Every signal that we have sent out requires them to be visually oriented. Do you think the TV signals we beam into space will make any sense to beings that communicate ultrasonically? An encoded 2D image interlaced with alternate lines 30 times a second won't be of much use to intelligent vampire bats.
Okay, first off...
1) Vampire bats do not work that way.
2) Humans take information that our senses can't perceive all the time and turn in into forms that we can. That's what false-color images and the like are.
3) A species that can pick up the signal (as the GP posited) is most definitely able to transform signals between mediums. It's pretty much a fundamental part of any receiver technology - you take a propagating signal, turn it into data, then turn the data into a form that you can perceive.
Obviously no species is going to inherently have the recipe for demodulating the signal just handed to them - they'll have to figure it out, even if their senses are precisely the same as ours. They'll have to recognize, "hey there's a signal here, and by its pattern it doesn't appear to be naturally generated and seems to be storing data in some manner". They'll then have to reverse engineer how to pull the data out of the signal. Then they'll have to figure out how the data is structured (probably the hardest part, esp. with modern compressed digital formats). All of these apply to all beings. But once you've figured all of that out, turning it into a form that you can perceive is the easy part.
Say there's a species with no vision that can only experiences the world through ultrasound echolocation, as in what you probably intended to be your example? Once you understand that the signal is, say, periodic frames representing an array of triplet values (what we know to be RGB) and know how to decode it to that, the species may play it back by, say, an "ultrasound screen" that creates the perception of a 3-dimensional surface, with the height representing pixel intensity. Maybe they might combine all three RGB values into one height, maybe they might present them as side by side heightfields, maybe they might use one value to represent height, another to represent surface roughness, another to represent sound absorptive properties of the surface, or somesuch. They'll pick whatever is most convenient for them.
I'm not going to humour your "liquid methane temperature" communication concept because that's far too low bandwidth for a sentient species to practically use. Pheromones also. And "interference patterns of UV radiation", that depends on what you mean by "interference patterns" - you're either talking about a UV equivalent of echolocation, as above, or just visible data shifted into the UV, which is just a frequency shift on the RGB image into their visual range. We as humans do frequency shifts of astronomical data all the time, that's what every image made from a UV, X-ray, IR, radio, etc telescope is.
For any species to be able to get to the phase of being able to receive and demodulate communications, it must have at least the concept and ability to perceive 2D orientation (if not 3D), because it has to be able to align receivers with the right patch of sky. That perception can be of some unthinkably bizarre form by our standards, but it has to exist. Whatever perception of 2D it has, 2d images can be presented in that form.
Your Pi/Tau example is clearly pointless. We as humans clearly know of both constants. Sure, Pi "stands out" more to us at first glance, but if we received something that appeared to be of non-natural origin, you really think nobody would notice if the data was Tau?
Circles are no more "rare in water" than on land. The cross section of a sphere is a circle. What do you think bubbles are? Rounded rocks? Round sea life? Heck, lava underwater, unlike on land, tends to produce round structures called pillow lava. And again, if this to the point of being able to isolate faint radio transmissions from the cosmos and recognize natural from manmade, then it's familiar with all sorts of other concepts - stars, planets, moons, orbits, and countless other macroscopic round things, as well as microscopic / subatomic round things, both 2d (circles) and 3d (spheres). And how exactly are their mathematicians (which they fundamentally must have to be able to be able to demodulate these signals) not figuring out what shape has the least surface to area (or in 3d, surface area to volume) ratios? How are they dealing with radio transmissions without understanding sines/cosines and the like?
There is no such thing as "a constant that describes the relationship of the volume of a sphere to its radius/diameter". To its radius/diameter cubed, yes, but not its radius/diameter. And you really think that mathematicians trying to figure out a transmission from another world wouldn't be able to figure out that a number was 3/4 Pi? Seriously?
In your last example (gravity waves), you've switched to something entirely else entirely. You're responding to a post based on the premise " If they have the ability to pick up the signal". So why are you talking about a situation where they can't pick up the signal because they communicate by gravity waves and we don't? And seriously, if they can manage something as difficult as gravity wave communication, radio wave communication is going to be laughably simple to them.