Spoiler Alert: First 3-D Printed Records Sound Awful | Hacker News Reader

Spoiler Alert: First 3-D Printed Records Sound Awful

wired.com

130 pointsreplicatorblog61 comments

Its interesting to note that the results of early stage enabler technologies are worse than the previous generation technology its trying to mimic and surpass.

e.g

1. Digital cameras producing lower quality images than film cameras

2. Computer memory chips holding less memory than actual books.

3. Digital Printers producing low quality prints compared to the printing press.

4. 3-D printed records sounding awful compared to actual records.

5. ?

This turns out to be a key component of Clay Christensen's original theory of "disruptive technology" (before it became a buzzword.) The theory is that the disruption comes in part because existing producers ignore the new technology, since initially it is only of use to a tiny niche in the market that doesn't care about performance. The niche is left to bit players, but it turns out to provide them enough oxygen that they can invest in R&D and improve the technology, eventually to the point that it's the best choice for all but the highest-end applications.

http://www.amazon.com/Innovators-Dilemma-Revolutionary-Chang...

Its also interesting that the technologies developed to close the gap between digital systems and their physical counterparts eventually create the efficiencies for them to surpass them in the long run.

For example: Noise, image sharpness and sensitivity algorithms applied to early CMOS 2 and 3 megapixel sensors eventually allowed 8-65 megapixels sensors to far surpass the quality of their physical counterparts (a current 8mp sensor is better than 35mm film, 65mp sensors are far better than medium format film even though their resolutions are roughly comparable)

Its like stepping backwards to digitize it forces the development of efficiencies required later to create a far better product.

congratulations (no snark) - you're Clay Christensen

http://www.amazon.com/Innovators-Dilemma-Revolutionary-Chang...

haha thanks. That book is on my list for years!

> Its interesting to note that the results of early stage enabler technologies are worse than the previous generation technology its trying to mimic and surpass.

This is nothing new - the question is whether the potential of the enabler technology is larger than the potential of the "old" technology (which, presumably, has been maxed out or is getting close).

I am sure there are technologies that died out in the labs because they looked promising at first but never became commercially feasible.

Mobile apps?

If you are comparing it to the PC Apps, they will always be much more powerful than Mobile apps simply because they have more screen real estate and processor power available to them.

Web "Apps"

Pretty impressive. Of course you could also make a 'needle' out of a solid-state accelerometer too these days.

But what is interesting is that the fidelity is more about bandwidth than resolution, so really if you wanted to you could make a very large diameter disk and turn it faster to improve the resolution that you are hearing. Sounds like a lot of fun to play with.

It blows my mind that it played anything. Also, the irony behind actually printing out physical records is delicious.

The future of piracy.

1) Download a 3D model LP file from BitTorrent 2) Create LP from model using 3D printer 3) ??? 4) Profit!

I know a lot of people (and am in that group) that would pay for high-quality vinyl 3D models once we all have a printer in our kitchens, especially for non-mainstream releases that have little chance of getting a wide distribution on vinyl. An example of there being a market for make-your-own-music: check out Beck's sheet music album, it's been very popular. http://songreader.net

Oh, the irony... People willing to pay for high-quality vinyl (analog recordings) built from a higher fidelity digital recording of the physical object.

Why not just listen to the digital recording in the first place?

Isn't pretty much all new vinyl produced from digital sources? It's all about the hipster factor.

Because it is the sound creation in the end that is analog I assume. You could create a musical instrument from a digital source -> 3D printer. Let's say a flute. It would still be played analogue and the means of creation do not matter at all.

I am not an "audiophile" but I assume this is the reasoning and it is perfectly understandable and logical.

That would depend on the eventual bit-depth I suppose. Maybe there is some micro-smoothing function on these new-fangled machines-of-the-future that will bring a new type of listening. It's not just the listening itself that gives pleasure to listening on vinyl, it's the sitting next to a turntable alone or with a small group of friends choosing tracks and have a physical connection to the source of audio.

Also, with vinyl you can always listen to your music without requiring electricity should you find yourself with only a needle, turntable and large sheet of paper.

Deleted comment

I'm not sure this makes sense - digitizing entails "not analog". Can you expound on what you mean by preserving "the entire medium"?

Isn't it a given that using a cartesian co-ordinates to encode polar data would sound awful (that said, I am very surprised at the quality)?

I'd love to see what they could do with a polar printer, depositing material on a spinning disc. I'm sure it could be done.

It could be, and has been, but now some fucker is attempting to patent it and ruin it for everyone else. See http://www.kickstarter.com/projects/wjsteele/ultra-bot-3d-pr... for details.

"Is it open source? No... this design is so revolutionary that I had to patent it to protect my IP. This design has the potential to change the 3D printing industry. It affords many benefits that make it stand out above all other 3d printers."

It's not a new design at all, one of the first Reprap prototypes was a polar printer circa 2006 and there have been several other attempts at it since. Of course, because of the variable accuracy (least accurate away from the center) they don't usually get much use.

Edison's first tin-foil phonograph sounded pretty crappy, too.

http://websnap08.lbl.gov/Tinfoil_files/foil_top_ewc-DC-FIRFo...

It will get better.

Is the improvement of 3D printer resolution simply a matter of time and engineering or are there new technologies that require significant R&D?

There's not really a difference. Maximum speed of an airplane in the 20th century improved at what looks like a fairly smooth curve if you plot it, but there were lots of new technologies that were required to do so (e.g. forced induction, jets, etc.)

Presumably that's because the new technology is deployed just as soon as it first overcomes the old. :)

Of course, that's why there is this illusion of "progress marches on" when in reality it's a bunch of breakthroughs under the curtains.

There are also some patents that are blocking innovation at the moment in the commercial 3D printer space.

Interesting - I didn't know that. Do you have some examples?

Crowd-funding website Kickstarter is being sued for its promotion of a new 3D printer.

More than 2,000 users contributed over $2.9m (£1.8m) to help Massachusetts-based Formlabs build the device.

However, 3D Systems - a leading maker of printers that turn computer design files into real-world objects - has alleged one of its patents was being infringed by the machine.

It has also filed a lawsuit against Formlabs itself.

From http://www.bbc.co.uk/news/technology-20434031

Many are held by 3D Systems (example: Zcorp has a patent on full color printing via inkjet), some by other organizations. UT Austin has the patent for SLS (Selective Laser Sintering) which is a nylon-based high-resolution 3d print. MIT has the patent for ceramic 3d printing.

Entirely anecdotal, but I was told that there are some commercial patents regarding air circulation (keeping a stable chamber is absolutely necessary for high-resolution) that are blocking some of the open projects as well.

Is there a difference? Or is your question really about what field the R&D will have to happen in?

I think he's asking if it's like Moore's Law (expect consistent geometric gains) or if it's like battery technology (consistent linear gains). Will next years' printers do 8 micron steps or 15.5 micron steps?

IIRC, Battery has been geometric gains since the 80s, at a single-digit percentage increase per year.

To be fair, you could never really understand the lyrics of Smells like Teen Spirit anyways.

prediction: people will download the digital files for the record models, then write software that will play them back from the file rather than having to print them out and playback on a physical machine...

Yes, and once they have circumvented the record files to play withou 3d printing, perhaps there will also be some manner of online store where one may purchase these digital files, a store for tunes! :)

I also expect people have software that can generate the model on the fly and just print a record straight from an audio file.

I thought there was already tech that allowed you to play audio from pictures (special?) of vinyls.

I haven't seen anyone image a whole record at once, but there are http://en.wikipedia.org/wiki/Laser_turntable units which image the track as it spirals by (without wear from contact).

he was being ironic. We already have that technology.

We need to figure out a way to 3D print with the plastic they use for Shrinky-Dinks. Print your record out 4 feet wide with giant grooves and pop it in a pizza oven.

So can someone here explain what exactly would be needed to improve the quality of audio on these? Is it just higher resolution or does it require better materials? Can the quality of the plastic dampen the vibrations somehow?

There's so many aspects to that question, but I'll have a go.

The first thing to point out from the materials aspect is that there are compliances everywhere with vinyl records. They are essentially lubricated with plasticisers which are needed to give some reasonable playing life given the accelerations involved. They also depend upon shrinkage in the pressing stage, the metal stamps engineered to allow for material relaxation. There is a slight shrinkage in cooling, and this helps shrink errors as well.

However, these processes lower fidelity. The original masters, or even direct to disk recordings, use chiselled stylii driven by a high precision lathe and electromechanics to carve the track in the material. Corrected phasing, given that a difference in track lengths is present between left to right, is also performed.

So materials that could do with some annealing after deposition could help, at least until higher resolutions become available. Following a correct RIAA curve of course matters and using a steep low pass filter rather than down sampling an already phase damaged mp3 would help.

At the moment all this is hobby stuff though. Allowing material reflow would probably be the best way to preserve the life of the stylus and reduce noise. It's first playing will be carving off the printer resolution deposition errors.

Higher resolution requires better materials - but most additive 3D printing has a max resolution of a few fractions of millimeters (hundreds of microns). A good record has a groove resolution on the order of single micron units - if not smaller. That's actually quite a ways to go. Even conceptually, it's not easy to deposit and solidify something with single micron precision without a well-regulated chamber, a very precise (and probably slow) nozzle. Even the expensive commercial units still have only sub-mm resolution.

They should print it for 78 RPM, they would get about double the bandwidth.

A 78 also has a wider groove than a 33 or 45.

If the 3D model file is too large, then can it instead be a piece of software that converts any given song into a record as it prints? Perhaps a 3D model of the record + software that knows where to put the grooves and takes an audio file that is already prepared for the disc but is more compressible than the full 3D model.

It may sound awful, but so does everything else at first :)

The Library of Congress used a scanner to restore and preserve old broken records:

"You Can Play the Record, but Don't Touch" (2007)

http://www.npr.org/templates/story/story.php?storyId=1185184...

This technology in future could also be used by criminals to print fake finger-prints. I guess then physical authentication needs to be updated to something un-printable. e.g retina scans, voice.

I wonder if an Epilog laser engraver could do any better.

That reminds me of this: https://www.youtube.com/watch?v=C4OV2UofPFg

Their high-end models have double the linear resolution (1200dpi), so it seems likely.

3D print the record then laser engrave the song on it?

If you can start with a blank record then printing it is a waste of time, for the same reason we don't print paper.

Good thing those records are durable: http://youtu.be/h4GYg-5AdRw

Oh the irony. Printing out records makes me laugh.

ghostery users: unblock brightcove.

How has nobody pointed out how awesome that girl is? 3D printing? Algorithms? The Pixies? <3

Reply on news.ycombinator.com