Helix 3.5 New Cab Engine is Awesome! Fractal needs to catch up!

[dean701]

Haha those bloody sliders are so sensitive that even a minor alteration changes the tone quite a bit.

I haven't done the Helix 3.5 update yet so can't say how good it is but there is a general concensus that overall sound quality is improved.
I'm going to go the other way on the updated stock cabs though because in my opinion once you're settled on your favourite IR's it's better to stick to them regardless of cab updates. Leon Todd did a good video on the update and summed it up the same way.

Editor interface is a different discussion but to me, I would always prioritise sound quality over anything else.

regarding the Helix sliders, just use the center mouse wheel. Hover over the field and the center mouse wheel alters the values by each gradient. No clicking needed.

Leon's video was good but no one IR fits all amps so you'll still need to search for something from amp to amp. Plus what if there is a better sound just waiting to be found. lol.

 

[FractalAudio]

The actual data is pretty "coarse". There are actually only a handful of IRs for each cab/mic combo and the IRs are linearly interpolated for the in-between positions. I didn't do any exhaustive tests but it appears that there are only three to five actual IRs in the x-axis. These IRs are then linearly interpolated to obtain the resulting IR.

Same goes for the distance control. Only several IRs and the in-between positions are obtained via linear interpolation.

Of course this is not surprising as the number of IRs adds up quickly and hardware modelers are inherently limited in the amount of non-volatile storage they possess.

 

[toneseeker911]

FWIW we've been working on something like this for the past six months or so. We have several robots that we use for automated IR capture, custom control software, etc., etc.

I didn't want to spill the beans this early but...

Your patience will be rewarded.

Holy cow !!! Yes. My biggest gripe has been navigating IRs, when they contribute 50% or more of the sound.
Ordering a 100 pack of F5 keys.

 

[axl-fx]

Great to hear they are working on something!

Fractal has the sound dialed but the interface could be streamlined. Make it simple so even someone like Dave Grohl can use it. The irs housed similar to what QC does would be the right step.

 

[dr bonkers]

The words Helix and accuracy with regards to modeling do not belong in the same sentence.

I found the sound and feel of the Zoom G3Nx more accurate than the Helix.

I'm still waiting for the Helix to add a Variac. It's only been on IdeaScale for 5 plus years.

 

[yyz67]

[T]he IRs are linearly interpolated for the in-between positions.

Only several IRs and the in-between positions are obtained via linear interpolation.

FWIW we've been working on something like this for the past six months or so. We have several robots that we use for automated IR capture, custom control software, etc., etc.

I didn't want to spill the beans this early but...

In some physical processes, linear interpolation (or any kind of appropriately fitted monotonic interpolation) can be a "reasonable" or good approximation.

Curious if scanning/mapping the FR vs. mic location of a speaker+cab you might be able to extract some kind of multivariate parametrization (e.g. PCA) that's better than simple interpolation between a few measured points. But please don't "spill the beans"...

 

[Sixstring]

The actual data is pretty "coarse". There are actually only a handful of IRs for each cab/mic combo and the IRs are linearly interpolated for the in-between positions. I didn't do any exhaustive tests but it appears that there are only three to five actual IRs in the x-axis. These IRs are then linearly interpolated to obtain the resulting IR.

Same goes for the distance control. Only several IRs and the in-between positions are obtained via linear interpolation.

Of course this is not surprising as the number of IRs adds up quickly and hardware modelers are inherently limited in the amount of non-volatile storage they possess.

At the risk of sounding stupid does this mean that the IR length would have to be shorter (standard rez IR) in order to accomplish the task with existing hardware?

 

[dean701]

The words Helix and accuracy with regards to modeling do not belong in the same sentence.

I found the sound and feel of the Zoom G3Nx more accurate than the Helix.

I'm still waiting for the Helix to add a Variac. It's only been on IdeaScale for 5 plus years.

There's no reason to bash other modelers. Countless Sonic Drive Studio videos have shown that most modelers including the Helix can be made to sound very close to the real amps. Competition breeds excellence.

As far as this thread goes, some of the more recent modelers/plugs have a speaker GUI that makes the choice quicker. Even if Fractal used a sample of the existing included IRs in the Axefx and made a GUi out of it, it would be positive ground from a streamlining perspective. Then of course leave the custom IR option open to use your own.

 

[DLC86]

There's no reason to bash other modelers. Countless Sonic Drive Studio videos have shown that most modelers including the Helix can be made to sound very close to the real amps.

That's the point, it can be made to sound close, but that happens after a good amount of tweaking, and that doesn't look like accuracy to me.
In the axe fx/fm9, to replicate the sound and feel of the few real amps I have experience with, I just select the model, set authentic controls roughly in the same position, select the closest impedance curve and it sounds for the most part identical.
Doing the same with the helix sounds nowhere near close to real amps and I need to set the knobs in very different positions, put additional EQs and compressors, tweak the (not enough available) advanced parameters and then I'll just get in the ballpark. Then I'll go on tweaking to try to make it closer.

So yeah, I think @dr bonkers is right. This is based on first hand experience and not youtube videos, ymmv.

PS: this amount of tweaking was once necessary on the axe fx too, but Cliff's work during these years has brought his models closer and closer to the real deal, something that hasn't happened on the helix. I think in terms of accuracy they're stuck at the level the axe fx II was in its first few firmware revisions.

 

[dean701]

That's the point, it can be made to sound close, but that happens after a good amount of tweaking, and that doesn't look like accuracy to me.
In the axe fx/fm9, to replicate the sound and feel of the few real amps I have experience with, I just select the model, set authentic controls roughly in the same position, select the closest impedance curve and it sounds for the most part identical.
Doing the same with the helix sounds nowhere near close to real amps and I need to set the knobs in very different positions, put additional EQs and compressors, tweak the (not enough available) advanced parameters and then I'll just get in the ballpark. Then I'll go on tweaking to try to make it closer.

So yeah, I think @dr bonkers is right. This is based on first hand experience and not youtube videos, ymmv.

PS: this amount of tweaking was once necessary on the axe fx too, but Cliff's work during these years has brought his models closer and closer to the real deal, something that hasn't happened on the helix. I think in terms of accuracy they're stuck at the level the axe fx II was in its first few firmware revisions.

I have both an fm9 and a helix. I have no problem making the helix sound like the amps. Formerly it started with using york audio IRs. The new cabs bring it more in line.
Regardless this thread is more about the cab functionality and not the sounds of the Helix.

 

[DLC86]

I have both an fm9 and a helix. I have no problem making the helix sound like the amps. Formerly it started with using york audio IRs. The new cabs bring it more in line.
Regardless this thread is more about the cab functionality and not the sounds of the Helix.

As I said, ymmv. Maybe some amps are more accurate than others, the ones I'm more familiar with are not (e.g. hiwatt).
And I play them both thru IRs and real cabs so I don't think the IR is the issue here.

 

[zedeloc]

My experience with the HX cab engine was a tone-chasing time sink that was immediately resolved by loading a good IR. Granted, the word on the street is one of L6's weaknesses are their cabs. So maybe the quality of the underlying IRs that make up the cab engine are the issue.

I spent a good amount of time with the new helix cab engine. My other guitarist (helix) and I (fractal) are prepping to go direct for a concert. After spending a good amount of time moving mics, changing cabs, changing mics, changing amp settings, and busting out the eq block... he just kept getting lost in the mix. His levels were louder than mine and his guitar still wouldn't cut through. So I loaded some IRs in the helix and his tone was immediately better. A couple tweaks, and the tone was great. Its anecdotal, and could totally just be my own user error. But it seems to me that bells and whistles just cannot beat a good IR.

 

[dean701]

My experience with the HX cab engine was a tone-chasing time sink that was immediately resolved by loading a good IR. Granted, the word on the street is one of L6's weaknesses are their cabs. So maybe the quality of the underlying IRs that make up the cab engine are the issue.

I spent a good amount of time with the new helix cab engine. My other guitarist (helix) and I (fractal) are prepping to go direct for a concert. After spending a good amount of time moving mics, changing cabs, changing mics, changing amp settings, and busting out the eq block... he just kept getting lost in the mix. His levels were louder than mine and his guitar still wouldn't cut through. So I loaded some IRs in the helix and his tone was immediately better. A couple tweaks, and the tone was great. Its anecdotal, and could totally just be my own user error. But it seems to me that bells and whistles just cannot beat a good IR.

The cabs are just IRs. It's just a matter of tuning to find what gets through the mix. In your case you have an IR already handy that worked for you. Now he can reference that IR and see what cab/mic combination gets closest or not if he just wants to use that IR. This is the same for all modelers. In the Fractal world it comes with a plethora of included IRs, some work for certain amps better than others. Some people buy more or craft their own. It's all what works for the nduvidual and the situation.
The main point is the cab gui built within as opposed to picking random file names with no idea how they sound.

 

[Will Chen]

My experience with the HX cab engine was a tone-chasing time sink that was immediately resolved by loading a good IR. Granted, the word on the street is one of L6's weaknesses are their cabs. So maybe the quality of the underlying IRs that make up the cab engine are the issue.

I spent a good amount of time with the new helix cab engine. My other guitarist (helix) and I (fractal) are prepping to go direct for a concert. After spending a good amount of time moving mics, changing cabs, changing mics, changing amp settings, and busting out the eq block... he just kept getting lost in the mix. His levels were louder than mine and his guitar still wouldn't cut through. So I loaded some IRs in the helix and his tone was immediately better. A couple tweaks, and the tone was great. Its anecdotal, and could totally just be my own user error. But it seems to me that bells and whistles just cannot beat a good IR.

Do you have experience micing cabs? The Helix cab engine are IRs wrapped in a GUI which may be more intuitive for those who are more comfortable doing that especially when dialing in a new amp. But certainly if one has a go to IR, especially a multimiced/post processed IR, it's not going to replace that. I'd love to see/hear Frcatal's take on a similar GUI.

 

[Rajnar]

Do you have experience micing cabs? The Helix cab engine are IRs wrapped in a GUI which may be more intuitive for those who are more comfortable doing that especially when dialing in a new amp. But certainly if one has a go to IR, especially a multimiced/post processed IR, it's not going to replace that. I'd love to see/hear Frcatal's take on a similar GUI.

Word on the street, or at least up-thread, is that Cliff et al are working on something big along those lines.

 

[Vio]

I love posts where people tell Fractal what they need to do...

but seriously

 

[Will Chen]

Word on the street, or at least up-thread, is that Cliff et al are working on something big along those lines.

Right. Looking forward to it.

 

[zedeloc]

Do you have experience micing cabs? The Helix can engine are IRs wrapped in a GUI which may be more intuitive for those who are more comfortable doing that especially when dialing in a new amp. But certainly if one has a go to IR, especially a multimiced/post processed IR, it's not going to replace that. I'd love to see/hear Frcatal's take on a similar GUI.

I am also excited to check out fractals innovation with cab simulation. I've miced cabs before, getting pretty good tones. But I am not an expert, and yes it could definitely be user error on my part. I'm just saying that I get better results, faster, from IRs. And in reference to hx vs axe-fx, I consistently got better results, specifically concerning IRs/cabs, from the fractal. When using external IRs on the helix, it seems to help out a lot and the quality difference isn't so drastic. This is before and after the HX cab update. My real question isn't whether the concept of the Cab Engine is useful. But whether the underlying IRs just don't gel with me. And of course the big what-if is whether it is user error.

 

[Will Chen]

I am also excited to check out fractals innovation with cab simulation. I've miced cabs before, getting pretty good tones. But I am not an expert, and yes it could definitely be user error on my part. I'm just saying that I get better results, faster, from IRs. And in reference to hx vs axe-fx, I consistently got better results, specifically concerning IRs/cabs, from the fractal. When using external IRs on the helix, it seems to help out a lot and the quality difference isn't so drastic. This is before and after the HX cab update. My real question isn't whether the concept of the Cab Engine is useful. But whether the underlying IRs just don't gel with me. And of course the big what-if is whether it is user error.

To be clear, I much prefer the FM3 to HX Stomp. All I'm saying is this style of UI with a visual of the mic position/angle makes a ton more sense to me than much of the current Fractal IRs with names like "4X12 Greenback 57 A". And knowing Fractal's dedication to accuracy, I know something like would be pretty special and allows an opportunity to do things like automate the impedance curve selection. Really looking forward to what they do.

 

[zedeloc]

To be clear, I much prefer the FM3 to HX Stomp. All I'm saying is this style of UI with a visual of the mic position/angle makes a ton more sense to me than much of the current Fractal IRs with names like "4X12 Greenback 57 A". And knowing Fractal's dedication to accuracy, I know something like would be pretty special and allows an opportunity to do things like automate the impedance curve selection. Really looking forward to what they do.

I totally agree with you. I am an inherent tone tweaker and am really excited to see this new approach.