Premium Membership - All Voxengo Plugins at a Fixed One-Time Fee - Click Here For More Info
Forums     Plugins     r8brain PRO upsampling and noise shaping

Hi Aleksey,

Interesting plugins..  I just bought the analogflux suite... great stuff!

I am having a bit of trouble understanding how exactly to use r8brain pro...

I read the manual and saw this:

"r8brain PRO's bit-depth conversion is limited to flat dithering.  We have decided not to implement noise-shaping dithering because pro audio production software available on the market usually offers the user noise-shaping dithering of some kind already. "

Let's say I wanted to convert from 96-32float to 44.1-16bit.

If my 96khz 32bit float file is in cubase, should I first convert it to a 96khz 16bit file using the apogee hr dither plugin and THEN use rbrain pro to downsample the sample rate to 44.1 externally?

I thought noise shaping was supposed to smooth out quantization distortion during bit rate conversion... would the apogee hr dither be preferrable sound quality-wise for bit conversion (than the "flat dither") but then use the r8brain for the sample rate processing part?



Noise-shaping is usually used to increase resolution in frequency-dependent manner.  By saying 'no noise-shaping' I mean that r8brain PRO does not offer this advanced dithering, but nevertheless it can be used for bit-depth conversions as far as you do not need noise-shaping.

In your situation I would suggest you to export 96-32 bit file and convert it to 44.1-32 file using r8brain PRO.  Then use Apogee HR dithering to convert 44.1-32 to 44.1-16 file.  You may alternatively convert 96-32 file directly to 44.1-16 using r8brain PRO to see the difference.

Ok, I will try that and compare the difference..



Hey Aleksey,

I got around to doing a couple tests...

All I can say is WOW !!!!!!

Honest it actually sounds like the 96khz-32bit file after converted to 44.1 at 16 bit !!! ... super clear sounding !!! .. the cubase 44.1 mixdown sounded dull in comparison...

What a program !!!

I admit it WAS a hassle doing a 96-32 mixdown in cubase, then downsampling with rbrain to 44.1-32bit, then importing and doing another mixdown with dithering to 44.1-16 ... but man was it worth it !!

Thanks Aleksey !!

... i still can't believe what i'm hearing !!!

I'm glad to hear this worked so well!


I currently own Elephant, which I see has the "Noise Shaping" option in it.  If I'm willing to take the time to get the best possible bit/sample-rate conversion - should I utilize Elephant's dithering, or opt for the Apogee HR dithering (included in Cubase SX 2)?

Second question is - in a nuthshell - can you give me some idea of the kind of sonic improvements one might hear when using r8brain as opposed to just using a program like Cubase SX2?  I read Disper (above) referring to clarity... anything else I might look for in particular?

Currently, my process is working at 44.1 in Cubase SX2, then mixing down to 24-bit, 96kHz, at which point I do post-mixing (or mastering) within Cubase, the chain ending with Elephant (using its dithering and noise-shaping).

I've still got any number of serious challenges in my mixes... and none of them may have anything to do with this subject!  However, I just thought I'd ask in case you might give me something in particular to look for.

I can't give any objective comparison between Elephant's and Apogee HR dithering methods.  This is where you are probably have to decide for yourself.

If you take Cubase SX2's sample rate converter then of course r8brain PRO should give you a considerably better quality.  This clarity is about clean high frequencies, no-noise overall conversion and identical to your original stereo field.  It is also possible to run 44.1 -> 96 -> 44.1 conversion without worrying about any drawbacks - there will be none.  I just suggest to perform final peak-limiting and dithering on the very last stage, after final 96 to 44.1 conversion.  In that scenario you may use Elephant or Apogee HR dithering algorithm - whichever works best.

Noise shaped dither is absolutely the last thing you should do to a file - period.

This is because it is possible - although unlikely - for additional processing after dithering with noise shaping to introduce high end artifacts like a kind of ringing.

Aleksey - what type of dither does R8Brain Pro apply - triangular or square?

Multichannel Audio Specialists

DVD-A, DVD-V Authoring

Mixing & Mastering to most formats

Neil, it is closer to gaussian distribution, but is not exact at that (otherwise it could be a lot slower to compute).
This topic was last updated 180 days ago, and thus it was archived.  Replying is disabled for this topic.