A Detailed Introduction to the Basic Operation of SerialEM.ppt

《A Detailed Introduction to the Basic Operation of SerialEM.ppt》由会员分享，可在线阅读，更多相关《A Detailed Introduction to the Basic Operation of SerialEM.ppt（28页珍藏版）》请在麦多课文档分享上搜索。

1、A Detailed Introduction to the Basic Operation of SerialEM,Image Buffers in SerialEM,Images are saved in buffers, referred to as A NCamera acquisitions and processed images go into buffer A, and the top few images are rolled into higher buffersTypical situations (enforced during tilt series) are:,A

2、B C D E G H ,New image -,Align reference,Read-in image -,(Home)(Insert) (Delete)(End),Normal operation,New image -,Record reference,Read-in image -,(Home)(Insert) (Delete)(End),Low dose mode,A B C D E F G H ,Trial reference,Image Display Panel,The available data range in the displayed image is contr

3、olled by the Black and White levels: everything outside this range is truncated To determine Black and White, it analyzes pixels in a central area and truncates a certain fraction Contrast and brightness are of limited usefulness because of this truncation Most useful option is to set “Area Fraction

4、” to smaller value, like spot metering on smaller central area To see contrast at extreme end of range, enter image minimum as Black value or image maximum as White value Zoom pretty much takes care of itself with default settings,Microscope Status Panel,Screen current is more useful than exposure t

5、ime; “Float” button floats a screen meter, good when tuning scope “Dose” button floats a cumulative dose meter The Defocus is based on a microscope readout, not the actual focus on the specimen On Tecnai, it matches TUI defocus reading, can be reset in TUI or SerialEM On JEOL, it is arbitrary and se

6、t to zero on program startup, can be reset IS is image shift in microns, important to know on 300KV Tecnai and JEOL 2100 without high power image shift,Tilt Control Panel,The delay time is the delay after tilting by basic increment delay will be longer or shorter for bigger or smaller tilts,Camera a

7、nd Macro Tools,Buttons for camera parameter setup and 5 camera acquisition modes (good for low dose) STOP is the one general stop button for all kinds of operations “Preview” morphs into “Montage” when montaging, unless low dose is on Macro buttons also morph into useful buttons during tilt series,C

8、amera Parameter Setup,Continuous acquisition is only mildly useful (no access to fast camera modes) Processing options useful for testing and troubleshooting; otherwise always use gain normalization Binning is used for speed. Tracking correlations are always binned down to 512 so there is no point t

9、aking Trials bigger than 512. A full-width image gives fastest readout per pixel; hence the wide options for fast subarea pictures A separate dark reference is kept for every binning and exposure time. “Force new dark reference next time” is the way to remove a dark reference Averaging dark referenc

10、es is helpful for a high-noise camera (not typical),Pre-exposure,Pre-exposure (drift settling) requires two shutters The Gatan shuttering mode that allows pre-exposure is inflexible and exposes specimen during readout so SerialEM uses beam blanking to provide selected pre-exposure while blanking dur

11、ing readout “Dual shuttering minimum exposure” mode provides for this and just does beam blanking when drift settling is 0 so it can always be used But the camera timing must be calibrated so the beam blanking does not occur during exposure time Tietz cameras require Shutterbox II for pre-exposure t

12、o work,Image Alignment & Focus,Autoalign always changes microscope image shift “Clear alignment” undoes that “Reset Image Shift” moves the stage to keep the same specimen point centered Moving stage for mouse shifts avoids having to reset image shift Happens automatically if option selected and move

13、ment is above either absolute (micron) or relative (fraction of field) threshold You can do it intentionally with Shift key Trimming dark borders keeps autoalign from being thrown off by beam edge or grid bar Used automatically in low dose and during “tasks” done at lower than working mag Will not d

14、etect less extreme contrast differences that can throw off autoalign alternative is to use smaller area (higher trim fraction) “Center image shift on tilt axis” will give fewer movements during tilt series but requires an offset calibration,Autofocusing with Beam Tilt,Autofocus is done by taking two

15、 images with the beam tilted in opposite directions and measuring the shift between imagesIt relies on the microscope being aligned such that the image of a specimen in the focal plane of the objective lens does not move when the beam is tilted.,A,Specimen in focal plane,Image plane,A appears here w

16、ith both directions of tilt,B,Specimen out of focus,Image plane,B or C appears here depending on direction of tilt,C,Focus Parameters and Operations,The target is the defocus it sets to after measuring The beam tilt needs to be big enough to get good image displacements, but is limited by objective

17、aperture Autofocus Offset lets you measure defocus at focus different from target Can protect against correlation failures by focusing at zero Can be used to focus at a defocus with more contrast Taking three pictures lets it subtract off drift “Drift protection” “Report shift and drift” takes the t

18、hree pictures and just tells you the beam-induced shift and the drift “Report on existing” reports the shift and drift from pictures just acquired “Autofocus” measures and changes to target; “Measure defocus” just measures without changing Autofocus sometimes works poorly at high tilt - “Check autof

19、ocus” assesses what fraction of a known focus change can be detected “Move Focus Center” lets you click on spot at desired focus and change focus so center of field is at that focus,File Operations,You can open new image files or existing files (up to 8 now) Opening a new file brings up File Propert

20、ies dialog, where you set the data type and items to save in header Anything selected for saving can be extracted with extracttilts (options -st, -ma, or -in for stage position, mag, or intensity) After opening an old file, the “Set” options can be used to change the properties that would be set in

21、the File Properties dialog “Save to Other” and “Read from Other” can be used to save one image or read from a file without leaving it open You can save an image to a specific Z value with “Overwrite” Saving a log is useful. “Read & Append” lets you accumulate a log across several runs of the program

22、,Camera Commands,The most commonly used item in the camera menu is “Prepare Gain Reference” Gain reference can be done at typical exposure for non-low dose work Having lots of counts averaged (e.g., 10 x 1/3 of dynamic range) is more important than matching the counts for actual images Binned gain r

23、eference works for normalizing images with even binning useful for slow 4K camera that is being used as 2K camera If camera counts per electron is calibrated, this can do automatic dose calibration It can be useful for troubleshooting to view gain reference or dark reference of image just taken “Pos

24、t-actions” controls whether image shift, mag change, or stage move are taken during camera readout Requires correct camera timing parameters and with that, it is surprisingly robust But it can be turned off to troubleshoot problems,Montaging The Camera Is No Limit,To start montaging, define the numb

25、er of frames in X and Y in Montage Setup dialog Parameters start at current state, based on current Record parameters The default overlap (10%) is fine for image-shift based montage Once data has been taken, or if you open an existing montage, most parameters in the dialog are locked After taking a

26、montage, you will see “overview” in buffer B, a subsampled version of the whole area A (composed) image of the center is left in A for alignment Either image can be mouse-shifted “Prescan” takes a fast highly binned montage of the same area without storing it Options are mostly important when using

27、montages as navigation maps, except: “Change focus with height” will take each picture at same focus, but this may make it harder to correct for distortions due to magnification gradients,Image Processing,In general, processing works on the active image, the processed image is put in A, and the top

28、images are rolled Binned FFT bins the image to 512 pixels It is faster, less noisy but only shows central part of frequency space Probably most useful for assessing drift in image Live FFT will do FFT immediately in Continuous acquisition mode (version 2.6.4?) X-ray items are to help set up removal

29、of X-rays in dark references and images X-ray removal from images should only be done for low-dose images from low gain cameras where X-rays screw up tracking “Show cross-corr” will do an autoalignment with current A and align buffer images, then show correlation,Tasks,Tasks are (mostly) operations

30、required for doing tilt series, involving a sequence of stage movements and tracking images Eucentricity has two different routines Rough eucentricity will work from 100 m away or more in Z Refine eucentricity does a 24 tilt series to measure Z height with lateral offset of tilt axis taken into acco

31、unt and estimated also Works up to 10 m away or more “Refine & Realign” runs it in a way that restores the image position at the end Can use result to “Set Tilt Axis Offset”,Tilt axis,Optic axis,Specimen feature,+24,-24,0,Axis offset,Z,Tasks #2,Walk Up goes from one tilt to another while keeping an

32、image feature centered “Anchor” is an image at intermediate tilt that can be a reference when coming back down in a tilt series “Reset IS & Realign” and “Reverse Tilt” keep an image feature centered when resetting image shift or reversing direction of tilt “Set Intensity” lets you change intensity b

33、ased on the counts in the current image Enter a factor to change intensity by (e.g., 0.75, 1.5) Or enter desired counts in a Record image (e.g., 4000) “Move Beam” - if you click at the center of the beam, this will move the beam to the center of the field,+,The Problem of Tilt Series Acquisition,The

34、 goal in acquiring a tilt series is to have every picture centered on the same specimen feature, and at the same focusThe basic sequence of operations required is thus:Tilt to a new tilt angleCorrect change in specimen positionCorrect change in specimen focusAcquire final (“Record”) imageFor full au

35、tomation, the desired features must stay centered during initiation and after interruptions of the series the tasks are components for ensuring this,Approaches to Automated Tomography,Traditional approach: track and focus at every tilt Used in original Agard/Koster software and in Tietz softwareMay

36、have to track at low mag: time-consumingPrecalibration: take a coarse tilt series to determine X, Y, and focus, then take full tilt series without tracking or focusingUsed in new Koster software for Tecnai and in FEI softwareCan be fast but is susceptible to drift and calibrations being off; hard to

37、 accommodate user interventionZ-prediction: calibrate tilt axis location and assume specimen moves in a circle around it (new Agard method) - Very fast, but many scopes do not meet the assumptionsRobust prediction: predict X/Y/Z position on next tilt from changes in position on previous tiltsMust ad

38、apt to non-ideal conditions and user interventionsGoal is to achieve the reliability of traditional approach and speed of precalibration,The Robust Prediction Method in SerialEM,Tilt 2,Y,Tilt 3,Y,Tilt 4,Y,Tilt 5,Y,Make and use prediction but track,Make and use prediction but track,Rely on prediction

39、 do not track,Actual position,Predicted position,Tilt Series Preliminaries,Adjusting beam and exposure parameters Tradeoff between specimen damage and signal-to-noise ratio in images and reconstruction For plastic section work, additional constraints of dynamic range of camera and difficulty of gett

40、ing drift-free high-exposure image Whether to let program start at zero or go to high tilt “manually” “Manually” means you refine eucentricity and run walk-up You would do this if You are not sure you can get to the highest tilt angle You may need to adjust drift settling at high tilt You are unsure

41、 of what counts you can reach at high tilt Pre-expose plastic sections to minimize shrinkage and warping during series - 2000 electrons/2,Tilt Series Setup the Mother of All Dialogs,Tilt and focus parameters are the same as the ones exposed elsewhere there is just one underlying parameter Some choic

42、es will be set/enabled depending on whether you are at zero or already at high tilt Low mag tracking needed only if field of view = 2 m field (all assuming well-behaved stage),Tilt Series Setup 2 - Intensity Control,Intensity control is important because otherwise higher tilt images will have lower

43、SNR Intensity control is complicated because it covers several situations; You might want constant intensity if doing cosine tilting 1/cosine or 1/cosine to a power is preferred for cryo Use a target # of counts for stained material: 1/cosine can give big variations in counts The problem with consta

44、nt counts is that it may require too bright a beam at high tilt. Two solutions: Taper counts down above an angle OK once you are familiar with usable values Use with “Do not increase intensity above value for first saved image” to protect against overconstricted beam at far end of series Set up inte

45、nsity before starting and use “Keep intensity below current value” Preferred method if you are starting at high tilt and setting up beam there Can be used even if starting at zero tilt as long as images arent too bright there,Tilt Series Setup 3 Autofocus,The autofocus interval sets how often it wil

46、l focus anyway when predictions are reliable 5-6 is good value Can focus every time above an angle not really needed at end of series (should have an option to do this only at start) Checking autofocus is designed to prevent runaway focus changes if autofocus doesnt work well enough at high tilt You

47、 can follow the advice, i.e. adjust focus manually, and set defocus target to measured defocus value so it will stop trying to change focus Using an autofocus offset that is the negative of the target defocus might also work The (rare) problem usually goes away by 50,Tilt Series Setup 4 Initial Acti

48、ons,It is good practice to have a centered image before you go into the dialog, then select the option to align to it You should refine eucentricity before a series, especially after pre-irradiating the specimen Rationale for anchor is that tracking is more likely to drift off at high tilt and is li

49、kely to be accurate below 45-50 Aligning to an image taken during walkup at such an angle will result in desired area being centered at low tilt There is no data loss from the misalignments at higher tilt, since the drift is lateral to the tilt axis and lost image area is not needed for backprojection You can have it take the anchor when it walks up, or you can to “Walkup & Anchor” before starting series and then select the anchor there,Tilt Series Setup 5 Tracking Control,