Dear Agistoft.

I have a very ambitious project where we sent plant roots to space aboard Polaris Dawn and are 3D modeling the root structures. We were very limited by the containers we used. They had to be plastic, square and frozen upon return prior to gene sequencing. I have over 118 models I need to generate.

I'm getting good alignments of the cuvette containers (Image 1). However, the root structures are being misaligned because of optical refraction when photographed through a planar surface from an oblique angle. I have to proceed with a dense point cloud method because the stereo method will not work because of this refraction problem (Image 2).

Image 1.

/wp-content/uploads/2024/12/screenshot-2024-12-20-at-115820am.png

Image 2.

/wp-content/uploads/2024/12/screenshot-2024-12-20-at-115833am.png

After a quick crop to isolate the root structure. I now need to proceed with a method that allows me to combine the four prominent structures into one accurate and unifying structure (Image 3). I have tried markers, but because each structure to twist. Im also batch processing over a dozen samples for my initial alignment, so I want to keep my workspace (Image 4) fairly clean without having too many duplicated chunks. How do you advise I proceed. What steps should I take to achieve one unified root structure considering the challenges I am confronted with?

Image 3.

/wp-content/uploads/2024/12/screenshot-2024-12-20-at-120040pm.png
Image 4.

/wp-content/uploads/2024/12/screenshot-2024-12-20-at-120051pm.png

I did a practice run over a year ago and published this here: https://www.youtube.com/watch?v=1zHiX2zy6a4

Some models to come together very nicely on their own, the majority of them do not.

You can see an example of a dense point cloud that merged exceptionally well below at Image 5.

Image 5.

/wp-content/uploads/2024/12/screenshot-2024-12-20-at-121610pm-1.png

I'm running into an issue where, about a quarter to halfway through the image stack (so 90* - 180* around from the start), my x-axis "start" is starting in the middle of my specimen.

I'm operating from an assumption that when I link the X and Y axes, as the stage turns and the specimen rotates about, the linking takes into account how the specimen might have moved and corrects for that displacement. But that's an assumption and I don't want to make an @$$ of myself so I'm open to being corrected on what it means to link the two axes.

Is the solution to cheat my start and stop by 10+ extra steps to account for how the specimen may move? Thanks in advance!

This video records a conversation with Dr. Michael Branstetter, where he and Mark discuss macro photography tips related to the Macropod PRO 3D. Mark and Michael talk about lighting, photoshop techniques, field use, basic camera settings, contrasting backgrounds and imaging specimens preserved in a liquid.

00:00 Intro
01:30 Godox Flash Connections
04:30 Lighting and Backgrounds
10:30 Boosting Contrast
17:00 Scale bars and Settings
23:30 The Macropod in the Field
25:45 Explaining Exposure, ISO, Aperture and Flash Fractions
38:00 100mm Macro or MP-E 65mm
40:00 Image Capture in the Field
44:00 Staging Specimens in the Field
55:00 Homemade or Bought Diffusers
58:00 Talking Turtles
1:04:15 Aquatic Samples

https://youtu.be/vhVrxTmIr-c

How to sync channels and use the Godox MF12 flash unit on the Macropod PRO 3D.
https://youtu.be/g8WEK4OF_w8

A quick video showing users how to convert the Macropod PRO 3D from the rotary stage to the translational stage.
https://youtu.be/MCRpdsA8Kms

Need a license? Click here: /product-category/downloads/software/

This is a complete walkthrough video about 3D modeling a plant roots, Arabidopsis) using Macroscopic Solutions Macropod PRO 3D focus stacking and photogrammetry system. The specimen is cultured in a plastic cuvette in agar and is the test subject for experiments to be conducted in zero gravity. Mark details methods to efficiently photograph, batch focus stack, post process and assemble the 3D model using Agisoft Metashape. Mark also demonstrates how to add a scalebar to the model.
https://youtu.be/1zHiX2zy6a4

Need a license? Click here: /product-category/downloads/software/

This is a complete walkthrough video about 3D modeling a fossilized jaw bone using Macroscopic Solutions Macropod PRO 3D focus stacking and photogrammetry system. The specimen is highly reflective, difficult to contrast and oddly shaped. See how we efficiently photograph, batch focus stack. post process and assemble our 3D model using Agisoft Metashape. Mark also demonstrates a few common masking techniques and will show you how to add a scalebar to your model.

See us 3D model something with a SIGNIFICANTLY HIGHER DEGREE OF DIFFICULTY: https://www.youtube.com/watch?v=1zHiX2zy6a4&t=48s

https://youtu.be/wznsRLOpSiw

This is a multi-step tutorial designed to teach you how to use key methods and strategies in Agisoft Metashape.
https://youtu.be/H6sN7iyPAzo

This video demonstrates how images may be batch edited, cropped or altered using photoshop prior to importing final images into Agisoft Metashape.
https://youtu.be/GRfsgiWQST8

2:1 Batch Processing

This tutorial explains how to batch process final images for focus stacking using Folder Axe and Zerene Stacker.

Note* Prepare Agisoft Metashape Professional to use Depth Maps using the following instructions:

Open Metashape Pro - Tolls - Preferences - create the following Tweak and set its value to False:

main/mesh_tvl1
Then use the Depth Maps as a source for the Build Mesh operation.
https://youtu.be/teMBpnCD6dQ

Macroscopic Solutions offers a step-by-step tutorial designed to help users capture and process their first few images using the Macropod PRO 3D

The Universal Stage for this system is now available at: /product/universal-stage-gimbal-for-manual-micro-macro-3d-imaging/

https://youtu.be/Vo5carREHyA

View: https://zenodo.org/records/10594883
Three dimensional models of insect specimens offer the potential for a variety of research applications, including taking morphological and trait measurements of insects and automated species identification. Yet high-quality 3d models are challenging to create because of the specimens’ small size. Here we overcome this challenge by creating 3d models using focal stacking to take high-resolution images of bee specimens. Sets of 3d images are processed using Zerene Stacker and Photoshop, and the 3d models are created using Agisoft Metashape. Our protocol can be used for both larger insect specimens, like bumble bees and carpenter bees, and small- and medium-sized specimens, such as mason and sweat bees. This protocol is being used by members of the Cheadle Center for Biodiversity and Ecological Restoration at the University of California, Santa Barbara, and is based on a protocol developed by members of the Big-Bee project.

Download: /wp-content/uploads/2024/02/Workflow-TPT_Pinned-Specimen-and-Label-Photography-Workflow-Template.pdf

This guide describes a basic image capture workflow for pinned, wet and slide mounted specimens using
the Macropod Pro system. Images of labels should be taken with a point and shoot or smartphone
camera, instructions included. Please review Macroscopic Solution’s Macropod Pro Manual and video
tutorials (/tutorials/) to acquaint yourself with the proper assembly,
operation, and storage of the Macropod equipment prior to continuing with this guide.

Macroscopic Solutions offers a brief walkthrough to navigate the software icons, workflows, manuals and directories associated with each Macropod PRO 3D.

https://youtu.be/z-ooSy3ITPs

This tutorial describes the proper method to configure the Macropod PRO 3D Imaging System.

The Universal Stage for this system is now available at: /product/universal-stage-gimbal-for-manual-micro-macro-3d-imaging/

https://youtu.be/Gj5OQeuuY50

Video tutorial describing the most important components included with the Macropod 3D system.
https://youtu.be/Ky8c3Vm-pfI

I want to clarify the difference between the 1) turtledove diffusers - for either the MT-24EX or the MT-26EX-RT flashes and the 2) MP-E 65mm diffuser or Mitutoyo diffusers.

The turtledove diffusers are for field use and are to be used when the flash needs to be pointed directly at the specimen your imaging. They do not perform as well as the MP-E65mm or Mitutoyo diffusers; however, they are very good in comparison with other off the shelf macro diffusers. They effectively eliminate glare without blocking too much light, they perform even better when used with a polarizer.

One of my clients who has used these diffusers very effectively shares ample imagery on his instagram account here: https://www.instagram.com/venom_macro/

Video in use: https://www.youtube.com/watch?v=DBWNvW29g-0

The MP-E and Mitutoyo diffusers are intended to be used as a reflected light source. That is, when the flash heads are NOT pointing directly at the subject you're shooting. Instead, you'll be posting your flash heads at the diffuser, which will then reflect onto the subject you're capturing. This is hands down the best illumination method to use.

Most of our images are captured using these diffusers and are shared here: https://www.flickr.com/photos/107963674@N07/

Video in use: https://www.youtube.com/watch?v=w_SB6AD7wdo

The diffusers are not used in tandem, only one or the other. Turtledove for field, others are for lab use. Thanks again for choosing Macroscopic Solutions!

I have a reoccurring problem when acquiring large stacks, hopefully some here might be able offer some advice. I am a newbie to this system, so I haven't a clue what I might be doing wrong. Any advice would be greatly appreciated.

When shooting stacks at high magnification I keep getting slices where the flash (Twin Lite MT-26EX-RT) does not fire. These black frames only occur towards the end of the stack suggesting it might be a power issue, or maybe the flash is getting too hot? The stacks are usually >100 slices where this is a real problem.
Has anyone else had this kind of problem?

System Details

Flash:
Canon Twin Lite MT-26EX-TR using 4 x Panasonic Eneloop Pro rechargeable batteries internally and the Quantum external AC adaptor. I have also used the external battery pack (CP-E4N) and had the same issue.

Flash settings:
Manual
Second-curtain sync
Flash head A:1/4
Flash head B 1/16

Camera:
Canon EOS 6D MK2
EF 70-200mm lens
M Plan Apo 10x

Shooting is controlled by the stackshot software through Zerene. Settling period of 3 seconds between slices.

We have tried the battery pack for the flash but had more success than with the Quantum external AC adaptor. The problem might be with the four batteries that are in the flash. I have tried both rechargeable and single use batteries and again, the problem gets worse after two or three shoots.

I have been trying out slabbing in Zerene and it really does make a big difference on “hairy” insects. Is there a way of automatically doing this on the acquisition side? This could be a way of getting around the power problem.

For example, is there a way of importing a script into the stackshot controller built into Zerene that will
1) split your stack into X number of sub-stacks
2) set a delay of Y minutes between sub-stacks
3) give files from each sub-stack a unique identifier to make focus stacking easier in Zerene.

On the subject of stackshot, it is very buggy when run through Zerene. The standalone version of the stackshot software is much more stable. The bugs also seem to be manifesting on the hardware too, the stage makes a high-pitched whistling noise when using the stackshot controller through Zerene, but does not when using the stand alone stackshot controller software. Have you had any experience of this?

There are two ways to add scale bars to 2D images.

Quick Note: Note the magnification you are using on the MP-E 65mm 1-5x to apply scale bars within that range. When using the MP-E 65mm lens, ensure the yellow xX line is just slightly, but completely revealed in order to properly match a scale bar (see photo below).

As a helpful reminder, I always insert the magnification using not the filename.

Method 1: The first method of applying a scale bar is to create the file yourself using photoshop. This method has a few nice perks:

perk 1) Have confidence your scale bar is correct
perk 2) You can shoot at any magnification you would like as long as you have a unit of measurement in the background
perk 3) The scale bars are forever available in photoshop for future use.

Macroscopic Solutions calibrates scale bars using the TedPella 5mm ruler: /product/5mm-scale-027-calibrated-to-0-1mm-5mm-range/#product_images[grouped]/0/

Step 1: Insert an image into photoshop with a unit of measurement in it. I'm going to use an image where a scale bar has already been applied on an image taken at 5x with the MP-E65mm. You can pretend the scale bar is a ruler or measurement grid in the background.

Step 2: Use the ruler tool to measure the length of the line (1mm) in pixels. The ruler tool can be accessed by right clicking the eye dropper tool (figure 1).

Step 3: Once selected, click and drag the ruler tool over the length of known measurement. In this case, the line indicates 1mm. The length in pixels is revealed in the toolbar as L1 (circled in red in figure 2). In this case, there are precisely 765.38 pixels in 1 mm.

Step 4: Go to "Analysis" > "Set Measurement Scale" > "Custom"
a) Insert the "pixel length" (765) rounded
b) Insert the "logical length" (1) representing no. of millimeters
c) Insert the "logical units" (mm) representing units (could be micrometer, millimeter, etc.)
d) Click "Save Preset" and name a unique identifier so that you can avoid these steps in the future. I've named this one "Forum Test." Then press "OK" and "OK." (figure 3)
Note: If the ruler you have is 5 cm... you can find the pixel length of the 5 cm rule and enter 5 cm.

Step 5. You will now see the calibrated scale bar in the list under "set measurement scale." Please make sure the scale bar you just created is checkmarked.

Note* In figure 4. you will see a list of scale bars made by me at each level of magnification I use specifically with the Canon 6D (5472x3648 pixel resolution). If you change camera bodies or use a different pixel resolution, you will need to repeat steps 1 - 4. Please ask questions below if you need me to explain further.

Step 6. Apply scale bar by going to "Analysis" and "Place Scale Marker." Choose your color, font size and font. Click "OK". (figure 5).

Step 7. A new scale bar will emerge as a layer that you can further manipulate if you wish (figure 6)

Method 2: This method is nice because you do not have to worry about changes in resolution, you simply have to match the size of the .psd overlay file to the overall size of the image.

The scale bars and an instructional video can be downloaded and found using the following link.: /product/psd-overlay-files-for-mp-e-65mm-1-5x-lens/

I you are a client of Macroscopic Solutions, contact us and we will provide you with these .PSD overlays without charge.

When using the 50x Mitutoyo objective on the Macropod setup, I find that when looking through the stack in Zerene, the images seem to jump around a bit. Everything is tightened down and the specimen (in this case, a piece of Baltic amber with two tiny wasp specimens) is embedded firmly in a piece of modeling clay. Locking up the mirror on the camera does not seem to help. Finally, if I click the "Show as adjusted" box in Zerene, which I assume shows the images after aligning, they still move slightly from image to image. I am not sure this is really causing lack of detail in the stacked image, but I expect it can't be good.

Anybody else having this problem?

Jim Woolley

The dreaded donut is a term coined by Dr. Jim Woolley of Texas A&M Entomology. It defines the effect caused by intense light reflecting from the face of a white diffuser. The diffuser is necessary for creating diffuse light, but can generate unwanted glare to reflect off of any glossy subject being photographed.

Images taken at the National Natural History Museum at the Smithsonian Institute. Samples provided by Dr. Lourdes Chamorro and Dr. Matt Buffington.

The image above shows the reflection clearly in the abdomen, thorax and head. Some research professionals such as Dr. Matthew Buffington believe that this reflection is critical for showing the reflective characteristics of the specimen; however, others would prefer to show textural detail of these parts instead of a prominent reflection.

There are four proposed solutions to this problem.

1) The first would be to reconfigure the diffuser so that it sits above or slightly behind the subject. The best way to do this is to utilize the Jansjo light that is becoming increasingly popular among scientists as a suitable replacement to Fiber Optic light boxes. The lights costs less than $20 on Amazon. Important, do not have the light on when capturing a set of images. Instead, use the flash and diffuser only. The light is a great tool for initial positioning of the specimen.

Figure A.

Change from:

Standard Diffusion Configuration

to:

Reflection Reduction Configuration

2) The second solution would be to post process your images in Photoshop. The shadows/highlight section in image > adjustments will allow you to effectively minimize glare and accentuate detail. For instance, the image shared above was edited using the following steps.

Step 1. Check "show more options"
Step 2. Move Shadows to 0%
Step 3. Move Highlights to 14%
Step 4. Move Highlights "Radius" to 2500px
Step 5. Move "Midtone Contrast" to +30

Lastly, use curves to further remove reflection by changing the following.

Figure B.

Change from:

Default

To:

Adjusted

Result:

Photoshop Edit

3) The last suggestion would be to increase the magnification. The added magnification is by far and away the most effective method of displaying detail as shown bu the following figure. The more magnification, the more detail you can resolve. You can see the "dreaded donut" is seen prominently on 20x, but becomes further and further out of view towards 50 and 100x.

Figure C.

10, 20, 50 and 100x objective viewpoints of shiny Tiger Beetle

4) The last solution would be to fully immerse your subject into something white. Leaving only a small opening large enough to resolve the subject you are looking at. This ball bearing was focus stacked inside a white cylindrical container leaving only the top exposed. It reflects prominently; however, to show exterior detail only>

Figure D.

Ball Bearing shows compete cylindrical diffusion and reflects only the glass omn the objective at the very center. This is the best way to achieve a high level of textural information >95%.