DIGITAL How To Temporarily Fix "Sparkle In The Rain" Surround Mix

[alk3997]

I had a few minutes this morning and so I thought I'd try a quick remix of the remix.

From what I've found out, the center channel and LFE are the only channels at the right levels. That was a surprise.

The corner channels are too high. So, my suggestion, if anyone wants to try this, is to use the following levels:

Center: 0dB
LFE : 0dB
Front Left: -4.5dB
Front Right: -4.5dB
Surround Left: -6dB
Surround Right: -6dB

Some may like the surrounds a little higher, so I tried -4.5dB on the surrounds and it still sounded good. I just liked everything a little more forward.

To be fair, I really can't tell if the corner levels should be -6dB or -4.5dB. That's a personal preference based on how the vocals should be mixed. We'll have to wait until the official corrected version is available to find out.

For those who would like to try this a slightly different way, you could spread the center channel to the front left/front right channels and not have to lower the corner channels as much. But that would take more time...

I'm listening to the 5.1-channel mix of East At Easter and it is very good with these corrections. I suspect most of us would have given this a "10" had it not been for the production errors.

I now understand Steven Wilson's comment about the 5.1-channel problems not being that difficult. It's just simple level adjustments and (maybe) center channel spread.

Hope everyone is able to enjoy the 5.1-channel programming of the disc now until the corrected version comes out.

Andy

 

[grill]

FYI
Here are the foobar DRM statistics of the mch audio:

foobar2000 1.1.15 / Dynamic Range Meter 1.1.1
log date: 2015-03-29 10:51:20

--------------------------------------------------------------------------------
Statistics for: ?-1_5_audio
Number of samples: 259951200
Duration: 45:08
--------------------------------------------------------------------------------

Channel 1 (FL) Channel 2 (FR) Channel 3 (C) Channel 4 (LFE) Channel 5 (SL) Channel 6 (SR)

Peak Value: -0.32 dB --- -0.32 dB --- -0.52 dB --- -19.29 dB --- -0.32 dB --- -0.42 dB
Avg RMS: -14.09 dB --- -13.87 dB --- -17.60 dB --- -37.08 dB --- -16.83 dB --- -16.54 dB
DR channel: 11.90 dB --- 11.29 dB --- 13.51 dB --- 13.86 dB --- 13.73 dB --- 13.10 dB
--------------------------------------------------------------------------------

Official DR Value: DR13

Samplerate: 96000 Hz
Channels: 6
Bits per sample: 24
Bitrate: 7211 kbps
Codec: FLAC
================================================================================

 

[JanBakker]

Can you tell me what programm you used? I used Audacity and couldn't find the setting to do it
but I just did. I used your settings except for the rears I did -5,0 and I like it that way. Thanks for pointing me
to another feature in Audacity :banana: this adaptation (?) made this a superb surround mix thanks!! 9 more songs to go :51banana:

 

[alk3997]

I'm using ProTools HD with a control surface. That allowed me to make changes in real-time and lock in good levels. -5dB should be very satisfying. Anything between -4.5dB and -6dB really becomes personal preference.

However, if you don't want to play with software, just play the disc and change the levels on your AVR. Make sure to write down the original correct levels first and then "undo" the change in levels on the AVR after playing the disc. For those using analog cables from a BD player, you could make those adjustments in the BD player instead.

 

[fredblue]

Up on the catwalk...



I forgot to label the wavs
from top down..

Front L,
Front R,
Centre,
LFE,
Rear L,
Rear R

 

[fredblue]

(You know its "getting serious" when I start cranking up the computer.. ..I hate the damn thing and do all this chit chat from the iPad/phone..! :ugham: )

 

[JanBakker]

That's the original right? Here's the corrected one:




I forgot to label the wavs too

from top down..

Front L,
Front R,
Centre,
LFE,
Rear L,
Rear R

 

[alk3997]

It's funny, on your original graph the corner channels are much more "dense" so they end up looking much louder on the graphs. In reality the center is peaking at around -0.5dB and the front left/right channels are peaking at -0.2dB. Not much difference. That is, of course, a large part of the problem.

The vocals are isolated in the center except for a tiny amount of delayed/reverbed vocals in the front left/right channels and slight delays in the rear channels. Very very slight.

The vocals aren't the only thing in the center channel but there is not much else. It's a nice effect if you have a good center channel and the center is at the right level relative to the corners.

Andy

 

[fredblue]

That's the original right? Here's the corrected one:

View attachment 17863

Yes, thanks I only downloaded DVD Audio Extractor for the first time the other day so I thought I'd play around with it now.

 

[JanBakker]

Yes, thanks I only downloaded DVD Audio Extractor for the first time the other day so I thought I'd play around with it now.

It's pretty easy to correct the files with Audacity if you like to play around with that one too . Makes it a much more pleasant listen.

 

[fredblue]

It's funny, on your original graph the corner channels are much more "dense" so they end up looking much louder on the graphs. In reality the center is peaking at around -0.5dB and the front left/right channels are peaking at -0.2dB. Not much difference. That is, of course, a large part of the problem.

The vocals are isolated in the center except for a tiny amount of delayed/reverbed vocals in the front left/right channels and slight delays in the rear channels. Very very slight.

The vocals aren't the only thing in the center channel but there is not much else. It's a nice effect if you have a good center channel and the center is at the right level relative to the corners.

Andy

In my layman's way I would imagine that's because almost everything is indeed happening in the two L&R pairs..?

Isolating the centre for the first time right now strikes me there's mostly "only" lead vocals in there..
and the sub isn't doing very much at all (isn't that as it should be? just sub bass reinforcement?)..

all these graphs and stuffs just not my thing (at all) so forgive me if I ask a really dumb question = do these graphs accurately portray the difference between musical energy/activity/content and sheer volume/loudness?

 

[fredblue]

It's pretty easy to correct the files with Audacity if you like to play around with that one too . Makes it a much more pleasant listen.

Ok, thanks! I might just do that (I know my way around Audacity a bit).

 

[JanBakker]

Ok, thanks! I might just do that (I know my way around Audacity a bit).

I'm a newbie with the programm so at first I couldn't find the right setting (amplify didn't seem right when you want to lower levels ).

 

[alk3997]

In my layman's way I would imagine that's because almost everything is indeed happening in the two L&R pairs..?

Isolating the centre for the first time right now strikes me there's mostly "only" lead vocals in there..
and the sub isn't doing very much at all (isn't that as it should be? just sub bass reinforcement?)..

all these graphs and stuffs just not my thing (at all) so forgive me if I ask a really dumb question = do these graphs accurately portray the difference between musical energy/activity/content and sheer volume/loudness?

You understand correctly. I'd just caution on the use of power or energy and stick with levels. They mean different things obviously. What you are looking at are simple level graphs. The levels are usually referenced in dB (logarithmic scale). Yours is normalized to max and min (+1 to -1).

It looks to me that what you have are peak levels over time, rather than RMS levels.

So breaking that all down, what you have are the difference in peak levels (not power) across all 6 channels at any time. From that you can get an idea of how loud the channels will sound relative to the other channels.

It is normal in a modern music mix to use the LFE sparingly and put the bass into the main channels, particularly the front. As you said, the LFE should function as a slight reinforcement for the lower bass. I believe this has become more popular because many people don't have their subwoofer levels correct or aren't using a subwoofer (have full frequency speakers instead).

I think I heard a small amount of guitar in the center channel as well. But, for the most part, it's vocals only.

 

[himey]

What is the "tool" in Audicity to accomplish this. TIA

 

[JanBakker]

Mine is in dutch but translated it would be amplify or something like that and then use alk3997's numbers.

 

[LizardKing]

Mine is in dutch but translated it would be amplify or something like that and then use alk3997's numbers.

Yes:

Effect -> Amplify

 

[himey]

Thanks much!

 

[LizardKing]

Manually tweaked the levels in my pre-processor... as a quick test...

It sounds like a SW mix now... phew...

Thanks!

 

[himey]

Manually tweaked the levels in my pre-processor... as a quick test...

It sounds like a SW mix now... phew...

Thanks!

Once you get used to the 80s production it sounds fine fixed.