I’ve read and do know that this has been wholly and/or partially discussed some times before, but…
Given the (CREATING) facts, that
I would like to create my songs only with Renoise, and only at 96kHz/24-bit (ASIO);
I would like to render the tracks of a finished song to separate 96kHz/24-bit WAV-files;
I would like to finally polish (‘master’) my songs in an other multitrack DAW and/or multitrack editor (usually 96kHz/24-bit (ASIO)) and make the final single-file mixdown from it (WAV or MP3);
… and given the (LISTENING) fact, that
I would like to listen to/share/send/store my songs in 44.1kHz/16-bit (320kbps) MP3-files…;
… which are usually played and listened to using DirectSound drivers (not only by me but most of the others who might want to listen to it);
My question is (THE QUESTION) that
…where and what I should I change or do to keep as much of my clear, sharp, nicely sounding, ‘full-bodied and -souled’ Renoise-song (that I like so much when I play it with Renoise) as I can when it becomes a 44.1kHz/16-bit (320kbps) MP3-file?
So far I did it according to one of the 2 routines below:
Create in Renoise (96kHZ/24-bit)
Render in Renoise (96kHz/24-bit WAV - Separate files for each track)
Multitrack mastering in an other DAW (96kHz/24-bit - using the separately rendered WAV-files of the tracks)
Mixdown from the DAW to
WAV (44.1kHz/16-bit) and convert to MP3 (44.1kHz/16-bit/320kbps) with an external encoder,
WAV (96kHz/24-bit), convert it to another WAV (44.1kHz/16-bit) and then convert this to MP3 (44.1kHz/16-bit/320kbps) with an external encoder
MP3 (44.1kHz/16-bit/320kbps) directly.
Whatever I do, whatever I change, the final 44.1kHz/16-bit 320kbps MP3 never sounds the same as my “perfectly” composed song in Renoise. And not just not-the-same - I’m aware of the data and thus quality loss caused by the conversion and compression - it’s absolutely different… muddy, messed-up, blunt, unenjoyable.
This is my first post and question here. Thank you for reading, and please be patient. I accept any suggestion, piece of advice or help.
What does it sounds like compared to the wavs? Make sure to use the same program and drivers for the comparison. It could be the mp3 coding artefacts you are finding objectionable and there’s not much you can do about this except try and different form of compression (and this can include using something other than LAME to create the mp3, assuming you currently use LAME, which most software I’ve come across does.)
If you want it to sound the same as when played in Renoise then render the full track (not the stems) and use Cubic interpolation (or select Realtime mode.) As soon as you import into another program to mixdown you will start getting aspects of their audio engine coming into play (although interpolation mainly occurs when changing sample rates or playing at a pitch other than normal speed so this can maybe be minimal.)
what audio rate converter are you using? they DO sound different… (yes shocking) for example r8brain and audioMove give different results, both good, but different…
also try 32bit (float) for renders instead of 24bit, 32bit is the same as renoise internal engine
You convert from 96Khz/32 bit to 44Khz/16 bit.
I would advise you to convert from 96Khz to 48Khz because that is an exact 2:1 ratio that is better to interpolate. 96Khz to 44Khz is a 2,1818181818181818181818181818181818181818…:1 ratio so how would you expect any conversion program to polissh and interpolate this properly?
(not even mentioning 24bit to 16 bit).
It is all about nice round numbers that does the magic.
Wow, it seems like I’ve found the wrong community to ask lame questions… too bad. Surely, next time I’m gonna give it a try at the local AA service or something like that.
Thanks to those who tried to help anyway, I’m gonna make some more attempts according to the suggestions.
)