Render to mp3 from Cubase or post Cubase?

An sound engineer friend of mine suggested I always render out of cubase to a 32 floating wav file and afterward process this file in a converter program (he recommended Switch) to mp3. He feels the fidelity of this process is superior to rendering my mp3’s directly out of Cubase. Are there any other opinions on this?

The only thing that matters are what encoder and settings you chose when you do the encoding.
Fraunhofer’s encoder doesn’t sound as good as the LAME codec. Cubase doesn’t support LAME afaik, so that is the reason you have to do it outside. You can definitly hear the difference between LAME and Fraunhofers codecs.
You can use the LAME codec inside Wavelab or Audacity if you want to.

Aloha m,

Very interesting.

Is the resulting LAME file larger in size?


Is the resulting LAME file larger in size?

Nopes, not from what I have seen anyways.

That’s a bold statement - it’s ALL very dependent on which versions are compared. And not only that, the used mp3 bit rate is also important when comparing. Years ago, Fraunhofer was far superior to LAME in 128 kbps, but it has also been the other way around. Even then, it could still be that the other codec is better at 320 kbps or VBR settings.

Luck, Arjan

Yea, it’s a bold statement. What I should have said is: the best thing to do is encode your own material and compare yourself. As you say, I shouldn’t be that categoric since I don’t know what kind of material and settings we are talking about.
Anyways, I was an early adaptor of the mp3 format (I encoded my first one in DOS in 1994 - took several hours encoding a couple of minutes back then :laughing: ). I have been comparing different methods, bitrates and material during the years. All MP3’s at 128 kbps sounds terrible to the original, Ogg Vorbis is better if the filesize is more important but that is another topic (for streaming for example). At 192 kbps or above I find LAME to be far better sounding, it is very obvious when I’m listening to the the highs (cymbals/crashes/hihats/clean HP-filtered type of sounds).

I also prefer LAME… But TBH I use whichever comes to hand quickest.

To directly answer the OPs post, I’d export in 32bit or 24bit WAV and convert afterwards. That’s because then you’ve got a full quality WAV on your hard disk for future use. If you export directly in MP3 then you’ve got nothing but the lower quality MP3. As for the fidelity I don’t think it’d matter apart from the aforementioned MP3 encoder differences.

Of course, what we’d really like is multiple output formats all in one click. I would then export in 32bit float for high quality plus 16bit44K for CD plus MP3 for my MP3 player testing. Job done in one hit would be great :smiley:


Thank you all for the discussion. It looks like I have some work to do comparing some different mp3 renders.