DIYstompboxes.com

DIY Stompboxes => Building your own stompbox => Topic started by: Vivek on March 31, 2021, 09:03:53 AM

Title: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: Vivek on March 31, 2021, 09:03:53 AM
Is there enough flexibility and power in the FV-1 to do a digital simulation of a Rockman X100, including

Compressor (with specific input output curve)
2 stage Distortion
Cab sim
Chorus
Reverb
Mixer

With switching of above internal parts to get the 4 Rockman sounds ?

I currently have zero experience on FV-1 and was wondering if the MIPS and Instruction space limits are sufficient

Option:
Or is it better to consider a hybrid solution with Cab sim, chorus and delay on the FV-1 and rest in Analog ? Would FV-1 be a better choice than BBD and Belton approach for the Chorus and Reverb application ?


Your thoughts, please !
Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: vigilante397 on March 31, 2021, 09:37:15 AM
I'm not an FV-1 expert (we do have a handful around here, hopefully they'll jump in as well), but you'll probably be wanting a hybrid approach to cover all of those. FV-1 is excellent for modulation effects and would do great for chorus and reverb, but it's awful (in my opinion) for things like drive. I would use the FV-1 for the chorus and reverb sections then throw in your favorite analog distortions, analog cab sim, and analog compressor.

I've started using FV-1 instead of BBD for chorus and delay because there are a huge range of sounds you can get out of FV-1 for either, and it takes much less circuitry (and biasing) than BBD, and provides a lot more flexibility and options in the reverb department than Belton bricks.
Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: Vivek on March 31, 2021, 10:06:14 AM
I've started using FV-1 instead of BBD for chorus and delay because there are a huge range of sounds you can get out of FV-1 for either, and it takes much less circuitry (and biasing) than BBD, and provides a lot more flexibility and options in the reverb department than Belton bricks.

Did you write your own programs ?
Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: vigilante397 on March 31, 2021, 11:26:55 AM
Did you write your own programs ?

I've done a couple originals on SpinCAD (Digital Larry, SpinCAD's creator, is on here and will hopefully drop in with some thoughts) but I use a lot of patches people have written and posted online. There are TONS out there for FV-1.
Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: Digital Larry on March 31, 2021, 02:03:48 PM
Compressor (with specific input output curve)
- probably, several compressor/limiter algorithms are freely available. 

2 stage Distortion
- there's really only one useful overdrive algorithm, the "T/X" which cleverly creates soft clipping.  I mean maybe you could do something else but it's beyond me.

Cab sim
- Using a single 2-pole resonant low pass?  Sure.  Implemented using "impulse response"?  Not practical.

Chorus
- yes

Reverb
- equivalent to the MN3011 multi-tap BBD used (I think) in the Rockman?  Yes

Mixer
- no problem

Can you get all of those at once?  Probably.
Can you control them the way you want with only 3 knobs and 8 settings using 3 switches?  You tell me.
Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: Mark Hammer on March 31, 2021, 03:09:58 PM
The EHX Holy Stain was probably one of the first FV-1 based commercial pedals, and it will do several things at once. Robert Keeley's various "workstation" pedals are also FV-1 based.
Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: Ice-9 on April 01, 2021, 11:19:32 AM
The EHX Holy Stain was probably one of the first FV-1 based commercial pedals, and it will do several things at once. Robert Keeley's various "workstation" pedals are also FV-1 based.
Mark, The Holy Stain was a nice pedal but it didn't really do several things at once, it used 4 of the internal FV-1 programs but only 1 at a time and it also only used two of the control pots. It did have a fuzz built in as well but that was not part of the FV-1 program it was a separate part of the pedal circuit.

I started a thread on modding the Holy Stain a long time ago and modified the pedal to add the 3rd control pot and to access all the 8 internal programs. It was also possible to modify it to use an external EEPROM which made the pedal a great tool for FV-1 development at the time.
Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: Mark Hammer on April 01, 2021, 01:35:02 PM
Thanks for that correction, Mick.  Much appreciated.  I knew there was a reason to keep my eyes peeled for a cheap one on the 2nd-hand sites, but had plum forgotten the reason.  If I do find one, I'll be sure to contact you.

But I do know that Robert Keeley was using the FV-1 for his various multi-function pedals.  That said, I've never tried one, so I don't know if, like the Holy Stain, they did a buncha tricks but only one at a time.
Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: Ice-9 on April 01, 2021, 03:27:15 PM
Hi Mark,

Yes indeed Robert Keeley has made quite a few FV-1 based pedals and I had 1 in for repair a while back but it was not the the multifunction one so I don't know what they are capable of. The one I had was a Hooke Reverb.
Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: ElectricDruid on April 02, 2021, 07:23:07 AM
The controls on the Rockman are pretty minimal:

(https://davesguitar.com/wp-content/uploads/2016/07/Rockman-2.jpg)

If you were doing it as an FV-1 based pedal, you could use the switches to switch between programs. So the code doesn't have to be able to do *everything* at once, only *all the things needed for this particular sound*. That helps a bit.

Most of the things on the list are simple for the FV-1. The chorus/reverb is easy. The basic cab sim filters are easy. Mixer is easy.
The FV-1 will do compression, but setting it up to get a specific curve might be hard. You'd probably end up tuning it by ear to try and get it close. The distortion is the place where the FV-1 will struggle the most, and it might be worth doing that stage with the original Rockman circuit. Note that the FV-1 is stereo, so you could have a analog distortion stage that you can fit in "between" any FV-1 based stages by going:

Input -> FV-1 Left input ->FV-1 Left output->Distortion Input->Distortion Output->FV-1 Right Input->FV-1 Right output->Output

That enables the firmware to decide what stages go either in front or after the distortion, and it could be different between different programs.
Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: amz-fx on April 02, 2021, 07:46:38 AM
But I do know that Robert Keeley was using the FV-1 for his various multi-function pedals.  That said, I've never tried one, so I don't know if, like the Holy Stain, they did a buncha tricks but only one at a time.

I contributed some bits to the early Keeley FV-1 pedals and they are a basic design with an analog path from the input buffer to the output mixer, and with the digital signal from the FV-1 added into the output mixer via the Blend control. The early versions for which I have the schematics did not allow the direct signal  to be faded out of the output, but I believe that may have changed in later revisions (but I'm not 100 percent sure about that).

Any of the Keeley FV-1 reverb designs would be a good platform for testing merely by changing out the 24L32 memory chip where the programs are stored.

Note that his digital pedals are being moved to a different platform and the FV-1 designs will probably be phased out in a few years. The AKM fire is likely delaying that changeover.

regards, Jack
Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: Mark Hammer on April 02, 2021, 01:26:55 PM
Note that his digital pedals are being moved to a different platform and the FV-1 designs will probably be phased out in a few years. The AKM fire is likely delaying that changeover.

regards, Jack
Yes.  When we met at NAMM in 2018, he mentioned that he was pretty much done with the FV-1 and would be moving toa different platform, though I never learned which one.

Best to you and the family for the holidays, Jack.
Mark
Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: Eb7+9 on April 02, 2021, 02:29:41 PM

Your thoughts, please !


not sure if you're referring to the cab sim as the 6-stage piece-wise linear filter section ... to me, the heart of the Scholz sound

https://patentimages.storage.googleapis.com/e2/54/b1/3a67892edaac53/US4584700.pdf (https://patentimages.storage.googleapis.com/e2/54/b1/3a67892edaac53/US4584700.pdf)

that in itself would take up quite a bit of DSP power ... but the advantage there would be lower overall noise/hiss ... the question becomes this, would an analog distortion signal bode well with a digital filter following it ...

might not be a bad thing to investigate

with any luck you would have succeeded in making the design cleaner in a sense, because lots of the noise in that design comes from those stages

in any event the front end limiter/distortion should be analog imo ... and the chorus/echo and gate ... well, that's another conundrum for you to figure out

I would leave all that out and make use of better DSP or whatnot instead ... in particular, expect to run into problems if you try building the gate as per his patent info ...

to recap, analog comp/distortion front-end followed by digital implementation of 6-stage Scholz tone shaping ... followed externally by higher quality chorus/verb/echo (gated or not) ... would potentially give you an X100 with improved specs
Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: Ice-9 on April 02, 2021, 06:22:20 PM

Note that his digital pedals are being moved to a different platform and the FV-1 designs will probably be phased out in a few years. The AKM fire is likely delaying that changeover.

regards, Jack
Yes the AKM fire and the EOL of the AK4556VT codec IC have a major effect on selecting a replacement, there are alternatives to this which I am looking into but at the cost per unit around X2 -X5 for similar quality parts. I have just been testing different I2S codec IC's tonight. There is also the complication of single ended in/out or balanced in/out which complicates the analogue circuitry. Along with the AKM factory fire and the worldwide shortage of chips it is a little bit of a wait and see thing at the moment for future availability.
Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: ElectricDruid on April 02, 2021, 06:53:00 PM
BTW, I think the the title of this topic is a great question. Very direct and to the point, and at just the right level to be something that is open to discussion. Something more complicated, no way. Something simpler, easy. This thing is at just that point where it's not so easy to say for sure and leads to an interesting discussion about pros and cons.

Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: amz-fx on April 03, 2021, 03:39:41 AM
Best to you and the family for the holidays, Jack.
Mark

And the same to you and yours. Stay safe.

BTW, I think the the title of this topic is a great question. Very direct and to the point, and at just the right level to be something that is open to discussion. Something more complicated, no way. Something simpler, easy. This thing is at just that point where it's not so easy to say for sure and leads to an interesting discussion about pros and cons.

I see a couple of limitations that might come into play. The Spin chip only has 128 instructions per fx patch. The reverb, in particular, will eat up a lot of the available code space. In addition, the memory has to be shared among all the fx that need it, and the chorus will need a chunk, then the reverb will eat most of the rest. A simple reverb will be about the best you can do, but fortunately, it will be in character for the Rockman, which did not have complicated reverb hardware.

I think the distortion and its related tone shaping are best done analog style with the dsp handling the rest. I like your idea of using the two available signal paths to manage the fx. I like this version with the FV providing the compression:

FV-1 Left input ->Compressor->FV-1 Left output->Analog Distortion/Tone->FV-1 Right Input->chorus->reverb->FV-1 Right output

This simplifies the compressor by eliminating the need to match/find jfets to produce the correct compression response. Also, analog tone shaping will free some code space in the FV-1.

regards, Jack
Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: Digital Larry on April 03, 2021, 05:42:51 AM
then the reverb will eat most of the rest. A simple reverb will be about the best you can do, but fortunately, it will be in character for the Rockman, which did not have complicated reverb hardware.
The Rockman reverb is just an MN3011.  Multi-tap delay is quite efficient in the FV-1.  I just slapped the MN3011 block into you-know-what and the code is below.  You don't even have to allocate all the RAM.  Now that I look at it though I realize that the MN3011 block I designed is a little silly in that I mix all taps to a single output.  All this time and nobody ever said anything about it.  That is easy enough to change even manually.

I've used the MN3011 block as a reverb in more complex patches.  Unless you are playing really isolated guitar notes, to my ear, you can get away with a really simple implementation in many cases.

Code: [Select]
;
; ----------------------------
;------ Input
;------ Feedback Output
;------ MN3011
RDAX REG0,0.5000000000
RDAX ADCL,0.5000000000
WRA 0,0.0
RDA 7839,1.0
WRAX REG1,0.0000000000
RDA 7839,0.28183829312644537
RDA 932,0.5
RDA 1559,0.31622776601683794
RDA 2812,0.3981071705534972
RDA 4065,0.28183829312644537
RDA 6571,0.22387211385683395
WRAX REG2,0.0000000000
;------ FB In 1
RDAX REG1,1.0000000000
WRAX REG0,0.0000000000
;------ Output
RDAX REG2,1.0000000000
WRAX DACL,0.0000000000
Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: ElectricDruid on April 03, 2021, 06:43:48 AM
The Rockman reverb is just an MN3011.

Didn't I see an extra delay stuck after the MN3011 in one of the block diagrams? It stuck me as a bit weird because adding a pre-delay *before* the reverb would make more acoustic sense, but it definitely looked like it was after.

Also, what's the clock rate for the MN3011 in the Rockman? How much delay does it actually generate?

Either way, I suspect the FV-1's two seconds is more than enough. The Rockman Reverb is going to be a few hundred milliseconds at best, and the chorus adds another 20-30msecs or so. We've probably got a second and a half of RAM left over.
Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: amz-fx on April 03, 2021, 10:11:58 AM
I believe that the FV-1 is one second of memory in the standard configuration, though there are tricks you can use to double it, but with corresponding loss of bandwidth. It should easily be enough to duplicate the Rockman delays as Larry noted.

There are several Scholz patents listed in my fx patent collection at http://www.muzique.com/misc/patents.htm

The extra MN3007 delay chip is used to extend the delay time of the right channel (only). It is acting as another tap on the MN3011 and enhances the stereo effect.

The MN3011 has 3328 delay stages, so if you clock it at 10kHz, you would get 166ms maximum delay time. The upper band limit then has to be less than 5kHz (as I am sure you know) but because of the heavy filtering, this is probably not a problem. With a 10k clock the MN3007 is adding an extra 51ms delay to the right channel.

Best regards, Jack
Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: Digital Larry on April 03, 2021, 10:31:29 AM
Didn't I see an extra delay stuck after the MN3011 in one of the block diagrams? It stuck me as a bit weird because adding a pre-delay *before* the reverb would make more acoustic sense, but it definitely looked like it was after.

Also, what's the clock rate for the MN3011 in the Rockman? How much delay does it actually generate?
You're right... though that extra delay is just a few more instructions.  From the text it's not described as pre-delay and it has its own feedback path.  Interesting approach to creating a longer tail I'm guessing.  I'm also somewhat surprised by the pre-MN3011 gating to trim transients.  Just my take on it but that seems more likely to have been to prevent distortion rather than some aesthetic need to remove transients.  IMO reverb is all about transients.  I mean Dick Dale anyone?

edit: I didn't even see that it was only on the right channel.

FV-1 only supports 1 second of delay without sub sampling, but still I think that's adequate for this application.

The complex filter could most likely be described as a cascade of first and second order blocks.  FV-1 does 1st order easily in 2 instructions.  2nd order (implying a resonant structure) take more like 6 or 8.  And there are more filters sprinkled throughout the design.

When you push the limits of what can be done with an FV-1 you may have to start dropping things off.  To me that is a fascinating part of such an exercise because it's not immediately obvious what you can get away with.  Much of the consulting I did for one firm used this approach.  Every program was packed to the gills and I was always on the lookout for some way to save 1 or 2 instructions.
Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: Ice-9 on April 03, 2021, 02:00:31 PM
I do think all this is possible within the 128 instructions but the problem would still be only 3 control pots. I would say it is definitely worth going over to the FX-Core DSP for the Rockman idea.
Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: marcelomd on April 03, 2021, 02:45:06 PM
Regarding memory/instruction limitations. Is it possible to use multiple FV1s? Just set them up as multiple pedals.
Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: Ice-9 on April 03, 2021, 03:05:53 PM
Regarding memory/instruction limitations. Is it possible to use multiple FV1s? Just set them up as multiple pedals.

yes it is.
Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: ElectricDruid on April 03, 2021, 08:31:29 PM
I do think all this is possible within the 128 instructions but the problem would still be only 3 control pots. I would say it is definitely worth going over to the FX-Core DSP for the Rockman idea.

Why? The Rockman uses a grand total of no pots at all, so it's not a big issue for a proper clone. Ok, there's a lot of knobs you *could* put on it, but then it's not a Rockman any more.
Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: Vivek on April 04, 2021, 02:10:26 PM
Since the FV-1 does 128 instructions on each Audio sample

I started to guesstimate the size of each part


FET Compressor (with specific input output curve) :
??? steps

2 stage Distortion : A saturated Opamp feeding Hard clip LEDs
About 30 instructions

Cab Sim "Complex Filter" : 3 opamps as filters
About 15 instructions

Chorus : 555 LFO at 1.7Hz modulating VCO at 50 Khz thats clocking a BBD
LFO = about 6 instructions
delay ????

Reverb : MN3011
One of the sample progs "rom_rev1.spn" at http://www.spinsemi.com/programs.php had around 80 instructions for the reverb. Is that about right ?

Mixer : to generate a L and an R from the dry and wet signals
2 or 4 instructions

Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: Vivek on April 04, 2021, 02:16:26 PM

The complex filter could most likely be described as a cascade of first and second order blocks.  FV-1 does 1st order easily in 2 instructions.  2nd order (implying a resonant structure) take more like 6 or 8.  And there are more filters sprinkled throughout the design.



yes "Complex Filter" or Cab Sim is 3 stages of filter blocks, they are resonant at about 400 Hz and 3K

There are many other filters to restrict the high frequencies before and after the chorus BBD, and also before and after the reverb MN

Will these be needed on a FV-1 ?


ie is the need to cut High frequencies related to deficiencies of BBD chips, or are they universal truths related to how delays work ?
Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: ElectricDruid on April 04, 2021, 04:07:27 PM
There are many other filters to restrict the high frequencies before and after the chorus BBD, and also before and after the reverb MN

Will these be needed on a FV-1 ?
Yes, and certainly Yes if you want it to sound the same.

Quote
ie is the need to cut High frequencies related to deficiencies of BBD chips, or are they universal truths related to how delays work ?

It's a universal truth related to how sampling works. BBDs are one (rather unusual since it's analog) example of sampling technology, and the FV-1 is another. In short, if you put any frequencies in that are above half your sample rate (that's your clock rate for a BBD) then the frequency that comes out won't be the one that went in. This is "aliasing".
Once you have the signal safely inside the FV-1, you don't necessarily need to add more filtering unless you're about to do something that would add harmonics (like a distortion stage), but in this case, you'd probably add the filtering to try and get the most authentic sound.


Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: ElectricDruid on April 04, 2021, 04:12:10 PM
Reverb : MN3011
One of the sample progs "rom_rev1.spn" at http://www.spinsemi.com/programs.php had around 80 instructions for the reverb. Is that about right ?

I'd say that's probably an over-estimate, because the Spin Semi reverbs are much more complex reverb algorithms with various stages of allpass filters and multiple delays.

The Rockman reverb is one MN3011 chip followed by a MN3007 chip, so it's just a multi-tap delay with a couple of mixers to mix the tap outputs and a bit of feedback. Not terribly complicated.
Of course, if you've got the code space, you could always start making improvements to the Rockman design and use a better reverb algorithm like some of the Spin ones. There's no reason your Rockman clone has to be totally authentic if you don't want it to be.


Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: Digital Larry on April 06, 2021, 03:54:04 PM
I already posted the code for an MN3011 block I made, it's about 10 instructions.  Adding another delay tap is just one more instruction to that.  If you try this on your FV-1 be aware that the feedback is turned ALL THE WAY UP UP UPUPUPUPUUUPPP ahhhhhhhhh!

Someone else seems to be interested in that awesome sound as well: https://line6.com/support/topic/58178-free-bostonrockman-irs-zip-updated-nov-19/

DL
Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: Cybercow on November 20, 2021, 11:28:14 AM

The Rockman reverb is just an MN3011. 


Actually, looking at a completely traced and verified X100 schematic, the MN3011 is directly associated with an MN3007 and are joined thru a stereo filtering network and tied to the same clock chip. So rendering an exact match to the X100 reverb/echo section is a bit more complex than just an MN3011 emulator.
Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: Vivek on November 21, 2021, 11:51:33 AM
In the FV-1 world

Adding MN3007 functionality to a MN3011 model could be 2 extra lines of code and change in one constant value.

"Stereo summing" could be 2 extra lines of code

And "same clock" doesn't need any code
Title: Re: Does FV-1 have what it takes to do a Rockman X100 simulation ?
Post by: ElectricDruid on November 21, 2021, 06:50:27 PM
+1 what Vivek said.

The 3011+3007+filters+mixers really isn't the hard part at all. It's pretty trivial on the FV-1, since that sort of stuff was absolutely what the chip was designed for.