pan law question and 64bit mixing question

Discussion about: tracktion.com
RELATED
PRODUCTS

Post

Well, there is no audible rounding-errors. The rounding-errors occur at a way too low level for anyone to hear it.

Post

Cascaded errors will eventually be audiable. Is it not good to know that you can reduce these on a technical level?

Post

Well, no they actually don't usually. We're talking about something like a 150dB SNR here. When adding two signals, both with the same SNR, you will end up with... The same SNR.

Post

The thing about floats is that they never have the same SNR. At least you cant count on that. A round-off is never the same.

Post

Please elaborate.

Post

In floating point qudio a tiny quatization noise is added at each and every gain stage - this could be quite a few times in each track. In 64-bit these artifacts will be smaller. Granted the quant noise is somewhere around -105db but it could in certain extreme situations leak into the audible range. So 64-bits will be at least some minor improvement. I work 32 then render 64. Why not?

Post

An what do you think, how much quant. noise is added with each filter stage?
Solution: don't use EQs, FXs and volume-sliders! :P

Post

semiquaver wrote:Granted the quant noise is somewhere around -105db

105? Is that a typo? Unless my math is failing me, at 24bit precision, the quantisation noise should be more in the -140dB area, which really is pretty quiet.
Someone shot the food. Remember: don't shoot food!

Post

I'm new to T2, just registered yesterday, and have a lot to learn (reading, searching, you know the drill!) as long as this thread was already here I thought I'd ask.

Does the 64bit depth get rendered when a track is frozen. The reason I ask is because Sonar4 renders the frozen tracks at 24bit even though the mixing engine is 32bit float, Adobe Audition renders the frozen track at 32bit float. I have both of those and am just comparing various things. Thanx !

Post

I really don't know foir sure, but I believe it is always 32bit.
Someone shot the food. Remember: don't shoot food!

Post

The 64 bit mixing is really just to be able to compete with other sequencers on paper.
Every day takes figuring out all over again how to f#ckin’ live.

Post

Cool beans - thanks guys!

Post

32bit mixing sounds like a 16k-encoded mp3 played underwater. 64bit mixing sounds like an old tape while someone is flushing the toilet in the background. 128bit is muffled like how south park's kenny would hear through his hood. 256bit lacks of bass and hisses enough to break your ears. 512bit mixing aliases. Only 1024bit starts to be good.
The next FL will use a revolutionary 1025bit mixing. The extra bit that makes the difference.

Post

Sure wish there was a pan-law switch in T2 - the current pan law makes things sound much too loud at extreme r/l, especially when automating panning accross the stereo field.

64-bit: don't really care. I agree with chagzuki - its a marketing thing mostly.

Post

floyd wrote:Sure wish there was a pan-law switch in T2 - the current pan law makes things sound much too loud at extreme r/l, especially when automating panning accross the stereo field.
you can always use a patchbay filter to achieve a more user customized panning response.

Also, there's nothing to stop you using 3rd party vol/pans that give the behaviour you are looking for.

I agree it wouldn't hurt for Tracktion to offer a few different options though.
Someone shot the food. Remember: don't shoot food!

Post Reply

Return to “Tracktion”