Have made a quick search of the forum for anything regarding the audio quality when bouncing the complete mix vs the audio being played.
Didn’t find anything so here goes…
The problem is:
Have tried multiple settings but everytime i bounce the mix down to a stereo track i always get an alterd sound.
Mostly it is noticeble in the stereo field and a general bland/flat sound.
Now is there anyway to get the EXACT same sound that i get playing the rns, when streaming the whole thing down onto a .wav ?
Once I got sound changes in the rendered file because I was using two VSTis in four different tracks… Maybe it’s your same problem…
Did you try to render the song in low prioity mode??
Did you use VSTis which read directly from the disc??
Do you encounter random sound/FX changes or are they always in the same parts of the song??
I’ve had the same problem when freezing VSTi+VSTfx tracks. Regardless of the samplerate and quality settings, the rendered sound was different not only in terms of volume, but the spatial characteristics as well - it sounded flat and stripped of the stereo image.
Fortunately, I don’t have to freeze those tracks anymore since PolyIblit has been updated and it’s CPU usage dropped down, but the problem remains.
Still, I hope that version 1.5 will be improved in terms of What-You-Hear-Is-What-You-Get…
I think that the problem may be hardware-wise. Rendering means absolute ideal mathematical exporting of the Renoise output, while when you stream the output live it might be trasfered through the D/A and A/D which can improve the quality in analogical way - adding the warmth like analogue hardware does. I think it’s not any flaw of Renoise, it may be a matter of your soundcard. I may be wrong, but you should check somewhere thw path of thw sound when recording the PCM out.
But anyway, if you feel that recording makes better results why don’t you just… record? (I know… the speed…)
Nah, DDSpeed could be right probably. Not that I made such an experiance but it could be that your soundcard sounds different once you use ASIO and while you play simple waves through Direct Sound or whatever it might be different again. But still this would depend more on drivers than on hardware. Maybe you could try this: Export your song as you do usually and load the exported wave as a sample in Renoise and play it there again. But I actually wanted to suggest something different:
If you’re sure that it’s Renoise which brings two different outputs, you can use TotalRecorder. It’s a virtual audiodriver which streams down the sound that would usually go directly into your soundcard. I suggest this because then you can be sure that your sound has been streamed lossless to disk.
And if Renoise lets you to do so, you can stream with full CPU-power (i.e. faster than realtime). However, most modern soundcards record digitally if you do internal recordings regarding the quality issue.
That’s exactly what I did. I froze a track, loaded it in Renoise and played in a song in place of the VSTi generated sound - and it did sound different, flat. (And no, it was not the matter of a missing reverb send )
You’re right, but the algorithm used when playing may be different to the one used during rendering - in terms of interpolation for example. So the output might not necessary be the same in both cases (especially if there’s a bug in the rendering mechanism).
Which interpolation mode have you chosen for the rendering?
The player uses the cubic, not the sinc (which is slower).
But that should not affect VSTi’s…
After trying out multiple .rns’s and the diffrent interpolation modes,
i do hear that the cubic is actually closer to the .rns audiofeed than sinc.
Still i guess i could try TotalRecorder but the problem,
is that our .rns’s are VERY cpu intense.
We usally finish the projects around 96-100+%.
So it’s not possible to get a skip free playback,
but works really fine if you bounce it to disk…
Now for some computerspecs:
Amd xp2600+, 512 mb 333mhz, a bunch of hd’s (some scsi).
Last and most important an - Ego sys WAMIRACK 24 sound/midicard
I should also mention that all the audio is accuratly played,
but the problem is the audio QUALITY.
Though this might not be obvious on a regular speaker system,
it sure is obvious in a studio enviroment with proper monitors.
As far as I remember, I used sinc + 16 & 32-bit float. The track I bounced contained PolyIblit chords played through some spatial effects (mda Stereo and built-in reverb if I remember correctly). The rendered output sounded as if it got monophonized - flat and muddy.
Unfortunately, as I’m working in a different country for the next month or so and haven’t taken my computer with me (well, it’s a desktop), I can’t send you an .rns with this track nor check if I can still reproduce this effect, but do that as soon as I get back home (with version 1.5 perhaps? ) and let you know.
just to hop in on this thread cuz i have similar issues…
I just got my new Emu 0404 card 2 days ago and sofar really happy with it. www.emu.com for specs…
After bouncing back sum tracks, i noticed a rather big difference in audio volume and flatness in quality. I dont remember having this when using my directx Hercules Fortissimo II…which obviously did not have Asio drivers.
Since im new to Asio and DSP i figured i could do some searching in Renoise Forum, resulting in finding this thread.
I havent done any further tests but the difference between ‘playing a song’ and ‘rendering a song’ is really there… which is somewhat of a bummer cuz i thought it sounded really good when playing.
Any further insights would be appreciated…
Looks to me its the Asio to Wave transfer causes the difference…
The rendered output quality has nothing to do with your audio drivers - they’re not used for rendering.
The drivers could be the case if Renoise playback sounded different from standard windows WAV playback - the ASIO sound might go through some extra processing (like EQ) that standard (or DirectX) playback doesn’t get.
But if you’re playing your rendered WAV files using Renoise (and I guess that’s what you meant by bouncing), then:
If the rendered WAV’s are simply quieter, the cause might be that Renoise uses some dynamic headroom (of 3dB if I remember correctly) to avoid clipping. You might try to increase the master level for the rendering and see if it fixes the problem (but watch out for clipping).
If there’s more to the perceived quality loss of the rendered wave than just volume - like narrowed stereo field, you might be getting the same effect I did once. It would then be best if you sent your RNS with the description of the rendering settings and the percevied change in sound to the devs so they can fix the bug.
The rendered output quality has nothing to do with your audio drivers - they’re not used for rendering.
i know i was just suggesting the difference between Wave and Asio volumes.
The drivers could be the case if Renoise playback sounded different from standard windows WAV playback - the ASIO sound might go through some extra processing (like EQ) that standard (or DirectX) playback doesn’t get.
asfar as i can tell i have no fx added in my Asio.
But if you’re playing your rendered WAV files using Renoise (and I guess that’s what you meant by bouncing), then:
If the rendered WAV’s are simply quieter, the cause might be that Renoise uses some dynamic headroom (of 3dB if I remember correctly) to avoid clipping. You might try to increase the master level for the rendering and see if it fixes the problem (but watch out for clipping).
could be but i remember running this track though Wavelab and Normalized it…but still it wasnt as loud as just playing the track with Renoise.
(by bouncing i meant just Rendering the track to wav and playing it in mediaplayer for instance, i didnt test it in Renoise but i will try this and listen for differences)
If there’s more to the perceived quality loss of the rendered wave than just volume - like narrowed stereo field, you might be getting the same effect I did once. It would then be best if you sent your RNS with the description of the rendering settings and the percevied change in sound to the devs so they can fix the bug.
there is indeed also some quality-loss but hard to tell (volume diffs is most pronounced) so maybe i will send the .rns and rendered wave.
(I know… the speed…)
At some point that also has to pass a D/A…
Nah, DDSpeed could be right probably. Not that I made such an experiance but it could be that your soundcard sounds different once you use ASIO and while you play simple waves through Direct Sound or whatever it might be different again. But still this would depend more on drivers than on hardware. Maybe you could try this: Export your song as you do usually and load the exported wave as a sample in Renoise and play it there again. But I actually wanted to suggest something different:
)
) and let you know.