Micro Cam Footage Restoration Workflow - Where to Start?

[taigi]

I'm initiating a project to restore footage from older, low-cost cameras, beginning with a micro cam circa 2010.
This particular camera is characterized by its rudimentary design: a non-standard lens with a 1.2 mm diameter, minimal battery capacity, compact form, low cost, inherent instability, and complete lack of user-adjustable settings. The camera was manufactured in China, and its name was never known and is still unknown.
The attached video is a direct stream copy created using VirtualDub in 2013.
Don't be alarmed by what you see; it's an art action, limited view.
I'm seeking guidance on the optimal restoration workflow.
MediaInfo output is pasted below:

General
Complete name : /home/oem/Downloads/2013-Sep-28.avi
Format : AVI
Format/Info : Audio Video Interleave
Format settings : BitmapInfoHeader / PcmWaveformat
File size : 55.0 MiB
Duration : 1 min 31 s
Overall bit rate : 5 048 kb/s
Frame rate : 30.000 FPS
Director : Generplus
Original source form/Distributed by : Generplus
Recorded date : 2010-06-29
Writing library : VirtualDub build 32842/release
Copyright : Generplus

Video
ID : 0
Format : JPEG
Codec ID : MJPG
Duration : 1 min 31 s
Bit rate : 4 683 kb/s
Width : 720 pixels
Height : 480 pixels
Display aspect ratio : 3:2
Frame rate : 30.000 FPS
Color space : YUV
Chroma subsampling : 4:2:2
Bit depth : 8 bits
Compression mode : Lossy
Bits/(Pixel*Frame) : 0.452
Stream size : 51.0 MiB (93%)

Audio
ID : 1
Format : PCM
Format settings : Little / Signed
Codec ID : 1
Duration : 1 min 31 s
Bit rate mode : Constant
Bit rate : 352.8 kb/s
Channel(s) : 1 channel
Sampling rate : 22.05 kHz
Bit depth : 16 bits
Stream size : 3.84 MiB (7%)
Alignment : Aligned on interleaves
Interleave, duration : 34 ms (1.01 video frame)
Interleave, preload duration : 500 ms

[taigi]

Would this be a suitable starting point?:

AVISource("2013-Sep-28.avi", audio=true)
ConvertToYV12() # Convert color space
QTGMC(Preset="Medium", InputType=2) # Denoise/stabilize for progressive footage
ColorYUV(autogain=false, gamma_y=0) # Adjust color
Tweak(sat=1.15) # Boost saturation
LSFmod(strength=50) # Gentle sharpening
Return(last)

[Selur]

Since the source does not really show details and tons of macro blocking, I have no clue why you would apply sharpening, since at least the sample, to me, doesn't show anything I would sharpen.

 ColorYUV(autogain=false, gamma_y=0) # Adjust color

That shouldn't do anything,...

=> don't really see what you are going for,... I blindly played a bit with it.

I first threw BasicVSR++ (not available in Avisynth) at it.
That totally killed the blocking and smoothed the fog a lot.
Then I tried some denoising with DPIR and smoothing the chroma with fft3dfilter.

Sorry, this all probably doesn't really help.

Cu Selur

[taigi]

@Selur, thank you. I am very impressed with how well the scripts cleaned the background of the video.
I may explore the possibility with something less harsh, as shirt really can't be so clean and ironed.

[Selur]

You can always add some grain afterward.