.. _changelog_v4_3_0:

v4.3.0
======

v4.3.0 adds hardware-accelerated JPEG decoding, FLIR Lepton high-temperature
mode and model detection, new JPEG color-space conversions, and MicroPython
1.18, plus a batch of RTC and WiFi fixes. The Lepton measurement-mode API
changed — read the breaking changes below.

.. contents:: On this page
   :local:
   :depth: 1

Highlights
----------

- **Hardware JPEG decoding** on STM32 (H7) — faster JPEG decompression.
- **FLIR Lepton high-temperature mode** (up to ~600 °C on Lepton 3.5) and
  automatic model detection (1.5–3.5).
- **New JPEG conversions** — decode YCbCr JPEGs to grayscale and grayscale
  JPEGs to RGB565.
- **MicroPython updated to 1.18.**
- **Breaking:** the Lepton measurement-mode ioctl changed — see the breaking
  changes.

New features
------------

- **Hardware-accelerated JPEG decoding** — the STM32 on-chip JPEG codec is
  now used for decompression.
- **JPEG color-space conversions** — the decoder can decode YCbCr JPEGs to
  GRAYSCALE and GRAYSCALE JPEGs to RGB565.
- **Lepton high-temperature mode** — a new ``high_temp`` measurement mode
  uses low-gain to extend the range to roughly 600 °C on the Lepton 3.5.
- **Arduino Nano RP2040 Connect** — the :mod:`onewire`, :mod:`dht`, and
  :mod:`neopixel` driver modules are now frozen into the firmware.

Other changes and improvements
------------------------------

- **MicroPython updated to 1.18.**
- The FLIR Lepton driver now detects and reports the specific Lepton model
  (1.5 / 1.6 / 2.0 / 2.5 / 3.0 / 3.5) to the IDE.

Bug fixes
---------

Image processing:

- ``Image.to_grayscale()`` / L-channel conversion no longer returns negative
  L values for dark RGB565 pixels; the JPEG decoder no longer faults
  (MemManage) on odd-dimension images.

Camera and sensors:

- Fixed GC2145 window subsampling ratios and Bayer pixel order for correct
  image quality at sub-QVGA resolutions and in BAYER mode.

Networking and storage:

- Corrected WINC1500 non-blocking socket timeout handling and added proper
  MicroPython errno mapping for socket errors; fixed the STM32 OpenMV
  USB-disk file handling.

Real-time clock:

- Fixed the RTC clock source / LSE-vs-LSI selection and LSE oscillator errata
  across the OpenMV Cam H7 Plus and Pure Thermal boards.

Hardware and board support
--------------------------

- **STM32 (H7)** — hardware JPEG decoding.
- **OpenMV Cam H7 Plus / Pure Thermal** — RTC clock-source fixes.
- **Arduino Nano RP2040 Connect** — frozen :mod:`onewire` / :mod:`dht` /
  :mod:`neopixel` drivers.

Breaking API changes
--------------------

User-visible API breaks between v4.2.3 and v4.3.0. Scope: Python C-modules in
``modules/`` and Python libraries in ``scripts/libraries/``.

Each change is tagged with its impact:

- *minor* — narrow API; only affects scripts that used it.
- *behavior* — same API, different results; re-check tuned scripts.

Both breaking changes affect the FLIR Lepton thermal sensor only. Each commit
hash links to its diff on GitHub.

.. _v4_3_0_lepton_mode:

Lepton measurement-mode ioctl *(minor)*
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

``sensor.ioctl(SET_MEASUREMENT_MODE, ...)`` accepts a new optional
``high_temp`` argument, and ``GET_MEASUREMENT_MODE`` now returns a 2-tuple
``(enabled, high_temp)`` instead of a single boolean. Code that read the
previous boolean return value must be updated to unpack the tuple.

*Commit:* `6b2cd1008 <https://github.com/openmv/openmv/commit/6b2cd1008>`__

.. _v4_3_0_lepton_range:

Lepton default temperature range *(behavior)*
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

The FLIR Lepton default temperature range changed from −17.78 °C…37.78 °C to
−10.0 °C…40.0 °C. The color mapping and reported temperatures for scenes that
relied on the old default range will differ — set an explicit range if you
need the previous behavior.

*Commit:* `bc07e0131 <https://github.com/openmv/openmv/commit/bc07e0131>`__

Migration checklist
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Only FLIR Lepton scripts are affected: unpack the new
``GET_MEASUREMENT_MODE`` ``(enabled, high_temp)`` tuple
(:ref:`the measurement-mode change <v4_3_0_lepton_mode>`) and set an explicit
temperature range if you depended on the old default
(:ref:`the default-range change <v4_3_0_lepton_range>`). All other scripts run
unchanged.