Blind A/B Shootout: Upsampling Your Mix

[Funkybot's Evil Twin]

I can't discern a favourable difference between either of them. If I had to guess you've done no drastic EQing in the mix and there are probably no saturation/distortion plugins there at all (which benefit the most), and very little compression - which makes the whole ordeal just too subtle.

See above, I actually did some seriously drastic EQ'ing with one of your favorites (I even posted the settings), a fair amount of compression (overheads, entire drum submix, and vocals, then full mix), and used a straight saturation plug-in, not to mention EQ's and comps that should add their own coloration/saturation.

I mean, at this point, the difference I hear (listen to the vocal and hihat) is either a placebo effect due to the result of knowing the correct responses, or this technique just isn't worth the time it takes.

I think this test works better than the methods mentioned in the previous threads simply because whatever changes occur after the upsampling, later get applied to the original 44.1k versions as well. In the previous threads everyone mentioned making some slight tweaks 96k, but no one went back and applied those same tweaks to the originals. The huge improvements you guys may have been hearing may have been the result of just adding some final polish to the mix due to working with the stems rather than the samplerate.

BTW, I'd of course love someone else to post their own version of this test with different material. I'm all ears.

[Kingston]

I'm using Audiomove 1.14, perhaps it's a bug in an older version?

you're using the right version. there's never been a bug.

Are you doing 88.2 to 44.1 with it, or just 96 to 44.1? I swear, I tried it twice, and the difference was obvious. BTW I was measuring with level meters from InspectorXL (formerly from EA now by Roger Nichols), I know better than to trust the hosts metering. Again, I assure you, this is really happening time and time again.

it produces identical RMS/peak volumes going to and from any direction, any bitrate. wavelab is showing it works perfectly every time.

what host are you talking about? something other than audiomove is screwy here.

[herodotus]

exactly. brutality: not nice.

Kingston is never nice, it's not his job.

He is the senior resident audiophile, and as such brutality is to be expected.

Subtle differences in audio quality are really f**king important, and deserve to be treated with all the precision and seriousness of a war crimes tribunal.

This is audio god damn it!

AUDIO!!!

[Kingston]

I think this test works better than the methods mentioned in the previous threads simply because whatever changes occur after the upsampling, later get applied to the original 44.1k versions as well.

yes of course. that's how it's supposed to be done.

[Funkybot's Evil Twin]

Well, your effects are of the higher quality variety! I would think they would benefit from it, but you may have squashed it beyond recognition...perhaps no mastering next time?

What? I was going for loud .

I tell you what, I'm still off work tomorrow, how about I upload an unmastered version tomorrow?

[Funkybot's Evil Twin]

what host are you talking about? something other than audiomove is screwy here.

I only checked on Sonar 4. I never even considered that the host could be adjusting the volume somehow.

[Edit]

Nope, just checked in CD Architect, the latter wave file is definately quieter. I agree that this is entirely bizarre, the upsampling went fine. Anyway, I don't think that R8brain destroyed the audio, so I'm just going to consider that a non-issue for the time being. If I discover anything about why Audiomove dropped that file's volume, I'll post it somewhere. If you come across anything, shoot me a PM and let me know.

[Kingston]

or might even be inspectorXL doing something funky (but that's doubtful for a measurement plug).

thanks for letting in on *all* the details. Good mythbusting going on here.

[bduffy]

Well, your effects are of the higher quality variety! I would think they would benefit from it, but you may have squashed it beyond recognition...perhaps no mastering next time?

What? I was going for loud .

I tell you what, I'm still off work tomorrow, how about I upload an unmastered version tomorrow?

Eh? Sorry, sonny: can't hear you! I have an L2, L3 and a MPL-1 on my master bus...

Yeah, if you got time, that'd be more useful, I think, to bounce without the master crush.

I totally agree, btw, that what I've been doing so far could just be the benefit of working with the stems...I just can't resist mixing some more, and I kind of have to, once my levels change. And to be organized enough to get the same 44k stems back in...phew...I will try it though. It'll probably sound exactly the f**king same!

[Funkybot's Evil Twin]

Kingston, see my edit point above.

[Kingston]

I don't think that R8brain destroyed the audio, so I'm just going to consider that a non-issue for the time being.

...which it is. just odd considering audiomove works flawlessly here. and 1db is rather drastic.

[Funkybot's Evil Twin]

BTW, Kingston. Just tried it a THIRD time. EXACT same result! This 16bit, 88.2k file keeps getting reduced in volume (it's audible, visible, and the meters see it) by 1.1 dbs. In Audiomove I'm using these output settings: Format: .wav Rate: 44.1kHz Width: 16bit Fixed Quality: best, Threads: 1.

I'm scratching my head...

[Kingston]

yeah well me too. same settings used here (except 4 threads so it can take all the cores it can when it needs to for fast total project conversions). wavelab is showing me equality, within the limits you can expect from such a drastic conversion. in this case it's 0.05dB RMS in both channels.



why are you working on 16bit files by the way?

[bduffy]

I was just going to say I tried this to: I got the same thing, about -0.5dB average difference.

[bduffy]

What are the "threads" for, anyway? Are those encoding passes?

[Funkybot's Evil Twin]

why are you working on 16bit files by the way?

I'm using 16bit, because I'm assuming the samplerate conversion should come last, and Sonar's Pow-R dithering is actually supposed to be good. I guessed it'd be best to let Sonar convert from 24 to 16 bit, then use something else (R8Brain) to convert the samplerate. The reason I'm going to 16 is because it's still the industry standard as far as I can tell.

Hmm..Bduffy's getting it too...ok, glad it's not just me now. BTW, guess who's currently exporting unmastered mixes ?