MusicTech review: Live 6 lacks audio quality!!

[TheAnimal]

The results:

- The warp modes Beats, Texture, and Repitch resulted in 100% cancellation, comfirmed by my meters.

- The warp modes Tones and Complex did not cancel, with Tones being the worse of the two (the most audio information was audible where total silence should occur).

Great idea to test this! I'll make repitch my default instead of complex.

[nolus]

the complex mode sounds dreadful on solo piano, even at the original tempo it gives it a kind of underwater sound, but then tones and texture deal with this much better.

you have to choose the warp mode carefully when shifting tempo. i find that complex and texture complement each other, where one falls down the other excells. texture really comes into it's own when reducing tempo.

[suburbanbather]

There's no secret sauce in mixing that would make one DAW sound crisper than another. A magazine called "MusicTech" should be capable of tracking down why it sounds different or at least finding a better description than "crispness and clarity" - that sounds suspiciously like VooDoo Hifi magazines telling me I should put my CDs in the freezer and buy gold-plated power cables.

Don't forget your audio enhancing knobs
http://www.museumofhoaxes.com/hoax/webl ... ents/4309/


Also on the subject of mono sounding better(for dance/club music) I agree with this statement when it comes to having good and punchy BASSSSSSS! Listen to the Mr. Oizo album "Analog Worms Attack". It is 100% mono and the album BANGS hard on even the tiniest or just plain shit quality speakers. Where as with other bass heavy albums you either have to turn the bass down or settle for unwanted distortion.

[melocoton]

An easy way to test this is to put the same clip on two tracks, set one to warp and one not to warp, set the master tempo to match the clip tempo, then invert one track with the Utility plugin. This way you can switch between the different warp modes on one track and hear the results instantly.

Doing that I am able to get Beats, Tones and Textures to cancel completely while Repitch and Complex do not. However there was some wierd behavior with tones that gave me a clue as to what might be going on. I am using a one-bar loop and if I change some of the parameters on Tones it will drastically un-cancel for a bit and then completely cancel when the loop starts again. It is obviously going out of sync for a moment when you change the parameters and then locking itself again. Something sort of similar is happening with Complex and Repitch. Those two sound almost completely cancelled at the beginning of the bar and then they sound like they are drifting out of time in a phasey way. I was able to get them to cancel slightly better by messing with the track delay in tiny increments which seems to verify that there is some sort of latency issue causing the difference. This would make sense with Complex since it's the most processor heavy but I'm not sure what's going on with repitch.

I thought this had been resolved and disproven already in the past but maybe that was before Complex mode was introduced. Plus who knows what changes go on under the hood with each new version of Live. I guess after all it is worthwhile to dredge up these old topics again and again.

[tylenol]

I don't have most of your tools, but it is possible to run a simplified version of this test in live using the new meters in session view + Utility.

LOL, I knew this was going to come up. Forgot to add in my original post that I repeated this test in Live with utility with the exact same results. sorry about that, I had a feeling someone was going to call me ont he SX bit, so I double-checked in Live too.

Tones comparisons seem to be audio dependant. The original file I used for the comparison was an arp synth bit I had recorded from my TI. With that, the Tones mode DID cancel. That was only an audio file that had mainly mid-range sounds in it through, which is why I redid the test with a full range mixdown audio file.

The file I originally used was in fact mostly midrange, so I tried again with longer mixdowns. I haven't managed to reproduce the behavior you seemed to describe (constant non-cancellation), but I have managed to get some wierd behavior in tones mode -- for a 5min song, about the first 4:30 cancel perfectly, and then all of a sudden it stops cancelling. This is completely reproducable, but when lining up the two versions in arrangement and scrubbing, I get sort of random results depending on where I start playing and how long I let it play. As far as I can tell there are no particular characteristics of the audio at that point that would lead to this behavior. I actually wonder if this is a bug (a rounding error building up?) or something.

[melocoton]

but I have managed to get some wierd behavior in tones mode -- for a 5min song, about the first 4:30 cancel perfectly, and then all of a sudden it stops cancelling. This is completely reproducable, but when lining up the two versions in arrangement and scrubbing, I get sort of random results depending on where I start playing and how long I let it play. As far as I can tell there are no particular characteristics of the audio at that point that would lead to this behavior.

Yep, that's why I think it must be a latency thing. It's just drifting off by a couple of samples or something. Otherwise it wouldn't cancel completely at certain points and then suddenly not cancel when the source audio hasn't really changed.

[frankie123]

I posted this in another thread but figured it was appropriate. I compared outputted tracks in pro tools and ableton, they sound the same. Maybe under some EXTREME recording circumstances like recording an orchestra in a huge room with tons of dynamics you would hear a difference.

When recording stuff, it is your interface that does all the work. It spits out the ones and zeroes that makes up the resulting file.

Whats your point. I'm talking about comparing a rendered track in ableton and bouncing a track in pro tools of the same recorded guitar. The commentary made by this magazine about the final rendered product seemed to be in reference to the actual software quality, not the hardware.

[Tarekith]

If you'll read further back in this thread, you'll see I made the exact same comparison with SX rendered files. Result was identical in Live and SX.

Tylenol- this is one reason I used SX for the comparison initially, I know it does not have any weird latency shifts or other unknowns when playing back audio. I just don;'t know how Live is, which is one reason I'm so interested in doing all these tests.

Regardless, the case is still made in your test too, that Tones mode is not what you want to use when transparency is needed and you want to retain Clip functionality. Whether it happens at 4:30 or right from the start, still proven that it's affecting the audio somehow.

[headquest]

I'm talking about comparing a rendered track in ableton and bouncing a track in pro tools of the same recorded guitar. The commentary made by this magazine about the final rendered product seemed to be in reference to the actual software quality, not the hardware.

From reading the review, what I suggest they are comparing is in fact the quality of the mix engine, i.e. not the individual rendered tracks, but the mix summing of them. They say that in Live this lacks the "crispness" and "clarity" of the old-school software that the magazine continuously pushes (specifically Logic and Pro Tools).

In terms of the tests people are doing on Live's warp modes, interesting though they are, they are not really relevant to the specific claims in the MTM review. There is no recognition in the MTM review that warp modes play a part: and the implication is that Live's audio engine (regardless of warping) is not as good as their prefered software.

In terms of Warp Modes however, they compare them to PT's Beat Detective. Given that BT is not a realtime, gapless, live process, it is once again an idiotic comparisson for them to make, and it simply betrays again that their reviewer was prejudiced.

[boraxx]

in any case, quality is overrated and many times just acts as a cosmetic to hide lack of ideas. all my best songs were made on fucking rebirth.

[headquest]

in any case, quality is overrated and many times just acts as a cosmetic to hide lack of ideas. all my best songs were made on fucking rebirth.

Some of Massive Attack's best stuff was done with ReBirth too.

[leisuremuffin]

in any case, quality is overrated and many times just acts as a cosmetic to hide lack of ideas. all my best songs were made on fucking rebirth.

this is a valid opinion, but remember, the people who *really* get into this stuff are usually more involved in the technical side rather than the writing side.


of course some of the best music in the world has horrible sound quality. I'd rather listen to a recording of Robert Johnson than the newest FM rock shit, but there are people out there who are more concerned with quality than content.



.lm.

[headquest]

The point in the MTM review that actually made me angry is that they quote Ableton as replying to their question saying that they are not aware of any "serious" problem with Live's audio quality...

I can just imagine the conversation -

MTM Journalist: Are you aware that there is a serious problem with Live's audio quality?

Ableton person, trying to give a helpful and precise answer to foreign journalist, not in their native tongue: No, we are not aware that there is a serious problem with Live's audio quality.

MTM Journalist: inwardly chuckles, knowing that the German guy failed to pick up the cynical undertone in the question.

The worst kind of gutter journalism, taking advantage of somebody speaking a foreign language in order to portray them badly.

[Tarekith]

Yes, I'm completely satisfied with Live from an artist standpoint, it's awesome. I want to use it all the time.

However, there's times I need accuracy in a DAW when working ion stuff for other people, and I want to know what, if any, limitation Live MIGHT be imposing in this regard.

If someone asks me to mix a string quartet for them (this happened recently), it'd be nice to know which is the best tool for the job. There's a reason it's called "audio engineering".

[Tarekith]

but there are people out there who are more concerned with quality than content..lm.

Or those who strive to do both