Which export setting preserve the absolute same sound without truncation?

One more additional note, if I don’t dither, one of the member in the forum, stated that, Cubase automatically ‘truncates’… so that made me to do some reading on the net…

Like even inaudible margin…?

32bit float (64bit float processing) - export 24bit <— truncate

So this got me consider, ‘let’s export as it is, as perfectly exact, then do the conversion outside’

It’s 24bit, but the internal mixer uses 42bit for its processing.

As already said, You really really need to stop overthinking this…
If you want to process you audio in another software that uses floating point, just go ahead and export as 32f, the main benefit being that you don’t have to worry about clipping if you don’t have a limiter on your bus.

If you convert to 16i, dither. If you convert to 24i, roll a dice whether to dither or not, doesn’t matter, the noise floor of your electronics (not to speak of that of your room) is higher. Or just dither, if it makes you happy, doesn’t cost anything.

If you’re worried about signal degradation, everything else you do to the audio (processing with plugins, sample rate conversion) has most likely more impact on that than truncating/rounding from 64f to 32f or even to 24bit. Don’t even look at your monitor speakers! They distort the crap out of your precious audio :wink:

1 Like

Right right! Totally! I’ve been overthinking about :slight_smile:
Just wanted to let (other member) Mattias know so no confusion there :slight_smile:
(Perhaps I could actually learn from other member’s comments :))

I learnt, and now I got my masters all exported :slight_smile:
So all good here :slight_smile:

Thanks again, sorry if I sounded too attached about this small thing haha…!
Have a great day.

Well, I don’t think I’ve ever seen or heard of a converter that somehow uses floating point when converting to analog. If you think about it the maximum level is 0dBFS which equals an electrical signal according to however the converter outputs are calibrated. 32-bit float is like 1500+ dB of dynamic range. If float signals were allowed and the signal could go above zero it’d be no bueno for speakers (and ears). So the only way would be down, but we already have enough bits for that. Noise will take over anyway.

So anyway, you can trust Fese if they say it’s 24 bit fixed because that sounds reasonable, and that’s what you hear.

The whole benefit of float is that you don’t have to worry about clipping and you get this massive dynamic range. Once you’re done mixing/processing it’s of little benefit though. So ‘no’, you didn’t do it wrong. It’s just that you can’t hear a 64-bit float signal out of the converter, you hear a 24-bit signal at most.

1 Like

Right, been overthinking about it too much…! lol
Thank you :slight_smile: