Serg Halen
New Member
Actually.. This is pretty ez, just put compressor on master and turn down threshold, then knob gain turn up. After compressor put limiter. Profit.
A little example
Last edited:
Reducing Transients Transparently
- Thread starter Daniel Petras
- Start date
Actually.. This is pretty ez, just put compressor on master and turn down threshold, then knob gain turn up. After compressor put limiter. Profit.
A little example
Sounds like some peeps here would really benefit from playing with a compressor/limiter on both mixes and individual instruments. Getting that into the ears is one of the most important parts of mixing.
Some of the applications:
Limiting (10:1 input:output ratio or higher) frees you from setting the level of the whole mix relative to the highest transients. You can also use a limiter to make percussive sounds crunchy, which works really well in a parallel compression setup as Kenk suggested - the original sound has the dynamics, but you mix in the hard attacks.
Bus compression - smoothes out the whole mix. Narrowing the dynamic range can also let you push up the bottom, which increases the density. Usually slow attack and release.
Individual instrument compression can get a smooth, even sound. It can let you shave off the attack, for example Paul McCartney's bass sound on a lot of songs. Or you can exaggerate the attack (fast release).
And then you get to multiband stuff... but the main point is that buying plug-ins is great, but it's only part of it.
Some of the applications:
Limiting (10:1 input:output ratio or higher) frees you from setting the level of the whole mix relative to the highest transients. You can also use a limiter to make percussive sounds crunchy, which works really well in a parallel compression setup as Kenk suggested - the original sound has the dynamics, but you mix in the hard attacks.
Bus compression - smoothes out the whole mix. Narrowing the dynamic range can also let you push up the bottom, which increases the density. Usually slow attack and release.
Individual instrument compression can get a smooth, even sound. It can let you shave off the attack, for example Paul McCartney's bass sound on a lot of songs. Or you can exaggerate the attack (fast release).
And then you get to multiband stuff... but the main point is that buying plug-ins is great, but it's only part of it.
Lessons from the "make everything loud as possible" branches of EDM would be:
- Multiband compression, working both downwards AND upwards, especially on the low end.
- Multiband soft-clipping, i.e., distorting each band separately in (you can usually crush the mids, low and highs need a lighter touch), then possibly the whole thing together.
- Perhaps most importantly, ducking! AKA side-chain compression. Make the rest of the track duck behind the key percussion elements. Vary the degree of ducking based on the material -- e.g. duck bass and pads a ton, leads just a touch. Not only does this create tons of headroom, but it plays a psychoacoustic trick -- any drum sounds yuuuge when it causes everything else to temporarily get quiet to make room for it.
dgburns
software surfer
So the thread title is reducing transients transparently, and I have yet to find something like the overkillers in software save for a few clipper plugins. Those overkillers are a simple diode circuit but they do wonders on the right material.
Also don't know if it's been mentioned, but the excellent Drum Leveller by Sound Radix is really very good at allowing you to set each drum hit as well. It's on my "to buy" list for sure.
In the picture below, same song, same section, one with overkillers on the output of my Neve8816, one without the overkillers. There's a bit of volume loss, but the transients are contained and it doesn't sound "compressed" in any way. Maybe it's not your thing, but it's a useful tool in the arsenal. And one that wasn't mentioned here till now.
Also don't know if it's been mentioned, but the excellent Drum Leveller by Sound Radix is really very good at allowing you to set each drum hit as well. It's on my "to buy" list for sure.
In the picture below, same song, same section, one with overkillers on the output of my Neve8816, one without the overkillers. There's a bit of volume loss, but the transients are contained and it doesn't sound "compressed" in any way. Maybe it's not your thing, but it's a useful tool in the arsenal. And one that wasn't mentioned here till now.
Attachments
whoot whoot
New Member
I'd suggest watching the FabFilter Videos on Compression Techniques, also the UAD Compression overview videos. There are standard ways of dealing with Transients - Transient shaper plugins - and then there are older more tried methods of dealing with the transients + opening moments of a sound source like the LA2A and 1176 emulations.
There are lots of great videos talking about how to handle, cut and boost tranisents by using compressors on Dave Pensado's youtube channel and the above mentioned.
I guess what I am saying is: it is more of a technique thing to learn than to buy.
There are lots of great videos talking about how to handle, cut and boost tranisents by using compressors on Dave Pensado's youtube channel and the above mentioned.
I guess what I am saying is: it is more of a technique thing to learn than to buy.
Last edited:
karelpsota
send nudes
This track is extremely well engineered. Attila spent almost a year on it. The production and sound selection is 90% of why it works.
Mastering wise, you need to control the low end very well if you want to achieve loudness. You also can't stack too many things. Less is more as usual.
Titan must fall has barely any low end. No bass line, just a glimpse of sub under the hits every bar. The hits sound huge due to harmonic exciters. All the energy comes from tasteful distorsion and good orchestration. The tonal anvil also help to enrich the hits, so it doens't sound like noise.
Try using pro-MB to control your lowend, softclip for a bit distorsion, then push as much as you can into a good limiter. (I believe Attila uses the AOM Limiter)
Last edited:
Nathanael Iversen
Active Member
Cubase has a look-ahead limiter built in - has a handy little graph that shows it pushing down the peaks. If you have Cubase, you don't even have to spend $$...
OP
OP
Daniel Petras
Active Member
Wow a year!?
Yeah, I've learned soo much about low end since I've made this post. And my mixes are constantly getting better and louder as well. A lot of my EQing now simply consists of cutting the low end where it is not needed.
One of the things I still find difficult is crafting really nice trailer hits. I've definitely started to do a lot more processing and parallel stuff, but it still requires really great sounds through the spectrum to get something hard hitting, but with good character.
Edit: Btw which clipper do you recommend? I started using this crispy present on FF Saturn which did really nice things for a hit I was working on. I tried it on this track I did and the hits sounded a lot snappier, though I think there are things I could do to make them bigger (if you have any suggestions).
Last edited:
OP
OP
Daniel Petras
Active Member
I just realized FF Pro - L has that and I've never used it. I'll mostly use a looked ahead if I'm side-chaining something and I want it to get out of the way quickly.
storyteller
Senior Member
For anyone interested in @charlieclouser's technique, the seemingly-never-ending-series-of-waves-sales has L3 complete for under $23 with CK901 discount code (on the plugin sale sites). Supposedly the sale ends today... but, its waves, so who really knows. 
Killer deal though for sure. 
Killer deal though for sure.
karelpsota
send nudes
GClip is really good. I know Joshua Crispin uses it too.
https://www.gvst.co.uk/downloads.htm
Also its free
try pushing single tracks, busses, and masters as much as you can before the distorsion is audible. You'll reduce the crest factor without any pumping or "compressed" effect.
Sometimes, if the master clipper is distorting. Leave the clipper ON and work in the mix to reduce clashing elements. (often comes from hit, bass conflict).
Last edited:
charlieclouser
Senior Member
I have recently figured out some settings on Ozone's Harmonic Saturator module (or whatever it's called) that don't wreck things too much: I use the "tape" model with intensity set to about 5 out of 15 and mix at 50% on all four bands, tweaking band crossover points, mix, and intensity a little as needed. This puts back a bit of grind and sparkle that compensates a little for what is lost by heavy limiting. Many of the other saturation models seem to break up too easily, but I like what the tape model does.
Saturn is also a great processor, but has so many options that it takes a minute to get the hang of it. I've used it on the short strings stem to get a bit of bite and get the shorts to cut through the mix in dense passages.
Saturn is also a great processor, but has so many options that it takes a minute to get the hang of it. I've used it on the short strings stem to get a bit of bite and get the shorts to cut through the mix in dense passages.
OP
OP
Daniel Petras
Active Member
I've recently discovered that people will do 2 stages of mastering where they will have mastering on the on the mixbuss (which If I'm correct, is the final output for all tracks) and then will do mastering on the printed track. Is this something that people often do? I've tried to do my mastering on the mixbuss stage as the final master to save time, but for some reason it has a different effect from rending and then doing the mastering on the printed track. I have not really figured out why.
Here's a quote from what Mick Gordon does:
Mastering: it depends on what we sort out with the Audio Director. Each game is different and games are still trying to figure out their ideal loudness standard, so whatever I do usually has to line up with that. Some games, I've made the action music quite dynamic with a loudness range between 6-12db. Other games, I've brick-walled the music and the volume of the music changes according to the in-game mix engine (so, when SFX play the music gets ducked, etc). Depends on the game. I master it myself - often there's multiple versions that get changed right up until we release the game (and even sometimes afterwards...) so I can't keep sending it out to someone externally. Mastering is really basic. I have two stages: MixBuss: usually a buss compressor into an EQ. I usually use either the Slate Buss Compressors but I'm loving the UAD2500 now. That goes into the two UAD Pultec Plugins, adding 3-5db at 60hz, a broad shelf of 2-3db from 3k upwards. Then, I have the midrange EQ adding about 2 db of 1k, or 1.5k (depending on the track). That's just the overall curve I like for most stuff. Mastering: Depends on the track. It always starts with the Waves Low Band LinEQ shaving off everything below 32hz or so. That could go into any combination of things depending on the track (usually the Shadow Hills, maybe another EQ, etc). It always finishes with a limiter. I audition several limiters each time, trying to find a nice balance of control and dynamics. I don't like limiters that "pump". The DOOM stuff was mostly clipped through a converter for limiting, just cause the music is stupid and aggressive.
Here's a quote from what Mick Gordon does:
Mastering: it depends on what we sort out with the Audio Director. Each game is different and games are still trying to figure out their ideal loudness standard, so whatever I do usually has to line up with that. Some games, I've made the action music quite dynamic with a loudness range between 6-12db. Other games, I've brick-walled the music and the volume of the music changes according to the in-game mix engine (so, when SFX play the music gets ducked, etc). Depends on the game. I master it myself - often there's multiple versions that get changed right up until we release the game (and even sometimes afterwards...) so I can't keep sending it out to someone externally. Mastering is really basic. I have two stages: MixBuss: usually a buss compressor into an EQ. I usually use either the Slate Buss Compressors but I'm loving the UAD2500 now. That goes into the two UAD Pultec Plugins, adding 3-5db at 60hz, a broad shelf of 2-3db from 3k upwards. Then, I have the midrange EQ adding about 2 db of 1k, or 1.5k (depending on the track). That's just the overall curve I like for most stuff. Mastering: Depends on the track. It always starts with the Waves Low Band LinEQ shaving off everything below 32hz or so. That could go into any combination of things depending on the track (usually the Shadow Hills, maybe another EQ, etc). It always finishes with a limiter. I audition several limiters each time, trying to find a nice balance of control and dynamics. I don't like limiters that "pump". The DOOM stuff was mostly clipped through a converter for limiting, just cause the music is stupid and aggressive.
charlieclouser
Senior Member
I do the "mastering twice" thing to some degree - stems delivered to the dub stage have per-stem limiting from Waves L3-LL Multi only, since the composite mix is not usually used on the dub stage, and there will probably be some additional dynamics processing on the final mix from the stage as well.
But when I prepare the composite mixes for release on streaming / CD / vinyl I take another pass with Ozone and other goodies - on my latest score (JIGSAW) this was a +3db hi shelf above 6k, an "infinite slope, zero resonance" hi-pass at 20hz, Ozone's Harmonic Saturator on the "tape model" setting with intensity at 5 out of 15 and mix at 50%, and up to about 3db of "peak shaving" from Ozone's Maximizer with a little transient recovery and the most advanced release algorithm (ARC IV in classic mode I think?). I set the Maximizer ceiling to -.5 db just in case of inter-sample peaks even though it's set to catch those. I automate the threshold of the Maximizer to "chase" the level of the mixes so I can bring up quiet passages and back off on loud stuff - the range might vary by up to 12db across a long cue.
My scores are always mastered by a real mastering engineer for release on CD / streaming / vinyl after I do this, and I always send my Ozone-ed versions as well as the dry mixes that are just a straight-fader composite of the already limited stems as delivered to the stage. On this last score, the mastering engineer really liked my Ozone-ed versions and used them instead of starting over with the dry versions, only adding "a little bit of presence" as he put it.
So I guess home mastering with Ozone passes the smell test.
But when I prepare the composite mixes for release on streaming / CD / vinyl I take another pass with Ozone and other goodies - on my latest score (JIGSAW) this was a +3db hi shelf above 6k, an "infinite slope, zero resonance" hi-pass at 20hz, Ozone's Harmonic Saturator on the "tape model" setting with intensity at 5 out of 15 and mix at 50%, and up to about 3db of "peak shaving" from Ozone's Maximizer with a little transient recovery and the most advanced release algorithm (ARC IV in classic mode I think?). I set the Maximizer ceiling to -.5 db just in case of inter-sample peaks even though it's set to catch those. I automate the threshold of the Maximizer to "chase" the level of the mixes so I can bring up quiet passages and back off on loud stuff - the range might vary by up to 12db across a long cue.
My scores are always mastered by a real mastering engineer for release on CD / streaming / vinyl after I do this, and I always send my Ozone-ed versions as well as the dry mixes that are just a straight-fader composite of the already limited stems as delivered to the stage. On this last score, the mastering engineer really liked my Ozone-ed versions and used them instead of starting over with the dry versions, only adding "a little bit of presence" as he put it.
So I guess home mastering with Ozone passes the smell test.
dgburns
software surfer
Logic's new loudness meter is awesome. You can now get to a definitive LUFS reading, say -14 (or whatever) without having to guess. Also the Level meter now has "True Peak" metering, so you don't have to guess on intersample peaks. The integrated let's you see infinite, or over the whole program. The Level meter comes up as "Peak" so you have to switch it over to "True Peak". (kinda dumb if you ask me)
I've found for hard hitting program material, it's been hard to stop intersample peaks if my brick wall limiters are anything over -1db ceiling. I don't know if -.5 will stop all peaks from happening. And apparently when it goes to mp3, it sounds better if there aren't any intersample peaks. Some mixes I've had to lower the ceiling even lower, like -2db or more even.
Course some like the sound of their converters clipping, so what the hell do I know??......
LaboratoryAudio
New Member
Try using parallel compression. Compressor set to super fast attack and fast release. You need a compressor that can grab the initial transients like The Glue. The mix that in with the "dry" signal. You should be able to bring up the overall volume but with less transients. Hope that makes sense.
