What do commercial production studios do on stereo imaging?

[Spyro]

Well I found that using those Multiband comps that has a "stereo" option you can make mixes a lot wider.

Thats why I really liked MDA multiband and the Buzzroom GranComp

I prefer the opposite, isnt that sounds bad, is just that sounds too mainstream to me (not on all cases anyways)

[popsych]

Obvious question, are u bouncing down all your vst midi tracks to audio. I used to mix vsts with bouncing down all my tracks to audio. Mixes sounded muddy with no stereo separation.

Now I render everything to audio - wow what a difference.

U r talking Bol***ks. With sequencers today bouncing to wav would be at best the same as MIDI through vst or worse. Reason ? MIDI through vst comes out as 32 bit audio and goes through everything as that until bounced. When you bounce to Wav, not all apps support 32 bit float export, it would uneconomical for disk space reasons and i'm certain it would still be bit for bit identical to what you would be getting out of the MIDI through vst.

On the topic now, i find that M/S techniques indeed work better than anything else, not in the sense of loudening the S channel, but as someone else has mentioned processing them individually. I find that compression with different settings for each channel, and an EQ boost at 11-20 khz and/or slight tube style distortion (a la PSP VW, Voxengo lampthruster, even antares) on the Side signal can really make for a wider image.

If you create your mix checking in mono every now and then, chances are you can create a decent mono mix, and an outstanding stereo one. With M/S it's like being able to work on two versions of a mix. One mono and one stereo and i suggest you really look into it

Finally i think brianbrian is right if you heard it over the radio. God knows what processes poor audio goes through in radio stations .

[brianbrian]

God knows what processes poor audio goes through in radio stations .

Seriously! I heard an all new low recently...I recognized that something sounded very different on the radio (compared to the album version), and then I realized the next song had that same bizzare quality, and it had nothing to do with stereo width, eq, or compression. After listening more carefully, I realized to my horror, that they had added a reverb to the entire signal!

[popsych]

God knows what processes poor audio goes through in radio stations .

Seriously! I heard an all new low recently...I recognized that something sounded very different on the radio (compared to the album version), and then I realized the next song had that same bizzare quality, and it had nothing to do with stereo width, eq, or compression. After listening more carefully, I realized to my horror, that they had added a reverb to the entire signal!

Oh the horror.

But IMHO i attribute the "abolute necessity" that the so called mastering reverb has become to Ozone. Not blaming the guys at izotope because they gave us another (bery good may i add) tool, but people just saw it there and thought : " It has to make my sound better for it to be on a (say in awe) 'PROFESSIONAL' mastering tool. I MUST use it " .

Damn you guys. Just use what sounds nice for your track(s), not : anything available/what you heard on the radio or tv/what this mastering engineer uses.

I'm wandering off-topic here but this is really irritating these days

[LoopFilth]

Obvious question, are u bouncing down all your vst midi tracks to audio. I used to mix vsts with bouncing down all my tracks to audio. Mixes sounded muddy with no stereo separation.

Now I render everything to audio - wow what a difference.

Why should it make a difference?
Bouncing tests I've been doing showed that the bounced file was identical to MIDI data playing through the VSTi in question.

Dunno why, maybe the load on the pc makes a difference. I dont know what host/conditions you did you bounce test on. Load up your host to 80% and then do your bounce test and compare. All I can say is I'll never mix raw vsti output to get final mixes of the tunes again.

[flex42]

[LoopFilth]

Obvious question, are u bouncing down all your vst midi tracks to audio. I used to mix vsts with bouncing down all my tracks to audio. Mixes sounded muddy with no stereo separation.

Now I render everything to audio - wow what a difference.

U r talking Bol***ks. With sequencers today bouncing to wav would be at best the same as MIDI through vst or worse. Reason ? MIDI through vst comes out as 32 bit audio and goes through everything as that until bounced. When you bounce to Wav, not all apps support 32 bit float export, it would uneconomical for disk space reasons and i'm certain it would still be bit for bit identical to what you would be getting out of the MIDI through vst.

I'll do a real world comparison and mix my lastest tune both ways for you and you can have a listen if you like. Seems to make the most difference when using hehe rewire.
A/B and you tell me

[bmanic]

There should be absolutely no difference, unless you monitor at 44.1kHz sample rate and bounce at higher rates which effectively double samples your effects/vsti which results in less aliasing which in case usually clears up the mix a bit.. but man.. I dunno, that just sounds plain voodoo/plasebo.

- bManic

[daverich]

I echo most of the sentiment here. Stereo widening tools are certainly not used here - you get a wider/more defined image by using decent converters/clock. Widening tools are a bit of an intrigueing subject. Personally I think they're pretty useless for mastering/mixing as they effect the sound in such a way which always muddies the water and makes more problems than it solves.

For live sound however wideners are a very handy tool.

Kind regards

Dave Rich

[LoopFilth]

Obvious question, are u bouncing down all your vst midi tracks to audio. I used to mix vsts with bouncing down all my tracks to audio. Mixes sounded muddy with no stereo separation.

Now I render everything to audio - wow what a difference.

U r talking Bol***ks. With sequencers today bouncing to wav would be at best the same as MIDI through vst or worse. Reason ? MIDI through vst comes out as 32 bit audio and goes through everything as that until bounced. When you bounce to Wav, not all apps support 32 bit float export, it would uneconomical for disk space reasons and i'm certain it would still be bit for bit identical to what you would be getting out of the MIDI through vst.

Ur probably right with the theory - I don't doubt it, but I cant get a decent overall mix with raw vsti outputs. It feels more difficult nailing the levels/eqs etc etc and controlling the sound for some reason.

It's probably me but like 'people' say, use what works best for you. If you're interested I've uploaded two mixes of the same song - tried to get the mixes as close as possible. Thru my ears one sounds better than the other. The stereo imaging on the drum loops sounds better to me individually bounced to audio than all trax coming thru rewire (ableton).
These trax sound very similar but I've had instances of completely muddy mixes, little stereo separation and vsts sounding thin? until I bounce them down.

Anyway, carry on making gr8 music...

http://users.on.net/dragank/Music/AllahFunk_A.mp3
http://users.on.net/dragank/Music/AllahFunk_B.mp3
[/url]

[popsych]

Obvious question, are u bouncing down all your vst midi tracks to audio. I used to mix vsts with bouncing down all my tracks to audio. Mixes sounded muddy with no stereo separation.

Now I render everything to audio - wow what a difference.

U r talking Bol***ks. With sequencers today bouncing to wav would be at best the same as MIDI through vst or worse. Reason ? MIDI through vst comes out as 32 bit audio and goes through everything as that until bounced. When you bounce to Wav, not all apps support 32 bit float export, it would uneconomical for disk space reasons and i'm certain it would still be bit for bit identical to what you would be getting out of the MIDI through vst.

Ur probably right with the theory - I don't doubt it, but I cant get a decent overall mix with raw vsti outputs. It feels more difficult nailing the levels/eqs etc etc and controlling the sound for some reason.

It's probably me but like 'people' say, use what works best for you. If you're interested I've uploaded two mixes of the same song - tried to get the mixes as close as possible. Thru my ears one sounds better than the other. The stereo imaging on the drum loops sounds better to me individually bounced to audio than all trax coming thru rewire (ableton).
These trax sound very similar but I've had instances of completely muddy mixes, little stereo separation and vsts sounding thin? until I bounce them down.

Anyway, carry on making gr8 music...

http://users.on.net/dragank/Music/AllahFunk_A.mp3
http://users.on.net/dragank/Music/AllahFunk_B.mp3
[/url]

Mix both with the same settings and i you are using vsts (and export 32 bit float), they should null. I am not familiar with the intricasies of rewire though.

[Debutante]

I seriously doubt it's some kind of post production thing. That would really damage the sound quality. For people who are hopelessly anal about sound quality like "high-end" studio engineers, I'm sure they would do things during the mix phase to get a nice stereo image.

Here's some ways to widen your mix

-Phase Delays: Simply a left or right channel delay of about 10-20ms, very short. This makes things wide without sounding "chorusy" or having an obvious echo. It's wider, but can still sound "dry." It's used all the time on guitars and pads. Nice if you don't want something to be on "one side."

-Stereo Enhancers: Almost all of these work by feeding the left channel into the right and inverting the phase, then the right to the left and doing the same thing. This makes the differences in each channel louder, and the similarities quieter. Tends to sound unnatural if pushed really hard.

-Double tracking: Record the same thing twice, compress, edit timing, pan opposite. It creates that intimate kind of acoustic guitar sound. Like it's really close to you but not "mono." Everyone's heard it at one time or another. You can do a similar thing with a delay.

-Slapback: Pan something one way, have a pretty quick and more subtle delay in the other channel. Do the same with a counter element in the opposite channel. This one's nice if you want things to sound "big and hall-y." It really widens the mix. I like it personally. Not that great if you wanna do the modern ultra dry thing.

-Something else: Every studio engineer has their own tricks. One of the keys to becoming a good one is to find your own. Learning from others is obviously a big part too. Records in 10 years won't sound like they do today, so just trying to copy someone else's style will only be so rewarding.

hth!

Is the volume of the combined tracks supposed to fall so dramatically?...with M/S

[greendoor]

Something i've found with VSTs and hosts - trust nothing. In "theory" they should behave logically, but often in practice they don't. If something appears to sound better for no logical reason, it probably does. It can be years before a fault is acknowledged and fixed. I notice some synths sound much better at 96kHz, therefore it makes good sense to export audio at 96kHz, and then reimport to your 44.1kHz project anyway. That's logical enough. For backup purposes, you are much safer if you have rendered your VSTis to audio. It wouldn't surprise me in the least if VSTi's that are rendered seperately might sound better than those that are rendered together with the mix. I don't think there is a universal answer to this, other than check with your specific combination of software and don't trust what should happen "logically".

[popsych]

Something i've found with VSTs and hosts - trust nothing. In "theory" they should behave logically, but often in practice they don't. If something appears to sound better for no logical reason, it probably does. It can be years before a fault is acknowledged and fixed. I notice some synths sound much better at 96kHz, therefore it makes good sense to export audio at 96kHz, and then reimport to your 44.1kHz project anyway. That's logical enough. For backup purposes, you are much safer if you have rendered your VSTis to audio. It wouldn't surprise me in the least if VSTi's that are rendered seperately might sound better than those that are rendered together with the mix. I don't think there is a universal answer to this, other than check with your specific combination of software and don't trust what should happen "logically".

I on the other hand know there are no magicall reasons in software, and i can tell you that what loopfilth said makes no sense, viewed as a bug or otherwise. The 96khz to 44.1 khz sounding better is a matter of algorithm working better. Mainly a) Pushing anti alias in inaudible frequencies and b) more accurate filters.

[LoopFilth]

Something i've found with VSTs and hosts - trust nothing. In "theory" they should behave logically, but often in practice they don't. If something appears to sound better for no logical reason, it probably does. It can be years before a fault is acknowledged and fixed. I notice some synths sound much better at 96kHz, therefore it makes good sense to export audio at 96kHz, and then reimport to your 44.1kHz project anyway. That's logical enough. For backup purposes, you are much safer if you have rendered your VSTis to audio. It wouldn't surprise me in the least if VSTi's that are rendered seperately might sound better than those that are rendered together with the mix. I don't think there is a universal answer to this, other than check with your specific combination of software and don't trust what should happen "logically".

I on the other hand know there are no magicall reasons in software, and i can tell you that what loopfilth said makes no sense, viewed as a bug or otherwise. The 96khz to 44.1 khz sounding better is a matter of algorithm working better. Mainly a) Pushing anti alias in inaudible frequencies and b) more accurate filters.

You're right computers don't make mistakes... Its the programmers that supply the 'magic'