Editing question, mixing parallel effects

[JWDubois]

Spent the weekend with my new Ultra. I'm impressed beyond my expectations. I have been poking around with the editing interface, and other than some minor gripes with the interface and layout that I won't get into here, it seems pretty straightforward.

However, I did run across one situation that I could not figure out at all. I ran across one patch that had configured the reverb and delay blocks in parallel, in the column just behind the cab block. Reverb was in Row1, delay Row2, the amp and cab were on Row3. The output from the cab block drove the input of the reverb and delay blocks. Reverb and delay blocks were set 100% wet internally. I understand this.

Now, if I had done this patch, I would have run shunts from the outputs of the cab, reverb, and delay straight back to the output mixer, on their individual rows, so I could independently mix the three effects in one place.

However, in this patch, the outputs from the reverb and delay blocks were routed back down to a shunt on the cab row, on the column just behind the reverb and delay.

So my dumb question is: where is the mixer for the shunt that combined all those blocks back together in a single row? I figured there must be a mixer in the shunt to combine the L+R from three rows into one, but I sure couldn't find it.

JWW

 

[Mike Snider]

there is not a mixer in the shunt it mixes them 1:1.
you could put a mix block in if you would like, but not necessary. The shunt will simply sum the input.

 

[Spellbound]

As far as I understand this parallel thing you don't need a mixer. It is mixed with the level/volume parameter of each effect. If you want less delay just turn the delay level down and the other way round for more delay. Same for reverb here. So it is giving more or less effect signal to the original signal.

As far as I know a block in the routing matrix has 4 inputs and if they are connected they are combined and will be sent to all connected outputs. The amount of each input signal is the level that it has.

 

[Matt_B]

I'm in the process of updating all my patches to use parallel effects chains I haven't found a real need for the Mixer block. As far as I can tell, it just makes controlling levels a bit easier because all the level controls are in one block instead of 3 or 4. I might use it though if the channels in the Mixer block used dBs instead of abstract values.

 

[JWDubois]

I think I get it now. Thanks.

I still think it would be more straightforward (to my way of thinking) to shunt the individual rows all the way back to the output mixer or use a mixer block as a summer. Otherwise you have to go back and forth between the cab, verb, and delay blocks to tweak the mix.

JWW

 

[TimmyM]

I think I get it now. Thanks.

I still think it would be more straightforward (to my way of thinking) to shunt the individual rows all the way back to the output mixer or use a mixer block as a summer. Otherwise you have to go back and forth between the cab, verb, and delay blocks to tweak the mix.

JWW

While this may be more straightforward if you are using only a couple of parallel effects, its when you start running out of CPU because you are using more shunts than needed, that you start trying the method suggested whereby it is all summed to one row of shunts and the level parameters of each effect are used to mix them in.

TimmyM

 

[JWDubois]

You loose CPU by adding shunts? If that's the case, then you've got a point.

I haven't asked this yet, or bothered to look it up, so here goes ....

Is there a sonic advantage to parallel vs series FX routing? How much difference is there between adding reverb on a parallel row and summing it with the dry signal or doing the mix inside a series reverb block, for example?

JWW

 

[Bakerman]

A shunt uses about 0.2% CPU. If you want to use fewer shunts and keep the output mixer functional for parallel signals, move the parallel blocks to the end of the layout and use one row of shunts before them.

Is there a sonic advantage to parallel vs series FX routing? How much difference is there between adding reverb on a parallel row and summing it with the dry signal or doing the mix inside a series reverb block, for example?

No difference in sound, just in level/mix knob settings required to get the same level/mix.

 

[JWDubois]

Actually I thought of one reason to go parallel. With delay and reverb in parallel with the dry path you can avoid having to either add reverb to the delay sound or delay to the reverb sound. Maybe the overall sound would be a little more clear this way.

I can't see loosing a little CPU to shunts being that big of a deal with the Ultra, none of my patches is going to be all that loaded up anyway. But I understand your suggestion of moving the parallel blocks near the output mixer to minimize the amount of parallel shunts.

Thanx,
JWW

 

[Dpoirier]

Actually I thought of one reason to go parallel. With delay and reverb in parallel with the dry path you can avoid having to either add reverb to the delay sound or delay to the reverb sound. Maybe the overall sound would be a little more clear this way.
...
Thanx,
JWW

Actually, another advantage is greater mix flexibility when you use the effect in parallel, especially reverb. When you use a reverb block in series, I find that anything above 50% mix (when shooting for dripping wet ambiant sounds) is not as useful as a parallel reverb where you adjust the perceived "mix" by adjusting the level. The parallel routing leaves your dry signal completely intact, whereas the serial reverb with a high mix value does not. I don't have that problem with the delay block, but certainly the reverb block allows great effects when used in parallel. Both approaches have their usefulness.

Daniel

 

[Bakerman]

Serial reverb at 50% mix and +6 dB will give the same dry & reverb level as 100% mix & 0 dB in parallel with a shunt. You can use the reverb input gain control to adjust reverb level without changing dry level.