Would you find a modeling pickup useful?

You are right, there would be some added coloration for not placing it even closer to the bridge. Placing in the bridge pickup position is a compromise for practicality. The main factors play out a bit like this: This effect only matters when simulating non-bridge pickup positions. The distance to the bridge has an effect mainly on standing waves, not much of the initial pick attack sound. The physics of the instrument and strings makes the effect most pronounced for the string with the lowest pitch, for the high strings the effect is not significant. The worst case scenario is the low E string of a single coil Strat, where this effect would start noticeably attenuating something like the 12th overtone of the low E string. For a HB Strat and other designs which work out better for this design it would start at maybe the 17th-20th overtone, which I doubt will be very audible.

This is one of the major error sources in the design. I honestly don't know how severe it will be in practice, time will have to tell.



Yes, it will be hex.



Thanks for the pointer! I don't know why but I had not considered selling a fully assembled control plate.



Thanks for your feedback, I really appreciate it. I am about to write another post that explains why I currently am looking at having a "breakout" pedal, but I can't really say much in response other than that this is a niche product. I hope to find a bunch of buyers but I don't think it will appeal to as many as for example the Axe Fx.
I suggest you to write about it also on vguitarforums.com (if you haven't already), I bet there's plenty of people down there which could be interested on something like this.
 
Come to think of it, for the Fractal crowd, a Fractal pickup with ultra-flat frequency response would be more valuable (coupled with a block inside AxeFX which lets you select whatever pickup model you want). Sort of like FRFR, but on the other end of the chain.

This is what I keep thinking when I see pickup modeling mentioned (I think I've said it in a couple pickup threads now :)). Flat response pickup then some block that's like a pre-eq type thing, some added dynamic control to model the different respsonse of magnet types, etc., at the beginning of the chain in the AxeFX. X/Y and/or scene controllable so you can "change pickup" types with your midi controller. Way more than I'd ever know what to do with, but it would be super fun for tinkering and some people would probably do some pretty cool stuff with it
 
I definitely plan to have a mode where pickup simulation is turned off, for an "FRFR-like" experience. It ought to be a super clear sound that makes for beautiful cleans, but will probably require some eq to work with distortion.
 
The Roland VG hex pickup modeling products sound like shit. Even so, they've been a staple in my rig for 20 years. I've found no better way to do Strat-type 2nd and 4th switch positions, acoustic sims, and altered tunings on my favorite dual humbucker guitars. And that is sad.

I was so impressed and excited when I first tried them, imagining what the tech would be like in 20 years. Here we are. The latest products are no better; even worse.
 
Last edited:
...Pretty much the only thing I know about lithium battery pack design is that if you do it wrong they occasionally explode. Burning down your house is not on my list of things I'm hoping to achieve with this project...

Gotta admit, having smoke come out of your guitar during a killer solo would sure impress the crowd, though. :D
 
  • Like
Reactions: Per
It’s an exciting idea with some major technological hurdles. If you position a single pickup that can gather enough information about a string to accurately infer its harmonic content at different locations (and different fretted lengths), and do that for every string, and know where the string is being fretted at the moment, you would have a fighting chance of making it work.

Remember too that commercial success is largely influenced by who is doing the marketing and how they’re doing it.
 
The Roland VG hex pickup modeling products sound like shit.

I've heard this sentiment from multiple places and it's also my impression from listening to demos. Since you have lots of experience using it, is it possible for you to specify in more detail how it fails to sound good? It would be interesting to get more perspectives on what to avoid.
 
I've heard this sentiment from multiple places and it's also my impression from listening to demos. Since you have lots of experience using it, is it possible for you to specify in more detail how it fails to sound good? It would be interesting to get more perspectives on what to avoid.
I think the main problem there is the lack of dynamics, they probably just modeled the frequency response part of the pickups, completely ignoring all the non-linear aspects (compression, distortion, how these change with the distance between the pickup and the strings, etc.)
 
While you mention that you work with Google, I almost got the feeling this was an outside/personal project. Is this true? If so I would be interested in speaking with you and would be happy if you would private message me ( some "outside the box" considerations that impact the hurdles mentioned ).
 
I think the main problem there is the lack of dynamics, they probably just modeled the frequency response part of the pickups, completely ignoring all the non-linear aspects (compression, distortion, how these change with the distance between the pickup and the strings, etc.)

I want to add something to this. The Variax was the "best" attempt at modeling pickups, I think, and it fell short of reality. Part of it, in my opinion, is the comb filtering that creates a nasal quality - just like a poor IR - that is extremely off-putting.
So in addition to a lack of immediacy and dynamic which are both essential elements of the "feel" of the guitar, there was this slight honk that told you this is not the real thing.

The following is entirely not based on anything but an idea:

There might be a way to tune the pickup to the individual guitar to avoid some of the (unwanted) interaction that occurs between the Hex Pickup and the guitar itself that might mess up the modeling. If you think about it: the modeling assumes the Hex pickup is absolutely still and the string the only thing that vibrates. But in real life there is an interaction between the pickup, the strings and the body which resonates. Possibly this interaction messes up the best of calculations and introduces phase errors and additional unwanted comb filtering? Just a thought.
 
Thanks Patzag for your thoughts. I'm not yet at a point in the project where I can actually test this kind of things but I will think about it when I get there.

While you mention that you work with Google, I almost got the feeling this was an outside/personal project. Is this true? If so I would be interested in speaking with you and would be happy if you would private message me ( some "outside the box" considerations that impact the hurdles mentioned ).

This is not a Google project, I could have been more clear about that. I only mentioned that because of the very explicit forum rules and I wanted to avoid breaking rules and upsetting people.
 
I want to add something to this. The Variax was the "best" attempt at modeling pickups, I think, and it fell short of reality. Part of it, in my opinion, is the comb filtering that creates a nasal quality - just like a poor IR - that is extremely off-putting.
So in addition to a lack of immediacy and dynamic which are both essential elements of the "feel" of the guitar, there was this slight honk that told you this is not the real thing.
The problem of the comb filtering IMO is that it's static.
In a real guitar with real pickups you have that same comb filtering that depends on the pickup position relative to the scale, but it's not static, it varies when the tension of the string changes (i.e. when you pick, fret or bend it).

Simulating this would not be easy though, you'd probably need an algorithm that tracks the pitch of each string and which modulates the comb filter based on how much the string tension changes for every fretted note, according to the string motion/envelope and which can recognize bends and vibratos too.

PS: you can play a bit with this simulator to see the effects
http://www.till.com/articles/PickupResponseDemo/
 
Last edited:
The problem of the comb filtering IMO is that it's static.
In a real guitar with real pickups you have that same comb filtering that depends on the pickup position relative to the scale, but it's not static, it varies when the tension of the string changes (i.e. when you pick, fret or bend it).

Yes, the comb filtering varies with string tension and fret position. But is the comb filtering of Variax/Roland stuff really static? I thought that the main reason they use a hex pickup was to be able to apply different comb filters for each string, and I assumed that they also varied that with at least the fret position. It requires pretty fast pitch detection to avoid weird artifacts during note changes so maybe they don't?

I haven't done the math on how much the string tension from plucking the string and/or vibration amplitude affect the filtering, that is an interesting point, I'll check.
 
Yes, the comb filtering varies with string tension and fret position. But is the comb filtering of Variax/Roland stuff really static? I thought that the main reason they use a hex pickup was to be able to apply different comb filters for each string, and I assumed that they also varied that with at least the fret position. It requires pretty fast pitch detection to avoid weird artifacts during note changes so maybe they don't?

I haven't done the math on how much the string tension from plucking the string and/or vibration amplitude affect the filtering, that is an interesting point, I'll check.
Yes, they apply different comb filtering on each string but I'm pretty sure it's static.

I'm glad to read you're considering all these factors for your product, it could be a huge step up compared to Roland/Variax/Antares stuff.
If it will also have good pitch shifting/alt tuning capabilities (Antares level) and the ability to recall patches and settings via midi I'm sold. ;)

PS: you should also consider adding a pitch-to-midi converter, that will make your product a nice all in one solution that could appeal many current Roland users.
 
Yes, they apply different comb filtering on each string but I'm pretty sure it's static.

Huh, ok. If that's the case, it could perhaps explain parts of the "nasal" sound quality Patzag mentioned.

I'm glad to read you're considering all these factors for your product, it could be a huge step up compared to Roland/Variax/Antares stuff.
If it will also have good pitch shifting/alt tuning capabilities (Antares level) and the ability to recall patches and settings via midi I'm sold. ;)

Do you think of something specific when you say "Antares level"? Is it the features (correcting the tuning of the guitar, microtuning), the sound quality, the controls?

PS: you should also consider adding a pitch-to-midi converter, that will make your product a nice all in one solution that could appeal many current Roland users.

That's definitely part of the plan. I will need pitch detection code to feed some filter parameters anyway. It seems a bit silly to not take the small extra step of emitting that as midi. What's the typical interface here? Old school DIN MIDI connectors? USB?
 
Do you think of something specific when you say "Antares level"? Is it the features (correcting the tuning of the guitar, microtuning), the sound quality, the controls?
I'm mainly talking about the sound quality of the pitch shifting, it's literally flawless.
No perceived latency and absolutely no glitches even on extremely difficult (for the pitch detector) situations where both variax and roland struggle a bit, for example on palm mutes and pinch harmonics.

IIRC the method they used for the pitch detection is decribed on the Antares ATG manual.
EDIT: attached below, chapter 1 and 2

That's definitely part of the plan. I will need pitch detection code to feed some filter parameters anyway. It seems a bit silly to not take the small extra step of emitting that as midi. What's the typical interface here? Old school DIN MIDI connectors? USB?
Great! Having both would surely be ideal for the use with hardware synths and vst plugins. But if I had to choose just one I'd prefer the din midi for sure.
 

Attachments

  • 74_ATG_Guitar_Feature_Packs_v3.32.pdf
    1.9 MB · Views: 10
Last edited:
Huh, ok. If that's the case, it could perhaps explain parts of the "nasal" sound quality Patzag mentioned.



Do you think of something specific when you say "Antares level"? Is it the features (correcting the tuning of the guitar, microtuning), the sound quality, the controls?

I completely agree with DLC86 re Anteres. Its very difficult to accurately convert audio data to realistic pitch/amplitude.

Anteres and Melodyne are in the pro realm; most of the Audio to MIDI stuff is not as accurate or sophisticated in its approach (another exception would be Symbolic Sound with Kyma) - but keep in mind that as soon as you try to apply a midi sound the loss of the more fluent actual vibrato and the loss of the audio's amplitude and timbre variations will be noticeable in most cases. Which is why I vote for a dedicated vst/AU instrument (if this is possible) that can do a better job of using the full complement of data available from the pickups.

Re MIDI connector vs USB, MIDI connector signal is so much slower - USB for MIDI would be sufficient for most persons - I think not many people will want to use your obtained data drive an old school MIDI keyboard without a USB connection.
 
Back
Top Bottom