Improving the LR Baggs Lyric sound

[dannyg1]

Just to be clear, and sans comedic comments, what you've made here is better than just good. This thing sounds fantastic and I want one.

[Cuki79]

Hi everyone

I did a comparison with a piezo

So I recorded my good all Yamaha FGX412 in many ways:
* DPA
* Piezo alone
* Piezo + Fishman Aura (Yamaha FG412C image) Full Blend (100% image)
* Piezo + Fishman Aura (Yamaha FG412C image) 50%
* Logidy EPSI IR I just made from the piezo (Full Blend 100% IR)
* Logidy EPSI IR I just made from the piezo (Blend 67% IR)
with struming also

Playlist (one missing sample you can find at the end: strum EPSI full blend, I think playlists are limited to 10 samples):


DPA Sample


Piezo


Aura 50%


EPSI Blend


EPSI Full


DPA struming


Piezo struming


Aura Struming 50%


EPSI Blend strum


EPSI Full strum


I think the logidy compares well to the Fishman Aura.
Tell me what you think.

[Cuki79]

Hi Dannyg1

I am on vacation tomorrow so I won't have time before I am back. Send me a PM if you are really interrested.

Don't buy the Logidy EPSI yet. When your IR is made, you can first try with a plugin (ex: free NadIR convolver).

Do you have a plugin host? (Cubase, Garageband, Ableton...)

If you have a Mac, peadlboard is great. It's free and you can set the latency to taste.
http://www.niallmoody.com/apps/pedalboard2

PS: I did not say no but I didn't make any promises. It can be made remotely but I need you to explain what to do.

Cuki

[dannyg1]

I'd be using the thing out on the street, so I'd need the EPSi in any event (Can't use my computer out there. It doesn't run nearly long enough). It would be fun to make images for each location I typically play. I've got the mics and a fairly good interface (I use a Zoom R24 as my interface these days, plugging a Focusrite ISA as my mic pre).

The strummed samples through the EPSi seem to be (To me at least) 6-10db down from the rest. Is that because of the image you're using or is it a compression hit?

[Cuki79]

Hi

The EPSI adds a lots of ultra low end (40-70Hz) because I make the IR from the DPA. The DPA is a clip on mic (see link)
http://www.dpamicrophones.com/microp...one-for-guitar

There is a huge proximity effect. So the "Ompfff" (40-70Hz) is strong and there is a lot of energy in there. Because of normalization, it makes the other frequencies sound weak (well that's what I think). The effect is also very noticable when you use the NadIR convolver (I have to put the volume and gain of the plugin to max).

This is of course tweakable. I made few tests and I noticed that cutting all these frequencies did not affect the end sound. But I did not normalized afterward, so I can't tell if you get back the volume. It would be logic.

For my own IR, I like to keep the "Ompfff" strong. I got a SR Tech JAM 150+ and it is pure joy in the lower register. Of course it's not practicable for LIVE use, because you get feedback sooner. But the feel of the air moving/ vibration is so cool, I keep it for home use. It's like asking a guy with a ful stack Marshall to put the volume down...

Anyway, for LIVE use I will make myself a more usable IR with all the unwanted frequencies cut. Those things really depends on personnal taste.

I forgot to ask you. If you want me to make an IR, you need to have a microphone and an audio interface (presonus, focusrite, motu...)... Because you're not going to send me your guitar of course I need you to records wav files and send them to me. I'll send you back the IR file, and you can try it on your computer.

[Mischief]

Originally Posted by Cuki79

Hi



The EPSI adds a lots of ultra low end (40-70Hz) because I make the IR from the DPA. The DPA is a clip on mic (see link)

http://www.dpamicrophones.com/microp...one-for-guitar



There is a huge proximity effect. So the "Ompfff" (40-70Hz) is strong and there is a lot of energy in there. Because of normalization, it makes the other frequencies sound weak (well that's what I think). The effect is also very noticable when you use the NadIR convolver (I have to put the volume and gain of the plugin to max).



This is of course tweakable. I made few tests and I noticed that cutting all these frequencies did not affect the end sound. But I did not normalized afterward, so I can't tell if you get back the volume. It would be logic.



For my own IR, I like to keep the "Ompfff" strong. I got a SR Tech JAM 150+ and it is pure joy in the lower register. Of course it's not practicable for LIVE use, because you get feedback sooner. But the feel of the air moving/ vibration is so cool, I keep it for home use. It's like asking a guy with a ful stack Marshall to put the volume down...



Anyway, for LIVE use I will make myself a more usable IR with all the unwanted frequencies cut. Those things really depends on personnal taste.



I forgot to ask you. If you want me to make an IR, you need to have a microphone and an audio interface (presonus, focusrite, motu...)... Because you're not going to send me your guitar of course I need you to records wav files and send them to me. I'll send you back the IR file, and you can try it on your computer.

I am sold on imaging. I think I would like to explore this avenue. But I have a few questions.

1. What is the benefit of using the lyric as your pick up source vs piezo when using an image?

2. This likely relates to question 1. But what what are the pros and cons of running say 50% blend vs 100% image?

3. Is there any advantage to using the EPI over buying a fish an and loading that with a custom image or is that not possible? The reason I ask is I can buy a Fishman Aura pro for the same price as the EPI box.

Lastly is it complicated to record properly for image creations? Is it similar to recording good studio guitar where room sense sounds etc play a large part or is it something that can be done with a descent mic and tracking in a bedroom sort of thing?

Thank you for your insight and great contribution with your project.


Sent from my iPad using Tapatalk

[Cuki79]

Hi Mischief,

> 1. What is the benefit of using the lyric as your pick up source vs piezo when using an image?

Well, I think the IR made from piezo is as good as the one from the lyric. Why did I use a Lyric then? Because of my personal trip in the acoustic amplification world. I first used the Yamaha piezo. Then the Aura, then I was happy. But then I wanted a better guitar and there is one thing I did not like with the Aura: It feeled "digital". My guess is first the Aura IR uses a blackfin DSP and has only 2000 taps, see link:
http://archive.eetindia.co.in/www.ee...P_7d5cc0de.HTM

So I went to the Taylor ES1, and then I was very frustrated, it sounded great in the low register and very magnetic up in the neck (because of the comb filtering). See link
http://www.till.com/articles/PickupResponse/

I was never convinved by the LR Baggs Anthem, because the element piezo pickup is one of the worst sounding piezo I know (Sorry for Baggs). So when the lyric came out I was sure I needed a try. So I bought one second hand and waited until I had a new guitar. I read everything on the forum about it, so I spent a lot of time EQuing Lyric's demo with Plugin on my computer. To finally conclude that it needed a wide mid cut (or multiple surgery mid cuts) and a huge boost around 80 Hz. So I also bought: a headway EDB1, a TC electronic DPEQ and the Dtar EQuinox... Then I realized that Imaging would do better.

But when martingitdave said in that thread that he would probably buy a Tonedexter with piezo pickups in each of his guitars, I thought I could make a demo of my IR with a piezo. Without martingitdave I probably would not even try.

Now I should answer the quesion PRO vs CONS
* Lyric
PRO: dynamic response (Mic response + preamp multiband peak/RMS Compression). Less digital correction than a piezo, because it senses the top vibration and a bit of air.
CONS: feedback. You can EQ out the bass but it should feedback sooner than any piezo. Well it depends on your playing. If you are playing in a loud band. It can/will be a problem (but In a loud band you don't need the bass)

* Piezo
PRO: better against feedback. Its cut through better.
CONS: More correction. Piezo dynamic. The piezo compress/saturate faster. no IR modeling can correct that. So the "feel" will be the same. Note that Baggs really worked well on the Lyric with the preamp compression that mimics the feel of the guitar top response.

For me the answer is in the use of the pickup. It depends on people.

[Cuki79]

Hi Mischief

Question:
2. This likely relates to question 1. But what what are the pros and cons of running say 50% blend vs 100% image?

This is something you can only feel with the pedal and an amp. Soundcloud does not help.

For The EPSI:
* With headphone, listening on the computer, I felt that 100% image sounded better. It was true to the DPA microphone.
* Playing LIVE with an amp, it lacked a bit of "air"/"directness". For example, the lyric is very midrangy with a lot of top end (I put the presence screw to max to have no signal loss). In that configuration, the guitar feels "hot", any tap on the guitar side will be amplified. I call it "air" feeling. With the EPSI, a lot of that is lost. Once again because the bass are back and because of normalization, the energy that is not in the high register is in the bass (and bass takes a lot of energy).

To get back more air, instead of raising the highs, I though adding a bit of a dry signal would to give back a bit of "directness" and it does. So playing the guitar feels much better with 33% dry signal blend. Note that it's not a matter of latency, since the dry path stills gets through the A/D and D/A converters.

For The Fishman Aura:
* The image of the Aura is nowhere as good in term of digital quality as the ones you make yourself. The reason why: 12 years of DSP evolution. I once read a scientific article (early 2000s) that predicted how much DSP calcualtion power you needed to fool humans ear for these applications. Well 15 years later, we are nearly there. That's for example the Kemper (see Chappers video)
https://youtu.be/INJ_H5PiuTE

The fishman has 2000 points IR and I feel a bit of latency (more than the EPSI for sure). The EPSI in "zero latency" mode computes 1.5s * 44100Hz= 66150 pts. Thats a 33 factor improvement (latency set aside). It's huge.

In my opinion, the fishman Aura does not sound as good with the image at 100%. Well it's not bad in headphones but in LIVE and amplified situation it does sound muddy.

So you need to blend usually 50% of the signal.

PRO and CONS 50%-100% Blend: It's just a matter of taste. There are absolue no drawback in term of feedback or usability. Except maybe latency with the Fishman: because if think the dry path is all analog so you'd better keep some dry path on.



Question 3:

3. Is there any advantage to using the EPSI over buying a fish an and loading that with a custom image or is that not possible? The reason I ask is I can buy a Fishman Aura pro for the same price as the EPSI box.

Fishman:
PRO: Great studio mic and preamp for imaging (I cannot afford Neumanns mic). You can send your guitar to have a custom image (You don't have to work but I am not sure about the cost: ask Fishman). In that case, you'd better use Fishman Matrix (great piezo when well installed). Moreover, if you buy a Fishman Spectrum DI you get EQ, Feedback supression, compressor and DI.

CONS: Digital quality is outdated. Sound is not that good (in my opinion but you can check the soundcloud Aura 100%). You have to use the IR they made with the EQ they've used. For example, they use a low-cut on every IR (I think). Some people also complained about the slight room reverb that is always there. (+if the Aura is inside the guitar like in Martin Retro or new Martin E standard series: battery eater).

EPSI (or tonedexter, AMT Pangea, Torpedo CAB...):
PRO: You can make your own IR... OK that's your "cheap" microphone, preamp, audio interface and room (In my samples: DPA 4099G+focusrite 2i2+living room, nothing fancy) but you get a much better digital quality from the EPSI. (And even more from a computer)

Note that I can't say how many taps are used by the Tonedexter. I asked in this forum or geaslutz and the guy from Tonedexter never answered. Well you can calculate from the size of the IR allowed.

Morevover if you have the knowledge, you can actually EQ everything. For example you can configure the thing in the "Anthem" way, with a crossover tuned so that the pickup handle the low end and the IR the high end. You can add a any number of parametric EQs on the dry or IR path.

Also if you get into it you can actually EQ out the room. I've never tried that. But in principle, it works. This pedal is meant to make reverbs. You make an IR of the place you like to capture, for example the natural reverberation of a church. You put it in the pedal. But the operation can be used in the reverse way. For example in my case I EQ out the boxiness of the guitar, by getting rid of the internal reverberation of the guitar and apply the response of a mic in a room (the DPA). So in principle, you can capture the resonances of let's say the restaurant you play every week-end, and correct that in a second. The only drawback is that the speakers position, amplification and restaurant... should be always the same (that is mostly the case). The only limitation is that you are allowed to have 999 restaurants in the EPSI pedal, or you have to buy an extra SD card

But in that last application, I recommand to have a second convolver because the effect should be applied to both vocals and guitar...

Finally, the EPSI can work in stereo. As I said before I don't have the audio interface or microphones to make an XY or MS demo. But it should work. Guitars have a beaming effect, see video (9:20)
https://interstices.info/jcms/c_6167...are-acoustique
Capturing that would be great.

CONS: Low live tweakability. Well you can have many IR, so you can make one for each Blend configuration. The only parameters are 3 band EQ but on electric guitar frequencies (on the EPSI)... For me it's not really a problem because Fishman Aura is an "always on" type of effect. I am not sure, you actually tune anything during a show.

[Cuki79]

The last question
> Lastly is it complicated to record properly for image creations? Is it similar to recording good studio guitar where room sense sounds etc play a large part or is it something that can be done with a descent mic and tracking in a bedroom sort of thing?

No and yes. At the begining I had a sustain problem with a lot of violoning effect due to unwanted digital resonances. I wrote to EPSI and the guy said that my recordings were not ideal... The good thing is that you d'ont have to play great!! IR is the same whether you are a great guitar player or a crappy one.

To conclude, all I can say is that for the demo you've heard:
* The recordings were made on the same couch you see in the youtube video. It's my living room, no acoustic isolation but a carpet full of kid toys...
* The DPA (which is a good mic but for LIVE use) goes straight into the focusrite 2i2 which is a cheap audio interface.
So if that quality is OK for you, my answer is yes, it is what you get from bedroom quality recordings.

Fishman gives access to great studio and great mic and preamp. But:
1) I don't know how much do they charge for that
2) You can't tweak anything because the guy from the studio makes the sound.
3) You have to send your guitar.
4) It's recorded by some stranger using I don't know which pick (It's not you playing).
5) at the end you get a 2000 pts IR (45ms) which is what you get with Line 6 Helix for example. For me nowhere as good as the EPSI.

For me your problem is really easy to solve because you don't need to buy anything to try a homemade IR because any decent computer can do that.

So you can borrow a Fishman aura from a friend and try. If you look on the net. I've seen a guy who actually made IR from the fishman Aura pedal and let people download it.

Make your own opinion.

Nota Bene: Fishman aura Pro is said to have 54h battery life with lithium and 27h with alkaline. The aura is known as being a battery eater. And you have to drill a huge hole to put that in your guitar.

PS: Don't think I have any hard feeling against Fishman. I do own a Fishman Aura. And the day I bought that thing really change my way of thinking. Fishman's idea is great. The pedal was released in 2003. The Kemper in 2012. Today we are 2016. Well the Aura is nearly vintage nowadays.

[Cuki79]

Oups

I've realized that you don't have the Aura Full link

[Cuki79]

I would like to add something.

I am no real musician. This thread is more about experiments than road worthy practical music tools.

I think the two people I've read the most in the forum are:
SpruceTop and Vancebo (and of course Doug Young).

Vancebo (I believe from what I read) has settled to soundboard transducers (Dazzo) after testing everything. SpruceTop has new Aura VT, Taylor ES2 and TA Amulet.

I'd like to quote a recent thread from Vancebo:

After owning numerous gear over the years, I settled on something real simple and it has been the best I have ever owned, and it's not expensive. I have a handmade Dazzo Pickup ($215) through a Red Eye preamp ($199). That's it. No matter how hard I strum, no quack. I get the benefits of a mic-like sound but it's on steroids because I can easily shape the tone with simple 3 band EQ (on my amp) for any situation I play in from rock band to solo fingerstyle.

http://www.acousticguitarforum.com/f...1&postcount=19

Sometimes there is a difference between what we think is good ("sounds like my guitar only louder") and what we actually need ("cut through a noisy environment without feedback").

If you gig a lot. What you need is something simple and efficient. I usually try not to add something in the signal chain if the added benefit is not way superior to the cost...

If you read Doug Young booklet on amplification (great booklet), you will understand that acoustic amplification is everyway a compromise. Doug did a great table with all PRO and CONs of every technology.

So before buying a modelling/digital pedal based on threads, I invite you to read what those three people wrote. You may save a lot of time.

PS: However another guy, I don't remember who, said that everyone has to make his own way... He said that whatever people say about guitars, you will have to try them all and enventually own them before finding the right one. I believe it is also true because: you are not chasing a guitar... You are chasing yourself as a guitar player... And only once you've find yourself, you know what you need.

Well I don't want to have dozen of people angry at me because they bought an IR convolver and did not get what they expected. (that's also why I suggest you tried first with a computer).

I can quote the guitarist of my band, last week he said "It's really great, we must now test in a loud gig". That's pretty wise don't you think?

[Mischief]

Cuki79
Thank you so much for your insight. This is absolutely fascinating. I see I have so much to learn yet.

I agree that what works or is great for one sucks for someone else.

So far on my journey to achieve a great plugged in tone. I've tried many pickups and combinations. I have let go the idea of my guitar only louder and just want a natural plugged in sound. I think the concept and principles of imaging are great. Especially if one can image their own guitar.


For me I now have a Tak pro with onboard CT-3 tube preamp and EQ. So is the imaging effected if using EQ or tube presence prior to the Imaging box?

I've been thinking of adding a second source to the Aux input of my onboard preamp. So I'm not sure how that effects things.

So far all I know is good imaging audio sounds natural to my ears.

[Cuki79]

Hi Mischief,

>For me I now have a Tak pro with onboard CT-3 tube preamp and EQ. So is the imaging effected if using EQ or tube presence prior to the Imaging box?
The tube or EQ is not a problem. But it's better that you make the image with everything flat and keep it this way. Of course nothing is forbidden. But when you make your "Image", it is made with the CT-3 knobs in a certain position. So for best performance, you should keep them in the position used when the image is done.

One can see the EPSI IR as a very complex EQ (as said in the description of the Aura in one of my previous post reference). In the case of the EPSI a 66150 bands EQ with amplitude and phase.

So if one adds or cut bass or treble before the EQ... well there will be less bass and treble. But there will probably be an extra mess in the phase.


> I've been thinking of adding a second source to the Aux input of my onboard preamp. So I'm not sure how that effects things.

I believe that if you choose this way, you don't neef an IR/image. Fishman uses the Aura as a second source (now they've even added a third one, with a soundboard transducer in the Aura VT enhance).

The cool thing about the CT3 is that it provides gain to the aux source. It can stand "medium" impedance which excludes any piezo (that would need a high impedance input) but not microphones. Since you have no DC voltage available you can't use a condenser microphone straight. So you have to find a dynamic internal mic... (don't know any)

Or you can use something like the highlander mic.
http://highlanderpickups.com/catalog/hmic.htm

it has built-in preamp and you only have to make a custom cable
RCA cable + 9V cable --> 3.5mm stereo jack
and use a standard battery bag


Then you get no digital stuff, no latency... and It would basicaly cost the same price (if you have access to a soldering iron). It is a solution that envolve very light modification to your guitar (if you can stand the battery bag in the guitar)

If it does not work for you, you can still resale the highlander. You only lose a battery bag, a homemade cable.

NB: Before doing anything, I would write to Takamine and ask how much AUX gain you have (+10, +30 dB?) and what is the input impedance of the AUX input. You should also have at least basic electronic knowledge.

On the other hand, you can easily try modeling in any shop. Just ask to try:
* Fishman Aura,
* Zoom A3
* TC BodyRez (or Voice Live Playacoustic)
* The new BOSS AD-2 (or AD-8 and VE-8)

If you want to try the IR thing send me a PM so I explain you what I need to make you an IR.

Basically you need
* An audio interface with two inputs. One mic input and at least one input instrument input (with high impedance).
* A microphone
* A google account to google drive me the files I need.

Then you can try your IR on any free IR convolver (ex: NadIR ingnite amp), with the latency of your computer/audio interface of course.

[Doug Young]

Cuki, this is a very cool bit of work. Can you elaborate on your process for making the impulses? What did you play? How long were the samples? I'm curious what the inputs and outputs of the matlab process looked like. I'm thinking of giving this a try myself.

[Cuki79]

Hi Doug,

First I have to say, I am very grateful for your work towards the acoustic guitar community. I've listen to your pickup tests, bought your booklet on amplification and read many of your articles.

For the MATLAB process, It is very simple.

I've recorded two samples at the same time thanks to the two channels of my audio interface.

* Channel 1 takes the DPA
* Channel 2 takes the pickup

Channel 2 must be set as instrument input with a high impedance. (This important because the EPSI pedal has a rather high impedance input: 500kOhms)

The two samples are then loaded in MATLAB as two separate wav files. (but you can use Scilab or Octave that are free)

Then it's quite simple. I apply a Fast Fourier Transform (FFT) algorithm to go into the fourier domain for both signal. Convolution is a standard product in Fourier space, so I only divide both Fourier transforms then apply the inverse Fourier transform to get the IR in time domain. If there is a bit of imaginary part left, I only keep the real part (from the complex numbers). I also normalize the signal.

Back in time domain you have to "fade out" the signal by multiplying by a chosen envelop. But you can also do it with audacity (free and simple). Long IR are well handled by a computer but not by the EPSI pedal (in CAB/zero latency mode).

Note than in the Fourier domain you can apply any filter, crossover functions you like. You just need to use fftshift because Matlab filp positive and negative frequencies.... Maybe I am going to fast, there are negatives frequencies because we use complex numbers. In Fourier space, a cosine (pure frequency) is the sum of two complex terms with positive and negative frequency.
Anyway, MATLAB also have a filtering functions that you can apply in time domain. If there are digital resonances that create weird artifacts, it can be corrected this way.
I forgot: The dry path is made by adding a dirac delta function.


Question:
>What did you play?
I played what I usually play. Some struming (cowboy chords), some arpeggio, The A and Am pentatonic scales along the neck (did not do all positions: only one run) and the intro of Little wing (well something a bit like the intro of little wing...).

As a I explained to Mischief in a PM, my point of view is the following:
=> I don't try to capture the sound of "a guitar", I try to capture "Someone's playing his guitar". So what he plays does not matter as soon as he is the one playing. So my advice to Mischief for his IR was: "Play what you like to play".

>How long were the samples?
around 4 minutes for the Lyric
around 5 minutes for the piezo

>I'm thinking of giving this a try myself.
Please send me a PM, I'll give you the address to google drive me the wav files. I would be glad to make you a personal IR.

Since you have guitars with multiple pickups, we could also compare applying the IR to piezo or microphone.

If I remember well, I think you mainly use a sunrise magnetic pickup that you blend with a DPA internal microphone aiming towards the soundhole using either a pendulum SPS-1 or the Grace audio Felix.

Note that the IR should not work properly with the Sunrise because of the comb filtering effect.

Also note that the EPSI is stereo and can treat two channels independantly. So you can run 2 IRs simultaneously for stereo reconstruction OR 2 IRs for 2 differents pickups.

The IR should work with:
* Piezo UST (Fishman, Baggs, D-tar...)
* Soundboard Transucers (K&K, Dazzo, Amulet, Shertler, Lyric...)
* Microphones (DPA...)

Best Regards,
Cuki