Skip to main content

I have some tracks that are originally 16 bit, wondering what the benefits or consequences would be to convert to 24 bit to mix and then reconvert & dither to 16 bit for CD burning? Thanks for all opinions!

Comments

Ammitsboel Tue, 08/31/2004 - 16:38

David, What about with regard to any plugins used during mixing? It seems the popular position is that effects tails, overall quality is better in 24 bit. If it won't adversely affect the audio, would it make sense based on the improved plugin performance? Thanks.

With regard to anything, if you put 16bit files into a mix session and process in 24bit or higher you will have a contense in 24bit or higher before you dither down to 16bit.

If you simply just converted the 16bit files to 24bit or higher they would still play like 16bit files and you would have the same result as above, just with bigger files on your HD.

There has to be some sort of processing done to a 16bit file before you can save it in 24bit or higher and make use of the extra wordlength.

Best Regards,

Ammitsboel Wed, 09/01/2004 - 02:07

Henrik, So I'm better off applying effects and mixing in 16 bit for 16 bit .wav files, do I understand you correctly? BTW, I love Copenhagen, one of my favorite cities! Thanks for your interest.

Always process at the highest bitrate!
What I wrote before was that a 16bit file and a 16bit converted to 24bit file would both play the same in any program.
But as soon as you start processing or editing a 16bit file your program will suddently play at 24bit or higher.
So it's important that your program is processing in 24bit or higher to be able to do a decent mix and edit.

When talking about digital format and processing there is 3 differnet things to have in mind:

-Bitrate of the file(s) (this is your 16bit files, they wont change by converting them to 24bit)
-Internal processing (at how high bitrate the program process the files)
-End format (Is it going to be on a CD with dither or something else...)

I hope this helps :roll:

Thanks for your kind words! :D

Best Regards,

jonyoung Wed, 09/01/2004 - 08:41

Thanks Henrik. I record on a 24 track, 24 bit hard disk modular deck, but 16 bit is an option. I did a very small voiceover project for someone in 16 bit before starting a second round of music recording for a client, several previous songs recorded at 24 bit/44.1k. The recorder defaults to the settings of the previous project, and I noticed/remembered this after most of the new recording was done :roll: I transfer files on a firewire drive to a DAW and mix in Sonar, which I believe processes at 32 bit float. I tried 48k for several projects, but the sample rate conversion algorythm in Sonar does introduce some noticeable sonic changes. The end product will be CD, but the client will be having professional mastering done, so I anticipate leaving mixes in 24 bit at -3 to -6 dbu. Give my regards to the Mermaid. :D

Opus2000 Mon, 09/06/2004 - 08:59

Good discussion. Something in which a lot of people have misconceptions about, Bit Depths.

Back in the beginning days of digital, 16 bit was the best that could be delivered...at that point you had to truly record as hot as possible to truly take advantage of the digital resolution. The bit rate in digital has a HUGE affect on your recording. The lower the resolution the worse it is...

24 bit converters are the choice in todays digital realm. With 24 bit converters you do not need to record as hot as you used to with 16 bit converters. The resolution of the signal is as good as it's going to get.

Now, there's 32bit float processing...this is a computer specific resolution only. Here's why...when sending 24 bit into a computer you are in theory only getting 20 bits! So the software can now take that and do some heavy duty mind numbing mathematical equations to bring it to true 24 bit!

Now, in regards to files recorded at one bit and processed to another bit... noise reduction/truncation and dithering...all forms of bit depth resolution/changes.

You definitely ALWAYS should record at 24 bit...plain and simple. Then at the end on the final mix use a dithering application to bring it back down to 16 bit.

By defaul all converters these days are 24 bit..so setting a 16 bit session creates automatic dithering within the application..usually a no no.

Understand that 16 bit recording from a 24 bit converter takes the last 8 bits and puts noise in it's place...the more noise you add the more grainy your recording will start to sound when you add it all together(Summing!)

Now take all the 16 bit tracks and change them to 24 bit...well, you have now taken the noise already there and simply shuffled it around with some newly added noise!

Now, the noise induced in dithering or truncating is not audible to a human being(thus why I put in my quote I mix for humans not dogs...also has to do more with sample rates than anything!) so don't think that you can just sit there and listen to it..

Here's a good example..if anyone has UV22 within their application do this...
Listen to the output of the application at 24 bit...crank your speakers...don't play anything, just listen...now add UV22 to your master output insert...listen again...you'll hear noise....this is dither...compare this dither to other dithering applications...Waves.. and whoever else makes a dither app...

Opus

x

User login