Mix down: real time or export function?

this thread just BLOWS my mind!

I have to try this tomorrow as well.. i use a lot of virtual instruments and compression.
 
/\ and i thought iwas asking a proper dumb-arse question.
the problem i anticipate is that i am currently on the edge of CPU load effecting the audio eg. pops and clicks. If i export realtime i guess there is more chance of the pop/clicks actually being presenting in the mixed down audio....but i havent had a chance to try it yet
 
I just started thinking to myself, maybe this is the reason why I always feel like my tracks sound a little different outside of Cubase. I bounce in offline by default, that's just how I've always done it. But more and more lately I have noticed things sound a tiny bit different when I play back the bounce in Quicktime or iTunes - and yes I have all the sound altering shit turned off on both. They sound identical between QT and iT, but different from within Cubase. In Cubase it sounds more...I dunno, just better haha. I think I will do a good test of this but I will most likely start bouncing in real-time from now on.
 
I noticed a few years ago that while exporting in logic in real time that when I played the tune and just recorded it with and external device (into another interface through its converters) that it (seemed) to capture more of the sound of the session than rendering it internally.

I haven't really noticed a difference between offline and real time though. Now I'm curious.:)


J.
 
When I got into recording we were using 4 tracks, old reel to reels, and ADAT machines. Real time mixdown was the only option. Our CD burners were 1x as well. What sucked worse is my band would try writing these "epic" metal masterpieces that were like 8 minutes long.
 
/ and i thought iwas asking a proper dumb-arse question.

hehe, and now it will become the most epic thread in the whole net ;)


Interesting that there was a difference with the Kore Player. But we still have to prove which method sounds better.

To me it's really logical that the non-realtime must be better, because I have some plugins like Elephant, Cubase 4 Reverb and even 8505 which upsample or show their whole efficiency in mixdown.

So I'm in the non-realtime camp.

With the UAD stuff though it can be true realtime is better, because its a different cpu.
Have to find this out in an own test I guess.
 
...buti don't get the real-time option in the export dialogue. Is that because i am running SE not SX?? :-((
i'll check the prefs...
 
...To me it's really logical that the non-realtime must be better, because I have some plugins like Elephant, Cubase 4 Reverb and even 8505 which upsample or show their whole efficiency in mixdown...

...mmh... to me it's not that logical though. Upsampling should be done offline anyway, that's why you get different latency times with different plugs (among other things). ie: the UAD Pultec has upsampling and you can listen to it's result when using it in real-time.

What I have experienced is that sometimes I get drops and artifacts when using almost 100% processing power (that is computer DSP not UAD DSP). So I try to be efficient with my use of plugs so I don't hit that threshold. But it makes no sense that because the CPU is being use in it's full capacity that it somehow starts making mistakes and you end up with a lesser quality render. That is not how digital processing works. If it were that way you would be in a situation where you could get jitter if you didn't maintain the cpu in it's sweet spot just because if it suddenly hit a bottleneck then it could drop a cycle.

Realtime rendering is in my estimation requires same processing that you need when rendering offline. The only difference would be the cpu doesn't have to keep refreshing meters and animations so it has more leeway to process audio. In that regards, if you can listen to your track without it dropping or throwing artifacts at you then you are good to go.

just my 2 cents
 
What if we use a file comparer, just like the one in Wavelab, to compare a real-time to a offline exported version of the same project? Wouldn't that solve the quality related questions?
 
I just started thinking to myself, maybe this is the reason why I always feel like my tracks sound a little different outside of Cubase. I bounce in offline by default, that's just how I've always done it. But more and more lately I have noticed things sound a tiny bit different when I play back the bounce in Quicktime or iTunes - and yes I have all the sound altering shit turned off on both. They sound identical between QT and iT, but different from within Cubase. In Cubase it sounds more...I dunno, just better haha. I think I will do a good test of this but I will most likely start bouncing in real-time from now on.

I made the same observations a while back but concluded that it must be my mind playing tricks on me. I don't know. Since then others have made the same remarks. There has to be some truth to it. Perhaps the mixdown/bussing procedure isn't as transparent as we like.
 
I'm serious, I know when I listen back in QT or iT it sounds different to me. I thought it was just the pause in between bouncing and opening the file in QT since reading that recorderman (I think?) stuff, a good trick to refresh your ears during a mix session is to basically stop for about 15 seconds then roll the volume back up and keep going, high end is way more apparent to you after that - it really works, try it!

And in my case I'm not doing any dithering or conversion to 16-bit. It's 24-bit/44.1kHz no matter what.
 
where do you see a real time export function in Nuendo? I don't recall ever seeing it?

I use a Finalizer, and the audio AFTER the finalizer is brought back into the session on a new track in real time (obviously), but other than that, I don't remember ever even seeing the option for real time export...Hmmm, now I wanna try it, too!
 
It's on the export window, the one where you name the file, check if you want it to be mono or stereo, what rates, etc. Should have a little box next to it that you can check for Real-Time. I'm not at home otherwise I'd just take a screen of Cubase 4 for you.
 
I just tried it and I did not hear any difference at all between real-time and normal export. When I imported the two files and flipped the phase on one the only thing left was the drums' reverb, no guitars, no bass, just drum reverb.

006, when you import the exported file into a new project does it still sound different to the way it sounds before exporting?