Macro Photos of Record Grooves

[Soandso]

Here are the first images with the new system. Arrival, last song on side 2 of Abba's Arrival album. Has good HF content. You can see the first quiet groove on the bottom in 2D, on left in 3D images.

I'm seeing the same type of "damage" as I saw on the previous album. I bought this album new back in 1978, and I am the only one who has played, it, but it has been played on a bunch of different systems with dubious stylus qualities. The first album I showed with this "damage" was purchased as NM from a record store about 2 weeks ago, so this is not an isolated phenom.

View attachment 290011

View attachment 290014


View attachment 290015


View attachment 290016

Considering repeated occurrences of what our eye's interpret as "damage' it seems, to me, those may actually be intentional imprinting for audio playback reasons which we don't immediately understand. At this time I am not ready to suggest what function the sections of those markings might serve.

Furthermore, the way those markings appear to be "damage" may in some way be artifacts of light scattering from variability of their immediate surface angles relative to the fixed angle of the photographing lens. I have seen published record groove microscope pictures where the author shows photographic variation depending on incident angle; although none of those done at the extreme OP magnification.

 

[ray_parkhurst]

Interesting hypothesis. One difference I'm seeing in the record, beyond the sheer number of damage sites, is that some of them occur on the trailing edges rather than leading edges of the high amplitude peaks.

I'll check other worn records as well as some new ones to see if these damage sites are present and to what extent. It's hard to imagine why an "impulse" response would be imprinted in the record, though.

 

[Thomas_A]

I wonder what it means. It seems like surface is rough but without stylus trace pattern.

 

[Soandso]

Interesting hypothesis. One difference I'm seeing in the record, beyond the sheer number of damage sites, is that some of them occur on the trailing edges rather than leading edges of the high amplitude peaks.

I'll check other worn records as well as some new ones to see if these damage sites are present and to what extent. It's hard to imagine why an "impulse" response would be imprinted in the record, though.

There's a difference in the photographed light we're seeing; what got polarized vertically is the incident light and what got polarized horizontally is the scattered light. Vinyl records are made of PVC and plasticizers in a blend with ordered moleclar domains in a matrix. PVC has a crystalline component that can be anywhere from less than 5% to 10% of it's volume. The crystalites are composed of ordered monomer units. While the distance between crystalline domains depends on the % of plasticizer; presumably the scattering of light is off crystalline vinyl facets.

Below scanning electron microscope picture is the interior of a primary PVC particle. And a representative sketch of it's composition.

 

[MaxwellsEq]

Fascinating. I'm learning a lot from your work, so thank you. I wondered if it was localised heating/cooling cycles, since the same sections would always be the ones that would be heated during play.

 

[Soandso]

The OP pictured illumination set-up seems prone to induce some scattering. Within photographed sectors appearing "damaged" there are different light intensity sub-sections. Such relative intensity of light scattering depends on the momentum of light energy transfer and that vector of energy is itself a factor of both the incident radiation wavelength (lighting source) plus angle involved.

Which is to say, we are not seeing a myriad of defects. But rather evidence of normally occurring 3D record groove angles we didn't otherwise know about that 2D photographs from a fixed position can only register as scattered light.

 

[ray_parkhurst]

The OP pictured illumination set-up seems prone to induce some scattering. Within photographed sectors appearing "damaged" there are different light intensity sub-sections. Such relative intensity of light scattering depends on the momentum of light energy transfer and that vector of energy is itself a factor of both the incident radiation wavelength (lighting source) plus angle involved.

Which is to say, we are not seeing a myriad of defects. But rather evidence of normally occurring 3D record groove angles we didn't otherwise know about that 2D photographs from a fixed position can only register as scattered light.

You seem very sure about this! I have to admit that these localized patches are new to me, and it's odd that no one has published such phenomena before, so I'm a bit perplexed. I'm pretty sure there is nothing going on with the lighting that is creating these patches, but I've been fooled before. Lighting at high magnification is tricky indeed. FYI the primary lighting for these is the same as I use for the stylus imaging, a ringlight shining down at ~45deg from all angles. This tends to eliminate the types of ambiguity you mention. I do use lights similar to what's shown in the pic above as "fill" lights, but most of the illumination is from the ringlight. I'll only know for sure after examining more records, and especially new ones with no wear. That's my next task.

 

[Soandso]

You seem very sure about this! I have to admit that these localized patches are new to me, and it's odd that no one has published such phenomena before, so I'm a bit perplexed. I'm pretty sure there is nothing going on with the lighting that is creating these patches, but I've been fooled before. Lighting at high magnification is tricky indeed. FYI the primary lighting for these is the same as I use for the stylus imaging, a ringlight shining down at ~45deg from all angles. This tends to eliminate the types of ambiguity you mention. I do use lights similar to what's shown in the pic above as "fill" lights, but most of the illumination is from the ringlight. I'll only know for sure after examining more records, and especially new ones with no wear. That's my next task.

Maybe those anomalies in OP pictures that we have been trying to identify are simply out of phase audio signals in the recording. An audio signal in phase on both the left and right groove wall is achieved by lateral motion during playback. Whereas an audio signal out of phase between the left and right grove wall is acheived by vertical motion during playback.

If we consider the possibility that a record producer may have introduced any out of phase audio signals then conceivably one side of the groove will instigate vertical motion during play. That implies structurally some of the PVC crystalline domains in the vinyl matrix are consequently orientated differently than the crystalline domains orientated for the more common in phase audio signals (hence scatter light from the camera rig's otherwise ideal fixed angle illumination). It is true that 2D high resolution photography accurately reproduces groove displacement, but 2D photography does not necessarily completely capture vertical cut grooves (3D confocal imaging is apparently better for imagery of vertical cut grooves).

 

[Soandso]

Maybe those anomalies in OP pictures that we have been trying to identify are simply out of phase audio signals in the recording. An audio signal in phase on both the left and right groove wall is achieved by lateral motion during playback. Whereas an audio signal out of phase between the left and right grove wall is acheived by vertical motion during playback.

If we consider the possibility that a record producer may have introduced any out of phase audio signals then conceivably one side of the groove will instigate vertical motion during play. That implies structurally some of the PVC crystalline domains in the vinyl matrix are consequently orientated differently than the crystalline domains orientated for the more common in phase audio signals (hence scatter light from the camera rig's otherwise ideal fixed angle illumination). It is true that 2D high resolution photography accurately reproduces groove displacement, but 2D photography does not necessarily completely capture vertical cut grooves (3D confocal imaging is apparently better for imagery of vertical cut grooves).

As to what purpose a record's out of phase audio signal might serve I suggest these are not 180 difference phase impulses (since that would cancel out 2 harmonic waves if they were of the same frequency and amplitude). Rather a partial phase shift of a harmonic tone creates in us significant perception of a pure tone when all other harmonics are in phase.

Moore & Glasberg's (1989) "Difference Limens for Phase In Normal and Hearing Impaired Subjects" provided a shift in phase of a complex tone with equal amplitude and "subjects reported that the phase-shift harmonic appeared to 'pop out' and was heard with a pure-tone quality."

Another audio production effect that depends on the degree of phase shift is that our hearing of higher frequencies (generally more so than lower frequencies) can briefly be delayed. Then too, phase shift can change timbre.

 

[ray_parkhurst]

Can you give any evidence to show that these anomalies are happening the way you are hypothesizing? Usually hypotheses are based on some evidence rather than just conjecture.

 

[Soandso]

Can you give any evidence to show that these anomalies are happening the way you are hypothesizing? Usually hypotheses are based on some evidence rather than just conjecture.

I am most certainly hypothesizing just as, so far, all related comments have been conjecture. Since we've seen light anomalies in several pictures we have no basis to decide they are simply random artifacts without some function. My conjecture is, due to multiple occurances, we are not looking at purposeless anomalies. Thus I tried to think what those non-anomalies (conjecture) might possibly be and proposed a hypothesis.

Do you think it would be interesting to ascertain if the light scattering artifacts are found in very recent newly recorded vinyl records? If they show up only on older recordings (and/or their re-issue from a master) then I think that may indicate it is evidence of an old production tactic. Namely, there used to be record production with knowingly out-of-phase incidents which are incidents that contemporary record producers no longer employ (whether for style or modern technological options).

 

[ray_parkhurst]

I am most certainly hypothesizing just as, so far, all related comments have been conjecture. Since we've seen light anomalies in several pictures we have no basis to decide they are simply random artifacts without some function. My conjecture is, due to multiple occurances, we are not looking at purposeless anomalies. Thus I tried to think what those non-anomalies (conjecture) might possibly be and proposed a hypothesis.

Do you think it would be interesting to ascertain if the light scattering artifacts are found in very recent newly recorded vinyl records? If they show up only on older recordings (and/or their re-issue from a master) then I think that may indicate it is evidence of an old production tactic. Namely, there used to be record production with knowingly out-of-phase incidents which are incidents that contemporary record producers no longer employ (whether for style or modern technological options).

For sure they don't appear to be purposeless nor random. Most seem to occur on leading edges of high modulation, or on immediate trailing edges, as if a resonance is set up and the stylus has enough energy to damage the opposite sidewall. The Abba album has many more high modulation grooves than the previous (classical) album, so it presents more of the damage sites. The damage extends above the sidewall, and/or into the and across the groove bottom at some sites, which shows that it is more than just a phenomenon of record production and therefore I believe disproves your hypothesis. I don't know what cause them yet, but will be doing more imaging to see if I can come up with an explanation.

 

[MaxwellsEq]

Do you think it would be interesting to ascertain if the light scattering artifacts are found in very recent newly recorded vinyl records? If they show up only on older recordings (and/or their re-issue from a master) then I think that may indicate it is evidence of an old production tactic. Namely, there used to be record production with knowingly out-of-phase incidents which are incidents that contemporary record producers no longer employ (whether for style or modern technological options).

Normal stereo recordings have content where the left channel is "out of phase" with the the right. Left and right positions in the image are a mix of intensity and phase, so a sound that is very much on the right, will have both intensity differences and phase differences with the left. This hasn't changed in recent years.

However (unlike digital), extreme intense content with large phase differences can require reductions to make an LP cuttable.

 

[JP]

I called this "wear" since it only occurs on the leading edges of amplitude peaks. Not sure why a pressing flaw would only happen there, but I am not sure. This was the worst patch I found, and it does go to the bottom of the groove, but many other smaller ones are more localized. Here's a 2D view of another section showing several of these in a row, all on leading edges:

View attachment 287912

Looks familiar. If you don't have the AES disc anthology I'd recommend getting it - there are number of interesting related papers.

 

[ray_parkhurst]

Looks familiar. If you don't have the AES disc anthology I'd recommend getting it - there are number of interesting related papers.

View attachment 290597

Fantastic! Thanks for that reference. Looks a lot like what I'm seeing and is probably the same phenomenon, just in a more controlled environment.

 

[Deleted member 48726]

View attachment 289992

View attachment 289997

Scanning electron microscope images of record groove segment(s) with respective scale bar(s): as per C. Supranowitz, Institute Optics University of Rochester OPT 307;

The mechanical world is, IMO, often much more impressive than the digital. Try to search for MEMS technology. It's fascinating how electromechanical devices we use every day is made and where you find them..

 

[JP]

Fantastic! Thanks for that reference. Looks a lot like what I'm seeing and is probably the same phenomenon, just in a more controlled environment.

IIRC most of the papers attribute it to miss-tracking. Pre-peak it's "welding" and post-peak it's "embossing". In one paper Shure showed that embossing could lead to amplitude increases by as much as 3dB.

 

[ray_parkhurst]

The mechanical world is, IMO, often much more impressive than the digital. Try to search for MEMS technology. It's fascinating how electromechanical devices we use every day is made and where you find them..

Indeed. My last job was Director of Engineering at a MEMS development startup. I also wore the hat of Quality Engineer, and developed some of my photographic techniques to do FA work on the MEMS. We built antenna tuners and RF switches.

 

[ray_parkhurst]

IIRC most of the papers attribute it to miss-tracking. Pre-peak it's "welding" and post-peak it's "embossing". In one paper Shure showed that embossing could lead to amplitude increases by as much as 3dB.

Glad this is already well-known, as it makes it much easier to move forward. I had posited mistracking as the cause but it was only conjecture. Seems that others had the same idea decades ago.

 

[Deleted member 48726]

Indeed. My last job was Director of Engineering at a MEMS development startup. I also wore the hat of Quality Engineer, and developed some of my photographic techniques to do FA work on the MEMS. We built antenna tuners and RF switches.

Really? That's quite a coincidence I wrote that then..

Sounds like an exciting field to work in.