Antenna thoughput, typo or my misunderstanding?

[SaturnV]

In the description, the com-16 has 0.4(second) interval, 0.8 thoughput; DTS-M1 has 0.3 interval, 0.6 thoughput; 88-88 with 0.2 interval and 0.4 thoughput.

With some test I found

All type of antenna can only transfer up to 2Mb per tick. So 10Mb goo data need 5 tick, 25Mb SC-9001 data need 13 tick.

The electricity on description panel is energy needed per tick. For DTS-M1, 25Mb data = 13 tick = 13*15 = 195E, while com-16 needs only 130E.

Interval is the delay between two tick, DTS-M1 needs 13*0.3 = 3.9 seconds, com-16 needs 13*0.4 = 5.2 seconds

So IMO, the throughput should be 2Mb/0.4 = 5Mb/s for com-16, 6.67Mb/s for DTS-M1, 10Mb/s for 88-88. Thus I can't understand what the 0.8/0.6/0.4 thing stands for.

Edited October 22, 2013 by SaturnV

[TheWanderer]

you looked at this a little more than i did. I was just left thinking the more advance "intermediate" dish is actually worse than the antenna.... ?

[SaturnV]

Uh I think I found why!

It must be a math mistake, look, thoughput = data per tick / tick cycle time, so thoughput of com-16 is 2Mits/0.4 = 5Mits/s, but what if you incorrectly multiply them up? Then you get 2Mits*0.4 = 0.8; So do DTS-M1, we got 2Mits*0.3 = 0.6, 0.4 for 88-88, that is exactly what shown in game! The dev must be wrong this time

[Gus]

So they are mislabeling a value as it's inverse?

[falofonos]

report it as a bug!

[Tassyr]

So wait- DOES this mean the 'best' one is worst? O.o if so I can just correct the dang thing myself... xD

[frenchie16]

There's a lot of weirdness with antennas. For one thing, unless I'm reading the specs wrong, the antennas get progressively worse in terms of electricity usage as well. They all have the same packet size, but the intermediate antenna uses 15 electric charge per packet, and the advanced one uses 20 (the basic is 10). This means the basic is the most efficient...?

Second, you can enqueue as many experiments to transmit as you want, and the antenna will work its way through them. This essentially means that you can do something like spam an experiment while passing through Duna's upper atmosphere, choosing to transmit every time, and then once you've landed, sit on the ground with solar panels out and warp until transmissions complete, thus getting nearly all of the available science for that experiment in Duna's upper atmostphere with a single brief flight through it and a single experiment part.

Finally, the time it takes to transmit seems to be unaffected by warp. This is kinda minor compared to the above, but weird...

[Tassyr]

Sadly the 'spam' option isn't an option. At least not that I've found. It seems that it halts the upload, and goes for the new one.

[Mykill Metal]

So wait....I'm totally lost. The farther down the tech tree you go you unlock WORSE antennas? That doesn't make any sense. As you unlock more you should get better ones that send more data faster, and don't lose as much. Leading to the final one you unlock that gives you 99.9% of your science you transmit because it's SO EFFICIENT. But nope. No one thinks like that, do they.

[Johnno]

Throughput isn't hardcoded and is just a calculated reference number. Only thing you really have to worry about is power usage and transmitting time, seeing as noted the packet sizes are the same. The later antennas eat more power but transmit faster, in some cases that is preferrable. Also mass is a factor, if you're happy queueing reports and transmitting them one at a time you can simply use one antenna. Efficiency per transmission doesn't change but the efficiency over time does increase with the later antennas, meaning you can transmit the same amount of data but faster.

I've found that having several antennas with faster transmission speed can be very useful when you need to transmit from several experiments and preferrably fast, like barreling through the atmosphere. If you have several antennas they can transmit at the same time, instead of waiting until the previous transmission is completed, and the faster you can get the data sent the more readings you can do.

All in all, as with everything else, it comes down to a judgement call of what you need for a specific craft/situation. Personally I find them quite well balanced.

[Eric S]

Sadly the 'spam' option isn't an option. At least not that I've found. It seems that it halts the upload, and goes for the new one.

My Eve lander probe was broadcasting high-atmosphere results for over an hour after it touched down, you can most definitely spam the science on the way down. It was still broadcasting "flying in eve's atmosphere" when I went to bed. I may have overdone the spamming (as in definitely, by the time it got towards the end, all the experiments were returning 0 science), but when you see 1200 points for a single experiment, you don't want to miss a single chance at getting more of that.

The way I do it is to put all my experiments on an action button. Hit the button, click transmit on each popup. Wait a few seconds after clicking the last one for all the animations to complete, then hit the action button again (the delay avoids a popup that you may be running into that does cancel the previous experiment).

I've found that having several antennas with faster transmission speed can be very useful when you need to transmit from several experiments and preferrably fast, like barreling through the atmosphere. If you have several antennas they can transmit at the same time, instead of waiting until the previous transmission is completed, and the faster you can get the data sent the more readings you can do.

I found that multiple antenna's didn't help much, so it may be buggy, or it may just not work. I found that it would start using the first antenna, then start using the second antenna. From the time it started queueing things for the first antenna, it never queued anything for the second antenna.

Ah, a little experiment, and now I understand better. The system that assigns an experiment transmission to an antenna doesn't load balance very well. It seems that if all the antennas are in use, it assigns the transmission to the first antenna it finds, even if that one antenna has a lot of items queued and the other antennas only have a single item to transmit.

So, as long as you're generating a constant stream of experiments, it seems that you're better off with multiple 16s instead of the dish antennas. Dang it, and I've got a Jool impactor already in system, and now I want to redesign it.

[Fr3Runn3r]

Basically people need to stop thinking of the transmitters as being better/worse. There are trade-offs. The small transmitter is slow, but efficient (exactly what you need in the early game) and the large transmitter is fast but costly (designed for when you have electricity to spare). Personally I tend to use multiple small antennas for parallel transmissions (quite often 6 or 8)

[frenchie16]

Except that due to the behavior of transmitters (with unlimited-length queues so that you can spam experiments), speed really doesn't matter. Power consumption matters somewhat more (especially with probes) but with solar panels, even that isn't a huge deal, so it's really down to just mass.

Also, it makes no sense that a "high-gain" antenna (which the description of the fancy dish-shaped one claims it is) would require more power to transmit the same data. A high-gain antenna requires LESS power, because the gain is higher.

[Galane]

And now there needs to be communication lasers. NASA just successfully tested a 600+ megabit/second laser transmitter from the Moon to Earth.

Steve Wozniak lives in a place where the fastest landline internet is 1 megabit DSL but there's 600+ megabit from the Moon. How @#%$@ up is that?

[KerbMav]

Looks like we need a middle part between science equipment and antenna. A computer built into pods and probe bodies preferably and with a limited capacity, to buffer all data and let the player flip through the stored reports and decide what to transmit and what to keep until back at Kerbin - like we already can with EVA reports and samples.

Still leaves open the debate if physical experiments should be (100%) transmittable (at all) and if pure data shouldnt be.