Ehh.. what is happening to my bounces !

[Ed Ryan]

Right. This one is driving me crazy.

 

At first I thought it my ears playing trickery. But basically when I bounce a project it doesn't sound the same as the source logic file.

 

There are a few problems. A lack of bass, the reverbs come through louder and the mix just sounds off when it's turned into a bounce.

 

What the devil is going on ?!?

[David Nahmani]

Turn off "Normalize" in the bounce window, and what you'll hear is what you'll bounce.

[Ed Ryan]

I never use normalise !! Forgot to mention.

[David Nahmani]

So how do you determine it doesn't sound the same as the source file? Where do you playback that file? Maybe the playback device changes the sound on playback?

 

If Normalize is unchecked, the same signal sent to your speaker is placed in an audio file, so it should sound identical.

[Ed Ryan]

So how do you determine it doesn't sound the same as the source file? Where do you playback that file? Maybe the playback device changes the sound on playback?

 

If Normalize is unchecked, the same signal sent to your speaker is placed in an audio file, so it should sound identical.

 

I'll let you know what I'm working with.

 

Macbook 2.2

Logic 8

Focusrite Sapphire

Tannoy Reveal 6d's.

 

Just by A/Bing between the same points in Logic & Itunes.

 

The playback should be the same. Both apps go out through soft outs 1&2 to the spdif output that hooks up my tannoy monitors. I have both apps outputs at full scale as well, so it's nothing to do with perception and loudness contours..

 

I'm wondering if the off-line bounce is having something to do with it. The on-line bounce should write data in the same way that bouncing internally via an aux would.

[Ozzi Blake]

I've read a few reports of problems with off-line bouncing. Have a go in real-time.

[Ed Ryan]

I've read a few reports of problems with off-line bouncing. Have a go in real-time.

 

Hey Ozzi. Can you pm me the details on those reports. I'd be really interested to read them.

[marcel72]

Just by A/Bing between the same points in Logic & Itunes.

 

Do you have all Itunes' Enhancers/compressors/EQs turned off?

[jharrylong]

I was having some problems with off-line bounce using version 7. Sounded like the the center material was being dropped. No bass, lots of reverb. No problems since I moved to version 8.

[Ed Ryan]

[quote="marcel72"

 

Do you have all Itunes' Enhancers/compressors/EQs turned off?

 

Yep. Checked and they are off..

 

Realtime bouncing seems to have sorted it. It's annoying as Logic is quite slow. To read, bounce, convert and dither a wav and mp3 of a 7.40 track took it 3.35 last night. So I'm guessing it does it at just over 1/2 realtime. It should be more accurate then I'm finding it..

[marcel72]

 

Do you have all Itunes' Enhancers/compressors/EQs turned off?

 

Yep. Checked and they are off..

 

Realtime bouncing seems to have sorted it. It's annoying as Logic is quite slow. To read, bounce, convert and dither a wav and mp3 of a 7.40 track took it 3.35 last night. So I'm guessing it does it at just over 1/2 realtime. It should be more accurate then I'm finding it..

 

Wait...

 

You're dithering... I assume down to 16 bits? And converting to mp3? You know that both of these things are going to have an impact on the sound and quality of your mix file as well, right?

 

Try this: Bounce a mix to whatever the session bit depth and sample rate of your project are. Compare that to the live playback of your Logic session, if you're mixing ITB and monitoring thru the same outputs, it should be virtually indistinguishable.

 

If you're mixing OTB, the quality of this file versus live playback will be dependent on the quality of the converters you use for mix capture.

[Ed Ryan]

 

Wait...

 

You're dithering... I assume down to 16 bits? And converting to mp3? You know that both of these things are going to have an impact on the sound and quality of your mix file as well, right?

 

If you're mixing OTB, the quality of this file versus live playback will be dependent on the quality of the converters you use for mix capture.

 

I've kept the the project at 16 bit 44.1 That's the standard of the vocal part I was given.

 

I know the mp3 would sound poor, but I've always thought dithering was a good idea. Even when there is no change in source against output, as the dither increases the usable dynamic range of the output..

 

The online bounce sounds more accurate then offline. I'm thinking of a few ways to try prove it. I may bounce a few samples of parts sing either method, and then use something like a Nuetrik and SMART to analyse them.

 

All of the instruments, processing and mixing is done within Logic. Logic feeds my sapphire, and that connects to my monitors via spdif. It should be a pretty clean D/A path. Same outs and path for Itunes. So it should be pretty indistinguishable. But I can hear the differences..

[fader8]

Ed,

Adding dither noise and then compressing to mp3 are two processes that usually don't get along. The mp3 process has its own methods of dealing with low level bits.

 

There are a bunch of reasons for offline bouncing to sound different from real time, eg MIDI and tempo controlled effects, arpegiators, sampler voice allocation, etc. Offline doesn't necessarily mean it's going to be faster than realtime either. An offline bounce of a heavy piano piece in Kontakt with a lot of sustain pedal can take 2-3 times longer than real time to complete. Since it's a single process, it only uses one of your processors too.

[davidpye]

Why don't you just bounce to disk, offline, then import the resulting file, so do it at the same bit rate/format.

If you put it at the same start point as your original bounce was done, Then open a Gain Plugin on the channel it's on and flip the phase. It SHOULD cancel out, you may need to shuffle it back to account for plugin delay, but probably won't.

If it doesnt cancel then something's different. It's normal for the reverb to still be there, as it's generated live every time it's played back.

[Ed Ryan]

I thought Logic split the bounce into stages though.

 

Ie. It passes the playhead over and generates the result. It then creates a data compressed version (mp3 or AAC) and goes back to dither the aiff/wav before writing the file. That appears to be the order of events in LP8.

 

This is proving be to a valuable eye opener. If you guys could share any more info I'd appreciate it 8)

[fader8]

Ie. It passes the playhead over and generates the result. It then creates a data compressed version (mp3 or AAC) and goes back to dither the aiff/wav before writing the file.

 

That's true if you're dithering in the bounce dialog, but not if your using a dithering plug-in, which I had assumed you were doing. Sorry.

[Ed Ryan]

 

That's true if you're dithering in the bounce dialog, but not if your using a dithering plug-in, which I had assumed you were doing. Sorry.

 

No need to apologise. I know a few people that do this, and in other daws like Cubase you have to insert Dither as a plugin...

 

I'd still like to know more though !

[dpodosek]

Post bounce- My mixes don't sound as clear and full or as they do in logic. iTunes sounds the worst. On CD sounds a bit better, and with the quicktime player they sound even a bit better than on CD. But the native Logic file is what sounds the best.

 

[sureno]

LOL if i bounce off line my timing is WAY out, i just stick to real time now. i had synths coming in a half second late, sounded horrible and no it wasn't my music

[davidpye]

Bizarre, I've NEVER had a problem with this!?!?

[Malk]

I have never had this happen iam using logic 7. something osx 10.3.9 and offline is identical to real time this is mad how many people got this going on?

[kwackman]

So, I was bored tonight......

 

I just finished a tune at the weekend, and read this thread.

I'd used a mic'd up acoustic guitar, EZ drummer for drums and Trilogy for bass for this tune.

I had bounced down and automatically added a copy to iTunes.

I work at 44.1Khz and 24bit.

The iTunes version is at the same sample and bit rate.

 

Tonight, I imported the itunes version into the original project, and pulled up the gain plug in and reversed the phase of both channels.

After playing with the gain to get the maximum phase- minimum audio- i got a strange result.

I got what I expected- almost perfect cancelation except for some reverb. So I was happy my bounce and original were the same EXCEPT for the drums.

The hihat and snares and cymbals were fairly much still there, not just the reverb.

I was going nuts until I remembered, EZ drummer plays different samples each time it's triggered!

DOOH!

So, now I'm happy that my bounces are the same as the Logic version.

 

So, anyone doing the phase reverse check- if any of your AU instrument do different samples randomly....... well don't do what I did!

[Ed Ryan]

See I tried this.

 

It did cancel out, but not at 0dB. I had to pull one of the faders down to get a cancellation....

 

So for me I'm staying with online bounces. Just to be sure...

[davidpye]

It did cancel out, but not at 0dB. I had to pull one of the faders down to get a cancellation....

 

And you're sure you're not using normalize? Are you running any plugins on your master output? Is your master fader set to 0dB?

 

I've always used Offline Bounces, not exclusively, but I'd never NOT use it if I wanted to, and I've NEVER had any differences between those bounces and the project itself.

You really should check the bounce within LOGIC if you're unsure though, any other way introduces way too many variables, especially iTunes!

[kwackman]

See I tried this.

 

It did cancel out, but not at 0dB. I had to pull one of the faders down to get a cancellation....

 

So for me I'm staying with online bounces. Just to be sure...

 

If it cancelled out- that's good!

The fact you had to move a fader IMHO ain't a big thing as long as we're not talking about about more than a few dB.

The gain thing coud be to do with the pan law in your project etc, and really isn't a big problem. The fact that you can get it to cancel out is the important thing, as that shows that your mix (releative volumes, EQ, comptression etc) IS the same.

 

I work with pro classical music balancers (I am NOT one of them- I'm never near the faders!!) who set stereo mics up using the phase cancel thing with someone talking into the middle of the mic. They're aren't too concerned about the fact that one half of the mic is a wee bit higher/lower than the other half of the mic, as long as they get the cancellation.