Thursday, November 26, 2015

Thanksgiving ATSC Tests

Yesterday I did some more ATSC experiments.


(above) Finally... the elusive WZRB ION channel 47 comes in. On the right edge is WACH Fox 57.

(I'm using the older copy of SDRSharp I mentioned in my previous post)

Since hard drive recording is spotty, I copied SDRSharp to a RAM drive, since it only records to its own program directory. SDRSharp kept setting the SDRPlay's PPM correction to a crazy value and changing it kept locking up the computer.

SoftPerfect RAM Disk also kept locking up when I tried to unmount disks. I would turn it off in Task Manager and restart it from the Start menu. That caused it to forget the RAM it had previously taken and I would have to reboot. I eventually switched to IMDisk and kept working.

Adding to my troubles, when I tried to decode it, I accidentally used the old flowgraph from my previous post, the one that expects 8-bit samples. Of course, my TS file stayed empty, and it took me a while to realize the issue. (Pictured below)



Even after fixing this, I wasn't able to decode WZRB, even though the TV had no trouble. However, I did manage to get a long successful recording of WRLK PBS.

Pictured below is WAGT NBC. As I write this, I can't get it to come in on my TV.

I've seen this kind of fading in other stations, most notable PBS. This picture was taken with my SDRPlay hooked to a rooftop Yagi. For some reason, a few cases of this fading can be cured by using indoor rabbit ears and rotating them until it goes away. Strangely, I've actually had far better results on PBS with indoor rabbit ears than an outdoor Yagi. I don't yet know why certain portions of the station would fade before reaching me, considering that it leaves the transmitter as a uniform distribution of power over the whole 6 MHz. I know that ATSC doesn't work for moving receivers, but I'm not moving, so why would only certain sections fade as opposed to others?

In summary, trying to decode ATSC is more of a hit-and-miss process since it's hard to know if a station will be readable. In several situations, software-based solutions perform better than embedded hardware, as is the case with DRM radio reception. But from what I've seen, if a TV can't receive a station, GNURadio certainly won't. Sometimes even if a TV is perfectly OK with a certain station GNURadio still can't decode it. I wonder if they'll update GNURadio's blocks for better results? In theory you should be able to decode just as well (if not better) in software as you can with a hardware decoder, since you have more control over processing.

No comments:

Post a Comment