RE: Setting Consistent Volume Levels

[monsterjazzlicks]

Hi folks,

I don't think this topic has been discussed at length (as far as I can see) and so I thought I would open a new thread here. Moderators feel free to relocate.

So, I have made some of my own programs on the Nord Electro 6 over the past few months. The specific issue I have is that the volume levels are not thoroughly consistent throughout all of the programs. By this I mean the Pianos and Rhodes are not at the same level; and esp. programs within the sample section - these have many volume discrepancies (even the factory-defaults). So I might, for instance, jump from a Grand Piano to a Flute preset and the volume drops by possibly 50%! Even if I set each and every 'section' volume (for every individual program) to '0dB' there are still noticeable spikes.

Do users of this forum just live with it or do they address the problem first-hand?

Would it be over-kill to hire a rehearsal room for a few hours, wind-up the volume a bit on the Nord, and play through each and every program monitoring the LEDs on a mixing-desk? I have found that if you leave it to your ears they are just not accurate enough, in my opinion at least. So it needs to be either a desk, or (in my case) using the meters in Cubase.

And once you start adding (MIDI) an external sound-module to the equation the issue becomes compounded!

In fact, this has always been a topic which has intrigued me but I never have gotten around to posing such a question.

Many thanks in advance for any kind assistance offered here . . .

Best,

Paul

[monsterjazzlicks]

Hi Paul,

I'm not a sound engineer, I'm not much experienced on stages, but this is clear: You will certainly not be able to measure the loudness of a programm by it's peak points. If you want to get them normalized, you might need special software/hardware tools, not a rehearsal room.

I'd consider a difference of 50% a bit too much. But they do exist.

As a newbie, I only learned that it's not a good idea to start programming sounds with a level of -0 dB - because there will always be sounds afterwards you need to push a bit.

Even if I actually take things like that into consideration, my goal wouldn't be to match all programs at exactly the same level. To me, it would be enough to get the monitoring sorted out so that you can always hear yourself without getting distracted by fiddling around with volume knobs during performance. Everything else is up to the FoH guy. And my job is to have fun and connect with the rest of the band.

I got that BeatBuddy virus from Normski (right?) and now am thinking about doing my own drum sets. The volume issue is even harder in that one-shot world.

But if you compare an organ sound to a piano sound - what's their loudness? - I'd bet, it also depends on the style you're playing, not only the sample and program volume.

I wish you a happy & healthy new year!

[monsterjazzlicks]

Many thanks FZ,

I'll reply in full tomorrow . . .

Have a great 2022!

[monsterjazzlicks]

. . . btw, for my own samples I use Cubase to normalize my Wavs to -2dB.

[monsterjazzlicks]

This is fully within your control. Set levels for each of your programs ahead of the gig and adjust as needed.

[monsterjazzlicks]

. . . btw, for my own samples I use Cubase to normalize my Wavs to -2dB.

If you adjust a pop or rock song to -2 dB, everything is fine; however, you would never do that to classical music as different parts will need to be softer or louder.
Talking about samples might be different. Or may be similar. I already said, I'm not a sound engineer, but I'd imagine those sample snippets need to be handled closer to my second example: It's up to your imagination what peak level you want to give to a sound; if it's an even sound without high peaks, you might need to level that sample down so it won't stand out against another sound that starts with a high peak and quickly gets down to a lower level.

In principle there are two different ways: a) set every sample to -2 dB; that way you will use most of the quality range available; but you then have no hint for the volume that sample needs to be fitted in with for a natural sounding; b) let go full quality and level off your samples to approximately the same loudness - then one sample will have -0 dB (drums) and another maybe -9 dB (legato strings).

I'm not sure which way is the better one to go, but I'd probably do a mix of both: Try to level the samples closer to each other so you might exchange without adapting volume -- then you can't set everything to a peak of -2 dB because the peak doesn't fully determine the loudness of the sample; I imagine it will not only depend on the instruments you are recording but also the style they are playing.

There are sample creators on the forum -- so I hope they will share their experience on this point. The only thing I can say: It's probably not done to just level anything to the same peak value. To make help easier, tell us what you had in mind when asking that question: Is it about the acoustic/electric piano example you used or is it about those brass samples you are working on? As I would handle sounds differently depending on the fact if they loose volume over time or not.

[monsterjazzlicks]

One more thing: Whether Cubase or Audacity - the perceived loudness of a sound, I guess, is closer to the surface the sound waves cover than to the peak points. This might be a simple approximation - but this does not take into account the different effects of different frequency ranges. So calculating loudness is a complex thing. If you have already mastered some sound events (speech or music), you probably know.

[monsterjazzlicks]

Do users of this forum just live with it or do they address the problem first-hand? ...

And to get back to your question: I definitely live with it - as I have no intention of quitting my job, abandon making music and only do sound engineering.

[analogika]

You CANNOT gauge the volume of a program without taking into account the context.

If you keep the horn patch for „Skyfall“ at the same level as the horn patch for „Sledgehammer“, the sound guy will rip your head off, if you’re lucky. If you’re unlucky, he’ll just turn you the f**** down to a tolerable level, never turn you up again, and file you under „amateur“ for the foreseeable future.

Add to that the fact that dB alone mean nothing without taking into account the nature of the sound — if you compare a full orchestra patch at -2dB to a distorted Clavinet at -2dB, the Clav is going to rip your head off.

So, short answer: getting your levels right is your job, and yes, it is tedious and difficult. I find it much more demanding than programming the actual patches.

Longer answer:
If you can get multitrack recordings of the live production without keyboards (or the band minus one), you can adjust to that mix in your practice space or on your home stereo to get the ballpark, and the live sound guy will take it from there.

I’ve found clear communication with the mix engineer (or with a competent audience member, if you’re doing your own mix) is key: let them know that you’re trying to make their job easier, and ask them to take notes if anything strikes out as too loud or inaudible. If they have time.

[monsterjazzlicks]

Happy new year FZiegler!

Thanks very much indeed for your helpful and lengthy replies. That is very kind of you to do so.

I know you are not a qualified Sound Tech. but you should not put yourself down because I think you have a great deal of knowledge and experience to offer to other Nord members such as myself.

Yes, I agree that in Classical music the elements are far more dynamic than in, say, Rock music (which is possibly why I dislike Rock so much! ).

It was very interesting to hear you views regarding a comparison between samples with, and without, peaks. Up to now I have been using the Nord pretty much in isolation - I mean the samples are played one per time and without other Nord panels being active. However, now I am looking at having it MIDIed to my (Roland) XV5050; and I can tell you that the latter module has ginormous discrepancies in volume between its many presets - the widest range of dynamics I have ever witnessed on a synth!

I think I will follow your first principal (which I was kinda' doing already) whereby I render all of my Wav samples to -2dB. That way, I will have at least some level of consistency throughout. THEN, go into my Organ and Piano sections and tweak those by ear (ie. to roughly match the Sample section components). I only use Normalize, never Limiting - and never Delay or Reverb (the latter two I apply using the Nord's FX section [if needed]).

I use my Nord Electro in a Jazz trio (keys, bass and drums). Up to now I have mostly just been using Piano, Rhodes and Hammond. But I have found that once you start adding (eg.) EQ and Drive, then these colourings have a significant impact on the volume levels of the related program. This alone is not too bad, however now that I'm starting to deploy the Synth section such elements are magnified even more so.

Regarding application of the samples, my intention (for many of them) is to utilize these as 'features'. By this I mean the band (bass & drums) won't be playing at this point. It will just be me playing a solo introduction triggering different samples for the start of certain songs. For instance, I might open the song with a sample of Darth Vader speaking for 30 seconds, and then we all segue into a version of the Star Wars theme (that type of application). Everything is live (we don't use sequences or anything).

Cheers,

Paul