Jump to content
amsdenj

IR referencing and hi cut

Recommended Posts

This is interesting. I'm guessing a too subtle slope would explain why when I try to tame the krinkle to where I like it but  then overall tone just gets too low middy and kinda dead. ?

 

 

 

Got it in one.

 

That is precisely the issue / problem  with how the current  2.12  Lo / Hi  Cut  "slopes"  are working.

 

Hopefully this is "fixed"  soon.

 

Ben

Share this post


Link to post
Share on other sites

i agree that the slopes seem pretty low and too find i get into muddy territory easily trying to balance out the tone- selectable slope would be nice

Share this post


Link to post
Share on other sites

My personal Gig experience: sunday had a TV show big orchestra an loaded only the new Celestion irs- alnico blue for clean fender tones and creamback for Plexi lead (sm57 royer121 hi and low gain type). I did'nt use any hi/low cut and compared to the 2 nd guitar player in the orchestra (he use Helix as pedals only in front of a Fender deville miced only with an sm57) my sound was more intellegible and crispy than my collegue. One year ago had another live TV show and I used only the helix cab with hi cut to 3 khz. I suspect Celestion irs are ready to go irs and my suggestion is cut or add only on the amp model and dist pedal models to taste. Please Digital Igloo we need your contribution and tips.

  • Upvote 1

Share this post


Link to post
Share on other sites

There's one thing that i don't understand: if you obtain the same sound that you can have from kemper/axe setting low values with the filters, why bother to match the numbers with other modelers?
It's only a number: if it sounds good there's no problem in my opinion.
It would be different if the sound will be different/less good.
Or am I missing something?

Share this post


Link to post
Share on other sites

Quantifiable measurements of how Helix processes IRs vs other reputable IR loaders would be really great. If anyone has the time and tools and knowledge to do that, please have at it, with much gratitude from all of us.

 

Even as it stands, the anecdotal evidence seems pretty strong. Maybe folks who've observed this first-hand should file an issue with support? Or maybe we need some IdeaScale entries? DI does hang here some, but this isn't an official Line 6 attention-getting, product-changing portal.

 

Lastly, what happens if Line 6 acknowledges some sort of IR processing behavior that they agree should be changed? The sound of everyone's patches would change, possibly for the better maybe, with potential for even better than that once you'd adjusted your patches for the new approach, but if I were a gigging musician, I sure wouldn't just press Go on that upgrade, without making certain I had time to revisit every patch my gigs require. I'd suggest that if an IR processing change becomes reality, there should be a per-patch setting for Old Style vs New Style IR handling, so you could migrate your patches one by one.

Share this post


Link to post
Share on other sites

Quantifiable measurements of how Helix processes IRs vs other reputable IR loaders would be really great. If anyone has the time and tools and knowledge to do that, please have at it, with much gratitude from all of us.

 

Even as it stands, the anecdotal evidence seems pretty strong. Maybe folks who've observed this first-hand should file an issue with support? Or maybe we need some IdeaScale entries? DI does hang here some, but this isn't an official Line 6 attention-getting, product-changing portal.

 

Lastly, what happens if Line 6 acknowledges some sort of IR processing behavior that they agree should be changed? The sound of everyone's patches would change, possibly for the better maybe, with potential for even better than that once you'd adjusted your patches for the new approach, but if I were a gigging musician, I sure wouldn't just press Go on that upgrade, without making certain I had time to revisit every patch my gigs require. I'd suggest that if an IR processing change becomes reality, there should be a per-patch setting for Old Style vs New Style IR handling, so you could migrate your patches one by one.

 

I'm wouldn't be too quick to advocate changing the IR block processing even if it is a bit brighter because you have the think about the bigger picture.  There are some combinations of amps, guitar, and IR that don't even need any high cut.  What happens to them?  Will they get darker or more muddy?  It's probably pretty easy to take out frequencies, but it becomes a lot more challenging and artificial if you need to add them back in for some reason.

 

If I were to advocate anything I might thing a better approach would be to add another parameter to the IR block that allows you to adjust the slope on the high cut and low cut parameters.  That would be the most flexible way to do it.

 

Therefore I took the opportunity to add it to IdeaScale.  Regardless of the discussions here, I think it would add a nice capability to the platform.  You can vote it up at the link below:

 

https://line6.ideascale.com/a/dtd/Adjustable-Slope-Parameter-for-IR-Blocks/869790-23508?submitted=1

  • Upvote 5

Share this post


Link to post
Share on other sites

From Ben Adrian over on TGP:

 

Just opened the code to check. The crossover filters are 4th order (4 pole) low pass and high pass. This is 24 dB per octave. IMO, this would be considered a pretty hard filter.

 

Cabs and IRs have 2nd order (2 pole), which is 12 dB per octave. I'd consider this a medium cutoff.

 

EQs have a 2nd order (2 pole) filter for the models which contain the Low and High Cut parameters, but there is also a little extra voicing thrown in with an additional filter to make it more musical.

Share this post


Link to post
Share on other sites

I'm wouldn't be too quick to advocate changing the IR block processing even if it is a bit brighter because you have the think about the bigger picture.  There are some combinations of amps, guitar, and IR that don't even need any high cut.  What happens to them?  Will they get darker or more muddy?  It's probably pretty easy to take out frequencies, but it becomes a lot more challenging and artificial if you need to add them back in for some reason.

 

If I were to advocate anything I might thing a better approach would be to add another parameter to the IR block that allows you to adjust the slope on the high cut and low cut parameters.  That would be the most flexible way to do it.

 

Therefore I took the opportunity to add it to IdeaScale.  Regardless of the discussions here, I think it would add a nice capability to the platform.  You can vote it up at the link below:

 

https://line6.ideascale.com/a/dtd/Adjustable-Slope-Parameter-for-IR-Blocks/869790-23508?submitted=1

 

Here is a related idea for high/low cuts in general: https://line6.ideascale.com/a/dtd/Tapered-high-low-cuts-for-more-realistic-amp-cab-modeling/840630-23508

Share this post


Link to post
Share on other sites

I'm wouldn't be too quick to advocate changing the IR block processing even if it is a bit brighter because you have the think about the bigger picture.  There are some combinations of amps, guitar, and IR that don't even need any high cut.  What happens to them?  Will they get darker or more muddy?  It's probably pretty easy to take out frequencies, but it becomes a lot more challenging and artificial if you need to add them back in for some reason.

Here's a simple solution. Change IR hi cut to hi shelf boost/cut. Fix the IR block so it faithfully reproduces the IR, not a brighter version of the IR. Set the hi shelf default to the boost in the current IR block so that there's no change to existing patches. Users can then adjust to taste.

 

But the key point is that the IR block should not need any filtering to reproduce the intended sound of the provided IR. The filtering should be to adjust from what the IR was designed to do to support the user's taste, or target FRFR. In this case, a shelf would be more flexible because it could make the IR brighter or darker depending on the user's needs.

Share this post


Link to post
Share on other sites

I like the idea of a more easily customized high/low cut with an adjustable curve, this gives you the option of a shelf or a more gently curved cut. I am not sure a high/low shelf is the answer due to the lack of ability to customize the rate of the cutoff.. A shelf cut may not be ideal to use for shaping IRs although for all I know it is what we have now. I agree with asmsdenj that if the current Helix IR implementation is flawed now is the time to fix it rather than to proceed with it broken and just add more customized cuts that essentially act only as a band-aid for a larger inherent problem. Along with normalizing the IR implementation if indeed it is flawed, it would also be great to have high/low cuts with a db curve parameter to provide greater flexibility in shaping the EQ for IRs.

Share this post


Link to post
Share on other sites

Yes please, particularly for loud drummer levels, greater or adjustable slope on the IR block cuts to tame piercing highs without getting muddy.

On the low cut too, I think it would tame the disparity I find between bridge and neck pickups. Is it just me? When I've removed enough low end for my neck single coil, the bridge humbucker starts to get a bit thin.

Share this post


Link to post
Share on other sites

Yes please, particularly for loud drummer levels, greater or adjustable slope on the IR block cuts to tame piercing highs without getting muddy.

On the low cut too, I think it would tame the disparity I find between bridge and neck pickups. Is it just me? When I've removed enough low end for my neck single coil, the bridge humbucker starts to get a bit thin.

 

Try adjusting your tone for the neck pickup. Then use the tone control on your guitar to fine tune for the bridge pickup, just to remove a bit of the ice pick. That should provide a good compromise.

  • Upvote 1

Share this post


Link to post
Share on other sites

Try adjusting your tone for the neck pickup. Then use the tone control on your guitar to fine tune for the bridge pickup, just to remove a bit of the ice pick. That should provide a good compromise.

My problem is more with the low end. Depending a little on the amount of overdrive and IR choice, I find my neck single coil is good with a low cut around 145Hz to avoid boominess, but then the bridge humbucker sounds a bit skinny and is better with low cut around 90Hz. I wondered if it was due to the low cut slope of the IR block in Helix being a bit gentle, similar to that being said about the hi cut. I never really felt disparity with my main guitar using helix or various real amps through real cabs.

 

I do appreciate your advice about the ice pick, thanks. I should be more judicious with the tone control.

 

I'm ok enough with with the cuts in the IR block. Perhaps they could be improved to avoid some compromise between beef and boom or icepick/fizz and clarity. Maybe i'm still unrealistically hoping to make an FRFR be a 2x12 !

Again, talking live and loud, for recording I can tweak the low end and Iike more brightness

Share this post


Link to post
Share on other sites

I filled out a support ticket letting Line6 know that there is a noticeable difference between how Helix processes an IR and how a number of different DAW plugins process the same IR. The Helix IR block processing is much brighter, more brittle and harsher than all the other convolution plugins I tried and requires a significant amount of high cut in the IR block to tame this high frequency boost.

 

Line6 support's response was essentially "All I can say is that there is no hard and fast rule that they need to be exactly the same. If you would like to make suggestions for improvements, you can submit your ideas here: https://line6.ideascale.com/".

 

Convolution is in fact a very specific mathematical process. IR producers like Rewirez, Ownhammer, Rosen Digital, 3sigma audio, Celestion and many others painstakingly capture IRs for speakers, cabinets and mics in order to faithfully capture the dynamic frequency response of the source. There certainly can be some art in how the IRs are captured, how they are mixed,  and how they are processed to produce a product. That's what distinguishes these different vendors. And of course the speaker instances they are capturing are also different. So we shouldn't necessarily expect IRs from different vendors of the same speaker/cabinet configuration to sound exactly the same.

 

The same can be true of convolution engines that process audio through the IR. These can be different too, adding features like Z or damping factor, phasing, delay, high/low cut, etc. However, these are typically subtle differences, not significant deviations from the sound the IR was intended to capture.

 

Line6 should consider doing some testing of its own to verify whether the Helix IRs are working as intended or if there is something causing the output to be overly bright. If Helix is working as intended, then Line6 should provide Helix users some guidance in how to use the IR block to produce the output as intended by the IR producer.

 

Otherwise we are not necessarily getting the intended value out of our IR purchases or creations (I've created a few myself). For example, I used my DAW and a convolution plugin to audition a large number of IRs to determine the set I wanted to include in Helix. But when I added those IRs to Helix and tried to use them at rehearsal, they were way to bright and didn't sound like they did when I auditioned them. This is what prompted me to do the testing described in this thread.

  • Upvote 1

Share this post


Link to post
Share on other sites

Thanks for pursuing this Jim.

 

Disappointing response from Line 6, I have to say. FWIW, I agree difference between IR processors should be subtle, not the obvious differences folks are reporting.

 

It'd be great to show them frequency response plots from several processors alongside Helix. Based on the reports here, I'd expect a cluster of similar graphs, with Helix as the outlier.

 

Of course it's unlikely anyone here is going to provide an improved IR rendering algorithm, so the only way I can see Line 6's response besides blind obstructionism is that we need an IdeaScale item saying that Helix should process IRs more like other software. I'd upvote that in a flash, with or without comparative response plots.

  • Upvote 1
  • Downvote 1

Share this post


Link to post
Share on other sites

I am not sure that this potential issue with IRs does not extend to the Helix native cabs as well which are essentially optimized customized IRs. Those tend to be a bit on the bright and woofy side as well although changing the mic selection can make a significant difference and help in adjusting the EQ. Would definitely love to hear DI or Line6 chime in with a technical explanation of their IR implementation and, as amsdenj mentioned, perhaps specific instructions on how to EQ their cabs and IRs for best results on the Helix. Maybe this is simply an issue where better educating the user base would do the trick. Or, perhaps there is a fix or EQ modification that needs to happen under the covers that would require less modification by the user to tame these beasts and have them responding in a fashion more akin to other products such as the Axe, Kemper, and various DAWs and their convolution engines. Still loving my Helix and able to get some great sounds out of it but I think this thread has brought an issue to light that perhaps requires addressing.

Share this post


Link to post
Share on other sites

Ben Adrian posted this on the TGP Thread where they compared an Fourier transform of an IR with the result of a frequency sweep through the same IR loaded into the Helix IR block without any filtering and online digital signal paths:

 

https://www.thegearpage.net/board/index.php?threads/line6-helix.1586637/page-1701#post-23640385

 

The identical response indicates that it is not the Helix IR processing that is causing differences.

 

Ben also in another response checked the code and said that the filters are supposed to be second order on both Cab and IR blocks but didn't dig in further because it wasn't something that he coded and I am sure that we would all prefer it if he was spending his time developing new models.

 

But that doesn't mean that the high and low filtering is working correctly or that some other component is not changing the sound - all that the measurements show is that the IR convolution process is giving as close to the ideal output for the IR as can be seen through the test process.

 

It is quite possible that the other IR Loaders all include a degree of high cut filtering that Line 6 in the interest of a perfect signal path hasn't included. Perhaps the same comparison for the DAW based IR should be attempted

  • Upvote 1

Share this post


Link to post
Share on other sites

Ben Adrian posted this on the TGP Thread where they compared an Fourier transform of an IR with the result of a frequency sweep through the same IR loaded into the Helix IR block without any filtering and online digital signal paths:

 

https://www.thegearpage.net/board/index.php?threads/line6-helix.1586637/page-1701#post-23640385

 

The identical response indicates that it is not the Helix IR processing that is causing differences.

 

 

This is great news and very helpful. This means that something else must explain what I observed - including the possibility that I observed incorrectly for some reason! I'll study this some more tonight with some additional tests and capture some recordings.

Share this post


Link to post
Share on other sites

Well, I'm the first to admit when I'm wrong. I did a more thorough test tonight of the Helix IR block and mixIR2, LAConvolver, Logic Pro X Space Designer, NadIR Convolver and S-Gear Pro Convolver, all using the same IR. 

 

I created a patch in Helix that uses the Archetype Clean amp with default settings except for Presence=5.0 and Bright=On (to put a little more high end into the cabs and IRs. The amp output is routed to USB 1/2. The output of the amp is also routed to another parallel path that has a Line6 Dual Cab (1x12 Celest 12-H) that is output to USB 3/4, and a third parallel path has thhe IR block with G12-65 212 C R-121 Balanced Celestion IR that is output to USB 5/6.

 

In Logic Pro X I created three tracks:

1. Helix Cab with input from USB 3/4 - Helix: Archetype Clean > Dual Cab with 1x12 Celest 12-H > USB 3/4

2. Helix IR with input from USB 5/6 - Helix: Archetype Clean > IR block with G12-65 212 C R-121 Balanced Celestion > USB 5/6

3. Helix Amp with input from USB 1/2 - Helix: Archetype Clean > USB 1/2 (raw amp)

 

I recorded a bit of my Variax Standard on the magnetic neck pickup on all three tracks at the same time. I then copied the Helix Amp track and added each of the convolution plugins above, all configured with the same G12-65 212 C R-121 Balanced Celestion IR. 

 

After carefully setting the gain of all tracks so they had the same output level, I compared all the tracks with the convolution plugins with the recording of the output of the Helix IR block simply using Option-Click on the track Solo button. I could not detect any difference between any of the renderings of the IR blocks. As expected, the Helix Cab track did sound different as its a different speaker. 

 

I'm not sure why I got the results I got before. I checked the original recordings before I started this test and they still sounded different. I must have had something wrong in the setup.

 

So I apologize to everyone, and especially Line6, for the distraction resulting from sloppy testing. However, it is nice to know that the Helix IR block is working as expected.

 

Now back to why my patches sound so bright into my EON610 monitor compared to a reference amp with the same Celestion G12-65 speakers. It may be simply where I was sitting, getting too much direct sound from that horn. I'm going to go back and do those references again this weekend and see what else I can learn. This time I'll reference just the speakers against the IRs using Helix amp model directly into a power amp into the reference speakers. That will eliminate any differences caused by going through my Showman tube amp.

  • Upvote 2

Share this post


Link to post
Share on other sites

Now back to why my patches sound so bright into my EON610 monitor compared to a reference amp with the same Celestion G12-65 speakers. It may be simply where I was sitting, getting too much direct sound from that horn. I'm going to go back and do those references again this weekend and see what else I can learn. This time I'll reference just the speakers against the IRs using Helix amp model directly into a power amp into the reference speakers. That will eliminate any differences caused by going through my Showman tube amp.

It could simply be where your ears are relative to where a mic would normally be placed. If you stick your face up to the speaker grill and listen to it in the near field, it will have much more high frequency content than if you stand a few away and 30 degrees off axis.

 

In the far field, a 12" speaker beams higher frequencies in an increasingly tighter angle as the frequency goes up, especially frequencies past about 1kHz. This is why FRFRs have a crossover to send those frequencies to a speaker with a smaller aperture and/or a horn. It's all about the changing angle of dispersion over the frequency response of the speaker.

Share this post


Link to post
Share on other sites

Well, I'm the first to admit when I'm wrong. I did a more thorough test tonight of the Helix IR block and mixIR2, LAConvolver, Logic Pro X Space Designer, NadIR Convolver and S-Gear Pro Convolver, all using the same IR.

 

I created a patch in Helix that uses the Archetype Clean amp with default settings except for Presence=5.0 and Bright=On (to put a little more high end into the cabs and IRs. The amp output is routed to USB 1/2. The output of the amp is also routed to another parallel path that has a Line6 Dual Cab (1x12 Celest 12-H) that is output to USB 3/4, and a third parallel path has thhe IR block with G12-65 212 C R-121 Balanced Celestion IR that is output to USB 5/6.

 

In Logic Pro X I created three tracks:

1. Helix Cab with input from USB 3/4 - Helix: Archetype Clean > Dual Cab with 1x12 Celest 12-H > USB 3/4

2. Helix IR with input from USB 5/6 - Helix: Archetype Clean > IR block with G12-65 212 C R-121 Balanced Celestion > USB 5/6

3. Helix Amp with input from USB 1/2 - Helix: Archetype Clean > USB 1/2 (raw amp)

 

I recorded a bit of my Variax Standard on the magnetic neck pickup on all three tracks at the same time. I then copied the Helix Amp track and added each of the convolution plugins above, all configured with the same G12-65 212 C R-121 Balanced Celestion IR.

 

After carefully setting the gain of all tracks so they had the same output level, I compared all the tracks with the convolution plugins with the recording of the output of the Helix IR block simply using Option-Click on the track Solo button. I could not detect any difference between any of the renderings of the IR blocks. As expected, the Helix Cab track did sound different as its a different speaker.

 

I'm not sure why I got the results I got before. I checked the original recordings before I started this test and they still sounded different. I must have had something wrong in the setup.

 

So I apologize to everyone, and especially Line6, for the distraction resulting from sloppy testing. However, it is nice to know that the Helix IR block is working as expected.

 

Now back to why my patches sound so bright into my EON610 monitor compared to a reference amp with the same Celestion G12-65 speakers. It may be simply where I was sitting, getting too much direct sound from that horn. I'm going to go back and do those references again this weekend and see what else I can learn. This time I'll reference just the speakers against the IRs using Helix amp model directly into a power amp into the reference speakers. That will eliminate any differences caused by going through my Showman tube amp.

Thanks for your continued testing. Your results this round are very reassuring. Ultimately what I would be curious about is how various modeling devices and DAW convolution engines compare direct out to a powered FRFR speaker. I suppose the fact that each device/computer may have different D/A conversion and signal/output processing makes this difficult to do an apples to apples comparison, especially as some devices or apps may be applying some kind of automatic behind the scenes correction via an EQ algorithm. With a little practice you get facile at EQ'ing the Helix but I still can't help but wonder why the IRs and cabs tend to be so strong on the low and high end and often tend to require so much EQ to get an optimal sound. It seems like they should be a little more restrained without modification but perhaps it is just the nature of the beast or an effort to offer maximum flexibility. Maybe we have just come full circle back to the fact that an FRFR speaker simply does not respond like the more limited range and different frequency response of a traditional guitar cab.

Share this post


Link to post
Share on other sites

I would add that excessive highs and lows from an FRFR speaker may be because it isn't really flat response - many PA speakers are "optimised" so they sound good for their most common use... i.e. Disco and other pre-recorded dance music and will have a boost for the highs and lows.

Even the Line 6 L series speaker have this, but fortunately you have to select the PLAYBACK mode while REFERENCE / P.A. attempts to be as flat as possible.

Share this post


Link to post
Share on other sites

Thanks for your continued testing. Your results this round are very reassuring. Ultimately what I would be curious about is how various modeling devices and DAW convolution engines compare direct out to a powered FRFR speaker. I suppose the fact that each device/computer may have different D/A conversion and signal/output processing makes this difficult to do an apples to apples comparison, especially as some devices or apps may be applying some kind of automatic behind the scenes correction via an EQ algorithm. With a little practice you get facile at EQ'ing the Helix but I still can't help but wonder why the IRs and cabs tend to be so strong on the low and high end and often tend to require so much EQ to get an optimal sound. It seems like they should be a little more restrained without modification but perhaps it is just the nature of the beast or an effort to offer maximum flexibility. Maybe we have just come full circle back to the fact that an FRFR speaker simply does not respond like the more limited range and different frequency response of a traditional guitar cab.

 

Yes, I've been wondering about this myself... It's my understanding, if you're using an L2t or L2m with L6 Link, the Helix is actually sending the digital signal to the amplifiers, and the D/A conversion is being handled on the StageSource Speaker, not in the Helix.  I wonder what the difference is between the D/A converters in the Helix vs the StageSource speakers?  Is this why some people like to use XLR connection instead, so the D/A is handled by the Helix?

 

One thing I've been wondering, and I posted this in The Thread, but nobody responded... as flat as the L2t is supposed to be, the published response shows

- A 3 db bump up between 60 to 180 Hz in Free Space

- An 8 dB bump up between 60 to 180 Hz in Half Space (seems to me this is the important spec in a room?)

- A 2 dB dip at 800 Hz

- A 4 dB dip at 2500 Hz

- An up and down craziness of up to 4 dB between 10K and 20K Hz

 

Am I wrong in how I'm interpretting/thinking about the Frequency Response Graph?

 

Should we be using an EQ to flatten this out some?

 

Seems to me there should at least be some shelving below 200 Hz and above 8 KHz?

 

How does coupling with the floor come into play with this if I have the speaker sitting on the floor?

 

https://www.dropbox.com/s/1pilqy77qvdc3rr/Line%206%20L2t%20Response2.png?dl=0

 

Line6_L2t_Response.jpg

 

(P.S. I can't find any help on how to show an image in a post.  Does it not like DropBox or something?)

post-1915414-0-01695800-1489759077_thumb.jpg

Share this post


Link to post
Share on other sites

Yes, I've been wondering about this myself... It's my understanding, if you're using an L2t or L2m with L6 Link, the Helix is actually sending the digital signal to the amplifiers, and the D/A conversion is being handled on the StageSource Speaker, not in the Helix.  I wonder what the difference is between the D/A converters in the Helix vs the StageSource speakers?  Is this why some people like to use XLR connection instead, so the D/A is handled by the Helix?

 

One thing I've been wondering, and I posted this in The Thread, but nobody responded... as flat as the L2t is supposed to be, the published response shows

- A 3 db bump up between 60 to 180 Hz in Free Space

- An 8 dB bump up between 60 to 180 Hz in Half Space (seems to me this is the important spec in a room?)

- A 2 dB dip at 800 Hz

- A 4 dB dip at 2500 Hz

- An up and down craziness of up to 4 dB between 10K and 20K Hz

 

Am I wrong in how I'm interpretting/thinking about the Frequency Response Graph?

 

Should we be using an EQ to flatten this out some?

 

Seems to me there should at least be some shelving below 200 Hz and above 8 KHz?

 

How does coupling with the floor come into play with this if I have the speaker sitting on the floor?

 

https://www.dropbox.com/s/1pilqy77qvdc3rr/Line%206%20L2t%20Response2.png?dl=0

 

Line6_L2t_Response.jpg

 

(P.S. I can't find any help on how to show an image in a post.  Does it not like DropBox or something?)

 

 

This matches my experience with my L2t.  I've currently got it in ref mode, on the floor, tilted back at 45 degrees on the kickstands.  When I go to my neck pickup on my PRS and hit a single note on the low E 5-12th frets the room rattles.  There is a huge bump in the 150hz range.   I've tried low cutting the IR around 180, which helps tame the boomy, but when I go back to my bridge pickup it's thin.   Settled on a compromise low filter around 140.  

 

Still trying to tame the ice-picky harshness.  If I set the the high cut down around 3k it sounds less ice-picky, but is missing some sparkle.  

 

Still struggling with this rig.  

Share this post


Link to post
Share on other sites

This matches my experience with my L2t.  I've currently got it in ref mode, on the floor, tilted back at 45 degrees on the kickstands.  When I go to my neck pickup on my PRS and hit a single note on the low E 5-12th frets the room rattles.  There is a huge bump in the 150hz range.   I've tried low cutting the IR around 180, which helps tame the boomy, but when I go back to my bridge pickup it's thin.   Settled on a compromise low filter around 140.  

 

Still trying to tame the ice-picky harshness.  If I set the the high cut down around 3k it sounds less ice-picky, but is missing some sparkle.  

 

Still struggling with this rig.  

 

How were you applying your filters?  From the Cab or an eq?  As Ben pointed out, the different tools act differently, and there are multiple locations and ways to apply filters.  It is my understanding that the cab filters aren't very aggressive, as I think is noted in a quote from Ben above.

 

From my post, I'm wondering if not only should we be applying cuts, but maybe also some Hi Q boosts in the midrange as well?  Perhaps it might quell some of the boom and ice pick by allowing you to use a lower actual volume to get a similar perceived volume.

Share this post


Link to post
Share on other sites

How were you applying your filters?  From the Cab or an eq? 

 

I've tried high/low cut from amp+cab blocks, and from the IR blocks.  So far I haven't inserted any EQ blocks or touched the global EQ.   I *thought* I could just create a blank preset, throw in an amp block followed by an IR block, and then audition IR blocks until I found one I like.

 

For years (and years, and years) I played through a Mesa Boogie slant with Vintage 30's, mic'd with a 57.  So naturally I thought I'd buy the Redwirez Mesa V30 IR, and then pick one.  It's never that simple, is it?  

 

I KNOW that the FRFR is never going to sound/feel like my Triaxis/50:50/cab.  I totally understand that.  But it's SO bright and boomy that I'm obviously doing something wrong.  Which knob to tweak?  Things that are new to me 1) modeling  2) IRs  3) FRFR.   So many variables.  

Share this post


Link to post
Share on other sites

I'm wouldn't be too quick to advocate changing the IR block processing even if it is a bit brighter because you have the think about the bigger picture.  There are some combinations of amps, guitar, and IR that don't even need any high cut.  What happens to them?  Will they get darker or more muddy?  It's probably pretty easy to take out frequencies, but it becomes a lot more challenging and artificial if you need to add them back in for some reason.

 

If I were to advocate anything I might thing a better approach would be to add another parameter to the IR block that allows you to adjust the slope on the high cut and low cut parameters.  That would be the most flexible way to do it.

 

Therefore I took the opportunity to add it to IdeaScale.  Regardless of the discussions here, I think it would add a nice capability to the platform.  You can vote it up at the link below:

 

https://line6.ideascale.com/a/dtd/Adjustable-Slope-Parameter-for-IR-Blocks/869790-23508?submitted=1

 

 

 

Voted both.

I think being able to choose the db per octave, or the curve of the slope would definitely help in getting more precision out of our tone sculpting tools within the Helix. I think it will help mitigate some of the post processing EQ I do when recording as well.

 

But I don't want this just on the IR block. I would like to see this implemented in the HX Cabs as well!

Share this post


Link to post
Share on other sites

I'm beginning to come to the conclusion that its not the IR's that's making things too bright and boomy, and its not the specific frequency response of the FRFR (if its reasonably flat). Rather I think its how we're listening and how a mic'd guitar cabinet sounds through a PA vs. how it sounds "in the room". Guitar cabinets have limited frequency response and the high frequencies are very directional. We generally listen to guitar amps very off axis and so we don't get hit with all those high frequencies coming out of a very narrow band from the center of the speaker.

 

But that changes when you stick a mic in front of a cabinet, or use an IR with a similarly positioned mic. The mic is going to hear exactly what you would hear if you put your ear right into the middle of that speaker. That's going to be very bright as you're hearing all the high frequencies the speaker is able to reproduce. This will not sound natural, so we shouldn't necessarily expect a mic positioned in the same place to sound natural. When a Helix amp model is played through an IR, the IR does indeed reproduce quite well the dynamic frequency response of the modeled cabinet, just like if you had that cabinet and stuck the same mic at the same position in front of it. What's different is the FRFR. Typically these have much wider high-end dispersion due to horns or tweeters. So you'll hear a lot more of that high end over a much wider range of positions.

 

It is that phenomena that we need to compensate for with high cut. We're essentially using the high cut to reproduce the amp "in the room" sound that we're use to because of the directionality and narrow dispersion of the high frequency response of a guitar speaker.

  • Upvote 4

Share this post


Link to post
Share on other sites

Amsdenj has a good point. I bought an amp stand to tilt my 2x12 up, and it was exceedingly bringht in a very narrow arc. You had to be standing in exactly the right place. My L2t seems to widen that "face melting" arc from a few degrees to 180.

 

Maybe that means I need to audition more off-axis IRs.

Share this post


Link to post
Share on other sites

I'm beginning to come to the conclusion that its not the IR's that's making things too bright and boomy, and its not the specific frequency response of the FRFR (if its reasonably flat). Rather I think its how we're listening and how a mic'd guitar cabinet sounds through a PA vs. how it sounds "in the room". Guitar cabinets have limited frequency response and the high frequencies are very directional. We generally listen to guitar amps very off axis and so we don't get hit with all those high frequencies coming out of a very narrow band from the center of the speaker.

 

But that changes when you stick a mic in front of a cabinet, or use an IR with a similarly positioned mic. The mic is going to hear exactly what you would hear if you put your ear right into the middle of that speaker. That's going to be very bright as you're hearing all the high frequencies the speaker is able to reproduce. This will not sound natural, so we shouldn't necessarily expect a mic positioned in the same place to sound natural. When a Helix amp model is played through an IR, the IR does indeed reproduce quite well the dynamic frequency response of the modeled cabinet, just like if you had that cabinet and stuck the same mic at the same position in front of it. What's different is the FRFR. Typically these have much wider high-end dispersion due to horns or tweeters. So you'll hear a lot more of that high end over a much wider range of positions.

 

It is that phenomena that we need to compensate for with high cut. We're essentially using the high cut to reproduce the amp "in the room" sound that we're use to because of the directionality and narrow dispersion of the high frequency response of a guitar speaker.

How do you mark a post as the answer?

Share this post


Link to post
Share on other sites

Hi all

 

All excellent points :)

 

In addition to this - and my apologies as I probably should have posted this here in this thread rather than in a new discussion - but please see here:-

 

http://line6.com/support/topic/26252-hi-lo-cab-and-ir-cuts-attached-audio-clip-proof-they-need-to-be-fixed/

 

The  exisitng  "non-standard slope"  of the current  FW 2.12  Hi / Lo Cut in the Cab / IR  Block is  i.m.h.o  a major contributor to a lot of these issues ... however  even a  db-octave-correct  [ 6db or 12db ]  slope will not address the  "amp in the room"  issue ... that is a cross-platform  [Helix/Axe/Kemper/AFire etc...]   FRFR-related issue

 

Ben

Share this post


Link to post
Share on other sites

 

 

 however  even a  db-octave-correct  [ 6db or 12db ]  slope will not address the  "amp in the room"  issue ... that is a cross-platform  [Helix/Axe/Kemper/AFire etc...]   FRFR-related issue

 

Ben

What they will do is give us more control over sculpting our sound. Could result in a bit less post processing, as well as getting your EQing closer without using an extra block.

Share this post


Link to post
Share on other sites

The exisitng "non-standard slope" of the current FW 2.12 Hi / Lo Cut in the Cab / IR Block is i.m.h.o a major contributor to a lot of these issues ... however even a db-octave-correct [ 6db or 12db ] slope will not address the "amp in the room" issue ... that is a cross-platform [Helix/Axe/Kemper/AFire etc...] FRFR-related issue

 

Now that I'm (hopefully) understanding this better, I would add two additional points. First, its possibly better to choose an IR that provides the tone you want before using high-cut to tailor the output. For me that means choosing IRs that use ribbon microphones, close mic'd and positioned at cone and cone edge, staying away from the bright cap or cap edge I typically used. This is actually how many sound engineers mic up electric guitars live. They often start with the mic at the cone edge then go out in the room and listen while the guitar player plays various different settings. Then they have an assistant gradually move the mic closer to the cap until the high end is about right, a balancing act for sure. This approach uses mic choice and positioning as the primary means of getting the desired tone result (getting it right at the source), and then using EQ to make smaller adjustments as needed.

 

The second point is I don't really mind the apparently very gradual slope of the IR high-cut on Helix. I feel like this helps dial in a sort of high-resolution high-cut that's gradual and smooth, doing what needs to be done while having a minimal overal impact on the IR.

 

6dB is pretty gradual too, so is 12. Much higher than that can start to sound unnatural and have resonance peaks. Bottom line, I don't really care where I have to set the high-cut knob as long as it does the job. For me, it seems to pretty well. My original issue was that I didn't understand that the high-cut was needed at all, or why.

  • Upvote 1

Share this post


Link to post
Share on other sites

It would be great to have two additional Helix Mic Cab Parameters. One for Mic Position (medial to lateral: Cap Center, Cap Edge, Mid Cone, Cone Edge); and one for Mic Aim/Orientation (0 to 90 Degrees).

 

Depending on whether Line 6 originally captured 3D acoustic data, it may be feasible to reimage the data based on these Parameters. Otherwise, it would almost certainly require the substantial task of re-capturing the cabs with each of the mics. Either way, it would vastly enhance the sound of Helix's built in cabs, likely with additional DSP overhead (subject to whatever algorithm optimization magic the coding teams are using). On the other hand, L6 may have intentionally left this for 3rd Party IRs.

  • Upvote 1

Share this post


Link to post
Share on other sites

Now that I'm (hopefully) understanding this better, I would add two additional points. First, its possibly better to choose an IR that provides the tone you want before using high-cut to tailor the output. For me that means choosing IRs that use ribbon microphones, close mic'd and positioned at cone and cone edge, staying away from the bright cap or cap edge I typically used. This is actually how many sound engineers mic up electric guitars live. They often start with the mic at the cone edge then go out in the room and listen while the guitar player plays various different settings. Then they have an assistant gradually move the mic closer to the cap until the high end is about right, a balancing act for sure. This approach uses mic choice and positioning as the primary means of getting the desired tone result (getting it right at the source), and then using EQ to make smaller adjustments as needed.

 

The second point is I don't really mind the apparently very gradual slope of the IR high-cut on Helix. I feel like this helps dial in a sort of high-resolution high-cut that's gradual and smooth, doing what needs to be done while having a minimal overal impact on the IR.

 

6dB is pretty gradual too, so is 12. Much higher than that can start to sound unnatural and have resonance peaks. Bottom line, I don't really care where I have to set the high-cut knob as long as it does the job. For me, it seems to pretty well. My original issue was that I didn't understand that the high-cut was needed at all, or why.

 

I totally agree with everything you said here.  It's nice to find out I'm not the only one having this experience.

 

I think most of this issue probably comes from people not having experience with how mic placement has such a significant effect on the tone you get from a cabinet.  It's that way with real cabinets and modeled cabinets work exactly the same.  If there's any issue Line 6 needs to address it would be greater flexibility in how their stock cabinets get mic'd.  Although I'm personally fine with using IR's, using stock cabinets with a greater range of mic placement would likely be more convenient.

 

I would echo and expand on what you've said that EQ is really only a last step tweak and may or may not be necessary depending on what you've done with the mic's and the cabinets.  I used to be one that assumed I would need to use a hi-cut/lo-cut until I began using IR's.  I still use them on some patches if none of the IR positioning or mic's sufficiently tames the highs or lows to my taste.  But it's never an assumption that I MUST do it.  Very often I can add a second different cabinet in a parallel mix and achieve a great tone with no EQ on either cabinet.  That may not be as DSL efficient, but iit;s certainly more authentic to the way things work vin the real world.

  • Upvote 1

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...