Page 1 of 2 12 LastLast
Results 1 to 10 of 15

Thread: Airspy (hardware) + ADSBSpy (software) out performs the FR24 receiver.

  1. #1
    First officer
    Join Date
    Dec 2012
    Posts
    452

    Airspy (hardware) + ADSBSpy (software) out performs the FR24 receiver.

    The Airspy, http://airspy.com/, is now providing better results on ADS-B reception than the standard FR24 receiver.
    Using this software: ADSBSpy, http://airspy.com/downloads/adsbspy-1.0.0.11.zip

    My own tests: FR24: 125 aircraft around 220nmi - Airspy 140 aircraft 230nmi.
    The Airspy is a clean receiver which uses no LNA.

    With a LNA some user has already reported 260nmi with an antenna inside the house behind a wall on the top floor: https://www.dropbox.com/s/ygr6wb0o8t...49.13.jpg?dl=0
    This is the result: http://i.imgur.com/UHvj3QX.png

    When using a preamp, http://ava.upuaut.net/store/index.ph...&product_id=85 the Airspy really becomes a premium ADS-B receiver.
    And of course, when you're not interested in ADS-B, you have a premium wide band receiver, 24 - 1800 MHz.

  2. #2
    Purser
    Join Date
    Mar 2015
    Location
    Mojave Desert, USA
    Posts
    133
    Jarod, I think you're comparing apples with oranges. I get about 20% more AC feeding FR24 box to VRS locally, about the same as you're getting with Airspy.
    F-KDAG1

  3. #3
    First officer
    Join Date
    Dec 2012
    Posts
    452
    Quote Originally Posted by Patrick Reeves View Post
    Jarod, I think you're comparing apples with oranges. I get about 20% more AC feeding FR24 box to VRS locally, about the same as you're getting with Airspy.
    ??
    I compared both. The airspy decodes faster, more and further. How is that comparing apples and oranges?

  4. #4
    Passenger
    Join Date
    Sep 2014
    Location
    Blue Mountains, Australia
    Posts
    32
    I have replaced my TV stick/Raspberry Pi module (T-YSCN8) with the Airspy via Windows 7. I get slightly more aircraft on the Airspy compared to the FR24 box (F-YSCN2) even though I am using a simple cantenna on the Airspy. I can imagine that the performance with the FR24 antenna will be even better.

    The FR24 box reports about 50% more messages decoded. This is probably due to the extra gain and lower noise from the antenna, but the Airspy beats the Raspberry Pi/TV dongle combination hands down.
    Last edited by stanno; 2015-05-05 at 10:42.

  5. #5
    Purser
    Join Date
    Mar 2015
    Location
    Mojave Desert, USA
    Posts
    133
    Quote Originally Posted by Jarod View Post
    ??
    I compared both. The airspy decodes faster, more and further. How is that comparing apples and oranges?
    Airspy is a multi-use wideband receiver. The FR24 box is a single purpose 1090MHZ receiver. If you scroll through the FR24 log you will see that about 20% of the aircraft are removed.

    2015-05-05 06:17:00 | [feed][i]removed 13 of 60 AC
    2015-05-05 06:22:00 | [feed][i]removed 15 of 51 AC
    2015-05-05 06:27:01 | [feed][i]removed 8 of 43 AC
    2015-05-05 06:32:03 | [feed][i]removed 11 of 45 AC


    Speed. The FR24 box ships the data to Sweden (in my case, 83.xxx.xx.xx - halfway around the world) merges the data (probably in MYSQL) then rebroadcasts it to the FR24 website. So you have to account for the latency when comparing your local Airspy traffic on your local host.

    Bottom line, I'm an Aviation buff, not a radio guy. I like watching AC traffic around the world. looking at local traffic is amusing, but the two aren't the same. Apples and oranges.
    F-KDAG1

  6. #6
    Passenger
    Join Date
    Sep 2014
    Location
    Blue Mountains, Australia
    Posts
    32
    I am comparing the data direct from my local decoders - apples and apples.

  7. #7
    Team FR24 Mike's Avatar
    Join Date
    Feb 2010
    Location
    Sweden
    Posts
    3,110
    FR24 receiver focus is quality and not quantity. There are so many different receivers and decoders out there giving very different results. I think most people has noticed that quality of tracks today is much better compared to 2-3-4 years ago and we will continue focusing on the data quality instead of getting the extra frame with possible bad data.

    You may also be comparing different numbers. For example Basestation software is showing number of tracked aircraft within the cache setting where default value is 60 seconds I think. Of course you will get a higher value if you count all contacts during 60 seconds, and contacts exactly right NOW that FR24 show.

  8. #8
    Passenger
    Join Date
    May 2015
    Posts
    4
    Quote Originally Posted by Mike View Post
    FR24 receiver focus is quality and not quantity. There are so many different receivers and decoders out there giving very different results. I think most people has noticed that quality of tracks today is much better compared to 2-3-4 years ago and we will continue focusing on the data quality instead of getting the extra frame with possible bad data.

    You may also be comparing different numbers. For example Basestation software is showing number of tracked within the cache setting where default value is 60 seconds I think. Of course you will get a higher value if you count all contacts during 60 seconds, and contacts exactly right NOW that FR24 show.
    Hi Mike et al.

    You're right. Many decoders just try to brute-force broken frames to get more throughput. This gives a lot of CRC valid garbage and it's not suited for a large network like FR24. What makes things even worse is the "confidence" setting that trusts ICAOs based solely on the number of occurrences, which, I agree, is not optimal. I did that on ADSB# like all other decoders for the el-cheapo dongles because there are no alternatives.
    Now things are different with Airspy :-) The detection process is 100% deterministic and makes full use of the 12bit @ 20MSPS sampling in a commodity CPU. I'm also concerned about data quality and I think the equation has changed quite a bit with the new algorithm. I invite you to give it a try and assess its suitability for serious ADSB work.
    During all the development phase, I was benchmarking against a FR24 receiver because it's a robust and solid performer. Now I'm just offering a "softer" alternative.

    Youssef

  9. #9
    Purser
    Join Date
    Mar 2015
    Location
    Mojave Desert, USA
    Posts
    133
    This ham radio guy bench tests Airspy. It gets interesting about 5 minutes in. Never heard of decimation before.

    https://youtu.be/ojC3QBdYMfk
    F-KDAG1

  10. #10
    First officer
    Join Date
    Dec 2012
    Posts
    452
    Some VRS statistics:
    FR24:
    http://x264.nl/fr24-vrs-stats-2-min.jpg

    Airspy + ADSBSpy 1.0.0.13:
    http://x264.nl/airspy-adsbspy-vrs-stats-2-min.jpg

    And the results in VRS:
    http://x264.nl/fr24-adsb.jpg
    http://x264.nl/airspy-adsb.jpg

    The Airspy reports 10-20 more aircrafts than FR24 in the list after within 2 minutes of run. The runs are done sequentially many times and give consistant results.

    piopawlu i suggest you'd contact prog. It would be really nice to use these wonderful receivers to feed FR24 directly.
    It would be awesome if you could just enter your FR24 key into the ADS-B program and start sharing.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •