Can't tell you how much I've learned from this brief yet very informative tutorials. I've been processing dialogues so blindly and randomly without any ideas what to do for my projects and this has just made me realize how much can be done by just playing with dialogue!!
Hunting for random frequencies is a fast way to over-process, especially if you're just starting out. Practically everything will start sounding like you should cut it if you're listening with a sharp 10dB boost. You should only seek out and cut ringing frequencies if they're audible enough to bother you in the first place.
While I do agree with you to a point, most people don’t have a well trained ear to hear these ringing frequencies in context of a mix. By sweeping you can identify the ringing frequency and reduce it.
Love your content!🔥 Question.... You used brickwall EQ cuts. Correct me if I'm wrong, but wouldn't that introduce problems with phase shifts in your audio? Especially since you're using the "zero latency" setting in the plugin.
Absolutely right. Those brick wall cuts are definitely introducing some resonance and phase issues. Never seen anyone go that extreme on what is generally a 12-24db/octave roll-off at max.
why add music in the background - its so distracting - let us hear what you are saying and listen to the examples, really no need at all to add music underneath
I much prefer doing manual spectral noise reduction in the RX editor. I haven't found any real time noise reduction plugin that stands up to the level of detail you can get manually denoising clip by clip using the editor. Also - I find multiband compression can be a a very useful dialogue shaping tool in conjuction with broad EQ.
I agree with you. Depending on the level of noise (how poorly the dialogue was recorded) I move that processing up the chain as it would need more extreme settings. I do have a multiband compressor and an broadband eq on my dialogue bus to help control the tone of my dialogue more, I’ve just found a single band compressor works good on the track itself.
RX Denoise is designed to be used in the RX Editor, not as a real time plugin. You must "teach" it the noise first by giving it a few second clip of the cleanest noise you can get, and then it will work as intended. All those noise reductions you showed sounded pretty terrible. Also AI is making huge advancements in this area. It will probably be the standard, in less than a year, to use AI powered tools in our editors to clean most types of noises.
i’ve been using multiband compression less and less lately. i heard this tip to try TDR Kotelnikov’s frequency dependent ratio feature and it changed the game. no phasing issues at the crossovers anymore either
Idk much about audio mixing but isnt doing those sharpe pointed cut in 3:14 generally not recommended? Because in any freq, when you do a sharpe bell curve, it sounds like there is a problem because you are hearing single freq over amplified
The probelm these frequencies cause is when you compress the signal. When you compress a signal these frequencies can be more pronounced and can make the voice harder/more tiring to listen to. These rining frequencies can be anything from room modes to just sibalane and should be controlled. A lot of tutorials you'll see on RUclips will tell you to cut the frequencies, which is not correct. You need to reduce them enough so they blend better with the dialogue. which can be anywhere from 5-10dB.
@@AftertouchAudio Not entirely true. When you're boosting a bell filter with such a small q-factor, you just completely take those frequencies you are sweeping across out of context. In this case, almost any frequency can sound "problematic", while in context it rarely isn't. The amount of narrow cuts you did and the severity of them is completely unnecessary. If you are really worried about certain frequency bands being problematic after compression, just adjust the eq after you've compressed it. You will see you'll need much less severe eq adjustments.
I’ve heard of this channel from Jessie and you helping with some of the audio for Total Drama oskay island great work and that’s so nice of you to help out the the beloved fan made season of total drama that a lot of people as well As myself are really interested in it’s Gonna be great and I just wanted to let you know I was one of his fans and cool video by the why it could help some people as creators for certain projects.
Hey thank you NoLJak 101, It is going to be a good episode, just been a slow grind at it with all of the other projects on the go haha. It will get done though.
i saw someone video he told that we should always use noise reduction before compression but you did exact opposite of that! Please could you explain why should we use noise reduction after compression??
So I've been meaning to do an update to this video for awhile, life is just getting in the way. Essentially there is no HARD rule that says do this, then this, then this, then this. Your goal with transparent processing (like dialogue processing) is to do as little as possible. Sometimes it is ok to have noise suppression before compression sometimes it is better to have it after compression. With that being said, my dialogue processing chain has really changed quite a bit since the release of this video that it does warrant an updated video.
Two questions. You said to bring each clip up to -24 or -23. 1. What about a dialogue part where the the actor is whispering but then talks loud? 2. Can you a tutorial on doing ADR in Nuendo, using the ADR Panel?
1) It is more of a starting point. If you have your dialogue at a consistent level, then you can make the choice to lower it slightly for whispering or louder for yelling. (Follow the natural dynamic flow of the dialogue) Where if you don't you end up with a mess where dialogue is too loud, too quiet, and then your plugins don't operate well. 2) I don't do much ADR, but I could go through an ADR Panel in a future video.
This is invaluable! One question, if we have a spec of -23 and gain stage like the first step to -23, after processing it'll be louder? do we just turn it down or should we gain stage to say -26? thank you
Do you have any problems when adjusting clip gain to each region with room tone? Wouldn’t the room tone change with each volume alteration and make it harder to process? Or do you simply remove all of the background noise and add an artificial one?
Not really, the only time where you run into issues like that is when the dialogue is recorded VERY poorly with levels not being set consistently. If I have noisy dialogue I don't try and process it completely away I instead aim to lightly process it and then fill it with noisier BGs instead.
So it really depends on how much noise there is to reduce. Sometimes I'll switch around the order so it is the following. Gain Stage - Subtractive EQ - Denoise - Compress - DS - Saturation - Additave EQ. It really just depends on what you need and what is the least amount of work each plugin will be doing.
I can appreciate the passion you must need to perform this much processing for dialogue, but it’s just too extreme to be practical. You could back off your processing by 60% or more and it would likely sound better and more natural.
Just discovered your channel, great videos! I have a question, after you set the lufs on the dialogue clips, when you start the processing with plugins, I guess you still keep an eye on the lufs? In order to keep the 23 lufs. Like what's your workflow, you gain stage each single time you add something? Hope it makes sense :) thank you!
Basically I still keep an eye on the LUFS Meter as I add plugins to the dialogue chain. As plugins get added to the chain I'm not really adding any gain other than maybe some EQ boosts in the upper register to help with Lavs, the compression is really just keeping the loud peaks in check so not a lot if any gain is added to the final output. But yes, when I am mixing and riding volume automation, I'm still keeping an eye on my LUFS meters.
would of been of use if you showed how to EQ the bassy parts of the voice Aint gonna really work when you got the music playing in the background of the game and dialog starts makin the music become all muddy and dirty sound duo to the bass from the voice
You know, I watch the video every few weeks. I've been working in audio post for a while....mainly short form and some of the tips in this video are genius so I really appreciate the help on this long form I've been working on for a few years. My NR procedure is very different as I use Acon mostly...nevermind that, this video has helped think outside the box a bit too, thanks for this inspiring, info packed gem!
I have been scouring youtube for weeks educating mysef about dialogue editing for my short film, and all I've needed is something that lists the step by step process as succinctly as this does. Thank you!
I usually automate things like the noise reduction, reverb reduction, automation, & the EQ preset. As people change rooms they will need different EQ grading as some of the frequencies you remove are room modes.
You can do it on the track itself, but I would not say FL Studio is aimed towards producing mixes for films. Can FL Studio do it..... yes, but there are other DAWs that are designed with features in mind for film & game workflows.
Your dialogue sounds over processed in terms of NR and subtractive EQ. Maybe you’re exaggerating the point for RUclips, but this is not how finished dialogue should sound. Maybe try using RX Isolate or Clarity for the broadband noise, even RX spectral de-noise could be considered. That amount of subtractive eq makes the dialogue sound ‘ringy’, don’t obsess over ringing frequencies, anything boosted with a narrow Q will sound weird/ringy especially if it’s a harmonic of the voice
The settings for the noise reduction were exaggerated intentionally, going from minimal to extreme settings, to demonstrate the differences between the 3 plugins. I'll end up updating this video as my template has changed drastically, and there is very little on my processing chain now. I agree that the examples I used could have been much clearer, and I did need better source audio as I did not have permission to use any other examples, which will be corrected in future videos. When I did this video I was still trying to figure out the whole RUclips thing and find my own flow to presenting, but I do appreciate the feedback.
If you are making your content for streaming would you make each clips LUFS -14 or whatever your end target is? You said -23 LUFS since thats for broadcast obviously but would you change that number for different deliveries off the bat? Do you do the same for the music. I mostly do talking head corporate stuff and some promotional videos.
nevermind i watched that part and answered my own question lol however I work in premiere while editing and I also do my audio workin there since I dont really understand the round trip workflow through audition or my other DAWs and my company's workflow doesnt really work for that but I am curious if the Auto loundness match in the essential sound panel is a fast way to auto level all the dialogue clips to -23 or -14 or whatever I need or if it's a bad idea.
i hope this question isnt too off topic. but im so confused say you have you dialogue track and youve got all your insets/sends plugins dialed in and sounding great. what do you do in the next scene with the same track when the dialogue is recorded in a different environment and with a different mic? you obviously have to use different settings in you plugins. is the answer automation? im using cubase too
Hey Carlos, Automation is what you would use for different scenes. Cubase offers an awesome suite of Automation controls. I would really look into all of the automation modes including preview as it will save you hours.
@@AftertouchAudio thanks man i have been scouring videos all day. super appreciate the response. im trying to find the preview in the automation panel. i guess its only in nuendo?
Hi, cool video ! For the 1st step of loudness normalizing, Reaper has a feature that does it automatically, which avoids you the pain of trimming each clip's gain knob manually. You should try it some time :)
@@AftertouchAudio Im curious is there is a plug in that can analyze clips individually that works in premiere. Without having to listen to the clip each time but rather click the clip and analyze like you do.
I don't understand the first thing you do, setting gain for every dialogue so they are perfectly even erase all the natural dynamic the talents performed and the mix made by the sound mixer on the set. By doing this you have the same level for shouting and whisper
Not quite, you average it out to be -23 or whatever your spec is. If your actor yells for a section preserve the dynamics and go louder, if you actor whispers preserve the dynamics and go quieter. It is just a gain staging step.
Can't tell you how much I've learned from this brief yet very informative tutorials. I've been processing dialogues so blindly and randomly without any ideas what to do for my projects and this has just made me realize how much can be done by just playing with dialogue!!
Hunting for random frequencies is a fast way to over-process, especially if you're just starting out. Practically everything will start sounding like you should cut it if you're listening with a sharp 10dB boost. You should only seek out and cut ringing frequencies if they're audible enough to bother you in the first place.
While I do agree with you to a point, most people don’t have a well trained ear to hear these ringing frequencies in context of a mix. By sweeping you can identify the ringing frequency and reduce it.
This didn't help much. You didn't explain how to help with DRC for movies. Audio sfx is too high but low voice sounds
an update to this video would be great!!
Which software is this ?
It is called Nuendo.
I like to process my dialogue like how anime does it. Anime dialogue is typically crispy.
+12 semi tones haha.
@@AftertouchAudio XD
Love your content!🔥
Question.... You used brickwall EQ cuts. Correct me if I'm wrong, but wouldn't that introduce problems with phase shifts in your audio? Especially since you're using the "zero latency" setting in the plugin.
Absolutely right. Those brick wall cuts are definitely introducing some resonance and phase issues. Never seen anyone go that extreme on what is generally a 12-24db/octave roll-off at max.
why add music in the background - its so distracting - let us hear what you are saying and listen to the examples, really no need at all to add music underneath
Cause I was new to making tutorials and still trying to figure out the best way to present information. New videos don’t have music like this.
I much prefer doing manual spectral noise reduction in the RX editor. I haven't found any real time noise reduction plugin that stands up to the level of detail you can get manually denoising clip by clip using the editor. Also - I find multiband compression can be a a very useful dialogue shaping tool in conjuction with broad EQ.
I agree with you. Depending on the level of noise (how poorly the dialogue was recorded) I move that processing up the chain as it would need more extreme settings.
I do have a multiband compressor and an broadband eq on my dialogue bus to help control the tone of my dialogue more, I’ve just found a single band compressor works good on the track itself.
Try Clarity VX from Waves, it's brilliant
RX Denoise is designed to be used in the RX Editor, not as a real time plugin. You must "teach" it the noise first by giving it a few second clip of the cleanest noise you can get, and then it will work as intended. All those noise reductions you showed sounded pretty terrible. Also AI is making huge advancements in this area. It will probably be the standard, in less than a year, to use AI powered tools in our editors to clean most types of noises.
bertom audio....try it
i’ve been using multiband compression less and less lately. i heard this tip to try TDR Kotelnikov’s frequency dependent ratio feature and it changed the game. no phasing issues at the crossovers anymore either
Yea, so.... hire a professional! Will do! Great video.
Idk much about audio mixing but isnt doing those sharpe pointed cut in 3:14 generally not recommended?
Because in any freq, when you do a sharpe bell curve, it sounds like there is a problem because you are hearing single freq over amplified
The probelm these frequencies cause is when you compress the signal. When you compress a signal these frequencies can be more pronounced and can make the voice harder/more tiring to listen to. These rining frequencies can be anything from room modes to just sibalane and should be controlled.
A lot of tutorials you'll see on RUclips will tell you to cut the frequencies, which is not correct. You need to reduce them enough so they blend better with the dialogue. which can be anywhere from 5-10dB.
@@AftertouchAudio Not entirely true. When you're boosting a bell filter with such a small q-factor, you just completely take those frequencies you are sweeping across out of context. In this case, almost any frequency can sound "problematic", while in context it rarely isn't. The amount of narrow cuts you did and the severity of them is completely unnecessary. If you are really worried about certain frequency bands being problematic after compression, just adjust the eq after you've compressed it. You will see you'll need much less severe eq adjustments.
YOU COULD HAVE NOT USED THE BACKGROUND MUSIC LIKE A TO TAL I D IO T.
I’ve heard of this channel from Jessie and you helping with some of the audio for Total Drama oskay island great work and that’s so nice of you to help out the the beloved fan made season of total drama that a lot of people as well
As myself are really interested in it’s Gonna be great and I just wanted to let you know I was one of his fans and cool video by the why it could help some people as creators for certain projects.
Hey thank you NoLJak 101, It is going to be a good episode, just been a slow grind at it with all of the other projects on the go haha. It will get done though.
i saw someone video he told that we should always use noise reduction before compression but you did exact opposite of that! Please could you explain why should we use noise reduction after compression??
So I've been meaning to do an update to this video for awhile, life is just getting in the way.
Essentially there is no HARD rule that says do this, then this, then this, then this. Your goal with transparent processing (like dialogue processing) is to do as little as possible. Sometimes it is ok to have noise suppression before compression sometimes it is better to have it after compression.
With that being said, my dialogue processing chain has really changed quite a bit since the release of this video that it does warrant an updated video.
Two questions. You said to bring each clip up to -24 or -23.
1. What about a dialogue part where the the actor is whispering but then talks loud?
2. Can you a tutorial on doing ADR in Nuendo, using the ADR Panel?
1) It is more of a starting point. If you have your dialogue at a consistent level, then you can make the choice to lower it slightly for whispering or louder for yelling. (Follow the natural dynamic flow of the dialogue) Where if you don't you end up with a mess where dialogue is too loud, too quiet, and then your plugins don't operate well.
2) I don't do much ADR, but I could go through an ADR Panel in a future video.
More please!!! Thanks.
Just what I needed! Thanks for sharing your process!
🔥🔥🔥🔥🔥🔥🔥
This is invaluable! One question, if we have a spec of -23 and gain stage like the first step to -23, after processing it'll be louder? do we just turn it down or should we gain stage to say -26? thank you
Do you have any problems when adjusting clip gain to each region with room tone? Wouldn’t the room tone change with each volume alteration and make it harder to process? Or do you simply remove all of the background noise and add an artificial one?
Not really, the only time where you run into issues like that is when the dialogue is recorded VERY poorly with levels not being set consistently. If I have noisy dialogue I don't try and process it completely away I instead aim to lightly process it and then fill it with noisier BGs instead.
So, why do you apply broadband noise reduction after processing? Why not before?
So it really depends on how much noise there is to reduce. Sometimes I'll switch around the order so it is the following.
Gain Stage - Subtractive EQ - Denoise - Compress - DS - Saturation - Additave EQ. It really just depends on what you need and what is the least amount of work each plugin will be doing.
I can appreciate the passion you must need to perform this much processing for dialogue, but it’s just too extreme to be practical.
You could back off your processing by 60% or more and it would likely sound better and more natural.
Just discovered your channel, great videos! I have a question, after you set the lufs on the dialogue clips, when you start the processing with plugins, I guess you still keep an eye on the lufs? In order to keep the 23 lufs. Like what's your workflow, you gain stage each single time you add something? Hope it makes sense :) thank you!
I had the same exact question, but specifically
for Protools and WLM Plus
Basically I still keep an eye on the LUFS Meter as I add plugins to the dialogue chain. As plugins get added to the chain I'm not really adding any gain other than maybe some EQ boosts in the upper register to help with Lavs, the compression is really just keeping the loud peaks in check so not a lot if any gain is added to the final output.
But yes, when I am mixing and riding volume automation, I'm still keeping an eye on my LUFS meters.
^^^
@@AftertouchAudio great thank you!
❤ such a compact and to the point video . Thnx man
would of been of use if you showed how to EQ the bassy parts of the voice
Aint gonna really work when you got the music playing in the background of the game and dialog starts makin the music become all muddy and dirty sound duo to the bass from the voice
wow, one of the best I've ever seen on the subject, great work!
You know, I watch the video every few weeks. I've been working in audio post for a while....mainly short form and some of the tips in this video are genius so I really appreciate the help on this long form I've been working on for a few years. My NR procedure is very different as I use Acon mostly...nevermind that, this video has helped think outside the box a bit too, thanks for this inspiring, info packed gem!
Thanks much, I'm learning a lot from your video.....
I have been scouring youtube for weeks educating mysef about dialogue editing for my short film, and all I've needed is something that lists the step by step process as succinctly as this does. Thank you!
Encouraging to see that others are using Nuendo on a pro level. Makes me feel good about myself.
...but have you tried REAPER? ;)
Hey man, great info in your channel! Is it possible to get that Nuendo template for post prod?
Thanks!
Hi Tyler! Great video! I have one question, do you automate any plugin in these process? Or just work clip by clip?
I usually automate things like the noise reduction, reverb reduction, automation, & the EQ preset. As people change rooms they will need different EQ grading as some of the frequencies you remove are room modes.
Big thanks, i have no idea how to apply plugins on a small clip in fl studio tho
You can do it on the track itself, but I would not say FL Studio is aimed towards producing mixes for films. Can FL Studio do it..... yes, but there are other DAWs that are designed with features in mind for film & game workflows.
Your dialogue sounds over processed in terms of NR and subtractive EQ. Maybe you’re exaggerating the point for RUclips, but this is not how finished dialogue should sound. Maybe try using RX Isolate or Clarity for the broadband noise, even RX spectral de-noise could be considered. That amount of subtractive eq makes the dialogue sound ‘ringy’, don’t obsess over ringing frequencies, anything boosted with a narrow Q will sound weird/ringy especially if it’s a harmonic of the voice
The settings for the noise reduction were exaggerated intentionally, going from minimal to extreme settings, to demonstrate the differences between the 3 plugins. I'll end up updating this video as my template has changed drastically, and there is very little on my processing chain now.
I agree that the examples I used could have been much clearer, and I did need better source audio as I did not have permission to use any other examples, which will be corrected in future videos.
When I did this video I was still trying to figure out the whole RUclips thing and find my own flow to presenting, but I do appreciate the feedback.
You should win some awards for this video
为什么这么官方的网站，会出这么业余的视频
Thank you!
If you are making your content for streaming would you make each clips LUFS -14 or whatever your end target is? You said -23 LUFS since thats for broadcast obviously but would you change that number for different deliveries off the bat? Do you do the same for the music. I mostly do talking head corporate stuff and some promotional videos.
nevermind i watched that part and answered my own question lol however I work in premiere while editing and I also do my audio workin there since I dont really understand the round trip workflow through audition or my other DAWs and my company's workflow doesnt really work for that but I am curious if the Auto loundness match in the essential sound panel is a fast way to auto level all the dialogue clips to -23 or -14 or whatever I need or if it's a bad idea.
Wat lufs do we need to keep for netflix Amazon prime and stuff
Netflix is -27 but most platforms are -23 aside from RUclips Spotify etc which are usually -14
Can you please point me to the direction of the Pro Loudness Control ? I could not find it and it in the description. Thank you!
The Pro Loudness Control is not a plugin, it is a feature in ProTools Ultimate
Good, very usefull video. Thank You for sharing your knowledges ))
Thank you for watching =^.^=
Hi, where in this chain would you put dereverb?
I would put the de-reverb after the de-noising.
i hope this question isnt too off topic. but im so confused
say you have you dialogue track and youve got all your insets/sends plugins dialed in and sounding great. what do you do in the next scene with the same track when the dialogue is recorded in a different environment and with a different mic?
you obviously have to use different settings in you plugins.
is the answer automation?
im using cubase too
Hey Carlos, Automation is what you would use for different scenes. Cubase offers an awesome suite of Automation controls. I would really look into all of the automation modes including preview as it will save you hours.
@@AftertouchAudio thanks man i have been scouring videos all day. super appreciate the response.
im trying to find the preview in the automation panel. i guess its only in nuendo?
Nice one
Hi, cool video ! For the 1st step of loudness normalizing, Reaper has a feature that does it automatically, which avoids you the pain of trimming each clip's gain knob manually. You should try it some time :)
Dose it adjsust the clip gain or dose it automate the volume within each clip? One is good the other is bad.
@@AftertouchAudio it adjusts clip gain
@@martinrabiller8048 Well that is cool, I have a program that does it in ProTools but nothing that dose it in Nuiendo "yet".
@@AftertouchAudio what’s the name of the program that you use for pro tools
@@AftertouchAudio Im curious is there is a plug in that can analyze clips individually that works in premiere. Without having to listen to the clip each time but rather click the clip and analyze like you do.
nice work
Cool techniques...thankss
I’m glad you enjoyed the video!!!
What is this software?
I'm using a program called Nuendo.
Good job
Thank you =^.^=
Awesome video! Thanks
Glad you enjoyed the video =^.^=
Wow, thank you for this video! Some of your analogies were very helpful in understanding some things like subtractive EQ.
Hey Anne, I'm glad you enjoyed the video =^.^=
Amezing video
Thank you Reshma, I'm glad you enjoyed the video.
excellent！
I don't understand the first thing you do, setting gain for every dialogue so they are perfectly even erase all the natural dynamic the talents performed and the mix made by the sound mixer on the set. By doing this you have the same level for shouting and whisper
Not quite, you average it out to be -23 or whatever your spec is. If your actor yells for a section preserve the dynamics and go louder, if you actor whispers preserve the dynamics and go quieter.
It is just a gain staging step.
It's so nice when someone who knows what they are doing and makes a video for the rest of us. Thank you.
haha thank you XD