Discussion in 'Post Production & Mixing Discussion' started by Puzzlefactory, Sep 28, 2017.
I've tried it a few times but realized that I prefer being able to tweak specific details myself.
I get the feeling Ozone 8 is going to put a serious dent in landr.
I agree. Over the past 20 years I've become a bit of a fetishist for mastering plugins, ever since my old hardware Finalizer and the MasterX plugin series. I do extensive shootouts between anything that even resembles or claims to be a mastering processor - including possibly dubious contenders like BeatSkillz's "Slam Dawg" (which is actually pretty cool on certain material).
I just delivered the scores for Jigsaw and a Saw Anthology release to mastering prior to release on download/streaming as well as CD and vinyl. When I deliver to outside mastering houses, I always provide my completely un-compressed, un-processed "dry" mixes as well as my "best guess" home-mastered versions. That way the mastering engineer can see what I thought the mixes "should" sound like, and start over again with the dry files if he thinks I'm stepping on things too hard, etc. In the past my "best guess" was often a combination of MasterX5 and other processors, but lately I've been diving pretty deep into Ozone v7 - and it is fan-freaking-tastic.
For the Jigsaw score there's a bunch of cues that are just slamming industrial crunch, more so than in previous films, as well as cues with pretty heavy drums, sine-wave sub-bass, and bit-crushed guitars PLUS chugging string ostinato patterns AND grinding low brass braams AND epic french horn melodies, all at once - and it all needs to be legible. I used the Harmonic Enhancer module in Ozone for the first time, on the "tape" algorithms with amounts set to about one-third and the mix set to 50% - followed by the Ozone Maximizer with automation controlling its threshold to "chase" the levels of the incoming mix as it changes.
The mastering engineer told me that my Ozone versions sounded great, and only needed some small (1-2db) level adjustments between tracks and a little bit of presence added - so those are the versions that are going to manufacturing. I'm amazed at how well Ozone works to get all of the elements in a track to speak correctly - the Harmonic Enhancer added just a bit of "bite" on the chugging strings and some "sizzle" on the high percussion without anything sounding harsh, and the Maximizer can shave a few db off the peaks and you don't hear a thing, except for the guitars and bass sounding more legible in between all of the hard drums. A buddy of mine was sitting in on a mix with Alan Meyerson recently, and saw Alan using the Black Box saturation plugin a bit on stems to brighten and thicken things, so I've just added that to the pile - but I think it's broadly similar to Ozone's Harmonic Enhancer's pentode and triode tube models. With Ozone I'm not really putting the boot to the mixes, no more than 3db of peak shaving on the loudest and most intense passages, but Ozone does its magic with no artifacts or unexpected wierdness. The waveform looks pretty "bricked" in some cases but it doesn't sound like anything's getting stepped on - it's all just louder. WAY LOUDER.
I just got Eventide's new mastering plugin but haven't had a chance to give it the workout yet, but I can't wait to see what Ozone v8 will give us. Those guys are way out in front of the pack in that department.
But Landr? Nah. Not for me. Maybe if you've got an absolute dumpster full of old tracks and don't feel like nickel-and-diming them into submission, but... half an hour with Ozone, auditioning presets and fiddling with the controls, and I think you'll be out in front of what any AI can guess your tracks should sound like.
^I agree. Charlie, have you checked out any clipper plugins? I currently use Ozone 7 (minus the maximizer), then KClip 2, then the Ozone Maximizer. It's a great way to get LOUD without over using the Ozone 7 limiter.
I also got Eventide's Elevate Bundle. Haven't used it myself yet but am looking forward to testing out the clipping section (and all the other stuff) it has to offer.
Any clown can make it louder..........every streaming service can make it quieter.
Hmmm maybe upgrading to ozone 8 is a better idea...
Dude, you are the most argumentative person I've ever seen. Why go to the trouble of posting the mastered version and then not post the original? Are you just constantly trolling or something?
I feel we need to address the music style when talking levels. Sure in industrial-slam score, you're gonna be wanting to bring on the hurt.
But the more intriguing discussion is about reference levels for stuff that needs to reproduce other sources like accoustic environments etc. I've kinda found that recently I'm far more interested in re-creating 'nice to listen to' masters.
For example, I have a compilation of cello music called 'romantic cello' ( ok, go ahead and laugh ) with the most beautiful rendition of ' la sicilienne' on it. It's not particularly hot, in fact down in the -20 dbfs rms range, but the recording is just 'right'. I could listen to that all day, a masterful recording. And after all, isn't that the goal of recording music? (not that I mind me some extreme heavy shit every once in a while)
Another point, more to the orchestra. I've recently found that having no dynamic clamp, or modest comps tapping the sources allows for far more level control at the source. It seems to bring about a more emotional response out of me to be able to let one section take over, and the power of the orch makes me work harder at the orchestration. More to that point, the natural power of the sections is easier to let fall in place from an orchestration point of view.
To hybrid scores, I've gotten into the habit of level controlling at the source, so hard drums wll be heavily clamped. SSL g channel with the dynamics side-chained hpf rolled up with quick response compressor will tame alot of stuff (old school but easily overlooked today) followed by the clipper plugs. Bass stuff I also heavily reduce the range. Sidechaining stuff can create the same pushdown effect as heavily clamping down at the stems, but you can always use 'control tracks' to create the envelope of the sidechaining effect, like quick sine bursts, or anything that creates the pull-back sidechained shape you need. It gives the music some push pull and movement.
Maybe I'm just tired of hearing overcompressed stuff. Nothing worse then a democratic mix- the mix where everything needs to be heard all at once all the time. Loud means nothing without the presence of soft imho.
My litmus test these days is 'do I need to hear that again?' if yes, then I got the balance right, if not, then I overdid it. Creating depth is far harder then anything else.
Recently did a test on a theme where the first version I slammed to full. After listening to the that music against the show, I felt I decided it sounded out of place, so after some revision, I went back and delivered the level down in the -20 dbfs range-I was taking a chance here cause I felt I might get some flack due to the lack of slam and volume. No feedback from up the line but positive, and certainly no concerns about subjective things regarding the sound. This from the iphone/tablet toting above and beyond crowd that have no tech inclination.
Obviously context is everything.
LANDR is at least worth checking out if one doesn't know how to master, wanting to release demos or for artists on a budget.
Just keep in mind if you're planning an album or collection LANDR isn't going to 'Master' them properly. You're going to get individual files based on how their algorithm interprets them. (If that's the case make sure all your source files are mixed close or relativity the same.)
I've tried it & got some decent results but never better than what I could do myself & I'm not an ME in a proper room.
The tech has gotten better since it first started but it will always have it's shortcomings compared to professional mastered music. YMMV
Here's a blind test @Puzzlefactory. It's a piece I'm working on for a short film.
Password is this beverage (all lower case) --->
Here's the key after you've listened. For the sake of fairness, I bounced the Ozone version to MP3 320 because I'm not paying LANDR for a WAV version. Let your ears decide for you
Well my ears aren't exactly well trained and I'm listening through earbuds on my iPhone.
But I would say test 2 is the better mix/master (although there's very little to choose between them).
My problem with AI based stuff like Landr, (and now Ozone 8) is you're letting a machine make choices based on an average. Mastering isn't intended to get you to average.
To quote a buddy who's a mastering engineer and Ozone 5/7 super fan, (Ozone 8 not so much, and not because of the the price hoopla... He'd pay in heartbeat if interested... I digress...) "It's a superhighway to average town".
I like Izotope, but they completely lost me with Ozone 8...
The entire point of a mastering engineer is to give you fresh ears, some intuitive aesthetic choices that a computer can't make now, if not ever, and ultimately help you achieve the artistic vision you want. Are there some things that can be automated? Sure... But no machine has the intuition to highlight an intriguing area to bring attention to an instrument or voice, or drop the dynamics down a hair just before a few key moments so the big ones have more impact.. Can you after the fact? Sure. But in my opinion letting a machine make most of your decisions from the start only serves to make you less secure about your own decisions. I personally think it colors your perspective, and considering how sucked in to technology we are it's a lot easier than you think to assume a machine is making the "right" choice...
Think about how often auto-complete ruins a text message or wrecks an email and you don't catch it until it's already sent... These essentially do the same thing; examine a portion of what you wrote, look up something it thinks is similar, and attempts to predict what the word is. Sure it guesses right pretty often, but some day your friend's going to get a text that says "My dog Spike's breeding a nasty bitch!" instead of "My dog Spike's being a naughty boy!"
I'd pay the upgrade no problem, but I don't like where they've taken Ozone. Ever since 6 they've watered it down with each new version.
I just test-drived their service, so would love to chime in over here.
My pros and cons:
-descent masterings for very low price - if you go their subscription route you can get descent quality masters for hundreds of your tracks for just 39.99 $ (this is an amazing value imo! professional masters can cost anything from 50-200 USD)
-very easy to handle dashboard; really: just upload, wait, choose if you want either a light, medium or heavy master, wait again, download!...
-...on the other hand the limited options you get wont give you a lot of free space. There are really always just only three mastered versions you can choose from. This also means that if you got a certain specific wish, like a more warm sound, or more high frequencies, etc. it can't be specifically fullfilled.
-limited number of uploads (should be 25)...
-....and just in generall the uploading/downloading thing is a bit of a hassle at this point (also as speed is just plainly slow at times)
Luckily I have worked with mixing/mastering pros before (like Townley and J. Rodd) - so I really have something to compare and really can say that LANDR is a descent service. Still you should never expect too much of it, but if you are on a tight budget I could just give a thumbs up for using their service.
...here's a jingle I just mastered with LANDR:
Let me know your thoughts.
Pro-tools-expert recently did a blind shootout with a mastering engineer, a mix engineer, and Landr. The mastering engineer won of course, but the outcome on one song showed Landr was favored over the mix and mastering engineers. But, while I do think there are a few metrics that an algorithm can determine, the fact that the musicality is being ignored in favor of averages is more than mildly frustrating. It is like the telephone game you play as a kid, whispering a phrase in someone’s ear and having that person do the same to the person next to them. By the end of the line, the original idea is largely lost. We are already well past the point of musicality being lost. But all that’s old will be new again... hopefully.
Separate names with a comma.