Pros and Cons of Send vs Insert?

[Mister Gibson]

It's always worthwhile to experiment for yourself and use your ears!

+1

[mholloway]

Serial FX processing (insert) and Parallel FX processing (send) are not necessarily (nor usually) the same thing. The results of putting a reverb as an insert are wildely different then sending to that same reverb on a send. Why? Because when you send (assuming you are Post-Fader sending) the volume of your original channel stays exactly the same, and the reverb gets added to it at the Master Bus level, from the send. Crank up the send, you crank up the verb -- but your original channel volume level remains exactly the same. Now lets say you go the Insert route. Must plugins process via dry/wet ratio. That means as you crank up the verb level, you are reducing the dry level. Your original signal is effectively getting quieter as your FX level is getting louder. Compare that to the previous description of sending, and you'll see the huge difference between Serial / Parallel processing.

It should be noted that some dry/wet knobs play games with how the ratios work and actually make the 50/50 spot = 100/100, in which case it's acting more like a parallel processing knob. But generally speaking, they are a simple dry/wet ratio. If you dropped a verb on your kick drum channel as an insert and cranked up the Wet, your kick is going to effectively disappear into the verb. If you just Send to a verb from that kick, your kick will remain solid (no loss of level) while adding in the verb from the send (added, technically, to the master mix, not back to the original channel).

All this assumes we are talking Post-Fader on any Sends. When you switch to Pre-Fader you can actually replicate the other half of what the dry/wet knob does -- e.g. dropping the volume of the original channel while increasing the amount of added effected.

I'm not trying to makes things overly complex here, I just think it's funny how many discussions of Insert Vs. Send pop up across audio forums and virtually nobody discusses the inherent differences between Serial and Parallel processing. To me, this is the most key factor in deciding whether I want to use an Insert or a Send. It's not so simple as just "Sends exist so you can have multiple instruments going to one effect!" though certainly that's a part of it.

-M

[btw]

The results of putting a reverb as an insert are wildely different then sending to that same reverb on a send. Why? Because when you send (assuming you are Post-Fader sending) the volume of your original channel stays exactly the same, and the reverb gets added to it at the Master Bus level, from the send. Crank up the send, you crank up the verb -- but your original channel volume level remains exactly the same. Now lets say you go the Insert route. Must plugins process via dry/wet ratio. That means as you crank up the verb level, you are reducing the dry level. Your original signal is effectively getting quieter as your FX level is getting louder. Compare that to the previous description of sending, and you'll see the huge difference between Serial / Parallel processing.

That's definitely something to consider vis a vis mixing workflow, and I think it's a great thing to point out. I would guess the reason people might often gloss over that point in discussions of insert vs send is that presumably, during mixing, one ought to be using one's ears to adjust relative levels, whether it be by dry/wet balance and channel fader, or by pre/post-fader send and channel fader. It's purely a question of workflow; there is no combination of dry and wet signal that cannot be achieved with either serial or parallel processing, right?

[aMUSEd]

Serial FX processing (insert) and Parallel FX processing (send) are not necessarily (nor usually) the same thing. The results of putting a reverb as an insert are wildely different then sending to that same reverb on a send. Why? Because when you send (assuming you are Post-Fader sending) the volume of your original channel stays exactly the same, and the reverb gets added to it at the Master Bus level, from the send. Crank up the send, you crank up the verb -- but your original channel volume level remains exactly the same. Now lets say you go the Insert route. Must plugins process via dry/wet ratio. That means as you crank up the verb level, you are reducing the dry level. Your original signal is effectively getting quieter as your FX level is getting louder. Compare that to the previous description of sending, and you'll see the huge difference between Serial / Parallel processing.

It should be noted that some dry/wet knobs play games with how the ratios work and actually make the 50/50 spot = 100/100, in which case it's acting more like a parallel processing knob. But generally speaking, they are a simple dry/wet ratio. If you dropped a verb on your kick drum channel as an insert and cranked up the Wet, your kick is going to effectively disappear into the verb. If you just Send to a verb from that kick, your kick will remain solid (no loss of level) while adding in the verb from the send (added, technically, to the master mix, not back to the original channel).

All this assumes we are talking Post-Fader on any Sends. When you switch to Pre-Fader you can actually replicate the other half of what the dry/wet knob does -- e.g. dropping the volume of the original channel while increasing the amount of added effected.

I'm not trying to makes things overly complex here, I just think it's funny how many discussions of Insert Vs. Send pop up across audio forums and virtually nobody discusses the inherent differences between Serial and Parallel processing. To me, this is the most key factor in deciding whether I want to use an Insert or a Send. It's not so simple as just "Sends exist so you can have multiple instruments going to one effect!" though certainly that's a part of it.

-M

Thanks - I learned something here.

[shonoob]

Yeah, thanks for the responses so far guys, lots of great info.

[ghettosynth]

In my opinion (and experience) this totally depends on the situation.

Usually:

Is the effect supposed to affect only one track in the song: Insert

Is the effect supposed to affect multiple tracks in the song: Send

So to me its not a matter of 'Pro and Con', but simply of 'What do i need here to achieve what i want to achieve'.

Yes, and, I agree with some of the previous posters. I think that the debate largely only matters when you are in the latter case, that the effect is supposed to affect multiple tracks. At that point we can ask, is there an advantage to using inserts over sends. In the former case, then there's no point to the computer power discussion at all really. If an effect is only affecting one track, then there's only one instance, so whether you put it on an insert or on a send is driven by other concerns.

Ok, so in the latter case, the debate centers around using multiple instances of a single effect as inserts vs using a single instance as a send.

The advantages of sends have been discussed. I generally use reverbs here because when I want to change the ambience of a piece I don't want to change the preset or plugin on every track. On the other hand, using the same reverb as inserts on multiple tracks allows you to slightly tailor the reverb for each track beyond just the send level.

For example, here's an article that discusses using three reverbs to place things in the soundstage:

http://www.musicradar.com/tuition/tech/ ... ace-604097

Sometimes having that level of control over individual tracks is important and putting the reverbs on sends just complicates things. While this seems like it's a reduction to the first case, it really isn't. For ambient stuff, I'll often not know in advance how I want to place things and it's often more convenient to just put reverbs on an insert so that I have the easy option of tailoring each channel. The practical end result is that sometimes it would have been more "efficient" to use a send, but that workflow improvement is worth the sacrifice.

Other time based effects being used "as effects", especially delay, I often prefer to use as inserts, even if they are going to mostly use the same patch/plugin. Largely this is for similar reasons as above. But it's far more common that I don't know whether I want an entire buss affected by a delay, or just certain tracks, or exactly what settings I want to use.

In short, it's a tradeoff of convenience vs flexibility. Sends make it easy to change the plugin or preset for the entire piece, while inserts make it easy to tailor the effect individually, at the cost of convenience.

Finally, there is some subtlety in having many tracks mix within an effect vs post effect mixing that may drive a preference for using send effects as inserts, or not, e.g., the aforementioned "glue", or absence thereof.

[ImNotDedYet]

I still say the advantage of using as a send is processing the reverb/delay/effect separate from the track's sound itself. Pan location, EQ, (look up the Abbey Road reverb trick to see how EQing the reverb different can bring out a deeper sound) compression or sidechaining of the effect.

[mholloway]

I still say the advantage of using as a send is processing the reverb/delay/effect separate from the track's sound itself. Pan location, EQ, (look up the Abbey Road reverb trick to see how EQing the reverb different can bring out a deeper sound) compression or sidechaining of the effect.

Yeah, that's an excellent point as well. EQ can be crucial to get big verbs sounding right.

[Fleer]

Thanx shonoob for starting this thread

[highkoo]

I guess I only necessarily use an insert if the effect causes latency.
A send is just more flexible.
If I want it to behave like an insert, I just stop sending the channel to the master.

Great stuff in this thread.

I still say the advantage of using as a send is processing the reverb/delay/effect separate from the track's sound itself. Pan location, EQ, (look up the Abbey Road reverb trick to see how EQing the reverb different can bring out a deeper sound) compression or sidechaining of the effect.

This is pretty much the reason why I just default to using a send.
I just take for granted being able to mix wet/dry, and treat the wet, on everything.

[rustman]

I think one exception that I've used with reverbs is to place a short, ambience type reverb on an insert to put a track into something like a small, live room. Later, that track might get a send to help put the overall song mix into the same environment (whatever it might be). I definitely lean towards inserts, but that's only because I haven't spent enough time yet focusing on this topic in my mixing journey. My views will always be evolving.

[prene]

Like posted before.

Insert when I want that ONE track affected.

Send when I want more than one track with that same effect.

[pixel85]

sends vs inserts topic is totally pointless. It's the same like car vs bike comparison or studio monitors vs Hi Fi loudspeakers.
Both have different purpose. There's no disadvantages if you use them properly

[Meffy]

sends vs inserts topic is totally pointless. It's the same like car vs bike comparison or studio monitors vs Hi Fi loudspeakers.
Both have different purpose. There's no disadvantages if you use them properly

I don't agree that it's pointless. Not everyone knows how to use them properly, or why you choose one or the other. mholloway's post contains plenty of information telling how and why. Pretty pointful if you ask me.

Now, if you said the way the topic's subject line was worded is off target, I'd agree there.

[shonoob]

Thanx shonoob for starting this thread

lol, thank you mate.