"Render Selection To Sample"-Output Is Approx -6 Db

Hey there,

had a short discussion 'bout it on IRC… I was asking why “render selection to sample”-Samples are always like 6-7 dB quieter than it’s original output. Is there any reason why? I really want to know what it is good for… furthermore I would really want to know how many dB difference there is. I guess it is -6 dB but I’m not sure on that, can anyone confirm it?

If it is -6 dB I think it should be mentioned somewhere in the renoise documentation and manual. And if it is already in the manual I obviously didn’t find it, so it maybe needs a bit more highlighting… (that is just a suggestion which could help other newbies with renoise)

the problem right now is actually that approx 6 dB difference can pretty much fuck up every mixing and mastering settings if you make anything with “render selection to sample” at final stages of a song…

Yeah, I wish that 6db headroom was optional. It’s not like you don’t know what levels your channels are at, or if the resulting rendered sampled has been clipped.

so what is the exact difference then, is it exactly -6 dB?
It’s not a big problem if I know the exact difference… because then I could always amplify the samples by this value to get the original value.

The magic number you’re looking for is a rather strange one, actually:
6.0205999132796239042747778944899 dB

…or simply (and very slightly less accurate):
6.021 dB

Why this number? Well, it’s a long story, but I believe it relates to the fact that gain levels in Renoise used to be represented as percentages rather than dB, so you would typically be making adjustments of 50% or 200% and things like that. (Edit: I should clarify, this is also quite common in other apps as well)

50% gain = -6.0205999132796239042747778944899 dB

200% gain = +6.0205999132796239042747778944899 dB

So I think that the “render selection to sample” feature currently outputs at 50% of the master output level, resulting in this odd dB level that we have to compensate for afterwards.

See this thread for reference:

It is rather unfortunate quirk, but at least now you know. Some of us have suggested in the past that this reduction in volume should be an optional thing - something that can either be adjusted or turned off altogether if the user chooses to. Hopefully it will change in the future so we can avoid this type of confusion.

thanx for the summury dblue.

but the open question still is, why is the “record selection to sample” output being reduced to lower gains? As long as I can’t get any plausible explanation I can only reguard it as an annoying bug…