Tifffile: Read and Write TIFF Files

2026.3.3 · active · verified Sun Mar 29

Tifffile is a Python library designed for reading and writing TIFF files, particularly those prevalent in scientific imaging. It offers extensive support for various TIFF formats, including BigTIFF, OME-TIFF, and multi-page TIFF stacks. Actively maintained, it receives frequent updates, with the current version being 2026.3.3, reflecting continuous development and improvements.

Warnings

breaking The `TiffPages.pages` and `FileSequence.files` attributes, as well as `stripnull`, `stripascii`, and `bytestr` functions, have been removed. Command-line interfaces were rewritten.
Fix: Migrate code to use current API for page and file sequence access. Consult the `tifffile` GitHub repository for specific replacements in the releases or changes log [6].
breaking The `TiffWriter.save` method is deprecated; use `TiffWriter.write` instead. Various other `TiffWriter` parameters like `compress` have been replaced by a `compression` parameter.
Fix: Update `TiffWriter` calls to use `TiffWriter.write` and the `compression` keyword argument. For example, `tif.save(data, compress=6)` becomes `tif.write(data, compression=6)`.
breaking The `multifile` parameter for `TiffFile` was removed. Support for Python 3.7 and earlier has been dropped, and Python 32-bit versions are deprecated.
Fix: Ensure your environment uses Python 3.8+ (64-bit recommended). Review usage of `TiffFile` to remove the `multifile` parameter if present [9, 13].
gotcha For very large TIFF files, loading the entire image into memory with `imread` can lead to `MemoryError`. Also, `TiffFile` objects should be properly closed.
Fix: Use `tifffile.memmap` for memory-efficient access to large files, or iterate through `TiffFile` pages to process chunks. Always use `TiffFile` within a `with` statement to ensure it is properly closed [3, 14].
gotcha Some TIFF-like formats do not strictly adhere to the TIFF6 specification, which might lead to unexpected behavior or incomplete data reading if not handled correctly.
Fix: Be aware that `tifffile` supports a large subset but not all possible TIFF variations. Consult the documentation for specific format support, especially for proprietary formats [1, 4].
breaking Decoding EER super-resolution sub-pixels and parsing EER metadata to dict changed significantly, introducing breaking changes.
Fix: If working with EER files, review code that decodes EER sub-pixels or parses EER metadata, as the API has changed. Refer to the latest documentation or release notes for the new approach [4, 6].

Install

pip install tifffile Minimal installation
pip install -U tifffile[all] Full installation with all optional dependencies
conda install tifffile -c conda-forge Conda installation

Imports

imread

import tifffile; image_data = tifffile.imread('image.tif')

Directly reads image data into a NumPy array.

imwrite

import tifffile; tifffile.imwrite('output.tif', image_data)

Directly writes a NumPy array to a TIFF file.

TiffFile
```
from tifffile import TiffFile; with TiffFile('image.tif') as tif: ...
```
Provides object-oriented access to TIFF file structure and metadata; should be used with a context manager for proper resource handling.

Quickstart

This quickstart demonstrates how to write a NumPy array to a TIFF file, read it back, verify its content, and access basic metadata using `tifffile.imwrite`, `tifffile.imread`, and the `tifffile.TiffFile` context manager. It also includes cleanup of the created file.

import numpy as np
import tifffile
import os

# Create some dummy image data
dummy_image = np.random.rand(100, 100).astype(np.float32)

# Define a filename
filename = 'example.tif'

# Write the image data to a TIFF file
tifffile.imwrite(filename, dummy_image, description='Generated by tifffile quickstart')
print(f"Successfully wrote '{filename}'")

# Read the image data back from the TIFF file
read_image = tifffile.imread(filename)
print(f"Successfully read '{filename}' with shape: {read_image.shape}")

# Verify data
assert np.array_equal(dummy_image, read_image)
print("Original and read image data are identical.")

# Access metadata using TiffFile context manager
with tifffile.TiffFile(filename) as tif:
    if tif.pages:
        first_page_description = tif.pages[0].tags.get('ImageDescription')
        if first_page_description:
            print(f"ImageDescription from file: {first_page_description.value}")
        else:
            print("No ImageDescription found for the first page.")
    else:
        print("No pages found in the TIFF file.")

# Clean up the created file
os.remove(filename)
print(f"Cleaned up '{filename}'.")

view raw JSON →