h a l f b a k e r yi v n i n seeks n e t o
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
Add an odour cartridge to an inkjet printer. Package smell information along with the image.
When you print a picture of your holiday destination, you can actually smell the sea or the pine trees. Pictures of cars actually smell like cars. Movie stills could smell of gunsmoke or burning rubber or
ozone or whatever. Images of clothes could have that "freshly-washed" smell.
The effect woudn't last long but it would be a neat gimmick for promoting some web services or products.
Perfume classification.
http://www.museesde...ng/crea_classif.htm
Barely scratching the surface. [angel, Jun 26 2002, last modified Oct 21 2004]
Patent for fragrant inkjet ink
http://patft.uspto....757&RS=PN/6,123,757
Baked for two years in the US Patent office. [smokeyjohnson, Aug 19 2002]
Please log in.
If you're not logged in,
you can see what this page
looks like, but you will
not be able to add anything.
Annotation:
|
| |
Scratch and sniff has been around for a long time. Is this different in any major way, other than the implied cheapness? |
|
| |
Angel: It's instantly and completely customisable. The smell can be sent as a data file along with an emailed image. There's no intermediate manufacturing process. Somebody could send you a picture of a pie thry just cooked and you could smell it as well as see it. It's a printout system rather than a mask so that it is highly portable once produced, requires no power, and can be shared with others. |
|
| |
And it would work how? Colours can be reproduced from four components. Scents can comprise dozens, sometimes hundreds, of individual elements. Do you have a cartridge for each?
//The smell can be sent as a data file//
Unless you give at least some idea of the mechanics of it, this is looking WIBNI-ish. |
|
| |
You could perform an exact chemical analysis, and reassemble the exact same molecules at the far end in the same quantities, in science fiction replicator style. Oh, no, that would be a WIBNI, would it not? |
|
| |
And it would not deal with the different components of the smell reacting with each other, evaporating (since most smelly compounds are by their nature highly volatile, else they wouldn't evaporate in sufficient quantities to be detected), decaying, etc. |
|
| |
There is a reason why scratch-and-sniff seldom smells like what it's meant to. |
|
| |
Like the "smelling aid". Existing inkjet cartridges have a matrix of tiny heat-activated jets. This works in the same way except the cartridge contains a hundred or so smell elements. The "scanner" at the other end uses thick-film chemosensor deposition technology to detect the principal smell elements and then encodes them in the smell equivalent of a JPEG. Since smell doesn't have to be precisely targeted onto the paper like the image does, the smell cartridge could have a very broad printhead that sprays over whole areas at once, each pass containing different protosmells as required.
The data encoding is in fact pretty easy once you've sampled the original smell. Just a table of the components and their relative intensities. Much less diffcult than an image. |
|
| |
This makes me wonder: How many different smell/taste receptors do we have? Do they form a continuum or are they distinct types? Anybody know the answers? |
|
| |
How many 'protosmells' do you expect to use? I ask because there are a lot. For instance, "over 100 different scent components of otter dropping have been identified. About 17 of these are thought to contain information on sex, age and even individual recognition, which can be used by other otters". (From otter.org.) Not that I'm suggesting you would want otter-dropping-scented postcards, although you might, but it does at least give some idea of the scale of the problem. It would be possible to encode a chromatogram of the scent, and transmit it, but re-assembling it at the other end into a copy of the original scent would be a major problem. |
|
| |
I read, about a year back, in Popular Science (I think) about a printer that would actually print edible, flavoured ink. So, the theory went, you could go to www.buyapizzaonline.com, and print out a "sample" before ordering. Or www.edibleunderwear.com for that matter. I can't find a link at popsci.com, though. Sorry. Great idea - smells like a croissant. |
|
| |
Angel: The minimum number of components to produce a reasonable facsimile of the source smell would have to be established by experiment. In the same way that a 2D paper image is an acceptable alternative to the true 3d thing for many purposes, so the synthetic smell would lack "depth". You couldn't send a wine smell by this method and expect to recognise the vineyard. But you should be abe to tell it's wine and not fruit juice. To be widely acceptable it would have to be cheap and simple; the "SmellCam" at the sender end and the "SmellJet" printer at the other. |
|
| |
Interesting about the otters. Human smell ins't anything like that precise. |
|
| |
Pottedstu: No, it's not a transporter or replicator. And it would get around the "scratch-and-sniff" ageing/reaction/oxiadation problem is that all the protosmells would be "fresh" when applied (sealed in tiny reservoirs until needed). The effect would age pretty fast, just like real smells. |
|
| |
Hey, there are worse possibilities. I may well take that as a compliment.
[8th], I think that your 2D / 3D analogy is faulty, but we could compare what you want to a grey-scale representation of a colour image. The 'resolution' of the smell would be extremely limited, perhaps to combinations of only 16 commonly required, easily-synthesized components. |
|
| |
Angel: Yes, the "Greyscale" analogy is the best, we think. Probably more than 16 components needed, too - this would be determined by experimentation. After all, a greyscale image can tell you most of what's really important about something - you don't need colour or 3D all the time. |
|
| |
Steve DeGroof: Yes - you might even find a way to do this with existing printers: |
|
| |
1. Print the image on SmelloPaper using the conventional printhead. |
|
| |
2. Swap for the SmelloPrint Head (could even be done with an old dot-matrix printer wnd no ribbon). Put the paper back in but upside down this time. |
|
| |
3. Print the smell image. The printhead cycles over the papaer, either spraying activator, or physically bursting the microcapsules, to release the scent. |
|
| |
4. Enjoy ! (Unless, of course, Angel has sent you a picture of Otter Turds. Or maybe you like Otter Turds ? There's no accounting for taste). |
|
| |
Check out US patent 6123757 "Fragrant Ink for Inkjet Printing" published Sept. 26,2000, US patent 6261347 "Scented Jet Ink and Printed Article Therefore" published July 17,2001 and "Scent Diffusion Apparatus and Method" published April 16,2002 |
|
| |
Thannx, blaise - will check the references. This idea may therefore be Baked although our initial searches didn't find an existing product. |
|
| |
Ooh, I just thought of this idea. |
|
| |
I too remember the popsci or popular mechanics article about smelloprinting. |
|
| |
The problem is that unlike other sensory inputs there's been very limited success dividing scents into individual components. Where as a wide gamut of colors can be recreated from red blue and yellow ink. And sounds can be recreated using wave form data at 44.1khz there's no easy way to divide scents into "smellements" |
|
| |
That is to say most recognizable smells have a very unique chemical signature/componenets. combining the smell of butter, flour and yeast won't produce the smell of a croissant. i'm sending you the calcium carbonate smellement of a fishbone right now. |
|
| |