Most of you know about the announced Ableton Live 7.0 by now, and you could visit their site to get all the details. I’m just now beta testing it, and I will post now and then findings about my tests.
One feature they announced, actually on the top of their announcement page, is an enhanced audio engine. According to the blurb they have precision 64-bit summing at all mix points through their program, POW-r dithering, optimized sample rate conversions and other advances.
Live has had a bad reputation about their mixing engine, it was not that bad, but all kinds of things out-of-control made the final sound not as clear and airy as with for example Logic Pro. Some things were not that obvious, for example, using warped audio clips caused dullness and less clear transients. Another issue that many have reported, and I noticed myself, was that Live was less forgiving about running hot tracks. Usually by keeping things below -3dB fixed the quality issues (bulky sound). And having no dithering support was not good, but I’ve used dithering tools separate from Live, such as Ozone, to fix that issue.
Anyway, what I did was to render out the same track with both Ableton Live 6.0.10 and 7.0b1. This is a new track from Genitronix called Funky Music, Up and Down (Level 5 remix), released very shortly, and it’s been a notorious production on my hard disk for a long time. The reason is that it has many dense tracks, and until a couple of weeks ago I could not tame it, until I did massive eq:ing across all the tracks. So I think this is a good example of a typical massive dance track, could it sound more airy with Live 7 compared with Live 6? Another possible test would have been a track with a lot of reverb, that usually also shakes out issues about clarity.
Anyway, I took the two exported 16-bit WAV files, and opened them up in QuickTime. This is at trick of mine to do A/B testing, I start and adjust the playback in the QuickTime player, and by switching the focus between the two tracks I could easily do A/B testing.
The verdict? Well, I had a hard time to hear the differences. I think with a full blind test I could have not heard any differences. It felt like Live 7.0’s rendered output was somewhat more clear, but it was so marginal. This all requires more testing, but I didn’t hear a big, huge difference. So anyone who has published tracks with 6.0, you didn’t miss out on anything big.
Some other thoughts, even if the mixing engine in Ableton Live is now fully 64-bit, there are all kinds of external plugins (might still be 32-bit or worse) and original sounds that will not benefit from this. The weakest links is still the weakest link. So this all is always marketing talk, so you need to be a little bit skeptical about all the talk about 64-bit sound engines. I might write something about the whole idea of sound quality, bits, floating point, and similar things shortly, as it’s a fascinating world, and musicians should know about it so they don’t fall for snake oil.
I will still release this track and some other Live-based tracks using Live 7.0, as the beta is stable, and maybe there are some small transients here and there that sound better, not to speak of improvements in the plugins, especially the built-in compressor and EQ8. Oh, the purist inside me.
Entries (RSS)
About a year ago I did my own personal test where I summed a huge amount of audio files in Ableton and then in Cubase and no one I played the files to could tell the difference. I’m not saying there isn’t something going on with Live 6 and crunching audio because I do hear that sometimes…. its just not at the basic render across summing level. Yes again keeping levels low really improves things.
I am very excited about Live 7. They really added tons of new stuff.
Have you tried to Sidechain using the Autofilter yet? That sounds very interesting. Please post here what you can without getting in trouble with the NDA.
Yes, I would like to test the side chaining next. I don’t think there’s an explicit NDA, any new features are published by Ableton Live, at their site, so it’s more of an exploration phase now for anyone testing out the betas. Try to jump in on the next beta seeding — it usually requires polling their beta web site to find out when the next round is opened up.
Interesting, I find it weird howerver to mention the “bad reputation” thing in the article, as I think its forum crap one should not give the exposure you gave here.
Well, I’ve been reading Ableton forum related postings for the last 4+ years or so, and this comes up from time to time. And I’m usually the first to justify that the audio mixer in Ableton is not that bad, you just need to avoid using excessive warp modes for most tracks as those eat up the transients, and be careful with hot signals.
Here’s actually a good link talking about audio and 64-bit mode:
http://mixonline.com/mag/audio_im_sixty_four/
I love ableton’s flow, I can layout the backbone of a track in minutes and quickly get to the interesting stuff. But the overall sound had often left me a little sad. The lack of quality in ableton’s mixing engine has driven me to become completely rewired. Routing all tracks into Cubase or Logic to maximize audibility, even while in early production stages.
Pillform,
I was having the same problem and didn’t think that I’d have to route tracks via Cubase so early in the production stages. Have you identified a work around for this yet?
If Ableton’s time stretching algorithms were as good as Melodyne’s… Meanwhile, the best you could do is to import content and use the RePitch algorithm but then you also need to keep track of the original BPM and use that as the baseline BPM when working.