This is literally the only video anyone wanting to learn about gain staging at the recording phase should watch. Absolutely simple and perfectly explained.
Normalized "good" -10 level 4:47 Normalized "bad" -40 level 6:57 I have to turn up everything pretty loud to be able to hear the noise over the signal. Is that bad hearing (of the high frequencies) or to be expected? I found the bad sample sounded slightly rougher/ harsh, but the noise is not audible at normal (or lower) listening levels.
So, -10 or -12 dbfs is a good peak signal for gainstaging, right? For basically any source? Same for percussive and non-percussive? And what do you think about doing the gain staging to 0 dbvu without going over 0? What is better?
It depends on how predictable your source is. -10 dBFS is good for voice. It doesn't have to be exact but the meter should hover around -10 on peaks. The aim is to be completely confident you will never get an over which will ruin your take. You can concentrate on helping your singer perform well. But if you were for instance recording from a drum machine, which is totally predictable, then there is no reason not to go all the way to 0 dBFS, just a little below clipping. The higher level you record, the better the signal-to-noise ratio, but with 24-bit audio it's hard to tell the difference between -10 and 0. DM
@@AudioMasterclass Hi, some people say that even with a predictable signal, this must be -12dbfl tops, because of the plugins you put afterwards. Is this so?
@@preciseaudioblog The saying that 'two wrongs don't make a right' comes to mind. 12 dB of headroom is OK in a 24-bit recording which makes it a 22-bit recording, but your performer might go crazy and the level goes up to -1 dBFS or maybe even -0.1 dBFS. Don't panic, everything is OK. Any plug-in that can't handle 0 dBFS at least has not been competently designed. But consider that a plug-in that emulates hardware may also emulate that hardware's clipping characteristics. Or in the case of tube equipment a progressive increase in distortion towards clipping. The key is to listen. If you like what you hear then great. If you don't and you suspect the issue is level then use a gain plugin before the plugin that doesn't sound the way you want. Other than for the purpose of headroom there is never any reason to record at a low level. DM
Celebrate! Your level is perfect with 6 dB headroom that kept you safe from clipping. With headroom, like many things, it's always better to have it and not need it than need it and not have it. Your RMS measurement compared to peak is governed by the dynamic range of the performance and it is only the peak measurement that you should consider when setting gain. DM
The one that was just about under the threshold of peaking. Imma say on most preamps, it will have distorted. Its mild, but its been measured on most interfaces.
Hello David. What about the track input gain control itself? Would you adjust that also to -18dB or leave it set at 0dB and just use the fader only? When would you opt for using the input gain control as abviously its there for a purpose. Thank you so much. PS - Will enrol in your course!
I am using Adobe Audition. Another associated question is how loud should my project files be? It would make sense for them to be all at a nominal level, say -3dB or -6dB perhaps, and normalised to those values? To help the mixing process? I totally get the -18dB target, but is that set with fader levels @ -18dB or by the channel input gain control, then raising the fader to 0dB? There seems to be a lot of variables in this process. In my current project I set my track input gain controls in the DAW @ -18dB, but the project files were much too low and I had to norrmalise them to -3dB. The fader default being 0dB. Thank you David.
I'm not an Audition user so this is a generic reply. If there is an input gain control in software that controls the level being recorded, then it is a convenience feature that saves the user having to reach for the real gain control on the audio interface. Since it can only work on the bits that the interface sends into the DAW it cannot improve the signal-to-noise ratio or distortion performance. Also, it can't control clipping. So I'd say don't use it. Set the gain on your interface. Regarding your other question, it doesn't matter. When you're recording you can set the gain (on your interface) as high as you like as long as there is no clipping. Setting a lower gain however gives you more headroom so there is less chance of an accidental clip. If your meters are bobbing around -10 dBFS then you're in the right zone. Or go for -18 if you want to be extra cautious. Go for -6 if you want to throw caution to the winds but watch out for clips. DM
Recording while playing games (for recording or on stream) I have to set normal voice at around -30 or so, cause the peaks are very high, but still, with the processing afterwards (afterwards in teh chain with VSTHost, so it's processed live), it works just fine for both live and recording. A bit of noise is much better than peaking which just sound painful to listen to if it is the norm in a recording. And for gaming, the game sound mostly covers the noise anyway.
@@Alexandra-Rex clearly you don’t know what you’re talking about. I’m not talking about a VST compressor. You use an analogue compressor BEFORE the signal reaches the ADC. It will limit the signal peaks and stop the ADC ever peaking.
@@pete3816 Do y ou mean VST? If the signal is clipped at the input stage, before any processing and before the ADC, it's still clipped regardless of what you do afterwards. You can't unclip a signal. Yes, it's less harsh than if it happens when it's digital, but it's still digital.
@@Alexandra-Rex bro. You clearly have no clue what you are even saying! Clipping a digital ADC happens WAY earlier than it does in the analogue world. Just because you have historically been clipping your signal in the digital realm does not mean that you are overloading the analogue circuitry of your pre-amp. It just means you’ve set too hot a signal to your ADC. Do like I said in the first post - place an analogue compressor limiter BEFORE the ADC and you’ll be good to go. Or, just keep playing your computer games and leave the audio engineering to those that understand it.
What if the master is clipping when my channels are at healthy levels, & that is causing me to have to lower all the channels extremely low, how do I fix this?
Pull down your master fader. Modern DAWs have a lot of headroom and are very flexible, as you can see and hear at www.adventures-in-audio.com/exploring-the-massive-headroom-in-your-daw/ If you have to lower the master by more than 10 dB, consider setting your channels lower next time you mix. If you have to pull down the master by more than 20 dB then there's a problem somewhere - a) Make sure your original recordings don't clip, b) Start off your mix with the first few tracks around -10 dB or so because the level is going to rise as you add more tracks, c) If a plug-in increases the level, like an EQ boost, consider lowering the fader to compensate, d) A plug-in, like an EQ boost, may add level in the master track. If your master fader is pre-insert then a level trim plug-in, placed last or just before your mastering limiter, can be used to compensate. The simple solution normally is to make sure your original recordings don't clip and to start off mixing with your tracks around -10 dB.
The reason the signal is weak is because insufficient gain had been used when recording. If you can't raise it enough with the fader, then you can use clip gain, a trim plug-in, or normalization. These will all raise the signal level digitally and there will be no noticeable effect on sound quality. If you send the signal out of the DAW and amplify it electronically then there may be a little more noise and distortion, but unless there is a fault then it should sound OK. There will be latency to correct for. It's just more trouble though and raising the level digitally is a lot easier.
@@gauravsharma-eh7sz It's the time it takes to get the signal out of the DAW, through the digital-to-analog converter, back through the analog-to-digital converter and back into the DAW. It will be several milliseconds, perhaps tens of milliseconds, which is enough time to throw things out of sync. You can correct it, but it's one more thing to do.
I seem to have forgotten about that. Yes there will be at some point, but for now I can advise that this whole issue is vastly inflated. Part 1 basically says 'no red lights on recording', and 'keep some headroom'. Part 2 would be 'no red lights when you bounce'. Part 3 is 'it's good to know what levels your plug-ins clip'. Part 3 would apply to a plug-in that emulates analogue equipment. For instance it would seem appropriate that 0 dBFS is the level the analogue equipment that the plug-ins is modelled on clips. But DAWs can handle much higher levels internally, so it's good to be aware of this, and also that it's easy enough to test clipping using sine waves. Keep your monitor level down though. There's more detail to add but these are the main points and with modern DAWs it's difficult to go wrong otherwise. DM
@@AudioMasterclass If you do post Part 2 I'll be interested to watch it. Gain Staging is an often discussed topic and I appreciated your video because it was one simple concept and not a long drawn-out discussion. Makes it easy to understand and apply to my recordings.
Thank you for the useful informations in the video. Can you recommend how to properly set gain for mixing from DAW to analogue console line inputs (in my case Focusrite Saffire Pro into old Studiomaster 16-4-2 Series 5, using Logic and Pro Tools). Regards.
This is generic for most audio interfaces and consoles. If your console has gain controls on the channel line inputs: Set the gains of the console channels you intend to use to minimum, EQ out. Connect the line outputs of the interface to channel line inputs of the console. Solo one channel and play your track. You should see some level on the main console meters. Increase that channel's gain until you get a good strong reading on the meter, not hitting the end stop or top LED on the highest peak. Repeat for the other tracks and channels. If your console doesn't have gain controls for the line inputs, and your interface probably doesn't have output level controls other than the main monitor outputs, then you'll probably be lucky and the levels are OK. If they are too high, you'll need to lower them in the DAW. Regarding the StudioMaster - Looking at photos it seems to have a clip indicator on each channel. If this is pre-fade which it should be, then it will be useful and should never come on. (There's another way to set up the console using sine wave tones, which you could use for any transfer. For this place a sine wave tone at 0 dBFS on one track, solo that channel of your console, then set the channel gain so that the meter comes within a hair's breadth of the end stop, or the top LED just comes on. Set all the channels similarly and this will work for any transfer.)
Normalise means to set the level to a specific value. So you would normalize loudness to -14 LUFS for TH-cam for example, which is their preferred level. There's more information at en.wikipedia.org/wiki/Audio_normalization DM
I wish you guys had made it clear in your video that the volume setting must be done before the computer. The proper gain staging needs to happen before the analog-to-digital-converter. Your video suggests that this can be done within the DAW, but there it is already too late. Therefore I would not recommend this video to anybody looking for advice for proper gain staging. A lot of what you are telling is true but some part is misleading and therefore the results may be unusable. Sorry.
@AudioMasterclass 0:27 "All I need to do is set the gain and watch the meter in my digital audio workstation". I feel you should have said that the gain needs to be set on the preamp of the interface (or elsewhere before the ADC). The way you say it can be interpreted as to use the pregain inside the DAW, especially since you are moving the DAW's fader afterwards. However, I feel you actually meant the gain before the ADC and it just came across in the wrong way for me. I agree with the rest of your video, though.
@@klauba Seeing as I said a mere five words between "audio interface" and "set the gain" I think a reasonable person would understand that it's the audio interface where the gain needs to be set. Having said that, if I update this video it wouldn't hurt to say "set the gain on the interface" to be doubly sure.
Get your 30-day FREE TRIAL of the Audio Masterclass Music Production and Sound Engineering Course www.audiomasterclass.com/free-trial?aid=12
This is literally the only video anyone wanting to learn about gain staging at the recording phase should watch. Absolutely simple and perfectly explained.
Normalized "good" -10 level 4:47
Normalized "bad" -40 level 6:57
I have to turn up everything pretty loud to be able to hear the noise over the signal. Is that bad hearing (of the high frequencies) or to be expected? I found the bad sample sounded slightly rougher/ harsh, but the noise is not audible at normal (or lower) listening levels.
This was straight to the point. Thank you
You're welcome.
very excellent - organized and clear presentation - nice production values
Do you stage on a track or the main track?
So, -10 or -12 dbfs is a good peak signal for gainstaging, right? For basically any source? Same for percussive and non-percussive? And what do you think about doing the gain staging to 0 dbvu without going over 0? What is better?
It depends on how predictable your source is. -10 dBFS is good for voice. It doesn't have to be exact but the meter should hover around -10 on peaks. The aim is to be completely confident you will never get an over which will ruin your take. You can concentrate on helping your singer perform well. But if you were for instance recording from a drum machine, which is totally predictable, then there is no reason not to go all the way to 0 dBFS, just a little below clipping. The higher level you record, the better the signal-to-noise ratio, but with 24-bit audio it's hard to tell the difference between -10 and 0. DM
@@AudioMasterclass Hi, some people say that even with a predictable signal, this must be -12dbfl tops, because of the plugins you put afterwards. Is this so?
@@preciseaudioblog The saying that 'two wrongs don't make a right' comes to mind. 12 dB of headroom is OK in a 24-bit recording which makes it a 22-bit recording, but your performer might go crazy and the level goes up to -1 dBFS or maybe even -0.1 dBFS. Don't panic, everything is OK. Any plug-in that can't handle 0 dBFS at least has not been competently designed. But consider that a plug-in that emulates hardware may also emulate that hardware's clipping characteristics. Or in the case of tube equipment a progressive increase in distortion towards clipping. The key is to listen. If you like what you hear then great. If you don't and you suspect the issue is level then use a gain plugin before the plugin that doesn't sound the way you want. Other than for the purpose of headroom there is never any reason to record at a low level. DM
Why is a minus sign missing from the values alongside the dB scale?
This is a good point that probably only Avid can answer. It wouldn't be the only meter scale to be incorrectly labeled though. DM
How do you deal with a situation where a vocal or guitar performance gives you -24dBFS RMS and -6dBFS peaks?
Celebrate! Your level is perfect with 6 dB headroom that kept you safe from clipping. With headroom, like many things, it's always better to have it and not need it than need it and not have it. Your RMS measurement compared to peak is governed by the dynamic range of the performance and it is only the peak measurement that you should consider when setting gain. DM
The one that was just about under the threshold of peaking. Imma say on most preamps, it will have distorted. Its mild, but its been measured on most interfaces.
Best video ever about recording gain setting! Great job 👍👍
This was awesome. Very informative after butchering the gain so many times in my life.
Thank you for your comment. We have more videos on gain staging coming up soon(ish).
Best video explained all the aspects..very helpful
You're welcome. DM
Hello David. What about the track input gain control itself? Would you adjust that also to -18dB or leave it set at 0dB and just use the fader only? When would you opt for using the input gain control as abviously its there for a purpose. Thank you so much. PS - Will enrol in your course!
If you can tell me which DAW you're using I can check this further. DM
I am using Adobe Audition. Another associated question is how loud should my project files be? It would make sense for them to be all at a nominal level, say -3dB or -6dB perhaps, and normalised to those values? To help the mixing process? I totally get the -18dB target, but is that set with fader levels @ -18dB or by the channel input gain control, then raising the fader to 0dB? There seems to be a lot of variables in this process. In my current project I set my track input gain controls in the DAW @ -18dB, but the project files were much too low and I had to norrmalise them to -3dB. The fader default being 0dB. Thank you David.
I'm not an Audition user so this is a generic reply. If there is an input gain control in software that controls the level being recorded, then it is a convenience feature that saves the user having to reach for the real gain control on the audio interface. Since it can only work on the bits that the interface sends into the DAW it cannot improve the signal-to-noise ratio or distortion performance. Also, it can't control clipping. So I'd say don't use it. Set the gain on your interface. Regarding your other question, it doesn't matter. When you're recording you can set the gain (on your interface) as high as you like as long as there is no clipping. Setting a lower gain however gives you more headroom so there is less chance of an accidental clip. If your meters are bobbing around -10 dBFS then you're in the right zone. Or go for -18 if you want to be extra cautious. Go for -6 if you want to throw caution to the winds but watch out for clips. DM
@@AudioMasterclass Thank you David. I won"t be using the DAW input gain any longer. That said, what dB level would you opt for re' the project files?
very well done
This is a great video. Clear, informative, engaging. Thank you!
Cleared some things up for me! thank you
Very good advice thnx
Recording while playing games (for recording or on stream) I have to set normal voice at around -30 or so, cause the peaks are very high, but still, with the processing afterwards (afterwards in teh chain with VSTHost, so it's processed live), it works just fine for both live and recording. A bit of noise is much better than peaking which just sound painful to listen to if it is the norm in a recording. And for gaming, the game sound mostly covers the noise anyway.
You’d be better off recording at a higher level (-18dbvu) and using a compressor/limiter to smooth out the crazy peaks and stop clipping.
@@pete3816 The peaking would happen before the compressor anyway, so it wouldn't change anything.
@@Alexandra-Rex clearly you don’t know what you’re talking about. I’m not talking about a VST compressor. You use an analogue compressor BEFORE the signal reaches the ADC. It will limit the signal peaks and stop the ADC ever peaking.
@@pete3816 Do y ou mean VST?
If the signal is clipped at the input stage, before any processing and before the ADC, it's still clipped regardless of what you do afterwards. You can't unclip a signal. Yes, it's less harsh than if it happens when it's digital, but it's still digital.
@@Alexandra-Rex bro. You clearly have no clue what you are even saying! Clipping a digital ADC happens WAY earlier than it does in the analogue world. Just because you have historically been clipping your signal in the digital realm does not mean that you are overloading the analogue circuitry of your pre-amp. It just means you’ve set too hot a signal to your ADC. Do like I said in the first post - place an analogue compressor limiter BEFORE the ADC and you’ll be good to go. Or, just keep playing your computer games and leave the audio engineering to those that understand it.
What if the master is clipping when my channels are at healthy levels, & that is causing me to have to lower all the channels extremely low, how do I fix this?
Pull down your master fader. Modern DAWs have a lot of headroom and are very flexible, as you can see and hear at www.adventures-in-audio.com/exploring-the-massive-headroom-in-your-daw/ If you have to lower the master by more than 10 dB, consider setting your channels lower next time you mix. If you have to pull down the master by more than 20 dB then there's a problem somewhere - a) Make sure your original recordings don't clip, b) Start off your mix with the first few tracks around -10 dB or so because the level is going to rise as you add more tracks, c) If a plug-in increases the level, like an EQ boost, consider lowering the fader to compensate, d) A plug-in, like an EQ boost, may add level in the master track. If your master fader is pre-insert then a level trim plug-in, placed last or just before your mastering limiter, can be used to compensate. The simple solution normally is to make sure your original recordings don't clip and to start off mixing with your tracks around -10 dB.
what would be the difference between reamping a weak signal vs raising it with clip gain
The reason the signal is weak is because insufficient gain had been used when recording. If you can't raise it enough with the fader, then you can use clip gain, a trim plug-in, or normalization. These will all raise the signal level digitally and there will be no noticeable effect on sound quality. If you send the signal out of the DAW and amplify it electronically then there may be a little more noise and distortion, but unless there is a fault then it should sound OK. There will be latency to correct for. It's just more trouble though and raising the level digitally is a lot easier.
@@AudioMasterclass got it thank you please make more such videos
@@AudioMasterclass in this case how do you define latency ?
@@gauravsharma-eh7sz It's the time it takes to get the signal out of the DAW, through the digital-to-analog converter, back through the analog-to-digital converter and back into the DAW. It will be several milliseconds, perhaps tens of milliseconds, which is enough time to throw things out of sync. You can correct it, but it's one more thing to do.
Is there a part 2?
I seem to have forgotten about that. Yes there will be at some point, but for now I can advise that this whole issue is vastly inflated. Part 1 basically says 'no red lights on recording', and 'keep some headroom'. Part 2 would be 'no red lights when you bounce'. Part 3 is 'it's good to know what levels your plug-ins clip'. Part 3 would apply to a plug-in that emulates analogue equipment. For instance it would seem appropriate that 0 dBFS is the level the analogue equipment that the plug-ins is modelled on clips. But DAWs can handle much higher levels internally, so it's good to be aware of this, and also that it's easy enough to test clipping using sine waves. Keep your monitor level down though. There's more detail to add but these are the main points and with modern DAWs it's difficult to go wrong otherwise. DM
@@AudioMasterclass If you do post Part 2 I'll be interested to watch it. Gain Staging is an often discussed topic and I appreciated your video because it was one simple concept and not a long drawn-out discussion. Makes it easy to understand and apply to my recordings.
1:37 free my boy
Thank you for this video. I'm wondering what the normalized level is. Did you apply any limiting /dynamics?
The narration was processed separately and is set to -21 LUFS using the Waves WLM Plus. The poem does not have any limiting or compression.
Thank you so much!
Thank you for the useful informations in the video. Can you recommend how to properly set gain for mixing from DAW to analogue console line inputs (in my case Focusrite Saffire Pro into old Studiomaster 16-4-2 Series 5, using Logic and Pro Tools). Regards.
This is generic for most audio interfaces and consoles. If your console has gain controls on the channel line inputs: Set the gains of the console channels you intend to use to minimum, EQ out. Connect the line outputs of the interface to channel line inputs of the console. Solo one channel and play your track. You should see some level on the main console meters. Increase that channel's gain until you get a good strong reading on the meter, not hitting the end stop or top LED on the highest peak. Repeat for the other tracks and channels. If your console doesn't have gain controls for the line inputs, and your interface probably doesn't have output level controls other than the main monitor outputs, then you'll probably be lucky and the levels are OK. If they are too high, you'll need to lower them in the DAW. Regarding the StudioMaster - Looking at photos it seems to have a clip indicator on each channel. If this is pre-fade which it should be, then it will be useful and should never come on. (There's another way to set up the console using sine wave tones, which you could use for any transfer. For this place a sine wave tone at 0 dBFS on one track, solo that channel of your console, then set the channel gain so that the meter comes within a hair's breadth of the end stop, or the top LED just comes on. Set all the channels similarly and this will work for any transfer.)
@@AudioMasterclass Thank you so much! All the best!
Hi! Great video. What do you mean with normalice?
Normalise means to set the level to a specific value. So you would normalize loudness to -14 LUFS for TH-cam for example, which is their preferred level. There's more information at en.wikipedia.org/wiki/Audio_normalization DM
@@AudioMasterclass Thank you.
I see mixcraft up there
Thanks, (finally clearly explained) thank you, ;-)
I wish you guys had made it clear in your video that the volume setting must be done before the computer. The proper gain staging needs to happen before the analog-to-digital-converter. Your video suggests that this can be done within the DAW, but there it is already too late. Therefore I would not recommend this video to anybody looking for advice for proper gain staging. A lot of what you are telling is true but some part is misleading and therefore the results may be unusable. Sorry.
I doubt if you're all that sorry. You're getting some kind of pleasure from criticising stuff I never said.
@AudioMasterclass 0:27 "All I need to do is set the gain and watch the meter in my digital audio workstation". I feel you should have said that the gain needs to be set on the preamp of the interface (or elsewhere before the ADC). The way you say it can be interpreted as to use the pregain inside the DAW, especially since you are moving the DAW's fader afterwards. However, I feel you actually meant the gain before the ADC and it just came across in the wrong way for me. I agree with the rest of your video, though.
@@klauba Seeing as I said a mere five words between "audio interface" and "set the gain" I think a reasonable person would understand that it's the audio interface where the gain needs to be set. Having said that, if I update this video it wouldn't hurt to say "set the gain on the interface" to be doubly sure.
🤯
if there only was such a thing as a floating point mic and interface haha
Is this the kind of thing you have in mind? www.zoom-na.com/products/field-video-recording/field-recording/zoom-f6-multitrack-field-recorder
Not a bad introduction :-) I did a similar video on gain staging just last week...
A truly lifeless performance .