|
How Do You Bounce Your Audio?
|
View this Thread in Original format
| Rodas |
One thing I notice is that bouncing to audio never sounds the same as the original. It looses a bit of it's lust, and reverb and other effects never sound the same as they do in the sequencer...
So I guess the question is, is there a way of getting around this? I'm curious to hear ways to do this and still make the production process not so tidious as it would be.
Rodas |
|
|
| emc^2 |
well, sure you'll probably lose some of the "luster" because you're probably recording in a higher resolution than you're bouncing. Say, you're recording in 24bit/48khz - when you bounce it it becomes 16/44. ONe way is record at higher resolution: 32/96 or 24/88. It can also varry on your plug ins. Keep in mind that hi res is there not for showing off but for "added headroom" - use it wisely. And if your mixes loose luster, do "post bounce" FX'ing. So, you'll be bouncing several times but this way you can tell where the problem is. I usually combine all my audio to 2 tracks and when I'm happy, I then FX it and bounce it again. If it sucked, I just need to tweak FX's/eq/compression/etc.
also make sure you're doing a clean bounce (e.g. no other plug ins are being used)
Also, try bouncing to 48 AIFF and test it out in WINAMP or other playback proggy. If it sounds the same, then it's your resolution and it is time for you to upgrade to higher res. |
|
|
| Rodas |
Hm, ok what's your definition of bouncing? I got a bit confused reading your post.
Boucing I thought was having your plug in's going, and sequencing the parts how you want them in your final track.. Then when you think there finished you bounce to audio to save cpu and to add more... Right?
Rodas |
|
|
| emc^2 |
| quote: | Originally posted by Rodas
Hm, ok what's your definition of bouncing? I got a bit confused reading your post.
Boucing I thought was having your plug in's going, and sequencing the parts how you want them in your final track.. Then when you think there finished you bounce to audio to save cpu and to add more... Right?
Rodas |
Ok, so I guess then you're talking about bouncing from plug ins. E.g. composition written with soft synths, later translated to Audio wave. I c. Well, I have a little trick I use to preserve some of that freshness (may not be for everyone). I route audio signal from my PC out to mixer, and then back into PC in form of audio track - so there's no "bouncing" involved - just regular audio track recording. But I did have the same issue you're describing with some VSTi's - Spectrasonics seem to be the biggest offenders. That's why I started doing tricks like that. |
|
|
| DigiNut |
| quote: | Originally posted by Rodas
One thing I notice is that bouncing to audio never sounds the same as the original. |
Sounds pretty much the same to me, you must not have your sequencer set up correctly for previewing.
Either that or your master is going over 0 dB and it only becomes obvious on the bounce when it actually clips. |
|
|
| Derivative |
1) the imported wav is playing back at a lower gain than the synth playing back in realtime. use a dB meter on both channels and make sure they both peak at exactly the same level.
2) if you record a synth directly from one channel, but there are processors and effects on the master bus, then those processors are ignored. you need to record the output of the master bus to include those. this sometimes results in bouncing the results for 1 channel but because you have a limiter or a compressor on the masterbus say, it always sounds quieter and weaker.
3) record in a bit rate/sample rate that is equivilent or greater than the bit rate/sample rate of the synth you are recording from. for a safe bet, record everything in 24/96 or keep it all floating point.
4) as digi says you may have configured your synth to audition incorrectly. ages ago, some dude on this forum kept asking why his virus b always sounded fatter when it was playing back than when it was recorded. turned out he had the output of the synth playing back through his soundcard mixer as well as playing back through fruity's mixer channel - he was listening to the same sound, playing back twice (and it sounds hella loud). recording the output of that channel in fruity will of course record only the input fruity's mixer receives. in that case, he just set things up wrong. |
|
|
| Biatchzxz |
I have the same problems as my sounds sounds terrible after bouncing. From let say Midi to Audio. It loses that Stereo Quality. I will try some of the things said..
Is bouncing to Aiff better then wav or just different |
|
|
| DJFreaq |
| quote: | Originally posted by Rodas
One thing I notice is that bouncing to audio never sounds the same as the original. It looses a bit of it's lust, and reverb and other effects never sound the same as they do in the sequencer.. |
=
Reason
(I'm guessing here, but I thought I'd share my two cents about it anyway)
But I remember the days when I used Reason heavily. For some reason, when I bounced. An intangible something was lost from the mix. Not sure what it is. I had my audio-card settings the same bit-depth and sample rate as my bounce settings.
I never have figured it out totally, but I attribute it to two things:
A. I sucked
B. Reason sucked (but honestly, for bouncing. I would considering using a different program to compress and mix your final track)
There. My two cents. |
|
|
| Diginerd |
No neccesarily either.
What you may be is the victim of Jitter. Assuming you have not made any of the above obvious mistakes.
All digital audio exhibits this to some extent, but most of the time it's undetectable.
Even my TDM Pro Tools rig actualy suffers from this problem, though why I can't assertain even when running around with Digidesign for hours at a time.
If I "Bounce to Disk" I loose stereo imaging and my dynamic range "Feels" more limited, also I loose some of the shine off my parts.
If I hook an AES/EBU cable from the main digital out and pipe it back into a spare stero track (Muted of course to avoid feedback)the problem is substantially reduced.
Piping that same AES/EBU out to a Tascam 24 Bit DAT machine and then recording it back in into a new session produces no loss.
All of the above is verifiable by an old Digital audio trick.
Take two supposedly identical audio files, line them up perfectly (By which I mean to the sample level) and invert the phase on one. If both are kept at unity gain then the resulting output file should be absolute silence (Pure zeros). This resulting file can itself be captured and then normalized. In a perfect world it will remain all zeros. If not then you get a world of uglyness as you can clearly hear the resultant out of phase audio.
Every one of the above bounce methods except the dubbing out to the 24 Bit DAT machine produces varying degrees of artifacting on analysis.
It shouldn't. But it does.
This is a subject of debat on a collection of high end audio boards for some time, with both sides of the fence getting very heated.
I don't really care what people say, I notice the problem, and I have found a solution. YMMV.
Point to note, the above only applies if you are working DIGITALLY. If you are doing an analog bounce as someone mentioned then you are introducing another set of issues by passing through an AD/DA conversion, which in iteslf poses it's own set of compromises.
Ultimately, do whatever you feel makes things sound best. |
|
|
| DJFreaq |
Diginerd pwns again.
Thanks for that enligthening post. Next time I do a bounce I'll use the AES/EBU outs on my Motu 2408. At least in reason anyway.
:) |
|
|
| Biatchzxz |
| I am using Logic 5.5 for PC. Everytime i bounce from audio inst. to Audio. Even after normalizing it still sounds a little worse. How do you Bounce your full Track after you complete it and with what setting 16 or 24 bit. Logic 5.5 Doesnt go past that as far as i know. I would usually just bounce my Output1-2 to Wav at 24bit. Is there any other way to do it better than that. ? |
|
|
| Diginerd |
Ive just fully twigged what the OTHER problem some of you are facing is..
Ok, if you are bouncing a VST down, make sure that the last plug that the signal goes through before hitting the disk is a dither set to 16bit if you are creating 16 bit files, or 24 bit dither if creating 24 bit files from a 32 bit source.
Waves IDR, type ultra 1, with relevant resolution should help get you started.
This actually holds true for most bounces, put the dither at the last thing on your master bus just before bouncing.
The quick and dirty explanation here is that your sequencer's internal bit depth is quite likely higher than your output file will be (As has been said before). that actually holds true if you are playing back 16 bit files, EQing and effecting them and bouncing out as 16 bit
your Sequencer's environment is probbably 24 bit, and maybe 32 bit.
Either way savagely changing bit depth without dither created truncation artifacts. This produce som ugly sounding effects, inlusing some of those described.
So my 2 top tips are:-
1. As in my last post, see if you can at least make a physical recording of your output (This means realtime peeps!), ieither out and back via digital I/O, or out to an external computer or in my case DAT machine
2. MOST IMPORTANT Whenever changing bit depths you need to dither your output to the recording depth. This removes (Well, hides) truncation artifacts and makes things sound much nicer.
Dither must be the last thing you do. Don't go trying to dither every track.
If you want a discussion on dither I'm happy to have an in depth talk about it... |
|
|
|
|
