Lets take a look at another commonly used mic from helix, and run a few tests on it to see if we can make it behave like the sm57 in the previous post I wrote about using a split crossover block to mimic the axis rotation of a mic.
Ok so now I want to run the same test with a Royer 121. So lets pull up a 121 in WoS and take a look.
Here we have a Marshall cab and a 121 mic. The green graph is with the mic on center on axis and on the grill cloth. the blue graph is with the mic 33% off center, about 30 degrees off axis and still pressed up to the grill. Now I this case its bass heavy, as is the case with most ribbon mics, and changing the position and axis makes it more so. So maybe we had better go a different direction. I generally like to pull ribbon mics back from the speaker a bit till they sound good, it helps tame that bass. But we already have the option to do that in helix. So lets go ahead and try that split crossover trick we learned from Jason Sadites anyway and maybe we can flip it and use it to artificially brighten up an otherwise dark sounding mic. Full disclosure I already know it works great. So looking closely at the graph we can see that the mic first crosses over at 600hz and last crosses at about 1khz with a small exception at about 1.7khz . So lets start with a crossover setting of 800hz, lets set a low boost at 1.5db and a high cut at 2.5 db. just to see if we can make them match.
Pretty darn close, with the exception of that strange 1.7k anomaly that seems to show up in all the WoS mics I’ve tested so far. ….Anyway. Adding bass isn’t really what I’m looking to do with this mic. Id like to brighten it up a bit, maybe tame a bit of bottom end, but without changing the character of a 121 into something un-recognizable. So lets keep the same crossover point at 800hz, and lets reverse the gain blocks cutting bass and adding treble. Its essentially turning the mic more on axis than 100% and only possible theoretically, or in modelers.
Now we get this, On axis and in center in green, and our theoretical “more” on axis in blue. We have maintained the character of the mic, and just dialed it more for our needs.
Play around with cutting more bass and adding more treble till it suits your needs.
And if you haven’t watched Jason’s video Here’s the link:
A lot has been made of the fact that Helix does not have options for adjusting Mic Axis, and distance. The seemingly simple little studio trick of changing the axis (angle) of an SM57 to “soften” it’s the hard edge, or pulling a Royer 121 ribbon mic back from the grill cloth to help reduce the boominess of the bass… How could this be missing, and is there a way we can simulate this effect in helix?
Well the first, and most obvious method is just use IRs They come with gobs of files that cover all of your bases. Find the one you like and bingo you are all set. Now not everybody likes dealing with IR’s, the sheer volume of files to chose from can be daunting. Plus there’s loading them and keeping them organized. Thankfully Helix updater handles most of this for us now, also there’s the issue of what do you do if you have a cab setup in helix you like but you just need more versatility. Well until recently here I was stuck just messing around with EQ blocks trying to soften the harsh bite of some mics, or tame the bottom end of others. That is until I saw this video from Jason Sadites
The technique uses a split crossover block with a gain block on the upper and lower path. By using those gain blocks you can essentially create a high and low shelving EQ within helix and come up with a very workable hack to simulate moving a mic around a cab.
The graph above is what that EQ looks like. Here in blue I have it set to a 750hz crossover, and a bass boost of 3.9db and a treble cut if 1.3 db. In green its just the reverse. I just picked these numbers at random but they do a pretty decent job of showing how this works. so now how do we apply it? Lets start by looking at a few examples.
Here we have a Marshall cab with an SM57 as measured from 2 notes wall of sound software. The Green graph is with the mic centered on the cone and on axis. The blue graph is with the same mic and cab, just with the mic 33% off of center, and approximately 30 degrees off axis. (The manual is a bit vague but I set that control to 33%) We can clearly see a difference. We have a pretty constant 3db Increase in bass from around 100hz all the way to about 1khz. where the difference starts to decrease. it first crosses over at about 1100hz and crosses back and fourth a few times until it crosses for the last time at around 2100hz. If I was setting up my helix to replicate this difference I would set the crossover at 1600hz I would put a 1.8db cut on the trebles and a 3b boost on the bass, and adjust from there. Lets try it and see what we get.
The pink graph is the result. It is the same SM57 model from wall os sound, set with the control on zero, so on the cone and on axis (ouch). So its essentially the green graph with the split block trick applied to it. lets look closely at only those 2
Here we have the Off axis 57 in Pink and the on axis plus the split crossover in yellow. its really pretty close, It just looks like were about 2 db high in the 1-2k range, so lets adjust our eq at the end of our patch accordingly.
Here it is with a parametric eq with -3db cut at 1.2khz and a q of 10. I really don’t think I can get it any closer than that using this method, but honestly its a fantastic workaround that’s already part of helix.
Big thanks to Jason Sadites for showing us how to do this.
I’ll be following up shortly with more posts on this subject including more mics, and how to create IR’s to do the same thing.
Ok so last evening I posted up this video of the Minotaur model in the line 6 helix. Today I want to do a little write up to go along with it.
So I figured I’d go online and find a schematic and start going thru it and step by step, really basically as that is the extent of my ability with reading schematics. …anyway I found that the Klon is a pretty darn complex circuit. I was expecting your typical dual op amp with soft clipping diodes in the feedback loop …easy peasy. Boy was I wrong. Well the good news Is I found this. which is a superb article explaining the Klon way better than I ever could, the comments on this article are outstanding too… so If you want to get your Inner Geek on Check it out.
I was messing around on FB this weekend. Just shooting the crap with people about my Mic deletion file. Just a quick aside, it doesn’t make much difference in sound because a Beyerdynamic M160 has a really flat response curve, which is the point of a good mic. At 4cm it has a maximum 6db variance at 70hz, and by the time you get to 300hz its under 2db variance where it remains all the way out to 15khz… It’s a good mic, A really good mic. The point of the entire microphone exercise was, firstly to see if I could do it, and second to have the ability to measure things like cabs and Mics in helix as accurately as possible. Why …because I want to, and I think its fun.
Newsflash.. I’m weird like that.
Anyway back to the point today’s post. Someone asked me If I could create an IR to basically convert single coils into humbuckers and vice versa. Now I wasn’t actually sure if it would work, how it would sound etc. but apparently this has been done, and it was pointed out in the post. Several of the IR companies have them available for purchase. Now I’m a guy who likes to see if I can make it better, smarter, faster, and more importantly cheaper. I make my own IR.s all the time, and on principle, not price, I hate paying for them. Here’s why
Go ahead hit play. You don’t even have to turn your phone volume down, the guy on the train sitting next to you won’t even notice. Hell you probably won’t notice, It’s a damn click, nothing more. Contains loads of data for sure, but its easy to make. I made a video on how to do it. which you can find here:
Perfect… no . Effective … mostly for my purposes.
Anyway I digress. I was game to take on the challenge of making this IR How hard can it be right? Its been done before tho, so I want to do it better, but how 🤔
Well the first thing that popped into my mind was all kinds of concerns about getting consistent level outputs from the pickups. I’ve dealt with this problem before when I was working on a project measuring the outputs of different strings to see what, if any differences there were in materials and construction methods. It was a huge pain in the ass. I literally had to build a machine to consistently pluck the strings exactly the same every time, in order to get consistent data, and that machine is long gone.
So the second thing I went for was a rig I made a few weeks ago to measure the differences that the impedance settings made in helix.
It basically uses 2 single coil pickups, sends a sweep tone into one, and the signal gets inducted into the other. Its a crude version of a transformer. I’ll just make a second one with humbuckers, and I’m all set! ….heh when will I learn it’s never that simple.
Ok so I have the 2 test units, I plug em into a loop in helix one at a time, and run a sweep thru em with RTA.
Ok… So thats encouraging. I see significant differences in the data. So I’m measuring something. Lets sweep these things and stick em in match eq, create a curve for each conversion H>S and S>H and then sweep those curves into IR’s and were good. I’ve made a video on how to do this here:
So I went and created this project in logic with all kinds of complex signal routings.
No . they’re really not that complex, I’m just really bad at doing this and it takes me a while. I actually made a video on the signal routing mainly so I don’t forget how i did it . ….But anyway. The IR’s sucked. One of em didn’t even produce any sound. I’m not really sure where they went wrong. It could be because i lined up the pickups with the magnetic poles attracting each other, It could be because its a faulty process, It could be …hell I have no idea what it could be. I just know it didn’t work. It was a colossal fail, resulting in dozens of unemployed aliens. 😂
Ok Eric… Exhale. Go to the grocery store, do the food shopping. Get beer cause you’re apparently going to need it. Regroup and try again. Ok open beer, take a swig, think…
Stop worrying about what went wrong. Lets start over. First thought… Does it really matter if the pickups are stimulated in the exact same way? Does this really require doing a sweep analysis? Maybe not. The match eq program seems to map some type of variant of peak output, average maybe? I have no idea, but maybe if i just play my guitar into each side of the match program that’ll work. (seems too easy right, I think so too) . So lets see what we get.
The 2 lower graphs are the second pair if IR’s I created. They are basically the difference of Humbucker curve minus single curve and single minus Humbucker, The upper two are the differences of the first two.
I sweep em into IRs and presto. I have a pretty darn believable Single to Humbucker IR, and a passable humbucker to single IR, especially if you put a compressor in front of the Humbucker IR to tame the spikey Highs a bit.
Today’s lesson… Sometimes the obvious (easy) way is the best way.
So I’ve been struggling for a while trying yo figure out the best way to accurately “delete” a mic in Helix, so I could accurately measure the eq curves the speaker cabinets in Helix, without having them “colored” by the microphone. Basically what I was looking to do was create a calibration file for a microphone, and then use it to measure speaker cabinets in helix. Once I had that calibration dialed in I could Use it to erase mics, or cabs or whatever I needed to In order to get the measurements I was after. The ticket was being able to eliminate one or other a mic or a cab.
I’ve been back and forth with this for a while. I started with speaker curves. I figured what the heck. I’ll just make an upside down version of a speaker response chart using a bunch of stacked EQ’s in helix, convert it to an IR and bingo. I’ll erase the speaker, be left with the mic curve, convert it to a house eq file, and go measure the speaker curves in helix. Here’s the problem…
This is a typical Speaker frequency response plot. I just figured even if it was smoothed out a bit, it would come out ok.. Right? Nope. The problem with trying to recreate a speaker curve and using it to reveal a Mic curve is 1 you can’t get the required resolution by using the availabe eq’s in helix, and 2 any errors you do make due to smoothing etc will probably be larger than the entire dynamic range of the microphone you are trying to measure. So needless to say then, using this erroneous curve to make determinations about other things …well lets just say it was a colossal fail and leave it at that. I gave up on it for quite a while.
My next attempt came closer. Lets try and recreate a Microphone eq curve, turn it upside down and Bingo! We can erase that sucker and reveal the speaker cabinet eq curves I’m after. So i quickly jumped onto google grabbed a SM57 microphone eq curve and dutifully recreated it using eq blocks, turned it upside down added it to a stock cab block set the mic to match and… It sucked too. Here’s the SM57 eq curve.
Problem 1.. It aint exactly Hi res. Problem 2 Its lacking critical info like how far from the speaker it is during this test. Oh hell its an SM 57 these are always pressed against the grill cloth right? Just set the helix mic distance as low as it goes and you should be good to go right? …Nope. More crappy data. I was bummed, and I gave up on this for a while. Until one day I was starting to collect EQ curve plots for my blog cause I’m planning on at least listing the mics in helix, even if I cant accurately measure them. For the sake of completeness I figured what the heck I’ll start collecting all the factory spec sheets and throw together a post. (i haven’t posted it yet btw) Thats when I found this.. It may as well be the rosetta stone as far as I’m concerned.
You really do have to admire german engineers for their thoroughness sometimes. I suddenly had the data necessary to proceed. To be honest I had seen this graph before, But I didn’t realize how it would assist me. I initially dismissed it as useless, Not knowing which bass response graph to choose. But when I took this second look at it i saw the Mic distance information, and decided to take another crack at it. I carefully reproduced the curve using the high and low shelving EQ and a couple parametric EQs from helix. and I came up with this.
Next step flip it upside down.
And lastly…. Convert it to an IR, for convenience sake.
Now to test it. But how do I do that? lets delete some cabs and look at the mic curves and see what they look like as compared to the factory spec sheets.
Ok so here in blue is a frequency sweep of a 1×12 Us deluxe followed by my IR. The data is then “calibrated” by the software and a new eq curve is created called a “house file” This file when Summed with the Input data generates a flat line shown in green. Every measurement I take now is subjected to that House file and thus anything I change is revealed in detail. Notice the deviations below 40hz and above 16khz. My test equipment has a really hard time measuring extremely quiet signals at the limits of its range.
Anyway… In my travels Looking for Mic data I also came across this beauty (once again thank you German Engineers) Unfortunately the resolution of this document doesn’t come thru in this screenshot, but Once again its a super detailed frequency response curve of a Mic in helix.
so lets switch Helix to this mic and see what we get…. Should be close I hope.
Ok well that was a bit unexpected, But it shouldn’t have been remember my test equipment is trash at measuring really low signals below 40HZ and above 15k or so. The good news is so are guitar speakers, and we generally use low and High cuts in these areas anyway. so lets try that and see what we get.
Remember its supposed to look like the U87 graph from up there 👆… and it does. First test looks ok. Lets check it against another mic I have a decent chart for.
Ok Sennheiser 421 more detailed graph.. Lets Switch helix to Dynamic 421 and see what we get. …Fingers crossed.
Ok time for one last test. Our old friend the SM 57 remember this guy
The first distortion block in helix is the Exotic Effects EP booster
It is a recreation of the gain circuit of Echoplex delay units from the 1960’s. Many famous musicians from that era used echoplexes, most notably Jimi Hendrix. Many of them found that when the unit was off their guitar sounded different, and they liked their sound better with it on. So even when they weren’t using any delay they tended to leave the gain on to get that sound. Its probably the earliest use of boost on guitar amps. Exotic has recreated this gain circuit in this boost pedal. As it was a level control to adjust the level of the output going in to an amp, or a board, or wherever they were sending the signal, the designers were looking to create a clean boost with a flat response. They achieved that (mostly) with only a slight roll off in the very low frequencies, and a gentle roll off above 1khz. Here’s the output of the Line 6 Helix model called the Kinky Boost.
As we can see we have about an 11 db boost at Helix’ factory settings a roll off of the subs below 20hz, and a gentle roll of of the highs above approximately 1khz.
Exotic gave us 3 options to control this pedal. a level knob and 2 switches labeled boost and Bright
The following graph is the Kinky with the level set to 5 and…
Purple > both switches off
Yellow > boost switch on
Blue > Boost and Bright on
This graph has the Kinky with the level set to zero, and the same combinations of switches.
Turquoise > Switches off
Green > Boost on
Blue . > both on
This graph has the same combo of switches with the level dimed.
The Kinky boost is for the most part unremarkable in its EQ curve. It exhibits no variance due to the position of the level control. and consistently rolls off a little of the lows and highs however you have it set.
It is also unremarkable in its waveform output. it exhibits Very little if any distortion, and relies on hitting the front of the next gain stage with a bigger ( albeit slightly eq’d ) input signal to do what it does. Here’s the waveform graphs. They exhibited no changes in waveform shape no matter how the controls were set, Indicating that the pedal has plenty of clean headroom.
Kinky at 200hz .
Kinky at 500hz
Kinky at 1khz
…And kinky at 2.5khz
As we can see the waveforms remain unchanged.
And if you’d prefer here’s a quick video run through.
In this Video I go through my tone matching process to create a Boss BD2 patch using only the blocks available to us in the line 6 helix.
This video has been in the works for a few weeks. I’m sorry its so long, but there is a lot of ground to cover, especially since this is the first time I’m going this in depth with this stuff.
If you are interested in trying this patch out for yourself use the link below to download this BD2 patch. It is the blocks that make up the pedal only. you will have to create your own patch around this,
To download the patch you heard me playing in the video use the link below. It has the BD2, a reverb, two Elephant mans in parallel, with the lix control on EXP2, and a Jason Sadites inspired Mastering section at the end.
“Impedance is witchcraft”
a phrase I’ve repeated often.
Ok so for a while I’ve struggled with getting accurate measurements of this with my spectrum analyzer. The output is a buffered signal, and every time I tried to measure the differences in the impedance settings I just came up with this.
Small differences in overall output volume, and a slight roll off in the really low frequencies. I know there has to be more going on here, But how do I measure it? Helix wants to see a pickup here, and my spectrum analyzer only puts out a buffered signal. Well I got an idea from watching a video interview of Tor from TC Electronic. He was saying that the way that toneprint actually works is that the pickup in your guitar actually picks up the magnetic field fluctuations from the magnets in your phones speaker. This got me to thinking of how I could replicate that. The best I could come up with was this. Its two pickups taped together. The first one is connected to the output of my spectrum analyzer, and functions as the speaker coil, and the other picks up the signal and goes into the guitar input of helix.
I can now switch the input impedance in Helix and see changes. While I realize I can’t accurately determine exactly how many dBu difference there is in each signal because this is far from a calibrated setup. I can measure the differences which is really what I’m after. So after testing this here is what I came up with, and my ears back up the data. Which is about as scientific as I can get with this experiment.
The solid blue line at 5dBu is the test tone, and ignore the small spike at 60hz as I know this to be an anomaly within my test equipment.
This is much different data than we saw with the buffered signal. The purple line at the bottom is the 10k Ohm setting and the yellow one at the top is the 1M ohm setting. Auto setting just changes the impedance based on what the first block in your patch wants to see. For example a fuzz face wants 10k Ohms, so helix switches it to that when you have a fuzz first in your chain. It is important to note that it doesn’t matter if the fuzz is on or off. Helix still switches the impedance.
On a more interesting note in addition to the 6dBu …ish average (it changes depending on frequency) difference, we can also see a pronounced high frequency roll off centered at around 2khz as we decrease impedance from 1M Ohm down to 10K Ohm.
Im just glad I was finally able to measure this and it verifies what my ears have been telling me.