Old 10-09-2020, 01:07 PM   #1
ReaperStudent1
Human being with feelings
 
Join Date: Jan 2016
Posts: 152
Default Proper gain structure question

This is a question about how to improve my mixing roadmap. I use samples and synths exclusively and generally try to bounce all tracks to audio for mixing.

1. I watch a video that says “print at -18db”, so, when I bounce my Kontakt tracks to audio, I follow the video, match levels to a tone set at -18, and print to WAV files.

2. I open them up in a “mixdown” session and I can barely see them. Also, my faders are now at “zero” so there is theoretically little or no headroom. I can barely hear anything either, though of course if I raise my headphone volume and/or put a limiter on the master buss I can hear fine

3. However, when I watch a “Master” class, like soundtrack mixer Dennis Sands, I see the tracks come in at max level, the waveform display fills the track!

I am sure I am missing a lot in this sequence of events and it would be extremely helpful if someone can explain what I am missing.

1. Are my audio tracks REALLY supposed to be so low in the monitor (headphone) level after printing at -18?
2. Are my audio files REALLY supposed to be almost invisible?
3. Am I supposed to maximize them after printing to audio?
4. How do I get any headroom when my faders are at 0db and my incoming tracks are maxing at -18?
ReaperStudent1 is offline   Reply With Quote
Old 10-09-2020, 03:34 PM   #2
juan_r
Human being with feelings
 
juan_r's Avatar
 
Join Date: Oct 2019
Posts: 1,075
Default

I'm sure someone more knowledgeable will chime in with useful advice, but first off:

Printing at -18 dBfs peak level (dBfs means 0 is the absolute maximum!) is not a hard rule. You can print hotter if you like, depending on the program material. -18 dBfs is sane, though.

Even at low levels, you should be able to see some meter movement. Are you metering pre-fader (original printed signal) or post-fader? There's a preference for this. As long as you don't clip (touch 0 dBfs, full scale), you're good to go. What matters more is what you hear. Raise your monitors/headphones if you need to. The way you're doing, you've got plenty of headroom! Headroom is the space between your signal peaks (smallish) and full scale (up above). If you use plugins, it is important to hit each one of them at its preferred level, especially for amp sims or other saturation-rich FX.

There is no one right way to do it. If you have even just a few tracks printed around -6 dBfs, the master channel will soon try to go over the ceiling at 0 dBfs. This must never happen! I'm betting Dennis Sands has most faders well below the 0 dB mark

Maximize/normalize: not really necessary, usually. If you don't like seeing wimpy meter readings or you "run out of fader" when trying to boost the signal, you can add some gain to the audio items on the tracks by using Item Properties (F2).
juan_r is offline   Reply With Quote
Old 10-09-2020, 03:40 PM   #3
vdubreeze
Human being with feelings
 
vdubreeze's Avatar
 
Join Date: Jul 2011
Location: Brooklyn
Posts: 2,613
Default

Quote:
Originally Posted by ReaperStudent1 View Post
2. I open them up in a “mixdown” session and I can barely see them. Also, my faders are now at “zero” so there is theoretically little or no headroom. I can barely hear anything either, though of course if I raise my headphone volume and/or put a limiter on the master buss I can hear fine
Are you able to raise the waveform view in Reaper's tracks (shift + up arrow key) to see them well enough? As far as visibility, that usually will let you see pretty low levels as full waveforms.

I don't think you can equate the faders being at 0 with the literal headroom of whatever is in the tracks. If there's a lot of range for the fader to go before the audio approaches clipping the master, then you have headroom. So I personally wouldn't be concerned about that. I think about how the combined tracks are hitting the master when the faders are at 0.

You can experiment with your results by selecting all items in the tracks and in item properties bring the level up 7 or 8 or so db (if you know they're at -18) and see how things respond. And take them back to their original state if desired.

One great feature in Reaper is the monitor FX inserts, where you can raise the level using whatever you'd like, even a simple gain change plugin, and these only affect the monitoring. They're not included in file renders or etc. Very helpful for monitoring of quiet files than strain the headphone amp before you've reached the stage of final leveling of the master output.
__________________
The reason rain dances work is because they don't stop dancing until it rains.
vdubreeze is offline   Reply With Quote
Old 10-09-2020, 03:58 PM   #4
serr
Human being with feelings
 
Join Date: Sep 2010
Posts: 12,562
Default

Quote:
Originally Posted by ReaperStudent1 View Post
This is a question about how to improve my mixing roadmap. I use samples and synths exclusively and generally try to bounce all tracks to audio for mixing.

1. I watch a video that says “print at -18db”, so, when I bounce my Kontakt tracks to audio, I follow the video, match levels to a tone set at -18, and print to WAV files.

2. I open them up in a “mixdown” session and I can barely see them. Also, my faders are now at “zero” so there is theoretically little or no headroom. I can barely hear anything either, though of course if I raise my headphone volume and/or put a limiter on the master buss I can hear fine

3. However, when I watch a “Master” class, like soundtrack mixer Dennis Sands, I see the tracks come in at max level, the waveform display fills the track!

I am sure I am missing a lot in this sequence of events and it would be extremely helpful if someone can explain what I am missing.

1. Are my audio tracks REALLY supposed to be so low in the monitor (headphone) level after printing at -18?
2. Are my audio files REALLY supposed to be almost invisible?
3. Am I supposed to maximize them after printing to audio?
4. How do I get any headroom when my faders are at 0db and my incoming tracks are maxing at -18?
That -18db suggestion is for recording live audio. If you don't know what stray peaks might come along with a signal, setting the record trim to achieve -18db rms is a good guess that usually avoids any surprise peaks hitting red and clipping.

It's a guess to help avoid distorting a recording. And if it turns out that you didn't ever have any surprise peaks and the whole thing (peaks and all) never hits above -18dbfs, then you've only sacrificed 3 bits and still have a 21 bit recording.

This doesn't apply after recording at all and it's not a starting point target for channels on the mixing board. It's purely a guide to avoid surprise peaks when recording live inputs.
serr is offline   Reply With Quote
Old 10-09-2020, 07:49 PM   #5
ReaperStudent1
Human being with feelings
 
Join Date: Jan 2016
Posts: 152
Default Re serr, juan r, vdubreeze

Serr - -this is super helpful. In point of fact I felt I was losing a bit (maybe my imagination) when I had to use a lot of gain to bring the mix back up to the level that I heard when I was composing. Your “3 bit” loss….what level would help me avoid that???

It did occur to me that the “-18db rule” had more to do with live recording. But all of the posts mentioning the danger of exceeding 0db on the master are really sinking in.

Vdubreeze – I searched in the manual for “waveform display” but never came across this shortcut. Thank-you so much

Juan r-- your post is very reassuring and I appreciate it a lot.
ReaperStudent1 is offline   Reply With Quote
Old 10-09-2020, 08:00 PM   #6
serr
Human being with feelings
 
Join Date: Sep 2010
Posts: 12,562
Default

That's what 24 bit format is for.
It's a very large dynamic range and quite forgiving.

If your worst case is making a 21 bit recording...
That raw track is higher quality than CD right out of the box with that. There's no problem to avoid.

You could make a recording that peaks at -36db and you'd still have an 18 bit recording. Still better than CD format. That would look like a blank flat line down the middle of the item at -36db. You have working room!

And for a final release format, you can have really really really quiet sections in the music and still have 8 to 12 bits of resolution. This is where 16 bit CDs maybe get grainy sounding. But I digress.

Anyway, the thinking is that if the meat of an incoming signal is -18db rms then it's very unlikely to get an over because that would be really really dynamic!
serr is offline   Reply With Quote
Old 10-09-2020, 08:17 PM   #7
valy
Human being with feelings
 
Join Date: Jan 2020
Posts: 1,927
Default

-18 dBFS works for RMS-heavy signals, but for transient stuff like percussion, aim for a max peak of -6 dBFS instead.

In fact, regardless of the source, I almost always gain-stage each track to either -18 average or -6 peak, whichever it hits first.
valy is offline   Reply With Quote
Old 10-09-2020, 08:46 PM   #8
serr
Human being with feelings
 
Join Date: Sep 2010
Posts: 12,562
Default

Quote:
Originally Posted by valy View Post
-18 dBFS works for RMS-heavy signals, but for transient stuff like percussion, aim for a max peak of -6 dBFS instead.

In fact, regardless of the source, I almost always gain-stage each track to either -18 average or -6 peak, whichever it hits first.
No argument there!

We used to have to nearly fully produce the 'raw' tracks to stay well ahead of tape hiss. Fill all those rust particles fully with music or they will just hiss at you! Slam those levels! Trying to drag people away from that kicking and screaming now that we have happiness and light in digital headroom is still difficult. Heh, then the volume war CD came along. Now beginners have their volume control way down for those shrieking loud CDs and they start trying to mix at that monitor level. Red lights going off before you even hear the track!

Anyway, the goal isn't to hit some cryptic number or meter reading. It's just to not fuck up. These are guides to help not fuck up.
serr is offline   Reply With Quote
Old 10-10-2020, 05:28 AM   #9
ReaperStudent1
Human being with feelings
 
Join Date: Jan 2016
Posts: 152
Default Re Gain structure - - faders???

Thank-you Serr...so...the other part of my question. I have bounced my Kontakt tracks to audio and opened them in a new mix session (my weird way of doing things). Now my faders are at zero. My audio is peaking at -18db So I am mostly pushing the faders UP? That's just strange to me but the goal is the appearance of the master buss fader, not where the other faders are? I guess I'm really looking to understand the next part of the routine. It is sounding like I need to bring up the gain in my headphones up to a very comfortable (full) level in a non-destructive way (headphone volume or gain on the output fx insert) and then I won't be pushing faders that much? Is there a preferred way that this is normally done?
ReaperStudent1 is offline   Reply With Quote
Old 10-10-2020, 07:00 AM   #10
valy
Human being with feelings
 
Join Date: Jan 2020
Posts: 1,927
Default

Quote:
Originally Posted by ReaperStudent1 View Post
Thank-you Serr...so...the other part of my question. I have bounced my Kontakt tracks to audio and opened them in a new mix session (my weird way of doing things). Now my faders are at zero. My audio is peaking at -18db So I am mostly pushing the faders UP? That's just strange to me but the goal is the appearance of the master buss fader, not where the other faders are? I guess I'm really looking to understand the next part of the routine. It is sounding like I need to bring up the gain in my headphones up to a very comfortable (full) level in a non-destructive way (headphone volume or gain on the output fx insert) and then I won't be pushing faders that much? Is there a preferred way that this is normally done?
You need to calibrate your monitors and/or headphones so that a full mix that averages -18 dBFS and peaks around -6 dBFS plays back through them at the volume you want to mix at. For me, that's 75-80 dBSPL.

Easiest way to do this is to play pink noise at -18 dBFS through your monitoring system, then hold a C-weighted dBSPL meter exactly where your head would be in the listening position and adjust your monitoring system's volume knob to the target dBSPL level.

If you do this properly and then gain-stage each track as discussed above before you start mixing, you will find yourself rarely touching the volume knob, and most of your track faders should end up in the -20 dBFS to 0 dBFS range when you set your initial levels.
valy is offline   Reply With Quote
Old 10-10-2020, 07:05 AM   #11
valy
Human being with feelings
 
Join Date: Jan 2020
Posts: 1,927
Default

Quote:
Originally Posted by serr View Post
No argument there!

We used to have to nearly fully produce the 'raw' tracks to stay well ahead of tape hiss. Fill all those rust particles fully with music or they will just hiss at you! Slam those levels! Trying to drag people away from that kicking and screaming now that we have happiness and light in digital headroom is still difficult. Heh, then the volume war CD came along. Now beginners have their volume control way down for those shrieking loud CDs and they start trying to mix at that monitor level. Red lights going off before you even hear the track!

Anyway, the goal isn't to hit some cryptic number or meter reading. It's just to not fuck up. These are guides to help not fuck up.
Yeah, even in the 16-bit digital days we had to worry about recording hot enough to not lose resolution. Thankfully those days are behind us as well!

Also, like you mention, the numbers are somewhat arbitrary: -18 dBFS is just a recommendation that I happen to use. Some people use -20 dBFS, some use -12 dBFS. The most important thing is a consistent approach across all the tracks in all your mixes.
valy is offline   Reply With Quote
Old 10-10-2020, 03:41 PM   #12
ReaperStudent1
Human being with feelings
 
Join Date: Jan 2016
Posts: 152
Default Thank-you to all

Thank-you for all of the replies. I have been putting your advice into practice today and feel much better about how the bounce to audio is being preserved, as well as enjoying plenty of headroom and flexibility. Here is a "next step" question. So...let's say I am getting a satisfactory mix but I am now hitting the master well below 0db. Sometimes, when this happens , I will use a mastering product (Slate FG-X or Ozone) but to my ears even the most benign settings, when adding, say, 10db or more of gain, seem to be coloring the sound of the master buss. Am I crazy? Or are there other secrets to bringing up the gain more transparently? remember, the actual faders are now only 6db below the top of the channel (at least on my mixer setup).
ReaperStudent1 is offline   Reply With Quote
Old 10-10-2020, 04:14 PM   #13
juan_r
Human being with feelings
 
juan_r's Avatar
 
Join Date: Oct 2019
Posts: 1,075
Default

If you don't want any coloration, you can boost your mix after the fact. Export it and normalize it later. Alternatively, turn all features off in Ozone/Slate FG-X and just tick normalization.

These mastering programs/plugins will usually add some "beautifying" EQ, multiband compression, and wide-band limiting. A tad of transparent limiting (high threshold, just kissing the peaks) won't change the sound that much; EQ and multiband compressor tend to be more obvious. Don't like that? Just turn them off.

Be aware that normally you won't be able to get 10 dB of apparent loudness just with normalization. You can only get "lossless" gain, meaning "leftover headroom": the distance between your highest peak and 0 dB. Anything more, there need to be compromises.

Valy's advice is golden.

Quote:
Originally Posted by valy View Post
You need to calibrate your monitors and/or headphones so that a full mix that averages -18 dBFS and peaks around -6 dBFS plays back through them at the volume you want to mix at. For me, that's 75-80 dBSPL.
And, may I add, if you don't have a suitable dbSPL meter to measure sound pressure, you can use your ears to ballpark a reasonable listening level; that's the main point of calibration anyway. It doesn't need to be 100% exact to be useful.

Last edited by juan_r; 10-10-2020 at 04:28 PM.
juan_r is offline   Reply With Quote
Old 10-14-2020, 03:52 PM   #14
Win Conway
Human being with feelings
 
Join Date: Dec 2010
Posts: 3,826
Default

Ignore pretty much everything and make up your own mind about stuff, I render tracks and normalize to 0, only time I worry about gain staging is if a particular effect has some kind of modelling that needs a specific level because it is pretending to be hardware.
Other than that, just don't clip the master, done.
__________________
Stop posting huge images, smaller images or thumbnail, it's not rocket science!
Win Conway is offline   Reply With Quote
Old 10-14-2020, 08:22 PM   #15
Steviebone
Human being with feelings
 
Join Date: Jul 2018
Posts: 809
Default

As it just so happens I tried an exercise the other day for just this purpose. To understand how to calculate headroom for a mix you need to know how many maximum tracks you are going to have playing at one time. You also need to understand the laws of logarithmic decibel summation. Without getting too mathematically technical try the following test project.

Create a project with 8 identical tracks each with a single instance of JS tone generator outputting a 440 signal at zero dB (sine). Put a copy of blue cats DP meter Pro on the master output channel. With the master fader set to zero dB the channel faders must be scaled back by approximately 16.8 dB to maintain 0db output at the master. Then try different track counts. Notice how the relationships change with different track counts...


Here is an excerpt from my test:

The offsets for various track counts is as follows:

440 hz

8 tracks: -16.8

16 tracks -24.9 db

24 tracks -28 db

32 tracks -30.3 db

40 tracks -31.6 db

48 tracks -32 db

Then try it with different frequencies. Note the energy put out by lower frequencies is a little higher.


220 hz

8 tracks: -18.5

16 tracks -24.1 db

24 tracks -27.7 db

32 tracks - 30.3 db

40 tracks -32.1 db

48 tracks -32.5 db


Finally a little white noise... (or pink take your pick, they produce slightly different results). After all music isn't just one frequency.

8 tracks -17.9

.... do your homework... you should go through this process yourself.

See how the logarithmic math changes as the track count increases the difference at the output becomes less.

So for me, what I do is calculate roughly what the total number of tracks playing at any one given time are going to be, then start there (nothing is ever playing at zero dB all of the time so I usually factor that in as a percentage also). After you've gone through your mix once or twice to get an additional balancing of levels look at where your headroom is at the final output at the loudest stage of the piece. Then use the SWS action "increase track volume envelope by x db" for all tracks (before busses) to bring all of your envelopes up to a level that gives you a final headroom sufficient for level with a little left over for mastering.

Last edited by Steviebone; 10-14-2020 at 08:27 PM.
Steviebone is offline   Reply With Quote
Old 10-17-2020, 03:09 PM   #16
ReaperStudent1
Human being with feelings
 
Join Date: Jan 2016
Posts: 152
Default cool test!

The test that Steviebone created is fascinating. and makes incredible sense. Thank-you!
ReaperStudent1 is offline   Reply With Quote
Old 10-17-2020, 03:11 PM   #17
ReaperStudent1
Human being with feelings
 
Join Date: Jan 2016
Posts: 152
Default Re: Normalize after rendring the mix

This is super fascinating. Is this being done commonly by this community? Is this the most transparent way to retain the flavor of the mix and maximize gain of the final track?
ReaperStudent1 is offline   Reply With Quote
Old 10-17-2020, 03:25 PM   #18
ReaperStudent1
Human being with feelings
 
Join Date: Jan 2016
Posts: 152
Default Re "increase track volume envelope"

<<<After you've gone through your mix once or twice to get an additional balancing of levels look at where your headroom is at the final output at the loudest stage of the piece. Then use the SWS action "increase track volume envelope by x db" for all tracks (before busses) to bring all of your envelopes up...">>>

I love this but I am unable to find any SWS action that affects "track volume envelope". I do see SWS actions for "Select all envelope points" and "increase selected envelope points..."

If there is a more "global" way to approach raising ALL volume envelope points at once...I'd be grateful if someone could point me to it! And Thank-you!!!
ReaperStudent1 is offline   Reply With Quote
Old 10-17-2020, 04:01 PM   #19
serr
Human being with feelings
 
Join Date: Sep 2010
Posts: 12,562
Default

Quote:
Originally Posted by ReaperStudent1 View Post
This is super fascinating. Is this being done commonly by this community? Is this the most transparent way to retain the flavor of the mix and maximize gain of the final track?
Not in my cage.

Set volume of listening system with non-volume war music.
Ears now indicate too loud.
Meters agree.
Mix hits the proper volume. No math.

Something in there hitting my limiter hard?
("Hard" is anything more than a stray db or two once or twice.)
And I can hear it being kind of crude for that?
Finesse that element.

Trust your ears, man!
serr is offline   Reply With Quote
Old 10-17-2020, 05:37 PM   #20
Steviebone
Human being with feelings
 
Join Date: Jul 2018
Posts: 809
Default

Quote:
Originally Posted by ReaperStudent1 View Post
<<<After you've gone through your mix once or twice to get an additional balancing of levels look at where your headroom is at the final output at the loudest stage of the piece. Then use the SWS action "increase track volume envelope by x db" for all tracks (before busses) to bring all of your envelopes up...">>>

I love this but I am unable to find any SWS action that affects "track volume envelope". I do see SWS actions for "Select all envelope points" and "increase selected envelope points..."

If there is a more "global" way to approach raising ALL volume envelope points at once...I'd be grateful if someone could point me to it! And Thank-you!!!
I posted a script over in the script forum for this:

https://forum.cockos.com/showthread....19#post2353219

If you need some help adapting it to your setup let me know.

Note: I made 10 copies of the script, changing the parameter at the very bottom for each of the SWS cmds, down01, down05, down5, down10, up01, up05, etc., named each one appropriately and then assigned them to buttons on my launchpad. Works great!

Last edited by Steviebone; 10-17-2020 at 05:49 PM.
Steviebone is offline   Reply With Quote
Old 10-17-2020, 06:10 PM   #21
ReaperStudent1
Human being with feelings
 
Join Date: Jan 2016
Posts: 152
Default Re: Great Script...question

Pardon my ignorance here. I am unfamiliar with Reaper scripting syntax. When you say "changing the parameter at the very bottom" you mean the very last appearance of :

adjust_vol('up5')

To the other values in your explanation? In other words
change adjust_vol('up5') to something like adjust_vol(up05) or adjust_vol(up10) ... or do I need the single quote too as in adjust_vol('up10')

But this is really cool. I just don't want to mess up because i missed a quote mark or whatever.

Sincere thanks!
ReaperStudent1 is offline   Reply With Quote
Old 10-17-2020, 06:55 PM   #22
ReaperStudent1
Human being with feelings
 
Join Date: Jan 2016
Posts: 152
Default Steviebone adjust track volume

I got the script loaded. I ran it "as is". it adjusted track volumes up 1.67 db . When I changed the param to 'Up10' it adjusted all points up approximately 3.3 db Should it have adjusted up 5 and 10? I simply pasted, named it, it auto set the suffix to .lua and it started working. I don't know anything about scripts. I just pasted exactly from your text.

It worked on multiple selected tracks which is so cool. I just don't knw if there is anything I need to do to get it to follow the adjust params.

Last edited by ReaperStudent1; 10-17-2020 at 07:09 PM.
ReaperStudent1 is offline   Reply With Quote
Old 10-17-2020, 10:47 PM   #23
Steviebone
Human being with feelings
 
Join Date: Jul 2018
Posts: 809
Default

Quote:
Originally Posted by ReaperStudent1 View Post
I got the script loaded. I ran it "as is". it adjusted track volumes up 1.67 db . When I changed the param to 'Up10' it adjusted all points up approximately 3.3 db Should it have adjusted up 5 and 10? I simply pasted, named it, it auto set the suffix to .lua and it started working. I don't know anything about scripts. I just pasted exactly from your text.

It worked on multiple selected tracks which is so cool. I just don't knw if there is anything I need to do to get it to follow the adjust params.
That sounds like a metering issue. The SWS actions up/down by .1 decibel, .5 decibel, 1 decibel, 5 db and 10db. You should be seeing those jumps. When I get back in the studio I will run the modified script and check the results. Have no idea how you could be getting those numbers. Which script did you copy? the first or second one?

Edit: I checked and rechecked both versions of the script... they work fine for me. Are you looking at the amount of gain on each individual track in the volume lane or are you looking at the master meter output?

Last edited by Steviebone; 10-18-2020 at 02:46 AM.
Steviebone is offline   Reply With Quote
Old 10-18-2020, 08:41 AM   #24
juan_r
Human being with feelings
 
juan_r's Avatar
 
Join Date: Oct 2019
Posts: 1,075
Default

Maybe this thread is straying far from the OP's needs, at least at this stage. If I had to choose by short-simple-useful, my favorite snippet would be this.

Quote:
Originally Posted by serr View Post
Set volume of listening system with non-volume war music.
Ears now indicate too loud.
Meters agree.
Mix hits the proper volume. No math.
Meaning "calibrate meters vs. monitor loudness, even if ballparking by ear"


Quote:
Originally Posted by serr View Post
Something in there hitting my limiter hard?
("Hard" is anything more than a stray db or two once or twice.)
And I can hear it being kind of crude for that?
Finesse that element.

Trust your ears, man!
^^ This. The last line sums it all up.
juan_r is offline   Reply With Quote
Old 10-18-2020, 01:20 PM   #25
ReaperStudent1
Human being with feelings
 
Join Date: Jan 2016
Posts: 152
Default Reply to Steviebone (and another thank-you)

I was simply looking at the volume envelope points on a basically "empty" (1 point) volume envelope set at 0.00db. when I hit "Run" and checked the value of the envelope those numbers I mentioned were what Reaper "said" the envelope values were. Again, apologies for my script skills. I actually write quite a bit of code in Tradestation and have done some financial VBA macros, but never did anything in Reaper.

Possible sources of my errors(?) a)The location of the parameter, b) the syntax of the parameter (although there were definitely changes when I changed the parameter I thought was specified). Also, amazingly, when I popped in some random volume envelope points, they definitely moved in sync during run mode. c) Possible faulty copy/paste? I assume it simply wouldn't run if I did that...
ReaperStudent1 is offline   Reply With Quote
Old 10-18-2020, 01:54 PM   #26
Steviebone
Human being with feelings
 
Join Date: Jul 2018
Posts: 809
Default

Quote:
Originally Posted by ReaperStudent1 View Post
I was simply looking at the volume envelope points on a basically "empty" (1 point) volume envelope set at 0.00db. when I hit "Run" and checked the value of the envelope those numbers I mentioned were what Reaper "said" the envelope values were. Again, apologies for my script skills. I actually write quite a bit of code in Tradestation and have done some financial VBA macros, but never did anything in Reaper.

Possible sources of my errors(?) a)The location of the parameter, b) the syntax of the parameter (although there were definitely changes when I changed the parameter I thought was specified). Also, amazingly, when I popped in some random volume envelope points, they definitely moved in sync during run mode. c) Possible faulty copy/paste? I assume it simply wouldn't run if I did that...
Again, I can confirm here that the script works as expected for me. There may need to be envelope points on the lane... You first load the script, then assign it a keystroke. Select one or more tracks, press the assigned keystroke. The selected lanes (not the master) should increase/decrease by the selected amount of db. Start with a single track. Put media on the track and write some envelope points to the volume lane. Then try running the script. If you have the vol lane visible you should see all the points move. Make sure the lane is big enough that you can easily see the change.
Steviebone is offline   Reply With Quote
Old 10-18-2020, 02:14 PM   #27
Steviebone
Human being with feelings
 
Join Date: Jul 2018
Posts: 809
Default

Quote:
Originally Posted by juan_r View Post
Maybe this thread is straying far from the OP's needs, at least at this stage. If I had to choose by short-simple-useful, my favorite snippet would be this.


Meaning "calibrate meters vs. monitor loudness, even if ballparking by ear"




^^ This. The last line sums it all up.
Not in any way trying to start a flame here but this is bad advice, especially for someone just starting out. No need to learn bad habits. Ears can do all sorts of weird things. For one thing, everyone's sense of 'loudness' is different. And then there is the issue of ear fatigue. Also, someone who is just starting out doesn't yet have properly trained critical hearing. That comes from thousands of hours actually finishing projects.

Sometimes when you're headroom isn't it right you tend to over use compression. A professional engineer pays close attention to gain structure right from the start and adjusts it constantly as needed. It's sometimes impossible to know exactly what gain structure is going to work best until you're well into balancing signal. As you proceed, things like EQ and compression can alter this.

A mix that is balance hot and then compensated for later in the chain will not sound the same as one that is mixed with proper gain staging all the way through the chain. It is sometimes a subtle difference but it can be the difference between an amateur mix and a truly professional one.

Also your gain staging, especially at your lower foundational levels, affects how all of your FX are driven. You should pay close attention to where all your sends are as well. An under-driven or over-driven input to an FX plug-in can make all the difference in the quality of the effect in the mix.

I am by no means an expert but I have learned from many hours of experience that proper gain staging is particularly important, part science as well as admittedly part art at times.

I see this every day watching people mix live where almost everything on their EQ board is boost rather than cut, for example. In a studio mix setting knowing how to get the right balance without pushing too many frequencies can mean the difference between a truly clean mix and one that is creating more harmonic distortion than necessary.

I understand where you're coming from with the whole "if it sounds good it is good" thing, but I think that can be a bad habit to teach right from the start. You should stick to the rules and only break them later when you really know what you're doing, and only then in special circumstances.

Until I created those projects and studied the logarithmic summation of signal rules in a digital world I truly didn't understand how gain staging works, especially at the lower levels. It's counter-intuitive to realize that the more tracks you add, the less of a difference there is at the final output. You want to believe that the more signal and tracks you add the more headroom will be encroached. But the math actually works differently, it's not something you're going to intuitively 'hear'.

Again, not trying to be in any way offensive or combative just stating my opinion -- my two cents perhaps only worth half a penny.

Carry on. Peace out. Rock on.
Steviebone is offline   Reply With Quote
Old 10-18-2020, 02:30 PM   #28
serr
Human being with feelings
 
Join Date: Sep 2010
Posts: 12,562
Default

I don't mean to suggest being cavalier. But to suggest this is so cryptic for someone just starting out that you need scripts and maths is very wrong IMHO. THAT is bad advice! I stand by my advice for a starting point. Absolutely follow up on further understanding as you go! Don't get me wrong. But also if I'm right and the issue is monitor volume, a deep dive into scripts is going to result in confusion and frustration and leave the main issue unsolved.

Good luck ReaperStudent1!
serr is offline   Reply With Quote
Old 10-18-2020, 05:43 PM   #29
Steviebone
Human being with feelings
 
Join Date: Jul 2018
Posts: 809
Default

I guess my response to that would be that it depends on the OP's endgame objective. If music is just a hobby and he doesn't have any time to invest then your suggestion makes sense.

But if the OP's long-term objective is that the music is extremely important to him and he wants to learn how to do it right, my suggestion would be to take advantage of the fact that he's just learning - to learn it right. Yes, it will require a small investment in time, not all that much really (especially just for gain structure), but that small investment in time will pay off in spades down the road. Any small investment made in understanding proper gain structure is a drop in the bucket compared to the amount of time it saves you down the road. Not to mention the fact that you will end up with much better sounding songs.

On the other hand, if it's just a hobby and you just want to scratch some tracks out without understanding what you are doing or why, by all means turn some knobs until you get something that sounds good.

But without understanding why, you will never be able reproduce these results from song to song and end up wasting a lot of time turning knobs searching for something that 'sounds good', instead of knowing what to do to get you where you want to go.
Steviebone is offline   Reply With Quote
Old 10-18-2020, 06:53 PM   #30
valy
Human being with feelings
 
Join Date: Jan 2020
Posts: 1,927
Default

Quote:
Originally Posted by Steviebone View Post
I guess my response to that would be that it depends on the OP's endgame objective. If music is just a hobby and he doesn't have any time to invest then your suggestion makes sense.

But if the OP's long-term objective is that the music is extremely important to him and he wants to learn how to do it right, my suggestion would be to take advantage of the fact that he's just learning - to learn it right. Yes, it will require a small investment in time, not all that much really (especially just for gain structure), but that small investment in time will pay off in spades down the road. Any small investment made in understanding proper gain structure is a drop in the bucket compared to the amount of time it saves you down the road. Not to mention the fact that you will end up with much better sounding songs.

On the other hand, if it's just a hobby and you just want to scratch some tracks out without understanding what you are doing or why, by all means turn some knobs until you get something that sounds good.

But without understanding why, you will never be able reproduce these results from song to song and end up wasting a lot of time turning knobs searching for something that 'sounds good', instead of knowing what to do to get you where you want to go.
I'll admit that your method is interesting, but it seems a bit of a stretch to state that something this complex is required for proper gain staging.
valy is offline   Reply With Quote
Old 10-18-2020, 07:39 PM   #31
juan_r
Human being with feelings
 
juan_r's Avatar
 
Join Date: Oct 2019
Posts: 1,075
Default

In other words, my ears might be the optimal target for initial training.

Yes, it's subjective, but borrowing an SPL meter for half a day will give me enough objective data to show if my "normal" level is a bit off - be it loud or quiet. Adjusting to the "proper" reference will be a breeze if I already have a working routine - imperfect, non compliant, whatever.

Of course, "left-brain" understanding helps too, but the brain must receive meaningful info from the ears in the first place. Learning to trust them is the name of the first game!

I'm sprinkling more water, even if have no hint of flames.
juan_r is offline   Reply With Quote
Old 10-19-2020, 02:02 PM   #32
Steviebone
Human being with feelings
 
Join Date: Jul 2018
Posts: 809
Default

Quote:
Originally Posted by juan_r View Post
In other words, my ears might be the optimal target for initial training.

Yes, it's subjective, but borrowing an SPL meter for half a day will give me enough objective data to show if my "normal" level is a bit off - be it loud or quiet. Adjusting to the "proper" reference will be a breeze if I already have a working routine - imperfect, non compliant, whatever.

Of course, "left-brain" understanding helps too, but the brain must receive meaningful info from the ears in the first place. Learning to trust them is the name of the first game!

I'm sprinkling more water, even if have no hint of flames.
Well it DOES take both skills arguably. It don't matter much if there is proper gain structure and it sounds like crap.
Steviebone is offline   Reply With Quote
Old 10-21-2020, 10:00 AM   #33
ashcat_lt
Human being with feelings
 
Join Date: Dec 2012
Posts: 7,272
Default

Quote:
Originally Posted by Steviebone View Post
See how the logarithmic math changes as the track count increases...
Your results are weird. IDK what that meter is showing you or why you’re not just using the Reaper’s Master meter which shows you exact sample peak and RMS levels.

For peak levels, the logarithmic math is “simple”. Assuming that all the sources are exactly the same, it will be 20 log10 (number of tracks) which works out so that every doubling of tracks adds about 6.02db. We usually just use 6 because it’s close enough. 2 tracks is 6, 4 is 12, 8 is 18... This will be true of any source, any frequency, even complex signals as long as every track is exactly the same and perfectly in phase.

This is also true of RMS measurements. The actual measured RMS level of a sine wave depends somewhat on the relationship of the measurement window to the frequency of the signal, but there will still be 6db increase for every doubling of tracks.

But if we are really just mixing the exact same thing on multiple tracks, we could just turn the one track up and it would be the same and that’s not what we call mixing to begin with. In actual practice, our each of the tracks is different. Different frequencies, dynamics, phase relationships... In that case it’s pretty tough to predict how mixing any number of tracks is going to affect the overall levels. There’s a sort of rule of thumb that we sometimes use where if all of the tracks are about the same RMS level, then every doubling will probably add about 3db to the overall RMS. Peaks will fall where they may, and since the crest factor of each of the tracks is probably different, it can be really tough to even start to eyeball. I guess if you wanted to be safe, you’d assume that all of the tracks might hit the same peak level as the most dynamic of them at the same time, and therefore end up using the double = 6db thing, but that would probably be way too much “headroom” in practice.

Note that when things are not the same level, the quieter one has less impact on the total on that same logarithmic scale such that when one is 9db quieter than the other, the total will be less than 1db louder than the loudest. This is the foundation of the often misunderstood (at least) 3:1 rule (of thumb).


To the OP -
1) Anything that you render (freeze, glue, whatever) that is not your final distribution master file should go to floating point format. That way you have the full resolution no matter where the levels actually fall. It can peak at -100dbFS and you’re still not “losing any bits”, and it can peak at +100dbFS and not be clipped off. You can literally just add or subtract gain to make it sit where you want/need without any issues. “Losing 3 bits” doesn’t sound like much when you start with 24 to begin with, but if you’re going to do things like add gain then compress/limit it back down (or vice versa, it’s ultimately the same), then you’re basically throwing away even more bits and in some not terribly extreme situations, you will start to notice some effects of that truncation distortion and with actually extreme processing (run it through a high gain amp sim) it will make a real difference. Basically, since much of what you’re likely to do is reduce the dynamic range from the top down, you should start with as much usable information toward the bottom as possible.

B) I agree with the idea of calibrating your monitors, but also it’s perfectly ok to just adjust the master volume to get it to a reasonable level. If it’s peaking above 0, turn it down. If it’s way below, turn it up. In future mixes maybe try to remember how you got to that point and start at more reasonable levels, but if your mix is sounding good, there’s no point in going back and adjusting everything individually. Just move that one fader.
ashcat_lt is offline   Reply With Quote
Old 10-25-2020, 11:39 AM   #34
Steviebone
Human being with feelings
 
Join Date: Jul 2018
Posts: 809
Default

Have no idea where you are coming from. Here's the screenshots of one of the tests at A440.

https://l.linklyhq.com/l/8rQD
Steviebone is offline   Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -7. The time now is 04:06 AM.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.