TIFF to PNG is useful for pulling a browser-friendly image out of scanner TIFFs, but our image engine writes one decoded frame rather than a page set.
From real runs
| Element | Result | What happens |
|---|---|---|
| First TIFF frame | Kept | Our 4,544 byte two-frame LZW TIFF produced one 1,378 byte PNG from the 300×200 first frame only. |
| Later TIFF frames | Lost | The 220×160 second page in the same TIFF was not written to a ZIP or a second PNG by the API run. |
| 16-bit grayscale | Kept | A 1,166 byte deflate-compressed I;16 TIFF became a 276 byte 16-bit grayscale PNG at 256×128 pixels. |
| LZW compression | Changed | LZW is decoded before writing PNG, so the output is a new PNG stream rather than a wrapped TIFF payload. |
| CMYK TIFF | Lost | A 180×120 CMYK TIFF returned conversion_failed through the API because PNG writing did not convert CMYK to RGB first. |
| Scanner tags | Changed | Resolution and compression tags are read for decoding, but the PNG is not a preservation container for TIFF metadata. |
Our 4,544 byte two-frame LZW TIFF produced one 1,378 byte PNG from the 300×200 first frame only.
The 220×160 second page in the same TIFF was not written to a ZIP or a second PNG by the API run.
A 1,166 byte deflate-compressed I;16 TIFF became a 276 byte 16-bit grayscale PNG at 256×128 pixels.
LZW is decoded before writing PNG, so the output is a new PNG stream rather than a wrapped TIFF payload.
A 180×120 CMYK TIFF returned conversion_failed through the API because PNG writing did not convert CMYK to RGB first.
Resolution and compression tags are read for decoding, but the PNG is not a preservation container for TIFF metadata.
We converted our TIFF sample (19 KB) with Image engine on 2026-10-20. It took 137 ms and produced a 21 KB PNG file, about the same size as the input. The engine raised 1 warning, listed below.
Input · sample.tif · 19 KB
1200 × 800 px, deflate
Output · sample.png · 21 KB
1200 × 800 px, 8-bit, colour type 6
A TIFF scan is still pixels after TIFF to PNG. If selecting text in the source fails or the page is a fax-style image, Deep Read is the route for searchable text rather than this raster conversion.
Pillow 11
from PIL import Image
im = Image.open("scan.tiff")
im.seek(0)
im.save("scan.png", optimize=True)This mirrors the first-frame behavior; iterate ImageSequence yourself when every TIFF page needs a PNG.
The HTTP API and the MCP server run the exact engine this page uses (Image engine); the response carries the same warnings the web converter shows. Check support first with GET /api/v1/pairs?from=tiff&to=png.
curl -X POST https://filetypeconverters.com/api/v1/convert \
-H "Authorization: Bearer $FTC_API_KEY" \
-F "file=@input.tif" -F "from=tiff" -F "to=png" \
-F 'options={"width":0}' \
--output output.png| Key | Applies to | Type | Default |
|---|---|---|---|
| width | Writing PNG | number | 0 |
| height | Writing PNG | number | 0 |
| fit | Writing PNG | contain · cover · fill | contain |
| quality | Writing PNG | number | 90 |
| background | Writing PNG | string | #ffffff |
| stripMetadata | Writing PNG | boolean | true |
No. In our API run, only the first frame became a PNG. Use a PDF route or split the TIFF first when later frames matter.
Yes for the tested I;16 path. The 256×128 gradient stayed a 16-bit PNG instead of being reduced to 8-bit color.
PNG does not accept CMYK pixels directly. The current our image engine path did not convert that mode before saving, so the API returned a conversion_failed error.