Keep live and historical data in view
Follow incoming readings, inspect an earlier range and return to Follow Live. Compare channels with stacked plots and multiple Y axes.
Instraxis Plot · part of Instraxis Control
See more of your experiment while it runs.
Control your instruments, follow growing measurement histories and investigate signals in one workspace. Instraxis Plot is our native visualization and analysis engine, developed for Instraxis Control.
Built into Control’s measurement workflow. The default for new normal and log figures.
From the next reading to the next insight
Acquire. Inspect. Analyze.
A chart is most useful when it helps you make sense of the measurement. These tools work inside Instraxis Control, alongside your sequence and recorded data.
Quantify a selected signal
Fit curves, inspect fitted parameters and standard errors, find peaks and measure FWHM. Run analysis on a fixed selection or a live trailing window.
Preserve the measurement
Analysis leaves acquired samples intact. Export numerical results with source context, and save figure projects to revisit data and results.
Understand signal quality
Explore noise statistics, filtering, FFT and time-frequency spectrograms. Irregular timestamps have dedicated methods; spectral tools report their sampling requirements.
Choose the range that matters
Rendering and analysis have separate limits. Exact raw analysis uses a 200,000-point default budget; the backend supports up to 2 million selected points. Oversized selections are rejected rather than silently reduced.
Feature availability follows the evaluated Control build. Existing explicit SciChart and ScottPlot figure selections remain compatible. Explore the complete Control workflow
12 September 2026 · all four backends rerun
Fresh measurements. Workloads you can inspect.
Instraxis Plot recorded the lowest p95 chart cycles in the three streaming cases, concurrent analysis and 100M-point replacement. ProEssentials recorded a slightly lower p95 for 10M-point replacement. Explore the workload closest to yours.
Every value below comes from the new campaign. All 72 processes across these six workloads passed acceptance: three fresh runs for each of four backends. The same workstation, input fixtures and application-level workload were used, with backend-specific adapters. See conditions and exclusions.
Stream · 100M points + 1M samples/s
Fresh four-backend campaign · p95 completed cycle · milliseconds · lower is better
Instraxis Plot recorded the lowest p95 chart cycle in this 100M-point +1M/s replay. Instraxis Plot and SciChart both accepted 99.87% of scheduled input within the window.
Instraxis Plot accepted 29.96M of 30M scheduled new samples within 30 seconds; 40,000 remained at the boundary. All four backends reached 130M retained points after catch-up, with no unconsumed input.
| Backend | Input accepted within 30 s | Final drain |
|---|---|---|
| Instraxis Plot | 99.87% | 36.29 ms |
| ProEssentials 11 | 96.70% | 622.41 ms |
| SciChart 9 | 99.87% | 142.89 ms |
| LightningChart 10 | 98.70% | 277.65 ms |
Drain measures final catch-up after the producer ends. A cycle crossing the 30-second boundary can finish before the drain timer starts, so this is not total boundary-to-completion time. The last input block is scheduled exactly at the deadline.
Stream · 10M points + 100k samples/s
Fresh four-backend campaign · p95 completed cycle · milliseconds · lower is better
Instraxis Plot recorded 37.79 ms p95, the lowest of the four backends in this replay.
All four backends retained 13M points after catch-up, with no unconsumed input. In-window delivery and final drain are shown separately below.
| Backend | Input accepted within 30 s | Final drain |
|---|---|---|
| Instraxis Plot | 99.87% | 19.08 ms |
| ProEssentials 11 | 99.53% | 92.65 ms |
| SciChart 9 | 99.83% | 23.35 ms |
| LightningChart 10 | 99.83% | 60.80 ms |
Drain measures final catch-up after the producer ends. A cycle crossing the 30-second boundary can finish before the drain timer starts, so this is not total boundary-to-completion time. The last input block is scheduled exactly at the deadline.
Stream · 100M points + 100k samples/s
Fresh four-backend campaign · p95 completed cycle · milliseconds · lower is better
Instraxis Plot recorded 38.70 ms p95, the lowest of the four backends in this replay.
All four backends retained 103M points after catch-up, with no unconsumed input. In-window delivery and final drain are shown separately below.
| Backend | Input accepted within 30 s | Final drain |
|---|---|---|
| Instraxis Plot | 99.87% | 38.67 ms |
| ProEssentials 11 | 97.90% | 543.70 ms |
| SciChart 9 | 99.80% | 108.64 ms |
| LightningChart 10 | 99.83% | 217.22 ms |
Drain measures final catch-up after the producer ends. A cycle crossing the 30-second boundary can finish before the drain timer starts, so this is not total boundary-to-completion time. The last input block is scheduled exactly at the deadline.
Replace all X/Y · 10M points
Fresh four-backend campaign · p95 completed cycle · milliseconds · lower is better
ProEssentials recorded 58.06 ms p95 and Instraxis Plot 58.74 ms. The small observed difference does not establish a statistically significant advantage.
Twenty complete A/B/C source-bank replacements per process; three accepted processes per backend. The first replacement is included in all 60 pooled events.
Replace all X/Y · 100M points
Fresh four-backend campaign · p95 completed cycle · milliseconds · lower is better
Instraxis Plot recorded 230.94 ms p95, the lowest of the four backends for complete 100M-point X/Y replacement.
All 60 replacement cycles per backend are included. Source-bank generation is outside the timer; validation, index or geometry rebuilding and completion work remain inside the measured path.
Stream + analysis · 100M points + 100k/s
Fresh four-backend campaign · p95 completed cycle · milliseconds · lower is better
Instraxis Plot recorded 38.39 ms p95 chart cycles while the common analysis ran. Analysis request-to-UI latency is a separate measurement.
A common C# algorithm analyzes a 10M-point Gaussian region once per second. All backends delivered 29 of 30 requested jobs within the window. This does not benchmark native analysis functions or Control analysis capacity.
Bars begin at zero and share a linear scale within each workload; the scale changes with the selection. P95 pools measured events across three accepted processes. These are descriptive observations on the tested configuration.
| Workload | Instraxis Plot | ProEssentials 11 | SciChart 9 | LightningChart 10 |
|---|---|---|---|---|
| Stream · 100M points + 1M samples/s | 38.47 | 667.67 | 46.00 | 275.66 |
| Stream · 10M points + 100k samples/s | 37.79 | 110.43 | 45.43 | 65.00 |
| Stream · 100M points + 100k samples/s | 38.70 | 557.53 | 46.18 | 236.18 |
| Replace all X/Y · 10M points | 58.74 | 58.06 | 91.62 | 321.87 |
| Replace all X/Y · 100M points | 230.94 | 415.39 | 533.26 | 3,184.96 |
| Stream + analysis · 100M points + 100k/s | 38.39 | 560.31 | 46.26 | 234.52 |
Bold marks the lowest measured p95 in each row. “100M” is initial retained history for streaming and the complete dataset size for replacement. The concurrent row measures chart updates while a common managed analysis runs. These are backend measurements, not complete Control throughput or a universal vendor ranking.
Beyond the headline · fresh 100M-point results
Prepare large datasets. Keep exploring.
Instraxis Plot’s preparation plus first completed cycle measured 235.76 ms. Inspect the loading and navigation results together: a fast attachment call alone does not describe the work needed to prepare a chart.
| Measurement | Instraxis Plot | ProEssentials 11 | SciChart 9 | LightningChart 10 |
|---|---|---|---|---|
| Attach APIMedian · ms | 64.91 | 1.04 | 726.57 | 569.02 |
| Preparation + first completed cycleMedian · ms | 235.76 | 349.20 | 936.48 | 1,992.08 |
| Pan completed cyclePooled p95 · ms | 33.55 | 33.20 | 34.27 | 295.45 |
| Zoom completed cyclePooled p95 · ms | 33.64 | 34.27 | 63.44 | 1,199.45 |
| Random viewport completed cyclePooled p95 · ms | 33.54 | 33.58 | 682.29 | 718.36 |
Follow incoming data
The 100M-point +1M samples/s replay recorded 38.47 ms p95 chart cycles. Instraxis Plot and SciChart both accepted 99.87% of scheduled new samples within 30 seconds. All four backends consumed the full input after catch-up.
Keep analysis results within reach
With a common managed algorithm analyzing a 10M-point region once per second, Instraxis Plot’s request-to-UI p95 was 75.11 ms. This measures calculation and delivery under chart contention; Control’s native analysis tools have separate capabilities and limits.
Compare memory and analysis delivery
| Measurement | Instraxis Plot | ProEssentials 11 | SciChart 9 | LightningChart 10 |
|---|---|---|---|---|
| Static: warm extra private memoryMedian · MiB | 74.82 | 3,066.84 | 1,597.67 | 3,077.09 |
| Static: total warm private memoryMedian · MiB | 1,624.30 | 4,668.14 | 3,284.24 | 4,666.98 |
| Replacement: extra private memory at endMedian · MiB | 83.61 | 3,069.83 | 0.11 | 9,300.59 |
| Replacement: total private memory at endMedian · MiB | 4,746.66 | 7,750.76 | 4,816.63 | 13,965.91 |
Memory depends on the operation and the baseline. Instraxis Plot used the least total private memory in these two snapshots; SciChart used the least additional memory for full replacement. Each extra-memory value subtracts that backend’s own empty-chart-plus-source snapshot. A small delta does not describe the total process.
Static source arrays alone occupy 1,525.88 MiB. Replacement keeps three caller-owned banks totaling 4,577.64 MiB. MiB means 1,048,576 bytes; these measurements do not isolate GPU allocations or establish leak freedom.
| Measurement | Instraxis Plot | ProEssentials 11 | SciChart 9 | LightningChart 10 |
|---|---|---|---|---|
| Concurrent analysis request to UIPooled p95 · ms | 75.11 | 566.90 | 132.17 | 273.99 |
The same C# mean, standard deviation, RMS and known-baseline FWHM algorithm runs for every backend. All delivered 29 of 30 requested jobs within the window; later completions remain recorded. The test does not compare vendor-native scientific routines.
Preparation sums timed attachment, explicit index construction where exposed, and the first paint/pump/compositor cycle. It excludes source generation, control creation and diagnostic gaps, so it is not continuous process-launch or file-loading latency. ProEssentials recorded the quickest attachment and a slightly lower 100M pan p95; no backend leads every operation.
Test conditions
One campaign. Explicit acceptance rules.
The campaign completed 126 planned attempts: 117 of 120 chart processes and all six chart-free analysis baselines passed. All six displayed extended workloads and all displayed 100M static cohorts are complete.
Hardware & software
- CPU / memory
- Intel Core i7-10700 @ 2.90 GHz; 8 cores / 16 threads; 32 GiB; configured 2933 MHz.
- Graphics
- Intel UHD Graphics 630; driver 31.0.101.2140.
- Windows / display
- Windows 11 Pro, build 26200; Balanced power plan. 1920 × 1080, reported 59 Hz; 1000 × 650 chart client.
- Host runtimes
- Release x64. Instraxis Plot static native Direct2D host and extended .NET 10 bridge; ProEssentials and SciChart use .NET 10, with WPF ElementHost for SciChart; LightningChart uses .NET Framework 4.8. Verified existing binaries ran in isolated host trees.
- Tested versions
- Instraxis Plot revision
1a967f8; ProEssentials 11.0.0.2; SciChart 9.0.0.29196; LightningChart 10.0.1.4001.
What the timer includes
Append/replacement or range API, library/window completion, UI pumping and DwmFlush. Completion mechanisms differ by backend. Not physical display latency, FPS or isolated GPU time.
P95 is the nearest-rank 95th percentile, pooled over three accepted processes. Faster runs can contribute more events. Repeated cycles remain correlated; three processes provide descriptive comparisons, not a statistical significance test.
Runs occurred on 12 September 2026, 07:09–08:40 UTC. Processes ran serially, with backend order rotated across cases and repetitions. Verified existing binaries were reused in isolated hosts; this was a fresh execution campaign.
Replay publishes precomputed immutable input every 10 ms and requests at most 30 chart cycles/s. It excludes device transfer, live source writes and waveform generation.
Acceptance, excluded results and image validation
The full campaign has 41 complete cohorts out of 42. Each needs all three original accepted processes. ProEssentials’ 100k static cohort is withheld: all three attempts passed numerical checks but failed selected deep-zoom geometry checks. This describes the tested integration, not a general library capacity limit. No older measurement fills the gap.
Of 528 retained capture records, 516 belong to image-accepted runs. Manual review covered 143 distinct original/fixture groups, including two rejected groups. Duplicate decisions are bound to exact hashes and matching fixtures. An internal independent raw-metric audit passed 15,258 checks with no mismatches; this is not third-party certification.
Eight focused source-matched correctness suites passed on 11 September and were not rerun on 12 September. Only selected checkpoints were reviewed, not every timed frame. Narrow raster slits remain in some dense Instraxis Plot waveform bands. SciChart runtime-key and ProEssentials evaluation overlays were present in the tested installations.
Workload definitions and backend adapters
One explicit double-precision X/Y series. Replay retains all history for 30 seconds. The view shows the latest 1M points and visits the first 10M-point region for 250 ms every 2 seconds. Replacement cycles three precomputed datasets across 20 changes per process, including the first. Selected captures and settling pauses occur between timed cycles.
Instraxis Plot borrows source arrays, validates X, builds a compact index and refreshes it incrementally during replay; full replacement rebuilds the index. ProEssentials rebinds a growing pinned prefix, reasserts Filter2D3D and forces geometry rebuilding. SciChart uses bulk Append into owned static/streaming storage and documented array-backed replacement. LightningChart converts X/Y into SeriesPoint values for PointLineSeries. These are matched application-level workloads with different supported integration paths.
SciChart uses DirectX 11 Visual Xccelerator, Auto resampling and large-capacity ImpossibleMode. ProEssentials uses its Direct3D path with pinned X/Y buffers. LightningChart uses hardware DirectX 11 and AfterRendering completion. Its PointsOptimization=None setting concerns hidden markers, not disabled line optimization. Completion mechanisms and inner plot margins differ; no private GPU fence equalizes physical presentation.
What to validate in your Control evaluation
Use your intended instruments, PC, channel count, data rate and measurement duration. This campaign does not establish multichannel capacity, fixed-history FIFO performance, long-duration reliability, high-DPI or cursor latency, GPU traffic, or complete acquisition and logging throughput.
Control’s data table, file loading, source preparation and analysis add their own work. Confirm the supplied Control build and test the full measurement sequence. Instraxis Plot is developed for that integrated workflow.
Local measurements and vendor demonstrations
Similar point counts can measure different work.
The comparison above uses our fresh local runs. Vendor-published demonstrations provide useful methodology context, but their numbers are not inputs to our charts.
ProEssentials 11
Gigasoft distinguishes rendering an existing 100M-point dataset from replacing it with fresh data. Our explicit double X/Y replacement includes update processing; retained-history replay also differs from fixed-history circular-buffer streaming.
Gigasoft performance contextSciChart 9
SciChart’s linked WPF comparison dates to 2021, covers many chart types and caps tests at 60 FPS. Our campaign tests installed WPF 9.0.0.29196 with completed-cycle timing and the workload definitions above.
SciChart benchmark methodologyLightningChart 10
Our installed 10.0.1 PointLineSeries path predates SampleDataBlockSeries. The vendor’s 2021 comparison uses 10.1.1, block-series streaming and SciChart 6.3, with a 10-second scrolling history. It is a different workload.
LightningChart benchmark methodologyResults apply to the installed versions and integrations listed here, not every vendor’s latest release or fastest possible architecture. Third-party names belong to their owners; no endorsement is implied.
Sources & references
- Instraxis, “Four-backend comparison: loading, replacement and responsiveness,” completed 12 September 2026 at 08:56 UTC. Final report, acceptance summary, original events and independent metric audit for the fresh campaign. Download the public results and conditions (JSON). This curated summary includes source hashes; the raw evidence remains in the research archive.
- Gigasoft. ProEssentials product and performance description; ProEssentials v11 release notes.
- SciChart. SciChart WPF v9 release; 2021 WPF performance comparison.
- LightningChart. LightningChart v10.1.1 vs SciChart v6.3 comparison; SampleDataBlockSeries performance tests.
Vendor methodology references were checked on 11 September 2026. All performance values displayed on this page come from the 12 September Instraxis campaign.
Bring your signals. Evaluate the complete workflow.
Tell us your instruments, channel count, data rate, measurement duration and target PC. We can focus the Instraxis Control evaluation on the experiment your lab needs to run.