Codec2: A Whole Podcast on a Floppy Disk • Hacker News Board

Re: Codec2: A Whole Podcast on a Floppy Disk

#111

Post by andai » Sun, Jun 24, 2018, 1:45 PM UTC

peterbmarks wrote: post #92

mrob wrote: post #33

Earlier quoted context omitted.

If we're abandoning accurate reproduction of sound and just making up anything that sounds plausible, there's already a far more efficient codec: plain text. Assuming 150wpm and an average 2 bytes per word (with lossless compression), we get about 5bps, which makes 2400bps look much less impressive. Add some markup for prosody and it will still be much lower. This codec also has the great advantage that you can turn…

Speech to text is certainly getting better but it makes mistakes. If the transcribed text was sent over the link and then a text to speech spoke at the other end you'd lose one of the great things about codec2 - the voice that comes out is recognisable as it sounds a bit like the person. A few of us have a contact on Sunday mornings here in Eastern Australia and it's amazing how the ear gets used to the sound and it…

Could you elaborate on "a contact"?

Are you using Codec2 over radio?

Re: Codec2: A Whole Podcast on a Floppy Disk

#112

Post by gsich » Sun, Jun 24, 2018, 1:46 PM UTC

IshKebab wrote: post #99

gsich wrote: post #61

Earlier quoted context omitted.

What do you mean with "HDVoice"? On landline connections this usually means G722. G711u/a is definitly not "HD".

I don't know what technology it is specifically, but it's a brand name they used for actual high quality calls. Think, 128 kB/s MP3, rather than the standard cups-and-string quality. It only seems to work on mobile.

I know the difference, used G722. On mobile its G722.2, a totally different codec, but with the same ~7KHz range.

But there were some companies that advertised a lower frequency range as "HD".

Re: Codec2: A Whole Podcast on a Floppy Disk

#113

Post by andai » Sun, Jun 24, 2018, 1:48 PM UTC

rspeer wrote: post #80

bmdavi3 wrote: post #47

Earlier quoted context omitted.

Or what if you scrunched the audio down to a bandwidth beyond what was still intelligible, but still captured some semblance of the speaker's voice. Use the original audio to compute subtitles and store them alongside the audio. That's your file. Then the player uses both as inputs to ai (some hand waving), which now has enough to put the pieces together and produce something intelligible again, in the speaker's voic…

Without the "intelligible" part, this makes me think of what the game Celeste does to give its characters voices without voice acting. They make voice-like synth sounds, different for each character, that are about the length of the text they're saying. It adds prosody and intonation to the text-based dialogue of the game. https://youtu.be/TZpQH8kSWNU?t=2m50s

This is how I imagine an intelligent car would sound if it figured out how to produce speech through the antigravity engine.

Edit: oh sweet there's intonation too. Were these all made manually?

Re: Codec2: A Whole Podcast on a Floppy Disk

#114

Post by andai » Sun, Jun 24, 2018, 1:54 PM UTC

carapace wrote: post #81

jancsika wrote: post #44

Would be a fun experiment to use something like 3 or even 1 sine to get unintelligible speech, but then pair it with subtitles where each syllable of the text is animated synchronized with the speech. (Like the "follow the bouncing ball" song lyric animations.) By pairing the audio with the text, you would almost certainly convince the listener that they can understand it. Edit: typo

;-) Sine-Wave Speech Demonstration https://youtu.be/EWzt1bI8AZ0?t=74 > Sine-wave speech is an intelligible synthetic acoustic signal composed of three or four time-varying sinusoids. Together, these few sinusoids replicate the estimated frequency and amplitude pattern of the resonance peaks of a natural utterance (Remez et al., 1981). The intelligibility of sine-wave speech, stripped of the acoustic constituents of n…

To anyone who listens to this, I recommend rewinding to the segment starting at 1:23 a few times and not letting it reach the spoilers. After a few rounds, my brain adjusted to the distortion and I could make it out perfectly, without ever hearing the original.

Re: Codec2: A Whole Podcast on a Floppy Disk

#115

Post by merinowool » Sun, Jun 24, 2018, 1:57 PM UTC

andai wrote: post #110

merinowool wrote: post #91

Earlier quoted context omitted.

When it was patented? When I was working with AI about 15 years ago I was experimenting with conv nn to generate audio. I wouldn't have expected for this to be patented as this is so friggin obvious thing to do. It is like patenting 2+2=4 once you discover numbers.

[Serious question] Does your prior art invalidate the patent?

I am not a scientist, just I was very interested in that space and it would be a long way to create scientific paper out of my experiments. Since patent law has been created for the privileged to reap profits I wouldn't stand a chance contesting that.

Re: Codec2: A Whole Podcast on a Floppy Disk

#116

Post by zouhair » Sun, Jun 24, 2018, 2:02 PM UTC

merinowool wrote: post #91

gwern wrote: post #66

Earlier quoted context omitted.

Codec2 might be patent-free, but Codec2 with a WaveNet decoder isn't because WaveNet (convolutional neural networks for generating audio sequence data) is patented: https://patents.justia.com/patent/20180075343

When it was patented? When I was working with AI about 15 years ago I was experimenting with conv nn to generate audio. I wouldn't have expected for this to be patented as this is so friggin obvious thing to do. It is like patenting 2+2=4 once you discover numbers.

Isn't that specifically what software patents are? Pythagoras could have become a billionaire in his time and don't get me started on Al Khwarizmi.

Re: Codec2: A Whole Podcast on a Floppy Disk

#117

Post by zouhair » Sun, Jun 24, 2018, 2:07 PM UTC

voxadam wrote: post #43

sp332 wrote: post #40

Earlier quoted context omitted.

Is Speex also in that category?

Speex has actually been superseded by Opus. Both are patent free as well.

Opus is patented[0] but just royalty-free patents.

[0]: http://opus-codec.org/license/

Re: Codec2: A Whole Podcast on a Floppy Disk

#118

Post by tototomtoboro » Sun, Jun 24, 2018, 2:17 PM UTC

carapace wrote: post #81

jancsika wrote: post #44

Would be a fun experiment to use something like 3 or even 1 sine to get unintelligible speech, but then pair it with subtitles where each syllable of the text is animated synchronized with the speech. (Like the "follow the bouncing ball" song lyric animations.) By pairing the audio with the text, you would almost certainly convince the listener that they can understand it. Edit: typo

;-) Sine-Wave Speech Demonstration https://youtu.be/EWzt1bI8AZ0?t=74 > Sine-wave speech is an intelligible synthetic acoustic signal composed of three or four time-varying sinusoids. Together, these few sinusoids replicate the estimated frequency and amplitude pattern of the resonance peaks of a natural utterance (Remez et al., 1981). The intelligibility of sine-wave speech, stripped of the acoustic constituents of n…

Wow this is amazing, after listening to this a couple of times the voice became super clear

Re: Codec2: A Whole Podcast on a Floppy Disk

#119

Post by ksec » Sun, Jun 24, 2018, 2:40 PM UTC

Let say we have Codec2 with WaveNet, its 3.2Kbps now does similar to may be 16Kbps EVS. ( EVS being the codec used in VoLTE, which is slightly better then even Opus in Speeches. )

What "value" / "uses" does this bring us?

It cant be used in podcast because as shown it isn't very good with Music. And many podcast has Music in it.

While Codec 2 with WaveNet can have a 2-4x reduction in bitrate. I cant think of a application that benefits from this immediately.

The other thing I keep having in my mind is convolutional neural networks on Codec in general, Music, Movies, etc. What sort of benefits it bring us.

Re: Codec2: A Whole Podcast on a Floppy Disk

#120

Post by madengr » Sun, Jun 24, 2018, 3:04 PM UTC

Would be interesting to combine this Codec2 with LoRa modulation. Of course the latter is patented, but it combines both chirped and direct sequence spread spectrum to yield some very resilient modulation.