add v9 spec based upon https://github.com/janelia-cosem/fibsem-tools/… #7
Conversation
specs/v9.tsv
Outdated
| 702 >f4 0 MillingPIDI | ||
| 706 >f4 0 MillingPIDD | ||
| 800 >S30 0 MachineID | ||
| 980 >f4 0 SEMSpecimenI |
There was a problem hiding this comment.
According to the docs Mark shared, JaneliaSciComp/jeiss_fibsem_labview_control#11 (comment) , SEMSpecimenI doesn't exist, or is possibly duplicated with offset 866. They seem to encode the same thing: in your v9 files, are they the same value? If one looks like null bytes (for f4, that looks like it would be 6.41e-10) we can probably trash it. That is, assuming the header is buffered with 0s before values are written into it.
There was a problem hiding this comment.
I spot checked v9 dats for one volume ... SEMSpecimenI at 980 is always 0.0 and SEMSpecimenICurrent at 866 varies with values like -1.2885212898254395. I'm happy to keep or trash the 980 data - I don't use either attribute.
There was a problem hiding this comment.
It sounds like it's not written to (not in the screenshot sent to you either) so probably best to scrap it.
There was a problem hiding this comment.
just committed change that scraps it
| 36 >f4 2,4 Scaling | ||
| 68 >u1 0 Restart | ||
| 69 >u1 0 StageMove | ||
| 70 >i4 0 FirstX |
There was a problem hiding this comment.
Do you know what FirstX and FirstY encode? If they're XY offsets between slices, that would be really helpful for alignment, but I'd expect them to be floats in that case.
There was a problem hiding this comment.
FirstX is the "X coordinate of the first pixel" and FirstY is the "Y coordinate of the first pixel".
They are integral values and are the primary reason for v9.
We asked for them to be added so that we could use them as the starting tile positions for alignment.
In case it helps, Shan emailed this screen shot of the v9 "Tracking File Headers" to Davis and me on January 27, 2022.
There was a problem hiding this comment.
If they're integers, I guess they're in pixel units, i.e. if slice 1 was at 0,0 with pixel size (464) 12nm, then slice 2 at 1,1 would be offset by 12nm, 12nm?
…0 was always 0.0 - so likely unwritten. At @clbarnes request, removed SEMSpecimenI at 980 and renamed SEMSpecimenICurrent at 866 back to SEMSpecimenI.
…blob/main/src/fibsem_tools/io/fibsem.py