Seeking advice in basic mixing (to Gain or not to Gain?)

[Dave G]

I'm an amateur MIDI musician who started composing in the 90's. Using Voyetra Digital Orchestrator, I created General MIDI tracks strictly within the confines of my PC: simply drawing notes into the Piano Roll, no live instruments or external hardware. I knew nothing about mixing; I just created the tracks, effects/controllers, set volume levels and called it done. Needless to say, when I listen to these tracks now, I could've done much better.

Twenty+ years later, I work in Sonar/Cakewalk in the very same manner as above. Synths used: TTS-1, Studio Instruments, Session Drummer, Cakewalk Sound Center, and Addictive Drums 2. I'm not a professional musician...I just simply wish to create the melodies I hear in my head. Target audience is the casual listener. I live in a small apartment environment which isn't suitable for acoustically-treated walls or a large assortment of equipment. However, once I get further along, I may purchase an audio interface and studio speakers/headphones. Otherwise, I have to work with what I have and be happy with it.

That being said, I will explain my experiences the best way I can.

I've recently begun the process of remixing my old MIDI songs in today's DAW. Having started in Sonar but also experienced in Studio One, I feel my place is in Sonar/Cakewalk. However, the drastic evolution and advancement in digital music software has left me consistently overwhelmed and questioning my processes. I've sat on this project for over six years, because I constantly get frustrated, annoyed, second guessing, overthinking -- and repeatedly abandoning the project to start up again. But now I feel determined and motivated to keep pushing, do it right and finish some day.

To the point: I've watched countless videos on mixing, volume levels, gain staging, and the like. I've taken notes of everything I learned, especially things that I can't find in the Manuals, documenting it in my personal reference guide. But I'm still confused, unprepared, and unsure of what I'm doing, worried about doing everything wrong. I seek guidance and education to help get my projects off the ground without fear of messing up.

To begin, here's my understanding of the gain staging process:

1.  Set every channel's fader volume at 0db. Begin setting track volume using the console's Gain knob, staying below -12db to -18db.
2.  Implement each track's plug-ins/effects. (I don't use EQ or compression, as these techniques are beyond my comprehension.)
3.  At some point, I intend to bounce these tracks to audio to preserve them. (Before/after step #2? And how do I convert the audio back to MIDI?)
4.  When the mixing is complete, set fader levels accordingly, preventing clipping.
5. I'll have to export and store my finished projects in both raw and final formats without losing any plug-ins...I have no idea how.

All my TTS-1 instrument tracks are gradually upgraded to SI Instruments, Sound Center, etc. Problem is, these other "synthtruments" are about 15-25% louder than TTS-1 instruments. But when I lower that track's Gain knob, the instrument becomes muddy, muffled, and distorted, and the Gain knob has reached a horrendously ridiculous level (< 50/60%). I've also heard that I can raise that synth's volume level to compensate lowering the Gain knob too much. (But by how much?)

The entire gain staging process has gotten me so conflicted and rattled that I don't know how do it right. On the other hand, I've seen articles that suggest simply ignoring the gain staging hype (don't worry about that Gain knob) and focus solely on preventing clipping throughout the entire process.

So, my biggest pain point is to gain stage/not gain stage...and that darn Gain knob. Properly balancing the volume levels without misusing the Gain knob. I don't think I even want to use that anymore. Can I create a balanced mix without even using the console's Gain knobs, and just use the faders in the end of the process? I know every musician's routine is a subjective personal preference, but I'm having a great deal of trouble establishing a process that works for me...after all, still an amateur. Because of the complexity, this project is literally taking me forever because I can't make a decision.

CONCLUSION

I apologize for the rambling, as I have to lay out all the details. It could be said that, after so many years, I clearly have no business making music and perhaps should be pursuing other hobbies. But I truly feel I can do this. It's been complicated, but I'm thankful for several users here on the Forums who've answered so many of my questions in the past as I work on these projects.

Therefore, I hope for some insight to get me going with confidence and hopefully more education. I could really use some help here, and dumb it down if you have to. Please provide your thoughts and guidance to above? Thanks in advance!

Edited October 7, 2021 by Dave G

[John Vere]

First I’m a bit confused by this 

20 hours ago, Dave G said:

But when I lower that track's Gain knob, the instrument becomes muddy, muffled, and distorted, and the Gain knob has reached a horrendously ridiculous level (< 50/60%).

This seems terribly wrong to me. Turning down gain or level has never changed the quality of the sound in my 60 years of audio.  It’s common knowledge that loud always sounds better but quiet  is just quiet  and there definitely should not be distorted sound as you reduce level. 
 

The better quality VST instruments are louder because they use samples that are optimized. 
I have never used the instrument channel gain on a VST. I always use the one supplied in it GUI. I set that so the loudest output is showing between -14db to  -8db on the meter. I then use the instrument track ( this is an audio track) to fine tune using the fader. This is also where I will use automation. I avoid using midi tracks to adjust level. There's no point in this as when you end up replacing the Instrument all will change anyhow. 
With midi data I do pay attention to velocities as each VST might respond differently. Velocity is for timber not volume . Example a piano will sound harsh at full velocity.  I find it's important to find the right velocity. Example if I use Ample or SI bass they seem best at 100 to no more than 110. 

Midi is the safest way to save projects as it has proven to be long lived and future proof. There’s always going to be better VST instruments to use. 
I have never needed to bounce midi to audio. As a matter of fact I do the exact opposite and convert audio to midi for a safety copy of all my projects. 
 

 

Edited September 26, 2021 by John Vere

[OutrageProductions]

Unless you plan to sell your final product for profit (ie: you're in this for personal enjoyment), then don't bother with the gain knob. Lots of mixes (even back on large format analog consoles) were done with some of the faders in the -40db range and below. Yes, keeping them in the 0db range allows for more finesse for the professional product, especially when automating flying faders, but it's a matter of personal choice.
If your initial instrument audio is printing in a healthy manner (not too quiet and not clipping) then just have fun.
Always print your final mix with effects, cuz they can't be added later.

And generally, a finished mixed audio track cannot be "turned back into MIDI".

Also check out the YouTube channel from Creative Sauce, he does really good, simple, detailed tutorials, in plain (well, Aussie ?) language.

Start here: 

 

[Starship Krupa]

I don't touch the Gain knob unless the virtual instrument on the track isn't putting out enough signal (or putting out too much, but I like to tweak this with the instrument's own output control) to allow me to keep the faders in their usable working range. So your Step 1 is alien and puzzling to me.

There is no "gain staging" that needs to be done before I engage my FX (unless, as I say, I need to get the faders in their sweet spot, which doesn't mean "sweet" as far as audio quality).

Gain staging in a DAW is done for different reasons than with hardware sources and mixers. In the hardware realm, you want to feed each stage with a signal that is neither so low that you get noise or so hot that you get clipping. In the DAW realm, the levels are virtual. So only processors that are level-dependent such as compressors and things that emulate the distortion characteristics of analog hardware care much what virtual level you send them.

What this means is that if you find that the input and/or output meters of a plug-in are "in the red" or barely registering, then adjust the input level accordingly. If you have to adjust the controls on a plug-in to extremes to get it to sound good, then you should adjust what level is feeding it. If you hear raspy distortion, gain staging is one thing to examine.

I find that it (and not only it) is something that novices tend to worry about out of proportion to its actual effect. Here is a good article on the topic.

For the kind of work you are doing, I'd suggest that one of the best things you could do is buy a license for XPand!2, which is currently on sale for $4.99 at Pluginboutique.

It's not a General MIDI synth, but it has all of the sounds that you would find in a GM synth, so if your MIDI tracks are split into individual tracks, you can use whatever sound you like. For the 2,000+ sounds that you get, which include some pretty decent drum kits, five bucks is the proverbial no-brainer. Bonus: it's really light on resource use.

I'm not sure why you want to or think you need to "convert the audio back to MIDI." Presumably you already have the note data in MIDI form. Why convert it back?

To get a track in audio rather than MIDI-to-virtual-instrument form, just get it to how you want it to sound, then freeze it. That creates an audio file with all of the track effects applied and disables the plug-ins on the track, both instrument and effect. Unless your system starts to bog down from a high number of plug-ins, there should be no need to do this.

[Byron Dickens]

https://mailchi.mp/soundonsound.com/basics-beyond-ebook?utm_source=soundonsound&utm_medium=weblink&utm_campaign=basics_sosnews

[John Vere]

This op has quite a few threads that he never responded to. Weird. 

[Byron Dickens]

Not so much weird as just damn rude.

[Dave G]

Thank you to all for your insight in this topic! I do apologize for not coming back to this sooner. I knew this would be a large field of discussion, so I wanted to give each response the time and attention it deserved to my best ability.

This is a huge subject for me as I learn the workings of the software -- how to properly assess and determine volume/gain levels, use all the functions correctly, and still create a solid musical piece.

There are occasional times I feel I might do something wrong and get in a rhythm, then realizing there was a more appropriate way to do something.

I just felt discouraged that, even after watching literally dozens of tutorial videos, that I can't/couldn't be sure of the right and wrong way of doing something -- forgetting that I don't have a professional target audience, so just do it and have fun. But I guess part of the complexity is that everyone has their own process and I simply haven't fully developed mine yet.

However, as I feel less overwhelmed and more prepared, I have been dedicating more time to "working, not worrying" -- actually getting things done on these MIDI projects without overthinking on any of the processes too much.

And so...

John: Your expertise is appreciated, as your information has always been useful in assisting others in their progress as well.

I had just simply noticed that lowering the Gain knob has produced unwanted results in the sound of said instrument. I can't explain why, either it happens or it's just the way I hear it.

This was first noticed that, as I lowered the Gain knob on a TTS-1 drums track, some kit pieces suddenly fell inaudible while others still stood out completely not in proportion to original level.

...I also have general-purpose desktop speakers (for now.)

OutrageProductions: Addressing your feedback in conjunction with that from John, for my needs, I've taken to using the faders for volume balancing. I tend to have developed a good eye (and ear) to monitor the volume levels and prevent clipping. This is just the path I choose to use.

Even so, I've learned to trust not only my own judgement but the expertise of those more experienced to guide me as I am still learning how this all works.

And thank you for your link to the Cakewalk MIDI Basics video. I have seen several of his videos; his tutorials are concise and well done.

Starship Krupa: Thanks for your wealth of information also. I seem to have gotten "bouncing tracks" confused with"freezing tracks". (Explained toward bottom.)

I am familiar with XPand!2, and have always been intrigued by having a bigger library of interments to choose from.

BDickens: Thank you for the Sound on Sound article. I have begun reading it and I will be sure to peruse that further as I go along.

ALL: I understand I wasn't clear when I first mentioned "bouncing my final tracks to audio" in step 3,  confusing that with freezing tracks. What I meant was, once my project is finished, the need to export it to a permanent format in which the instrument tracks, effects and elements therein can be "preserved", if possible (i.e. if, one day, I go back to a project and the instruments, plugins, or other parts are no longer available?)

I just want to have a solid format (along with the final MP3) saved so that everything will always be there.

In the old days (the 90's), when I finished a MIDI file, I would have my PC record the tracks via the Windows "What U Hear" setting and then save that file to an MP3. Ahhh...the old days. LOL.

On another note, I've also learned from several sources that a synth's built-in volume control can be applied to compensate for having to adjust the Gain knob too much.

I am well aware that, in order to take full advantage of this powerful software, I could use at least some knowledge of more of the controls, functions, plugins and effects. For now, I'm working with what I have and applying my learnings slowly and steadily.

Thanks again to all. ?

Edited October 2, 2021 by Dave G

[57Gregy]

I have never touched any Gain controls in any of my projects.
That may be reflected in the final mix. ?