[hamdani yusuf (OP)]

Evidence.

What kind of evidence do you need?
Can you be more specific?

[hamdani yusuf (OP)]

But you probably need to start by saying what you think quantization means.

I asked Google Bard. I think it gives pretty good answer.

Quantization is the process of converting a continuous signal into a discrete signal. This is done by mapping the continuous signal to a finite set of values. Quantization is used in a variety of applications, including digital signal processing, image processing, and machine learning.

There are two main types of quantization:

Uniform quantization: This is the simplest type of quantization, where the continuous signal is divided into equal intervals and each interval is mapped to a unique value.
Non-uniform quantization: This type of quantization is used when the continuous signal has a non-uniform distribution. In this case, the intervals are sized differently to better represent the distribution of the signal.
Quantization can be a lossy process, meaning that some information is lost in the conversion from a continuous signal to a discrete signal. However, the amount of information lost can be controlled by the number of quantization levels used. More quantization levels will result in less information loss, but will also require more storage space.

Here are some examples of quantization:

When you take a digital photo, the camera quantizes the continuous light signal into a discrete set of pixel values.
When you encode an audio signal for MP3 playback, the audio signal is quantized to reduce the file size.
When you train a machine learning model, the model's weights are quantized to reduce the model's size and improve its performance on embedded devices.
Quantization is a powerful tool that can be used to improve the efficiency and performance of a variety of applications. However, it is important to understand the trade-offs involved in quantization before using it.

Digital cameras and human eyes have inherent structures which cause spatial quantization. They can't detect the image of an object that can't activate one pixel.

[hamdani yusuf (OP)]

Here are some more screenshots.

This one shows the sparkling effect with various size.


This one shows the sparkling effect while also showing some diffraction effect. The small bright spots don't seem to show the diffraction effect.

[alancalverd]

Use a calorimeter or photocell to determine E.

I'm sure I heard about someone doing that.
It gave the wrong answer.
I think it's because they forgot about chemistry.

Not entirely.

I built an x-ray calorimeter with graphite because it has about one sixth of the specific heat capacity of water (the target material for radiotherapy) and thus a larger temperature rise per unit dose, making the electronics fairly simple*. This is the UK primary measurement standard, and similar devices are used in all national standards labs, but oncologists would like to have a direct measure of dose to water rather than use theoretical conversion factors.

Then several labs built water calorimeters with more sensitive thermometry, and discovered that the discrepancy between water and graphite measurements was (a) greater than the error range of the conversion factors and (b) decreased at very high continuous doses, so AFAIK (I haven't worked on calorimetry for several years) the decision was to stick with graphite and physics rather than try to understand and circumvent the chemistry.

*we needed to measure 0.001K temperature rise to better than +/- 0.000001K. I guess it's an idiosyncratic interpretation of "simple", but it was still easier than resolving 0.0002K to the same  uncertainty.   

[hamdani yusuf (OP)]

Then several labs built water calorimeters

What was the expected advantage of water calorimeters over the graphite?

[hamdani yusuf (OP)]

There are several modes of sparkling effect that can be observed from the videos. But basically, it's caused by reflection of bright light source by rippling surface of water, which produces several images of that light source on the surface of photosensitive detector. Depending on the curvature of the ripples, and the distance between the reflecting surface and the detectors, the size of the image formed on the detector may vary.
The image of an object reflected by a convex surface is smaller than the same image of the same object reflected by a flat surface. A concave surface may produce larger image at certain distance, but for significantly larger distance than the radius of the curvature, it also produces smaller image. At large enough distance, the image can be even smaller than the size of a single pixel in the photosensor. But since the light source is excessively bright, it can still saturate that photosensor pixel.
These diagrams show some of the modes. Each grey rectangle represents a pixel of a photosensor. The first may explain how a single bright spot can be seen steadily lit up for several frames until it suddenly turn off.

[hamdani yusuf (OP)]

Image size and intensity also affect the recorded image on the sensor.

[hamdani yusuf (OP)]

Finally, here's how diffraction effect doesn't seem to affect small sparkling spots.

[alancalverd]

Then several labs built water calorimeters

What was the expected advantage of water calorimeters over the graphite?

The effect of radiation on living tissue is principally due to its interaction with cellular cytoplasm - mostly water. So for radiotherapy, in particular, we would like to know the dose (amount of energy deposited per unit mass) delivered to water by our radiation beam. We can make a theoretical estimate from the electron density of graphite but it would have been nice to eliminate all the approximations inherent in that calculation.

I did some work with very large initial doses and doserates, so that the water had reached something of a chemical equilibrium before the actual measurement phase, but the results were still too variable to be clinically useful. Pity, because the apparatus was delightfully simple - just a bucket and a thermistor!

[alancalverd]

Your pixels seem to be binary, whereas real camera pixels are linear. 

[hamdani yusuf (OP)]

Your pixels seem to be binary, whereas real camera pixels are linear. 

I think you missed to see my previous post.

Image size and intensity also affect the recorded image on the sensor.

[hamdani yusuf (OP)]

The pictures I posted here are screenshots of my next videos. The first will show a closer look to the phenomenon, while the second one tries to offer some explanations.
The existence of several effects at once in a single piece of evidence may prevent us from identifying the phenomenon in the first place.

I'd like to have some constructive feedback to improve my videos before uploading them. Does anyone notice some unusual effects I haven't mentioned yet? Or think that the effects I did mention here implausible?

[hamdani yusuf (OP)]

Here it is. A closer look at sparkling water surface.

[alancalverd]

Everything you see is entirely plausible. Not every explanatory hypothesis is plausible. That's science!

[Bored chemist]

Here it is. A closer look at sparkling water surface.

What needs explaining?

[hamdani yusuf (OP)]

Everything you see is entirely plausible. Not every explanatory hypothesis is plausible. That's science!

Do you have an explanatory hypothesis for the observation?

[hamdani yusuf (OP)]

What needs explaining?

Anything that's seem unusual from a typical image reflected by water surface.

[Origin]

Anything that's seem unusual from a typical image reflected by water surface.

Nothing is unusual, so no explanation is necessary.  Easy-peasy.

[Bored chemist]

Do you have an explanatory hypothesis for the observation?

Water reflect light.
The surface of water is disturbed by things like the wind and fountains.

What else needs explaining?

[Bored chemist]

Anything that's seem unusual

What do you think is unusual?
Everyone else here seem to think that what you see is exactly what we would expect to see.