ToneDexter vs Baggs VoicePrint ?

[jonfields45]

Originally Posted by guitaniac

I recall that it was mentioned, in one of the VP live demos, that the single-knob resonant frequency cuts target the three most troublesome resonant frequencies of the guitar being analyzed and used to create the voiceprint. I'm hoping that those three cuts will be more effective than what can be done with a single fixed depth notch. In any event, I'm looking forward to the user feedback on TD's, and the VP's, anti-feedback systems.

The resonant frequencies, any number of them and all of them, will all be represented in any IR. It is the low frequency ones that are the feedback problem. Being able to slowly back off the low frequency only IR, is exactly what you want to do to balance the tone enhancement of the IR against a livable feedback tolerance for that performance. It could in reality be for your guitar and pickup one peak, three peaks, half a dozen, but no hardware limit due to backing off the IR instead placing exactly three notches.

Just stepping on the three loudest resonant frequencies is not guaranteed to sound musical and there are plenty of auto feedback pedals already out there which do exactly that, and the ones I've tried sounded bad to me.

If the guitar prior to the IR is already feeding back, then the bass knob or notch or changing the monitor EQ is going to be needed. AND maybe in that situation the VP auto feedback notching is going to be more helpful than backing off the the low frequency IR in ToneDexter.

[SongwriterFan]

Originally Posted by guitaniac

I recall that it was mentioned, in one of the VP live demos, that the single-knob resonant frequency cuts target the three most troublesome resonant frequencies of the guitar being analyzed and used to create the voiceprint.

Are those supposed to be fixed frequencies, or do they vary with each guitar/voiceprint?

If fixed, I wonder if what's being done is similar to what the Align Session (or Session DI) is doing with the "compression" knob? I like the Align Session, and I typically dial in a lot of the "compression" and just a bit of the "saturation" knob (too much of THAT can be a bad thing, IMO . . . but a little bit of it is kinda nice). It'd be nice if the VoicePrint had a bit of "saturation" you could dial in, too, without having to have the separate pedal.

[SpruceTop]

Frankly, and going against the tide, and given that I have a ToneDexter and a Fishman Aura Spectrum (and I've had all versions of the Aura onboard and pedals), and haven't used either in two years while playing only at home, I prefer the amplified tone of a Dazzo, a Trance Amulet, an Ovation, an LR Baggs Element, a Taylor ES2, and the Takamine Palathetic transducer systems without the IR stuff added into the mix. If needed, just add a little bass and subtract a little mid-range, and all these systems to me sound usable without the addition of something else in the mix. If we have to talk IR and best-sounding tone with IR enhancement, let me say that my long-gone Ovation Adamas 2087 NWT with VIP-5 Preamp (onboard Aura Images) sounded the best of any IR-enhanced guitar I've yet to hear.

[guitaniac]

Originally Posted by SongwriterFan

Are those supposed to be fixed frequencies, or do they vary with each guitar/voiceprint?

If fixed, I wonder if what's being done is similar to what the Align Session (or Session DI) is doing with the "compression" knob? I like the Align Session, and I typically dial in a lot of the "compression" and just a bit of the "saturation" knob (too much of THAT can be a bad thing, IMO . . . but a little bit of it is kinda nice). It'd be nice if the VoicePrint had a bit of "saturation" you could dial in, too, without having to have the separate pedal.

When you run the protocol for creating a voiceprint, the VP DI analyzes the specific resonances for the guitar being used to create the voiceprint.

With the newest TD software, TD also does the same thing. It analyzes the low range resonant peaks for the particular guitar that you are creating the wavemap for, and you can subtract a variable amount of those peaks (using a single knob) from the low end frequency range. To use the single-knob anti-feedback feature effectively, one obviously needs to use TD with the guitar which created that particular wavemap. Apart from that caveat, it can sometimes be fun to try mixing and matching guitars and wavemaps. James May has gotten excellent results using a Yamaha Silent Guitar with wavemaps created from a nylon string acoustic/electric.

If I recall James's explanation correctly, I believe TD's notch knob is the knob used when TD is put in the mode for using the new single-knob anti-feedback feature. If that's correct, the knob wouldn't be available as a manually sweepable fixed-depth notch when in the aforementioned mode.

[guitaniac]

Originally Posted by SongwriterFan

Are those supposed to be fixed frequencies, or do they vary with each guitar/voiceprint?

If fixed, I wonder if what's being done is similar to what the Align Session (or Session DI) is doing with the "compression" knob? I like the Align Session, and I typically dial in a lot of the "compression" and just a bit of the "saturation" knob (too much of THAT can be a bad thing, IMO . . . but a little bit of it is kinda nice). It'd be nice if the VoicePrint had a bit of "saturation" you could dial in, too, without having to have the separate pedal.

I took the trouble to find James May's initial explanation (on this forum) of how ToneDexter's single-knob anti-feedback feature works with the Version 2.0 firmware. You'll find it on the 1st page of AeroUSA's thread entitled Tone Dexter Version 2.0!.

The pertinent post is post #11 on the 1st page of the thread. The post is dated 11/28/2021. I'll copy and paste James's words and the initial question from shufflebeat, but the original post includes a very helpful graph which I can't transfer with a simple copy and paste. I seem to recall that you can also find the graph on the Audio Sprockets website.


Originally Posted by shufflebeat
To clarify:
The notch now applies a filter at a frequency pinpointed during the training process and is unique to each wavemap, is that correct?

Almost. It's more sophisticated than that. The low end resonance profile is learned during training. Whatever is peaky below about 440Hz. It's not just one frequency but typically is several an an acoustic flat top guitar. And, it's not just the frequencies but also the exact shapes that are learned. Some are narrow, some are broad.

Below is a graph of the measured response through ToneDexter showing the low end resonance profile of my D-18 style Voyage Air. You can see the feedback reduction steps from 0% none, to 100% full in 10% increments. At none, the two main peaks are sticking up 12-13dB. At 100%, they are essentially flattened. With the notch control, you can dial up as little or as much feedback reduction and tone shaping as you need or want.

Name: vairable feedback plot.jpg
Views: 313
Size: 30.1 KB

What's mind blowing is how useful this is live. If you're having feedback, turn up the notch dial till it stops. You don't have to hunt and find the offending frequencies. If you aren't having feedback issues, then adjust to suit your preference in how much of the acoustic guitar bloom you want to hear.

I should add that up until V2.0, this same feedback mitigation was baked into every WaveMap (except those started from slot 22). With V2.0 and going forward, it is no longer baked in, but rather adjustable during live playback. Slot 22 no longer has special significance when creating new WaveMaps.
__________________
James May
Audio Sprockets
maker of ToneDexter
James May Engineering
maker of the Ultra Tonic Pickup

[SongwriterFan]

Originally Posted by guitaniac

I took the trouble to find James May's initial explanation (on this forum) of how ToneDexter's single-knob anti-feedback feature works with the Version 2.0 firmware.

Thanks for finding that.

I still wonder, though, what happens (in version 2.0), if you put the Character knob fully CCW (which is zero wave map, 100% pickup, from my understanding) what happens when you engage the Notch knob?

I think it would still make sense to do exactly what it's doing (as you quoted James as saying above), in that those same frequencies are probably the offending ones even if JUST listening to the raw pickup.

[guitaniac]

Originally Posted by SongwriterFan

Thanks for finding that.

I still wonder, though, what happens (in version 2.0), if you put the Character knob fully CCW (which is zero wave map, 100% pickup, from my understanding) what happens when you engage the Notch knob?

I think it would still make sense to do exactly what it's doing (as you quoted James as saying above), in that those same frequencies are probably the offending ones even if JUST listening to the raw pickup.

Here's the relevant section, regarding the use of the notch, from the Version 2.0 Users Guide. You can find the various user guides in the support section of the Audio Sprockets website. I've put the most relevant paragraph in bold.

Modes of Operation
RUN Mode
This is the normal mode of operation for playing. The display
shows R (for run) along with the WaveMap number being used. If
the number has a dot next to it, it means there is a WaveMap in
that slot and it will be heard. If there is no dot, the slot is empty,
and the signal passes straight through unaltered except by EQ
and/or the feedback notch if engaged. Use the WaveMap Select
knob to choose between 11 WaveMap storage locations in each
bank. The Option A/B switch allows selecting between the lower
bank 1-11 and the upper bank 12-22. When changing the selection, the sound will immediately change to the WaveMap stored in
that slot.
During RUN mode, the Notch control will provide smart feedback
reduction, adjustable from none to full. This control defaults to a
simple notch when in bypass, playing from an empty slot, or playing a WaveMap created prior to V2.0 firmware.

By the way, reviewing the user's guide reminded me that the bass and treble controls are shelving EQs. Check the user's guide for more detail on that.

[Chriscom]

Originally Posted by guitaniac

I took the trouble to find James May's initial explanation (on this forum) of how ToneDexter's single-knob anti-feedback feature works with the Version 2.0 firmware. You'll find it on the 1st page of AeroUSA's thread entitled Tone Dexter Version 2.0!.

The pertinent post is post #11 on the 1st page of the thread. The post is dated 11/28/2021. I'll copy and paste James's words and the initial question from shufflebeat, but the original post includes a very helpful graph which I can't transfer with a simple copy and paste. I seem to recall that you can also find the graph on the Audio Sprockets website....

Here's the graph fwiw. And what the heck, link to the thread guitaniac labored to find https://www.acousticguitarforum.com/....php?p=6562319

[guitaniac]

Originally Posted by Chriscom

Here's the graph fwiw. And what the heck, link to the thread guitaniac labored to find https://www.acousticguitarforum.com/....php?p=6562319

For some strange reason the link takes me to the first post in the thread and then jumps to post #14. However, its easy enough to just scroll up to post #11 which contains the graph and the James May commentary. I appreciate the assistance because posting a link to the thread makes it much easier to find the graph, and the graph is quite helpful to understanding Version 2.0's single-knob anti-feedback feature.

Oops! I see now that you've figured out how to transfer the graph to this thread. (It wasn't there initially.) Its good to note that its possible to do (for someone who knows what he/she is doing). Thanks again.

[Chriscom]

Originally Posted by guitaniac

For some strange reason the link takes me to the first post in the thread and then jumps to post #14. However, its easy enough to just scroll up to post #11 which contains the graph and the James May commentary. I appreciate the assistance because posting a link to the thread makes it much easier to find the graph, and the graph is quite helpful to understanding Version 2.0's single-knob anti-feedback feature.

Oops! I see now that you've figured out how to transfer the graph to this thread. (It wasn't there initially.) Its good to note that its possible to do (for someone who knows what he/she is doing). Thanks again.

My pleasure, and fwiw I tried fiddling with that link behavior but ¯\_(ツ)_/¯

For the image, the simplest path on my Windows laptop is to hover your mouse over it, do a right-click to access some options including "save image as," and since it's a .jpg image, select that and save. (the forum appears to accept jpg an .png formats). Then at the bottom of the screen you're posting a message on, you have to scroll down for the "manage attachments" option for a little more noodling to upload.

Edit: I believe this works only for "Charter members."

[guitaniac]

Originally Posted by Chriscom

My pleasure, and fwiw I tried fiddling with that link behavior but ¯\_(ツ)_/¯

For the image, the simplest path on my Windows laptop is to hover your mouse over it, do a right-click to access some options including "save image as," and since it's a .jpg image, select that and save. (the forum appears to accept jpg an .png formats). Then at the bottom of the screen you're posting a message on, you have to scroll down for the "manage attachments" option for a little more noodling to upload.

Thanks for the tips for transferring images, Chris. They should come in handy on this forum.

[James May]

Originally Posted by SongwriterFan

Thanks for finding that.

I still wonder, though, what happens (in version 2.0), if you put the Character knob fully CCW (which is zero wave map, 100% pickup, from my understanding) what happens when you engage the Notch knob?

I think it would still make sense to do exactly what it's doing (as you quoted James as saying above), in that those same frequencies are probably the offending ones even if JUST listening to the raw pickup.

The variable smart feedback mitigation works on the WaveMap signal only. If you are using a blend such as 40% WaveMap / 60% raw pickup, then only the 40% WaveMap portion will be subjected to the smart feedback control.

This is in contrast to running in bypass or using software earlier than V2.0. In that case the notch control is a single sweepable notch and applies to everything, no matter what the Character/Blend control is set for.

[SongwriterFan]

Originally Posted by James May

The variable smart feedback mitigation works on the WaveMap signal only. If you are using a blend such as 40% WaveMap / 60% raw pickup, then only the 40% WaveMap portion will be subjected to the smart feedback control.

Thanks for the clarification. So, in this case, the 60% of the signal coming from the raw pickup might still require a notch filter somewhere (another pedal, or at the main board) to help eliminate feedback from the monitors?

[James May]

Originally Posted by SongwriterFan

Thanks for the clarification. So, in this case, the 60% of the signal coming from the raw pickup might still require a notch filter somewhere (another pedal, or at the main board) to help eliminate feedback from the monitors?

If it's a UST, then probably not. If it's an SBT, then perhaps. Depends on how feedback prone your SBT is to start with.

Based on experience, we don't expect most of our guitar playing users to use the blend function, since it generally sounds much better not too. In which case you won't have the potential issue.

Our upright bassist are another story, as they tend use the blend much more often, for more complicated reasons, but generally don't have the need for a notch with their raw pickups.

[guitaniac]

Originally Posted by James May

The variable smart feedback mitigation works on the WaveMap signal only. If you are using a blend such as 40% WaveMap / 60% raw pickup, then only the 40% WaveMap portion will be subjected to the smart feedback control.

This is in contrast to running in bypass or using software earlier than V2.0. In that case the notch control is a single sweepable notch and applies to everything, no matter what the Character/Blend control is set for.

Thanks for that important detail, James.

Is it correct that one will get 100% of the anti-feedback notch's cut potential with the character knob at 12 noon (CH0)? If that's the case, a noon character knob setting would provide the driest sounding signal that the user can get without sacrificing some of the anti-feedback notch's cut potential.

On the other hand, 40% of the anti-feedback notch's cut potential (by rotating the anti-feedback notch fully clockwise for a 40% wet signal) may be enough of a cut (or more than enough) to do the job with a 40% wavemap/60% dry signal blend.