Telstra's 4G Vs 3G: Is It Worthwhile?

4G is fast, or so the traditional orthodoxy goes. But how fast is it, and is the jump over 3G on the same network actually noticeable? There was only one way for me to find that out. Elly's review of the HTC Velocity 4G got me thinking about data speeds. There's little doubt that 2012 will be the year of 4G -- whether it's real 4G or not -- with Telstra's service already up and running, Optus due to trial through to April and Vodafone... well, Vodafone was originally intending to launch by the end of 2011 and then didn't. They've been rather quiet since the start of the year, but presumably Vodafone-branded 4G services will emerge at some point in 2012.

For now, though, if you want 4G (or at least Telstra's LTE variant thereof, an argument I'll address another day), then Telstra is your only bet. But what if you're already on a Telstra 3G product? Is the difference enough in the real world to make the upgrade worthwhile?

Testing a 4G USB modem (as I've done in the past) or a 4G phone (as Elly did) in isolation is one thing, but what I was interested in seeing was the level to which actual 4G coverage makes a data difference on the same network, and with the existing devices on the market. So I gathered up a Velocity 4G, Telstra 4G USB Modem and two 3G smartphones running on Telstra's network and got to checking out some real world speeds.

The Test Kit

1x Telstra 4G USB Modem (connecting to a Macbook Air) 1x HTC Velocity 4G Smartphone 1x Samsung Galaxy S II 1x Apple iPhone 4S

The key thing here is that all four devices were running from Telstra SIMs on Telstra's own network, with the latter two smartphones obviously being 3G only. I tested in three locations. The first two were within Telstra's 4G coverage zone. Firstly, Gizmodo's Circular Quay office, where typically the reception speeds are average to poor. Secondly, outdoors at Sydney's Martin Place to minimise the amount of building interference with the signal.

Telstra's 4G implementation drops to 3G (and in the case of the HTC Velocity 4G, dual channel HSPA+) outside of its stated 4G zones; it's fair to presume that the other carriers will follow the same drop-down-to-3G strategy, but what does that do to speeds?

To ensure that tests would use 3G only, my final tests were conducted north of the Harbour Bridge in Hornsby, well outside Telstra's claimed 4G coverage area.

Tests were run using connecting to the same Sydney-based server. For the USB modem, Speedtest was run on Safari, while the iOS and Android apps were used for the smartphones.

As a side note: Every time I do this, somebody comments that they don't trust's figures. In this case, it's not just the figures that I'm interested in; it's the difference in scale, and for that, using the same tool is a vital testing component. Likewise, the test 3G phones were used not to prove a point between them, but because they were what I happened to have at the time of testing.

Each test was run three times to get an average ping, download and upload speed. As with any mobile test, there's an amount of variability at play that I can't entirely accommodate for; at the same time, these are real world recorded speeds.

The Results

Firstly, the results from within Gizmodo's own walls, where mobile signal is (to put it politely) not always the best.

Device Average Ping (ms) Average Download (Mbps) Average Upload (Mbps)
HTC Velocity 4G 89 11.73 5.84
Telstra 4G USB Modem 37.66 12.84 10.74
Samsung Galaxy S II 685 0.57 0.09
Apple iPhone 4S 159.33 1.1 0.383

Then the results from Martin Place, where the outdoor setting should deliver optimal results.

Device Average Ping (ms) Average Download (Mbps) Average Upload (Mbps)
HTC Velocity 4G 55 11.9 10.89
Telstra 4G USB Modem 33 11.56 13.41
Samsung Galaxy S II 101 2.43 2.41
Apple iPhone 4S 71 2.91 0.51

And finally the '3G-only' Hornsby scores.

Device Average Ping (ms) Average Download (Mbps) Average Upload (Mbps)
HTC Velocity 4G 80.33 7.959 2.44
Telstra 4G USB Modem 35.66 8.29 2.4
Samsung Galaxy S II 96 2.17 0.86
Apple iPhone 4S 96 5.14 1.84

The Conclusion

Once again, just to be clear: As with any wireless test, conditions and results can vary, as can congestion on the network. This is particularly true of 4G, where fewer devices will be filling up the network compared to 3G.

The 4G scores are, indeed, faster and better than their 3G counterparts; there's no great shock there, although it is worth noting that if it's just speed and not smartphone ability you're after, the USB connection was not only faster, but consistently so; the single fastest download speed was achieved in the Gizmodo offices at 18.78Mbps, whereas there was more variance when using the Velocity 4G.

Switching over to 3G still shows a decent level of differentiation; while as Elly noted there are new 4G phones on the horizon, if you're in need of a little extra speed, whether it's on a smartphone, USB or tethered, Telstra's current 4G solutions do appear to offer value for now.



    Was that gs2 running the latest baseband version and android version? Seen much quicker then that on telstra with my s2

      Of course you have... Just as I have with my iPhone and with the velocity I tried...

      The point isn't to test the phones speed but to test the different network speeds at the locations he tried. He even explained as much in the intro...

      Frankly comments like this just sound fanboyish...

      As Alex noted, this is all dependent on many variables. But, as long as the tests were performed in the same location and at the same time (give or take a few minutes), they are a true reflection of the differences.
      My iPhone 4S speed tests differ massively depending on the location, time of day etc.
      Also worth noting: I get faster speeds on sloptus than what has been shown here. Although, I would expect a telstra iPhone to shit all over mine if it were to be tested at the same time. ;)

        so did yours btw LOL

    Results from the Velocity 4g in Brisbane:
    Woollongabba (3g) dl. 8.71 ul. 0.76 ping 169
    Toowong (4g) dl. 25.72 ul. 7.51 ping 141
    Toowong (3g) dl. 2.06 ul. 0.21 ping 148

    from the android app.
    *note in Woollongabba I should be getting 4g reception.

    I think you need to learn how mobile telephony works my friend.

    These are in no way Apples for Apples tests.

    Channelisation? F/W versions? H/W classes?

    IPhone 4s is faster in a 3G area than 4G- ?? How very strange !

      In none of the results does it show the iPhone being faster then the 4G

        maybe so... BUT if you turn the results around, i think TImmy may be onto something...
        Yeah, you're right!

      Not hugely odd -- the areas where 4G signal is present happen to be some of Sydney's busiest for 3G users; while undoubtedly there's more in the way of towers, there's also a lot more users. Hornsby is quieter ,and as such there (may) be more signal to go around.

      Which is not to say that it would always be that way, but it would certainly explain those figures and make a certain amount of sense.

    With mobile, what is more important than all out speed is connection reliability. On mobile, even your worst average of 0.5 on the sgs2 is ok (not great) if it is stable. 99% of browsing, VPN, VOIP, media streaming etc will still function fine. Its the dropouts that get ya. A comparison using something common like skype would be interesting. there's not an incredible difference in experience between 0.5 and 20mb/s connections (for general things).

    I got a 4s on Vodafone and it is dial up slow most times in seafood tethered to
    Apple tv and a movie on Sunday was going to take 5 hours plus to watch

    Ah, as I suspected, Telstra selling fake 4G.
    Thanks for saving me the time of switching then suing.

      You're all idiots, 4G is a marketing term ALL around the world. LTE is a 4G network, that much is not debatable. And to the person below me stating that its a non-standard frequency, just because America doesn't use it doesn't make it non-standard. I think there are/soon will be around 50 countries around the world running on the 1800 spectrum that Telstra uses. Including countries heavily into Tech like South Korea and some of the bigger European countries (I think even Spain).

    This test is flawed because your are using different types of devices. The Samsung and iPhone are limited to a maximum of 14.4mbs. The real test is the Telstra 3G Elite dongle us the 4G Dongle remembering that Telstra 3G is rated at a theoretical 42mbs.

      Samsung g s2 is not limited to 14.4mbps theoretical you should get 21

    It is not Telstra 4G but Telstra FakeG created by them Ericsson retards. Telstra's LTE Basic network operates on a non-standard frequency meaning compatible handsets and dongles have to be custom made for their network and will not work anywhere else in the world.

    hahahaha hows the reception over in Hornby

Join the discussion!