The db scale is logarithmic but doesn't match how our eyes sense light. So small changes in db may result in large visible light changes, or vise versa. Have you ever looked at a raw image, it doesn't look correct untill a gamma is applied, even though roughly the same information exists in both raw and gamma images. This is a similar concept.
To do do this: db > linear > gamma
This can be precomputed into a lookup table with the number of entries being the number of current steps, 8-bit 256 entries.
No influencer school, no video experience whatsoever actually
I watched TONs of youtube videos growing up and I guess I took inspiration from what I liked in each: fast pace, always shows what you are talking about on screen (no talking head), include tiny projects in the main one, explain the science, use an intro to captivate and an outro to nail the point, use music to drive rythme
Thanks! That kind of makes sense. It's always interesting to me when I see patterns that make videos work well, and I usually have a lot of questions about the production.
Another question that I hope is not disrespectful: does PCBWay and JLCPCB pay for brand placement during the video or was it just a tribute from your side?
My wife has worked in marketing for years and setting up photo shoots/video shoots has changed massively. A huge amount of the video and pictures that don't require a specialized lense are just done with feature phones these days.
Seems like it wouldn't be hard to filter this for something like the 2.4 or 5GHz range then put a different color filament at the demarcation point for each of the channels.
Yes each LED channel has an inductor. I needed this because I wanted the LED to be constant current driven, to reduce flicker and improve their longevity
amazing project ! well done ! inspiring to see it this morning the youtube algo recommended it for me. I was thinking what would be a budget way of creating this ? maybe a $20 SDR instead of the hack rf , seems to be a waste of the hack rf to use it as a lamp :)) I think the way you did must have cost you over $1k. Thanks for the inspiration !
I do think most people implementing something like this, especially in the exact way this one was implemented would really just be interested in the Wi-Fi ranges.
When I was doing wireless stuff we'd show customers the output of RF analysers to show that even with their wifi turned off their spectrum was packed full of noise.
Having a large viceral display of this would quickly enlighten them.
You do not need a full SDR to make a waterfall. The nrf52840 can function as a coarse spectrum analyzer, because it lets you tune to any frequency within its range, and then measure the approximate received signal strength over a period of time[0]. I have tested this myself, and it works pretty well. The only downside compared to an SDR is that it can only listen to one frequency at a time, so it will sometimes miss short signals.
I know it can switch frequencies, but I predict it wouldn’t make a very pleasing visualization if the updates are slow, but maybe you are right it would look fine for this implementation which is already far off from a waterfall
I'm personally especially interested in 'Latent Reflection'. I've tried to make something similar never got to a point where I was happy with the output the AI model gave me.
Lots of tuning to get the model to not immediatly spiral into nonsense. But small models are getting better by the minute, maybe i'll revisit it with a better model and share all the code
reply