The Tech Blog

Advanced Tech: SynthID detection

Posted under Advanced Tech, Educational on May 31st, 2026 by Amaldev

I wanted to discuss SynthID today because AI image detection has reached a level where even pros aren’t able to detect if its real or not. Its bound to get worse over time. Google launched SynthID in 2023 for AI-generated images. Now the ecosystem is expanding with OpenAI and NVIDIA adopting the same.

So SynthID is an invisible watermark for AI-generated content. Google built it for images, text, audio and video, but the embedding method changes for each medium. For today, let us stick only to images. SynthID adds a weak hidden signal into image pixel data. You do not see it, but a detector trained for that signal can scan the image and return a confidence score.

Technically, SynthID-Image is described as a post-generation encoder-decoder system. The AI model first creates the image. A neural encoder then makes tiny, distributed pixel changes while keeping the image visually aligned with the original. A neural decoder later checks whether that hidden pattern is still present. Its a trade off basically. If the signal is too strong, image quality drops. If it is too weak, JPEG compression, resizing, filters or cropping can damage it.

Google’s paper gives the general design idea but does not publish the exact encoder architecture, secret keys, threshold logic used in production. Based on what I have seen and tested, it’s clear that the watermark is content-dependent and tested against real image edits like compression, resizing, cropping, rotation, blur, noise etc. The detector is trained to recover that weak signal without needing the original image.

BTW, you can now upload supported media to the Gemini SynthID Detector or OpenAI’s verification tool to check for these signals. So you really can’t pass off images as real from the top AI image model providers. The Chinese and open-weight models haven’t implemented them yet.

If there is interest in this, I will cover the latest reverse-engineering work on SynthID next, specifically how some folks are trying to reverse these watermarks. Let me know.

If you liked the post, Share it with your friends!

Tech Explained: BLE 6.3 Inline PCT Transfer

Posted under Educational, Tech Explained on May 24th, 2026 by Amaldev

Bluetooth 6.3 spec was released a couple of weeks back, and I finally got time yesterday to go through the changes. The most interesting one is Inline PCT Transfer. Lets discuss that today.

I had written about Channel Sounding for distance measurement in older posts. In that, 2 BLE devices exchange known radio tones and watch how the phase shifts across frequencies. That phase shift carries distance information, so it can do better than signal-strength guessing.

Now the tricky part is the oscillator error. Each device has its own local oscillator. When the reflector receives a tone, the measured phase mixes the useful air-path phase with the phase difference between the two oscillators. Earlier, the reflector reported Phase Correction Term(PCT) data back. It used I and Q values, which describe amplitude and phase. The initiator combined its own phase reading with the reflector’s data and cancelled the oscillator error digitally.

BLE 6.3 moves that correction closer to the radio itself. With Inline PCT Transfer, the reflector looks at the phase of the tone it received and uses that information to shift the phase of the tone it sends back. So instead of sending all the correction details to the initiator for later math, part of the correction is built into the returned signal.

When that signal comes back, the initiator gets a cleaner phase measurement. It effectively sees the phase change from the signal travelling to the reflector and coming back again, so the useful channel phase appears doubled. That is helpful because the initiator needs less extra PCT data over HCI, which is the interface between the Bluetooth host and controller. So what’s the use case? Less reported data means less processing, less procedure time, and potentially lower power for repeated ranging.

The other BLE 6.3 changes are useful too. PHY-specific RTT accuracy improves timing and BR/EDR RF limits are relaxed for dual-mode radio design. Overall 6.3 is a good update.

If you liked the post, Share it with your friends!