Clipping issue if input 2 is not connected

[Rex]

But then there is no difference to the current way it works. You just save the SND and RTN block for something else.

The difference is that SND and RTN aren't used. You take away the feedback loop, which takes away the problem.

The only thing that makes the RTN -> FXL -> SND thing dangerous is that the FXL block works as a shunt as soon as no cable is connected.

There are two things that make that configuration dangerous: the automatic FXL shunt and the SND-RTN loop. Take away either one, and the danger goes away.

 

[Zwiebelchen]

There are two things that make that configuration dangerous: the automatic FXL shunt and the SND-RTN loop. Take away either one, and the danger goes away.

Not if the FXIN and FXOUT will still have the shorting behaviour.

lets say you have a seperate FXIN and FXOUT block:

[FXIN]--- [ stuff ] ---[FXOUT]

As you can see the actual problem remains:
If the FXOUT is connected to a speaker and the FXIN is not plugged in and the OUT is shorted to the IN because of that, you still have a feedback loop (The noise floor will be amplified to clipping value). This is not a solution to the problem.
The only solution to this problem is getting rid of the shorting behaviour (which is not common sense to begin with).

 

[GiRa]

If you take a delay and you loop itself at 100%, 0dB of course, you will have the same problem. It's an user responsibility!
EDIT: [SND]---[RTN] is another example.


You could post a wishlist item asking for a selectable no-input-2 behaviour.

 

[Rex]

lets say you have a seperate FXIN and FXOUT block:

[FXIN]--- [ stuff ] ---[FXOUT]

As you can see the actual problem remains:
If the FXOUT is connected to a speaker and the FXIN is not plugged in and the OUT is shorted to the IN because of that, you still have a feedback loop (The noise floor will be amplified to clipping value).

Sorry, I'm not seeing the feedback loop in what you've described. Without SND/RTN in the picture, no feedback is happening.

The only solution to this problem is getting rid of the shorting behaviour (which is not common sense to begin with).

If you have separate In2 and Out2 blocks, there will be no shorting behavior—no routing at all, in fact—unless the user explicitly puts it there.

 

[Zwiebelchen]

Sorry, I'm not seeing the feedback loop in what you've described. Without SND/RTN in the picture, no feedback is happening.

There isn't a feedback loop in the diagram I posted either. It just happens because the FXL shorts itself as soon as no input cable is detected.
If the same behaviour would be implemented to your seperate IN and OUT blocks, it will be exactly the same. The problem is not SND and RTN, the problem is the shorting behaviour of FXL, no matter wether the FXL block would be splitted or not.

If you have separate In2 and Out2 blocks, there will be no shorting behavior—no routing at all, in fact—unless the user explicitly puts it there.

They could also just remove the shorting behaviour from FXL. In the end it all boils down to the same thing: the shorting behaviour is both dangerous but could also be beneficial, so it should have a switch.

If you take a delay and you loop itself at 100%, 0dB of course, you will have the same problem. It's an user responsibility!

That's a different issue, because it will create feedback due to logical reasons that can be avoided 100% of the time by doing the setup right.
The situation I described is different, because it can happen when an accident happens and a cable is pulled out during use.

 

[Rex]

There isn't a feedback loop in the diagram I posted either.

There is, though. You're using the Feedback SND and Feedback RTN blocks. That's a feedback loop.

It just happens because the FXL shorts itself as soon as no input cable is detected.

It happens because two conditions occur at the same time:

1. The FXL shorts itself as soon as no cable is detected.
2. There's a feedback loop feeding signal from the FXL output to its own input.

Take away any one of those conditions, and you break the feedback loop.

If the same behaviour would be implemented to your seperate IN and OUT blocks, it will be exactly the same.

It wouldn't be the same, because there would be no need for Feedback SND or Feedback RTN—there would be no feedback loop. And there would be no short unless you put one there, because In2 and Out2 would be separate blocks.

 

[Zwiebelchen]

It wouldn't be the same, because there would be no need for Feedback SND or Feedback RTN—there would be no feedback loop. And there would be no short unless you put one there, because In2 and Out2 would be separate blocks.

Whaa, I'm really getting desperate; why won't you just understand what I mean?

There's no need to split up the FXL block into two when the shorting behaviour is gone, as you already said, because then one of the two conditions are gone.

You somehow assume that splitting up the FXL block will also get rid of the shorting behaviour. I'm curious, why would you think that? The shorting behaviour was obviously implemented on purpose, because when using the FXL block as a real effects loop like in real amplifiers, it's actually a desireable thing.

Look, I don't care if they split up the FXL block into two blocks or not. Actually, I'd love to see that because it improves readability of presets a lot. But the real deal is the shorting behaviour. Even if they split up the FXL block into two and the shorting behaviour is still there, the problem will still be present:

[IN] --- [stuff] ----[OUT]

-> no cable connected to IN
-> OUT routes itself back to IN
-> infinite loop

So no matter if the FXL block is split up into two or not: as long as the OUT shorts itself to the IN, theres always going to be a feedback loop, no matter if you use Feedback SND and RTN or not.

 

[Rex]

Whaa, I'm really getting desperate; why won't you just understand what I mean?

I'm trying to, my brother.

There's no need to split up the FXL block into two when the shorting behaviour is gone, as you already said, because then one of the two conditions are gone.

Agreed. If the goal is strictly to eliminate the feedback loop, then eliminating the shorting behavior would be enough. The idea is that with separate ins and outs, you get increased versatility. The fact that you wouldn't need to use the SND/RTN blocks or risk a feedback loop is an added bonus.

You somehow assume that splitting up the FXL block will also get rid of the shorting behaviour. I'm curious, why would you think that?

Because they'd be separate blocks. There's no connection between separate blocks unless the user connects them.

Look, I don't care if they split up the FXL block into two blocks or not. Actually, I'd love to see that because it improves readability of presets a lot. But the real deal is the shorting behaviour. Even if they split up the FXL block into two and the shorting behaviour is still there, the problem will still be present:

[IN] --- [stuff] ----[OUT]

-> no cable connected to IN
-> OUT routes itself back to IN
-> infinite loop

So no matter if the FXL block is split up into two or not: as long as the OUT shorts itself to the IN, theres always going to be a feedback loop, no matter if you use Feedback SND and RTN or not.

I see your point. But there'd be no shorting behavior unless the user explicitly connected it that way.

 

[Bakerman]

Because they'd be separate blocks. There's no connection between separate blocks unless the user connects them.

I guess he's just saying that if the Out2 & In2 blocks were coded to act like feedback send & return blocks w/ no cable at In 2L, you'd have the same issue as with the current FXL block. So for this way of using them, ideally they wouldn't be coded that way, which is what you're assuming already.

 

[Rex]

Not an assumption, but a consequence of my original proposal. Having two blocks with an implicit but invisible connection would be a one-of-a-kind deviation from the norm for the Axe.

 

[Zwiebelchen]

Not an assumption, but a consequence of my original proposal. Having two blocks with an implicit but invisible connection would be a one-of-a-kind deviation from the norm for the Axe.

Yeah, but we gotta assume that the shorting behaviour is there on purpose.
The reason for that is, that when using the FXL block in the way a real effects loop in real amps works, then it's a desireable thing that send shorts to return.

-> You are able to use the send to feed other devices while still having a signal going to the cab.
-> The rig automaticly adapts itself to wether the effects loop is used or not - so if you use an external delay on one gig, but not on the other, the rig doesn't need to be changed

Having seperate IN and OUT blocks will not allow for this. If you connect OUT to IN manually with shunts and have a mix parameter on both, you'd have an effects loop in parallel, which is not exactly the behaviour of loop shorting. Of course you could design OUT and IN so that the shunt connecting both in that situation would work different, but it would be very counter-intuitive.
I can see why they put both blocks together, for that reason.

 

[Rex]

I don't see the value of the shorting behavior; it sort of works for some situations, sort of doesn't work for other situations.

-> You are able to use the send to feed other devices while still having a signal going to the cab.

You could still do that if In2 and Out2 were separate blocks. No shorting needed. Just drop in Out2 before the Cab block.

-> The rig automaticly adapts itself to wether the effects loop is used or not - so if you use an external delay on one gig, but not on the other, the rig doesn't need to be changed

If I unplug an external delay, I don't want it to be the same as jumpering the loop. Particularly if I'm running that external delay in parallel.

->Of course you could design OUT and IN so that the shunt connecting both in that situation would work different, but it would be very counter-intuitive.

Agreed.

 

[Zwiebelchen]

I don't see the value of the shorting behavior; it sort of works for some situations, sort of doesn't work for other situations.

I don't either, but some people might, like GiRa:

The actual behavior of the FX Loop block is the more flexible, changing it will affect other users.

... though I can't really imagine a situation where you just couldn't short the send and return by yourself...

You could still do that if In2 and Out2 were separate blocks. No shorting needed. Just drop in Out2 before the Cab block.

Know that. I was talking about real amplifiers.

 

[Rex]

I don't either... though I can't really imagine a situation where you just couldn't short the send and return by yourself...

Yeah, that's what I was thinking.

Know that. I was talking about real amplifiers.

Then I'm not getting what you meant about sending a feed to other devices while still having a signal going to the cab.

 

[Zwiebelchen]

Then I'm not getting what you meant about sending a feed to other devices while still having a signal going to the cab.

A lot of guitar combos only have the effects send output in order to feed other devices like onstage monitors or effect processors.
In this case, the effects loop will usually short itself so to that you don't need to put a 'redundant' cable back to the effects return.

 

[Rex]

A lot of guitar combos only have the effects send output in order to feed other devices like onstage monitors or effect processors.
In this case, the effects loop will usually short itself so to that you don't need to put a 'redundant' cable back to the effects return.

Now I get what you're saying. No problem, though, if you've split the FXL into separate Out1 and Out2. All you'd need is a parameter in the Input2 block that, when you switch it on, will sense the presense of the cable (that functionality is already available in the Axe). If there's no plug, it will accept input from the block's grid input. Then you can wire up the shorting behavior on the grid, with no feedback loop.