New Features

​

• Spectral analysis: Compute power spectral density (as periodograms or using Welch’s method) and spectrograms.  Analyze an entire recording or selected regions.

• Improved plotting performance: New adaptive display downsampling provides improved plotting performance for millions of data points. Large single-record files are now plot in their entirety, the ‘Limit view range for performance’ option and automatic splitting of large recordings on load are now turned off by default.

• Display Groups in Events Overlay: Display only a selected event group in the Events Overlay window, or exclude the selected group from display.

• Updated appearance and Dark Mode: The Windows application theme is upgraded to ‘Fusion’, and a Dark Mode is now available (under ‘View’). General improvements to widget spacing, and relative window position of child windows on macOS.
   

Key Fixes and Notable Changes

​

• Corrected an issue for HEKA files with stored per-record zero-offsets when saving all AP Kinetics phase-plot results. If the zero-offset correction was applied and data saved to disk, the selected record’s offset was used for all records, introducing a constant voltage shift where offsets differed. Displayed and clipboard-copied data were unaffected.

• Improved the Events decay-time calculation when the ‘Decay time baseline’ setting was changed to ‘Decay endpoint’. The endpoint remains the reference level, but the percentage amplitude now uses the peak-to-endpoint difference rather than the peak-to-baseline difference. This setting is now also documented in the manual.

• Corrected time scaling for the third derivative threshold in ‘Cutoff’ mode in AP Kinetics. The calculation divided by the sampling interval rather than its cube (Δt rather than Δt³). Previous cutoff values were internally consistent but saved values will need adjusting for use in this version.

• Loading a new file now requires the Record Arithmetic window to be closed first. Previously, the window could remain open after loading, but clicking ‘Apply’ had no effect on the new recording.

• Fixed displayed record numbers in the Average Records window being decreased by one when the window was reopened after previously being saved.

• Changed the behavior of the AP Kinetics threshold calculation when the threshold was not crossed by any sample. Previously, the first available sample in the spike waveform was used. Now, estimation fails with a warning message, and no threshold value is provided for that spike.

• Fixed best-score assignment in Events for events matched by Templates 1 and 3, where an event could remain assigned to Template 1 despite being a better match to Template 3.

• Fixed an issue in Events when using ‘Batch Mode’ where switching between ‘Monoexponential’ and ‘Biexponential’ event fitting across files could cause the results table to become misaligned when viewed in 'Row' mode.

• Corrected RMS threshold region behavior in Events when ‘Link Regions Across Records’ was off and standard deviation was calculated ‘Within selected region’. After moving the region bounds on a later record, the calculation could use a previous record’s bounds instead of the current record’s selected region.

• Now when using ‘Fit biexponential curve, calculate kinetics on fit’ in Average Events, measurements use the fitted waveform’s peak rather than the original waveform’s peak for consistency.

• If analyzing AP Counting or Input Resistance in ‘Batch Mode’, loading a new file and changing ‘Round Im’ before analyzing it could overwrite the previous file’s results.

• Improved the Filter window’s display of design order and effective order for clarity. To reproduce a previous low/high-pass final order setting of 8, use a design order of 4.

• Corrected calculation of rise midpoint from events fit with a biexponential function when using ‘Adjust start point to improve fit’, used for Average Event and Events Overlay alignment. In our testing, shifts were 1–2 samples with little effect on alignment. Calculated half-width times were unchanged.

• Applying Data Tools to an analyzed file in batch mode now clears any calculated results for that file.

Other fixes
 

• Improved ‘Round Im’ in Input Resistance mode to recalculate resistance using the existing analysis. Other settings changed since ‘Analyze’ was pressed are no longer reapplied.

• Manual event-kinetics edits in Events now use the event’s own template direction, rather than the direction of the currently selected template.

• Fixed an issue where manually moving an event’s decay endpoint could cause the baseline value to shift.

• If AP Counting was restricted to a subset of records and an AP was manually selected in a record prior to this range, the displayed Im values could be shifted up by one record. Rheobase and other calculations were unaffected.

• Improved handling of empty Events analyses (i.e. no detected events) when recordings were analyzed in ‘Batch Mode’ where viewing frequency and summary tables could result in crashes. If not event is detected on an analyses, manual detection is now deactivated.

• Corrected recording-length and time-offset handling when using ‘Concatenate’ mode in ‘Load Multiple Files’. Automatic splitting on load now uses the full combined recording length.

• Reset to Raw Data now preserves the current analysis if the original recording cannot be reloaded.

• Fixed a crash when reopening saved datasets of different lengths in the Kolmogorov–Smirnov window.

• When working with HEKA files with per-record zero-offsets, the phase-plot offset option is now disabled after using Data Tools that invalidate the stored offsets, such as record concatenation.

• When a file’s header contained missing or unrecognized channel units and units had to be entered manually, unit-preference settings in File Loading Options could incorrectly override that entry. Manually entered units now take precedence.

• Column headings for stored Curve Fitting event results now follow the fit method used for those results, rather than the current Events setting.

• Improved polynomial-fitting stability for detrending and curved event baselines and thresholds by rescaling time values to the range −1 to 1 before fitting.

• Recalculating an Events analysis RMS threshold after changing records now displays the current record’s value. Changing event polarity also refreshes the threshold plot.

• Improved Omit Times validation for relative times on offset or cumulative-time recordings, invalid text and malformed pasted tables. Pasted intervals must have exactly two columns, with one Start/Stop pair per row.

• Fixed crashes in ‘Batch Mode’ after attempting to load an incompatible recording type, such as a current-clamp file into a voltage-clamp batch, while the Events analysis window was open.

• Rerunning burst analysis requires the existing results table to be closed first, with a reminder to save anything needed.

• Fixed a crash when deleting multiple events (with the D shortcut) while viewing a later event in the individual-event browser.

• If Events Overlay is open and events are subsampled, the window is refreshed.

• Unsupported multichannel IGOR binary-wave files now produce an import error.

• Baseline and threshold labels for Events analysis now adapt the units for the recording type, rather than always displaying pA.

Interface improvements
 

• Extended the Generate Template width slider from 200 to 1000 ms

• Enabled accelerated Events template-trimming button repeat in both directions.

• Fixed sag and hump markers reappearing after rerunning Input Resistance analysis with sag/hump calculation disabled.

• Removed the unsupported legacy frequency-bin option.

• Filter FFT previews now exclude only DC, previously the first 10 bins were excluded

• Blank Kolmogorov–Smirnov input cells now remain empty rather than displaying “nan”, and labels explicitly identify the comparison as “Dataset 1 CDF Greater” or “Dataset 1 CDF Less”.

• Renamed the input-table “Cut” action to “Delete” as it does not copy.

• Corrected the band-pass warning to state that the lower cutoff must be below the upper cutoff.

• Upper-plot x-axis visibility is now a checkbox rather than a button in Graph Options.

• Fixed unintended zooming at the plot edge when using 'Track Values at Cursor' with 'Displayed Record scaling' in Graph Options.

• Corrected typographical errors in labels, including “Lilliefors”, “Empirical CDF” and “Cubic Spline Interpolation”.

• The random event-subsampling option is now labeled “Uniform random” rather than 'Uniform'.

• Changing the event number by entering a number or using the mouse wheel now moves the main plot to that event when event-navigation lock is enabled.

• Increased the deconvolution-threshold display limit from 99 to 999.

• Curved event baselines and thresholds now recalculate immediately when the polynomial order changes.

• Record Arithmetic now rejects “r0” as invalid to avoid confusion, and handles additional invalid operations, including division by zero, without changing the recording.

• CSV import now places a newly enabled time column first and requires at least one signal channel, with safe handling of empty reorder selections. The sampling-rate field also explicitly states Hz.

• Manually entering an odd anti-alias filter order in Downsample now restores the previous accepted even value.

• Main and burst-results tables are now read-only.

• AP Counting summaries now show average counts to three decimal places.

• Changing average-event filtering in Average Events now clears kinetics and closes the associated results table until recalculation. Events Overlay also refreshes after event-kinetics edits.