Calcium signal in stacks

I am a new user of Cellprofile 2.0 and I would like to quantify a calcium dye staining in time-lapse movies. The images have been collected using metafluor and the result is a directory with one image/time point and an .inf file that contains the time infos. The files are NAME.001, NAME.002 etc and to my understand this is just a different format of .tif file. I tried to load them in cellprofiler but the files aren’t recognized. I used imageJ and I made a .tif stack that is recognized by CellProfiler but lacks of any metadata.
A) In a first pipeline I identify the objects (Cells and nuclei) in an image obtained from a projection of the stack (made in ImageJ) and I saved them in two .png files as ojects
B) In the second pipeline I upload the stack as .tif (I cannot group by metadata because I don’t have them) and in a second LoadImages module I upload the two files with the cells and the nuclei (as objects). Everything works fine in the Test run mode but when I run the analysis I get an error message that i cannot load the objects file anymore. I was just wondering if this was due to the fact that I try to apply the objects identified in a single image on a stack (it doesn’t work if I used the track objects either).
C) In a third test pipeline I loaded the stack, made a projection, identified the objects and tracked them in the stack. Here the pipeline repeat all the steps for each of the images of the stack (between 1000 and 4000 images/stack) and the quantification works fine but it takes a really long time. Because the cells don’t move much between frames is it possible to apply the same identified objects to the all stack without recalculating them for each image? Or when processing a stack each image is considered as a separated entity?
Any suggestion will be greatly appreciated!!!
Thanks a lot

It may be that having two LoadImages is not typical for a CellProfiler pipeline. It would be better to use LoadSingleImage to load the object images.

If the cells don’t move much, using the object images would be a good approach. If you are doing something like measuring the per-object intensity, you could simply use these loaded objects directly as input to MeasureObjectIntensity along with the image frame, without needing to perform object detection first.