You are not logged in. Please login or register.
RME User Forum → FireWire & USB series → 32 bits - 32 float
Hi
I record with Sound Forge but when I try to set it in 32 IEEE (float) it doesnt allow me. It only let me when I do it in 32 bits. Would someone clarify me whats the problem, or should I only attempt to do it in 24 bits ?
I own a Fireface UC
The hardware supports only 24 bits audio data because there is no way to receive/transmit more over ADAT/SPDIF/AD-DA. This has nothing to do with the internal mixer or audio engine of SF, which might be able to run at 32 bit float independent from the I/O audio word length.
May I ask what SF does when it records in 32 bits but the rme hardware only allows 24 bits ? how should I turn this audio I recorded in 32 to 24 bits ?
thanks very much for your help
Quick and easy explained a 32 Bit float is a 23 bit signal with additional bits (exponent) to scale it over the complete dynamic range. In contrary to a fixed point engine a floating point engine don't lose wordlength (bits) with a volume reduction and does not add noise. Rounding errors are not a problem like in a fixed point systems.
Recording to 32 bit transfers the fixed point numbers to the floating point format (mantisse 23 bit, exponent 8 bit, 1 bit for sign of the mantisse). When Soundforge handed the signal over to the driver of any audio hardware it gets truncated to 24 bit fixed point.
best regards
Knut
I think it mainly has to do with plug-ins as in like a VST plug-in whatever it is.
It still ends up 24 bit like in Sonar (Cakewalk) but the plug-in is handled originally in 32-bit plus mine is set for 32-bit engine but I record normal tracks (takes) in 24-bit. The software automatically handles that just for rendering purposes and such!
That's also to handle effects used for those plug-ins, but in the end it ends up 24-bit.
(software rendering engine of the program)
RME User Forum → FireWire & USB series → 32 bits - 32 float
Powered by PunBB, supported by Informer Technologies, Inc.