Alignment of the 2D stacks images

Hi all,

I am currently working on measuring the 3D PSF(stack several 2D images on different focal plane) of the confocal microscope by using PZT stage to scan the sample. However, it seems that there are shifting between image and image(see following GIF). To be specific, the psf pattern should be in the same position as I change the Z axes position. However, in my result , the psf pattern is running around . I already checked the servo of the PZT stage, it is working normally. Has anyone encounter this kind of problem before? Or is there a proper way to solve this problem?

GIF of the 2D stacks PSF

My equipment are listed below:
-0.1 micrometer latex beads as a specimen
-473nm light source
-100x,0.8NA Air lens
-Piezo stage to scan the specimen
-Single-photon avalanche diode as the detector

Thank you all for reading.


I am not sure, but it seems that there is something fundamentally wrong. When focusing through a PSF, you should see a sharp point in the center and diffraction rings above and below the center. In your case, the points appear more or less unchanged in many images and the rings appear only one side (near the end of your movie). Can it be that the beads are very close to the working distance of the objective and that the objective bumps against the specimen rather than focusing through the entire PSF? 100x/NA 0.8 objectives have working distances around 1 mm, so you might accidentally have your sample mounted upside down, imaging through the slide (with a thickness around 1 mm) rather than the cover glass.


Thank you for your comment, however ,I don’t think mounting the sample upside down is the case. I follow the Nature paper to prepare the sample and I am pretty sure that I place the sample correctly. I think the reason why I can see the diffraction ring on only one side is because of the spherical aberration.

See the link below for reference:

Do you spread the beads on the cover glass (and then mount it upside down on a slide)? If so, spherical aberration should be negligible, since the path between cover glass and object is virtually zero. Hence, light rays cannot deviate significantly due to a refractive index mismatch. In any case, you should be able to focus through the entire PSF, starting in darkness (out of focus), passing the central spot of the PSF, and then ending again in darkness (again out of focus, on the other side of the PSF). This is obviously not the case in the movie provided above (the first slices are not dark, but already contain some spots, corresponding more or less to the center of the PSF).


Thank you for your reply again, really appreciate that. My latex bead sample is prepared like the following picture:image

So I am pretty sure that I prepared my sample in the correct way, didn’t place the sample upside down and spherical aberration is not an issue like you just said, what could be the reason that I can only see the Airy pattern on one side? This really confused me. Is there a paper or article discussing about this problem?

Well, as the next step you could try to exclude (or confirm) the Piezo stage as the source of the problem. If you have an eyepiece attached and some widefield light source, just use the eyepiece and focus through your sample with the focus drive of the microscope (not with the Piezo). There should be no xy shift of the beads while focusing and you should see diffraction rings (usually even better as in cofocal mode, as there is no pinhole in the light path blocking out-of-focus light). Some confocals also allow to choose between Piezo focus and internal focus drive of the frame (to repeat this exercise in confocal mode). Then, try to replicate what you have seen with the Piezo.


1 Like

@lightmicroscopy has a great idea to try the motorized z stage if available.

So a lot times I have the experience something is loose. A thread adapter or something that needs tightening? If you capture the PSF several times with the piezo stage, and it is different each time, then I would tend to think that is the case.

In the case of a z-stack, the piezo stage is generally moving quickly. If the stage itself makes a “strange buzzing noise” or moving back and forth after a command signal, and the controller is in closed-loop (servo) mode, that would be more indicative of a problem.

If the stage is directed to a certain position and stopped, but cannot maintain a fixed position with the controller in closed-loop mode also shows a problem (image seems to be jittering slightly, but not enough time for thermal expansion). So that is sort of a another way to check the piezo stage in this case. It would also work for an auto-focus system, where the piezo controller is in open-loop mode but feedback is provided by a laser and sensor, or some image processing to form a closed-loop system. Changing the piezo controller back to closed loop and directing it to a certain position, stopping and checking for “jitters.”

These simple/quick tests may often be done with lower magnifications and some test slides, to help troubleshoot.

1 Like