Bouncing and Conversion question

[guy.fi]

I'm not sure where I picked up this idea (probably from the DUC or an engineer friend) but I have been bouncing my tunes twice from 24/48...First to 41k and then dithering down to 16 bits. I thought that by seperating the processes the sound would be better. Yesterday it occured to me that even when I do it in one step, the sample rate conversion is done after the bounce, so the computer isn't doing all the math at once.

Is there any real advantage of doing it in 2 steps? I've never heard any artifacts when doing it in one step.

[DigiTechSupt]

I'm not sure where I picked up this idea (probably from the DUC or an engineer friend) but I have been bouncing my tunes twice from 24/48...First to 41k and then dithering down to 16 bits. I thought that by seperating the processes the sound would be better. Yesterday it occured to me that even when I do it in one step, the sample rate conversion is done after the bounce, so the computer isn't doing all the math at once.

Is there any real advantage of doing it in 2 steps? I've never heard any artifacts when doing it in one step.

None whatsoever - the only 'benefit' being that, if you're close to running out of CPU power, doing the 'convert after' may give you a little extra cpu time to make it through the session.

[logicmidiman]

YOUR FINAL CONVERSON RATE WILL BE 16 BIT...BECAUSE YOU WANT TO PUT IT ON CD.....WELL REGARDLESS HOW MANY STEPS YOU TAKE TO BOUNCE IT DOWN IN THE END YOU WILL STILL HEAR A 16 BIT FILE MEANING LESS DYNAMIC RANGE....RIGHT?.......WHEN EDITING....YEA...YOU MIGHT WANT 24 BIT AT 48 kHZ PS....BUT IN THE END FOR WHAT YOU ARE DOING 16BIT IS THE GOAL....SO UNDERSTAND YOUR SIGNAL FIRST OF ALL....WHY ARE YOU CONVERTING A/D AT 48 KHZ RATHER THAN 44.1KHZ...24 BIT RATHER THAN 16BIT ...CAN YOU ANSWER THESE QUESTIONS?...BECAUSE IF THE LOUDEST FREQ YOU HAVE IS 20 KHZ OR LESS...THEN A SAMPLE RATE OF 44.1 KHZ WILL GIVE YOU 100% INFORMATION WITHOUT DISTORTING....24 BIT SLOWS YOUR COMPUTER DOWN AND MAKES BIGGER FILES...NOW IF YOU WANT BETTER PERFORMANCE YOU NEED TO GO 16 BITS..UNLESS YOUR COMPUTER CAN HANDLE THOSE HIGH BIT DEPTHS....YOU HAVE TO UNDERSTAND WHAT YOU ARE DOING....WITH 16 BIT DEPTH AND A SAMPLE RATE OF 44.1 KH PS...YOU CAN GET 100% OF YOUR ANOLOG SIGNAL'S INFORMATION AND A STANDARD DIGITAL DYNAMIC RANGE WITHOUT OVERDOING YOUR PROCESSING...
P.S.-IF YOU DON'T UNDERSTAND ANYTHING I JUST SAID...YOU SHOULDN'T BE TALKING BITS AND SAMPLE RATES...LOL...GOOD LUCK

[albee1952]

YOUR FINAL CONVERSON RATE WILL BE 16 BIT...BECAUSE YOU WANT TO PUT IT ON CD.....WELL REGARDLESS HOW MANY STEPS YOU TAKE TO BOUNCE IT DOWN IN THE END YOU WILL STILL HEAR A 16 BIT FILE MEANING LESS DYNAMIC RANGE....RIGHT?.......WHEN EDITING....YEA...YOU MIGHT WANT 24 BIT AT 48 kHZ PS....BUT IN THE END FOR WHAT YOU ARE DOING 16BIT IS THE GOAL....SO UNDERSTAND YOUR SIGNAL FIRST OF ALL....WHY ARE YOU CONVERTING A/D AT 48 KHZ RATHER THAN 44.1KHZ...24 BIT RATHER THAN 16BIT ...CAN YOU ANSWER THESE QUESTIONS?...BECAUSE IF THE LOUDEST FREQ YOU HAVE IS 20 KHZ OR LESS...THEN A SAMPLE RATE OF 44.1 KHZ WILL GIVE YOU 100% INFORMATION WITHOUT DISTORTING....24 BIT SLOWS YOUR COMPUTER DOWN AND MAKES BIGGER FILES...NOW IF YOU WANT BETTER PERFORMANCE YOU NEED TO GO 16 BITS..UNLESS YOUR COMPUTER CAN HANDLE THOSE HIGH BIT DEPTHS....YOU HAVE TO UNDERSTAND WHAT YOU ARE DOING....WITH 16 BIT DEPTH AND A SAMPLE RATE OF 44.1 KH PS...YOU CAN GET 100% OF YOUR ANOLOG SIGNAL'S INFORMATION AND A STANDARD DIGITAL DYNAMIC RANGE WITHOUT OVERDOING YOUR PROCESSING...
P.S.-IF YOU DON'T UNDERSTAND ANYTHING I JUST SAID...YOU SHOULDN'T BE TALKING BITS AND SAMPLE RATES...LOL...GOOD LUCK

Sorry to rain on your parade, but with 9 posts and this response, you probably should not be talking at all. First off, PLEASE PLEASE PLEASE turn off caps lock. It is geek-speak for yelling and is extremely hard to read. Next, learn to keep your mind open so you can gather information before you offer up info that is just messed up at best. Sample rate has nothing to do with distortion. The frequency (you mention 20KHz) has absolutely nothing to do with loudness. Some people work in 24 bit because they know that it yields much better headroom and any extra load on the computer is very minor(unlike the difference between 48K and 96K). 24 bit files are about 10$ larger than 16 bit files. As for 44.1K vs 48K, if the end product is meant for video/DVD, then 48K makes perfect sense. I do agree that if CD is the final goal, then I work in 44.1K to avoid needing a sample rate conversion(although some feel that the conversion can actually smooth out the rough edges of digital).

[mano111]

I've been working in 24 bit since I started using PT (5.1) on my Pentium III 500. I don't think its too tasking for the processor, maybe slight HD usage increase but the benefits are huge IMO. Not only in terms of dynamic range and detail in recording but in the way plugins process the audio.
44k to 48 is not much of a diffrence and really should depend on your target media (CD/DVD).

Is there any real advantage of doing it in 2 steps? I've never heard any artifacts when doing it in one step.

One step bounce is all you need.
Yet I preffer recording to a track and then exporting, makes it easy to create accapella and instumental mixes (as playlists) and you can monitor meters and everythig else as it plays/records .

[IntelDoc]

I too used to think that a 48/24 to 44/24 to 44/16 bounce was needed but have long gotten past those days. Now I do not "bounce" at all but rather sum back in through my API stuff and then simply export out the stereo track to whatever session format I was using (96/24, 88/24, etc) and send that off to the mastering house, or into a new session if I am mastering it. I think that the Pro Tools Bounce engine has come along way since the 6.0 era and really sounds better now.

Though, I do still work at 24 bits as I like the headroom and also at track at 48K. 96K to me does not sound good and like 88k much better. Math during the conversion is better.

On thing to think about is this...

If you are mastering the project yourself bounce the file down to the format you are in, so lets say 48/24. Now, open a new 48/24 session and import that audio into the session. Apply Mastering plugs, etc and then bounce that down to 44/16 respectively. Allow yourself the extra headroom that 24 bits allows when mastering.

Same goes for if you are sending the material out to a mastering engineer. Keep it at the session rate you started with and do not apply mastering plugs when sending it out.

Best,

Doc