I wonder who was drunk while designing this camera, that made them so inefficient. The camera runs on 2x AAA batteries, and the life is about 5 or 7 minutes. I try Heavy duty, alcaline and many rechargable batteries, but doesn't last more than 7 minutes. That's an old camera, so at the time lithium batteries wasn't as common as now.

My Canon PowerShot SX150 IS, began to do this problem at its end of life. It used 2 AA batteries, and I had to use two Fujitsu low self discharge NiMH batteries, to have longer battery life from this camera.

I wonder who was drunk while designing this camera, that made them so inefficient. The camera runs on 2x AAA batteries, and the life is about 5 or 7 minutes. I try Heavy duty, alcaline and many rechargable batteries, but doesn't last more than 7 minutes. That's an old camera, so at the time lithium batteries wasn't as common as now.

The motivation was to squeeze everything into as small as possible camera body, it was the era of "the smaller the better" hype (until people realized you can not make decent quality photo if you can not get good grip...).
The consequence was, the cameras draw ~2A current bursts, which very few battery types could handle. Definitely out of reach for even the "heavy duty" Carbon-Zinc, only about 5..10 shots possible from alkalines

I had Nikkon Coolpix (now it is broken) 3100 from this era, it was originally shipped with a Lithium CR-V3 battery (it was the recommended battery type, rated for 300+ pictures, the NiMHs were just "capable", other were really mentioned as "don't use, not able to work with"; contained two 14500 size 3V single use Li cells in parallel in one package), could use two HR6's (capable of barely 100 pictures per charge), but these last just few 10's of charging cycles before reaching the maximum tolerated internal resistance limit (the high current both caused the accelerated wear, as well as meant it tolerated only limited wear). The Eneloops seem to last longer (they had lower internal resistance, which was the main difference), but still not that much impressive (barely 100 cycles, but nearly 150 shots per charge).

What matters a lot are the convenience features, namely the continuous focusing (it consumes about 1A continually when activated), bright screen (not much efficient backlighting back then; about 0.3A).

This was the reason, along the requirements for a good grip from the market mandating bigger camera body, most makers switched to 4 cells supply (so all currents become half - that meant quadruple life of the cells charge and even their number of charging cycles)

My camera doesn't do much but still "short circuit" my batteries. The video recording codec is MJPEG at 640x480 very compressed and some kind of PCM mono audio all in an AVI container, that format doesn't need much prosessor. The picture quality is too compressed, I think 50 or 60% JPEG at 8MP. The focus is fix, the zoom is digital, the screen have an L*D backlight and you can apply some basic fancy color filters. For those bare functions the camera should not waste much power, even I don't feel any heat from the camera processor. That camera was a total degradation.

A bit off topic but related with cameras: Why cameras from 2010 and past burns a vertical line when you point the camera directly to a lighting source?

Because of the use of CCD sensor. This is its beat effects.

Because of the use of CCD sensor. This is its beat effects.

Actually CCD sensors are practically not used anymore since quite some time even before 2010.
They got replaced by CMOS sensors. First motivation was to make it cheaper. CCD process is an odd-ball, not used for anything else (they were used for some discrete time domain analog filters and delay lines, but these got replaced by digital processing), so only very few foundries are able to produce anything in it and in most cases they use quite old, small wafer so very expensive production.
The CMOS is used for virtually everything, it is optimized for large gate count digital (so computers, memories, digital processing), it is under constant development inctreasing the density on the chip at the same time moves production to large size wafers, both leading to drastic cost reductions. So although CMOS is not inherently as optimized for analog or image sensing (it needs more clever design techniques and usually larger complexity for the same analog or sensor function), because it is mainstream, at the end it becomes the cheapest use it for the given design.

And later image sensor development took advantage of the CMOS process: Allows more advanced signal processing within small subsections of the image sensor array (e.g. dynamic shutter time over the image field improving e.g. dynamic range between bright vs dark parts of the image or mask out bad sensor sub-pixels). These techniques were initially used to "just" compensate for the difference in the inherent properties of the main sensor elements (CCD inherently offers greater dynamic range so did not need that processing), but over the time the technology progresed so far these local correction circuits lead to the CMOS offering way better picture quality than attainable with CCD (e.g. pixel resolution - for CCD it means extremely long analog shift registers, capturing way too much noise when shifting the image data out of the chip, vs possibility to use many ADCs across the image field integrated into the sensor chip, as CMOS allows that; the possibility of local exposure control with CMOS means even when the sensor element may be worse, the overall dynamic range over the single picture could be way better than possible with CCD, where such local corrections are not possible to implement at all).

As a result, CCD are used only in really special applications and even there the CMOS are taking over...

A bit off topic but related with cameras: Why cameras from 2010 and past burns a vertical line when you point the camera directly to a lighting source?

The local overload may block the signal collection routed across the overloaded pixel, so blocking the whole line/column (depends, how the sensor is organized).
But there are design techniques to handle this problem, apparently the designers of your sensor did not use them (because it is either old or cheapeese design chip).
When the light source is really intense (e.g. the direct sun), it may even overheat the chip there and so cause a local latch-up (the CMOS structure tends to form a parasitic thyristr like structure, which when activated, stays on till the power gets interrupted; it kills all signals "going around"; normally the only solution is to make sure it never activates, in some chips where it is not avoidable  extra transistor switches are used in the internal supply lines designed to break the supply periodically, so that way shut these parasitic thyristors off before they damage the chip; but this does not recover the picture signals), again junking all signals that have to pass over the affected spot, so usually the whole column or line.

Interestingly, the same effect easily happens in the "preview" (the live display on the LCD screen) but never in the final photo. Is the sensor switched between different operation modes ?

By some reason my cheap sport camera from 2016 is generating a weird labyrinth pattern when it's exposed to bright colors like the sky.

Could be these older, digital cameras have failed electrolytics in them. Instead of the electrolytic cap exploding (possibly due to limited current from batteries) it just kills the batteries in such short time.

I have a XBOX from early 2000's and had to replace many of the electrolytics in it recently.

The one with the very obvious labyrnth looks like a glitch of the JPEG algorythm. Anything with just a "grid" could be few things, including scaling of a picture that have labyrinth in it to a different resolution (aliasing)

I shooted the picture at 8MP, the original resolution of the camera, so the internal processor doesn't do any resizing. What you see in "Zoomed in picture" is a section of the picture of 1440x1080 resolution resized with paint to 640x480, wich doesn't have any anti aliasing. What you see in "1X" is how the pattern looks in a 100% scale in the screen, and "2X" a 200% scale. What I wonder is if that's a glitch of the sensors or of the internal processor compressing the image.

By some reason my cheap sport camera from 2016 is generating a weird labyrinth pattern when it's exposed to bright colors like the sky.

Is this pattern constant in all images?
If no, it can maybe be some dithering.

Areas like the sky in the pic always have a noise pattern in them, used for the dithering. Without it there would be edges visible between areas with similar colors (like in very early digital/digitized photos from 30 years ago)

I guess that there is some use of a random number generator to generate the noise. But what if it failed and essentially returns 0 every time it is called ? The algorythm then will still try to do its thing (encoding deltas between pixel colors) but without noise....

Also, what if in this camera it is based on RTC and the RTC doesnt work due to a dead battery ?