Quick wins for a faster PC:
Scan for outdated or missing drivers - takes under a minuteDriver Scan →Repair Windows errors before they cause bigger problemsFix Now →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Hidden data embedded with least-significant-bit (LSB) methods can disappear when a chat app re-encodes a photo. The most direct measurement found for Telegram, published on 25 May 2026 by Lu’lu’ Atik Fitriyani and Fahmi Fachri in Jurnal Teknologi Informasi dan Multimedia, reports that LSB messages were extracted perfectly when the images were sent as a document, but extraction failed when the same images were sent as an image. Robust watermarking can improve recovery under the attacks its authors tested, but its results depend on the algorithm, the processing applied, and the route the file takes. No single study shows that every chat app destroys LSB data or that every robust watermark survives every route.
Why LSB data breaks when a photo is re-encoded
LSB embedding writes each bit of a message into the lowest-order bit of pixel values. Extraction is exact: the reader needs the same bits back, in the same positions. One flipped bit is one wrong bit, so the method has little tolerance for anything that rewrites pixel values, even slightly.
As an Amazon Associate I earn from qualifying purchases.
JPEG compression is a common cause. It discards fine detail by quantizing frequency coefficients and then rebuilds pixel values from the result, so many low-order bits no longer hold what the sender wrote. Resizing resamples pixels and can do the same. Whether a chat app triggers either step depends on how the upload is handled, which is why the same file can survive one sending option and fail another.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Fix the driver behind crashes, sound loss and screen glitches3Repair Windows errors before they cause bigger problemsSteganography and watermarking are different jobs
Pengfei Wang and coauthors, in their 2024 article Covert Communication through Robust Fragment Hiding in a Large Number of Images, separate the two goals directly: “Steganography schemes mainly focus on capacity, invisibility, and security, and their protected object is confidential information; watermarking schemes mainly focus on robustness and invisibility, and their protected object is the carrier.”
#1 Best Overall
- Used Book in Good Condition
That split explains the chat-app problem. A scheme tuned for capacity and secrecy is not designed to survive re-encoding, while a watermark built to persist through processing trades some capacity for durability. Real designs often borrow from both, so the label alone does not predict behavior.
What the measurements show
The four studies below use different images, algorithms, and test conditions. Read each row on its own terms. The numbers are not a ranking, and the cells marked “not stated” are values the accessible sources do not report.
| Study | Method | Images and route | Payload per image | Reported result | Caveats |
|---|---|---|---|---|---|
| Fitriyani and Fachri, 2026 (Jurnal Teknologi Informasi dan Multimedia, 25 May 2026) | LSB | 15 PNG, BMP, and JPG images sent through Telegram as a document and as an image | Not stated | Document route: 100% extraction, bit-error rate 0. Image route: extraction failed. | The authors attribute the image-route failure to compression. The journal page states that the experiment’s data were not yet downloadable. |
| Bunzel, Chen, and Steinebach, 2022 (ARES 2022) | F5 as a proof of concept | Telegram, with settings chosen around its API limits | Average 81 KB per image at 2560 × 2560 pixels and JPEG quality 82 (reported optimum) | Not stated as a recovery rate in the accessible record | F5 is a JPEG-domain method, not LSB. The optimum is a setting from this study, not a published Telegram specification. |
| Wang and coauthors, 2024 | Robust fragment hiding across a large image database, using DFT and DCT stages, redundancy, and coding | Rotation, scaling, cropping, and a combined attack of JPEG quality factor 80 plus cropping | Limited per image; no figure stated in the accessible text | Full recovery under the tested attacks; average PSNR 41 dB; full recovery with up to 30% of images lost; 61.1% recovery with 80% lost | Results describe the authors’ own scheme and setup. Weak to contrast and luminance changes, which the authors call out directly. |
| IEEE conference authors, 2026 | Hybrid DCT and Reed-Solomon coding | 50 natural HD images sent through Telegram photo mode | Not stated | Recovery when resolution was preserved | Sensitive to resizing. Only the abstract was accessible, so the full protocol is not stated. |
The Telegram LSB result: the sending route decides
The 2026 study is the only one here that tests LSB in a chat app, and its key finding is about route rather than app. The same 15 images behaved differently when sent as a document and when sent as an image. The authors’ explanation for the failure on the image route is compression. The study does not establish what happens on other Telegram clients, other versions, or other apps, and it does not cover WhatsApp.
PC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchThe 2022 Telegram study: settings matter more than the method
Bunzel, Chen, and Steinebach used F5 to probe what Telegram’s limits allowed. Their most useful output for a reader is the image geometry and quality setting they found workable. It is evidence about one embedding method under one set of conditions, not a comparison with LSB.
Rank #3
The 2024 robust scheme: spread the payload to survive losses
Wang and coauthors do not hide the whole message in one image. They split fragments across many carriers, add redundancy, and encode the pieces so that losing some images or some signal does not destroy the message. The transform-domain stages are what give the design resistance to geometric changes and recompression in their tests. The trade-off is visible in the table: capacity per image is limited, and the design is vulnerable to contrast and luminance changes.
The 2026 hybrid method: promising, but resize-sensitive
The IEEE abstract describes combining DCT-domain embedding with Reed-Solomon error correction. Error correction lets a receiver rebuild data that has been partly corrupted, which is the core reason such schemes can recover where raw LSB fails. The abstract reports that recovery depended on preserving resolution, so any app or workflow that rescales images removes that advantage.
Why capacity and robustness pull against each other
Every extra layer of protection costs room. Redundancy repeats data, error correction adds parity bits, and spreading fragments over many images means each image carries only a small piece of the message. The 2024 authors state that their scheme’s hiding capacity per image is limited, and the design shows that a system can gain resilience to losing whole images at the price of per-image payload.
The same design does not solve every failure. The authors state: “Our watermarking scheme is not very robust to contrast and luminance changes.” A chat app that alters tone curves, brightness, or contrast during processing could therefore affect a watermark even when its geometry survives.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Does a robust watermark survive a chat app?
It can survive specific transformations under specific tests. Wang and coauthors report full recovery under rotation, scaling, cropping, and a combined JPEG quality factor 80 plus cropping attack. Those results do not show that the same watermark survives every quality setting, every resize ratio, or every chat app’s pipeline. The safe reading is that robustness is a property of the algorithm, the attack, and the route together, and a claim about one of them does not transfer to the others.
How to test a chat route yourself
If you need to know whether a particular app and route preserve hidden data, measure it rather than assume it. A simple protocol:
- Create a test set of at least 15 images in each format you plan to use (PNG, BMP, JPG), at the resolution you intend to send.
- Embed a known bit pattern with the method you are testing, and save the embedded files. Record each file’s dimensions and size.
- Send each file through each route you care about: as a document or file attachment, and as a photo or image. Record the app version and client (phone or desktop) for each run.
- Download the received file on the receiving device and save it without editing. Confirm whether its dimensions and size changed.
- Run extraction on the received copy and compare the recovered bits with the known pattern. Record the extraction rate and the bit-error rate.
- Repeat with a second sample before drawing conclusions. A single successful or failed run does not establish that a route is reliable.
Write up each result with the app, client version, route, image format, and whether the resolution changed. Results recorded without those details cannot be compared with other tests.
Free tools Windows power users keep installed
One-click scans. No signup required.
What the current evidence does and does not establish
- It establishes that, in one 2026 Telegram test of 15 images, LSB extraction depended on the sending route, with the document route succeeding and the image route failing.
- It does not establish a Telegram-wide rule. The sample is small, the journal page states the data were not yet downloadable, and the same test was not repeated across clients or versions in the accessible sources.
- It does not establish WhatsApp behavior. None of the reviewed studies covers WhatsApp’s image handling, so no conclusion about it follows from these results.
- It does not allow direct comparison of PSNR values from different studies, because each used its own images and protocol.
- It does not show that a robust watermark is secret. Surviving processing and resisting steganalysis are separate properties, and the 2024 authors treat them as separate design concerns.
No official chat-platform specification for current image processing was found in the accessible sources, so any statement about how an app treats uploads should be checked against the app’s own behavior under the conditions you use.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
