PRS muses on pickup analysis ...

[timtam]

Fireside Chat: A Conversation with Paul Reed Smith, Jamie, and Jack
Dec 31, 2021
"i think one of the things we're going to look back on this year and say what did we accomplish ... i think the progress on pickups actually was probably a turning point for us i would say. i agree. look pickups are black art. it is a very very hard to understand thing. that machine that we have that tells us what it's going to sound like before we plug it in and every time we do this we think uh we're going to go from a to b and we go backwards to z. i mean it's all we call it opposite land. and Hedge started he walked into the office with the sheets. he goes you're not going to believe this one. and that you know once we could measure what was going on we would make a move and then the opposite would happen. and listen it's a real black art the pickup thing. because you've got this inductance and this capacitance and this coil and there's resistance and you're and it's un-model-able really. like you're going to have to do it in the analog domain. it's an eq basically. they have a whistle to them. and we finally started the beginnings of understanding what was really going on. yeah i think we have a long way to go. but from like five years ago to where we are right now boy i feel like we've really come a long way in understanding how to change the tone. i think that's a really big deal for instruments going forward "

Does this inspire confidence in a major manufacturer's understanding of pickups ? Or not ?

[antigua]

Did you transcribe that? If so, thanks!

I like to think people have become more aware of how pickups work the past few years, what with that "inductance and capacitance and resistance" as he says. I'd rather see some spiel about "understanding" pickups as opposed to what was more common a few years ago, like "I spent months YEARS researching what made those original 50's pickups so special and unique". I really do think the mystique has been dying down, not necessarily because of these efforts to myth bust, but because of cheap Chinese pickups on Amazon, probably witling away at people appetitive for very expensive pickups, slowly but surely. 

A couple videos have shown what'd going on with the PRS Silver Sky, a high priced Strat copy, for both domestic and import pickups, and both were sort of unusual, but nothing about them seemed especially beneficial. They reminded me of DiMarzios, having strange differences for the sake of strange differences. 

But this little clip reminds me why I don't like PRS. He seems to have a long standing modus operandi: if one assumes most guitarists are superficial and uninformed, then if sell to the superficial and uninformed, and you will sell a lot of guitars. 

[timtam]

No, I didn't transcribe it ! I used Youtube's auto-transcription (3 dot menu under title:Open Transcript, then toggle timestamps to turn them off, then select all text and Copy. Cleanup in MS Word).

Youtube also does a half-decent job of auto-translation. Useful for Zollner's two German YT channels. Turn on Closed Captions. Then under tools (cog icon), set subtitles to Auto-translate and select English. It does mis-translate "saite" (string) as "side".
www.youtube.com/channel/UCCZxuMUpvAQut978zlKU8Bw
www.youtube.com/channel/UCprM_QgP3ojPVp829olqeNQ

I wonder what PRS's "machine that we have that tells us what it's going to sound like before we plug it in " is. Lemme's box ? Willmott's rig ? Your Velleman approach ? Or maybe just an LCR meter ?

The fact that they think it's telling them the opposite of how the pickups actually sound perhaps suggests they're doing something wrong, or they don't understand it. And the last section suggests they have maybe gone back to just empirical listening/trial and error/looking at DAW EQ curves.

[antigua]

Feb 14, 2022 7:54:57 GMT -5 timtam said:

No, I didn't transcribe it ! I used Youtube's auto-transcription (3 dot menu under title:Open Transcript, then toggle timestamps to turn them off, then select all text and Copy. Cleanup in MS Word).

Youtube also does a half-decent job of auto-translation. Useful for Zollner's two German YT channels. Turn on Closed Captions. Then under tools (cog icon), set subtitles to Auto-translate and select English. It does mis-translate "saite" (string) as "side".
www.youtube.com/channel/UCCZxuMUpvAQut978zlKU8Bw
www.youtube.com/channel/UCprM_QgP3ojPVp829olqeNQ

I wonder what PRS's "machine that we have that tells us what it's going to sound like before we plug it in " is. Lemme's box ? Willmott's rig ? Your Velleman approach ? Or maybe just an LCR meter ?

The fact that they think it's telling them the opposite of how the pickups actually sound perhaps suggests they're doing something wrong, or they don't understand it. And the last section suggests they have maybe gone back to just empirical listening/trial and error/looking at DAW EQ curves.

I didn't know about that YT channel, thanks for pointing it out.

A LCR meter should be able to tell them what a pickup will sound like. The fact that he called it a machine, and not "computer equipment" or "a device" is making me think a Lemme box, it looks the most like a "machine" and some have been spotted in pickup maker's workshops, though maybe not PRS specifically. Once upon a time Seymour Duncan listed peak f for their whole line of pickups, and it's most probable they used the Lemme analyzer to quickly gather all that info, since doing so with an oscilloscope is so much more time consuming.  

The Lemme pickup analyzer device reports back the resonant peak with a built in capacitor dial, and I've had good luck predicting how pickups will sound based on a resonant peak with a test load of 470pF. I think in real life, because I use wireless units most all the time, that my perceived frequencies are above the 470pF benchmarks, but never the less, having all pickups normalized to that benchmark helps me predict the end result with high accuracy, and the inductance values are similarly correlated. If PRS has a Lemme unit, but they don't have any sort of system for comparison, then they would get a lot of useless results. 

As for how they really do it, testing aside, I don't think they do anything special. They just decide the pickup will have 4000 turns or whatever on the coil, and they call it a day. What makes a pickup sound good or bad is subjective, so it's probably arbitrary and random, but leave it to a salesman like Mr. PRS to make it out to be so much more complicated than that. 

When he talks about the black art of pickups, I don't think it's the pickups that are the mysterious block hole, it's the customers, and why anyone likes one pickup more than another, when those pickups are effectively identical. I guy like PRS doesn't know why customers are irrational, nor probably cares, he just happily accepts that they are and exploits the fact for profit, where as some of us make a living in an unrelated industry and only serve to be frustrated by the "black art" status quo.

 

[gpdb]

I don't have the bookmark, but when I was doing my research on TCI and other PRS pickup things, I read that their new way of measuring pickups had to do with Paul Reed Smith's other company, Digital Harmonic (https://digitalharmonic.com/2019/04/04/how-paul-reed-smith-went-from-guitars-to-spy-technology/). Apparently, they have a patented way to extract individual sound waves and analyze them. This is probably what they're referring to - I would assume they had been using bode plots and LCR meters since the early days.

I would love to see (or do myself at some point) a deep analysis between the 57/08 and the 59/09. When looking at the listed specs, these pickups are virtually identical. They both have brass baseplates as well. I believe it's very likely the only difference is that one is covered, the other is not. The Vintage Bass neck pickup is also nearly identical to the 57/08 and 59/09 neck pickups.

[timtam]

I think it's hard to see the connection between PRS's signal processing patents - all with his mathematician dad (and several other mathematicians) - and pickup analysis. Those appear to be basically just different mathematical ways of handling short-time FFTs to improve time/frequency resolution (which may also have application to image processing). How they could be applied to pickup development in ways not otherwise-achievable is obscure. And I'm not sure that PRS has ever claimed to have used that software for pickup work has he/they ? There are well-established methods for pickup analysis, which others have found to yield mostly sonically predictable results. From what PRS says in the video I linked above, they've had a hard time understanding pickup analysis, with sonic results opposite to what they thought their measurements predicted. Although the "progress" they say in the video they have made in the last few years could mean they now understand it properly.

Given what PRS says about his lack of maths skills and lack of qualifications (he dropped out of liberal arts college to build guitars), it's also hard to see his involvement in the complex signal processing work as more than bankrolling/getting investors on board (for whom patents are a sign of innovation, legitimate or not, and thus profit potential).

patents.google.com/?inventor=paul+reed+smith&oq=paul+reed+smith

digitalharmonic.com/precision-measuring-matrix/

PRS expands below in 2020 on his notion of pickups "having a whistle" (mentioned above in the 'chat' video), by which he seems to mean something akin to their resonant frequency. But it does not appear that he's talking about anything that is not established pickup analysis principles.

A bit more detail in this 2019 article, but still nothing out of the ordinary. It does show a pic of "the machine", which looks like it might be an integrated real-time frequency/bode analyser (ie something his dad would probably have used - it has a 3.5" floppy drive !).

www.pressreader.com/australia/guitarist/20190531/282278141800576

Edit: Found "the machine" from that pic. It's an old Hewlett-Packard 35665A Dynamic Signal Analyser (later Agilent, now Keysight). AFAIK it dates from the early 1990s. It was a high-level device in its time, and still commands US$2500+ on the second-hand market. Probably does much the same as the Velleman PCSGU250 can now do for 1/10th that price, given technology development over that time. Not purpose-built for pickup analysis like the Lemme box, so PRS would have needed to devise a protocol for that. Maybe that's what they struggled with ?

www.keysight.com/au/en/product/35665A/dynamic-signal-analyzer-dc-to-1024-khz.html

[antigua]

Thanks for finding all that info about PRS's side business and pickup analysis, that's good stuff. Those bode plotters used to cost thousands of dollars, or you would have to buy a function generator that would somehow integrate with a plotter, and each device could cost over a thousand dollars all by itself, and so the fact that you can buy a similarly capable device for around $150 now is really amazing, and this has only been the case for about ten years. 

[antigua]

Feb 15, 2022 11:20:25 GMT -5 gpdb said:

I don't have the bookmark, but when I was doing my research on TCI and other PRS pickup things, I read that their new way of measuring pickups had to do with Paul Reed Smith's other company, Digital Harmonic (https://digitalharmonic.com/2019/04/04/how-paul-reed-smith-went-from-guitars-to-spy-technology/). Apparently, they have a patented way to extract individual sound waves and analyze them. This is probably what they're referring to - I would assume they had been using bode plots and LCR meters since the early days.

I would love to see (or do myself at some point) a deep analysis between the 57/08 and the 59/09. When looking at the listed specs, these pickups are virtually identical. They both have brass baseplates as well. I believe it's very likely the only difference is that one is covered, the other is not. The Vintage Bass neck pickup is also nearly identical to the 57/08 and 59/09 neck pickups.

I was also thinking their tech sounded like making use of FFT processing to isolate frequency bands of interest. I've actually played around with that with guitar signals, guitarnuts2.proboards.com/thread/7998/tonal-effect-pickup-height , I was able to make one pickup sound like another by manipulating the harmonics with Adobe Audition's really powerful FFT audio processor. If you're analyzing a pickup in isolate of the guitar, there's no practical use for FFT processing, since a guitar pickup imparts to raw signal all by itself. 

If that table had inductance values, then you'd have no need to guess, it would tell you unambiguously how the pickup performs in it's capacity as a passive filter. But all we have to go on is the DC resistance, and it just looks like they said "let's put 50 more turns of wire on that 57 set", it's of no tonal consequence when you consider that the varying capacitance of a guitar cable is going to be more significant than the difference of inductance between those two sets of pickups. It would have been more interesting if they diverged the neck and the bridge to a greater degree, for the two sets, but they didn't even do that. 

[gpdb]

Feb 16, 2022 1:11:10 GMT -5 antigua said:

I was thinking about this yesterday and found a Philip McKnight video comparing both sets, and they sound identical. More than 0.1kohm of error happens all the time between the different pickups of the same model, so I'm sure there's plenty of 57/08 pickups that have less DCR than the 59/09, and vise versa. I'd like to test them at some point, but for my purposes it's more useful to go through SD and Dimarzio's popular options before worrying about PRS pickups which are more expensive and purchased aftermarket less frequently. My hunch is that it's simply an uncovered version. 

[timtam]

Edit: Oops ... I missed the earlier thread today on this:

guitarnuts2.proboards.com/thread/10138/prs-shows-bode-plots

===========================================

PRS now talking about bode ("frequency response") plots and more ...

And still using the old Hewlett-Packard 35665A Dynamic Signal Analyser shown up-thread. They're also measuring inductance and capacitance at 100 Hz, 1 kHz, 10 kHz, and the peak values.

"We can know what the pickup's going to sound like before we even plug it in". (4:00)

Hard to tell if that's an LCR meter here, and if so which one (red face) ....

[Yogi B]

Nov 19, 2022 0:23:18 GMT -5 timtam said:

Hard to tell if that's an LCR meter here, and if so which one (red face) ....

It's an Amprobe 37XR-A, so no not really. While it does do L & C measurements, they're at (semi-)fixed frequencies: 200Hz for 40mH < L < 40H; and 1.3Hz for C < 40μF. In the clip it's set to resistance.

[antigua]

I'm not sure what you call those LCR meters that have the large dial like a multi-meter, but they all cost like $30 and they're almost never suitable for pickup testing, due to the test frequencies being who knows what. On the other hand, the suitable LCR meters like the DE-5000, the Extech or the East Tester ET430 usually have a calculator-like button layout and cost more than $100, often closer to $200. 

This just makes me all the more curious about what they're doing with the data, because it's hard to imagine that it's representative enough to serve any purpose in a production shop. 

[heardsoundcircuits]

Amazing how these are just bog-standard pickup design practices. If he's only recently discovered a bode plot then this is more self-incriminating than it is successfully marketing their brand.

[timtam]

Just noticed what appears to be the rather large exciter coil in the video ...

[timtam]

The configuration of the old Hewlett-Packard 35665A Dynamic Signal Analyser in the video appears broadly consistent with its Swept Sine mode (as described in its documentation below), whereby the analyser generates sine waves at a range of swept frequencies, which appear at its Source output. That output is connected to the exciter and also to the analyser's Channel 1 input, with the pickup output connected to the Channel 2 input.

xdevs.com/doc/HP_Agilent_Keysight/HP%2035665A%20Concepts%20Guide.pdf

xdevs.com/doc/HP_Agilent_Keysight/HP%2035665A%20Operator%201991.pdf

xdevs.com/doc/HP_Agilent_Keysight/HP%2035665A%20Operator%201993.pdf

xdevs.com/doc/HP_Agilent_Keysight/HP%2035665A%20Quick%20Start%20Guide.pdf

[antigua]

Nice detective work. I believe you can use the Velleman PCSU200 software in a similar manner by using the "Spectrum Analyzer" feature. It will also yield the impedance curve of a pickup, but it's not as clean looking, nor convenient as using the bode plotter.

[gpdb]

Amazing catch Timtam. I didn't see that when I was watching the video either. 

[stratotarts]

Two interesting things that actually support each other... the HP probably does not have any integration method. Hence (my theory anyway) the reason for the world's largest test coil, with what appears to be a several kilogram steel core. The huge inductance is necessary to provide an integrated response, the same way that Moore did it. I think Moore's coil was not so physically large though. I wonder how that affects the accuracy? Everyone here has bent over backwards to try to make the test coil essentially invisible.

Also, the plot has the appearance of a linear plot with frequency vs. the log frequency plots that almost everyone else uses. Both things seem like "had to be invented here" syndrome in play. Which from a PR perspective, makes perfect sense.

Note also the apparent lack of simulated load.

[timtam]

Some more info just out from Paul Reed Smith on the history of their analyser above ...

"the tools weren't available so we had to make them. what we showed you was a machine that we built ... it says how a pickup sounds before we ever plug the pickup in. but you can't buy that. so an engineer made it for us ... you know, electrical engineering. he says I can make you a tool that'll show you the sound before you ever plug it in ... 'yeah right sure' ... it took us 20 years to learn how to use that thing. now we live by it"

Is it perhaps conceivable that PRS thought/still thinks they are the only people who can do a pickup bode plot ?

[ms]

"it took us 20 years to learn how to use that thing. now we live by it."

20 years?  Utter nonsense or incredible stupidity!

[antigua]

I just wonder why a company that has been around for forty years would care about pickup analysis, ever so slightly at the twenty year mark, and then finally live by it at the forty year mark. Well there's nothing I can say about PRS that I haven't said before.