September 16, 2014 – 10:05 am | No Comment

Audiofanzine put the Waldorf Streichfett digital synthesizer to the test. Released earlier this year, the small box is designed to replicate the sound of vintage string machines.
Read the review
SynthGear TagsSynths keyboards and anything with a keyboard …

Read the full story »
Home » Audio Gear

Record grooves under an electron microscope

Submitted by on April 21, 2014 – 7:24 pm 62 Comments

Chris Supranowitz is a researcher at The Insitute of Optics at the University of Rochester. Along with a number of other spectacular studies (such as quantum optics, trapping of atoms, dark states and entanglement), Chris has decided to look at the relatively boring grooves of a vinyl record using the institute’s electron microscope. Well, not boring for me.

From what I read, it’s not just a simple matter of sticking a record under a fancy microscope, as there is a lot of preparation (such as gold-sputtering the surface) and post-processing to be done. Having said that, the results are very cool:

Here is a shot of a number of record grooves (the dark bits are the top of the grooves, i.e. the uncut vinyl):

Here’s the grooves closer up – the little bumps are dust on the record:

And here’s a single groove even closer still, magnified 1000 times:

Chris also did the pits in a CD – here’s what they look like, just for contrast:

Chris decided to take the whole electron microscope image one step further, and created a blue/red 3-dimensional image of the record groove! So, if you have a pair of 3D glasses (sorry, the ones you got from watching Avatar won’t work – you need red on the left, blue on the right), throw them on and take a look at this amazing picture:

Maybe these vinyl grooves are only beautiful to an audio geek like me, but I think that these images are truly spectacular. I wonder what we’d see if it was magnified further still? Thanks to noiseforairports for the tip.

SynthGear Tags

  • Recordgroovesunderanelectronmicroscope
  • record groove
  • record grooves
  • record grooves magnified
  • magnified record grooves


  • Obbop says:

    With enough practice, just as a maestro can view sheet music and hear a symphony play the music in his mind, an engineer can scan the vinyl grooves and hear the song inside.

    Admittedly, it IS more difficult doing so viewing a quadraphonic vinyl record.

  • That is so neat. You can actually SEE the waveforms cut into the record. The term “cut a record” has so much more meaning at this level of detail.

  • Realistic Speakers says:

    How organic the analog music looks compared to the sterile CD.

    • Ben says:

      Cd’s look so fake, like the music is in pulses. Records, the sound is continuous. I always enjoy looking at records, seeing the wavy, squiggly grooves. Plus, with records, you can always tell how loud of soft the sound will be, by the space between the grooves.

      • Steven Sullivan says:

        This is silly. Analog tape has better audio performance than LP does, but you don’t see ‘wavy squiggles’ it. Does that make it ‘fake’?

        And if you can ‘see’ the pits on your CDs, you must have superhuman vision.

  • Cici says:

    I know this seems odd, but I love these photos. Do you know of a way I could get/buy prints of these? The resolution of the photos in the article wouldn’t be quite what I need for the size print I’d want.

  • James says:

    Now I want to see this done to a hard disk, which, compared to a CD, stores 1000x as much data in the same area. What kind of detail would a SEM be able to pick up at that level of magnification?

  • Tim says:

    It makes me wonder what the Grand Canyon would sound like.
    Probably should clean it first.

    • Steve says:

      Awesome awesome post. Maybe the planets are just LP’s stored in some dusty back cabinet of a stoner god, waiting to be made sonorous by a gargantuan crystal.

    • paul says:

      That is a fantastic idea. If there was a way of transferring the shape of the Grand Canyon into the grooves of a record then you could hear what it sounds like.
      Does anyone know how this could be done?

      • jigs says:

        it can be done, but its a lot of work — with the technology now, 3D image of the canyon can easily be copied, tranfer it to a graphics/imaging software (for enhancement and preparation), then to a CAM (computer aided manufacturing) system, then convert the data to be readable by a vinyl record cutting machine (where master copies are made)…then you have have your Grand Canyon Music..

  • anothermike says:

    @james, Not much. Hard disk drives store infomation in a localized magnetic distortion. Optical media like CDs store information in physical distortions on the surface of the disk. There’s nothing to see on an HDD but the bad sectors, areas of physical damage to the disk. It’s more fun looking at circuit traces inside microprocessors; sometimes the developers will hide “easter eggs” and funny pictures in the layout.

  • Ben says:

    @James Hard disk drives use magnetic polarization to store data, not physical deformation of the disk surface. Magnetic force microscopy (phase data from a microcantilever is used to calculate magnetic attraction or repulsion) has been done on HDDs however to show the bits on the disk. And yeah, much higher density than a CD.

  • dtenczar says:

    Really amazing images. I’d love to see how the needle looks inside the groove at this magnification.

  • Gene Haynes says:

    Lunar photos look very similar to photo #3 > I wonder what kind of informational matrix exists there?

  • Tim says:

    Simply amazing pictures.

  • Frips says:

    This is a reply to James. I work in a lab with an old SEM and a nearly new Zeiss FE-SEM. We can image at 1kV, low enough voltage to not have to gold coat the sample, and we can show minute details in a sample when imaging all the way to about 15k times magnification. I think a BluRay disc would be cool, as well.

  • john says:

    now a bonus would be to see the same thing, but only with the needle sitting in the groove

  • Even more incredible is the sound.

  • Steven Sullivan says:

    “How organic the analog music looks compared to the sterile CD.”

    It’s just different ways of transcribing information.

    The CD pits looks like bacteria to me — nonsterile and quite organic. ;>

    The difference is we can’t ‘read’ them by eye in any way whereas we can at least get a sense of amplitudes from waves carved into plastic. (It only goes so far; I don’t think anyone could ‘read’ a complex waveform by eye that they’ve never seen before, and correctly deduce the pitch, rhythm, tempo, and timbres of the music)

    One thing you can see is how prone the analog is to entropy — cracks, fissures, and dust. It would be interesting too to see an SEM of a brand new pressing versus the same after it’s been played a few dozen times. Also one could compare the waveforms of the original audio signal coming off the master recording (which is usually on tape or digital), to the waveforms that end up carved into the LP , and to the waveforms encoded into the CD.

  • Eric Porter says:

    These images are impressive–by far the best electron microscopic images I have seen of record grooves. There were some pictured in old Stereo Review magazines many years back, but I’m lovin’ these!

    High would love to have some high-quality images like yours. They would make cool artwork.

    Thanks for sharing!
    PS I hope you find a clean quadraphonic lp to view. I believe you would get a kick out of these, too!

  • Oli says:

    I would like to meet the engineer who can hear music from looking at the grooves of a record! I’m sure one can estimate the beat but much more than that I am highly dubious.

  • Al Pacheco says:

    wow! This is awesome. Just yesterday I was talking to my wife about how neat it would be if a record were to be magnified to the point you can be inside the record grooves and this morning I stumble on this!

    I looooove vinyl!

    I think I’ll e-mail the Science channel so they can do a show on “How that made”. Vinyl Records.

    Thanks for this!

  • maxer says:

    How winyl is made was already been made by Discovery channel
    Heres the link:

  • A.R. says:

    How’s about a high-res group of pictures? This would be an amazing desktop wallpaper.

  • rew says:

    Yes, very high res. files would be much appreciated!

  • Richard says:

    High res images would be great, but unlikely.

    If you look at the white text on the black band below each pic you’ll see that the resolution is 512 x 480 on the original pics. I have no knowledge of SEM equipment but I presume that if he could have taken higher res images he would have.

    Google “Chris Supranowitz” and you’ll find his own website with more SEM images, all at 512×480

  • Richard says:

    … and having read the article on his web page a bit more closely, the reason the images are not higher resolution is actually a physical one:

    “Currently, the Institute of Optics’ Transmission Electron Microscope (TEM) only has the capability to record micrographs onto photographic film; digital file acquisition is not possible. Since the TEM is under high vacuum, a digital camera will most easily be integrated into the TEM via one of the available window ports. Because of the harmful x-rays that are emitted by the TEM, the viewing windows must be made out of thick (at least 0.75 inches) leaded glass. One such glass is Schott F2. However, if one attempts to do high resolution imaging through a plane parallel 0.75 inch thick window, image quality will suffer greatly.”

  • cube says:

    here’s a cross-eyed 3d version of the red-blue image, for those of you who (like me) don’t have a pair of red-blue 3d glasses around.

  • Lira says:

    Thanks for such an inspirational imagery on such a dear subject.
    High resolution pictures?
    Yes, Grand Canyon would surely rock!

  • Crompton says:

    Beautiful! I have to know though–what records are we looking at?

  • Adam says:


    There’s no way an audio engineer could look at the grooves of a record and know how it sounds. The images we are seeing here probably aren’t even the frequencies of the notes of the song but the harmonics that more or less wouldn’t give us any relevant information of what the melody or chords are, or what instrument it is. At best you would find flaws such as scratches, but that’s about it.

  • amber says:

    ooh, my favorite song! LOL great post!

  • Jake says:

    That is truly is remarkable and what I find the most remarkable is the ability to see how incredibly true analog media is at this level. Analog media is at major upward trend this past decade and these photos really show you why.

    My roommate is big time audiophile and owns some very expensive record players. You really can hear a difference. I really wish people would start funding more research into better analog mediums.


  • Zosimus says:

    I actually did happen to have a pair of anaglyph 3D “glasses” lying around and the 3D image of the grooves is pretty awesome.

  • Dunc says:

    “Also one could compare the waveforms of the original audio signal coming off the master recording (which is usually on tape or digital), to the waveforms that end up carved into the LP”

    Well, you’d have to correct for RIAA equalization first… The signal carved into the vinyl is heavily equalized so that low-frequency signals can fit into the space available.

  • Bryan says:

    How organic the analog music looks compared to the sterile CD.

    I don’t mean to start a digital/analog debate but keep in mind that the CD is magnified 20,000x while the comparable vinyl image (the first one) in which the grooves are about the same size as the pits, is magnified 200x. That means the CD grooves are 1,000 times smaller than the vinyl. So, even though the CD pits are ‘sterile’ there’s a LOT more of them. You add compression algorithms, signal processing and copy handling in there and suddenly you’ll see why we’ve gone digital (or ‘sterile’ if you prefer).

    Still, I’d love to buy the high-rez prints!

    I wonder what the vinyl looks like at 20,000x…

  • Rob Anderson says:

    Hey1 i recognize that song!.
    just kidding, but I am 49 years old and still own 1000 vinyl LP’s and even play them.
    i ran a mobile disco and played thousands of hours of great music that way. (now i am an mp3 player addict)
    in my house it is my kids who ask me to turn down the music. phhha-kids!
    Thanks for the great photos.
    I probably do know the song – by the way.

  • Jamie Dell'Apa says:

    Hey, that looks like A Partridge Family song. I guess he didn’t want to ruin any of his top shelf records.

  • Lyn says:

    One of Chris Supranowitz’s vinyl groove photos is all over the Web at the moment, probably thanks to a post at Digg, but yours is the first I’ve found that actually acknowledges the source.

    Chris’s page at University of Rochester is here:

  • Wes says:

    I’ve been cutting (recording) analog disc masters since 1963, and I can tell you that it’s easy to estimate tempo and volume changes in a vinyl record without hearing it, although I can’t have any idea of the melody. For example, in a 45 RPM London pressing of the Rolling Stones’ “Satisfaction,” one can easily see that there are consistently three beats per rotation of the record (great drummer). I can tell from these SEM images that the middle groove picture (with dust particles) is a stereo recording because of the asymmetrical groove walls. Also, I can see that there isn’t much high-frequency information–mostly mid-to low frequencies, a mellow sound. I would love to see the striking of a cymbal. Depending on the pitch and depth of the groove, the curved tip of a stylus will normally ride along the groove walls a little less than half way up from the bottom, leaving the valley untouched. Nice work. Thanks also for the anaglyph.

  • aljuk says:

    Obbop says: “With enough practice, just as a maestro can view sheet music and hear a symphony play the music in his mind, an engineer can scan the vinyl grooves and hear the song inside”

    As someone with 25 years of top level professional experience, I can tell you this is utter nonsense.

  • John Eric says:

    The digital industry is betting that the
    human ear can only hear to twice
    22 kilohertz, twice using Nyquist sampling.
    Some individuals can experience more
    feeling from a continuous recording without
    sampling, such as an old audio cassette.
    Others cannot experience this. Just
    assume for an instant that the human ear
    could detect a rhythm up to a millionth
    of a second. Well guess what: an old
    80’s cassette or record even with a motor
    that drives inconsistently would give a
    better feeling experience for each second
    you listen to it because on the microscopic
    level, there are things there… Curves in
    the air pressure.

    • Steven Sullivan says:

      No, the digital industry ‘bet’ in the early 80s that adult human hearing only went to around 22 kHz — a stance well-supported by science at the time (and still well-supported today). So it chose a 44 kHz *sampling rate*. Of course, today we have options of sampling up to ~200khz (PCM) or even megaherz ranges (DSD). This is to solve implementation issues
      in converting to and from digital, NOT because there is strong evidence that humans need to hear frequencies above ~22 kHz.

      As for cassette tapes: tape doesn’t ‘sample’ the way digital does, but it does use discrete magnetic particles to represent information, not waves carved into plastic. Yet, like digital, THE OUTPUT IS CONTINUOUS.

      • laird says:

        Steve, you’re absolutely correct. In fact, many people can not even hear to that level (22kHz). A simple blind test shows that no one can tell the difference between music encoded in either medium. An experiment I perform with friends is to take a ‘high quality’, unplayed analog LP and digitize it with a cheap AD converter and then set volume levels as close as possible on both sources (critical to listening response as you know) and then collectively try to distinguish the difference while ABing back and forth. An easy set-up for distinguishing differences. But no listener, whether having young, sharper ears or old and slightly worn ears can distinguish the difference. What makes it so difficult (and humorous) is that now the digital recording has the same surface noise and equalization as the LP, both of which were previously aural clues to the presence of analog sound and its ‘purity’, ‘warmth’ or just higher fidelity.

      • laird says:

        Furthermore, those grooves wear, cartridges and styluses, too. And the high-frequency passages are the most difficult to accurately track and are the first to break up. I have many LPs where the soprano vocal parts just crackle and fizz with distortion – very unpleasant, as these can be the most expressive passages.

      • laird says:

        As to your last point, one wonders why there was never an argument about the ‘discontinuities’ of tape, which is the source for almost all analog LPs, exclusive of direct-to-disc. If such a ‘discretely particular’ medium can transfer to the ‘continuous groove’, the waves carved into vinyl, you’d think people would be suspicious of most LPs. I happily use all mediums and can be enraptured listening to a piece on my clock radio in the morning, as the most important part of the listening experience, that’s rarely mentioned, is the brain.

  • MarK Darvin says:

    That is so neat. You can actually SEE the waveforms cut into the record. The term “cut a record” has so much more meaning at this level of detail.

    I stumbled onto your blog and read a few post. I like your style of writing.

  • Niki Tag says:

    “Also one could compare the waveforms of the original audio signal coming off the master recording (which is usually on tape or digital), to the waveforms that end up carved into the LP”

    I stumbled onto your blog and read a few post. I like your style of writing.

  • Richard Wentk says:


    Vinyl grooves – they really are full of dust, distortion, and other crap.

    It’s amazing to me that vinyl was ever considered hifi. (Or apparently still is considered hifi by some.)

    • Kevin says:

      You, sir, are very ignorant.

      • Steven Sullivan says:

        I think he just underestimates the ability of humans to ‘listen through’ things like surface noise and minor pops/ticks. Some people are very bothered by this stuff; for others, the sound of LP itself — the ‘euphonic distortions’ that aren’t there in an all-digital recording — is more than enough compensation for any of its shortcomings. And using excellent gear and scrupulous care of the plastic discs can minimize (but not eliminate) the ‘bad’ noise to the point where listeners don’t find them intrusive.

        What amazes me is that vinylphiles still try to argue that LP is *more* ‘high-fidelity’ or ‘natural’ than digital, rather than just accepting that they like the way it sounds. Especially since ‘the way LP sounds’ can be easily captured in a digitization of an LP.

  • Wes says:

    Certainly the vinyl disc record has its limitations, but one shouldn’t forget advantages it possessed over analog tape recording–yes, really. By the late 1970’s it was evident that recordings cut live directly to disc of a performance had a clarity and warmth, when played with a fine elliptical stylus/cartridge combo, that magnetic tape recording could not match. So, on some occasions the recording amps, lathe, etc. would all be hauled with tender, loving care to the venue to cut the actual master on location. JVC’s pioneering efforts with discrete four-channel surround sound in the early ’70’s required a 30 KHZ subcarrier to be recorded along with the audible sound, much like an analog FM stereo radio station broadcasts.


    Tape recorders would have a hard time recording this with any purity because of resulting heterodyne noises caused by the necessary bias oscillator for analog tape media. This was no problem for disc recording. The issue didn’t exist. Near the end of vinyl’s “glory days” some manufacturers began recording 12-inch albums at 45 RPM, which obviously reduced recording time, but increased dynamic range, frequency range, transient response, and lowered distortion further.

    I agree that those wiggly lines that make up the sound on a record are an engaging visual account of what was happening at the time. It has no equivalent in other media except, perhaps, a bi-lateral variable area optical film soundtrack, which isn’t usually found in the home. The closest you can get to this with magnetic tape recording is to resolve the magnetic pulses visually with iron powder resolving fluid–not nearly as much fun, I think. Sure, each format has advantages, but it is interesting how nowadays analog vinyl records are experiencing an amazing resurgence. I wouldn’t have expected it, but it warms my heart. (ºۍº)

    • Steven Sullivan says:

      I wouldn’t say it was *clear* that direct-to-disc was superior to tape by the late 70s — I would say it was just another audiophile opinion in contention with other audiophile opinions.
      There were no rigorous listening tests of this proposition that I’m aware of. Objectively (measured S/N, pitch stability, crosstalk, etc), the best tape still would beat the best LP, DTD or otherwise. And of course, DTD is wholly impractical for most popular music.

      Quad LPs, of course, derive from quad analog master tapes. And among quad-o-philes, reel-to-reel (tape) was considered the best-sounding consumer format, as noted in that wiki article.

      Btw, vinyl 45s are still made, e.g the ones Steve Hoffman mastered in the past few years. So are reel-to-reel tapes, an even more specialized market.

  • Peter says:

    Exactly, what if we could see even closer? Not really an audio geek but more of an image geek and this was very well done. As you mention there’s a lot of preparation to capture these images and I hope people can appreciate their work and perhaps we can get them to zoom in further 🙂

  • good sound says:

    who cares how it looks. what matters is how it sounds.
    cds. I prefer to listen to see them 😛

  • Frequency is one thing, but rhythm is another.
    And the start and end of a note is being
    hindered by our slow thousanths of a second
    sampling. The human brain likes rhythm,
    and a note that starts and stops at the
    billionths of a second would be much more
    pleasing, hence digital technology will never
    win. That is why people are so down and
    and out this decade, the music sucks.
    All the great artists are losing all their
    recordings into trash, lost forever!!

    • laird says:

      So silly, so very silly. Some people’s logic is the same as Zeno’s proof that Achilles could never catch the tortoise that had a small headstart in a race.

  • Chris says:

    Can watch for hours to these images, beautiful. thank you. (laughed about the last one, with the mars rover

  • Everyone seems to be concerned with
    moving from 8bit to 12bit to 16bit
    on the amplitude of the CD music.
    This is stupid. The thing they are
    missing is the x-axis, not the y-axis.
    The y-axis is good enough, but the
    x-axis is not good enough for the
    start and stop timing of musical
    notes and drum and cymbal crashes.
    The y-axis is the pressure in the
    air as it changes, but the x-axis
    is the timing, and the Nyquist
    frequency is a bare minimum
    requirement, not the best we can
    achieve or even any good at all.
    In summary, both records (vinyl)
    and CD’s are worth continuing
    with, however in Japan, they have
    CD players that do have a higher
    frequency and Blue-Ray players
    also have a higher frequency than
    musical CD’s. So there.

  • VinylGuy says:

    Awesome pics dude, love to “hear” the Grand Canyon! I checked out the 3d pic…. sooooo cool! I was reaching out to pick the dust chunks off lol. Old Beatles… gimmie vinyl, old Bowie… gimmie vinyl, anything modern… cd or mp3 will do nicely.

Leave a comment!

You must be logged in to post a comment.