Interference colors on a vinyl record

[Orthoceras]

It is well known that interference colors can be seen on a vinyl phonograph record in sunlight. However, those colors appear only if the plane of the record is viewed at an angle of about 45°. That is strange, assuming the record is a reflection grating. There are 8 grooves per mm, so the calculated order of a spectrum at an angle of 45° is n = (d/λ) sin θ = (0.125 mm / 0.0005 mm) sin 45° ≈ 200. Such a high value is very unlikely for a brightly colored spectrum.

For comparison, the grating constant of a cd is 100 times smaller (600 lines per mm), so the order of a spectrum at 45° is about 2. This value is exactly in accordance with the observed colors on a cd.

So, why do the interference colors on a vinyl record appear at such a large angle?

 

[tech99]

Maybe it is caused by the sound modulation of the grooves rather than their spacing.

 

[A.T.]

Maybe it is caused by the sound modulation of the grooves rather than their spacing.

The sound modulation seems to be of the same order of magnitude as the spacing, but it could be a low pitch sound:

But there might be finer structures on the vinyl surface:

 

[Orthoceras]

The wavelength of a 4 kHz modulation is 0.1 mm in the groove, so the 'grating constant' of the sound modulation is greater than the distance between the grooves. The wavelength of low pitch sound is even greater. In addition, the colors are brightest if the lightrays from the sun and the observation direction are perpendicular to the groove, so I don't think the sound modulation of the grooves could explain the colors.

 

[hutchphd]

The record is not just a simple grating but contains grooves of a particular angular reflective geometry. In the language of diffraction gratings this represents a blazed grating and certain input geometries will select for higher order diffraction channels. These will be angularly separated more than the low order peaks. So it is both the spacing and the blaze that matters.

 

[Orthoceras]

A blazed grating is fine for selecting a first order spectrum, because a first orde spectrum does not overlap with higher order spectra. But in this case it should select the 200th order spectrum, which overlaps with the 201th, 202nd, .. and 199th, 198th order spectra. All those overlapping high order together are white light, I imagined a blazed grating could not disentangle them.

 

[hutchphd]

It depends upon the blaze, the angle, and the order. The angular separation between two "adjacent" colors gats larger at large orders. I have not worked it out but this is a good working hypothesis. Do you have a laser pointer and an LP record? Play with it.

 

[marcusl]

Lp’s have variable groove spacing and independent modulations on the two angled walls that encode stereophonic information—in other words, there’s a lot going on. Try experimenting with an old 78 instead. The grooves have a constant spacing and there’s only one (monaural) modulation.

 

[Andy Resnick]

Summary:: Why do the interference colors on a vinyl record appear at such a large angle?

It is a very interesting optical effect. I don't think the colors are due to interference, because the groove does not act like a diffraction grating. Higher-magnification images help show this- here's a relatively low magnification:

The light source wasn't the sun (I took these images in my lab), but the illumination geometry and spectra were similar to being outdoors. At higher magnification the individual grooves become visible:

Slightly higher magnification, showing the detailed groove structure:

And finally, the high-magnification view from a lower viewpoint angle:

I confess I can't think of a mechanism for the chromatic effects, but I think it's clear that the record groove does not act like a diffraction grating.

 

[Orthoceras]

Your microscopic images are fascinating. Amazing that each groove has a particular color, and that adjacent grooves have different colors.

I used a laser pointer to observe the scattering of light perpendicular to the groove, and summarized the findings in the figure below. In figure 4, A is the record, and B is the laser pointer. Fig 1, 2 and 3 are photos of the scattered light on the wall, for different values of angle β. The groove is a retroreflector. The light beam which is backscattered towards B has a few nodes and antinodes. The angle between two nodes, φ, is about 2°. I agree the cause for these nodes and antinodes is still unclear.

 

[A.T.]

I confess I can't think of a mechanism for the chromatic effects, but I think it's clear that the record groove does not act like a diffraction grating.

Could it be, that the grooves have smaller grooves along their sloped walls? When the master disc is carved, irregularities on the cutter could create fine groves along the groove walls. These would be transferred to the stamper and then to the pressed copies. Alternatively, maybe the pressing process could create them. But it seems like they are there in the images below.

From: https://commons.wikimedia.org/wiki/File:Langspielplatte.jpg

Here a similar picture:
http://4.bp.blogspot.com/_6NMjq-GAJNo/S42IWx0JDXI/AAAAAAAABP8/IZ08jEFFwpA/s1600-h/grooves.jpgWhat about the vinyl-video-disc shown in the video below at 5:30? Does it have tight enough grooves to explain the chromatic effects?

 

[Andy Resnick]

The groove is a retroreflector.

Thanks for the kind words- retroreflection is interesting, as it implies the groove is a 90-degree cut.

 

[Andy Resnick]

Could it be, that the grooves have smaller grooves along their sloped walls? When the master disc is carved, irregularities on the cutter could create fine groves along the groove walls. These would be transferred to the stamper and then to the pressed copies. Alternatively, maybe the pressing process could create them. But it seems like they are there in the images below.

I don't know about that- whatever irregularities/microcracks that may appear during carving don't seem to be present on the replicated records, or have worn down after playing the record a few times. I'll try and take some higher-magnification images today to show details of the groove.

 

[A.T.]

I don't know about that- whatever irregularities/microcracks that may appear during carving don't seem to be present on the replicated records, or have worn down after playing the record a few times.

Playing the record is another potential source of those fine grooves within the main groove. Probably the most likely one, now that I think about it.

 

[tech99]

Thanks for the kind words- retroreflection is interesting, as it implies the groove is a 90-degree cut.

The groove should be 90 deg cut because the Right and Left channels are cut at right angles to each other and 45 deg to the plane of the disc. This is the Stereo system, invented by Alan Blumlein at EMI.

 

[Andy Resnick]

Tough to get good microscopy images of a record groove, I couldn't pipe in sufficient light from the side. Here's a high-magnification image of a groove using epi-illumination:

The colors down inside the groove are (I think) the interesting bit we are discussing- it makes me think the phenomenon is due to a combination of partial transmission and reflection off a highly dispersive material.

Bonus points for figuring out what the song is- can you name that tune with two grooves? :)

 

[A.T.]

Tough to get good microscopy images of a record groove,

There is plenty of it online. For example here:
https://aphelis.net/representing-sound/

And again you can see, that the grove slopes have many much finer grooves themselves.

 

[Orthoceras]

Another idea: couldn't an imperfect retroreflector angle explain the color of a groove? Suppose the angle between the left and right wall has a small deviation from 90°. Then beam A and B would get slightly different directions, causing interference nodes and antinodes (if the deviation is less than the beam divergence of A and B). It would resemble a double slit pattern.

 

[.Scott]

The photograph you are showing demonstrates a type of "rainbow" effect.
Although the intensity of the light is clearly a function of the grooves and the encoded sound, the actual color is a function of the reflection angle.

 

[cmb]

Just a throw-in thought; Maybe the surface colour of the plastic after being pressed at dissimilar pressures (the louder sounds are deeper grooves, so there is more modifying pressure/surface area, possibly light penetrates the top most layer of plastic which has been differentially stressed) is an additional factor?

Possibly more importantly, can you tell which track is the best one from its colour?

 

[Baluncore]

The photograph you are showing demonstrates a type of "rainbow" effect.
Although the intensity of the light is clearly a function of the grooves and the encoded sound, the actual color is a function of the reflection angle.

The area between the grooves is reasonably flat, so it forms a slightly undulating plane reflector. At high angles of reflection, close to the normal, the differential path length, via any two adjacent surfaces, separated by the groove, will generate the wavelength dependent optical interference pattern. That is a blaze grating. The sound amplitude, or width of the groove, inversely modulates the width of the reflectors and therefore the wavelength selectivity. So a quiet period results in wider reflectors, so it is brighter, but with less wavelength discrimination.

 

[Orthoceras]

I could not find a confirmation that grooves might deviate from the perfect retroreflector angle. Instead, I did find that the bottom the groove is not a sharp corner, but has a radius of 6 μm.(link) The diagram below shows how the flattened bottom and the retroreflected beam could result in an double slit interference pattern. In the diagram, β=25° and the center distance of beam A and B is 18 μm. For a double slit and green light with λ=0.5 μm this would result in a φ=1.6° angle between nodes. That is of the same order of magnitude as the observed φ=2° angle in my laserpointer experiment when β=25°.

at .. angles close to the normal, the differential path length, via any two adjacent surfaces, separated by the groove, will generate the wavelength dependent optical interference pattern. That is a blaze grating.

Close to the normal on a vinyl record, interference colors do not appear, in reality. At larger angles with the normal (~30°) they do appear. The distance between grooves is 1/8 mm, which is about 200 wavelengths. That high number of wavelengths is unfavorable for brilliant interference colors. Compare it to soap bubbles, in which interference colors disappear when the distance of the reflecting layers exceeds a few μm.

 

[hutchphd]

could result in an double slit interference pattern.

What are the two paths (source to eyeball) that are interfering? Your diagram is very sketchy.

 

[Orthoceras]

For simplicity, the beams A and B interfere at a very distant screen, in the far field, as is usual in theory. For simplicity, the observer's eye is in the far field as well.

The two paths of the retroreflected beam are shown in post #10 (fig 4 and 5).

 

[hutchphd]

Neither the camera nor your eye is moving during the experiment. Yet there are colors.
Please trace two paths showing the interference you are attempting to describe that will give interference to which you allude. I understand and appreciate the drawing in #11 but fail to understand the diagram in #22. You need to accrue phase along two paths to show interference. This is the semiclassical (eikonal) approximation for light.

 

[Orthoceras]

In post #22 and #18, A is the left retroreflected beam (which is reflected from the right to the left wall of the groove), and B is the right retroreflected beam (which is reflected the other way around). They are divergent beams due to their small diameter in the groove, so they overlap and interfere in the far field.

In post #11 fig 4, next to the letter C, four antinode lines of the retroreflected beam are shown. One of the antinode lines hits B, that is the zero-order antinode; the others are 1st and 2nd order antinodes. Obviously, for the 1st order antinode, the path difference for the left and right retroreflected beam is one wavelength (and the phase difference is 360°), etc.

If somebody's eye would be positioned at one of the antinodes, it would see bright green light emanating from point X. This is also evident from from fig 1 2 3. When using a red laser pointer, the spacing of the antinodes is almost doubled. When using white light instead of the laser, the retroreflected beam would project different colors at the wall.

Below is a very schematic diagram with different light traces. The phase is zero at the blue line near the groove.

 

[Andy Resnick]

There is plenty of it online. For example here:
https://aphelis.net/representing-sound/

And again you can see, that the grove slopes have many much finer grooves themselves.

I apologize if it wasn't clear I meant optical microscopy.