Fun fact: most FPV equipment alternates between even and odd fields for each frame it shows to the user. This is why you don't see the interlacing artifacts when there's motion on the screen. It's also why the "resolution" of analog feeds seem so low. It's also why the image bounces by one scanline if you look closely.
I came across your channel quite recently and I have to say that I really appreciate your explanatory series. I feel like this is exactly the content that was missing in the FPV community for so long because prior to your videos this technical information was hardly accessible and not very structured. Many other RU-vidrs therefore base their explanations on feel rather than facts which often leads to confusion and no technical understanding. So that is exactly why I enjoy your professional way of producing these videos. Keep up the great work!
Thanks for several "oh wow" moments in your video. I sort of knew how it worked at a high level, but had never dug into the details. Thanks for another great video. I'm looking forward to the DJI video one.
This is the best channel for newer pilots for sure. (Maybe experienced guys too, but im newer so idk) In 4 months ive learned how to tune a scratch built quad with a free education from here and others. I cant even imagine the time involved in learning all that on my own, thanks for condensing this stuff so that a simple pipefitter can go from zero to building and tuning quads , you are really doing a great thing here.
@@ChrisRosser thanks for the reply. I still have so much to learn but im not scared of anything quad related anymore. I have alot to learn about pids and blackbox but there is hours of videos for me, and until then im flying everyday with a good enough tune so no complaints 🚀🚀🚀
Thank you for another absolutely amazing and informative video. After your explanation, analog video signal conversion, transmission and reception technology amazes me even more, especially considering the fact that it has been researched such a long time ago!
Thank you for another great topic. You filled in some areas I either forgot or was hung over in my undergrad engineering classes. Please keep up your amazing videos.
Sir as always a wealth of concise information. Armed with this information, picking out equipment is much easier and a lot cheaper. No more guessing and trying things if you know what you want this information lets you choose the equipment that suits your needs best the first time. Mr. Rosser, I will become a Patreon in the new year, for both yourself and J Bardwell. Can not thank you enough for all the hard work. Happy Holidays
@@ChrisRosser I am looking forward to it. It is interesting that the analog signal is converted from digital then back to digital. I am interested in why the Dji system has latency issues. Thank you.
Great video again Chris. Didn't know I wanted to learn about this but hey all knowledge is good eh ! Looking forward to the DJI video, although Good as it is I haven't won the lottery yet. Keep it up my man. Analogue for the time being. 😁
Great content! More information laid out in clear form in one place. Very nice :-) (Joke) Question: "How does analogue video work in FPV?" Answer: "Badly"
This makes me wonder about having a digital interface between camera/goggle and vtx/vrx, but still analogue ota. Might cleanup the signal abit by reducing interference from the esc/fc. Analogue vtx with mipi in? Nudge nudge, TBS.
this proofs im full blown stupid xD how amazing it would have been having these kind of knowledge taught at school ... about applying mathi this way... so amazing.. one day i will make money out of the hobby and support your patreon.. keep on doing this!! :D
Thanks for the informative video! One question, any solution/chipset (Camera -> VTX -> Receiver) out there that can allow 120fps or higher frame rate? Are there actually other industries improving this technology other than FPV community?
Great question, the analogue video standard is incredibly prescriptive. You can have 525 lines at 60 fields per second or 625 lines at 50 fields per second. That is it. The cost to do anything else is prohibitive because you need to redesign the whole encoder and decoder from scratch and "ain't nobody got time for that" because this tech was developed in the 1940s and 1950s and no-one is working on it anymore! The only reason you have two options is that USA and EU decided to do things differently. DJI delivers 120fps using digital technology.
Another great video, Chris. I was wondering what the differences were between analog and digital, so it's great that you're starting this series. I have a PID question, unrelated to this video, that I hope you can answer: First, do higher kv motors have better PID handling/ are they easier to tune? My thinking is more power means the quad can respond to PID errors quicker. Second, if the previous is true, could (for example) having 1950kv motors with a throttle limit that mimics the feel of a 1750kv motor mean that a pilot could have the throttle feel that they would prefer while also having superior PID control? The reason I ask is I currently run a 2550kv motor with a motor output limit of 70%. But if I could raise that to 75% and have a throttle limit of 5%, I wonder if the quad would indeed fly better. I tested this myself, but frankly, I don't feel a difference. Perhaps there is a difference, but at my skill level, I can't tell. I don't have blackbox logging, so maybe you could see whether there truly is a difference or not. Looking forward to your response. Whether I'm right or wrong, it will be fascinating to hear what you have to say
Handling is mostly down to motor mechanical torque at low RPM which is to do with motor stator volume. So bigger motors handle better and KV doesn't really matter (in fact too much KV is bad news for handling). I have a video on choosing motors that goes into more detail.
This is a cool video series! Also interesting to point out, that analog video is just digital video transmitted via an analog signal. Didn't know that the actual transmission is frequency modulated only, but I now understand why we got this black and white image in times. That's actually a nice built in failsafe! Could that CCD and global shutter thing actually be a reason, steele likes his ccd so much? I want to do an A+B test now :P
Umm, I think Steele just prefers certain gear because he's used to it not because of any significant performance benefit vs any other setup. CCD sensors are arguably worse than CMOS sensors for FPV because of the lack of WDR and other fancy features.
Very informative and educational video - thanks! When analogue FPV cameras are advertised as 800TVL or 1200TVL what does this really mean? Is it the initial (digital) image resolution supplied to the image processing chip and is more TVL necessarily better? I'm trying to understand the factors determining the final image resolution you see in your goggles.
Yes the 800 or 1200TVL is the resolution of the digital sensor in the camera. That is the resolution that is then downsampled to standard analgoue resolution. No matter the number of TVL in the camera the resolution of the video is always 525 lines for NTSC and 625 lines for PAL.
The analogue info on the CMOS sensor is just the charge on every R,G and B element in the display (all ~3 million of them!) . The analogue signal from the ISP has had a lot of image processing done has been converted to a single analogue voltage on the wire formatted in a certain way.
Unfortunately not. There were some very old protocols used with higher resolution but they were discontinued decades ago and components are no longer in production. In the future we may struggle to source components for analogue video systems now that CCTV is going digital...
Just curious, why isn’t there an HD analog used in FPV then? If the latency of analog is say 15ms and NTSC/PAL is ~500p then you can send a ~1080p analog video in about double the time which would be comparable to DJI latency and a 720p video in barely any more time.
You could transmit twice the resolution but then you would only get half the frame rate (30 fields per second) 15 frames per second. Because you can only send a certain number of lines per second.
I doubt analogue video in ntsc/pal in our goggles is only 30/25 fps. When I'm flying the frame rate looks silky smooth to me, and when watching back the dvr in goggles or on my computer, it is no longer so smooth, it becomes very choppy. On my pc the pal files have a frame rate of 25fps. So I assume in goggles while flying we are seeing a higher frame rate, 60/50 fps I assume. thanks for the video :)
The goggle screen will update 50/60 times per second but only half the lines will actually change! This actually looks smoother (to my eye at least) than a full frame update only 25/30 times a second.
@@ChrisRosser That is not true. The interlaced signal is deinterlaced to 50 or 59.94 progressive frames and all the frames are displayed. So all the lines of the image change every 1/50 or 1/59.94s. This is exactly the same process when you properly process 59.94i DVR recording to 59.94p distribution format. Example: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-Na9MdQliljU.html
you are right. You see fully deinterlaced 50p or 59.94p video in your goggles. The downside of most DVRs are that they deinterlace to 25p or 29.97p, so they throw out the half of the information.
PAL might be part of your problem.. Most computer monitors don't run at an multiple of 25fps, which will result in choppy playback. 30fps NTSC DVR will likely look much better on your PC - albeit still not as silky smooth as 60 fields or frames per sec!
@@ChrisRosser Most of the panel control boards incorporate advanced deinterlace mechanisms similar to yadif, bwdif, tomsmocomp, etc. Even the most stupid bob deinterlace is way better that what you describe which is weaving resulting with lost of artifacts that we simply don't see on most goggles.
Because its not that much easier to do SD than HD for digital video! And there would be little benefit to the consumer to justify the development cost.
how to make video 10 times shorter and in one episode: analog signal which is transferred through the air is a very complex signal, it has sinusoids which carry information about colours and brightness and impulses which carry information about frame and lines. Digital signal is just impulses with high frequency. Impulses are easy to synchronise on Rx and TX, but sinusoids are not available to synchronise. So when you have some obstacles the analog signal from tx reflects from them and Rx receives a great mess of sinusoids with high amplitude of main signal and a little bit lower amplitude signals from reflections which are out of phase with main signal (reflected signal is mainly inverted signal and that causes interference of sinusoid or a previous frame which came to the Rx at the same time with current frame but because it's had to move longer distance because of reflections than the main signal). So digital is better))
