Idea for creating a cloud capturing solution

[aminadav]

As you know, everything those days goes to the cloud.
And I thought about a new idea.

Users will buy small hardware for less than $50 that just converts an analog signal to bytes.
If it NOT do any processing. Only sampling the analog. It will support many inputs, composite, s-video, RF and others.

Those bytes can't be viewable since it does not encode it yet to pixels and frames. But those bytes can be uploaded as is it to the cloud.

Then on the cloud will be a pay-as-you-go service that has TCB, and many other fixes and gives the users the video in the highest possible quality for an affordable price.

Users will not need to buy expensive equipment, just a convert analog-to-digital, and the cloud will do all the hard work of understanding the signal and created video from it.

The good thing is that during days if the AI improves, new algorithms can use to process the same RAW signal for better videos.

What do you think about such an idea?

[dellsam34]

Show me a hardware that can do all that for $50 without any problems, And how do you upload a 100GB/hour footage and who would want to download that much data anyway? And if the customer will be mailing in the tapes anyway why don't you just send him the footage with the tapes on a flash drive and be done with it? This does not sound like an idea at all.

[lollo]

The price of a high-end S-VHS player is easily 10/20 times your 50$ budget. The you add a good capture card.

The levels adjustment must be done in the analog domain; the lineTBC and frameTBC corrections must be done in the near-analog domain, in the sense that the signal is A/D converted by the devices and often converted back to analog (specific card may process this data flow internally without D/A conversion). Accessing and storing the intermediate bytes in a cloud is a hard task.

4:2:2 YUV uncompressed data requires lot of space.

A card converting analog to "generic" bytes does not exist.

This does not sound like an idea at all

[aminadav]

Thanks lollo. You saved years of development
I didn't calculate the price of the player.
Regarding the card that does not exist, this is what I thought we might want to develop together.

But if it's an axiom that TBC must be done in the analog domain, so the idea is not relevant.

Do you have any link to "What exactly TBC's do to the signal..."

[lollo]

Do you have any link to "What exactly TBC's do to the signal..."

In short and not fully accurate: a TBC acts rebuilding all scan lines of an analog signal in order to provide a stable picture

Basic informations to start here: https://www.freevideoworkshop.com/time-base-corrector-tbc/ and here https://www.digitalfaq.com/forum/video-restore/2251-tbc-time-base.html

There is a lot information on the web and youtube, but a good book on Analog TV may better answer your specific technical questions

[Cornucopia]

You are throwing around words without understanding the implications behind them.

Just bytes? The cloud has improved in its bandwidth and throughput but it doesn't do miracles. Analog to digital means sampling, which means samples, aka pixels if you like (they aren't the exact same, but for discussion they can be).
Sampling without ANY processing means UNCOMPRESSED signal, so the birate is ginormous!
Data rate for uncompressed is color bit depth * Horizontal* Vertical * framerate, and for StdDef (D1) it is 275Mbps (V + A), and that's not even counting sampling the blanking, which it sounds like you want to do (for TBC purposes). So lets round up to say ~400Mbps.

Not even counting source playback device requirements, which has already been mentioned, those kinds of required guaranteed realtime data rates would bring your internet connection to its knees, even today.

A similar thing has already been discussed here on these threads before, back in ~2016 or so (about cloud editing, processing), and those same real world constraints are why it never got off the ground then either.

Facts. Science. You can pretend to ignore them but do so at your peril, like Wile E. Coyote realizing gravity.

Scott

[jagabo]

There are two projects that do this already -- except for the cloud aspect. One for VHS and another for Laserdisc. They record the entire analog signal including the active picture, front porch, back porch, sync pulse, all 525 (625 PAL) scan lines. Both of them have threads here.

[dellsam34]

I see this "re-inventing the wheel" threads all the time in A/V forums, Mostly because the thread posters don't have a complete knowledge about how the whole system works, Once they start learning slowly and information start to sink in drop by drop they come to their senses and start looking for a good capture device and a good VCR. Sometimes it's the other way around, out of frustration of finding good gear with decent price they start thinking about these ideas.