Hacker News new | past | comments | ask | show | jobs | submit login
Ode to Internet – V.90 56k dialup modem handshake for orchestra (1996) [video] (youtube.com)
87 points by CharlesW on Jan 24, 2024 | hide | past | favorite | 24 comments



To hear an example of what V.90 56k really sounds like:

https://www.youtube.com/watch?v=VaWpi9o_hHI


Yes, at higher speeds (like towards the end of that recording once the link is fully established) dial-up modems sound like white noise.

There's an interesting reason for that. A digital modulation that is going fast enough, if it is well-designed to make use of a channel, will approximate white noise. A primary frequency or harmonic can be thought of like a "hot spot" and the more relatively powerful it is, the more it will distort the other frequencies. So, an ideal encoding will have constant average power across the entire bandwidth. Essentially, white noise.

Modems got there by the late 1980s with the jump to 9600 bps, when a single bit started being spread out over many symbols using convolutional modulation.


Nah. Does not sound like that at all. 56k has that peculiar "ding" in the handshake, which makes very different from v34 handhake.


> 56k has that peculiar "ding" in the handshake

Depends which 56K “standard” you are using. There were two before v90 standardised things, they had different handshake modifications for negotiating the faster link, and v90 was different again. Some ISPs supported only one of the earlier “standards” for some time after v90 (most v90 modems were firmware switchable between it and one of the older methods or were soft-modems where that switch was in the driver).


K56flex was the one I had. USR modems were pretty expensive. But I think the next one for me was V.92 - not that it mattered greatly, because cable internet was around the corner.

https://www.inetdaemon.com/tutorials/computers/hardware/mode...


I've been trying to find a recording of "proper 56k" as I remember it for years but none of them seem to have the "ding" (or "bong" as I describe it) sound.

My understanding is that it required digital equipment at the exchange end which would make it impossible to do with two standard modems wired back to back. Unless someone has some telco kit laying around, it might be a lost sound.


It's not quite that bad. The thing you need on one end is a digital modem. You can buy a card containing a bank of these for the Cisco ISR G2 platform. The ebay cost is about $60. Then in the same router, you can put a T1 card, which the digital modems will use to accept calls (costs about $15). Finally you can buy an analog modem card (costs about $20) to accept calls from another modem. All of this would be contained in 1 or 2 rack units.

There are other methods but that's the one I would use. To see one of the other methods, try here: https://www.youtube.com/watch?v=yQ_hWS6ked8


Really appreciate the pointers. Sadly as none of the modems in the video you linked sounded right either, I might be about to go down a hardware buying rabbithole.


I had this same problem a few years ago and went down the rabbit hole. I was looking for the US Robotics X2 and K56Flex handshake sounds and eventually found them.

V.90/92: https://goughlui.com/2016/05/03/project-the-definitive-colle...

X2: https://goughlui.com/2018/06/10/project-record-a-us-robotics...

K56Flex: https://goughlui.com/2018/06/10/project-record-a-k56flex-dia...


Thank you - you've literally just saved me £500.

Turns out I've been trying to track down the USRobotics Sportster voice modem sound from your first link. Even better as I can't find my 56k USRobotics, only the 33.6k one, so probably wouldn't have ever found that sound with hardware either.

Thanks; that's been lodged in my head for a couple of decades.


Back in the day, when preparing for one of the Cisco certifications, we had to be able to identify handshake sequences based on sounds... ("ah, the V.90 bit starts here").


Reminds me of this bit from Monkey Dust https://www.youtube.com/watch?v=dTos3CsvnLQ


Fun! But I think using the western 12 tone scale really hurts this. I would love to hear a more precise interpretation using violins, trombones, slide whistles etc, DTMF emulated in 12-TET loses a lot of it's charm. A choir arrangement could also sound really epic.


> DTMF

Your critique falls flat (pun intended) because there is no DTMF here.

Also, this video is a chopped-up edit from a session of real music performance, rather than someone's original composition. Many commenters seem oblivious to this distinction.


The very first notes are supposed to be a dial tone followed by the note run on flute+clarinet representing the dialing of a phone number. That is supposed to be DTMF, with the flue being the high tone and the clarinet being the lower tone.

It's hard to recognize as not only would do the frequencies of the DTMF standard likely not line up with the fundamentals of the 12-note western scale, but these wind instruments don't produce anything close to a pure sine wave- they have lots of harmonics, and with two of them overlapping it's hard to discern what the phone number was supposed to be.


Now the Internet is ever-present, silently.


It is everything I ever wanted yet I wish it had never happened.


we want a lot of things when we're young that we only realize later were trivial and probably harmful.


Dependable.

Twitter, Reddit and HN can sometimes be pretty loud.


And here’s a 14.4k handshake for piano: https://youtu.be/niENQH--vXQ?si=meJkSVmL9i0sw3Qy


It sounds closer when it's sped up.


How have I not seen this before? I love the versions where it's slowed down, but that's something anyone can do. This. This took effort. And a hell of a lot of creativity.


Music truly connects us.


The youtube comments are hilarious.




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: