It seems like any color or sound can be accurately encoded (and recreated) with just a few numbers. Yet that doesn’t seem to be the case with smell or flavor. You can take a photo or sound recording and it’ll be a faithful recreation, but there’s no way to do that with smells or flavors. Is it a technology limitation or is there something fundamental to them that makes it harder to encode?
In: Physics
Both light and sound are a frequency spectrum of a specific thing. In the case of light a spectrum of photons, and in the case of sound a spectrum of pressure waves in a medium (usually air). This makes them easy to both capture and reproduce – you simply need a device capable of emitting photons/pressure waves and the ability for that device to alter the frequency of the above.
Smells and flavors, in contrast, are millions of different discrete molecules. Sugar and salt, for example, are completely different things at a molecular level. We can’t scan those molecules easily (outside of lab environments) nor do we have a machine that can easily force the chemical reactions required to make those molecules.
Because colors are just different frequencies of light, and sounds are basically just different frequencies of pressure waves. It doesn’t matter what color you’re talking about, it’s still just light. Different pitch sounds that you’re hearing are just variations in how quickly the air around you is moving back and forth, but they’re all the same phenomena.
Scents and flavors, on the other hand, are caused by the way our bodies interact with a huge range of different substances. The molecules that smell like a grilled steak aren’t just slightly modified versions of the molecules that smell like strawberries, there’s a whole bunch of substances in each one of them, and while some of them might overlap, a lot of them are just different.
You can, it’s just way more complex.
For sound, we only have 1 type of receptor: our eardrum.
For color, we only have 3 types of receptors: red, green, and blue cones.
For taste, we only have 5 types of receptors: sweet, salty, sour, savory, and bitter.
For smell, we have up to 400 receptors.
It’s a lot easier to reproduce something when there’s only a few inputs we need to replicate.
You might ask, “Wait, if we only have 5 taste receptors, why can’t we do it easily with flavors?”
That’s because flavor is actually a combination of taste and smell — mostly smell. It’s a lot harder to replicate 405 inputs than just the 4 for audio visual
Smells have hundreds of receptors. Tastes have a reasonably limited number of receptors, but there’s significant carry-over from smell to accomplish the full experience of taste, so it effectively also needs the smell receptors.
In theory they can be encoded similarly to colour and sound, but we’re not there yet due to the complexity.
Vanilla is primarily dominated by one chemical, vanillin. It has a bunch of other minor volatile compounds, but you can ‘fake’ it with just the one. Raspberry is another, the primary ketone 4-(4-hydroxyphenyl) butan-2-one (RK) can be perceived as raspberry even without the minor supporting chemicals. However, you can’t realistically produce Vanillin and RK from the same chemical “speaker” – they’re totally different things.
The smell of strawberries, on the other hand, is made of up to 360 individual volatile chemicals, none of which can stand alone and need to be at specific concentrations in relation to each other. Even if you have the same chemicals, if you get the concentrations wrong, your brain will not perceive it as strawberry, e.g. it might come off as rotten/fermented.
Even the same chemical can smell differently at different concentrations – “Dimethyl sulfide is important for fruit flavours, but also at certain concentrations gives the smell of the sea.”
Sound is one thing, the perceived pressure of vibrations at various frequencies and amplitudes in the air against one receptor, our eardrums. Color is the wavelength and amplitude of photons hitting our eyes, which only have to condense information from two different types of receptors – rods and cones.
Smell is the interaction of tens or hundreds or thousands of chemicals, all of which are made up of different molecules which are made up of different atoms, all at various concentrations, interacting with hundreds of different types of smell receptors.
To put it another way, think of those pin impression toys that you can press your hand or face into. Each sound from a speaker could leave an impression in the same board, and you can tell them apart. You can do the same thing with a smell, but instead of just storing the depth of the pins, you’re also changing the color of the individual pins, and their size, and their shape, and their material, and it changes with time, etc. And we can and do store all that information in various ways, but there’s still no existing technology that can reproduce the million billion trillion possible combinations of stuff on the fly out of some universal base material like sound or light can.
The sensation of smell happens when a volatile compound enters your nose. The schnoz analyzes the makeup of the volatile compound, and produces a sensation of smell in response.
There are hundreds and hundreds of building blocks that could potentially make up a volatile compound caught by your nose. It’s how these building blocks are combined and in what quantities that determines what the smell smells like.
You could make smell-o-vision (or more accurately a smell printer) if you wished, but you would need:
* An “ink” cartridge for each scent building block
* A database of recipes for printing each smell
* A devices that preciesly mixes and vaporizes the fragrance for your sniffing pleasure
If you think printer ink is expensive, imagine how costly it would be to get a few hundred vials of cosmetic-grade fragrence bases, some of which degrade over time!
Latest Answers