. Curiously, there are no large areas of such, perhaps implying a lot of tiny specular reflections.

Belay the comment about specular reflections, I had failed to notice that the original post was a fused stack, duh.

So maybe the fusion process did it:



Black = warning level of 255 ...

[DanK]

So maybe the fusion process did it:

Very interesting. I have occasionally noticed specular highlights appearing in stacked composites, although nowhere nearly on this scale. I didn't keep track, but my vague recollection is that it was more common with the PMax method. I have never gotten around to investigating why. The author of Zerene has a lot of technical information online, so I'll look. With Boston largely locked down, I have more time on my hands these days. If I find something, I'll post.

[DanK]

Well, that didn't take long. Rik (the author of Zerene) has the explanation on his FAQ page. My recollection that this is primarily a PMax issue was correct. I'll paste the relevant text from http://zerenesystems.com/cms/stacker/docs/faqlist explaining the problem and solution.

I suspect I did read this years ago, because I usually keep this option checked. The cost, explained just above it, is that the resulting composite can look washed out, which you can deal with in subsequent editing.

What does "Retain extended dynamic range" mean?

As background, you need to know that the PMax stacking method often causes contrast to increase, pushing darks darker and brights brighter. If your source images are already high contrast, then the increase can internally push pixel values to “darker than black” or “brighter than white”. Such values cannot be saved in ordinary image files. By default, Zerene Stacker clips these pixels to exactly black or white when the file is saved, thus throwing away some information you might like to keep. Placing a checkmark on “Retain extended dynamic range” essentially does a “levels adjustment” that reduces contrast and possibly brightens the image, exactly enough to occupy the full range of allowed pixel values, 0-255 in an 8-bit image. This preserves all the computed pixel values so that you can apply your own levels or curves adjustment in Photoshop or any similar tool to get whatever appearance you like best. When using “Retain extended dynamic range”, it's also a good idea to use 16-bit TIFF output, so as to preserve good gradation that might be lost if the extra dynamic range were compressed into 8 bits.

[Round Tuit]

Wow! Go away for a few hours and you have a stack(pun intended) of comments to address. Let me start at the top.
Harsh light: The apparent harshness of the light I believe is a result of poor post processing. I tend to like my pictures on the high contrast and crunchy side. In this case, easing off on the contrast and microcontrast would have given better results. I say this because my north facing window is four feet by four feet and the carnations were about five feet from the window. That I believe would make it a diffused light source.
PMax vs DMax: I have used both algorithms on every stack that I have made so far. I must admit that I havn't seen much difference between the two as far as colour goes and that I usually end up using the PMax result.
Posing the flowers: I have not been very active lately photographically speaking. All I can say is "mea culpa, mea culpa, mea maxima culpa" of a stupid rookie mistake. Concentrate on the flower; ignore the leaves.
PMax increasing the contrast: I did not have the "retain extended dynamic range" box checked. Something that I will do in the future.
Thank you all for your inputs. I will go back to the drawing board and see what I can come up with.
Dave - It is nice to have you back.

[xpatUSA]

... my north facing window is four feet by four feet and the carnations were about five feet from the window. That I believe would make it a diffused light source.

Not really, André. Say your window has a "radius" of 2ft and the distance to the subject is 5ft, that puts the lighting in Inverse Square Law territory, see P.26 here:

http://kronometric.org/phot/lighting...20handbook.pdf

To be "diffused", a 25ft window would work, I reckon.

[Manfred M]

Not really, André. Say your window has a "radius" of 2ft and the distance to the subject is 5ft, that puts the lighting in Inverse Square Law territory, see P.26 here:

http://kronometric.org/phot/lighting...20handbook.pdf

To be "diffused", a 25ft window would work, I reckon.

I was thinking that putting a scrim (diffuser) in front of the window would be an even easier solution. A plain white shower curtain works quite well in these situations.

[Round Tuit]

Not really, André. Say your window has a "radius" of 2ft and the distance to the subject is 5ft, that puts the lighting in Inverse Square Law territory, see P.26 here:

http://kronometric.org/phot/lighting...20handbook.pdf

To be "diffused", a 25ft window would work, I reckon.

Very interesting book for a nerd like me. But, to be pedantic, the referenced section is not relevant to the problem at hand. I am not measuring irradiance but rather am using the broad angle that the light illuminates my subject to soften the shadows. With the geometry of my setup, this angle is roughly 40°. The sun at ½° would definitly be a point source.

[Round Tuit]

I was thinking that putting a scrim (diffuser) in front of the window would be an even easier solution. A plain white shower curtain works quite well in these situations.

I don`t think that a scrim would make any difference as the light coming through the window is from a uniformly illuminated sky.

[Manfred M]

I don`t think that a scrim would make any difference as the light coming through the window is from a uniformly illuminated sky.

That would suggest that you are dealing with either north light or an overcast sky?

The right hand side of your image looks "hot" so there is something going on there that needs to be toned down.

[DanK]

I don`t think that a scrim would make any difference as the light coming through the window is from a uniformly illuminated sky.

I don't think that's right. A diffuser is not just a way to deal with point sources of light.

First, a diffuser would slightly lessen the difference in luminosity between the areas illuminated by direct sunlight and the rest. That in itself would help a bit.

Second, and more important, a diffuser scatters the light. This makes it softer even when there is not a point source, and it therefore softens shadows. Check out the photos of a frog here: https://en.wikipedia.org/wiki/Diffuser_(optics)

Re the differences between PMax and DMap: although you haven't seen them yet, you probably will. I've traded notes with Rik about this, and he has a brief description on the FAQ pages: http://zerenesystems.com/cms/stacker..._pmax_and_dmap:

What is the difference between PMax and DMap?

PMax is a “pyramid” method. It is very good at finding and preserving detail even in low contrast or slightly blurred areas. It's also very good at handling overlapping structures like mats of hair and crisscrossing bristles, nicely avoiding the loss-of-detail halos typical of other stacking programs. But PMax tends to increase noise and contrast, it can alter colors somewhat, and it's liable to produce fuzzy “inversion halos” around strongly contrasting objects.

DMap is a “depth map” method. It does a better job keeping the original smoothness and colors, but it's not as good at finding and preserving detail.

The two methods complement each other. Some types of subjects look good when they are processed automatically by PMax, but not by DMap. Other subjects are just the opposite. For particularly challenging subjects like bugs and flowers shot through microscope objectives, neither method is ideal by itself. In that case the best results are obtained by using human judgment and the retouching tool to combine the best aspects of both algorithms.

For further discussion of these issues, see “DMap versus PMax” on the Zerene Stacker: How To Use It page.

For this reason, in the case of flowers with color--not wilted flowers and seeds--I most often use a DMap composite as my base image and touch up areas with halos from a PMax composite or individual images, unless the halos are so bad and numerous that I can't fix them with a reasonable amount of effort. I've never tried working in the other direction, that is, using a PMax composite as a base and painting on onto it from a DMap image for areas with a lot of color, but I will probably try that soon.

[xpatUSA]

Originally Posted by Round Tuit I don`t think that a scrim would make any difference as the light coming through the window is from a uniformly [luminous] sky.

That would suggest that you are dealing with either north light or an overcast sky?

The right hand side of your image looks "hot" so there is something going on there that needs to be toned down.

Agreed.

An alternative would be to move the subject much closer to the window and, if possible, add in a couple of reflectors. I find the white side of 8x10" Kodak R27 "gray" cards quite good for that purpose.

As Manfred has said in the past, getting the lighting correct before the shot is a good thing ...

[xpatUSA]

Originally Posted by Round Tuit I don`t think that a scrim would make any difference as the light coming through the window is from a uniformly illuminated sky.

I don't think that's right. A diffuser is not just a way to deal with point sources of light.

First, a diffuser would slightly lessen the difference in luminosity between the areas illuminated by direct sunlight and the rest. That in itself would help a bit.

Second, and more important, a diffuser scatters the light. This makes it softer even when there is not a point source, and it therefore softens shadows.

Absolutely and well said.

I went through all that malarky when shooting wrist-watches on a bench. I ended up with two LED floods and two home-made diffusers (A4 tracing paper with added aluminum-foil borders).

Unlike André, I found relative distances to be very important - and it also helped that I could curve the diffusers around a watch if I so desired.

[xpatUSA]

Very interesting book for a nerd like me. But, to be pedantic, the referenced section is not relevant to the problem at hand.

The relevance lies in relationship of the light source size to the distance of the subject therefrom.

I am not measuring irradiance but rather am using the broad angle that the light illuminates my subject to soften the shadows. With the geometry of my setup, this angle is roughly 40°.

As to angular measure, it is not commonly used in this context but it's the same metric as is used in the reference. The reference uses the ratio of size and distance which is an angle, albeit in radians. So we could say that 4ft wide at 5ft away subtends 0.8 rad = 46 deg. Which means that the reference remains relevant to the issue at hand.

The sun at ½° would definitely be a point source.

A bit of a red herring, André. You could move something hundreds of feet towards or away from the sun and the illuminance would not change a bit, as I suspect you know already.

[DanK]

The proof is in the pudding, as they say on this side of the pond. Take a subject like that one and photograph it with and without a diffuser, keeping the average exposure about the same. The effect can be huge. And as Manfred and Ted pointed out, the diffuser can be just something lying around the house. I recommended the Westcott diffuser because it packs into a tiny pack that you can easy carry along, but at home, my diffuser for the halogen lights I use for most flower macros is just two sheets of baking parchment paper.

[Round Tuit]

Ted & Dan,
I am beginning to think that my communication skills may not be as good as I though they were. Both of you missed a point that I was trying to make on two different topics. Ted on the window being a diffused light source and Dan on the use of a scrim as suggested by Manfred. So let me try to be more specific.
When I mentionned the window, I was addressing Dan's comment that my lighting was harsh. I outlined the geomety of my set up to justify my conclusion that the problem lay elsewhere since the diffused light from a relatively close by north facing window should not have cause the flaws in my photo. The technical definition of a diffused source as opposed to a point source was definitely not something that entered my mind. Ted brought up the illuminance measurements aspect of these sources and on this topic, I will gladly defer to his expertise.
I must add that I am EXTREMELY grateful to Ted for uncovering the blown pixels in post #6 and to Dan for digging up the explanation for it. That is the root cause of the problem with this stack and every other stack that I have ever made. That is also the reason that I have never seen much difference between the PMax and DMax algorithms. The first step in my workflow when I stack is to set up the exposure so that the brightest channel,usually the red, is as far to the right as possible without blowing it out. The old ETTR! It turns out that this step blows out the hightlights in every stack. Back up the exposure by half a stop and the problem is solved.
As for the use of scrims, I shoud have made it clear that I was refering specifically to Manfred's suggestion to put a shower curtain in front of the north facing window. I do know that scrims can be very useful.

André
P.S. I like the frog!