The update to version 15.72 incorporates the following changes:
1. There were no new features in this update.
1. The scan function for Image Tools plots (MIT) did not work properly if the lens was AFOCAL. Also, the c/mm scale was incorrect.
2. The third-order pupil aberration DIP was not correct for AFOCAL systems.
The update to version 15.71 incorporates the following changes:
1. The Zemax conversion program ZMC has been modified. Some zmx files come in with a conic constant of zero, but SYNOPSYS treats any surface with a declared CC to be aspheric, which affects many other features, such as graphics rendering. Now it checks, and if the CC is zero, it does not add a CC directive to the RLE file. So the surface remains purely spherical.
2. DSEARCH now allows an AFOCAL directive in the GOALS section. Before, it only assigned that property if you used the current front end and it was declared there. This is more friendly.
3. The program now reorders your wavelengths if you do not observe the protocol: long first, short last. Converting from a different program can come in reversed. The only major difference is the sign of the chromatic aberrations -- but it is s good idea to be consistent everywhere.
4. PAD and DWG now illustrate an immersed object type OBA by drawing the object surface and connecting it to surface 1. That area is presumed filled with the immersion medium.
5. The program that prepares a step file of the lens, STFILE, has been enhanced. Now the native spherical shape and rotationally symmetric edges are all used where appropriate, and we have tested the results on eDrawings and ABViewer. You should not have to use the panel option anymore, which will reduce the size of the step file by almost a factor of 50. The PANELS mode is still available, in case your CAD program stumbles on the new files.
1. There were no bugs reported this update.
The update to version 15.70 incorporates the following changes:
1. The PANT file now accepts input of the form VLIST parameter jsss THRU jsps, which lets you specify a range of surfaces over which the parameter will be varied. This only applies to certain variable types, as you can read in the UM.
2. The input parser now ignores characters not in the expected character set. Some customers have sent us files with strange characters that produce an error message. Now those are changed to blanks, and if the rest of the file is okay, it will work.
3. Switch 91 is now on by default. That switch designates the current glass properties by coloring the elements in the PAD drawing. This is so useful it makes sense for it to be on.
1. FVF did not work properly for object type OBD.
2. The search routines have an option to run in QUIET mode to avoid endless text scrolling in the Command Window. But certain error conditions changed the mode, to make it possible to read the message -- and the mode was not reset on the next case. Now it is.
The update to version 15.69 incorporates the following changes:
1. The Zoom Slider now has a Stop button. Click it and the zoom scan freezes at the current position so you can examine it if you wish.
2. The logic related to the stopsign button has been revised. Before, the program would wait for certain routines to finish before halting -- and some of those reset the condition. Now optimization halts almost immediately if you want to stop it.
3. The default switch setting now places switch 1 OFF. That switch drops bounded variables, which makes optimization run faster since they almost always want to continue violating, which takes time to sort out all over again each pass. But if a TH variable is dropped, then when you run the annealing program, it will ignore edge violations via the AEC monitor, since that only applies to current TH variables, and once dropped they no longer apply.
4. ZSEARCH now has a new ZSPACE ITEMIZE feature that lets you specify the exact object parameters and target GIHT at every zoom position. The other spacing options do this automatically, which is better for most users, but if you really want explicit control, now you have it. You are reminded that if your input violates Lagrange's law, the results might make little sense.
5. The default thickness limit parameters (TLIM) are no longer also set by changing space limits (SLIM). It seems to make sense to keep them separate.
6. Drawings in PAD and DWG now show if the object is immersed. If so, they draw a diamond shape touching the object surface, the color (in PAD) derived from the index and dispersion of that medium.
1. OPD calculations were incorrect if the lens was in AFOCAL mode and had an odd number of reflections.
2. The ZMC command (to convert a Zemax file) did not work properly on zoom lenses.
3. The SCAO aberration did not work properly
4. If you ran a multiconfiguration Monte-Carlo tolerance analysis more than once, it could crash the program.
The update to version 15.68 incorporates the following changes:
1.You can now direct MC to save all of the sample Monte-Carlo cases (up to 100) in case you want to examine them yourself afterward.
2.The THERM program now accounts for the change in surface radius as well as the aperture of retaining rings when adjusting a thickness.
1.The new ATS analysis that works for almost all surface shapes did not do standard aspherics correctly.
2.THERM analysis did not work properly if the lens had hard CAOs.
3.The program did not clean up the many files called CORESUBnnn.RLE, which are created by some multicore features and are supposed to be temporary files.
4.BTOL could crash if a lens had a real pupil.
The update to version 15.67 incorporates the following changes:
1. ATS (Activate Thermal Shadowing) has been improved. Now it works with almost all surface shapes and with zoom lenses. It also removes any FILLSTOP or CSTOP attributes, following the pupil in ACON 1, scaling the aperture if WAP 2 or 3 is in effect.
2. The precision of RLE files has been increased. This reduces the difference in results from optimizing with switch 99 on and off in multicore mode.
1. AEI could in some cases change a paraxial pupil into an implied pupil, which is not exactly the same thing.
2. The sequence PAD / ZERO would loop.
3. The HELP buttons in some dialogs did not link to the current help file topic.
The update to version 15.66 incorporates the following changes:
1. The new version of object OBD has been revised. The pupil from a real stop is rotated with skew field, and the new OBD will now follow that same rule, to make things simpler to understand.
2. ACON PICKUPS ... now transfers all ZFILE object data if you specify 1 POB 1. Before, it only transferred data from zoom 1 .
3. SYNOPSYS can now run on a remote desktop. The dongle must be plugged in to that PC for it to work.
1. AED worked on all lenses, including aspherics. But that makes no sense, since the target to make the paraxial power go to zero ignored the higher orders. Now, AED ignores aspheric elements and the problem goes away.
2. ZDWG did not properly draw rays if the lens used the new OBD object.
3. Multicore optimization with the new switch 99 did not work properly if the optimization failed on the first pass.
4. Hyperhemispheric surfaces were not drawn correctly by CDWG.
5. If you gave a FIXed tolerance to an RFR in BTOL, the summary at the end did not repeat the same value if the element had EFILE data where point "A" was not at the CAO of the lens.
The update to version 15.65 incorporates the following changes:
1. ZSEARCH now lets you specify PLASTIC materials.
2. AED now honors switch 99.
3. The behavior of object type OBD has been revised. Before, the size of the entering beam would change if the first radius of curvature changed -- which made no sense. Now the default pupil with this object type traces a beam with constant size measured perpendicular to the chief ray. This is similar to the effect of the WAP 1 pupil, although the position of the rim rays still varies somewhat if the first radius is steep. Once you change to a real pupil and implement WAP 2 or 3, this size is adjusted, as before.
1. The TN function crashed in the previous version.
2. The lens layout tool (LLT) crashed if a previous optimization used the SNAP option.
3. Problems were discovered that involved certain combinations of ZSEARCH, OBD, VFIELD, ACONS, PAD SCAN, and optimization with switch 99 turned on.
4. The quick mode in ZSEARCH did not honor all of the SPECIAL AANT entries correctly.
5. The automatic aspheric assignment AAA did not return the best case if any of them encountered ray failures.
6. REVERSE did not work properly on zoom lenses.
The update to version 15.64 incorporates the following changes:
1. DSEARCH now puts color "M" in the optimization file if you requested GSHEAR and multicolors.
2. A new system flag REPUPIL B is now implemented. This is intended for the rare case when the paraxial chief-ray path yields a real ray that traces but comes out on the wrong side of a caustic.
3. You can now MAP the hologram surface frequency on a requested surface. Before, it analyzed the last HOE in the system.
4. DSEARCH now can model aspheres as Forbes type A surfaces.
5. A new monitor ADM will control the minimum diameter/thickness ratio. (Monitor ADT controls the maximum value.)
6. We have reinstated switch 65, which some found useful. It adjusts the entrance pupil for PAD and some other drawing
The update to version 15.63 incorporates the following changes:
1. A new surface shape is available. Type USS 29 is an 8th-power fully-populated polynomial.
2. ZSEARCH now supports object type OBD, as well as ZFOCUS with that object type.
1. An input error in the DSEARCH file could make multicore operations hang.
2. Entering WAP 2 and then STOP FREE in DSEARCH (which is not allowed) now aborts nicely.
3. The AGROUP option in ZSEARCH was not properly initialized if you had run a previous case.
4. The monitor AZA (automatic zoom airspaces) did not work on group 1 if that group started at surface 1.
The update to version 15.62 incorporates the following changes:
1. A new surface type (USS 29), a fully-populated 8th-order polynomial for freeform surfaces, is implemented.
2. ZFILE zooms lenses with object type OBD are now supported by CAM ZMAG.
3. Two new construction-parameter operands are now available. GSA and GSC will control surfaces with a "gull-wing" shape, which can be difficult to mold in glass.
4. Automatic DOE assignment ADA now supports multicore operation.
5. It is now possible to declare a lens system IMMERSE, which makes all air spaces take the index of the immersion medium. That can be WATER, SEAWATER, or a custom material modeled by GLM coefficients.
6. The program will now support up to 32 cores, doubled from the previous 16 cores.
1. Multithread raytracing of toric surfaces was incorrect.
2. The CAP listing for systems with object OBD that were assigned VFIELD data was not correct.
3. The Air Force target type in the Image Tools dialog (MIT) displayed incorrect frequency numbers if the reference dimension was specified in lp/mm. The displayed values were divided by 25.4.
4. If you ran ZDWG and then made a FRINGES / PUPIL plot, the plot was not centered correctly on the page.
The update to version 15.61incorporates the following changes:
1.Switch 65 has been deactivated.The effect was similar to the WAP 3 option on lens drawings, and we see no reason to maintain both features.
2.The drawing produced by DSEARCH now automatically ignores feathering, so you can see where the rays go even if they would be stopped. Lenses at this stage often are imperfect, so it makes no sense to delete those rays from the picture.
3.The GIHT parameter in ZSEARCH, used when the several zooms have different values, in AFOCAL mode, now controls the paraxial angle UB,not the height YA.A separate directive controls YA.Also, if the GIHT differs between zooms,the weighting of ray aberrations goes inversely with GIHT, reflecting the fact that smaller field angles require smaller images.
4.A new option is available in ZSEARCH.TSONLY will cause ray grids to correct only the TFAN and SFAN rays instead of a full grid of rays.This runs faster on systems where you don't need the full grid.
5.The program no longer scissors plots before they are displayed.This was a leftover from the batch-mode days, when the Calcomp plotter drivers did not deal gracefully with data that extended off thepage.Since modern graphics drivers are much better, we now allow the picture to extend beyond the page boundaries, so you can see what is out there.(If you print the plot, you still only get the data that fit on the page.)
6.The real-pupil search algorithm has been improved.The search used to fail with some of our examples; those now work properly.
1.The image tools feature (MIT) now takes into account the incident angle of the chief ray and stretches the image as needed.
2.An input error in the SPECIAL PANT section of DSEARCH input would crash the program.
3.The CSLOPE aberration was listed in the help file in two places. It is a construction parameter aberration, not a ray aberration.
4.The ATC monitor weight would override weighting on subsequent rays.
5.An implied real pupil,designated with APS -1 and a nonzero YPP1,did not work if the implied location was somewhere in image space.
6.The OPD calculation for afocal systems,if not in the major color, had a constant bias if the chief ray was not centered at the final surface.The image analysis was correct, since the shape was correct, but the PAD display showed the bias and was confusing.
The update to version 15.60incorporates the following changes:
1.A new aberration is available.CSLOPE will return the slope of the requested surface at the current CAO rather than along a given ray.This is useful for zoom lenses, where the CAO is taken at the largest aperture required for all zooms.
2.Some new monitors are now available. ACS will monitor the steepness of all surfaces whose curvature is varying, at the current CAO of that surface. This is similar to ASC, which only checks along the current upper and lower rim rays.This is intended for zoom lenses, where the CAO is set to the maximum aperture needed for all zooms.
3.Another new monitor is ATC, which checks the incidence angle of all rays in all colors at all zooms. This is to prevent critical-angle ray failures in zoom lenses.It is similar to ACA, which only checks a few rays at the current zoom.
4.The Automatic Asphere Assignment feature (AAA) now varies only aspheric term G 3 on a target surface.Before, it varied the conic constant -- but that did not work very well if the surface was nearly flat.This version should work with those surfaces too.
5.The program no longer scissors plots before they are displayed.This was a leftover from the batch-mode days, when the Calcomp plotter drivers did not deal gracefully with data that extended off the page.Since modern graphics drivers are much better, we now allow the picture to extend beyond the page boundaries, so you can see what is out there.(If you print the plot, you still only get the data that fit on the page.)
6.The GDIS command now lets you specify a size factor that can reduce the plot scale in the event that eliminating the scissoring makes the paraxial grid overfill the page.
1.The optimization program could go into a loop if rays failed on the initial pass and you did not request the FIX option.
The update to version 15.59 incorporates the following changes:
1. Two new materials are added to the U glass table: KIRO types CG1 and CG2. At the moment, those materials have no transmission data. When the vendor supplies them, we will add them.
2.. ARGLASS has been enhanced. Now it will vary the glass models to duplicate a potential real glass, even if that glass is outside of the current boundaries. Before, it honored those boundaries and sometimes could not match a glass that was outside. When it adjusts the model now, the real glass should work better when reoptimized.
3. If you assign a CORE number, the program now shows the result in the tray. That way you always know how many cores are authorized.
4. The logic of the PAD scan has been changed. For some reason, if a process sleeps for more than an instant, the Windows OS decides not to show the changing display until it is finished. We believe this behavior has been evaded.
5. The PAD drawings will now show the path of vignetted rays up to the point of failure, if you select the custom rayset for that display.
1. MTF calculations did not always work on a system that contained a real DOE surface. (The USS types of simple DOE were not affected.)
2. The Image Tools MIT did not display the correct image structure in COHERENT mode if the exit pupil was not circular.
3. The abscissa of CAM plots did not show the axis or tic marks.
The update to version 15.58 incorporates the following changes:
1. DSEARCH and ZSEARCH in multicore mode now run faster than before. Some jobs run twice as fast.
1. If you used the FLAG designation in a lens and then optimized it in multiconfiguration mode, the program only picked up the flagged surface if it was in configuration 1. Now it applies to the current ACON and works for any of them.
The update to version 15.57 incorporates the following changes:
1. The automatic symbols LB0, LB1 ... have been enhanced. Now you can reference a surface that is up to nine surfaces before the last, up to LB9. You can also reference a surface relative to a FLAG surface, up to nine surfaces removed, with FM9,FM8, ... FP8, FP9.
2. AEI can now add elements adjacent to most aspherics. Before, only spherical surfaces were treated and the shell initially got the same radius as that surface. Now shells will be added adjacent to an asphere but will start out flat.
3. The Excellent Editor EE will now display command words with a contrasting font and background. This will help you keep track of what is going on in your MACros. Comments are displayed in green type, as before.
4. The command to make an environment model (EMODEL) has been renamed to ENM, to avoid a collision with the command EM filename. The GRIN calculated to model a thermal nonuniformity has also been renamed to GRIN ENM.
1. The MTF portion of the GSS plot was not displayed if no MTF analysis had been done previously.
1. The option to load a bitmap picture from the clipboard in the MIT dialog does not work as it should. On Windows 7 it was perfect, but in Windows 10 it usually aborts partway through. Then, if you run the same procedure a second time, it usually works. We attribute this to a bug in Windows, over which we have no control. The same issue affects the MFB dialog.
The update to version 15.56 incorporates the following changes:
1. MIT and MFB have been enhanced. Now they show a progress bar when scanning a picture, so you know you have to wait a minute before clicking the Process button. Also, the MFB display was centered at the axis of the image surface, but if the axis ray did not go to the center (because of a prism, for example) the whole picture would show up decentered. Now it is centered at the axis ray.
1. The thermal shadowing feature ATS displayed the zoom bar if the temperature change was negative.
2. Input errors in the DSEARCH file could make the program go into a loop.
3. The automatic aspheric assignment (AAA) did not work properly in multicore mode.
The update to version 15.55 incorporates the following changes:
1. Switch 55 is now ON by default. (This samples diffraction patterns at 1/4 of the Airy disk radius instead of 1/2.) PCs are so fast these days that it makes no sense to revert to the less-accurate form.
2. DWG output with OBG systems now default to a single object height.
1. A SPT command did not work properly on a system with object OBG if you did not enter a scale factor.
2. The coherent analysis option in the Image Tools Menu (MIT) did not work correctly if the lens had the RPUPIL option in effect.
3. A problem was discovered with AFOCAL systems with a non-circular pupil, and with AFOCAL OPD calculations.
4. The value of an OPL aberration in the FINAL listing was incorrect.
5. ELD drawings of elements with rectangular apertures no longer display a "diameter" dimension.
The update to version 15.54 incorporates the following changes:
1. The automatic symbols LB0, LB1, etc. as well as APS, are now honored globally. The only exceptions are in routines where the symbol is a valid mnemonic, such as in an RLE file. Before, they only worked in AANT and PANT files.
2. The PAD display does a better job of remembering the zoom and pan settings of the several views as they are adjusted by clicking and dragging.
3. A new noun, NZOOMS has been added to the AI vocabulary.
4. There are now several new flavors of the PUPIL command. These allow you to view the wavefront at the exit pupil, at the last surface, and to superimpose the wavefront view from several different systems.
1. If rays failed after a successful quick search, ZSEARCH reported that case when it was finished. It should only have reported cases that optimized successfully.
2. GSEARCH could hang if you clicked the stopsign button to abort the process.
3. The AANT operands STX and STY returned the slope divided by 5, not the raw slope.
4. Object OBD would return erroneous OPD results if the object distance was less than 1.0E9.
5. FAMC could loop if you clicked the stopsign button when it was doing certain operations.
6. The MUT dialog would do an unwanted operation if you clicked the SCALE button.
7. Lens arrays did not display properly with RSOLID.
8. Surface type USS 19 (Y-cylinder) was not properly honored by XBEAM.
The update to version 15.53 incorporates the following changes:
1. MPE now has an option to create RSOLID drawings that can rotate much faster as you drag with the mouse. This is done by a coarser sample of the surfaces. (You can specify the sampling yourself with the SPART option in the RSOLID file.)
2. We have updated most of the lessons in the Online Tutorials, to reflect the current behavior of the program. Users who want to study those examples and run the MACros are advised to download the new versions.
3. The program no longer tests and adjusts the derivative increments if the starting merit function cannot be calculated or if FIXRAY has been requested. In the latter case, it is assumed that rays are going to fail and the problem will be corrected by that feature.
1. The new object grid option in MFB took the entered side length of a square pattern as the half-size, as with a circular pattern.
2. We found that if you copy a line from a PDF file that contains the apostrophy character "'", and paste it into the EE editor (from an online tutorial file, for example) the program would crash. Now it accepts that character.
The update to version 15.52 incorporates the following changes:
1. AEI now honors the LB0, LB1, etc. and FLAG, FB1, etc shortcuts as they work in an optimization MACro.
2. A surface can now be declared STATIC in the RLE file. This causes all variables related to that surface to be ignored in the PANT file, so you can use the handy VLIST ... ALL format and still keep certain surfaces and elements unchanged when you optimize the lens.
3. The DSEARCH optimization MACro now contains a much shorter PANT file, using the new STATIC feature to keep an entered front end static if it is not declared ALL.
4. The CAM program now has a new ZSPACE variety. The ENDS selection will distribute the zoom positions in a manner that concentrates them near the ends of the zoom range. This often works better than the other options if the zoom range is large.
5. Three plastics have been removed from the Unusual materials catalog. Types SP-1516, SP-1715, and SD-1414 are no longer on the vendors website and are presumed discontinued.
6. The Field Blur Menu (MFB) now has options to create a custom array of circles or squares, with user-entered size and spacings.
7. ZSEARCH now has an option (AGROUP) that causes the AANT file in the optimization MACro to contain a single ZGROUP ALL collection of aberrations that will apply to all zooms. Otherwise, it creates a separate set for each zoom, which can be useful if you want to adjust weights on an individual basis, which was the case before.
1. The DSEARCH combination USE CURRENT + COVER + BACK SET ... did not work properly.
2. The RPO plot created an incorrect plot if the numbers of fields was greater than the maximum allowed four.
The update to version 15.51 incorporates the following changes:
1. DSEARCH now has a FILTER directive, which lets you specify that certain elements should be either positive or negative, while the others can be either. If you know what a portion of your lens should look like, you can specify it this way -- and shorten the search time by a factor of two for each element so declared.
2. SYNOPSYS now has a STEP file converter. There are limitations, however, which you can read about in the UM.
1. The sequence IFR ... / IFP ... did not work if COMMON was requested.
2. ARG did not work properly with plastic elements.
3. Index values varied with GLM did not let the value of Nd rise above 1.9, even if you specified a higher limit with the CUL and FUL directives.
The update to version 15.50 incorporates the following changes:
1. SOLID and DSOLID now accept the mnemonic SPART, which sets the partition number for rendering surfaces. The default partition is very fine -- making a beautiful smooth curve -- but the DXF file from DSOLID was huge. Now you can request a coarser partition, and the file will be smaller.
2. We have enhanced the algorithm that finds the real pupil. It can still fail, in cases where the chief ray is close to a caustic, but it is more robust now.
3. The SINGLE input for ZSEARCH now accepts an integer number, as does DSEARCH.
4. We have updated the Ohara glass catalog, adding some new glass types.
5. Also, we have implemented a new feature: sn MOLD rate will alter the index of most of the low Tg glasses according to the rate at which pressed lenses are cooled. The index depends on this rate, and is somewhat lower than the catalog value. The data from Ohara are fully implemented, while Schott provides only a single molded value, from which one cannot extract the effect of the cooling rate. In any case, the change in index is modeled as MELT data, which apply to a specified surface and become part of the RLE file.
1. The previous version did not provide a way to turn off a selection in the WorkSheet.
The update to version 15.49 incorporates the following changes:
1. It is now possible for your annotated text to be displayed in Courier font in the graphics windows. This is intended for when you want to copy some printed text, such as Zernike terms, to paste on the drawing. The default text does not line up columns the same way, since it is proportional spaced..
2. DSEARCH now accepts an optional DAMP input and will add it to the optimization MACro. Sometimes a high damping makes it work better.
3. A new variable is available. The mnemonic PGM declares a surface PLASTIC and varies the index and Abbe number so that they both move along a curved line that passes through the area where plastics are to be found. If a surface is already declared PLASTIC, the usual GLM variable does the same thing. This should avoid problems with the very small area previously modeled with boundary conditions.
4. If your lens has aspherics defined as spherical or conic with G-term coefficients, a new variable DCA will vary the coefficients, so you don't have to list them individually, as before. You can select to vary only the axially-symmetric or rotationally-symmetric terms if you wish.
5. DSEARCH ... SINGLE now accepts the case you want to process as a decimal number. It still accepts a series of 1's and 0's, or P's and N's, as before.
6. PAD now lets you define a command link string for each of the five views in each of the six configurations. Before, a single string applied to all.
7. The Shape Option dialog for converting to or editing a Zernike surface now has a check box where you can select to make the surface go through the vertex. (Zernikes can have a nonzero sag at the origin, if this is not selected.)
8. We have modified one of the common blocks of the source code, and users who have implemented the USERDLL.sln to customize the program for their needs will need a new fortranstructures.h file. Please contact OSD and request this file before you use this new update.
1. ELD element drawings did not display mirrors whose thickness was declared CMIRROR correctly.
2. If switch 1 is off, the optimization could fail.
3. The SPS spreadsheet display did not show up properly on small notebook computers. Now it shows the lens data in groups of 25 instead of 50, so it works on small screens too.
4. If you ran AAA (Automatic Aspheric Assignment) and killed the process before it was finished, it loaded the current contents of the requested library location, which would have been overridden had the job finished.
5. The new protocol for selecting text in the Command Window would turn off the selection as soon as you pressed any key. But that disabled the <ctrl>C copy and <shift>extend-selection functions. Now those are allowed.
The update to version 15.48 incorporates the following changes:
1. The Automatic Aspheric Assignment feature (AAA) now supports multicores and will run much faster if that is implemented.
2. The PAD Scan button now shows the current HBAR in the PAD data box. So you can see exactly which field has a problem.
3. The FEATHER calculation now has an optional PLOT argument.
1. The SEL slider in the WorkSheet added one additional mouse move when you stopped moving your mouse. Now it doesn't.
2. The new Comment button in EE could miss the last line in the selection.
3. The fans plot in PAD would fail if the chief ray at the first HBAR analyzed did not trace.
4. The SEDGE edge declaration did not work properly if the second side of the element was tilted, and it did not display correctly in RSOLID if one side had a hard CAO but the other was soft.
5. The annotation editor deleted current annotation if you clicked the Copy button. It was supposed to add to, not replace it.
6. The DWG option to use BTOL data did not position the letter designation correctly if the element was declared SEDGE.
The update to version 15.47 incorporates the following changes:
1. We have added Lesson 48 to the Online Tutorials. This shows how to design a 22-element microlithography lens, from scratch, in about 45 minutes. Need we say no other code in the world can do this?
2. FFBUILD now honors a CFOV directive, if entered, and corrects only over a circular field. Otherwise it corrects over a rectangular field.
3. The EE editor toolbar has two new buttons. These will comment out or uncomment any block of text that is selected when the buttons are clicked. Now it is much easier to skip sections of the MACro, without making a SKIP block or editing each line in the selection separately.
4. The new commands SPL and RPL will save the definitions of the current PAD layout for all ACONS, and recall them later -- but will not save or restore the lenses therein. This is different from SSU and RSU, which save and recall both the layout and the lenses.
5. The EE editor option to color comments has been improved. Now it does not update the display after every keystroke, which could be slow if the file was large. Now it only updates after selected changes, and runs faster.
1. FFBUILD did not honor the LRAYS directive, if entered.
2. DSEARCH defaulted to a single field of view if the system was declared AFOCAL.
3. Some of the PAD labels could overlap others.
4. The default CAO on the image surface was incorrect if the aperture was not centered at the vertex.
The update to version 15.46 incorporates the following changes:
1. We have been tasked with analyzing a system with an isolated defect, such as a small bump or hollow on a surface. To do so, we implemented a new USS shape, number 28, which can model the bump.
2, A new GRIN form is also available, type GRIN LOCAL, which can model a local departure from a uniform index value. These two features can then be used to analyze the effect of these errors on the wavefront or image quality.
1. ZSEARCH did not always load the optimization MACro into an editor if the run specified explicit glasses for the elements.
The update to version 15.45 incorporates the following changes:
1. The top menu bar has been extended. For new users, this makes it a little easier to find the dialog you want. For experienced users, of course, it is entirely unnecessary, because the mnemonics are so easy to remember.
2. The DMASK ... ZONE analysis has been improved. Before, it sampled the aperture at 10 000 points and located the nearest zone with that precision. For fine fringes, this was not always accurate enough. Now it interpolates between those locations and is more precise.
3. A new command, CDIST, will analyze the distortion pattern over the field and return a set of interpolation coefficients that model that pattern. This is intended for use with digital displays, where one can compensate for lens distortion with an opposite distortion there. The coefficients tell you how to adjust things. The analysis can be done in each color, for cases where you also want to compensate for lateral color.
4. The Text Select button is gone. Now you can always select text lines from the Command Window. Typing anything in that window will automatically scroll to the bottom, where input has to go.
5. We have added another lesson to the Online Tutorials. Number 47 shows how to design a 30X zoom lens with three moving groups and 15 corrected zooms. We provide this lesson in response to a new user's request to support up to 45 zoom positions, which may indeed be necessary when using a competitor's primitive product. Unaware of the power and sophistication of SYNOPSYS, he assumed the same would be true here. But such a large number of zooms is not required when one has the sophisticated tools for the job provided by SYNOPSYS.
1. BTOL did not work properly if PFTEST was active and the lenses were extremely small.
The update to version 15.44 incorporates the following changes:
1. If you run ELD with a negative surface number, which disables the drawing but fills in the data boxes, the program will also disable listing aspheric coefficients. The latter are shown by default and make sense if the surface is rotationally symmetric -- but only symmetric terms are then shown. Now you can simply copy the terms from the ASY listing and paste them into the ELD drawing, along with PER drawings, as instructed in Section 7.7. Then all terms will appear on the picture, along with the drawings, which you can further annotate as needed. This is useful with free-form decentered elements, which the default ELD drawing does not render properly.
2. BTOL has been much improved. Now, if you declare a surface or collection of surfaces a GROUP, it will defeat the default tilt and decenter tolerance on those surfaces and instead calculate tolerances of all six degrees of freedom (X, Y, X, and alpha, beta, and gamma). This is intended for off-axis and free-form surfaces and mirrors, where one needs all six tolerances. It will also defeat the TH tolerance before the group, since the ZDC tolerance does much the same thing, but not after, since there are two airspaces involved.
3. The file saved by the BTOL SAVE command has been restructured. It now includes the new group tolerances, and it has also been reduced in size. Before, a .BTO file was over 30 MB in size; now for a typical system it is around 500 KB. This means that all previously saved BTOL files are now obsolete, and if you need them, you should run BTOL again and save a fresh copy. The previous file form will no longer open in BTOL.
4. The MOF command (MTF over field) now accepts the color number "S", which causes the MTF to be calculated in each color separately. This is intended for systems using a digital display of some kind, where one can compensate for substantial lateral color with software, thus eliminating the requirement to correct that error in the lens itself. This can sometimes result in a better image in each color, since one can then use the GPR raygrid request in the merit function (correct relative to the principal ray) instead of GNR, which corrects relative to the major-color chief ray.
5. The Worksheet dialog is now somewhat larger, so you don't have to scroll up and down as often as before.
6. The FPU command (PUPIL plot over field) now has a form FFPU, which covers the full field, from -1 to 1, instead of 0 to 1.
7. MC now treats decenters differently: it only perturbs the lens in Y, according to the budget, unless an explicit XDC range has been declared. This makes more sense, since one often only monitors the image in the Y-Z plane, and a decenter in X would appear less sensitive than it really is. If a system is not rotationally symmetric, the new GROUP tolerances, if applicable, get tolerances in X, Y, and Z, and that option is recommended in that case.
1. The BTOL summary listed the tolerance on XDC the same as on YDC, even though they were actually different in the budget listing.
2. The TFG plot drew the horizontal scale too wide.
The update to version 15.43 incorporates the following changes:
1. The routine that calculates the intercept of a grazing-incidence ray with an aspheric surface has been improved. Before, it often found the wrong intercept when there was more than one.
2. FIS (fringes at an intermediate surface) has been improved and works better at a decentered surface. It also now lists the Zernike coefficients that model the displayed wavefront.
3. A new surface shape is now available. Surface type USS 27 models a plane grating with a single surface. This is easier to use than the older GRATING specification.
4. The ORDER parameter on a GRATING surface has been reversed, bringing it into agreement with the new USS 27.
1. If you requested CSTOP and had a decentered CAO on the stop surface, the WAP 2 calculation did not work. Now it deletes any decenter when you use CSTOP.
2. The BTOL entry RANGE XDC range did not work.
3. The default RANGE for PFTEST surfaces in BTOL was sometimes too large, causing ray failures in the tolerance calculation.
4. Some kinds of ray failures during a BTOL run would loop indefinitely. Now it aborts and tells you which parameter was at fault.
The update to version 15.42 incorporates the following changes:
1. MTF over field (MOF) now has two new flavors: FMOF analyzes the full field, from HBAR = -1 to +1. And FZMOF does the same but with a Zernike fit to the wavefront. (MOF goes from 0 to +1.)
2. The ASY command now lists the group size along with the tilts and decenters on each surface.
3. The GOUT option has been improved. Now it sets the index at all wavelengths equal, so when the value reaches 1.0 you can just remove the element. Before, it only changed the index at the major color, and the others were sometimes not reasonable.
4. The FNUM entry in DSEARCH is now optional with AFOCAL lenses. You can also control the magnification with targets on selected rays.
5. The RZS command, which recasts a surface to a Zernike, has been improved. Now it has two new flavors: RZP tries to duplicate the slope of the result to match the target, while RZB targets both the sag and the slope. It also now runs more passes and then anneals, which usually gets a better fit.
1. MAP TRANS did not work properly in multicolor mode.
2. The shape options dialog would crash with USS 16 if you clicked the USS 15 selection.
3. OPD calculation was incorrect if the system had a DOE element that was also the UNDO surface for a previous tilt or decenter.
4. The glassmap disply MGT option to show the data at the current wavelengths in the lens did not show the VNUM correctly.
5. The SLOPE aberration came back with the wrong sign.
6. Surfaces designated SEDGE (simple edge) did not draw correctly if the CAO was decentered.
The update to version 15.41 incorporates the following changes:
1. A new shape is available: type USS 26 models a 3-layer DOE. This is similar to the 2-layer version USS 20, but has a layer of plastic between the two kinoforms and yields higher diffraction efficiency.
2. A new RLE entry LOG log will set the log number assigned to the lens to the entered value. Similarly, the command LOG LENS log will assign the number, as will the AI sentence LENS LOG = log.
1. Object type OBD did not support LRAYS.
2. ZDWG did not always position the lens drawings correctly.
3. The PAD SCAN button did not work properly with ZFILE zoom lenses assigned object type OBD.
4. Diffraction analysis did not work properly for DOEs in polarization mode.
The update to version 15.40 incorporates the following changes:
1. The spreadsheet SYS has been enhanced. Now you can change wavelengths directly in the edit boxes; before, they did not take effect until you pressed a second button.
2. DSEARCH now lets you specify selected surfaces as Zernikes. You can choose the coefficient set: all, bilateral symmetric, or radially symmetric.
1. The USE CURRENT option in DSEARCH did not clean up aspheric coefficients that were present on surfaces past the target point.
2. FINAL listed FLUX aberrations with a current value of zero.
The update to version 15.39 incorporates the following changes:
1. The flood-fill option in PAD has been enhanced, and it is now possible to declare a surface NOFFILL, for rare cases where the fill color leaks out and that feature got turned off.
2. DSEARCH now lets you USE CURRENT and vary the tilt and decenter on new elements.
3. A new operand, RCLEAR, can be used to keep adjacent surfaces from intersecting along a given ray. (The AEC monitor only works on default rays.)
4. The footprint dialog MFP now has an option to create rays over a grid of object points.
5. A new surface shape, USS 25 is now available. This models an extended DOE, with 32 terms in both X and Y, for systems where an axially-symmetric kinoform will not do.
6. The FINAL command has been enhanced. Now you can request FINAL NUMBER number, to show only the numbered aberration, or FINAL NAME "name", to show only the aberration with that assigned name.
7. Some new entries are available for the AANT file: The operand NURBS works for shape USS 15 and will target each of the Z-coordinates of the control points to the entered target. This is intended for cases where the point locations oscillate wildly -- still producing a smooth surface -- but are hard to use as variables. Just assign a small target and low weight, and the surface should be better behaved. The form XNURBS works on USS 24, the extended NURBS shape.
8. A similar operand, ASPHERIC, works on power-series aspherics and Zernike surfaces. If you want to see if you can remove aspheric terms, but rays fail if you just delete them, this lets you move them gracefully to zero.
9. The ACA monitor has been enhanced. Now, if you enter a negative weight, it monitors only the rays at the field point HBAR = -1. (A positive weight monitors the rays at a value equal to the diagonal distance to the corner of the field, which is appropriate for centered systems.)
10. You can now declare a surface SEDGES, which results in a very simple edge for most of the drawing programs. This is for cases that are too complicated for the usual edge-drawing routines.
11. A new online tutorial, Lesson 46, is now up. This shows how to design a converter for a laser diode whose divergence is different in X and Y and you want to make the beam circular.
1. If you entered a zero value for the Zernike unit radius, it could crash. Now it won't let you.
2. The Worksheet edit pane did not process the bottommost line unless you pressed the Enter key. Now it works either way.
3. Assigning aspherics to a DSEARCH run only worked for the first 21 surfaces.
4. Varying the RAD on a toric surface did not work.
5. If you ran RSOLID and then tried to print a window, the program would crash.
The update to version 15.38 incorporates the following changes:
1. AED now has a new option that will do an additional optimization run, after you agree to delete the selected element, where the targets on that element are controlled via "C" aberrations rather than by "M". This should reduce the thickness and power to very close to zero, so when that element is removed the ray paths are unchanged. Also, the former PIA parameter has been renamed PEL. The power is now controlled by taking the difference between the curvatures on both sides. Before, it used the first-order power, which was not appropriate if the element was cemented.
2. The number of allowed variables is increased from 200 to 400.
3. A new surface shape, USS 24, models a NURBS surface, as does type USS 15 -- but in this case the data area can contain up to 400 points instead of the 81 of the other shape. This form has a new restriction: only five slots are available, and you must select one of them when you assign this shape.
4. If switch 95 is turned on, KICK and ANNEAL will not alter GLM variables. (It still does not change the conic constant or general aspheric terms.) This should avoid ray failures that might occur in some systems.
5. PAD now shows an element with an aspheric on side 1 with a green surface number circle, and with blue if the aspheric is on side 2. If both sides are aspheric, the color is black or white.
6. A Zernike surface can now be assigned a new ZVZ setting. This will make term G 37 (a constant term) automatically adjust itself so the vertex of the surface actually goes through the axis point given by the preceding thickness. Since some Zernike terms are nonzero even at the axis, it is possible for the surface to wind up a distance from this vertex point otherwise -- which is an unwelcome surprise.
1. A DO MACRO loop that ended up in one of the image analysis features did not loop.
2. DWG did not correctly render a flat surface assigned a decentered aperture.
3. DSEARCH did not utilize an existing system (with USE CURRENT) if that system had aspheric surfaces.
4. SOUT operating on a surface within a tilt or decenter group did not work correctly.
5. The polarization phase difference was not properly calculated by IFP
6. ADEF ... FRINGES produced a blank picture.
7. The WorkSheet SEL slider sometimes produced a corrupted edit pane.