What a fantastic explanation. This video could have been twice as long and I'd still be sad it was over. RUclips abounds with in-DAW tutorials, but what I really crave are these deep explanations of the theory, the why. Thank you so much.
The only podcast I will nerd out to. Every episode is a milestone👌🏼Thank you for posting awesome and useful tips that give perspective to this process.
Thanks for addressing how different sample vs analog/live performance usage can be. A lot of people forget that and just think about techniques for an instrument type not the source origination.
It was the exact thing I needed to watch.I was just thinking and questioning all these to myself this afternoon and now you come and answering them all and more in a great point of view
This is a really excellent video. I still struggle with compression and figuring out exactly what problem I'm even solving in the first place, so being specific with "these settings to deal with transients, these settings for sustain" is way more helpful than the blanket "do what sounds good" advice you hear parroted. Because amateurs don't really know if what they're hearing sounds good, or how to fix it. Really appreciate this, gonna be checking out the rest of your videos for sure.
So glad to be useful! We’ve got tons of free videos on this channel that I hope you enjoy. If you ever want to go super deep on compression, we even have a full length course on it at CompressionBreakthroughs.com Hope that helps! -Justin
If anyone is looking for a cool trick, following the compressor on the SSL channel (making sure the compression is happening after the EQ) with the LA2A is absolutely fantastic on vocals. I've been doing this for quite some time, and it really keeps the vocals right in your face.
OD’n with one compressor will give you the sound of that compressor but at the expense of the performance. Try clip gain, automation, & peak limiting on vocal material. Use your DAW meters as a measurement tool. Try gates, envelope shapers & transient shapers on rhythmic material. Go there first before assuming a compressor is what you need.
just read about this today in mixerman's book "musicians survival guide". He does this on vocals and uses EQ before and after each compressor. He was talking to another engineer and they both agree that analog compressors do much better than the plug-in versions in software packages.
Tasty serial compression! Bound to be a popular ‘take-away’. Serial, cereal, surreal. Even comes with a free ‘triple-entendre’. All jest aside, always grateful for your interesting lessons and this was particularly informative to this nerd 🙏😎
great video, it just makes me wonder how different this could sound with analog vs digital compressors, especially with the examples given of 1176 into LA2A and various plugin versions of the hardware.
Commenting on another older video, Justin... but I was wondering, with stacking compressors, where do you put an EQ? I know the "correct" answer is probably "it depends" - but most people I know in recording/mixing prefer EQing into compression; but I can see reasons to EQ after on the master bus (as long as you don't get crazy with boosting lol). Is there a signal chain you set up just as a matter of course, even if you don't use everything; and is it EQ > compression? Compression > EQ? Compressor 1 > EQ > Compressor 2? Or is it different each time you roll up to your desk?
Ha. Me ten years later i still dont quite know what im doing with compression. I do use a lot of buss compressors and a couple on my master bus. The master is always to glue,warm and gently compress. Im just never quite sure if I’m using them right. Lol.
today i tried stacking compressors on every channel and mix bus. every compressor ssl bus less than 1 db gain reduction and triple stacking the master bus. sounds overkill but worked well.
Loved it! Serial nerd stuff. I had thought your example would be 1176 into 160/165, instead of 1176 into LA2A? I really thought also in reverse that you would leave all the transients for SNR with slow attack into 1176 fast attack using your ears for balancing that instrument. Good explanation.
(1) How do you eliminate phasing with parallel compression? (2) How do you high pass mix bus compression if the compressor doesn't have a high pass setting?
1. You don’t and don’t need to! The reality is that it’s not nearly as big of a “problem” as some make it out to be. Yes. If you EQ parallel tracks in phase shift will cause some minor changes in the EQ shape. But the changes that occur are generally minimal. For instance, if you use bell filters, the difference amounts to a slight change in the bandwidth of that filter. So it’s really a negligible difference. It’s a similar case with shelves. The only place where the change is somewhat significant is if you use low and hi pass filters with very steep slopes. But the effective difference there is that you have to listen instead of look at your low or high EQ settings to make good decisions. (Which is a good idea anyway). Because of this, it’s a best practice not to use super steep low and high pass filters in parallel...if you want to make things simpler and have the EQ curve you get be more like the EQ curve you see. Here’s a great video with some real audio examples of this: ruclips.net/video/efKabAQQsPQ/видео.html 2. If you have a side chain input on your plugin, you can have the sidechain’s input be a bus/aux that you put an EQ on. Hope that helps! -Justin
Really trying to learn this technique more than ever. Currently working on a track where I'm using the SPL IRON and then the Elysia Apha on the mix buss. The SPL adds some thickness and the Alpha is more transparent and precise, and it has the soft clip 😉
that's a common misconception to think that Teletronix LA-2A has a slow atack. It's atack varies from 100 μs to 1 ms program dependant. The slowest atack on UREI1176 RevE is 1 ms. So the same job on processing atacks can be done with either of them. But the release of the LA-2A is super slow and the release curve is different from the 1176. That what makes a perfect combination. Using LA-2A for drums creates much deeper and more distinct room effect.
I believe that the LA2A can have up to a 10ms attack time depending on program material. If you don’t mind me asking, where did you see that it only goes as low as 1 ms? I can shoot you a few links about it being as slow as 10ms if you like! While not exactly “slow” compared to how slow VCA or digital compressor can potentially go, this is certainly slow enough yo allow transients through on things like cymbals, acoustic guitar, high percussion, and even some drums. You also have to take the knee into account as well. Even with a nominally fast attack time, a softer knee will tend to give the impression of a relatively slower attack time and then a hard knee would with the same nominal attack time. The release of the LA2A is also program dependent. It can be almost moderately fast release 50%, but pretty darn slow to released all of the way. But that too depends on what kind of material is being fed into it! But whichever way you want to describe it, yes they can make a good pair on many types of instruments, thanks to the differing default envelopes! Thanks for the comment, Justin
@@SonicScoop sorry to say this, but, no, it can never be as slow as 10 ms whatever material you pass through it. If you look in the manual the specification uniquely says the atack time is essentially instantaneous -- 10 μs [manualslib.com/manual/828938/Teletronix-La-2a.html?page=2#manual], which is even faster than 1176, which atack time is 20 μs. But the atack time is too complicated, it gives us too little information about its character and what it will sound like on different material. What the manual says is more likely reaction time, but the atack slope is program dependent. I've measured the atack time on plenty of different LA-2As. The slowest case was around 600 μs [Gray face], which is 63% of the full slope. The full slope would be around 1 ms. You are absolutely correct concerning the knee factor. Yes, it gives the impression of a slightly softer invisible compression, but its too short compared to the attack time to take into account in our case. Thank you for your comments, Justin. Your videos are awesome. They give clear ideas on how things work. I would appreciate the links you are talking about. Best, Kirill
If the volume is consistent throughout the entire arrangement then there's no need for you to use a compressor (unless you want to shape the envelope or use a tube compressor for tone and attitude). If the volume of those VST instruments keeps changing (has dynamics), especially if you used a MIDI controller, then you can use serial compression or one compressor. Does that answer your question?
I made the mistake of using the fastest attack possible on vocals (Rcomp .5 attack and 10 or 11ms release @ 3 or 3.5 to 1 with up to about 7 to 10 DB of gain reduction on the big ass notes) because I was told you were supposed to do that and the mic had esses that were strong and unattractive and that helped catch and round out and sculpt the tone which was nice. The other settings I don't mind...but I wish I had learned to run it with a slower attack setting to keep from stealing the performances. Not only that but I learned to sing overemphasizing certain sounds knowing how they were going to push into the compressor and work in the mix. Then when you get new better mics and equipment you're in hella trouble. You have to learn to resing. It's weird. I'm thinking about using a fabfilter multiband from like 5k up with a fast attack and release to cover the top end transient problems that come with compression, and then a slower comp for the meat of the vocal. Maybe vice versa even. Frequency dependent attack and release. Lots of people just use a desser but i've never found one that sounds good to me and I have just about all of em except the Weiss one that looks hella nice. I've heard people love the 1176 into an la2a but...that la2a tone has never really worked for me. These are ALL great ideas. I love serial compression done the right way. I usually run an L2 or true peak limiter for extreme crest factor peaks of fast notes and transients and then a slower one for the rest of the volume differences which are less aggressive and don't cause as much compressor strain. Long comment but I hope it helps someone. Peace.
Hi Justin. My name is Mike kelley. You are the best. Do you have some tricks to controlling sibilance. I've been doing this a long time and i still have a hard time with de-essing or just taming the sibilance. I dont feel like i ever get it perfect. one guy i record always wants his lead vocal pitched down and i feel like that also adds to the problem because now my sibilance is lower in the freq range. should i use the original sibilance before the pitch down. i am working on a vocal right now and i tried clip gainging with de-essing and it sounds lispy, so now i undid the clip gain decreases and just stuck with de-essing with the waves f6 with 3 points at which i am controlling in the sibilance range. one for ess, one for f's, and ch's and its the closest ive got to sounding natural although natural went out the window with the pitch down but i have to make it work. thank you
Here are a few ideas: - The best way to control sibilance is to use the clip/pre gain before applying any processing, this is time consuming but it will give you more flexibility and control. - If you're relying on plugins then you'll still need to automate the threshold to avoid making the vocalist sound like they have a lisp. - You can also stack different de-essers, with one taking care of, let's say 2-6kHz and another one for the 6-12kHz range (these numbers are just examples). - A dynamic EQ can also be the best choice in some situations. - You can try using Soothe2, it's a great plugin that can fix your particular problem really easy. You can also use a combination of all the things I mentioned but in most cases, it will be overprocessing. I find that using clip gain and Soothe2 always does the job. If I'm working on a mix for a client who's in a hurry and I can't use clip gain since it takes time I'll stack 2 different de-essers and use the Soothe2 plugin to polish everything. Hope that helps.
This would have been so much better and more interesting if he just could show some examples of how it SOUNDS along with the talking. (That should be no problem on a podcast.) The theory is fine, but show us the proof of the pudding... before I fall asleep. ;-)
We have a TON of videos like that too. Here's one on this topic: ruclips.net/video/tnKgDAImZtg/видео.html The best audio examples though are going to be the ones you make for yourself using these concepts in your own DAW right away. Hope that helps! -Justin
@@SonicScoop Thanks, Justin. I see your point in making your own examples. But I must say (in a video like this) it is easier to also hear the effect it creates. But I would like to thank you for all the good videos you are making! :-)
What a fantastic explanation. This video could have been twice as long and I'd still be sad it was over. RUclips abounds with in-DAW tutorials, but what I really crave are these deep explanations of the theory, the why. Thank you so much.
So awesome to hear! That's what we're going for. If you like this one, we're announcing something REALLY deep on this subject on Monday. Stay tuned!
Flatten the transients and bring them up together. 🤦♂️This is the first time I’ve understood the logic of it. Thank you!!!!
That’s what I’m here’s for! Hope to see more of you around the channel :-)
-Justin
The only podcast I will nerd out to.
Every episode is a milestone👌🏼Thank you for posting awesome and useful tips that give perspective to this process.
Stacking limiters, in mastering, is a trick I use to get my masters loud - without crushing the audio and causing pumping
Thanks for addressing how different sample vs analog/live performance usage can be. A lot of people forget that and just think about techniques for an instrument type not the source origination.
It was the exact thing I needed to watch.I was just thinking and questioning all these to myself this afternoon and now you come and answering them all and more in a great point of view
This is a really excellent video. I still struggle with compression and figuring out exactly what problem I'm even solving in the first place, so being specific with "these settings to deal with transients, these settings for sustain" is way more helpful than the blanket "do what sounds good" advice you hear parroted. Because amateurs don't really know if what they're hearing sounds good, or how to fix it.
Really appreciate this, gonna be checking out the rest of your videos for sure.
So glad to be useful! We’ve got tons of free videos on this channel that I hope you enjoy.
If you ever want to go super deep on compression, we even have a full length course on it at CompressionBreakthroughs.com
Hope that helps!
-Justin
Couple days ago I checked on Creative sound lab's foldback compression trick. Stacking and feedbacking into a single compressor rack... I'm loving it.
Thank you for this, I'm a major benefactor of your podcasts.
Great vid! I knew all this but most don’t and you explained it perfectly- nice job!
If anyone is looking for a cool trick, following the compressor on the SSL channel (making sure the compression is happening after the EQ) with the LA2A is absolutely fantastic on vocals. I've been doing this for quite some time, and it really keeps the vocals right in your face.
I keep finding time to watch your videos, my friend.
Justin is our Superman in the world of Music Production .
saturation before compression also helps, basically a limiter anyway
Dude, you're the best. Thanks for the wisdom.
OD’n with one compressor will give you the sound of that compressor but at the expense of the performance. Try clip gain, automation, & peak limiting on vocal material. Use your DAW meters as a measurement tool. Try gates, envelope shapers & transient shapers on rhythmic material. Go there first before assuming a compressor is what you need.
So well done, Justin. Thank you again!
Great explanation, might have to get these courses you offer!
just read about this today in mixerman's book "musicians survival guide". He does this on vocals and uses EQ before and after each compressor. He was talking to another engineer and they both agree that analog compressors do much better than the plug-in versions in software packages.
Great information. Glad I do a lot of this already, but learned a few new things too. Thanks for all your great content!
Subscribed. This is such good information.
Thank you Kevin! Glad to have you with us.
Tasty serial compression! Bound to be a popular ‘take-away’. Serial, cereal, surreal. Even comes with a free ‘triple-entendre’. All jest aside, always grateful for your interesting lessons and this was particularly informative to this nerd 🙏😎
Great stuff Justin 👍🏻
I believe you can benefit from stacking compressors in electronic music too since compression is not only about transients
Thanks
Wow!! Thanks so much!! ❤️
-Justin
great video, it just makes me wonder how different this could sound with analog vs digital compressors, especially with the examples given of 1176 into LA2A and various plugin versions of the hardware.
Commenting on another older video, Justin... but I was wondering, with stacking compressors, where do you put an EQ? I know the "correct" answer is probably "it depends" - but most people I know in recording/mixing prefer EQing into compression; but I can see reasons to EQ after on the master bus (as long as you don't get crazy with boosting lol). Is there a signal chain you set up just as a matter of course, even if you don't use everything; and is it EQ > compression? Compression > EQ? Compressor 1 > EQ > Compressor 2? Or is it different each time you roll up to your desk?
Super helpful Justin!! I feel like I’m actually starting to almost understand compression! Lol
Ha. Me ten years later i still dont quite know what im doing with compression. I do use a lot of buss compressors and a couple on my master bus. The master is always to glue,warm and gently compress. Im just never quite sure if I’m using them right. Lol.
Lmaooooo.. we share the same feeling in this regard for sure
ty
today i tried stacking compressors on every channel and mix bus. every compressor ssl bus less than 1 db gain reduction and triple stacking the master bus. sounds overkill but worked well.
Definitely not overkill to get a few dB of compression out of a few compressors! Sometimes, you can push them harder than that too 😈
-Justin
What hardware fast compressor can you think of that does very transparent compression for tracking?
Loved it! Serial nerd stuff. I had thought your example would be 1176 into 160/165, instead of 1176 into LA2A? I really thought also in reverse that you would leave all the transients for SNR with slow attack into 1176 fast attack using your ears for balancing that instrument. Good explanation.
(1) How do you eliminate phasing with parallel compression? (2) How do you high pass mix bus compression if the compressor doesn't have a high pass setting?
1. You don’t and don’t need to! The reality is that it’s not nearly as big of a “problem” as some make it out to be.
Yes. If you EQ parallel tracks in phase shift will cause some minor changes in the EQ shape. But the changes that occur are generally minimal.
For instance, if you use bell filters, the difference amounts to a slight change in the bandwidth of that filter. So it’s really a negligible difference. It’s a similar case with shelves.
The only place where the change is somewhat significant is if you use low and hi pass filters with very steep slopes.
But the effective difference there is that you have to listen instead of look at your low or high EQ settings to make good decisions. (Which is a good idea anyway).
Because of this, it’s a best practice not to use super steep low and high pass filters in parallel...if you want to make things simpler and have the EQ curve you get be more like the EQ curve you see.
Here’s a great video with some real audio examples of this: ruclips.net/video/efKabAQQsPQ/видео.html
2. If you have a side chain input on your plugin, you can have the sidechain’s input be a bus/aux that you put an EQ on.
Hope that helps!
-Justin
Really trying to learn this technique more than ever. Currently working on a track where I'm using the SPL IRON and then the Elysia Apha on the mix buss. The SPL adds some thickness and the Alpha is more transparent and precise, and it has the soft clip 😉
You should read the manual for the SPL IRON, you're likely missing a trick or two, it's capable of so much, take a read and you will see
@@UncleBenjs oh yes, I've looked at the chart explaining the rectifier settings. It's been a while so I need a refresher 😉👍
hi I'm jindu a mix engineer in Kingston Jamaica luv di vybz man respect
some compressors add harmonic content as well - even without dynamics.
Really helpful
that's a common misconception to think that Teletronix LA-2A has a slow atack. It's atack varies from 100 μs to 1 ms program dependant. The slowest atack on UREI1176 RevE is 1 ms. So the same job on processing atacks can be done with either of them. But the release of the LA-2A is super slow and the release curve is different from the 1176. That what makes a perfect combination. Using LA-2A for drums creates much deeper and more distinct room effect.
I believe that the LA2A can have up to a 10ms attack time depending on program material.
If you don’t mind me asking, where did you see that it only goes as low as 1 ms? I can shoot you a few links about it being as slow as 10ms if you like!
While not exactly “slow” compared to how slow VCA or digital compressor can potentially go, this is certainly slow enough yo allow transients through on things like cymbals, acoustic guitar, high percussion, and even some drums.
You also have to take the knee into account as well. Even with a nominally fast attack time, a softer knee will tend to give the impression of a relatively slower attack time and then a hard knee would with the same nominal attack time.
The release of the LA2A is also program dependent. It can be almost moderately fast release 50%, but pretty darn slow to released all of the way. But that too depends on what kind of material is being fed into it!
But whichever way you want to describe it, yes they can make a good pair on many types of instruments, thanks to the differing default envelopes!
Thanks for the comment,
Justin
@@SonicScoop sorry to say this, but, no, it can never be as slow as 10 ms whatever material you pass through it. If you look in the manual the specification uniquely says the atack time is essentially instantaneous -- 10 μs [manualslib.com/manual/828938/Teletronix-La-2a.html?page=2#manual], which is even faster than 1176, which atack time is 20 μs. But the atack time is too complicated, it gives us too little information about its character and what it will sound like on different material. What the manual says is more likely reaction time, but the atack slope is program dependent. I've measured the atack time on plenty of different LA-2As. The slowest case was around 600 μs [Gray face], which is 63% of the full slope. The full slope would be around 1 ms.
You are absolutely correct concerning the knee factor. Yes, it gives the impression of a slightly softer invisible compression, but its too short compared to the attack time to take into account in our case.
Thank you for your comments, Justin. Your videos are awesome. They give clear ideas on how things work. I would appreciate the links you are talking about.
Best,
Kirill
What about VST instrument, especially superior drummer 3 or omnisphere?, it has already have a great sound
If the volume is consistent throughout the entire arrangement then there's no need for you to use a compressor (unless you want to shape the envelope or use a tube compressor for tone and attitude). If the volume of those VST instruments keeps changing (has dynamics), especially if you used a MIDI controller, then you can use serial compression or one compressor. Does that answer your question?
I made the mistake of using the fastest attack possible on vocals (Rcomp .5 attack and 10 or 11ms release @ 3 or 3.5 to 1 with up to about 7 to 10 DB of gain reduction on the big ass notes) because I was told you were supposed to do that and the mic had esses that were strong and unattractive and that helped catch and round out and sculpt the tone which was nice. The other settings I don't mind...but I wish I had learned to run it with a slower attack setting to keep from stealing the performances. Not only that but I learned to sing overemphasizing certain sounds knowing how they were going to push into the compressor and work in the mix. Then when you get new better mics and equipment you're in hella trouble. You have to learn to resing. It's weird.
I'm thinking about using a fabfilter multiband from like 5k up with a fast attack and release to cover the top end transient problems that come with compression, and then a slower comp for the meat of the vocal. Maybe vice versa even. Frequency dependent attack and release. Lots of people just use a desser but i've never found one that sounds good to me and I have just about all of em except the Weiss one that looks hella nice. I've heard people love the 1176 into an la2a but...that la2a tone has never really worked for me. These are ALL great ideas. I love serial compression done the right way. I usually run an L2 or true peak limiter for extreme crest factor peaks of fast notes and transients and then a slower one for the rest of the volume differences which are less aggressive and don't cause as much compressor strain. Long comment but I hope it helps someone. Peace.
Hi Justin. My name is Mike kelley. You are the best. Do you have some tricks to controlling sibilance. I've been doing this a long time and i still have a hard time with de-essing or just taming the sibilance. I dont feel like i ever get it perfect. one guy i record always wants his lead vocal pitched down and i feel like that also adds to the problem because now my sibilance is lower in the freq range. should i use the original sibilance before the pitch down. i am working on a vocal right now and i tried clip gainging with de-essing and it sounds lispy, so now i undid the clip gain decreases and just stuck with de-essing with the waves f6 with 3 points at which i am controlling in the sibilance range. one for ess, one for f's, and ch's and its the closest ive got to sounding natural although natural went out the window with the pitch down but i have to make it work. thank you
Here are a few ideas:
- The best way to control sibilance is to use the clip/pre gain before applying any processing, this is time consuming but it will give you more flexibility and control.
- If you're relying on plugins then you'll still need to automate the threshold to avoid making the vocalist sound like they have a lisp.
- You can also stack different de-essers, with one taking care of, let's say 2-6kHz and another one for the 6-12kHz range (these numbers are just examples).
- A dynamic EQ can also be the best choice in some situations.
- You can try using Soothe2, it's a great plugin that can fix your particular problem really easy.
You can also use a combination of all the things I mentioned but in most cases, it will be overprocessing. I find that using clip gain and Soothe2 always does the job. If I'm working on a mix for a client who's in a hurry and I can't use clip gain since it takes time I'll stack 2 different de-essers and use the Soothe2 plugin to polish everything.
Hope that helps.
Volume automation!
@@audiospectra thank you. I am looking into sooth2.
@@mykal92684 You're very welcome. All the best.
Cool bro
wow you amazing me man
Same surname, with an i added at the end. Could we be distant relatives??
Who knows! If your family from Sicily originally?
Could be! I've heard of others I don't know at all from around the country.
-Justin
How many minutes till your personal sale pitch ends...?
This would have been so much better and more interesting if he just could show some examples of how it SOUNDS along with the talking. (That should be no problem on a podcast.) The theory is fine, but show us the proof of the pudding... before I fall asleep. ;-)
We have a TON of videos like that too. Here's one on this topic: ruclips.net/video/tnKgDAImZtg/видео.html
The best audio examples though are going to be the ones you make for yourself using these concepts in your own DAW right away.
Hope that helps!
-Justin
@@SonicScoop Thanks, Justin. I see your point in making your own examples. But I must say (in a video like this) it is easier to also hear the effect it creates.
But I would like to thank you for all the good videos you are making! :-)
video doesnt start til 4 minutes in
"let's get right into it" -- 3 minutes later, after shout outs, promos and digressions. Very good content, but you don't get right into it.