How loud do patches need to be and how to measure them?

[Marco Raaphorst]

I am working on a ReFill. Synth patches. Wondering how loud they sound be and how to measure them.

Are you checking peak level only? That will make some sounds, sound way lower than others.
How many notes do you play? One note makes less noise than several sounds, and playing high notes can make a hell of a difference.

I am a perfectionist, so I am just asking. Any sound designers who'd like to comment? Thanks anyway!

[Loque]

I try to touch -12db peak when i make patches. Everything depends on the patch, velocity->gain, poly/mono.

I try to make the average velocity hit at -12db. If a 3 or 4 key chord gets too loud, i reduce the volume.

That is, what i observer on PH patches too.

loudmax.jpg (51.89 KiB) Viewed 574 times

[Marco Raaphorst]

I try to touch -12db peak when i make patches. Everything depends on the patch, velocity->gain, poly/mono.

I try to make the average velocity hit at -12db. If a 3 or 4 key chord gets too loud, i reduce the volume.

That is, what i observer on PH patches too.

loudmax.jpg

thanks. I'm checking the Europe patches but they are all over the place. some peak at -11 others at -2.

danger with peak is enormous volume differences when you preview patches one by one.

[fotizimo]

And VSTs seem to be so much louder by default than rack extensions. I have to keep reminding myself that any time I open a VST instrument, to turn down the gain for the channel by 6db or so.

Does anyone know why VSTs generally are louder?

[selig]

I try to touch -12db peak when i make patches. Everything depends on the patch, velocity->gain, poly/mono.

I try to make the average velocity hit at -12db. If a 3 or 4 key chord gets too loud, i reduce the volume.

That is, what i observer on PH patches too.

loudmax.jpg

thanks. I'm checking the Europe patches but they are all over the place. some peak at -11 others at -2.

danger with peak is enormous volume differences when you preview patches one by one.

I see it less as a danger and more as an advantage, as it shows instantly the relative loudness between the patches (which is extremely useful).

When I worked on patches for the FSB on Reason 6, the rules were peaks of -12 dBFS for patches. Solo/lead/bass patches were one note was the most you could play were measured by playing one note, while poly/pad/etc patches were tested with 4-5 note chords.

FWIW, every time I’ve heard patches jump in level, I noted the peak level also jumped. I’ve not heard many patches with greatly varying crest factor, at least in my experience.

And as far as I’m concerned, peak levels jumping around is far less useful to me than RMS level slightly varying (especially in the digital domain), which is natural and extremely informative as previously mentioned. It’s great to have patches that are not all at the same crest factor/loudness, and when you do I prefer similar peak level for all the advantages already mentioned.

And since I already set ALL audio signals to peak around -12dBFS, it's second nature for all patches to hit the same level.


Sent from some crappy device using Tapatalk

[Kalm]

I started mines too. PPM -12dBFS cause the PPM average metering is usually the closes to average level metering to my ears. I can switch between sounds 90% of the time without having to adjust for volume. It also makes my effect routing that much easier.

[bitley]

Dynamic patches like EPs require heavy playing with loads of voices to ensure it never crashes into the roof

[Marco Raaphorst]

thanks all!