TD vs Compressor

Nola

Well-known member
Hey guys, I'm trying to understand when to use a Transient designer. Since they have no threshold settings, what exactly are they doing? Say you put one on a snare and reduce -3db. Does the TD know to only effect the transient of the snare by 3db? Does it release as soon as the attack has passed?

Also, can't you setup a compressor to do that exact same thing? So is a TD just a quicker/simpler compressor?

Just trying to figure out why one would ever use a TD since it seems to do the same as a compressor yet has less control.
 
Last edited:
It is my understanding that TD will affect the transient of something regardless of the signal level. That is why there is no threshold and the big difference between it and a compressor.

It can also add transient level without reducing the level of the sustain. Another difference between it and a compressor.
 
It is my understanding that TD will affect the transient of something regardless of the signal level. That is why there is no threshold and the big difference between it and a compressor.

Cool, thanks. Is there a practical, real-world example where you would use a TD over a compressor? I learn best by that method and just trying to understand via example.

It can also add transient level without reducing the level of the sustain. Another difference between it and a compressor.

Maybe I answered my own question...
So would a use for this be say, you want more attack on a snare but without the compressor's release affecting the sustain?

What about the reverse. Say you have a loud pinging snare and want to soften the transient. Would a TD be better in that case even for reducing since you can keep the sustain?
 
Maybe I answered my own question...
So would a use for this be say, you want more attack on a snare but without the compressor's release affecting the sustain?

What about the reverse. Say you have a loud pinging snare and want to soften the transient. Would a TD be better in that case even for reducing since you can keep the sustain?
Yes both. Not being tied to -level- makes for a very cool tool option.
I use UAD's SPL emulation, like most has +/- for attack and sustain.
There's some do have release and some other features I believe. (Sounds interesting, haven't had the desire to go there :>)
Guitars (acoustic, clean electric) bass / kicks -When 'tightening the kick/bass relationship is of sustain/time overlap more than freq response. This goes for various rhythm instment/mix combinations -guitars again.. Drum bus sometimes, snares. What else?
 
Cool, thanks Mixsit. Do you know how it actually works? Like the mechanics/computer programming behind it. Does it have an algorithm that isolates the front (transient) from the back (sustain) and kind of splits it there, then allowing you to tweak each separately?

Are TDs only digital tools, or did guys have them back in the hey day of analog?
 
Are you on Reaper? If you really want to "know" you could drop the JS Transient Controller on a track and click the Edit button. ;)

I have questions about what I'm seeing there, but best I can tell it derives three different rectified and integrated envelopes from the loudest sample between the left and right at any given time. One envelope is really fast, one is just kinda fast and acts I think as a baseline, and the other is pretty pretty slow. It then applies gain on a sample to sample basis based on the ratios of those envelope levels. If the faster one is bigger than the baseline, then the attack percentage tends to dominate. If the slower one is bigger, then it will dominate. If they're both bigger it explodes! But not really. :)

Basically it's looking at how far*fast the signal changes and turns it up or down based on that. If it's going really far really fast - changing a whole lot all of a sudden - then it's a transient. Otherwise it's sustain.

Edit - except I said integrated above and now that I think about it, it's more like it's differentiated. It looks like a low pass filter, but there's a minus sign where there should be a plus. This is basically finding the slope of the waveform over a period of time, which is the real question.
 
Back
Top