The $10 Echo | Hacker News Reader

The $10 Echo

sammachin.com

375 pointsespadrine123 comments

Thanks for all the upvotes folks,

I should clarify that there is a bit of artistic license on the title calling this a $10 Echo, there are 2 caveats;

Its only about $10 the CHIP is a "$9 Computer" plus tax & shipping etc. and this hack then requires a speaker, microphone and a button, however you may well have these lying around of have another device like an old speakerphone or something that you could gut and put the CHIP into as a new brain.

The other point is that its not really a full Echo, the Echo is a hardware appliance that implements Far Field Voice Technology, Wake Word Detection, A High Quality Bluetooth Speaker, Support for various media service (Prime Music, Audible, Pandora etc) and finally the Alexa Voice Assistant. This hack is really just access to the Alexa Voice Assistant without the wake word or far field tech.

So its about 90% of the functionality for a much lower price. Personally I have an Echo in my dining room which is great when walking around the house but I built this to use on my desk when I want to trigger home automation features without shouting downstairs.

do you have the source for alexa in the browser on github ? that one was a very cool hack!

Yeah I just need to strip out the auth keys etc. and rebase it before i make it public. I'll try and get it posted this week sometime, check back on weds.

One of the more impressive things about the Echo is its far-field mic. I've used a lot of different voice recognition technologies, but Echo blows everything else away in terms of being able to speak in a normal voice from the other side of the room and have it still understand you. The actual intelligence behind the service may be lacking (or at least behind some of the competition), but for a general-purpose home application, the mic (in my opinion) is a significant part of Echo's reliability and success.

This is what is going to make Echo the next big computing platform. The voice recognition just works. It's like being on the Enterprise.

That's how I felt when using Siri and Cortana the first time. "Wow! Voice recognition actually works now??"

If you haven't tried the latest rev of Echo software, it blows Siri out of the water.

Very specifically, you can speak in a conversational tone with Echo, vs barking commands at Siri.

I don't have experience with Cortana, but the Voice Recognition capabilities in the XboxOne is garbage.

I've used Siri on the iphone and Cortana on a Nokia Lumia 939 and they seem about equal. I'm sure the echo is better due to the superior mic.

> Normal voice from the other side of the room.

High end smartphones have a similar high quality mic. It's no problem at all talking or using e.g. GoogleNow.

How often are you talking to your smartphone from the other side of the room? Usually it's a few inches from your face.

In my kitchen and in my office, it lies on the table. It very practical to just speak, I can walk around and do other things at the same time.

Not to mention smartphones seem to be explicitly designed to not provide this use case. Where is my always listening Google Now? If I have to take my phone out of my pocket and unlock it, I may as well type the question myself.

It's called loudspeaker or speakerphone mode. Of course it's not the default mode if you speak next to your ear.

The Moto X and its successors have had this feature for a while. Your phone can be sitting on the other side of the room, completely off, and you can say a certain catch phrase to activate Google Now.

I can usually activate Google Now from a few feet away on my Moto X (2013) but it can't understand what I'm asking after that. I have to get much closer for it to work.

Makes sense I guess. The "listen for the trigger" bit is done locally on those Motos (I've got the 2014 variant) but really all it needs to do is keep an "ear" open for that one phrase so it's probably got a better margin, only helped by local processing. The bit of audio that comprises the query has to get compressed and interpreted on the server side (not to mention the fact that it consists of a much wider variety of possible content to be parsed) so that can't help. I guess a dedicated chip that's just constantly asking "did anyone say 'OK Google?" has an easier job than a remote process asking "wtf does that fuzzy bit of compressed audio mean?"

(obviously being silly and oversimplifying but same idea)

All recent Nexus phones have always-on hotword support. I just tested it from 20 feet ("OK Google, what's the weather like tomorrow") and it worked fine.

In my limited experience using Echo at my parents house over the holidays I was distinctly not impressed by the Echo's abilities to understand voice from across the room. We often had to repeat ourselves, only louder and slower to get the Echo to understand us. We do not have strong accents.

Exact opposite for me, does a great job at recognizing speech but a horrible job providing answers to even simple questions.

> a horrible job providing answers to even simple questions.

If you think of it as a device that performs specific actions (i.e., plays audio, gives a weather forecast, sets a timer, etc.) via an voice interface, rather than a general intelligence that can answer random questions, you're going to have a better time.

THIS. It baffles me how people expect devices such as the Echo to understand every nuance of human language, then get annoyed when it can't answer your overly-complex or vague question. Voice interfaces for me are always reduced to clearly enunciated and simple commands such as "set timer 10 minutes" or "play album Anjunadeep 4". You'll have a much better time than saying "alexaplaymesomealbumsbyaboveandbyond"

Well, in fairness, Google does a much better job than Alexa at 'doing what I want'

Echo has gotten much better at simple questions, and by prefacing "Wikipedia" prior to a query, I find the results satisfactory.

As anecdata goes, I have an Echo. It was an impulse buy when it came on pre-sale, then didn't know what to do when I actually received. I've been using it for quite a while now and the voice recognition is amazing, but the software is not really there yet. I can tell it to set up a timer from the kitchen (which is around a corner from the device) without much of an issue, even with my, somewhat strong, Argentine accent.

Many of the commands it "fails" at is when it tries to "guess" what I might have meant to say because it can't find something in its database. For example, if I give it a wrong name for a song (say "Alexa play She's Leaving by The Beatles" it might play some song called "She's Leaving" by some other band because it can't find that song under The Beatles.

Everyone has an accent. Do you mean that you speak General American (i.e. the North East non-coastal accent)?

Of course you're technically correct, but I find it quite common for Americans to understand that "no accent" refers to General American, even if they're not aware of the term "General American."

It's not just pedantics. I live in the UK, and some people who speak something between Received Pronunciation and the BBC accent also think they have no accent. People can't hear their own accents and presume they don't have any.

But this is more than just my bias for my own accent (which is very close to GA). When I meet other Americans with other regional accents, they will often describe me as having "no accent."

Yeah, Echo is a good listener. In fact, when I ran the video on the OP's site and he said "Alexa, what time is it?" my Echo in the other room woke up and told me the time :)

That's actually one of my concerns about voice-driven interfaces, absent some sort of voice print authentication or tactile confirmation button. Imagine standing up in a crowded room and shouting "OK Google, delete everything!", or "OK Google, recite the first thouand digits of pi!"

We used to do this at work. "Hey Siri! Call [person's partner]". Everyone turned their voice activated stuff off quite quickly after that.

There was a demo a while back where some researches created radio interference in headset cables to trigger Siri remotely.

It would also be interesting if you could create sounds that sound like commands to the algorithms, but are not easily recognizable for the user.

Something similar to your idea was done for images recognized by neural nets and posted to HN recently. https://codewords.recurse.com/issues/five/why-do-neural-netw...

That could be a fun programming exercise for one of these devices you can hook into Alexa, see if you can catch unauthorised voices quickly and effectively without badly impacting response times.

e.g. you could do it by buffering the request, checking and submitting once you're confident about authenticity, but then you'd be delaying submission and response, depending on how much you'd need to buffer to gain sufficient confidence that the individual is authorised.

This was a common way to troll Xbox players: If they used speakers, shout "Xbox Off" loud enough via the voice chat (in a multiplayer game) and their devices turned off.

https://www.youtube.com/watch?v=mWZLa4AnN5k for a more humourous application.

Isn't that exactly why iPhone / iOS will ask you to say 'Hey Siri' several times when you activate the feature? I assumed it was tuning the recognition to key on just [approx] that voice.

It's not "secure" in any sense, but it cuts down on false positives and perhaps some very juvenile hacking attempts. Any professional hacker will generate a close-enough voiceprint that you can't use voice recognition inherently (purely passively) as authentication any time soon.

> I've used a lot of different voice recognition technologies,

You've probably not used the Kinect controller? It has a very sophisticated distant speech system.

I believe the Kinect has a 4-element microphone array. Not quite as good as the Echo's 7-element array but it uses the same idea.

I don't know about the specifics of characterizing beam-pattern, etc. I am pretty sure, Kinect 1.0 can be more directional in spite of having fewer microphones because the microphones are cardioid, as opposed to the uni/omni-directional ones used in the Echo. Also, some parts of the design of the Kinect speech processing pipeline is public. You can read about in Ivan Tashev's (of MSR) book [1], IEEE Signal Processing Magazine article [2], and some presentation videos he has online [3]. Also, see his publication list, a lot of papers are relevant to the Kinect speech pipeline http://research.microsoft.com/en-us/people/ivantash/publicat...

Also obligatory, why the downvotes?

[1] http://research.microsoft.com/apps/pubs/default.aspx?id=1022... [2] http://research.microsoft.com/apps/pubs/default.aspx?id=2012... [3] https://www.google.com/search?q=tashev+video+lectures&oq=tas...

Agreed. It's an array with 7 MEMS mics, which allows the Echo to do beamforming so they can essentially focus on the speaker while rejecting most of the other sound.

There have been advances in deep learning based speech recognition to reject noise - http://usa.baidu.com/deep-speech-lessons-from-deep-learning/ for instance, but clean audio source helps massively, especially if you are trying to talk from any distance instead of being right next to the mic.

I haven't seen open source libraries for state of the art RNN/LSTM based speech recognition, anyone know if they exist?

MEMS mics have been taking over electret mics recently, they are way smaller and have better SNR (at equivalent size).

I'm building a 4 mic array at the moment, tetrahedral/ambisonics mic for VR use, the mics themselves are very cheap. Mouser and Digikey stock them - http://eu.mouser.com/Sensors/Audio-Sensors/MEMS-Microphones/...

Haven't looked into the software/beamforming part yet - does anyone know if there are open source libraries for that?

Some MEMS mics can also do ultrasound range, opening up pretty interesting possibilities for cheapo LIDAR type echolocation with a mic array - but that's for the next project :)

One of the simplest beamforming algorithms are delay-and-sum. This explanation has example code in C: http://www.labbookpages.co.uk/audio/beamforming/delaySum.htm...

PS: interested in your project, anywhere I can follow it?

This is a pretty cool experiment. As someone that recently worked on an app that was doing speech to text in near and far field, calling this a $10 echo is like calling a Honda Civic a $15K Ferrari. I'd love to see the video of the author speaking to this microphone from 10' away, or even 5' away in a room with no padding.

The Echo's ability to do voice to text is like nothing that's ever come before it based on my experience building, and speaking with a consultant in the field that worked on the Echo team.

I totally agree, hence my comments about artistic license in the title! One of the issues right now is that people kinda assume that Echo and Alexa are the same thing hopefully as more 3rd party devices come to market using Alexa the difference will emerge. I'm thinking of doing a follow on post explaining the differences. Also there's not really a generic name for the type of appliance that the Echo is like smartphone, tablet, browser etc. I wonder if echo will become like hoover in years to come?

From your lips to Amazon's ears, they sure hope so :)

http://www.cobaltspeech.com ?

Yes sir

I'm not comfortable with the idea of everyone having far-field microphones in their homes. Cell phone spying is bad enough. If these become widespread, then ten years from now instead of a political debate about call metadata it'll be a debate about literally every word ever spoken.

That's a worry that I share. The good thing about hacks like the OP's is that they push us further in the direction of this technology being open and controlled by the user. We need more of that.

> I'm not comfortable with the idea of everyone having far-field microphones in their homes.

That's the problem with progress and technology. It will happen regardless of how comfortable you are with it. In fact, if it helps those with power to maintain or extend their power it will happen. If it tends to limit their power it will be squashed and marginalized.

False. Technology can proceed with privacy in mind. The Echo didn't have to push things to the cloud. It was, however, constructed to do so both for Amazon's benefit (data) and the consumer's benefit (ease of setup and use).

The trend you cite will happen because of ignorance, not progress.

It adds convenience to most services to connect to the Internet, so most of the do by default. So long as the device is connected, state-sponsored actors will find ways to eavesdrop even if the device wasn't originally designed to send audio over the wire.

Citizenfour was eyeopening on this point.

I wouldn't be so upset if the divice firmware was completely open source and easily user modifiable (android does not meet this, 100gb source download is not easy to modify).

I don't see how it's different than a smartphone. Everyone carries with them a little microphone and GPS now. Who cares about using voice recognition in the home?

And the real voice recognition should be turned off 99% of the time. It should just passively listen for a keyword before using any cloud service.

>I don't see how it's different than a smartphone. Everyone carries with them a little microphone and GPS now. Who cares about using voice recognition in the home?

"You already have the flu, why worry about getting gastroenteritis?"

>And the real voice recognition should be turned off 99% of the time. It should just passively listen for a keyword before using any cloud service.

On tools like OP's link, or Jasper, etc. you have the source and you can verify that it's actually not sending out data permanently. With the Echo, you're just hoping Amazon didn't do exactly that. You could watch the network activity, but that would just tell you that some data is sent back, sometimes.

it's like this scene from batman: https://youtu.be/kIvEu-XjZ40?t=21

and no different than a telescreen (1984). No one will leave home without one.

Any recommendations for microphones that are (a) high-quality, (b) cheap (under US$5.00), (c) small (under a cubic centimeter), and (d) with analog output (such as a 1/8-inch plug)?

Note that high-quality and cheap are not contradictory requirements. iPhones and other smartphones have extremely high-quality microphones (and they are tiny) and must certainly cost much less than $5 as a part, but they seem impossible to find. Every off-the-shelf mic under $5 that I've tried has been garbage in terms of sound quality, and even much more expensive ones have not been great.

Good mics certainly exist. Here's a compact analog microphone I own that has fantastic sound quality:

http://www.amazon.com/Sony-Electret-Condenser-Microphone-ECM...

But it's ridiculous that this microphone costs 6x as much as the CHIP and 10x as the RPi Zero. You can't build a product around the CHIP or the RPi if an essential part is that expensive.

This is my favorite cheap microphone. It does need to be fairly close, I'd say within 5 feet:

http://www.amazon.com/Cyber-Acoustics-100Hz-Microphone-ACM-1...

Electret microphone capsule with couple of transistors preamp can give very good results, but requires some electronics skills I guess...

If the author of this is reading, it might help to make the first reference to "Alexa" be a hyperlink to "what the heck is Alexa". (I know I can google it, but...)

"Alexa" is the assistant-in-the-machine that comes on the Amazon Echo (just as an iPhone 6 comes with Siri). You activate the Echo by saying "Alexa". The author is probably guessing that in context, you'll make the connection between the title "The $10 Echo" and the Echo's hotword (Alexa).

I agree it'd be useful if there was a brief explanation at the beginning of the article.

Here's the service that this depends on: https://developer.amazon.com/public/solutions/alexa/alexa-vo...

If you want to try this in your browser, the guy who built this $10 device also used the same APIs to put Alexa/Echo in the browser.

http://sammachin.com/hacks-and-projects/alexa-in-the-browser...

That's great from the price perspective - I still would not want another active listening device tied to a mega corp within my vicinity.

However, I would love to have an Echo like device which does the voice recognition/algo crunching/etc on premise, i.e. on a server/device I control.

Is there anything out which does this nicely?

> I still would not want another active listening device tied to a mega corp within my vicinity.

I wish more people were like you and voiced their concerns about this ongoing problem around centralization of services. However...see http://www.cultofmac.com/264381/hardware-siri-runs-puts-new-...:

> Over on Reddit, a post suggests that there are 3 main ‘instances’ (or server farms) for Siri in the United States, and at least one for every other country.

> This information allegedly comes — albeit second-hand — from Apple’s lead cloud architect, who says that every instance of Siri runs on 32 powerful HP servers with a total of 1024 cores and 32 terrabytes of RAM apiece. That certainly makes the new Mac Pro look long in the tooth.

> Specifically, each instance of Siri is made up of 4 HP c7k enclosures made up of 8 HP server blades each, with memory upgrades to 1TB of RAM.

> According to the post, if one server dies, it’s simply removed and another one slapped in, with no downtime.

Granted, there are a lot of people yacking at Siri at a given time, but still. You probably need a fair amount of horsepower to pull this off.

They talked a bit more about the Siri setup at a talk they gave for the Mesos guys (Siri runs on Mesos with a custom schedule Apple wrote - one of the biggest Mesos installs out there)

https://mesosphere.com/blog/2015/04/23/apple-details-j-a-r-v...

They say the clusters have 1k worker nodes

Apple has also said that Siri gets 1 billion requests per week, so by those numbers each core can handle ~10k requests/week, using its share of 32GB of RAM. Doesn't seem that much.

According to Alexa Voice Service page (https://developer.amazon.com/public/solutions/alexa/alexa-vo...)

Amazon is starting a $100M fund for anyone building voice related services. This is very interesting. Has any company been funded via this fund?

The Alexa Fund page lists a sub-set of the companies that have received funding from the fund so far: https://developer.amazon.com/appsandservices/solutions/alexa...

So how does Amazon make money from Alexa running on non-Amazon devices? (Not obvious to me, sorry.)

It's going to evolve into a massive app ecosystem, surely with paid apps

It looks like you can use it for non-commercial purposes but I imagine that they'd make money a few ways: indirectly through publicity and familiarity (the more people are used to having it, the more might use Amazon's commercial version or just Amazon stuff in general), directly (through licensing it for commercial use), and via improvements they make to their service as a result of people using the free version for non-commercial applications (better voice training and interpretation).

That's just a guess but it's similar to what Google does with AOSP and Google Android.

I'm surprised that no one has mentioned Jasper [1]. Its a cool little but very well organized project that I put together this Christmas.

[1] http://jasperproject.github.io/

You're surprised nobody has mentioned a project you organized less than 10 days ago?

No, I didn't organize it. The awesome people at Princeton did last year. Its really awesome. I just downloaded and made my own this Christmas.

Thanks for the shoutout! Glad you enjoyed it.

I love my echo (alexa), especially the convenience of "no button but wake word" when I have my hands full in the kitchen. That said, google voice is so much better at parsing my questions that I'd pay a bunch of money for another standalone device with wake word which would integrate with with google voice. Any chance of hacking that, or linking your device with both? (Ask Amazon / Ask Google)

Buy an high end phone with a good mic and you can choose between GoogleNow and Siri (and more). Or if you have a SmartTV in your kitchen, you can use that too (Samsung, LG, etc),

I have all of those, but it really doesn't compare. Echo does an amazing job recognizing every voice (without training) from a great distance.

Google's question parsing is ridiculously good. You can also do neat things like say "When I get to work, remind me to email John" and you'll get a notification when you arrive. I don't have an Echo, but I imagine it can't do such contextual things.

You could get a dedicated Android phone/tablet, and leave it always plugged into a charger somewhere. I just tested it on my Phone (LG G3 Android 5.0.1), and I'm able to get my phone to respond to "Ok, Google" questions, even with the screen off.

Though the microphone probably won't be nearly as good as the Echo, and from the setup it seems like you have to train "Ok, Google" to respond to your voice, so I don't know if it will work with multiple users.

I love my Nest Thermostat, and I know they've got "Protect" as the smoke/CO detector. I'd LOVE if Google would release an "Assist" product that was the Echo form factor powered by Google's agent engine, and I could mount it on the ceiling in each of my rooms similar to how smoke detectors are (hard wired to mains with battery backup).

Too much to hope for I suspect.

That would be both awesome and terrifying.

I have an Echo and the mic array is fantastic. And the idea of having one in every room for localized interaction is great. But the idea of such a good mic in every room is scary if you consider bad and/or state actors.

I've got an old Android phone always plugged in, connected to a THX system on top of my fridge/freezer, for doing things like automatically playing music in the morning. Exploring what other things I can get it to do with Tasker atm.

OK Google works, but the mic really isn't as good as the array on Echo. I've been looking into external mics for Android - will test with simple omni mics first but ideally an external array mic would be the best choice.

I've heard that the "$10" CHIP has a $20 "shipping" charge. Is that still the case?

Can't quite remember what I paid for shipping, taxes etc but I just tried adding one to my cart and it was only $6.22 shipping to the UK now.

I backed for just one board, and then they offered to let me add more without paying additional shipping, so I added a second, $18 for the CHIPs, $5ish to ship them. They are en route from China. To be fair, I only just succeeded in ordering a Pi Zero from Adafruit, $5 for the Pi, $4 shipping...

Cool clock, Sam.

I have to be honest, I got an echo for Christmas (thanks Mom!), and while it's really cool, it has only served to highlight how much amazon prime music sucks. In fact, it has only served to highlight to me how disconnected cloud music services are.

Google music has just about everything (now that it is connected to youtube), but amazon doesn't want to let its speakers play google music's music service because they want to sell you their own.

The future is a strange place right now.

The $10 CHIP will be a $10 CHIP when and if it ships. Right now, it's a Kickstarter claiming they will ship the thing in June 2016.

Actually the kickstarter did ship hence why I had my hands on one (well two actually) to do this hack. I backed the kickstarter back in May 2015.

The June2016 date is for orders placed on their store today which I think is the earliest they currently expect to have filled all the current pre-orders and later stage kickstarter backers. IIRC they're currently producing 5000 units a month.

Was a Kickstarter. Got mine in my hands last week.

This is another thing that makes you wonder why companies expose APIs to their services that enablenothers to make competing services.

Amazon doesn't care about selling the hardware. They'd be ecstatic if this caught on. It'd be like owning the kindle ecosystem without having to take a loss on the hardware.

They provide Alex-as-a-service: https://developer.amazon.com/public/solutions/alexa/alexa-vo... ("Build for Free. Using AVS to power speech experiences on your devices is completely free.") (Note: I'm guessing that this will be lower accuracy than the echo, since the echo's acoustic models are probably tuned for their hardware.)

They can revoke your access at any time.

Which is why no one is going to use it.

Reminds me of twitter awhile back.

I thought it was about experimental startup providing echo service for $10/month subscription :)

I'm being a pedant but doesn't an Echo have a mic built in? Buying an equivalent one would be ~£75 I think.

OP explained in a comment here that he was taking a bit of an artistic license calling this an echo.

Echo also is a pretty nice wireless speaker.

compared with ok google and siri, echo is not much better (if not worse) in terms of intelligence. What made echo was the physical form factor.

Intelligence is hard to mimic, but physical form factor is not. All google or apple needs to do is come out with with a similar speaker/mic combo and it will totally blow echo out of the water.

My impression of it was less for answering random questions and more for predefined tasks like home automation,listening to news ect.

I agree. I wish Google would build their own echo. The ability to do things like manage a calendar, text people, and all the other "Ok, Google" functions through an Echo-like device that also does what Amazon's Echo currently does would be fantastic. Alexa can't even tell me what is the stock price of Amazon and can't give me the current temperature without prattling off the entire weather update.

I guess they've just focused on getting the functionality available on devices you can already find (Android and iOS devices, TVs like the newer Sonys that run AndroidTV, etc). I see the appeal of a standalone device but I find that lots of times, the response may not translate well into audio-only so having a screen helps when you ask something and the best answer is a graph, text, or image.

Just a guess though. I'm surprised that wasn't included in that rather pricey wifi router they launched recently. It has a decent sized speaker in it but perhaps they decided not to enable it just yet. My guess is that they wanted to start selling the wifi routers but didn't think the voice stuff was up to snuff for a consumer application/selling point just yet.

Reply on news.ycombinator.com