So I am not alone in thinking that the synths ( or whatever ) I record through here sounds a bit better. I think it does slightly sound better after I record bounce and import back into sonar.
I too am finding going to Voxengo products A LOT more than any other plugs.
Thanks again for the verification, and of course, Aleksey for the awesome products !
Yes, when you set 24-bit output, it dithers incoming 32-bit floats to 24-bit as well. When your output is at 32-bit, no dithering takes place.
However, this seems to be very strange... If bouncing even to 32-bit floats sounds worse than through the Recorder, something strange is going on in Sonar for sure, or it can be an effect of several gain/pan stages that are present in the signal path (with Recorder in-track they are obviously eliminated).
Sure - makes sense to me now.
Reason - I hear a difference; Sonar - not so much ( freezing )
Maybe I just 'think' I hear a difference in Sonar because the reason synths do
due to 44.1K/16bit ( reason) and 44.1-88.2K/24 bit (sonar)
Maybe I dont know how to explain what I'm hearing ( and doing!)
It either sounds the same or better, not worse, to my ears anyway.
Either way I like the alternative and it works for me !
Stratcat - I would describe the difference between bouncing and "recording" (this will now be the word for using the recorder) like as if the bounced version has...a bit more smeared high end. If I have any cymbal in my mix, the recorded version gets the cymbal right while the bounced/exported makes it kind of phasey.
I also noticed that a recorded version feels like it has better dynamic ratio - I think it is subject to an overall clearer clone of what's really exits the Sonar mixer, and this imply that there's something weird about bouncing and that it kind of shrinks and flattens and harshens everything a bit.
I don't think there's anything wrong with the Sonar Bounce in contrast to other DAW's bounce-functions, but I think in general that the freezing and bouncing functions are subjects to not being clocked in real time and therefore gets a slight jittery effect.
Could be that the host is kind of "placing samples" when calculating the mix and this might give me this impression of jitter when played back. but I can't say for sure.
Make sure you are bouncing at the same sample rate as your project's samples are. If bounce ends up resampling your audio data, then you may experience some loss in quality.
Bounce or playback of any audio host has an ideal clock-rate, so to say - no reason to think it's jittery. Jitter may happen only on audio input or output.
I think you explained it
Phaseyish when bounced..
Clearer- when recorded and imported
Anyway - recorded and imported sounds a bit better to my ears,
especially the reason synths.
and Aleksey - I'll keep the rates equal - good suggestion to remember
I'm bouncing and exporting at the same rate...
R8brain takes care of my possible conversions :)
Well. The trials will continue.
I'm making one "export audio" and one "recorder" version of the same track. Everything is exported/recorded at 32 bit float in both cases. I have 96 K samplingrate in my project and at export.
When importing the tracks together in Sonar and flipping phase on one of them there is NOT silence in the beginning. After the low volume intro there is silence.
I open two instances of Cool Edit to analyse both tracks at the same time (selecting the same timerange by manually enter start and stop time for the portion). In parts where there is silence with flipped phases (where the files null out) Cool edit gives different values for highest peak, lowest RMS, DC offset and so on. I carefully make sure that the same portion of the tracks is analysed.
When I loop this one second of music (declining waveform) over and over again there is a subtle difference between the two versions; the recorded sounds like it has more bottom to the midrange and overall have a softer and rounder sonic character.
When I loop the bounced portion, the loop seem to have a higher speed than the recorded loop of the very same second of music.
All this, I tell myself, might be the result of 32 bit floating points and that two versions won't be the same. But I don't know.
Could it be that a bounced version is completely linear while the recorded is not? Can they still cancel each other out and causing silence if one of the versions have its phase flipped?
Investigations might continue, but there's also music to be made so maybe the case is closed now.
This topic was last updated 180 days ago, and thus it was archived. Replying is disabled for this topic.