This video really saved me from a lot of obstacles I put myself through trying to edit my voice over audios with minimal knowledge! Thank you so much for sharing what you know, seriously I could cry with relief!
I've just started a contract doing voice over for a nonprofit, and because of your videos, they are surprised I am a newbie :D Thanks for making these things easy to understand and follow!!
Life saver, had a combined effect of Hard Lim, SBC and Normalise and just wasn't fixing my levels at all and the speech leveler before the combined favorite was perfect. Cheers Mike, best audio expert on YT.
Oh thank you so much, you have just saved my sanity as I'm learning how to be my own podcast audio engineer. When I had a radio show, the station provided the engineer... so your tutorials (especially this one) are priceless!
Thank you very much for your free service. The Sir who is training me on Podcasts Production gave me some links to your videos to learn more. And they are really helpful. I am grateful for both him and you Sir Russel.
Used an old (legal!) AA v3 for years now and now bringing myself uptodate with how AA has developed. I was trying out Soundforge - the lack of tutorials and support for that put me off, never mind the interface. Your vids persuaded me to go with AA - seems like almost anything I need to know, you have the answer. Great channel, thank you.
Thanks Phil. And Adobe CC is a great suite of apps too. Audition 3.0 was a great editor but the advances between there and CC2020 are huge and you'll find yourself being able to produce better and quicker with the latest release.
I totally understand your concern! If you're looking to even out volume levels in normal recordings without amplifying breaths using a hard limiter in Adobe Audition can be a great alternative workaround but remember that’ll still bring breaths in line with speech volume if used incorrectly.
Again, great and clear video! Thanks a lot for your content! Fast question: what happens if you don't normalise before using the speech volume leveller? What is the reason behind normalising it if then you are going to level the volumes?
When you set the speech volume leveler up you assign a target volume level. By normalising the audio first you're leaving the speech volume leveler less to do and it can concentrate on making the audio a more consistent level.
sweet, this was the most thorough video I've seen on this thus far. Thanks! I'm using the Zoom F3 and loving it. But, I'm usually in less than ideal shooting circumstances. This may be very helpful. But, I'm feeling like I may like to first do some manual bringing down of hot peaks before trying to normal, maybe... Anywho, thanks again!
@@MikeRussell Quick question is it better to edit the audio in audition before audition before ever bringing it into premiere? Because when I right click the audio on premier to edit it in audition premiere says "render and replace". Does this inhibit some of my quality potential is sleep potential with 32 bit head room and what not? If I just go straight to audition 1st the audio goes straight into audition fast without having to be rendered and replaced or extracted by premiere.
I actually had a question. I know recently you spoke about De-Noise. This is just an audio of myself, of course there are some noise backgrounds and the eventual friend gets picked up. Should I be using De-noise, THEN Normalize & Speed Volume Leveler? Or is it overkill? I also have yet to put in effects and I'm also not sure as to what order is best (effects after my noise reductions or before it). I feel like every time I'm trying to learn a new way to make my audio sound nice and crisp I get scrambled with the order of the process.
Hi, the order would be to repair the audio if necessary then apply the enhancing effects. You'd want to remove background noise first before applying the speech volume leveler as otherwise the speech volume leveler will suck up the background noise also while leveling the speech.
Thanks for sharing this video Mike! Your videos have really enhanced my audio production. A question I had about this process is that if you have an audio clip with 5 different voices (at different levels), do you need to normalize each voice, or do you just normalize the whole clip? Cheers. Tom
Depending on how your clip is, if each person says their bit then I would suggest splitting the clip between each clip, using match clip loudness and then the speech volume leveller. If the voices are all together in the clip then run the speech volume leveller to get a much more consistent level out of your audio.
Speech Volume leveler is an extremely harsh tool, I'd say. It's best to just slice the track on the Multitrack with a Razor, and even the pieces there with Match Clip Loudness, and then adjust them by hand.
Thank you for putting out these tutorials. Really useful. I have a couple of questions if you don't mind. If you are cutting together lots of different takes (for a scripted project), would you apply the normalization / speech volume leveler to each clip before beginning editing? Or wait until you have a finished cut so as not to adjust clips which might not be used? Many thanks.
I would suggest putting all the clips into multitrack and then applying match clip loudness as a start to all the clips. You can then add the speech volume leveler in the effects rack.
The speech volume leveler will not introduce noise but may amplify noise that already exists quiety in the recording. Best to ensure the recording is quiet at source or remove any background noise before applying the speech volumer leveler.
Thank you, great trick! Can I ask a question? What is the difference between speech volume leveler and match loudness? I use match loudness for my podcast (with 3-man talkshow) with -19dB target loudness (mono podcast). Last question: can I use on voice bus track? Thank you for your answer!
Hi Robert, the speech volume leveler can be assigned to a bus. It's a powerful tool which will compress the audio based on the settings you specify to try and produce a more level sound. For a podcast I recommend not using the speech volume leveler too harshly, don't overdo the leveling amount and keep the dynamic range on the higher side too. Match loudness will match the the selected clip(s) to a target set. It's a good starting point before processing your audio. I would recommend running the final mix down through match loudness to ensure you hit the podcast loudness standard which is generally accepted as -16lufs for podcasts or -14lufs for youtube.
Hi Mike, 1st question - for that podcast clip with background noise - why don't you just edit that clip alone in Waveform mode using the noise reduction process effect? wouldn't that be better? 2nd question, now that all clips are even, since they are placed in one track instead of separate tracks, how do we export them individually?
Yes, you should have tidied up the clip before processing it but I was using the clip to demonstrate how the speech volume leveller could be problematic when used with clips recording in less than perfect conditions. To answer your second question you can select a clip and mix down just that clip.
I can't seem to do this without also decreasing the volume of the quieter speech in the same track. I only need the loud peaks to be brought down. Nothing else. How do I do this?
To preserve the quieter speech while reducing only the loud peaks you should use a technique called "peak limiting" instead of normalizing or a speech volume leveler. In like Adobe Audition you can apply a hard limiter to the track.
Hi Mike! Podcaster/audio newbie here....This video is very helpful but do you have a video that explains more about how to balance interview audio levels on the same track? You demonstrate this a bit in the "podcast" clip where one person is louder and the other person is much quieter. Any suggestions on how to balance extreme audio recordings in the same track? Been digging through your other content but can't seem to find the right one. Thanks and keep up the great work!
DO you think that doing what maule mode is good idea? ( selecting part of Speech and reducing or increasing to for example -3db) I want to do mix to my film...
I gave up on the speech leveler for podcasts because I'd find some soft sounds in speech end up being leveled up to sound noticeably unnatural, even with very tame settings. You don't find that happens?
Mike, fantastic videos, especially for someone like me who is drinking from the proverbial firehouse! I have a question. I’ve heard you say a billion times that waveform is destructive, so I’m never edit in waveform. But in your normalization at the beginning of this video, you go into waveform for the three tracks to normalize each at -3dB. Is there a way of doing this easily in multitrack so that I’m never modifying the original recorded track?
You cannot normalise in multitrack, only in waveform. Normalising is perfectly safe, as long as you don't exceed 0db, as it just changes the amplitude of the track but does not alter dynamics. You can however match clip loudness in multitrack.
@@MikeRussell thanks for the fast response. So, I’m not certain how to execute to your ENCN workflow. EQ in multitrack, then normalize the result in waveform, then import back into multitrack to compress, then normalize again in waveform for the final version of the audio?
It all depends on what the intended use for the audio is. For example radio imaging would be more processed than a podcast. But radio imaging can't be over processed though as it will go through a processing chain between the desk and the transmitter.
Hi Mike, thank you for these videos, they helped a lot. I was wondering if there is a way to extreme the audio waves automathic. What I mean is, I have my records, and for after effects purpose I would like to generate a sample of my track were the more silent space between the sylabs of a speech go verry low in volume, while the highs keep the same consistent, to pop out the difference bettween sound up and sound down. is there a way to force it like that?
Would you loudness normalize over peak normalize? Just because there could be a large peak somewhere that would throw off any peak normalization. Loudness normalization seems to make more sense then apply the leveler. Just curious. As when I do podcasts someone could laugh or something and will be crazy higher than the regular speech. If this is the case what level would you loudness normalize too. One last thing, say you loudness normalize but that makes something like a laugh peak. If you apply a compressor or volume automation would that laugh still be peaked or does the software know the original clip isnt peaked and bring that laugh level down without peaking. Sorry this is so technical but it gets into more advanced stuff.
When it comes to normalizing audio levels in Adobe Audition, loudness normalization is generally preferred over peak normalization to ensure a more balanced overall volume, especially for podcasts where speech levels can vary.
Ok cool thats basically what I thought as well, so if you do loudness normalize and like a laugh or cough peaks over 0 would the software know if you add a compressor that the clip isnt clipped or just by normalizing would it still be clipped@@MikeRussell
Hi, Mike! Until I found your channel I was very good at video but absolute rubbish at audio! It simply went over my head completely and I couldn't get good results. Now, well, I'm decent but learning, thanks to you! *I was wondering if you prefer the Audition's SVL over Accusonus' ERA 5 Voice Lever or vice versa?* I'm looking to level a person's voice who sometimes really emphasises some a couple of syllables in the entire recording which preclude the use of normalisation, as it often even drops down the rest of the recording in order to get those momentary under control. When using Accusonus' VL I don't quite get the result I'm after, as it tends to make the rest of the audio sound a bit weird. The "low" parts of the audio are correct, it's those outlying peaks that I want to get rid of. And when I use Audition's SVL on it, no matter what I do (in the range that does decently balance the level) I get thos magnified breaths at the end of sentences and such. So I'm wondering if there's any way of bringing just the peaks down, instead of passing the entire audio though a process that brings up the level of the rest of the audio. Again thanks a lot for these wonderful tutorials!
Each has their own use. The speech volume leveller if set up correctly is a fantastic tool but if set badly can ruin your audio. ENCN is a method of applying "polish" to your audio via EQ and compression. Both the speech volume leveller and "ENCN" will alter the dynamic range of your audio. Match loudness however will use a perceived loudness technique to match the audio to a target loudness on the LUFS scale. This does not apply compression and therefore doesn't alter the dynamic range of you audio. How you process voiceovers depends on whether or not you have a spec to produce to. If you do it's likely you'll be told what, if any processing is to be done and what target the audio should be at. If you are recording for someone else to produce then as little as possible processing is best as it's the producer's job to produce the audio and he'll want clean audio to work with. If you're producing the voiceover then I've used match loudess, speech volume leveller and ENCN all on the same audio before but there is no right or wrong answer as it depends on what your voiceover is and it's target destination.
Hey, thanks for the video. I'm a real newbie with this world but I just started now fixing some old recordings of speeches and I'm trying to understand the ropes. Why would you also use a normilzer and a speech volume leveler, wouldn't the speech volume leveler be enough? I also wanted to ask about the hard limiter, is there any point of a normalizer if I apply a hard limiter? And again is there any difference between hard limiter and the speech volume leveler?
There's a difference between all three. Normalise doesn't affect the dynamic range of the audio at all, it only find the loudest part of the audio and increases/reduces all the audio between the level of the loudest part of the audio and the level you set as to be normalised to. The speech volume leveller is basically a compressor. It levels the audio within the settings you specificy but reducing the dynamic range making the loudest and quitest parts of the audio closer in volume. The hard limiter at it's simplest basically justs any audio that goes above the level you tell it to cut at. However with the input boost it can be set as a virtual compressor. Withouth hearing the quality of the old audio recordings I can't really advise you much further as if the audio quality isn't the best the speech volume leveler could potentially highlight any unwanted noise.
Hi Mike...been watching your videos for years. You do a great service Sir. One question, if I normalize to -3, then I start clipping. Is that normal? Would that mean that my audio has been recorded at a high enough level where a normalization of say... -8 is more appropriate? I'm confused as to why the presets in the Favorites are automatically set to -3 or -1.5. Thanks!
Clipping does not occur until you exceed 0db. What this tends to indicate is that there is already analogue clipping in your audio where, at some stage of your audio chain you’ve exceed 0db.
Great video, Mike! I had a question.. I am putting together a gun fight scene with music playing over it. I want to make sure that the gun sounds don't overpower the music or vise versa. Is this the type of process I would use to avoid that happening? If not, could you possibly do a tutorial on that or point me in a direction? I'd greatly appreciate it.
Hi Phil, I would suggest running match clip loudness to make all your clips the same perceived loudness and then use essential sound. You can assign tracks as music, voice or fx and Audition will then create the final mix for you. Hopefully this helps ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-Lnp-6sR_5Iw.html
Asking for some help here, I have an audio file from a movie that is distorted in ANY program I play it, like a robotic voice, but when I play it on VLC PLAYER changing the "Preamp" levels on the EQ it solves the problem (downing the levels to be specific) the thing is that I can't find that "Preamp" option in any other program or editing program. my question is, is there any way to fix the audio permanently? By editing it so I can play it on any other video/audio program or my TV etc thank you .
Unfortunately, without more specific information about the audio file it is difficult to provide a precise solution. There are DeClipper effects in Audition. That's a good start.