When you break it down to the very basics, it’s really just numbers being added together. When working in floating point, it generally doesn’t matter too much how those numbers are added together. It only becomes problematic when dealing with very rare cases where nerdy shit like the numeric precision of floating point numbers comes into play. This is when people like to argue until they’re blue in the face over topics such as ‘summing’, but we’re not gonna get into that nonsense.
Let me put it as simply as possible…
Imagine I have four tracks and a master track.
Scenario A - Similar to how Renoise currently works.
Track 1 outputs a signal of 0.5
Track 2 outputs a signal of 0.5
Track 3 outputs a signal of 0.5
Track 4 outputs a signal of 0.5
Master Track volume is set to 0.5
0.5 + 0.5 + 0.5 + 0.5 = 2.0
2.0 x 0.5 = 1.0
Final signal is normalised and doesn’t clip.
Scenario B - How I would like things to work.
Track 1 outputs a signal of 1.0
Track 2 outputs a signal of 1.0
Track 3 outputs a signal of 1.0
Track 4 outputs a signal of 1.0
Master Track volume is set to 0.25
1.0 + 1.0 + 1.0 + 1.0 = 4.0
4.0 x 0.25 = 1.0
Final signal is normalised and doesn’t clip.
The end result in both cases is the same, but for me Scenario B makes a lot more sense.
this is not true
audio channel track volume is set @ 0.0db
impulse global vol is set @ 0.0db as each of his channels, too
simpler’s output volume is set @ -12db and display this in his vol slider
also
reaper channel volume is set @ 0.0db
samplomatic vol is set @ 0.0db, too
there is no other daw with this ‘babystting’ behaviour, also this concern VSTIs, too. their output is also attentuated
In light of all this discussion about track gain and keeping in mind the possibility that the behaviour might change, I wonder if it would make sense for each individual track to have its own clipping indicator and other related functions, such as auto-gain and soft clipping?
For example:
Obviously the master volume would still have final control over everything and you could lower this to prevent any further clipping at the final stage, but I still think it would be useful for people to see when they are pushing a particular track too hard.
i would definitely find clipping indicators + autogain per track useful.
i imagine this isn’t too hard to implement either. (ok, sorry, i know we shouldn’t bother with what’s hard and not… but still)
I did some tests, gmm04e is right about Ableton Live, don’t know about other DAW’s
making 0db 0db, will solve the rendering problem in a more elegant way as rising the sample volume.
So I think Dblue has a good point here. Now Renoise is just, lying to us about the db level
while the 32 float is at work it’s only the master where the small headroom will be, right?
so lowering the master level will solve this, right?
and so, IF we change this behavior, how will it affect songs that are written already?
Will the levels just pop up 6db lower? in the mixer?
Not really liking that idea of auto-gain or soft clip on the channels: will only make sloppy mixing habits sound even worse. Clip light might be a good idea though, but hell people learn how to use the master leveling meter and mix with some headroom! Damn it’s so simple: mix your sounds lower and turn up your monitors!
Internally this isn’t true! With the 32bit float calculations you have introduced into Renoise you could add hundreds of dB to a signal then remove it in the Master with no clipping on output if you so desired (although obviously you would if you tried to Render Selection.)
Do you see the key work setting in that quote? That is all most people have been asking for.
And then having to add it to every extra Track and Send as you expand your song beyond your Template… Yeah sure, it should be up to the user to do that at all times to get unity gain :-/
What you see should be what you get and currently that isn’t the case.
(Disclaimer: For those that don’t get it the above is sarcasm. MMD has shown himself to be a wealth of knowledge on this site at times, especially with respect to mixing, and provided a fair number of wonderful In Depth articles covering a fair few of the basics.)
As the -6dB is done at the start of the signal chain the affects the level going into any DSP, things like Compressors, Distortion, Wave Shapers and more will sound vastly different with the 6dB per channel in or out (although it shouldn’t be a problem to get the sound you want with either method, just different settings would be required.)
Also having it on the Master still wouldn’t let you do Render Selection to Sample would it? Didn’t think that went through the Master Channel…
not sure if i understand question completely… however, you can do both no? I mean, if you are rendering something that includes sends you have to include master channel… however, you can just render a track by itself…let me know if i m missing your point…
Why? Surely you should be stopping the DSP chain where you are sampling, so if you were to highlight Sends it should Render to end of the Sends’ chain with the tracks that feed them.
Highlight a section within a pattern and Render Selection. Surely the result is different if you have the 6dB Gain on each Track or on the Master.
Due to currently running off a Linux Live CD due to hardware failures unfortunately I can do zero testing to make sure I have all points straight though.
sorry…i m probably not following you correctly…anyways my point was responding to:Also having it on the Master still wouldn’t let you do Render Selection to Sample would it?
I have used 6 db gainer on master for several reasons and it renders fine…but maybe your question had other factors I am not reading into…
sorry about hard drive btw…hope you get it fixed!!
edit:Why? Surely you should be stopping the DSP chain where you are sampling, so if you were to highlight Sends it should Render to end of the Sends’ chain with the tracks that feed them.
i should respond to this…what i meant was if you are rendering something that includes send track fx etc, by default you highlight the master fx thus including them(unless you disable them)…i do this a lot to create fx patches and such…
No panic. Obviously it would be great to make this a setting. Seems that this is the only way to make it work for everyone.
