How to get samples/synths to "melt" into the mix.

[MeteoXavier]

Howdy. I've been trying to figure this out for years and to date, no one has given me a straight answer on it, so I want to try here.

One thing I'm trying to learn to do as I get better at mixing and producing my tracks is getting the sounds I use to sound more naturally occuring - in that it seems to me like all my sounds have too much punch in the mix and often do not leave the soundscape in the natural-sounding way. I thought, for a long time, the answer was a Transient Designer, but having gotten the one from Komplete 8, I can see that is not the answer to it. It only seems to work on drums anyway.

Everywhere I listen, even to the old Super Nintendo tracks, it seems to me like there is this microscopic amount of added attack and release on the transients that makes it so that each PCM sample comes in and leaves without sounding too much like an obvious sample. It seems like everyone knows how to do it but me - what am I missing?

In the years I've been trying to figure it out, I've heard the answer is in:
* Compression
* Reverb
* EQ
* Transient VSTs

And so far I haven't found it.

Can anyone help me out?

[ducter]

Everyone else seems to be trying to get their tracks "punchier".

How about doing quick automation fade-ins?

Cut the attacks like a guitarist does with a volume pedal.

[Perimeter Sound]

it seems to me like there is this microscopic amount of added attack and release on the transients that makes it so that each PCM sample comes in and leaves without sounding too much like an obvious sample.

If you are working with synths, then there you go, adjust the ADSR on the patch until it's smoother.

Most hosts have fade in/fade out for -samples- as well, although I've not had a ton of luck that way myself at least with percussive sounds (not sure if you mean all sounds, or just certain kinds). Perhaps some e.q. might help, to remove some of the (perceived) punch or aggression from certain sounds, to mellow them out, and then they might just naturally fit in the mix better/smoother.

[scavenger8]

Try Mda Envelope or Fruity Envelope Controller.

[Trakstar]

Schwa offer plugins that insert on your track of choice and you get envelopes which you can draw around every audio object to shape and mute and do what ever you like with. It was in answer to the very question you asked that the plugin was made.

http://www.stillwellaudio.com/?page_id=36

http://www.stillwellaudio.com/?page_id=35

Dyno and sculpto, one draws the wave form for the entire track and allows you to edit at any place along the waveform drawn out in the editor, while the other works as a normal plug in type dynamics processor. I would check them out if you are serious about what you are saying

[MeteoXavier]

It's not the envelopes. I've already done that and the results were ridiculous. I definitely believe it's in the transients themselves because it's in how the sounds are processed the moment they hit and when they leave, but it's such a microscopic amount that the ADSR can't touch them without sounding unrealistic and weird.

It's something in the whole mix or in the mastering process or something, and clearly most professional and even non-professional people know how to use it, and yet still no one can give me a straight answer on what it could be.*

I'll post some examples when I get home.

* - Sorry if that came out ungrateful and obnoxious, that wasn't my intent, but yeah I've been trying to solve this for years now and its pretty frustrating because I need to add that to list above now. I'm extremely sure it's not the envelopes.

[Trakstar]

the fx work specifically on transients and utilize what ever curve shapes you need to treat the dynamic structure, especially sculpto, otherwise I dont see what the point is unless you use a noise gate

[scavenger8]

I see; You want to change The character in its first few milliseconds. There's all kinds of debates how this effects the sound. so i don't know how come you can be so sure Transient shaping is the answer. Ever consider it could be an "Audio illusion" similar to an Optical illusion like the "Figured ground Principal". Swing is another thing knowbody knows what it really is but has a huge impact on the perception of what you hear in a number of ways. Stuff that human eyes and ears are'nt fast enough to distinguish,so the mind switches makes things up. There's all kinds of tricks painters use to fake deapth,space,etc. Why not the same in the audio world. I'd experiment around with swing. Also Google -Figured Ground Principle- super easy concept to graspyou'll see you've known this all along.

[Eric C.]

Try cutting the attack and or using a high cut EQ on some tracks.

[VibraSound]

Mix is 80% of the job.

Mastering is 20% of the job.

[Sendy]

Honestly, when I started out I really underestimated the power of amp envelopes. Take for example the attack, the difference between an instant attack and a slight bit of attack, and when to use each or go even longer into pad territory, is one of the last things I learned and what seperates a lot of beginner-ish programming, IMO.

Then there's mixing and mastering as mentioned earlier, and using delays and reverb to pull certain parts "back into" the sound stage.

[Tricky-Loops]

One thing I miss in your list is Filter and Filter Modulation (with Filter Envelopes). Maybe this could help to fade the samples in...

[JoeCat]

I'm no expert (for sure), but assuming my two cents can't hurt (and I think I know what you're describing):

Instead of focusing on the envelopes/compression/standard EQ or even transient designers, try a tape emulator, putlec EQ emulator, valve sat, etc.

The reasoning is this: A lot of the punch/attack is often distributed across the range of high/low frequencies (for natural sounds, think of the pluck on a guitar, etc.). Those analog-style EQs and tools like Vintage Warmer, etc. often do a great job of bringing clarity to the body of the sound. It's not so much that the transient attacks are overtaking the mix, it's that when you dial it back you lose the body, and standard compression across the entire frequency spectrum doesn't always produce the desired results (parallel/multi-band can help, but that's more of a black art to me).

I actually love the sound of a lot of digital synths and drums. I bought the Nomad pack when it was super-cheap (I think $100 for Magnetic I and II, all the British "stuff", Pultec EQs, etc.). I'm not great at figuring out the Pultecs for example, but often will just boost some of the nice mids (and attenuate the odd resonances, which are often found in the attacks), then dial it all back - key seems to be to use just enough so you can barely perceive the difference. Combine all those tracks in the mix, however, and you get a bit of the affect of analog channel strips (of course, if you can afford the UAD stuff, etc., you can play "Neve board"), and (too me at least), more natural punch with body and less of that annoying "drums falling down stairs" mix.

Anyway, can't hurt to try, especially with a demo of PSPWarmer or some such.

[tranel]

Mix is 80% of the job.

Mastering is 20% of the job.

Don't mean to sound like a wiseass but I'd make that about 90 - 10. Or even 95 - 5.


If I understand this correctly, the OP wants sounds to blend in naturally in the mix without being too dominant yet still be clearly audible and have impact. Is that right?

If so, I don't think it's about transients. It's about:

Choosing your sounds extremely well ie contrasting sounds well depending on the impact you want.

Good use of EQ, particularly mid-side EQ can work wonders.

Thinking carefully about what sonic element should "drive" the music at any given time, ie what you'd want the listener to focus on at any given time, and automating volume envelopes of other elements around it to make sure it stands out.


Among other things.

Again, I'm not sure if I understood the OP's post properly. It'd be helpful to have sound file as a reference.

Best,

Anders

[deastman]

I agree with Anders. Also, be sure to carve out space in the frequency spectrum and stereo field for each sound.

I'm curious to hear some examples of this problem. If you've been struggling with it for years, and yet every random non-pro on KVR has already solved it, you must be doing something terribly wrong. Either that, or its all in your head.