Choosing a hiding method
The Method selector in the Hide panel offers four options. Two work on lossless carriers (PNG, BMP), two on JPEG photos. Within each pair the trade-off is the same: more space or harder to detect.
| Method | Carrier | Trade-off |
|---|---|---|
| PNG or BMP — more hiding space (LSB-2) | PNG, BMP | Uses two low bits per channel: roughly double the capacity, slightly easier for statistical analysis to spot. |
| PNG or BMP — harder to detect (LSB-1) | PNG, BMP | One low bit per channel: half the capacity, subtler footprint. |
| JPEG photo — standard | JPEG | Hides in the JPEG's frequency coefficients. Good capacity on busy photos. |
| JPEG photo — harder to detect (F5) | JPEG | The F5 algorithm: less capacity, significantly more resistant to steganalysis. |
Rules of thumb
- Match the method to the carrier. An LSB method needs a PNG/BMP carrier; the JPEG methods need a JPEG. Mismatches fail with a format error.
- Busy beats flat. A detailed photo (foliage, texture, crowds) hides data far better than a flat logo or screenshot, in every method.
- When in doubt, LSB-2 on PNG. It's the default: dependable, roomy, and PNG survives file-based transfer perfectly.
- Choose "harder to detect" when the image will be scrutinised. If your threat model includes someone actively running steganalysis, take F5 (JPEG) or LSB-1 (PNG/BMP) and keep the payload small relative to the carrier.
How much fits?
Run the carrier through the Capacity panel — it reports what each method can hold for that specific image. As a rough guide, LSB-2 on a 12-megapixel PNG holds several megabytes; F5 holds the least of the four.
Tip
Payload too big? Compress it first (zip), split it, or use a larger carrier. A payload near the capacity ceiling also concentrates changes, making detection easier — headroom is stealth.