Watching you work and hearing the subtle changes you make as you go is like watching a colorist go through their grade and seeing incremental changes to the image. This is elite stuff man.
Im a firm believer that you should always use a compressor and not normalize. Then put a Limiter on Bus 1 to control total peak volume. You wont bring up the low volume parts without compression. Great stuff as always Jason!
Awesome explanation and demonstration of the difference. I am going to continue working on my skills in DaVinci Resolve and your tutorials are helping a lot. Thank you!
If you where to «render out» the video with the audio affected by the compressor and makeup gain, and re-import it into the timeline, showing what the waveform looks like with peaks compressed and the makeup gain applied it might show people more clearly what the dynamics processing does to the waveform vs what normalizing does to the waveform :)
Great video about the distinction Jay. One thing I recently learned from the engineering team on the settings like the new TH-cam normalize preset that has both a loudness and a true peak preset - It pulls the volume up to whichever one of those it hits first. In this case, and most of these it will usually just be the dBTP level. Since it doesn't do any dynamics it can't actually hit -14 LKFS. I think having that loudness part in there is kind of confusing since it would almost never be used since the dBTP would be hit first.
Thanks Chadwick! Good info to know and makes total sense. I didn't think about it hitting one of those first. Since I saw people had questions on it, I figured I'd drop a video about it. Thanks for checking it out man!
Chadwick I certainly agree with the word confusing especially as there is things like Auphonic around that will automatically adjust the sound to -14LUF (in the case of YouTUbe). I at first thought when DaVinci introduced the normalisation for Netflix and TH-cam it would do this and frankly now I am not at all sure what it is doing or why I would in fact want to use it.
Do you mean the Engineering team of Da Vinci? I neither understand how you could do a normalization of both parameters (TP and Loudness) without doing some dynamics. It is confusing to have bopth parameters on the same interface.
Awesome, glad to hear it was helpful! What type of audio settings would you like to know more about? I’ve got tons of videos about DaVinci Resolve’s audio but I’m always looking for more ideas for videos. 😜
Thanks Jason, I had no idea the make up was for the quiet parts only I thought that was another type of gain slider that affected everything on the track.
Well that's a coincidence Jason. Tonight I am doing some testing to see what effect ticking the Normalise Audio check box has on the TH-cam render preset. Well now I know. My present videos are basically Game sound output on 1 audio track and my mic audio on another and I want to make sure my levels are good and don't get turned down by TH-cam. I think getting my levels set before recording my footage is the way to go, followed by checking and adjusting the dymanics if required to make sure things are where I want them to be before I upload. Thanks for another detailed video explanation. 👍
I think you are certainly thinking about your process in the correct way. Basically you won't need to use the normalize audio on export if you set your levels and check your loudness before you render.
Hello Jason, I appreciate all the great content you are making for years, I learned a lot from your tutorials, from basic to some advanced techniques and improved my audio a lot. The question that I have is would you recommend using "Normalize audio" before or after using Dynamics? In my head it makes sense to use it after dynamics just to set it in certain loudness range instead of measuring it and doing it manually for some quick delivery interviews and similar, but I might be very wrong... Thanks once again for your videos its amazing library of knowledge!
audio is definitely the hardest part of recording anything. getting mic settings right in the first place so that it's usable, and then balancing everything out in post. i've been trying to make highlight clips for years and still run into audio issues because i just dont understand it
I hear ya! For your mic settings, I'd recommend to try setting it around -18dB, that way you've got plenty of headroom to boost it a bit in post. In Post I like to aim for -10dB for my dialogue track and at the end of the audio process I make sure I'm just under -14 LUFS for overall Loudness for TH-cam.
What a great help, as always, Jason. I wonder if you have experienced some problems in the Fairlight tab after the last mac update to Sonoma 14.1, for me, it's almost impossible to use that tab, quite difficult.
Thanks! I haven’t updated to Sonoma yet. I’m still on Ventura. I usually wait quite a while before I update the OS. That gives the software companies time to catch up.
In my case, I'd be a bit concerned when making a music video with a track that has already been mastered. I would think normalizing a mastered track that has some dynamic parts that are -1 to -2dB in volume (such as in the chorus sections) and if the song has already been keyed to TH-cam -14 LKFS target level. I think that nomalizing that audio would flatten out the dynamics because they are already so close together.
I hear ya, keep in mind that normalizing is just adjusting the overall volume, so it doesn’t really change any of the dynamic parts of a music track. It’s essentially the same as turning the volume on your radio up and down.
Thanks as always Jason! If I look at how the graph moves, the Make Up slider in the compressor doesn’t just make the quiet parts louder, but it makes all parts louder including the loud parts. Am I wrong? I didn’t post this comment to nitpick, just because I hope to learn and because others might benefit too.
For some reason i always thought Normalize was doing some compression too. Good to know it's useless. I always compress and adjust volume afterward anyway.
I wanna make sure i understand the workflow or chain of audio manipulation. So for audio work, especially voiceovers, the first step of the process should be to normalize LUFS of the entire project according to TH-cam standards for example and then and only then do you move on to using compressors/limiters to fine tune the dynamics?
You’ve actually got it backwards. First set proper/desired levels for all tracked based on track meters in mixer. Then move on to eq, dynamics, then effects. After mixing all tracks - dialogue, music and sound effects, then you check the overall loudness (LUFS) to meet TH-cam standards. You would use the loudness meters in Fairlight and/or the graph loudness meter on bus1 in the timeline tracks section of Fairlight. Does this help?
Again, very clear. Questions: 1) I assume it's worth normalising the whole track at the end (after dynamics etc) to hit a sensible TH-cam level? (I see this can now be done at the render stage) 2) I saw you also moving some expander settings - would love a video on this at some point. I think you did one before but I am finding these newer ones much easier to understand.
Thanks! I don’t think there’s any reason to normalize if you set all of your audio levels properly and use dynamics properly. You’ve essentially done a better job at getting your levels, correct then simply normalizing would do, in my opinion. I do have some videos on the gate and expander, but I also have a video that I recently put out, showing how the expander works, here’s a link to it: th-cam.com/video/tqgUA44Nd9k/w-d-xo.htmlsi=uDb8JghrAM7hK6f6 you will see the expander in action in my first example.
@@JasonYadlovski Hey Jason I've got audio from a zoom 32bit F2, and it's really low volume out of the SD card. Would you recommend normalizing it first? Even after the compressor and limiter it struggles to hit -14 lufs, sitting around -23. What's the best thing to do in this case?
Thank you this is what I was looking for! I'm new to Resolve and watched a bunch of normalization videos only hitting a dead end. And I thought there has to be a compressor like you have for a guitar. Great video! Edit: 45mins later I realize if you have a 20+ minute video, it is a pain in the ass to create key frames for each little section that needs raising/lowering. So record the audio in Reaper instead and use the compression there on the entire track in mere seconds. Before pulling it into Resolve and you will save a lot more time. 😎
You can use the compressor right in Resolve for a quicker workflow, it works great. Here's a video about the dynamics panel and the compressor: th-cam.com/video/H08721S3Z38/w-d-xo.html
Dialogue leveler can do a decent job too. I would probably use the dialogue leveler over using normalization. But as an audio guy, I’m always gonna go to the Dynamics and use the compressor and the make up slider versus using the other two tools.
@@JasonYadlovski I’ve been using Dialogue Leveler on my videos. But not an ‘audio guy’ so by the sounds of it, it is a good option. But if you’ve got an ear for it, the method you just demonstrated is better
Hi, can you make a tutorial for reels, on long videos users adjust the volume, but on short videos if it is too high or too low, users will skip the video. it is very important to set the correct volume according to the other reels. I have seen many videos but everywhere they teach how to decrease volume.
I always set my audio levels the same regardless. You might adjust the overall Loudness (LUFS) based on the platform if you need to. Instagram loudness levels is -14 LUFS from what see. All my videos aim for that as well for TH-cam.
I asked that same question on your last video - clearly I wasn't alone. Thanks heaps for the follow up Jason. So I could use normalisation and then fiddle with the dynamics or would that be superfluous use of the normalize? It'd be great if Resolve could introduce an automatic compressor type of thing to accomplish this. Maybe one for the BlackMagic boffins to consider :)
As I do not yet know enough about editing audio this was very helpful. I usually, for TH-cam, normalize to -14 LUFS and then in the Equalizer, turn on Band 1 (100) and Band 6 (13K1) and then as I go through the video tweak as necessary with any key frames. I haven't used the compressor yet (my poor computer might explode).
Thankfully, the compressor does not use a lot of resources from your computer. You should be just fine if you decide to use it. Little by little, keep learning your audio stuff and before you know it, you’ll know as much as I do. 😜
Haha, I've got a 2015 MPB and I Run DR 18 on it. It's not as good as my mac mini, but is still usable. You should get a mac mini, even the base model would be an upgrade for you. 😜
Yeah 18.6 runs in my 2015, but it’s not my daily workhorse. I mainly use it for multitrack recording to resolve. My daily machine is an M1 Mac mini with 16 gigs of RAM and 2 TB of storage.
the maaaaan, the Jasssoooonnn, the loudness war fighteerrrrr , the warrriorrrrr On topic: i found out in time that by the time i add a EQ/preamp plugin and a pair of a 1176 and a 2a compressors and a good limiter/maximizer - there is no more need of normalization because those plugins naturally add the required volume and beautifully shape the sound
Is it wrong? I would first do normalize audio levels to bring everything to the same loudness (LUFS) and then do dynamics to raise the (too) quiet parts and compress/limit the loudest parts.
Sorry for bothering you Jay but can you tell me regards to your other video where you told us about 14 LUFS TH-cam standard if I understood correctly that to make our video loudness comply with TH-cam standards we need to do the following: 1) increase Gain in Dynamics panel of Voiceover Audio track, 2) if it’s not enough then increase gain for the whole Bus1 (by Bus1 I understand all of the audio in our video including voice, music and sounds), 3) use Compressor for Voiceover Audio to make our video sound better after changes we done and to reduce peaks. Did I understood you correctly? Do you have a special video about it (how to make our video loudness comply with 14 LUFS TH-cam standard)?
No bother at all! Here's the video about loudness: th-cam.com/video/SyWFLS4VWvA/w-d-xo.htmlsi=I1CoJrMPc4ECTtdV I've got to make an updated video, but the concepts are still the same as shown here. What you really want to do is set your track(dialogue, sfx and music) levels to where they should be - for example dialogue around -10db. Once you get everything set up on a track level, including dynamics, EQ, and FX, and your mix is good, as the last step, you want to check the overall loudness of your video. Think of 'loudness' as part of the mastering process, not part of the initial track editing. Hopefully that video will help you out.
Hi Jason, another great video. I have a question. When delivering (to TH-cam) quiet nature audio, birds, a quiet stream, gentle breeze, how should the levels be set for that because I don't want them loud, I want them quiet to reflect the stillness of the scene but I do want people to know there is some audio there. I would very much appreciate your take on situations like that.
Hi and thank you! For something like that, while I’m mixing it in my project in DaVinci Resolve, I would probably aim to have the level somewhere around -20 dB or so. If you don’t have anything else going on in your video, you may want to boost the final output (Bus1) just a little bit but you wouldn’t want to bring it all the way up to -14 LUFS because it would be too loud. I think as long as you set your track levels to a good listening level, your export should probably be OK but you’re gonna want to use your ears to really tell you where to set the final output. Maybe even do a few tests and put videos on TH-cam just to see how it sounds, you can always keep them private.
You're welcome! So normalizing is just setting the basic levels. Dynamics would always come after setting levels - and to set audio levels, you'd do it either using 'normalize' or just set it manually by adjusting the gain line or the volume in the inspector. Does that make senese? So it doesn't matter if you 'normalize' or just 'set the levels', it's kinda the same thing.
@@JasonYadlovski Yes, that makes sense. I know you were referring to Davinci (I use Davinci and premiere) but would you happen to know the difference between the "Normailze Max Peak" and "Normalize All Peak" options in Premiere? 🙏
I'm not 100% sure, but I believe that 'normalize max peak' looks at all the clips in the track and normalizes based on the single highest peak and 'Normalize All Peak' will look at each clip in the track, find the peak of that clip and then normalize that clip individually. But not in comparison the the entire track, just relative to that one clip.
A limiter sets a hard limit (depending on the kind of limiter you use) at whatever threshold you set. If you use -3, dB, anything that goes over that he’s going to get stopped at - 3dB
Question, When I LOUDNESS normalize I normalize to -14 or sometimes -18 depending on the application. But recently I was working on a podcast with 2 people and when I normalize to -14 with a peak at -2 it seems to work but then in fairlight the loudness integrated meter reads like -30 LUFS??? Im reading 1770-4 on all the options. Just curious since I figured if I loudness normalized to a level and perform the analysis it should read that level on the integrated meter.
Hey! I'd try changing the meter in Fairlight to be the TH-cam preset if you are measuring to -14. I think the scale might be throwing you off. Also keep in mind that normalizing an audio track is not the same as setting proper loudness levels after all audio is mixed, especially if you have more that 1 track of audio.
@@JasonYadlovski yeah of course, I’m an audio guy at heart so I probably know more than the average Joe. I tried the TH-cam preset and the meters looked correct but I’m not sure why it would be reading -30 or so on the integrated if I had just normalized to -14. And you’re also correct I do more fine tune limiting at the end once everything is mixed. I just like to normalize all my main pieces of audio to get it in the right ballpark of loudness from the start.
Sure you can use both. Normalization is just a volume adjustment. Dynamics works with the dynamics (loud and quiet parts) of the clip and balancing it.
Jason Hello, tell me in version 18.6 in the tab "Progect settings/Fairlight/Audio Metering/Meter type... " Which type (Meter type:IEC, Digital, Costum) of the three should I choose for TH-cam?
I have a question. Would it make sense to normalize the audio after setting up the dynamics to ensure your audio fits the parameters of, for example, youtube? 🇨🇦🦍👊🏻
Hi! Not exactly, by 'normalizing' or setting levels, you're making sure that the signal of your audio that goes through the rest of the chain is at the correct starting point. The dynamics is going to help with the overall dynamic range of the audio. To meet the parameters of TH-cam, you would check that at the very end of the audio process and it's called measuring the overall 'loudness' of your audio. This is a combination of all the audio - dialogue, music, sfx, etc. Hope this helps! I've got videos on all this stuff too if you are interested.
The worst thing is, that the applied compression does not show up in the wave forms. So, it's like flying blind! Question to Blackmagic: Why tf wouldn't you design the compression like in QBase or ProTools, where you can actually see the applied compression visually? That's a huge flaw! So, what I do in these cases is cutting between the loud and soft parts, normalise these parts separately and then use audio transitions. Like in medieval times. Unbelievable, isn't it?
You can see your compression - especially if it’s a plug-in, you just need to cache the audio clips. It’s the same if you have other FX applied to audio clips, if you cache the clips, you’ll see the changes reflected in the waveforms.
I got a question and cannot find the answer to. If you have a low level recording, how to you fix that on resolve? When I boost the gain I get that wild background noise and to get the audio where you can hear it good it jumping into peak red. What are the steps to fix these issues?
You will have to boost your levels either on a clip by clip basis, using the faders or any number of other methods. As you mentioned, it will likely introduce some background noise which you can fix by using something like voice isolation if you have the studio version. If you don’t have the studio version, you can try using the expander to help reduce the background noise while you’re speaking You can also use a little EQ to help get rid of any nasty frequencies. You can also try using the compressor in the dynamics panel by enabling the compression and boosting the make up to help bring back the volume you need.
@@JasonYadlovski think you could do a video on it one day? Lol Imma try this out, but I’m a visual guy. Lol Be some new content for you. Haha Thanks for your help!
Because normalizing only adjust the volume of the audio, it does not apply any dynamics like compression or remove any quieter things you don’t want like the expander does.
When i apply dynamic and through compressor cut high level of audio then on TH-cam when i here audio of that video then it gives that cracking annoying some glitch type sound as and when that high level of audio reduced by compressor or limiter so can you suggest what is the solution for smooth output?
After applying dynamics can you then normalize it also so you get both effects? Or does normalizing it after the fact cause the dynamics to be disabled?
@@JasonYadlovskithank you. Is there any norm how much to increase the volume, the make up in compressor and the make up in limiter? I can go up to -14 but I got lot of noise with my shure :(
First, when you editing your audio, you need to set the track levels properly. I’ve got a video on setting levels in Davinci resolve if you want to check it out. After you’re done with all of your audio editing, added, sound effects, added, music, You need to check the loudness levels. You need to set the loudness level you want to reach then open up the graph on the main bus that will show you where your levels are. That will tell you how much you can raise or lower the main bus out. I do have loudness video that’s kind of old but the process is still the same. Chadwick also has a great video on loudness.
@@JasonYadlovski thank you, yes I got through your video related to levels, they are great and I’ve learned a lot from them. but to get -14 loudness from -10db voice I need to go like 10 on the bus level. So I thought maybe it’s better way to pump it up like little bit with compressor little bit with manual knob and little bit with limiter for better results
If you only have a voice or dialogue track, you can set the levels on the track. Loudness is the measure of all of the audio, including sound, effects, music, dialogue, etc. So it’s gonna be different than whatever your track says. Measuring loudness at the end of the video is the right way to do it, you can boost the overall loudness in a whole, lotta different methods. It’s essentially just raising the volume, using whatever method you want.
In my mind, I don't even think to 'normalize', I just start with adjusting my levels, then working with EQ and Dynamics. I suppose there may be a reason to do both, but if you are balancing your levels already, there's no need to normalize because it's just a volume adjustment....which you'd already be doing by setting your levels, etc.
If setting up dynamics correctly there really shouldn’t be a need for normalizing, because the combination of the compressor (flattening the peaks) and the make up gain should, when set correctly, give you an even sound level and bring it up further. However, the normalize presets for different «end points», like TH-cam or Netflix, are based on their preferred levels, so using dynamics to flatten things out, and then normalize with a preset for the target delivery level, might be a good idea - at least that’s my understanding of why a normalize preset might be a good idea before delivery :)
Doing dynamic equalization and normalizing audio, are two very different thing, that serves different purposes. You should ALWAYS do normalization, while adjusting dynamics is only necessary when needed, like in the clip. And what the manual says, is not entirely correct. It normalize the audio to a given level, say - 14 DBFS for TH-cam, but it also lowers the max. peaks to not go higher, than - 1 dB, so essentially it does a kind of dynamic adjustment of the peaks. In a software like Adobe Audition, the normalize feature only adjust the target level, and not the target loudness, which makes it useless. If you use that software to adjust to, say, -1 dB target level on two different videos, the overall target loudness can end up being very different.
@@akyhne I wish someone would do a video on the DaVinci normalisation and how to use it for making a programme correct for say TH-cam. To me the normalisation has always been a mystery because maybe what they call it for to my mind if you normalise something I would have assumed it did more than just reduce any volume over the setting you put in.
@@DaveKnowlesFilmmaker Each site (TH-cam, Netflix etc.) have different requirements. They are the settings provided by Resolve. So you just choose the right normalization for whatever need you have. For US TV, the target is - 24 dBFS, for Europe it's - 23 dBFS. For non TH-cam, like for personal use, I choose - 23 dBFS, as then my videos are on the same loudness, like most consumer "video", like broadcasting, playing Blu-rays, etc. If you need to deliver to a customer, you'd have to ask them, but in general - 23 or - 24 dBFS is the norm.
Watching you work and hearing the subtle changes you make as you go is like watching a colorist go through their grade and seeing incremental changes to the image. This is elite stuff man.
Im a firm believer that you should always use a compressor and not normalize. Then put a Limiter on Bus 1 to control total peak volume. You wont bring up the low volume parts without compression. Great stuff as always Jason!
I'm with ya! Thanks for watching and sharing your thoughts.
Very good explanations. You're a great tutor. Keep the flame.
Thanks! I appreciate your kind words!
You truly are. "The Audio Guy" and I for one am very grateful you are 😁🤘.
Fantastic as always, Jason 🤘🙏
😜😁 thanks Gary!
Awesome explanation and demonstration of the difference. I am going to continue working on my skills in DaVinci Resolve and your tutorials are helping a lot. Thank you!
Thanks! I appreciate you checking out the video and dropping a comment! If you have any questions on Resolve, drop me a comment somewhere. 😁
If you where to «render out» the video with the audio affected by the compressor and makeup gain, and re-import it into the timeline, showing what the waveform looks like with peaks compressed and the makeup gain applied it might show people more clearly what the dynamics processing does to the waveform vs what normalizing does to the waveform :)
Might be worth a try! I'll think about that. 😜👍
Alternatively, you can just bounce the mix out to a new track. Just make sure you mute the tracks you don't want included
Great video about the distinction Jay. One thing I recently learned from the engineering team on the settings like the new TH-cam normalize preset that has both a loudness and a true peak preset - It pulls the volume up to whichever one of those it hits first. In this case, and most of these it will usually just be the dBTP level. Since it doesn't do any dynamics it can't actually hit -14 LKFS. I think having that loudness part in there is kind of confusing since it would almost never be used since the dBTP would be hit first.
Thanks Chadwick! Good info to know and makes total sense. I didn't think about it hitting one of those first.
Since I saw people had questions on it, I figured I'd drop a video about it. Thanks for checking it out man!
Chadwick I certainly agree with the word confusing especially as there is things like Auphonic around that will automatically adjust the sound to -14LUF (in the case of YouTUbe). I at first thought when DaVinci introduced the normalisation for Netflix and TH-cam it would do this and frankly now I am not at all sure what it is doing or why I would in fact want to use it.
Do you mean the Engineering team of Da Vinci? I neither understand how you could do a normalization of both parameters (TP and Loudness) without doing some dynamics. It is confusing to have bopth parameters on the same interface.
Glad you're making these kind of videos as I'm still learning audio and I'll get there following your tutorials. Thanks!
You’re welcome, glad they’re helpful!
This is the exact video I needed to help me understand the how audio works. I'd love to see more like it explain in detail the other audio settings.
Awesome, glad to hear it was helpful! What type of audio settings would you like to know more about? I’ve got tons of videos about DaVinci Resolve’s audio but I’m always looking for more ideas for videos. 😜
This is one of best places for Davinci Resolve audio tips, so many times my search has led me here and problem solved. Thanks!
Awesome! Doing my best to help the community take their audio to the next level. 😜👍😁
Beginner here, excellent information about dynamics and how to use the Fairlight page, thank you so much!!
You’re welcome! Thanks for watching!
Thank again Jason, I'm learning a lot from you. May be a quick explanation, but it opens doors to more knowledge.
Epic E.T. Intro
Thanks Jason, I had no idea the make up was for the quiet parts only I thought that was another type of gain slider that affected everything on the track.
You’re welcome! The makeup is a great tool, I use it all the time, now you can use it too! 😜👍
Well that's a coincidence Jason.
Tonight I am doing some testing to see what effect ticking the Normalise Audio check box has on the TH-cam render preset.
Well now I know.
My present videos are basically Game sound output on 1 audio track and my mic audio on another and I want to make sure my levels are good and don't get turned down by TH-cam.
I think getting my levels set before recording my footage is the way to go, followed by checking and adjusting the dymanics if required to make sure things are where I want them to be before I upload.
Thanks for another detailed video explanation.
👍
I think you are certainly thinking about your process in the correct way. Basically you won't need to use the normalize audio on export if you set your levels and check your loudness before you render.
Hello Jason, I appreciate all the great content you are making for years, I learned a lot from your tutorials, from basic to some advanced techniques and improved my audio a lot.
The question that I have is would you recommend using "Normalize audio" before or after using Dynamics? In my head it makes sense to use it after dynamics just to set it in certain loudness range instead of measuring it and doing it manually for some quick delivery interviews and similar, but I might be very wrong...
Thanks once again for your videos its amazing library of knowledge!
audio is definitely the hardest part of recording anything. getting mic settings right in the first place so that it's usable, and then balancing everything out in post. i've been trying to make highlight clips for years and still run into audio issues because i just dont understand it
I hear ya! For your mic settings, I'd recommend to try setting it around -18dB, that way you've got plenty of headroom to boost it a bit in post. In Post I like to aim for -10dB for my dialogue track and at the end of the audio process I make sure I'm just under -14 LUFS for overall Loudness for TH-cam.
What a great help, as always, Jason. I wonder if you have experienced some problems in the Fairlight tab after the last mac update to Sonoma 14.1, for me, it's almost impossible to use that tab, quite difficult.
Thanks! I haven’t updated to Sonoma yet. I’m still on Ventura. I usually wait quite a while before I update the OS. That gives the software companies time to catch up.
@@JasonYadlovski That’s my lesson to learn!
You just taught me what I’ve been trying to learn for 2 years ty
Awesome!
Hey Thanks! This helped a lot in my Resident Evil Video! Preesh the knowledge drop!
You're welcome!
In my case, I'd be a bit concerned when making a music video with a track that has already been mastered. I would think normalizing a mastered track that has some dynamic parts that are -1 to -2dB in volume (such as in the chorus sections) and if the song has already been keyed to TH-cam -14 LKFS target level. I think that nomalizing that audio would flatten out the dynamics because they are already so close together.
I hear ya, keep in mind that normalizing is just adjusting the overall volume, so it doesn’t really change any of the dynamic parts of a music track. It’s essentially the same as turning the volume on your radio up and down.
Thanks as always Jason! If I look at how the graph moves, the Make Up slider in the compressor doesn’t just make the quiet parts louder, but it makes all parts louder including the loud parts. Am I wrong? I didn’t post this comment to nitpick, just because I hope to learn and because others might benefit too.
Thank you for the explanation, Jason! I love it!
Glad it was helpful!
For some reason i always thought Normalize was doing some compression too. Good to know it's useless. I always compress and adjust volume afterward anyway.
Sounds like you are on the right track. No compression w/ normalization, it's just a volume adjustment. 👍
I wanna make sure i understand the workflow or chain of audio manipulation.
So for audio work, especially voiceovers, the first step of the process should be to normalize LUFS of the entire project according to TH-cam standards for example and then and only then do you move on to using compressors/limiters to fine tune the dynamics?
You’ve actually got it backwards. First set proper/desired levels for all tracked based on track meters in mixer. Then move on to eq, dynamics, then effects. After mixing all tracks - dialogue, music and sound effects, then you check the overall loudness (LUFS) to meet TH-cam standards. You would use the loudness meters in Fairlight and/or the graph loudness meter on bus1 in the timeline tracks section of Fairlight. Does this help?
Again, very clear. Questions: 1) I assume it's worth normalising the whole track at the end (after dynamics etc) to hit a sensible TH-cam level? (I see this can now be done at the render stage) 2) I saw you also moving some expander settings - would love a video on this at some point. I think you did one before but I am finding these newer ones much easier to understand.
Thanks! I don’t think there’s any reason to normalize if you set all of your audio levels properly and use dynamics properly. You’ve essentially done a better job at getting your levels, correct then simply normalizing would do, in my opinion. I do have some videos on the gate and expander, but I also have a video that I recently put out, showing how the expander works, here’s a link to it: th-cam.com/video/tqgUA44Nd9k/w-d-xo.htmlsi=uDb8JghrAM7hK6f6 you will see the expander in action in my first example.
@@JasonYadlovski Thanks Jason
@@JasonYadlovski Hey Jason I've got audio from a zoom 32bit F2, and it's really low volume out of the SD card. Would you recommend normalizing it first? Even after the compressor and limiter it struggles to hit -14 lufs, sitting around -23. What's the best thing to do in this case?
@JasonYadlovski Thank you, I ruined my previous videos by not applying those audio settings, that will do it.
Thanks 🙂
You’re welcome!
Hey bud, I'm new to DaVinci and I enjoy your videos muchly.
Awesome, glad you're enjoying my videos! Thanks for being a part of the community!
@@JasonYadlovski I’m Scottish and I call you The Baldy Bard. It’s definitely a compliment.
Haha! Nice. I love being bald, hair is a pain to deal with! I just got a watch out for sunburn. 😜
@@JasonYadlovski Being Scottish and as pale as a bucket I’m completely with you on that
Great video, thanks!
You're welcome!
Thank you this is what I was looking for! I'm new to Resolve and watched a bunch of normalization videos only hitting a dead end. And I thought there has to be a compressor like you have for a guitar. Great video!
Edit: 45mins later I realize if you have a 20+ minute video, it is a pain in the ass to create key frames for each little section that needs raising/lowering. So record the audio in Reaper instead and use the compression there on the entire track in mere seconds. Before pulling it into Resolve and you will save a lot more time. 😎
You can use the compressor right in Resolve for a quicker workflow, it works great. Here's a video about the dynamics panel and the compressor: th-cam.com/video/H08721S3Z38/w-d-xo.html
What about the "Dialog Leveling" (if I remember the name of it right)? The control sits right next to the dialog isolation control in the inspector.
Dialogue leveler can do a decent job too. I would probably use the dialogue leveler over using normalization. But as an audio guy, I’m always gonna go to the Dynamics and use the compressor and the make up slider versus using the other two tools.
@@JasonYadlovski Thanks! 🙂
@@JasonYadlovski I’ve been using Dialogue Leveler on my videos. But not an ‘audio guy’ so by the sounds of it, it is a good option. But if you’ve got an ear for it, the method you just demonstrated is better
Awesome explanation ! Thank You for sharing Your knowledge :)
Glad it was helpful! Thank you for watching!
Hi, can you make a tutorial for reels, on long videos users adjust the volume, but on short videos if it is too high or too low, users will skip the video. it is very important to set the correct volume according to the other reels. I have seen many videos but everywhere they teach how to decrease volume.
I always set my audio levels the same regardless. You might adjust the overall Loudness (LUFS) based on the platform if you need to. Instagram loudness levels is -14 LUFS from what see. All my videos aim for that as well for TH-cam.
I asked that same question on your last video - clearly I wasn't alone. Thanks heaps for the follow up Jason.
So I could use normalisation and then fiddle with the dynamics or would that be superfluous use of the normalize?
It'd be great if Resolve could introduce an automatic compressor type of thing to accomplish this. Maybe one for the BlackMagic boffins to consider :)
As I do not yet know enough about editing audio this was very helpful. I usually, for TH-cam, normalize to -14 LUFS and then in the Equalizer, turn on Band 1 (100) and Band 6 (13K1) and then as I go through the video tweak as necessary with any key frames. I haven't used the compressor yet (my poor computer might explode).
Thankfully, the compressor does not use a lot of resources from your computer. You should be just fine if you decide to use it. Little by little, keep learning your audio stuff and before you know it, you’ll know as much as I do. 😜
@@JasonYadlovski I’m still using DR 16 as I’m scared to upgrade, my computer is ancient (2015 MBP i5 8GB RAM) 😂
Haha, I've got a 2015 MPB and I Run DR 18 on it. It's not as good as my mac mini, but is still usable. You should get a mac mini, even the base model would be an upgrade for you. 😜
@@JasonYadlovski you run 18.6 on a 2015!? Oh the mini has been considered for sure.
Yeah 18.6 runs in my 2015, but it’s not my daily workhorse. I mainly use it for multitrack recording to resolve. My daily machine is an M1 Mac mini with 16 gigs of RAM and 2 TB of storage.
Would normalize and at the same time have a limiter that pushes the loudest parts back down achieve something similar to compression?
Kind of, but a limiter is like a wall. A compressor is more gradual reduction. I'd always go w/ a compressor, not a limiter for dialogue/vocals.
the maaaaan, the Jasssoooonnn, the loudness war fighteerrrrr , the warrriorrrrr On topic: i found out in time that by the time i add a EQ/preamp plugin and a pair of a 1176 and a 2a compressors and a good limiter/maximizer - there is no more need of normalization because those plugins naturally add the required volume and beautifully shape the sound
Ahhhh! 💪 Haha, just trying to answer everyones questions. 😜
Yeah, if you are working with your audio already by adjusting levels and dynamics, there is no need to normalize in my mind....
Does the compressor work in the free version of Davinci? The red button/slider didn't appear to activate.
Great tutorial.
Yes it works in the free version also. 👍
Is it wrong? I would first do normalize audio levels to bring everything to the same loudness (LUFS) and then do dynamics to raise the (too) quiet parts and compress/limit the loudest parts.
Thx. I'm climbing the audio mnt.
Awesome! Audio is a lot of fun to learn. 😜
Sorry for bothering you Jay but can you tell me regards to your other video where you told us about 14 LUFS TH-cam standard if I understood correctly that to make our video loudness comply with TH-cam standards we need to do the following: 1) increase Gain in Dynamics panel of Voiceover Audio track, 2) if it’s not enough then increase gain for the whole Bus1 (by Bus1 I understand all of the audio in our video including voice, music and sounds), 3) use Compressor for Voiceover Audio to make our video sound better after changes we done and to reduce peaks.
Did I understood you correctly? Do you have a special video about it (how to make our video loudness comply with 14 LUFS TH-cam standard)?
No bother at all! Here's the video about loudness: th-cam.com/video/SyWFLS4VWvA/w-d-xo.htmlsi=I1CoJrMPc4ECTtdV
I've got to make an updated video, but the concepts are still the same as shown here. What you really want to do is set your track(dialogue, sfx and music) levels to where they should be - for example dialogue around -10db. Once you get everything set up on a track level, including dynamics, EQ, and FX, and your mix is good, as the last step, you want to check the overall loudness of your video.
Think of 'loudness' as part of the mastering process, not part of the initial track editing. Hopefully that video will help you out.
@@JasonYadlovski thank you!
But if I use 10 different songs for a 40 Minute slide show, normalizing should be the right tool ?
You can do that, it would work just fine. 👍
normalize on my end is pretty buggy for some reason it blasts random clips to +30 without any spikes or super quiet parts being present really weird
It might be a factor of the method of normalization.
Very good as always thank you
You’re welcome!
Hi Jason, another great video. I have a question. When delivering (to TH-cam) quiet nature audio, birds, a quiet stream, gentle breeze, how should the levels be set for that because I don't want them loud, I want them quiet to reflect the stillness of the scene but I do want people to know there is some audio there. I would very much appreciate your take on situations like that.
Hi and thank you! For something like that, while I’m mixing it in my project in DaVinci Resolve, I would probably aim to have the level somewhere around -20 dB or so.
If you don’t have anything else going on in your video, you may want to boost the final output (Bus1) just a little bit but you wouldn’t want to bring it all the way up to -14 LUFS because it would be too loud.
I think as long as you set your track levels to a good listening level, your export should probably be OK but you’re gonna want to use your ears to really tell you where to set the final output. Maybe even do a few tests and put videos on TH-cam just to see how it sounds, you can always keep them private.
Thanks! Is there ever a case where you would do both or would you always just do the dynamic compressor and not normalize?
You're welcome! So normalizing is just setting the basic levels. Dynamics would always come after setting levels - and to set audio levels, you'd do it either using 'normalize' or just set it manually by adjusting the gain line or the volume in the inspector. Does that make senese? So it doesn't matter if you 'normalize' or just 'set the levels', it's kinda the same thing.
@@JasonYadlovski Yes, that makes sense. I know you were referring to Davinci (I use Davinci and premiere) but would you happen to know the difference between the "Normailze Max Peak" and "Normalize All Peak" options in Premiere? 🙏
I'm not 100% sure, but I believe that 'normalize max peak' looks at all the clips in the track and normalizes based on the single highest peak and 'Normalize All Peak' will look at each clip in the track, find the peak of that clip and then normalize that clip individually. But not in comparison the the entire track, just relative to that one clip.
@@JasonYadlovski Thanks again. That helps a lot!
Hello, is there a way to apply Limiter to the whole timeline instead of individual tracks? ... Let's say I just to my whole video to be just -3?
Sure, drop it on Bus1, the main out.
@@JasonYadlovski thank You! Will it reduce the whole waveform ir just cut out the peaks?
A limiter sets a hard limit (depending on the kind of limiter you use) at whatever threshold you set. If you use -3, dB, anything that goes over that he’s going to get stopped at - 3dB
So to answer your question, just a peaks that go over where are you set the limiter
Question, When I LOUDNESS normalize I normalize to -14 or sometimes -18 depending on the application. But recently I was working on a podcast with 2 people and when I normalize to -14 with a peak at -2 it seems to work but then in fairlight the loudness integrated meter reads like -30 LUFS??? Im reading 1770-4 on all the options. Just curious since I figured if I loudness normalized to a level and perform the analysis it should read that level on the integrated meter.
Hey! I'd try changing the meter in Fairlight to be the TH-cam preset if you are measuring to -14. I think the scale might be throwing you off. Also keep in mind that normalizing an audio track is not the same as setting proper loudness levels after all audio is mixed, especially if you have more that 1 track of audio.
@@JasonYadlovski yeah of course, I’m an audio guy at heart so I probably know more than the average Joe. I tried the TH-cam preset and the meters looked correct but I’m not sure why it would be reading -30 or so on the integrated if I had just normalized to -14. And you’re also correct I do more fine tune limiting at the end once everything is mixed. I just like to normalize all my main pieces of audio to get it in the right ballpark of loudness from the start.
helpful
Glad it helped!
Can you use both ? use the dynamics and then normalize to the level you like or vice versa??? thanks
Sure you can use both. Normalization is just a volume adjustment. Dynamics works with the dynamics (loud and quiet parts) of the clip and balancing it.
Thjanks a bunch@@JasonYadlovski
Jason Hello, tell me in version 18.6 in the tab "Progect settings/Fairlight/Audio Metering/Meter type... " Which type (Meter type:IEC, Digital, Costum) of the three should I choose for TH-cam?
I have a question.
Would it make sense to normalize the audio after setting up the dynamics to ensure your audio fits the parameters of, for example, youtube?
🇨🇦🦍👊🏻
Hi! Not exactly, by 'normalizing' or setting levels, you're making sure that the signal of your audio that goes through the rest of the chain is at the correct starting point. The dynamics is going to help with the overall dynamic range of the audio. To meet the parameters of TH-cam, you would check that at the very end of the audio process and it's called measuring the overall 'loudness' of your audio. This is a combination of all the audio - dialogue, music, sfx, etc. Hope this helps! I've got videos on all this stuff too if you are interested.
@@JasonYadlovski thank you for the information.
The worst thing is, that the applied compression does not show up in the wave forms. So, it's like flying blind! Question to Blackmagic: Why tf wouldn't you design the compression like in QBase or ProTools, where you can actually see the applied compression visually? That's a huge flaw! So, what I do in these cases is cutting between the loud and soft parts, normalise these parts separately and then use audio transitions. Like in medieval times. Unbelievable, isn't it?
You can see your compression - especially if it’s a plug-in, you just need to cache the audio clips. It’s the same if you have other FX applied to audio clips, if you cache the clips, you’ll see the changes reflected in the waveforms.
I got a question and cannot find the answer to. If you have a low level recording, how to you fix that on resolve? When I boost the gain I get that wild background noise and to get the audio where you can hear it good it jumping into peak red. What are the steps to fix these issues?
You will have to boost your levels either on a clip by clip basis, using the faders or any number of other methods. As you mentioned, it will likely introduce some background noise which you can fix by using something like voice isolation if you have the studio version. If you don’t have the studio version, you can try using the expander to help reduce the background noise while you’re speaking
You can also use a little EQ to help get rid of any nasty frequencies. You can also try using the compressor in the dynamics panel by enabling the compression and boosting the make up to help bring back the volume you need.
@@JasonYadlovski think you could do a video on it one day? Lol Imma try this out, but I’m a visual guy. Lol Be some new content for you. Haha Thanks for your help!
If you wanna send me a clip of your audio, I’ll see if I can make a video showing how to edit it. 😜
@@JasonYadlovski yeah for sure.
you can email it to me or send me a link to it: jason.yadlovski@gmail.com
if u solo a track u don't have to mute the other track
Yup, different strokes for different folks. Lots of ways to do things in Resolve. 😜
Very informative, thank you
why not normalize and then apply the expander and compressor on the normalized audio?
Because normalizing only adjust the volume of the audio, it does not apply any dynamics like compression or remove any quieter things you don’t want like the expander does.
well, this just saved me a ton of disappointing results. Thanks for the clarification.
When i apply dynamic and through compressor cut high level of audio then on TH-cam when i here audio of that video then it gives that cracking annoying some glitch type sound as and when that high level of audio reduced by compressor or limiter so can you suggest what is the solution for smooth output?
After applying dynamics can you then normalize it also so you get both effects? Or does normalizing it after the fact cause the dynamics to be disabled?
What’s the best order for eq fx dy?
I like to do this - Set Levels, EQ, Dynamics, FX.
@@JasonYadlovskithank you. Is there any norm how much to increase the volume, the make up in compressor and the make up in limiter? I can go up to -14 but I got lot of noise with my shure :(
First, when you editing your audio, you need to set the track levels properly. I’ve got a video on setting levels in Davinci resolve if you want to check it out. After you’re done with all of your audio editing, added, sound effects, added, music,
You need to check the loudness levels. You need to set the loudness level you want to reach then open up the graph on the main bus that will show you where your levels are. That will tell you how much you can raise or lower the main bus out. I do have loudness video that’s kind of old but the process is still the same. Chadwick also has a great video on loudness.
@@JasonYadlovski thank you, yes I got through your video related to levels, they are great and I’ve learned a lot from them. but to get -14 loudness from -10db voice I need to go like 10 on the bus level. So I thought maybe it’s better way to pump it up like little bit with compressor little bit with manual knob and little bit with limiter for better results
If you only have a voice or dialogue track, you can set the levels on the track. Loudness is the measure of all of the audio, including sound, effects, music, dialogue, etc. So it’s gonna be different than whatever your track says. Measuring loudness at the end of the video is the right way to do it, you can boost the overall loudness in a whole, lotta different methods. It’s essentially just raising the volume, using whatever method you want.
Is there a case to be made for dynamics + normalize meaning once you balance your audio can you then normalize it ?
In my mind, I don't even think to 'normalize', I just start with adjusting my levels, then working with EQ and Dynamics. I suppose there may be a reason to do both, but if you are balancing your levels already, there's no need to normalize because it's just a volume adjustment....which you'd already be doing by setting your levels, etc.
If setting up dynamics correctly there really shouldn’t be a need for normalizing, because the combination of the compressor (flattening the peaks) and the make up gain should, when set correctly, give you an even sound level and bring it up further. However, the normalize presets for different «end points», like TH-cam or Netflix, are based on their preferred levels, so using dynamics to flatten things out, and then normalize with a preset for the target delivery level, might be a good idea - at least that’s my understanding of why a normalize preset might be a good idea before delivery :)
Doing dynamic equalization and normalizing audio, are two very different thing, that serves different purposes.
You should ALWAYS do normalization, while adjusting dynamics is only necessary when needed, like in the clip.
And what the manual says, is not entirely correct. It normalize the audio to a given level, say - 14 DBFS for TH-cam, but it also lowers the max. peaks to not go higher, than - 1 dB, so essentially it does a kind of dynamic adjustment of the peaks.
In a software like Adobe Audition, the normalize feature only adjust the target level, and not the target loudness, which makes it useless. If you use that software to adjust to, say, -1 dB target level on two different videos, the overall target loudness can end up being very different.
@@akyhne I wish someone would do a video on the DaVinci normalisation and how to use it for making a programme correct for say TH-cam. To me the normalisation has always been a mystery because maybe what they call it for to my mind if you normalise something I would have assumed it did more than just reduce any volume over the setting you put in.
@@DaveKnowlesFilmmaker Each site (TH-cam, Netflix etc.) have different requirements. They are the settings provided by Resolve. So you just choose the right normalization for whatever need you have.
For US TV, the target is - 24 dBFS, for Europe it's - 23 dBFS.
For non TH-cam, like for personal use, I choose - 23 dBFS, as then my videos are on the same loudness, like most consumer "video", like broadcasting, playing Blu-rays, etc.
If you need to deliver to a customer, you'd have to ask them, but in general - 23 or - 24 dBFS is the norm.
Thanks for that video Jason. That helped me a lot with understanding the different concepts!
Very welcome!