R/E/P Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: 1 [2] 3 4 ... 6   Go Down

Author Topic: WUMPV - Post Your Techniques Here  (Read 11713 times)

chrisj

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 959
Re: WUMPV - Post Your Techniques Here
« Reply #15 on: July 07, 2006, 08:23:56 PM »

I'm getting ready to do the full listening tests- should be good- I've also written a level matching program that I'm using to do it with, which is giving me +db measurements for people's entries by averaging every sample in the song Smile how's that for a big measurement window? I can post it if people want, perhaps on yousendit. It's a simple OSX program with a bunch of movieplayer controls that you drag and drop files onto. The level matching part is trickier- it only reads 16, 24 and 32 bit AIFFs and WAVs at 44.1K stereo, and some people's files (notably Patrik's) give it problems as Patrik's WAVs seem to not have a data block marker in the place the program expects it to be. I had a similar problem with UnderTow's, but then it went away. WTF???

This time around I controlled tone issues with only extreme treble and extreme bass EQs. The treble is the same one that threw up all over Rugged Cross, except for I've got some of the internals (a second FIR tap) onto their own controls now, so it's like I have a variation on a FIR filter with two taps which can give me more shaping of the highs boost. Seems to have mostly worked. I also wrote code to depict the brightness of the tone relative to pink noise. It seems that many natural phenomena have the same frequency weightings as pink noise, and whenever I've gone way under or way over the brightness of pink noise I've been scolded for it, so I'm giving myself a guideline.

The bass was weird- I'm doing basically a subsonic boost. No touchy anything in the actual bass, midbass, midrange, or mid treble. That's why it seems true to the sound of the mix. With respect I've totally ignored Cerberus's sensible advice on account of I think I just work better when choosing not to doctor things too much.

I got back into compression with this one, after many WUMPs where I was experimenting with doing it all with limiting/saturation. I'm doing 2.3/1 compression with the threshold at -6.75 db. I'm not getting more than one db of limiting over the entire verse, and not more than three db at any point including the chorus. In retrospect this was indeed real conservative, and I've been adjusting my various guidelines to give me a better idea of where most people seem to need the body of the sound to be, and where the sound starts to go seriously 'scrunch'. For that reason, this was an extremely useful learning WUMP for me.

I've experimented and I find that I can get the levels right up to most other entries without a lot of mess, with the following exceptions: I can't get peaks hotter than Sunny and Matt without making things scrunch too much, and I can't get the body of the sound hotter than Norbert and Matt without scrunching everything. This is not to say they didn't, what I'm saying is that I think I'm discovering areas of density where you CANNOT get things hotter, no matter who you are. I am not seeing pro examples exceeding these levels either. The areas are as follows as RMS loudness: remember, this is in a context of rock music that's being mastered hot, NOT classical or some form of accentuated dynamics. This is 'real world' as I see it.

'Blue', -18 to -15 db: very open, don't drop below -18 with music content or it'll sound unnaturally empty and quiet.

'Green', -15 to -12 db: okay for verses or low intensity stuff. Will feel open.

'Yellow', -12 to -9 db: choruses and loud bits, will feel dense and solid no matter how cleanly it's done. Should be reserved for dynamic effect.

'Red', -9 to -6 db: Sounds in this area will scrunch. The density is simply too high. It works very well for drum sounds as long as they do not ever EXCEED this zone.

Beyond 'Red': forget it. All that will happen is a big scrunch and it won't really rock any harder. Beyond red is hotter than -6db remember! This is already right at the brink of superhot commercial work that still sounds like the ME was working miracles. Instead of going hotter than -6db RMS, ever, stick to -6 as an absolute max and shape the EQ better.

With any of these, as you approach the brink of the next level up, it starts to seem like that level. A relaxed open verse at around -13.5db will really come across nice and open, one at -15 borders on being maybe too sparse, one at -12 is starting to get dense enough to lose that open feel. This is independent of how distortedly or cleanly you're doing it- it's purely a mechanical perception of nothing more than the RMS relative to peaks

Thanks everybody for all the raw materials to speculate upon! The next thing will be to see whether next WUMP I turn out a track everybody thinks is awesome- by observing these guidelines closely Very Happy

Matt_G

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 648
Re: WUMPV - Post Your Techniques Here
« Reply #16 on: July 07, 2006, 11:03:48 PM »

ericjenson wrote on Sat, 08 July 2006 02:38

Quote:

i never noticed a ringing at 18k. if you find that flaw in my master, please say it!


so far the half of the masters I've listened to , the ringing is still present.



Hi Eric, what monitors are you using? nothing weird showing up on the Dyn's, HD 600's or the spectrum analyzer (the dogs aren't restless either lol). There is a spike at 38kHz that is -90db down, but I doubt anyone could hear that & it would be lopped off when SRC'd to 44.1kHz. Actually I don't know too many guys who can hear 18kHz very well, myself included, so if you are hearing something weird there then kudos to you! Wink

Matt
Logged
Matthew Gray Mastering

Brisbane Australia

cerberus

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2651
Re: WUMPV - Post Your Techniques Here
« Reply #17 on: July 08, 2006, 12:32:00 AM »

index.php/fa/3111/0/
the flow chart to the left shows the basic chain that i set up during the first hour or so. this is not too different from what anyone would do except:  i set a limiter up first and get a hot signal immediately... i already can hear inside my head how i want this to go...once the limiter is set up for rms in the target neighborhood  i can begin to hear exactly what i need to do to make the client's needs and desires into reality. if possible, i will always try to exceed their expectations.

first off the mix was already overcompressed to death. that i would take care of in every process in one way or another but incrementally rebuilding the "push-pull" of ..ultimately air molocules!  until it sounded like a band playing music again, it's already about 20 db short of the full dynamic range of a cd. that is not a lot to chew on once the top gets bitten off.

but actual first process would be a transfer function from an old tape deck's line amp... this was the main "colorant" i used,    it has tubes and transformer qualities, but no dynamic effects, no noise such as transformers can make. so i will also be adding noise later because i want the dynamics to seem more like the natural way tape saturates, and to somewhat evoke tape because after all..something has to sound responsible for the flattened dynamics.... if only i had a phoenix for vst.

the ssl e-channel has the odd order transistor-like harmonics covered, i find it helps with edge, but is not nearly as prone to outright distortion as the g-compressor..much smoother, it can be subtle, so i almost always use it coming out of the first part of chain and into the "refinement" section, which i have many options already set in a template, but don't always use all...here for example i used 2 renbass busses (named after the matt g who gave me the idea), but one is folded to mono...they are set about exactly an octave apart and don't add much, just a touch to fill in some holes where there just isn't solid bass that an eq could easily find.

if only i could make feeback loops, but my mixer can't... notice the signal appears to move in one boring general direction: forward: bah!   the complexity you see is for making people forget i did it with only software that has some limits...i do not try for complexity. but it happens i am experienced enough to know when i have reached the point of diminishing returns with one stage and it's time to move forward in the overall process. so i am very comfortable with it, it is systematic; although admittedly too advanced and too flexible to teach an algorithmic way to produce results by my methods at this moment, but i am working on simplifying it.

i should note that the dark purple processes in the bottom section of the complex chain also employ all the dark blue processes in the top section here, but i didn't show it because that would make for a busy flow chart...  otherwise this is also very similar to the chain i used for wump4 (rugged cross) but there was less parallelism there and no renbass, but a stereo lnmb instead doing a smash thing for parallel low level compression (bringing up details)...on "don't wait up. there was a c4 and a g-compressor in those slots..here i felt that kind of thing would make too much distortion, so none of those appear in this chain. there is also another transfer function (urei 530, flat) i applied to some of the sub-busses which helped the dullness from becoming oppressive.

also this time i spewed a noise source to three points in the chain at once...and used some dithers too...(idr, but my final dither to 16 bits is always mbit+, no noise shaping ever) it's my psychoacoustic noise pillow!  and for the time you listen, yours too.

some of the unusual techniques that appear in this chain (and i've described or tried to describe in the past but not under these names) are:  "perverted side channel" "yin yang continuum" , "ersatz k-stereo" "wave warping",  "airwindow slating", "noise pillow". "dna injections", "microdynamic reconstruction", "pre-staging the smash" and "reverse feed processing". some are "foundation techniques" that make some of the others possible for me....but i don't think we have enough time to review all of them.


jeff dinces

chrisj

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 959
Re: WUMPV - Post Your Techniques Here
« Reply #18 on: July 08, 2006, 01:20:13 AM »

Jeff is scaring me  Laughing

I thought I would mention that Pingu's entry inspired me to drag out my old transient designer code. I had a version of this effect that tended to center around a frequency range depending on how fast it was operating, and was pretty hard to control. I've been changing interface elements over to a simple text entry of dB amounts, and it seemed to me that maybe it would be better if I had the transient designer working that way too.

I ended up with a dual parallel transient designer where I have two separate versions reacting to the same initial signal (in other words, not in series) and putting in the effect based on dB amounts, plus a 'frequency center' adjustment for each of the stages. I hyped up the transientness of the 25 hz area (also affecting sort of the 'volume surge' capacity of the mix) by about 1.5 db and cooled off the transientness around 2.5K by -0.75 db. Also did the whole thing hotter than before in line with what I learned.

It came out NICE and I entirely owe it to Pingu's willingness to experiment and be heard. I have a new favorite way to tailor frequency phenomena- it's very organic or something, seems to really get out of the way or maybe I'm just nuts after listening to this track fifty million times Very Happy

So thanks, Pingu Smile

cerberus

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2651
Re: WUMPV - Post Your Techniques Here
« Reply #19 on: July 08, 2006, 01:21:48 AM »

chrisj wrote on Fri, 07 July 2006 20:23

I can post it if people want, perhaps on yousendit. It's a simple OSX program with a bunch of movieplayer controls that you drag and drop files onto.
yes please, and your master rocks dude!

chrisj wrote on Sat, 08 July 2006 01:20

I ended up with a dual parallel transient designer where I have two separate versions reacting to the same initial signal (in other words, not in series) and putting in the effect based on dB amounts, plus a 'frequency center' adjustment for each of the stages. I hyped up the transientness of the 25 hz area (also affecting sort of the 'volume surge' capacity of the mix) by about 1.5 db and cooled off the transientness around 2.5K by -0.75 db. Also did the whole thing hotter than before in line with what I learned.

It came out NICE and I entirely owe it to Pingu's willingness to experiment and be heard. I have a new favorite way to tailor frequency phenomena- it's very organic or something, seems to really get out of the way or maybe I'm just nuts after listening to this track fifty million times :dSo thanks, Pingu Smile

you guys are the sultans of slew...


jeff dinces

chrisj

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 959
Re: WUMPV - Post Your Techniques Here
« Reply #20 on: July 08, 2006, 01:27:44 AM »

cerberus wrote on Sat, 08 July 2006 00:32

some of the unusual techniques that appear in this chain (and i've described or tried to describe in the past but not under these names) are:  "perverted side channel" "yin yang continuum" , "ersatz k-stereo" "wave warping",  "airwindow slating", "noise pillow". "dna injections", "microdynamic reconstruction", "pre-staging the smash" and "reverse feed processing". some are "foundation techniques" that make some of the others possible for me....but i don't think we have enough time to review all of them.


In case anyone thinks Jeff is joking, I know what ersatz K-stereo is (a Haas effect delay with left and right swapped). Noise pillow means 'having noise there to make the background softer'. Pre-staging the smash means doing stuff to counter for a known amount of limiting- like expanding to get a similar dynamic swing, or distorting so that the limited bits will come out hotter.

I'm not sure what airwindow slating is unless it's synching up the parallel feeds properly Very Happy

I'm curious what microdynamic reconstruction is- anything like the fooling with transient designers I'm doing?

I must know the secrets of perverted side channel Very Happy explain, I beg of you!

cerberus

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2651
Re: WUMPV - Post Your Techniques Here
« Reply #21 on: July 08, 2006, 01:33:20 AM »

chrisj wrote on Sat, 08 July 2006 01:20

Jeff is scaring me  Laughing

you've been to your own website i take it...

jeff dinces

cerberus

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2651
Re: WUMPV - Post Your Techniques Here
« Reply #22 on: July 08, 2006, 01:39:17 AM »

chrisj wrote on Sat, 08 July 2006 01:27

I'm not sure what airwindow slating is unless it's synching up the parallel feeds properly Very Happy
you know it.

Quote:

I'm curious what microdynamic reconstruction is- anything like the fooling with transient designers I'm doing?
that and the yin/yang backwards compression, interesting things happen at the fringes of the envelope. push and pull, in small increments. running the transient designer through m/s and mixing a back to front reversed copy puts it into warp drive, youi'd have to use airwindow slating though  or ir might get a little mushy.  (cubase does easily allow tapping the chain -anywhere-, ymmv.)
Quote:

I must know the secrets of perverted side channel Very Happy explain, I beg of you!
that information is proprietary, i'd have to "slew" ya.

jeff dinces

chrisj

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 959
Re: WUMPV - Post Your Techniques Here
« Reply #23 on: July 08, 2006, 01:41:59 AM »

cerberus wrote on Sat, 08 July 2006 01:21

chrisj wrote on Fri, 07 July 2006 20:23

I can post it if people want, perhaps on yousendit. It's a simple OSX program with a bunch of movieplayer controls that you drag and drop files onto.
yes please


Done- http://www.yousendit.com/transfer.php?action=download&uf id=9DBADBC613A072B6

I make no guarantee that this thing will work, even for me, but it's what I was using. You drag the reference file onto the top movie player, and the other ones onto the other movieplayers. If it doesn't like the file it may say so and politely refuse to use it. If it locks up hit command-period and it should quit scanning (happens when it can't find a data block in the file). It might be pretty slow for many people seeing as it scans every sample of the file, and not that quickly- and the RMS loudness readings may not be relative to anything more than each other, if that Very Happy

Whaddaya want for nothin- rubbbbbba biscuit?

bow bow BOW...

Anyway, there you go, hope it amuses somebody...

cerberus

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2651
Re: WUMPV - Post Your Techniques Here
« Reply #24 on: July 08, 2006, 01:51:29 AM »

it appears to run ok here (g5). thanks!

jeff dinces

Pingu

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1196
Re: WUMPV - Post Your Techniques Here
« Reply #25 on: July 08, 2006, 04:23:34 AM »

chrisj wrote on Sat, 08 July 2006 13:20

Jeff is scaring me  Laughing

I thought I would mention that Pingu's entry inspired me to drag out my old transient designer code. I had a version of this effect that tended to center around a frequency range depending on how fast it was operating, and was pretty hard to control. I've been changing interface elements over to a simple text entry of dB amounts, and it seemed to me that maybe it would be better if I had the transient designer working that way too.

I ended up with a dual parallel transient designer where I have two separate versions reacting to the same initial signal (in other words, not in series) and putting in the effect based on dB amounts, plus a 'frequency center' adjustment for each of the stages. I hyped up the transientness of the 25 hz area (also affecting sort of the 'volume surge' capacity of the mix) by about 1.5 db and cooled off the transientness around 2.5K by -0.75 db. Also did the whole thing hotter than before in line with what I learned.

It came out NICE and I entirely owe it to Pingu's willingness to experiment and be heard. I have a new favorite way to tailor frequency phenomena- it's very organic or something, seems to really get out of the way or maybe I'm just nuts after listening to this track fifty million times Very Happy

So thanks, Pingu Smile





Hi Chris.

I don't know weather you read my technique or not.

I used algorithmix blue
then psp mastercomp.

I played with this tune for ages trying transient plugs and what have you but a simple chain did it for me.

Now the mastercomp has a mix function as we know and of course i tried pulling it back a little but it just didn't do it for me so i kept things where they are.

The settings roughly as i have lost the presets

ratio = 4.1
threshold around -17
auto make up.
I tried to set the attack and release to groove and move with the tune.
Attack was around 65
release was around 450
Stereo linking 100% (Might have been a bit more musical if i turned it to 70%)
Sidechain was set fairly low around 120.(I never set it that low on mastercomp but this time around the lower i went it just felt right so i went with it.?
Mix was about 85 - 90% cant remember.


Sorry i cant give more details about the eq i rushed this one as i wasn't going to join in until the last moment.

For the first time in my life im mastering without har-bal and visual gimics and its soo much more creative and true to the source.
Logged
If I defend myself I am attacked. But in defenselessness I will be strong, and I will learn what my defenses hide.

Patrik T

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 833
Re: WUMPV - Post Your Techniques Here
« Reply #26 on: July 08, 2006, 04:37:52 AM »

NoWo wrote on Fri, 07 July 2006 23:05

Hi Jeff,

good shot of you.
this astonished " I can not believe it" mimic is certainly based on the fact that you are listening to some Wump masters. But the hand/finger is not mine I swear  Very Happy

So blast the damn thing off...we don
Logged

NoWo

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 458
Re: WUMPV - Post Your Techniques Here
« Reply #27 on: July 08, 2006, 07:04:54 AM »

Hi Patrick,

I owe you one because of your insisting that my DAs were crap at my hot levels. You were so right and I have changed it.

So if I have some time to spare I try to be with you again on Wump VI, but you might know that handy programming is not the easiest thing on earth. To be true it is a boring killer. So I can
Logged

Patrik T

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 833
Re: WUMPV - Post Your Techniques Here
« Reply #28 on: July 08, 2006, 07:49:55 AM »

Hi Norbert.

The trouble is that you seem to have defined what world class is by referencing to ONLY very loud records out there. Each and every time, the really loud releases seems to be the only ones that are "word class" to you.

The source mixes for the latest WUMP's has not been sources that are even comparable to the releases you refer to. Those world class productions does not only becomes world class at Sterling (or whatever), but in 100% of their cases they are very much world class prior to the mastering. Did we have any WUMP source that felt like world class in itself? Frankly, the sources we've got so far have had more than a little to wish for from the MIXING standpoint.

So, if you're trying to extract a world class master from a mid- or low-class source, you will never get anything close to what you reference to, and eventually just end up with tiny, tiny balls of fire situated nowhere near "reality" or "class".

You have to treat the source from what's in that specific source. Not treat it from what was present in a completely different source and lead up to a completely different master. You can not build a skyscrape with just a few pieces of cardboard. You better make sure that the source is made of concrete before even starting to consider building a skyscrape.

Therefore WUMP is a good thing; it allows us to figure out what methodologies that works in a very specific case. It allows us to pinpoint maybe one or two masters that we personally feel just got things right, with regards to THAT source and not any other source or any other master out there.

Best Regards
Patrik
Logged

NoWo

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 458
Re: WUMPV - Post Your Techniques Here
« Reply #29 on: July 08, 2006, 08:22:57 AM »

Patrik T wrote on Sat, 08 July 2006 12:49

Hi Norbert.

The trouble is that you seem to have defined what world class is by referencing to ONLY very loud records out there. Each and every time, the really loud releases seems to be the only ones that are "word class" to you.




Hi Patrick,

this has been true for the past as I was not good enough to reach it without massive artefacts but this has changed recently. As I described I am asking for concrete aims that have to be met, and that could be others than simply being loud. To be honest as I can reach that loudnes now at ease (and without distortion  Smile ) I am no longer so much interested in it. And of course loudness can very from genre to genre, so the importance of matching a certain world class master becomes even more important, but this also still means meeting it
Logged
Pages: 1 [2] 3 4 ... 6   Go Up