What's new

My tracks are too quiet...

Wow, may I say how grateful I am for taking the time to explain all these concepts to me. Thank you so much! I know that what I am asking is very basic to most of you, but indeed for me this simple matter has been a headache for weeks now. I will definitely go through your posts multiple times in order to absorb and understand everything you said.

So, if I understood it correctly reducing the dynamic range makes the softer tones louder and the louder a bit quieter, right? So that even at an average volume both extremes can be heard comfortably. After fixing the problem with my track being very quiet (by normalizing or by turning the kontakt fader to 0 from -6 dBFS and also increasing the GAIN afterwards) that was my next issue, because I noticed that the part where the piece was getting louder was too loud compared to the rest of the piece. The sample library I am using has a built-int function to increase or reduce the instrument's dynamic range. Is it advisable to adjust the dynamic range that way or do it as you said using compressors and EQ?

I did try using a compressor by the way and noticed a considerable increase in loudness, but my knowledge is still very basic as to how to use it correctly. I will have to experiment more with it

Do you think it is wrong if I turn up the Kontakt fader to 0dBFS? So far, I have not seen any clipping happening. Also, should I increase the GAIN, if I have to, before doing the EQ and adding the compressor or afterwards? Sorry, if my questions are redundant. Just trying to understand the whole thing a bit better! Thank you again!

Cheers,
Alex
 
Right, but he (I think he's a he) got the answer he was looking for early on.

Your posts about level in general are very good, and far more than he was asking. I only got frustrated at the ludicrous posts in this thread!

We all inevitably have to go through the learning process of asking questions and/or giving answers that sound stupid. For this reason, many people are afraid of asking questions which I think it is a pity. A lot of times, especially for those who are just beginners, the answer is right in front of them, but because it is something new for them, they simply cannot make the connection. For example, before asking this question, I didn't know what audio bouncing is and had never heard of it. I had heard of normalization, but had no idea what it does or how to use it. Now, some of these techniques and terminologies are a bit clearer to me and I hope that any other beginners who might happen to read this thread might also be benefited.

All in all, thanks for baring with me and for your help!

Cheers
 
by turning the kontakt fader to 0 from -6 dBFS
Just want to double check if this was a typo. 0 to -6 dBFS would be a volume decrease (i.e. gets quieter). turning the kontakt fader from 0 to +6 dBFS would increase the volume.

I noticed that the part where the piece was getting louder was too loud compared to the rest of the piece. The sample library I am using has a built-int function to increase or reduce the instrument's dynamic range. Is it advisable to adjust the dynamic range that way or do it as you said using compressors and EQ?

This is probably an artistic choice! It also depends on what the built-in dynamic range function is doing, that can help guide your decision. I can imagine the built-in dynamic range function working in two different ways:

Maybe that function is just using underlying Kontakt FX to work like a typical audio dynamics compressor. In that case, it will work similarly to any compressor FX plugin, and the difference between the built-in and your own compressor FX would just be the depth of control/options you have to configure the compressor FX, versus and the convenience of having it as part of the Kontakt instrument.

But the second possibility, that built-in function may be working like a "MIDI dynamic range" feature, and this could be really useful for intimate piano pieces. Explanation - most virtual instruments will be "velocity sensitive", i.e. the volume of each individual sample being played depends on MIDI note velocity. So if you increase this "MIDI dynamic range" option, the virtual instrument could play low MIDI velocities even quieter, and high MIDI velocities even louder. So in other words its like you have a larger dramatic volume difference between the quiet MIDI velocities and strong MIDI velocities. On the opposite end, if you reduce the MIDI dynamic range, then low and high MIDI velocities will have a more similar volume level.

the effect of an audio compressor versus the effect of a MIDI dynamic range control is very different. The audio compressor will apply to all notes/sounds simultaneously, and it has no concept of separate notes. Take a piano note for example, which can start loud, then it gradually decays over the sustain. A compressor would have no idea that this is one piano note, and so it could do some kind of awkward volume increase that counteracts the piano's natural volume decay. That can be a desirable effect sometimes, but not always. On the other hand, a MIDI dynamics compressor has the ability to keep the natural sustain/decay of each piano note, but it can change the overall volume of each piano note completely separately. If the piano library you're using has recorded different tone color of the piano playing softly and loudly, then this kind of MIDI dynamic range compression could work really well to create an intimate sound.
 
Do you think it is wrong if I turn up the Kontakt fader to 0dBFS? So far, I have not seen any clipping happening. Also, should I increase the GAIN, if I have to, before doing the EQ and adding the compressor or afterwards? Sorry, if my questions are redundant. Just trying to understand the whole thing a bit better! Thank you again!

When working with virtual instruments and mixing, there will always be a thousand different places where you can adjust gain / fader / volume level, some examples include
- the Kontakt instrument may have gain/volume knobs in the instrument's interface
- the Kontakt volume slider
- the channel fader on your DAW
- many FX plugins including compressors, EQs, etc, will have input gain and/or output gain knobs
- etc. etc.

So your question about deciding which place you change volume - mainly will depend on three things:
(a) if you have any FX that change behavior depending on volume.
(b) your personal workflow
(c) if you are worried that the output of some software may cause clipping

about option (a) - so for example, if you have a compressor FX, or a distortion FX - the output of those effects will change a lot depending on the input volume. So in those cases you will want to make sure you set the input volume to get the effect you want, and then re-adjust the output volume after the effect, to get the desired volume level you want.

Other than this FX detail, it shouldn't really matter where you adjust the volume, and you can do it where you feel it's the most useful - if you want easy access to change the volume, or if you want it to be a set-it-once-and-forget-it hidden somewhere in your template, etc.

about option (c) - 99% of the time this should not be an issue. We could to discuss it if you're concerned about it though
 
Do you set your levels up before recording? Play your piano so the loudest peaks are around -6/-8 db, record & mix.

Now have it mastered so it fit's in with the relative volume other any tracks in the project, album or based on where you plan on having it streamed or broadcast.
 
Thank you Shawn again for two very informative posts. The only thing I am reluctant to touch are the channel faders. I find myself only doing it if I am using prerecorded effects that are either too loud or too quiet.

I will experiment with both the built-in dynamic range and compression and see what works best. My goal is to achieve an intimate sound that doesn't lack in brightness if possible. I am not sure if The Giant is the correct piano library for that though... The piano by itself sounds very percussive and is overloaded with harmonics/overtones (probably due to the length of the strings).

Do you set your levels up before recording? Play your piano so the loudest peaks are around -6/-8 db, record & mix.

I saw that being mentioned before and I am not sure how to do that. I use my stage piano, that currently doubles as a midi keyboard to record. My audio interface has a gain knob for input levels, but since I am not using a microphone I though it wasn't relevant. The piano (a Roland FP-7) has a volume knob and a balance knob, but both don't seem to affect the sound at all...

I am planning on uploading my tracks on Youtube for the time being. My knowledge is not sufficient enough for an album release.

Cheers,
Alex
 
Last edited:
We have to go through the process of asking questions that sound stupid, but giving stupid advice... not so sure about that. :)

I might be guilty of doing this in the past, too. Sometimes we think we know something and it turns out that we don't or that there is a better way. It happens all the time with interpersonal relationships. Life is all about failing and learning, making mistakes and learning from them (or not).

Cheers
 
I saw that being mentioned before and I am not sure how to do that. I use my stage piano, that currently doubles as a midi keyboard to record. My audio interface has a gain knob for input levels, but since I am not using a microphone I though it wasn't relevant. The piano (a Roland FP-7) has a volume knob and a balance knob, but both don't seem to affect the sound at all...


you're right that is not relevant in your case, if you are using the keyboard only for midi. If there is no audio coming from the keyboard, and nothing bring recorded from your audio interface, then those volume knobs would not affect the volume of what you are doing

In your case, the piano sends midi to your computer and the midi eventually reaches kontakt. The sound is finally being generated from kontakt.

Sorry if you already mentioned earlier, but are you using Kontakt as standalone software? Or are you using it as a plug-in in your DAW? Which DAW are you using?

What you want to look for are some kind of audio meters in your software which show you the level of your audio signal. At least there is one in kontakt, but if you are using a DAW, the visualization on the meters from your DAW will have more clear dB units labeled and have a clearer visual. These dB meters will the somewhere in the mixer part of the DAW software.

Once you find the right meter that visualized your audio, and once you find a reasonable fader that affects the volume, ceemusic's recommendation of setting levels so they peak at approx -8 dB - I think that was mentioned with the assumption that mastering will be able to increase the loudness further. But if you are self producing without mastering, you'll want to consider a the other advice too, about using a compressor/limiter. Cheers!
 
Last edited:
My audio interface has a gain knob for input levels, but since I am not using a microphone I though it wasn't relevant. The piano (a Roland FP-7) has a volume knob and a balance knob, but both don't seem to affect the sound at all...

Learn how your audio interface operates!- YES, not only is the gain knob for input levels relevant but it's KEY!.

Getting nominal levels set is the first & most important factor you need to address before going forward. If not suggestions like normalizing might make the track louder but the floor noise will also be increased.
 
Last edited:
YES, the gain knob for input levels is very relevant.

But this only applies when actually capturing audio from the inputs of the audio interface. My understanding is that Agondonter is just using midi to control Kontakt with The Giant piano instrument loaded.

Agondonter, I think part of the difficulty in answering your question is that we don't know the exact setup you have :). So you are getting a lot of great advice about a lot of different scenarios. Hopefully that's not causing to much confusion and you'll be able to figure out which advice is actually applicable in your case.
 
When I mix, I often start with a gain control on my mix bus. I use this to raise the volume of the track to a healthy level (depending on the piece) and then add other plugins after that for other processing.

I think Nick is right - turn it up. When you don't get enough signal by moving the fader - use a gain plugin.

I don't think normalization is a good idea if you are trying to make a pro mix.
Are you implying I was suggesting normalization? I wasn't. I was suggesting increasing midi volume which is different from audio volume. It's a parameter to tell the sample player to play louder (or softer).
 
Sorry if you already mentioned earlier, but are you using Kontakt as standalone software? Or are you using it as a plug-in in your DAW? Which DAW are you using?

Both. When I am trying things out I use the standalone version. I then move to Sibelius (planning on changing to Dorico at some point) and write down my piece where I use Kontakt as a plug-in and then in Cubase 5 (I am planning to upgrade to 10 when there is a sale... hopefully soon) I will either perform and record the piece or import the midi file.

What you want to look for are some kind of audio meters in your software which show you the level of your audio signal. At least there is one in kontakt, but if you are using a DAW, the visualization on the meters from your DAW will have more clear dB units labeled and have a clearer visual. These dB meters will the somewhere in the mixer part of the DAW software.

Once you find the right meter that visualized your audio, and once you find a reasonable fader that affects the volume, ceemusic's recommendation of setting levels so they peak at approx -8 dB - I think that was mentioned with the assumption that mastering will be able to increase the loudness further. But if you are self producing without mastering, you'll want to consider a the other advice too, about using a compressor/limiter. Cheers!

I have SPAN from Voxengo, YouLean Loudness Metter and VUMT Deluxe. I check the LUFS reading from YouLean and the VU reading from VUMT Deluxe apart of course from the dBFS reading of the master channel. I am still unsure about the VU reading. I don't quite understand it, but to be honest I have also not taken the time to read the manual that came with the software.

I composed another piece for a single pad and two voices the other day and managed to bring the volume to a sufficient level by increasing the Kontakt fader and then the Gain by 6. Then I tried compressing the track using Supercharger from NI, but I noticed that the continuous and smooth sound of the pad started being uneven creating a crackling effect after applying the compression (I just used the preset with the least impact). The built-in limiter from the DAW had no other immediate effect other than increasing the volume output. Is it always advisable to use a compressor and a limiter?

But this only applies when actually capturing audio from the inputs of the audio interface. My understanding is that Agondonter is just using midi to control Kontakt with The Giant piano instrument loaded.

Agondonter, I think part of the difficulty in answering your question is that we don't know the exact setup you have :). So you are getting a lot of great advice about a lot of different scenarios. Hopefully that's not causing to much confusion and you'll be able to figure out which advice is actually applicable in your case.

Yes, I am not capturing audio from the inputs of the audio interface. My mic is not even connected. I use my stage piano to record.

I have an old notebook that has Sibelius, Cubase 5 and several plug-ins and libraries installed. My audio interface is Komplete Audio 6 from NI. I also have a Korg nanoKONTROL 2 for automating, which I haven't used at all so far. I am still very confused about automation and how it works, but that is another story for another thread maybe. :)

To record, I have a Shure mic and my old stage piano (a Roland FP-7). At the corner in my room sits a cello that misses a string and a megabass waterphone from the inventor himself. :)

Cheers
 
To the op, I'm sorry but someone should have told you immediately to just do a Google search for "gain staging". It's critically important to understand and there is a plethora of tutorials on the web on how to do it.

Long story short there are two kinds of volume adjustments for a kontakt instrument.

One is the performance dynamics which should always be controlled by the mod wheel, CC11, and/or velocity. These controls are used to simulate the natural dynamics of a performer, and they often use methods other than just turning the volume up or down. So if you need volume changes in the performance, these will be the most realistic.

The other kind of volume control is where gain staging comes in. And for this you use either the instrument's master volume, or Kontact's master volume which is at the top of the kontakt interface and may have to be unhidden.

The way to gain stage Kontakt is simple. Set it's channel fader in your DAW to zero. Then using Kontact's master volume, adjust it so that, at the performance's loudest point, it peaks at around -12db.

Why -12db? For one reason, this is the happy spot for most insert plugins you might want to use. This isn't hard written in stone, and some plugins might need to be a little hotter, like saturation plugins.

But generally speaking, -12db is a good starting point and it should leave plenty of headroom in the master bus after all of your tracks are assembled.

The main thing to remember is that all gain staging should be done before the signal hits the insert bus. This way you're not clipping your plugins. Unless of course that is the desired effect.

On a personal note, I generally frown on using the channel's fader to control volume. Many Kontakt libraries do some quite lovely things with their internal dynamic controls, like switching layers or changing timber.

Also, some libraries have their own built-in reverb and it sounds really unnatural to change the volume of a reverb return when you change the volume of the instrument that's feeding it.

So when you change the volume or automate the output of a Kontakt instrument that has its own built-in reverb send, you're essentially automating the reverb return on that instrument, which is whacked. Or, really cool depending on how you use it.

General rule of thumb is that if your reverb is meant to simulate a real natural space, then that reverb's output should be left alone.

So what I do whenever I start a new Contact track, is all turn up the mod wheel or CC 11 to the maximum volume that I'm going to use, which is not always 100%, then I'll gain stage to -12db using Kontact's master volume, then I'll record my part.

When mixing, if I need to tweak the volume, including adjusting the dynamic range , instead of using compression or gain automation, I'll try to do so using the instrument's built-in dynamic controls. Something not loud enough? Crank up its velocity levels or turn up CC 11.

Cubase even has a cool feature that allows you to "compress" the dynamic range of a part by reducing the difference between the highest velocity level and the lowest velocity level. Very useful.

So I rarely ever use fader automation and pretty much only use compression as a tonal effect, not to fix dynamic problems.
 
Last edited:
It seems there is some confusion between analog and digital audio here, and I would like to share some things I learned over the years, maybe someone find it usefull. And please correct me if I'm wrong.

There is no "efficient, healthy" capturing audio from a sample library. OP is not recording anything, there is no "gain" in terms of analog amplification of a signal, and there is no changing noise to signal ratio which is set once those samples were recorded. The samples are 24bit, if your signal is around -40db, just apply a clean digital gain aka PT Clip Gain, Cubase has the equivalent as well, or there are numerious free and paid gain plugs you can insert. BUT, it has to be pure digital gain in order not to modify the sound (Logic Pro Gain that sits in the panel on the left is NOT clean but it simulates an analog desk saturation, more you add it, more it modifies the sound - not sure if it can be turned off, but this is a negative thing in my book of daws). Also, volume/gain, in daw, those are all the same, level changing algos, so wheter you change it in Kontakt, in a plugin or your daw fader, if they are clean, they are all doing the same thing (they are probably different algos as there are probably more than one way to acomplish a level change, but that's not relevant here)

Our daws are computing with 32bit floating point which means they have a dynamic range of 1500db. There is nothing lost if you bounce (32bit or 24bit) your instrument at a low level, you can change that level up and down in your daw with no consequences (as long as you dont bounce above 0db digital scale as that is the top). I can send you my piano track (24bit wav) at -40 db and you can add gain to it in ur daw and get the same exact thing as I had before I lowered it in my daw.

We have to pay attention about gain staging when introducing other processes (plugins) which are balanced to work with certain signal levels. AND ofc when we want to reach certain level with our final track. But digital audio had simplified many things by introducing so much headroom, and as far as I know, many rules from analog age doesn't apply anymore.

But ofc, when we record live instruments using mics and preamps, signal levels are crucial.

Cheers.

Milos
 
Last edited:
@labornvain and @shomynik thank you so much for your added clarifications. Since I started the thread, my understanding has improved significantly and you have just added a lot to it. I am now researching more about Gain Staging and digital gain.

@labornvain I used to increase the gain from the mixer in Cubase, but after your suggestion I am going to do that using the Kontakt Master volume. As for the reverb, I recently bought VVV and Blackhole, so I turn the library's reverb off and use one of the two. Reverb is another area I need to learn a whole lot. Right now I am just using the presets with the exception of the Blackhole reverb where I managed to create a nice "natural sounding" reverb for the piano sound I mostly use.

@shomynik I am researching digital gain right now. What is your suggested amount of headroom, by the way? I mostly write ambient, minimalistic music. So far, I have never experienced clipping even with heavily increasing the gain.

Cheers,
Alex
 
@labornvain and @shomynik thank you so much for your added clarifications. Since I started the thread, my understanding has improved significantly and you have just added a lot to it. I am now researching more about Gain Staging and digital gain.

@labornvain I used to increase the gain from the mixer in Cubase, but after your suggestion I am going to do that using the Kontakt Master volume. As for the reverb, I recently bought VVV and Blackhole, so I turn the library's reverb off and use one of the two. Reverb is another area I need to learn a whole lot. Right now I am just using the presets with the exception of the Blackhole reverb where I managed to create a nice "natural sounding" reverb for the piano sound I mostly use.

@shomynik I am researching digital gain right now. What is your suggested amount of headroom, by the way? I mostly write ambient, minimalistic music. So far, I have never experienced clipping even with heavily increasing the gain.

Cheers,
Alex

No worries Agondonter, glad you found it helpfull.

In 32 bit floating point there is no clipping while you are working in a clean daw that doesn't have any analog simulations going on. Cubase is such a DAW. The red channel flashing is just an indicator that you have peaks over 0db digital scale, and you have to worry about that only when bouncing/exporting in order to avoid nasty digital clipping. That is very easily done, you can even lower you stereo master fader in the end until the clipping is gone.

As far as the recomended levels, you got very good advices in this thread, many plugins are balanced to work from -12db to -16db, that would be great if you could maintain that level all the way through. But to make things easier with no consequences, you can only think about levels when:

-you use your plugins: you can adjust input gain of that plugin or if the plugin doesnt have the input control, you can insert a separate gain plugin before it, or use any other level control before the plugin.

-when you bounce/export to audio file, nothing should go over 0db digital scale (which is the scale on the cubase faders). Again, for this you can use plugins on the master channel, either level controls in your existing plugins (eq, comps, limiters, etc...) or a separate clean gain plugin.

Also, you got tons of great info regarding the final track levels as well as many ways how to get to a desired one.

Here is a good pdf on digital audio, it's a very easy read:

https://redirect.viglink.com/?format=go&jsonp=vglnk_155499789167515&key=57ed2afae0b472fc7ec991a58f1b72c9&libId=jucsk43p0102uz1c000MAcb3whibl&loc=https%3A%2F%2Fwww.logicprohelp.com%2Fforum%2Fviewtopic.php%3Ft%3D65875&v=1&out=http%3A%2F%2Fwww.popmusic.dk%2Fdownload%2Fpdf%2Flevels-in-digital-audio.pdf&ref=https%3A%2F%2Fwww.google.com%2F&title=Can%20my%20low%20fader%20position%20degrade%20sound%20quality%20%3F%20-%20Logic%20Pro%20Help&txt=http%3A%2F%2Fwww.popmusic.dk%2Fdownload%2Fpdf%2Flev%20...%20-audio.pdf

Milos
 
Top Bottom