Fantastic video, I wish this was around when I was learning to use this! I'd add a few additional tips to save new users some headaches with calibration of the detector alignment. First, if your sample is sensitive to photobleaching/phototoxicity, try to avoid aligning the detector using the same cells/region that you plan to image. Continuous scanning can cause photobleaching pretty quickly in some samples. To avoid this, you can calibrate the alignment in one area, and then acquire data in a new region with new cells. Second, if your signal is dim, or the scan area contains lots of dark/empty space, you should not attempt to calibrate the alignment with it. If you do, the detector may actually drive itself farther out of alignment in a way that can be surprisingly difficult to fix. When the alignment gets too far off, you won't be able to detect any signal, even in bright samples, so the automatic re-alignment won't work (this happened in my lab, and it required a service visit from the Zeiss technicians to get it realigned. For some reason, manual realignment also wasn't working). This can happen even if the "weak or sparse samples" box is checked. If your sample is sufficiently bright but there are large areas of dark space between small labelled structures, zoom in to an area with a higher density of labelled structures to reduce the amount of empty space before calibrating. If the labelled structures are too dim, small, or sparse, you should calibrate the detector alignment using a separate standard. Our lab now keeps the "adjust in continuous scans" box unchecked to prevent accidental misalignment of the detector.
Hey Collin - thanks for the additional tips! There are certainly a number of scenarios that can get you into trouble - and little/no signal would be at the top of the list. I do have a few comments on your suggestions that hopefully other users will find helpful. I would caution calibrating the array with a separate standard. Unless you are using high precision coverslips (including measuring their thickness) and imaging directly at the coverslip/media interface, there is too much variability from sample to sample to rely on a single alignment from a separate sample. It might get you close - but you'll still need some fine tuning on your specific sample for the best accuracy. This will be especially important in low SNR images. Leaving the 'adjust in continuous scans' box unchecked forces users to manually adjust the alignment. As a core facility - we generally discourage unchecking this box as scenarios where the automatic alignment fails are the exception rather than the rule. The only way to activate an alignment leaving this box checked would be to use the 'Continuous' scan button - which we proactively discourage in favor of Live anyway (mostly for speed purposes). Of course there are many nuanced scenarios that are outside the purview of these types of videos and are best covered during in-person trainings.
I wish I had also read your comment in addition to watching this (superb) video. I had a dim sample and screwed up the detector alignment to the extent of wasting my first session. I subsequently used very high laser power on part of the sample I wasn't going to image and got the alignment and then could move on to the part of the sample I used for my image.
Thanks Andreas! The stripes are an artifact of how the preview image is calculated. The preview image is just a sum without any pixel shifting (a fast calculation that can be updated quickly), and in the FAST mode, the illumination profile is not perfectly homogeneous over all 4 pixels. If you look close you can see this alternating 2 pixel bright/dark lines in all channels - you can just pick it up easier in the 405 channel here because the DAPI is more uniform. The pixel reassignment during processing corrects for this.
I keep coming back to this great video & recommending it to others.
Thanks a lot for sharing your valuable insights!
Great tutorial! very knowledgeable and in details
great video, thank you so much! Looking forward to next vides
Fantastic video, I wish this was around when I was learning to use this! I'd add a few additional tips to save new users some headaches with calibration of the detector alignment. First, if your sample is sensitive to photobleaching/phototoxicity, try to avoid aligning the detector using the same cells/region that you plan to image. Continuous scanning can cause photobleaching pretty quickly in some samples. To avoid this, you can calibrate the alignment in one area, and then acquire data in a new region with new cells.
Second, if your signal is dim, or the scan area contains lots of dark/empty space, you should not attempt to calibrate the alignment with it. If you do, the detector may actually drive itself farther out of alignment in a way that can be surprisingly difficult to fix. When the alignment gets too far off, you won't be able to detect any signal, even in bright samples, so the automatic re-alignment won't work (this happened in my lab, and it required a service visit from the Zeiss technicians to get it realigned. For some reason, manual realignment also wasn't working). This can happen even if the "weak or sparse samples" box is checked. If your sample is sufficiently bright but there are large areas of dark space between small labelled structures, zoom in to an area with a higher density of labelled structures to reduce the amount of empty space before calibrating. If the labelled structures are too dim, small, or sparse, you should calibrate the detector alignment using a separate standard. Our lab now keeps the "adjust in continuous scans" box unchecked to prevent accidental misalignment of the detector.
Hey Collin - thanks for the additional tips! There are certainly a number of scenarios that can get you into trouble - and little/no signal would be at the top of the list. I do have a few comments on your suggestions that hopefully other users will find helpful.
I would caution calibrating the array with a separate standard. Unless you are using high precision coverslips (including measuring their thickness) and imaging directly at the coverslip/media interface, there is too much variability from sample to sample to rely on a single alignment from a separate sample. It might get you close - but you'll still need some fine tuning on your specific sample for the best accuracy. This will be especially important in low SNR images.
Leaving the 'adjust in continuous scans' box unchecked forces users to manually adjust the alignment. As a core facility - we generally discourage unchecking this box as scenarios where the automatic alignment fails are the exception rather than the rule. The only way to activate an alignment leaving this box checked would be to use the 'Continuous' scan button - which we proactively discourage in favor of Live anyway (mostly for speed purposes).
Of course there are many nuanced scenarios that are outside the purview of these types of videos and are best covered during in-person trainings.
I wish I had also read your comment in addition to watching this (superb) video. I had a dim sample and screwed up the detector alignment to the extent of wasting my first session. I subsequently used very high laser power on part of the sample I wasn't going to image and got the alignment and then could move on to the part of the sample I used for my image.
such an excellent resource !
Thanks for helping me understand the Airyscan principle.
BCM is amazing! Thanks for this video!
Thanks so much for this superb video. It has been an enormous help, especially with covid19 restricting in person training.
Good stuff. Thanks a lot!
Great video, thanks a lot. What's the reason for the stripes in the 405 channel for the airy fast acquisition?
Thanks Andreas! The stripes are an artifact of how the preview image is calculated. The preview image is just a sum without any pixel shifting (a fast calculation that can be updated quickly), and in the FAST mode, the illumination profile is not perfectly homogeneous over all 4 pixels. If you look close you can see this alternating 2 pixel bright/dark lines in all channels - you can just pick it up easier in the 405 channel here because the DAPI is more uniform. The pixel reassignment during processing corrects for this.
@@OiVM
Thanks a lot for the answer. That makes sense