Snapshot latency when controlling other parameters

[mikeolivercgp]

I need to preface this question with praise for the new "snapshots" feature in Helix -- I really like having that capability! Creating "snapshots" is much quicker than using virtual stomp switches to bypass different combinations of blocks in a particular preset.

 

I did encounter something that surprised me. I wanted two different gain/volume configurations for a Minotaur OD block in a preset for a particular song (I use snapshots to setup what I need for different sections of the song; I was previously doing something similar with virtual stomp switches). It was easy and quick to indicate that the volume and gain for that particular block should be controlled by snapshots. 

 

For one snapshot, I enable a Soldano crunch amp head model, disable, a Cali Lead amp head model, and shift the gain and volume ratio of the Minotaur block to 0.6 and 6.4 respectively (I'm just slightly pushing on the amp with some additional harmonic color from the Minotaur).

 

For another snapshot in the same preset, I disable the Soldano amp head, enable the Cali lead amp head, and shift the gain and volume ratio of the Minotaur block to 4.2 and 5.3 respectively (for little more smooth saturation). 

 

What surprised me is that when I change snapshots, I experienced just a slight hit of latency accompanied by a brief white noise shift as the Minotaur settings readjust for their respective snapshot. Granted, in full band context, I doubt anyone would notice. However, I had hoped that this would be a transparent transition. 

 

So... I simply added another Minotaur block to the preset. Each Minotaur block is configured for the respective amp head model and enabled by respective snapshots. That approach is completely transparent.

 

I was fortunate that I had sufficient processor capacity to add the second Minotaur block. However, I had hoped that snapshots would enable me to have more flexibility in retasking the same FX block to conserve CPU capacity. For example, I'd hoped to use snapshots to change the tempo note value and feedback on the same delay block as opposed to what I do today, which is to: add two delay blocks with tails on (depending on the song the preset is configured for) and enabling one while disabling the other. [bTW, I've not tried this with the delays; it might a non-issue with them, especially if the little noise anomaly I experienced with the Minotaur had more to do with sensitivity of the gain stage of those amp heads.]

 

Not complaining... just curious if anyone else has encountered this. I love running the Helix direct in the FOH (and not hauling around 250 lbs of gear) and knowing that the sound I hear at home is what I will get with an FOH that's tuned to the venue and the channel parametric EQ set flat (other than whatever tongue-and-grove EQ the FOH engineer finds necessary to situate my guitar in the overall mix).

 

Thanks!  :)

[mkornell]

I experienced much the same thing this weekend, though it doesn't involve changing parameters on a Minotaur. It is a fairly complex patch giving me two eambient and two driven snapshots. I noticed a clear visual latency when stepping on a snapshot - the effects that were being activated or bypassed changed visual status one at a time albeit very quickly. I could not hear if there was any audio latency as there are heavy delays and verbs with trails.

 

There was a clear burst of noise when going to the drive snapshots, though not the other way. Noticeable when I was playing on my own, but pretty much lost in the mix when the band was playing. I heard it in my IEMs, but did not get any grief from the band or the sound man so couldn't have been bad.

[Digital_Igloo]

It's rare, but this can happen if blocks have notable gain changes between snapshots. Although it seems like everything is happening simultaneously, there may be a few milliseconds between the first snapshot change and the last one. If one change involves a huge boost in gain, and another change drops the gain somewhere else to compensate, you may experience a very short volume spike while everything is moving into place (or a very short drop if those settings appear in the opposite order). Totally normal; it just requires a bit less extreme settings when going from snapshot to snapshot.

[mkornell]

DI, is there any way to control the order in which the parameter changes are applied when going from snapshot to snapshot? If possible, being able to make sure there is a short drop in sound would be preferable to a risking a short spike.

 

It's rare, but this can happen if blocks have notable gain changes between snapshots. Although it seems like everything is happening simultaneously, there may be a few milliseconds between the first snapshot change and the last one. If one change involves a huge boost in gain, and another change drops the gain somewhere else to compensate, you may experience a very short volume spike while everything is moving into place (or a very short drop if those settings appear in the opposite order). Totally normal; it just requires a bit less extreme settings when going from snapshot to snapshot.

[Digital_Igloo]

DI, is there any way to control the order in which the parameter changes are applied when going from snapshot to snapshot? If possible, being able to make sure there is a short drop in sound would be preferable to a risking a short spike.

 

I don't think so. It might be related to the order in which blocks are added to a preset. Or where they are in the signal flow. It's not arbitrary, but I wouldn't expect a consistent procedure.

 

It's definitely something our QA guys saw, but that's because they were trying to break snapshots with parameter changes that are considered way outside the norm. FWIW, this is something that happens with all products with scenes/snapshots. It even happens with iPods/iPhones with SoundCheck enabled.

[mikeolivercgp]

DI, thanks for the additional information!

[ThreeDogNate]

Swapping amps has proven to be the better way to do large gain / volume changes in amp blocks too for me. If I have a clean setting for say plexi with the gain down and channel vol up... switch to a higher gain snapshot with the same amp (Higher gain / Lower Channel Vol) it's fine  but going back to the clean snapshot it appears that the channel vol change happens either before or so suddenly that there's a few milliseconds of a volume jump... that little volume blip is keeping me from doing anything but subtle changes in things that can affect volume... seems to be consistent with your OD Stomp situation.

[HonestOpinion]

It seems like at some point it might be worthwhile for the Line6 developers to refine the hierarchy in their code for which parameters get changed first in a snapshot switch to minimize perceptible dropouts, volume changes, or other latency 'artifacts'. There might even be a way to have a kind of 'smoothing' algorithm, sort of like a cross-fade, assuming they have not done this already and that people would not then complain that the change was not fast enough.