DaveClark
Max Output Level: -71 dBFS
- Total Posts : 956
- Joined: 2006/10/21 17:02:58
- Status: offline
RE: Robert Babicz on mastering
2008/05/15 10:26:29
(permalink)
Greetings all, I don't have time to get involved in some of the topics being discussed here except to point out something that I find very misleading: ORIGINAL: UnderTow
The dynamic convolution part does not have any effect on the dynamics. The only thing it does is look at the signal level and uses that to determine which of the 128 IR's to apply to the signal. This way the level of distortion in the emulations follows the signal levels just as the levels of distortion of the emulated devices is dependant on the signal level. Again, the actual dynamics are controlled by a regular old digital compressor... First, it would be quite unusual for convolution (dynamic or otherwise) not to have any effect on the dynamics. One of the biggest problems I have with regular convolution in my own work in binaural processing is the effect on the dynamics. Large spikes with little energy are produced all the time, and periodically the IR convolved with the signal produces large swells with lots of energy, for example. One of the very first things one should expect after convolution with IR's is that the dynamics will have changed --- not "compressed" or "expanded" as we typically use these terms, but definitely changed. The idea that multiplying spectra together will not change the dynamics in the time domain is hopelessly optimistic! Dynamic convolution in audio as described by Michael Kemp <1> and others is aimed at modelling nonlinear devices whereas normal convolution is applied to linear problems; essentially the nonlinear problem is approximated with a time series of linear problems. The word "dynamic" does not refer to the signal level as Kemp clearly points out. IR's are often measured only with respect to the signal level as UnderTow claims, but as Kemp suggests, they could also be measured according to the sign of the signal (one for postitive, another another for negative). The work "dynamic" has to do with the fact that more than one IR is used; i.e. as time passes, the IR being used is exchanged for another one. Each IR represents a response to a linear system; as combined, they represent the response of a nonlinear system. Because of the changing out of IR's, if one models a compressor as a nonlinear system with a collection of IR's, then yes, compression will result from dynamic convolution that uses those IR's. That's the whole idea of using dynamic convolution to model a compressor. If you do that, you don't need a compressor. To simplify: One of the many attributes of an IR is amplitude. If the amplitude suddenly stops rising so fast as a sequence is produced for a system as increasing impulses are applied, then when an actual input signal is convolved with those IR's, compression results as a byproduct. Similarly for an expander. (This is not to say the people don't subsequently use compressors; I'm talking only about the expected effects of the convolution.) Regards, Dave Clark -------------------- <1> Michael Kemp. Dynamic Convolution. AES, 1999, Munich. Formerly titled: Analysis and simulation of non-linear audio processes using finite impulse responses derived at multiple impulse amplitudes.
|
John
Forum Host
- Total Posts : 30467
- Joined: 2003/11/06 11:53:17
- Status: offline
RE: Robert Babicz on mastering
2008/05/15 10:38:51
(permalink)
The idea that multiplying spectra together will not change the dynamics in the time domain is hopelessly optimistic! What you say in the bulk of your post seems reasonable to me except the part above. Here are you not talking about phase distortion? This would not be a dynamics problem but a phase problem. Correct me if I am misunderstanding you here. Phase can impact the dynamics if I understand you here.
|
DaveClark
Max Output Level: -71 dBFS
- Total Posts : 956
- Joined: 2006/10/21 17:02:58
- Status: offline
RE: Robert Babicz on mastering
2008/05/16 07:31:10
(permalink)
Hi John, ORIGINAL: John
What you say in the bulk of your post seems reasonable to me except the part above. Here are you not talking about phase distortion? This would not be a dynamics problem but a phase problem. Correct me if I am misunderstanding you here. Phase can impact the dynamics if I understand you here.
No, I'm talking about --- just to take a silly example for illustration --- feeding a constant sinusoid into a system that has no decay. There are no "dynamics" here at all, yet the energy builds without limit; the result is (again being silly) "infinite dynamics." More realistically, and as you know from experience, inappropriately long reverb times for loud music lead to long blasts of sound without relief, or limited dynamics, even for punctuated styles of music. At the risk of causing confusion, I was merely re-expressing what convolution in the time domain is in the frequency domain (i.e. multiplication) to give another perspective. If one cannot see the effect on dynamics in the time domain --- which I personally think is more obvious --- perhaps they can see it in the frequency perspective. Anyone who knows that multiplication in the frequency domain causes smearing in the time domain would be surprised by the claim "there is no change in the dynamics." ------------------------- My intent was not to make a big deal out of this, but to correct the possibly mistaken impression that dynamic convolution cannot model compression; it certainly can. The purpose of introducing dynamic convolution into audio was to model nonlinear processes such as compression. Regards, Dave Clark
|
guitartrek
Max Output Level: -47 dBFS
- Total Posts : 2842
- Joined: 2006/02/26 12:37:57
- Status: offline
RE: Robert Babicz on mastering
2008/05/16 07:48:53
(permalink)
DaveClark - you bring an interesting perspective to this topic. I'm curious what you think of multiband compression and mastering plug-ins like Ozone3?
|
Danny Danzi
Moderator
- Total Posts : 5810
- Joined: 2006/10/05 13:42:39
- Location: DanziLand, NJ
- Status: offline
RE: Robert Babicz on mastering
2008/05/16 08:51:42
(permalink)
ORIGINAL: guitartrek
DaveClark - you bring an interesting perspective to this topic. I'm curious what you think of multiband compression and mastering plug-ins like Ozone3?
guitartrek: I know I'm not Dave, but I wanted to give you my take on that stuff if I may as well as a few other points? :) First let me start by saying I think what Robert mentioned in his video about saying no to a project that sounds bad is important. I do the same thing myself as I do not feel it's my job to polish a turd unless the project can't be fixed by way of a remix. That said, I like to do individual processing for mastering instead of using "all in one" type gear. I think multi-band comps are pretty much last resort type plugs when you are out of things to try to fix a mix. The same with the Ozone type stuff. I've never been a fan of it myself. I think it's important for everyone to use what works for them. Having talked to Bob Katz at great length both on the phone, emails and during the mastering of my album, I learned quite a bit from him but still do not take his words as "the voice of God" as much as I respect him. I've also spoke with and worked with 80's producer Beau Hill who has some awesome methods (and he used mulitple compressors and still does to this day since the 80's) he's shared with me as starting points. If a multi-band comp or Ozone plug works for you or anyone else and you can hear results that you are happy with, I sincerely feel you are right where you want to be. Another thing to keep in mind is a mastered project should not sound drastically different than the original that is sent to you. Sure, we make sane eq adjustments to where it is sonically in the right place and compressed properly and spacious if need be, but the mix should not come back sounding like something completely different than what was orignally sent. This of course should only happen in the event the client is stuck with the mix they have and it can't be fixed. Then of course you have to go through the pains of hell to make it sound as correct as possible. The problem we face these days are as Robert and Bob Katz have mentioned several times...loudness. The loudness wars have completely stripped dynamics out of music because record company big wigs could care less about quality. I have several big name friends in the business that have informed me that they were instructed to make things as loud as possible at times and quality was not an issue. This is the stuff Babicz and Katz are talking about. There are songs that are so loud at times, you can literally hear digital distortion and on some LED read-outs, you'll see and hear clipping. But when these plugs like the Waves L-2 etc are used in moderation after a nice eq curve is drawn and a slight compressor is used to keep the little peaks and valley's sane, the results are quite good and every bit as pro as what the real pro's are doing. I also agree with the one German poster that mentioned Babicz possibly losing us a bit due to the language barrier. The guy has something that works for him and the stuff he masters. Though I felt some of his statements were a little harsh and biased, we can't argue with someone that has a system that works just like the people on this forum. If we listed all our likes and dislikes with plugins, recording techniques, mastering, big guitar sounds and drums etc, we'd probably all learn a lot about stuff we had never thought to try. Some of it would work well, other things may not be for us in our particular applications. So I try to accept when someone has a system that works even if I'm not down with it. It's just like a song...sometimes you love the song, sometimes everyone else does and you hate it. ;) But in my opinion, if something works for you and you're content with the results you're getting, there really is no right or wrong if you the creator is happy with the outcome.
post edited by Danny Danzi - 2008/05/16 08:52:21
My Site Fractal Audio Endorsed Artist & Beta Tester
|
Dr. Mac
Max Output Level: -78 dBFS
- Total Posts : 618
- Joined: 2006/07/19 22:50:18
- Location: Upstate New York
- Status: offline
RE: Robert Babicz on mastering
2008/05/16 09:47:22
(permalink)
I love people who purchase millions of dollars worth of vintage equipment just to add upper-harmonic distortion... wow... "now it sounds warm!"
RME FireFace 800, 3.4GHz quad-core AMD-64, 8 Gigs RAM Sonar X2a Producer, Fav. Plugs: Ozone 5 Advanced, Waves, Sonnox, Melodyne, Voxengo, SSL Native, Drumagog 5 Platinum
|
keith
Max Output Level: -36.5 dBFS
- Total Posts : 3882
- Joined: 2003/12/10 09:49:35
- Status: offline
RE: Robert Babicz on mastering
2008/05/16 10:11:39
(permalink)
ORIGINAL: ...wicked
It seems like the backbone of his mastering technique is to stack compressors and apply them in milder doses. Okee, that's pretty badass I reckon (if you've got gobs o' antique compressors) I bet Voxengo has a product that does something like that...  Oh wait, that's cheating!
|
keith
Max Output Level: -36.5 dBFS
- Total Posts : 3882
- Joined: 2003/12/10 09:49:35
- Status: offline
RE: Robert Babicz on mastering
2008/05/16 10:35:44
(permalink)
ORIGINAL: jamjar
Trust me, there is no difference between a $50 cable and a $500 cable. Yes, but here's the difference between a $5 and a $25 cable: you'll still be using the $25 cable 10 years from now.
|
Dr. Mac
Max Output Level: -78 dBFS
- Total Posts : 618
- Joined: 2006/07/19 22:50:18
- Location: Upstate New York
- Status: offline
RE: Robert Babicz on mastering
2008/05/16 10:42:32
(permalink)
I bet Voxengo has a product that does something like that... Oh wait, that's cheating! Ha ha! Cheating indeed! How dare we?
RME FireFace 800, 3.4GHz quad-core AMD-64, 8 Gigs RAM Sonar X2a Producer, Fav. Plugs: Ozone 5 Advanced, Waves, Sonnox, Melodyne, Voxengo, SSL Native, Drumagog 5 Platinum
|
21doors
Max Output Level: -87 dBFS
- Total Posts : 170
- Joined: 2007/11/27 11:38:12
- Status: offline
RE: Robert Babicz on mastering
2008/05/16 13:58:48
(permalink)
But when these plugs like the Waves L-2 etc are used in moderation after a nice eq curve is drawn and a slight compressor is used to keep the little peaks and valley's sane, the results are quite good and every bit as pro as what the real pro's are doing. Nicely put. I started to get much better masters ITB when I reduced their affect. But I also have to agree that tape is very improtant if you want that something extra. Now if my phone range off the hook, and my studio looked like his, yeah I'd want the very best cables at any cost. If for any reason, just to secure the clients with bottomless budgets who want 'the best'.
|
UnderTow
Max Output Level: -37 dBFS
- Total Posts : 3848
- Joined: 2004/01/06 12:13:49
- Status: offline
RE: Robert Babicz on mastering
2008/05/16 19:40:05
(permalink)
I think I need to clarify my comments about dynamic convolution.
The way the Sintefex FX2000, FX8000 and their derivative products (Liquid Mix, Liquid Channel) work is by looking at two different aspects of any modelled compressor in two steps. One step is the dynamic convolution aspect which measures the phase distortion, frequency response etc of the device at different levels by sending a series of pulses into the device (with ratio set to 1:1) and looking at what comes back out. It also allows you to model the compression curve of the modelled device by sending and analysing the dynamics of the return signals for different compression setting.
This second aspect, the compression curve, and the way it is modelled is no different to how many digital compressors model analogue devices. Of course it is very nice that the Sintefex product allows to do this all for you and turn it into an actual usable model rather than having a plug-in with a few pre-set compression curve models programmed by the developers.
So indeed, changing the model types on FX or Liquid products will give you different dynamic responses but this is no different than many other products on the market. The special thing about them (besides the ability to model devices you own) is the dynamic convolution which model other aspects of a device. Not the actual dynamic response.
I hope this is all a bit clearer than my previous post.
My point remains: Someone criticising digital compression should not use one of these products because in the end, they use fairly standard digital compression. They just have an added dimension to them due to their ability to also model the (non-linear) phase distortion and frequency response of a device at different levels.
UnderTow
|
UnderTow
Max Output Level: -37 dBFS
- Total Posts : 3848
- Joined: 2004/01/06 12:13:49
- Status: offline
RE: Robert Babicz on mastering
2008/05/16 20:12:56
(permalink)
ORIGINAL: DaveClark
No, I'm talking about --- just to take a silly example for illustration --- feeding a constant sinusoid into a system that has no decay. There are no "dynamics" here at all, yet the energy builds without limit; the result is (again being silly) "infinite dynamics." More realistically, and as you know from experience, inappropriately long reverb times for loud music lead to long blasts of sound without relief, or limited dynamics, even for punctuated styles of music.
Saying that feedback (or very long reverberation decays) affects "dynamics" is disingenuous. That is a bit like saying turning the volume up has changed the dynamics. It hasn't. It has just changed the level.
The purpose of introducing dynamic convolution into audio was to model nonlinear processes such as compression.
The only products that I am aware of using dynamic convolution are the Sintefex and Focusrite products. (All using the same system). Here are some lines from the Sintefex manual:
The SINTEFEX FX8000 REPLICATOR is a unique product which can capture an analogue signal path and reproduce it digitally, accurately simulating the frequency response and distortion characteristics of the original.
Replicator’s unique and patented non-linear technology takes it beyond this to reproduce the analogue characteristic of the sound changing when you drive the unit harder.
The word LINEAR is used in Replicator in a special way and simply applies to a sound process in which the output follows the input exactly. If an equaliser output doubles exactly when you double the input and otherwise sounds exactly the same it is LINEAR. This is the way most digital equalisers have worked (until Replicator came along).
Most real analogue equalisers and effects do not do this and so are NON-LINEAR (NON-LIN). That’s why your valve EQ sounds different when you drive it hard. This is one of the effects of analogue processing that Replicator was designed to replicate.
The heart of the SINTEFEX FX8000 REPLICATOR is a new processing system known as DYNAMIC CONVOLUTION. We won’t go into technical details here except to say that ordinary convolution has been known for a long time but DYNAMIC CONVOLUTION is new. This takes into account not just the way things respond to sound at one sound level like ordinary convolution, but how they respond at a whole set of levels. This means that when you drive a REPLICATOR effect harder, it sounds different – just like an original piece of analogue equipment.
...
If there is any automatic gain control in the system being sampled this must be switched off for the signal analysis, as it depends on varying level pulses being accurately assessed. Set a compressor to 1:1 or raise its threshold and keep an eye on its gain reduction. Gain reduction is sampled separately.
...
Snapshot sampling can capture the basic sound of the compressor signal path and this is a necessary step in compressor sampling. However in addition you need to capture the gain reduction curve – and this is what we mean by sampling the compressor curve.
In fact, Replicator can capture a whole set of curves for different gain reduction slopes (also known as ratios) and allows you to use them when you are simulating the compressor. In addition Replicator can work out curves for intermediate slopes and also down to 1:1 so after sampling a compressor curve you can simulate any setting from the highest slope curve you have sampled down to 1:1.
You can also choose threshold settings after sampling so you end up with a completely programmable compressor using the curves of the compressor you sampled.
In addition you can apply the signal processing characteristic of the compressor signal path to make the simulation very close to the original sound.
UnderTow
|
The Maillard Reaction
Max Output Level: 0 dBFS
- Total Posts : 31918
- Joined: 2004/07/09 20:02:20
- Status: offline
RE: Robert Babicz on mastering
2008/05/16 20:53:14
(permalink)
But it's true, long(er) reverberation decays reduce dynamic range.
The very definition of the term "reverb" supports the notion.
I'll admit I don't understand the context of what's being said with regards to this thread...
But taken at face value there seems to be some fuzzy audio going around.
best regards,
mike
|
The Maillard Reaction
Max Output Level: 0 dBFS
- Total Posts : 31918
- Joined: 2004/07/09 20:02:20
- Status: offline
RE: Robert Babicz on mastering
2008/05/16 21:02:48
(permalink)
I was gonna add...
Turning up the volume... does in general, reduce dynamics. (unless it's really quiet to start with) Once you begin raising the noise floor while confronting the diminishing returns of energy conveyance (in both electronic and acoutical mediums) in the loudest passages.... you are starting to compress. If you start really quiet you can increase dynamics for a while... until you start boosting the noise floor as fast or faster than raising the maximum amplitude.
It's not just the electronics... so it's not like you can say you'll just buy mo-better gear.
best,
mike
|
UnderTow
Max Output Level: -37 dBFS
- Total Posts : 3848
- Joined: 2004/01/06 12:13:49
- Status: offline
RE: Robert Babicz on mastering
2008/05/17 05:57:45
(permalink)
ORIGINAL: mike_mccue
I'll admit I don't understand the context of what's being said with regards to this thread...
It is just about someone criticising digital compressors and using one himself. Nothing more than that.  (Which proves yet again that people tend to hear what they want to hear). UnderTow
|
The Maillard Reaction
Max Output Level: 0 dBFS
- Total Posts : 31918
- Joined: 2004/07/09 20:02:20
- Status: offline
RE: Robert Babicz on mastering
2008/05/17 06:56:29
(permalink)
I'm going back to my cave....
:-)
|
deleter47
Max Output Level: -83 dBFS
- Total Posts : 379
- Joined: 2005/11/25 20:24:59
- Location: Rio Grande Valley
- Status: offline
RE: Robert Babicz on mastering
2008/05/17 08:48:15
(permalink)
In the GREAT QUEST TO ADD WARMTH.......IMHO some pluggies do add a little too much distortion, or smearing, or something (not an expert but I can hear it)....you throw in a couple of compressors here, a little convolution reverb there....and all that (purposeful) smearing, and distortion starts to add up (over many tracks). Maybe my ears have gone bonkers over the years but to me it seems to get worse when you mix it down from 24 to 16 bit. Before digital came along, when shopping for outboard gear, ie compressors, effects, amps, or whatever......I was always looking at the THD (total harmonic DISTORTION, and the S/N (signal to NOISE ratio....noise, distortion, smearing.....these were the enemies back then. Now were are paying guys to put this into our tracks.......Go figure.
" For what shall it profit a man, if he shall gain the whole world, and lose his own soul?"
|
John
Forum Host
- Total Posts : 30467
- Joined: 2003/11/06 11:53:17
- Status: offline
RE: Robert Babicz on mastering
2008/05/17 09:08:23
(permalink)
In the GREAT QUEST TO ADD WARMTH.......IMHO some pluggies do add a little too much distortion, or smearing, or something (not an expert but I can hear it)....you throw in a couple of compressors here, a little convolution reverb there....and all that (purposeful) smearing, and distortion starts to add up (over many tracks). Maybe my ears have gone bonkers over the years but to me it seems to get worse when you mix it down from 24 to 16 bit. Before digital came along, when shopping for outboard gear, ie compressors, effects, amps, or whatever......I was always looking at the THD (total harmonic DISTORTION, and the S/N (signal to NOISE ratio....noise, distortion, smearing.....these were the enemies back then. Now were are paying guys to put this into our tracks.......Go figure. I totally agree.
|
Tom F
Max Output Level: -48 dBFS
- Total Posts : 2749
- Joined: 2007/07/08 05:56:12
- Location: Vienna (the one in Europe)
- Status: offline
RE: Robert Babicz on mastering
2008/05/17 09:15:50
(permalink)
id say people just want the best of both worlds....actually i dont like all those "tubey" plugs ...
still i think you cant compare it the way you did...back in the "old" days my setup was pretty noisy...even if i had some good piece there...now there is exactly silence everywhere...but noise is important for psychoaccustics and personal soundinterpretation...actually i prefere noise over distortion...i even often add some very low noise in my tracks and they sound better than (my own hardcore dithering :-)
|
John
Forum Host
- Total Posts : 30467
- Joined: 2003/11/06 11:53:17
- Status: offline
RE: Robert Babicz on mastering
2008/05/17 09:32:30
(permalink)
psychoaccustics People use this word a lot. Most of the time with out knowing anything about it. What we often fail to keep in mind is that we are dealing with recorded sound. We may strive for sonic perfection but it is still going to be within the realm of recorded sound. It is what it is. I have a background in psychology yet I have no idea what psychoacoustics are. I don't think many do either. Using the term and being comfortable with it without a thorough background in its meaning and real word application is rather silly.
|
Tom F
Max Output Level: -48 dBFS
- Total Posts : 2749
- Joined: 2007/07/08 05:56:12
- Location: Vienna (the one in Europe)
- Status: offline
RE: Robert Babicz on mastering
2008/05/17 10:31:52
(permalink)
well john as usual you are very critical here - actually it seems that you might say that i am stupid because i use this word (thats what could be read between the lines) actually there are a lot of phenomena related to psychoacoutics and certainly i wont know them all...but i know a some and therefore i feel free to use this word...actually effects like the shepard effect, the cancellation of sound perception in the brain when sounds have an unantural short delay (dualstereo sounds played in a headphone), the relativity of the perception of loudness related to frequenciy ...and some other of those effects (very basic ones like the doppler effect) are no real mistery to those interested in the subject...actually there were several tests showing that very noisles records are percieved as more unpleasant as noisier ones (within a limit) because - and thats what the doctors say - the brain interpolates the noise info with the musical info creating a deeper perception (which obviosly comes from the subjective feeling of interpretation....) actually i have no grade in this but its funny how you (being also helpful very times) seem to be a bit arrogant...i know you say sonar is best (everyone claiming different must be a fool) / you can master like babic within minutes / and so on...personally sometimes i wonder where all your confidence in yourself comes from - i dont wanna be cocky - but actually sometimes you act like you were bob katz and quincy jones in one person... since you are very nice (sometimes) i would just ask you to be a bit less smugly cheers 
post edited by info@tomflair.com - 2008/05/17 11:08:34
|
UnderTow
Max Output Level: -37 dBFS
- Total Posts : 3848
- Joined: 2004/01/06 12:13:49
- Status: offline
RE: Robert Babicz on mastering
2008/05/17 10:37:47
(permalink)
ORIGINAL: deleter47
I was always looking at the THD and the S/N ratio
Some distortions are more objectionable to the ear than others. The typical numbers given in most hardware specs won't tell you this. Some very clean devices (according to S/N specs) don't sound very nice. Other devices with much higher S/N ratios sound sweet. Numbers never tell you how a device sounds. UnderTow
|
John
Forum Host
- Total Posts : 30467
- Joined: 2003/11/06 11:53:17
- Status: offline
RE: Robert Babicz on mastering
2008/05/17 11:00:31
(permalink)
ORIGINAL: info@tomflair.com
well john as usual you are very critical here - actually it seems that you might say that i am stupid because i use this word (thats what could be read between the lines)
actually there are a lot of phenomena related to psychoacoutics and certainly i wont know them all...but i know a some and therefore i feel free to use this word...actually effects like the shepard effect, the cancellation of sound perception in the brain when sounds have an unantural short delay (dualstereo sounds played in a headphone), the relativity of the perception of loudness related to frequenciy ...and some other of those effects (very basic ones like the doppler effect) are no real mistery to those interested in the subject...actually there were several tests showing that very noisles records are percieved as more unpleasant as noisier ones (within a limit) because - and thats what the doctors say - the brain interpolates the noise info with the musical info creating a deeper perception (which obviosly comes from the subjective feeling of interpretation....)
actually i have no grade in this but its funny how you (being also helpful very times) seem to be a bit arrogant...i know you say sonar is best (everyone claiming different must be a fool) / you can master like babic within minutes / and so on...personally sometimes i wonder where all your confidence in yourself comes from - i dont wanna be cocky - but actually sometimes you act like you were bob katz and quincy jones in one person...
since you are very nice (sometimes) i would just ask you to be a bit less smugly
cheers
I wasn't trying to be smug or saying I am a know it all just that we often say things that we are not as knowledgeable as we should be when we say things. I can include myself here too. I do know sound though. I do know what sounds good to me. That is all anyone really can say about sound. The why is best left to the theorist. One thing that pops up from time to time is that word being used as a reason for going to 96 k for sampling. Yet in this very thread the acknowledged expert cuts everything above 17 k in his mastering. Don't you think its sort of odd that all the expense and effort is so easily discarded by this very prominent person? Could it be that before we had spectrum analysis at the tips our fingers none of this had any real meaning? I was trying, I thought, to be a bit of a really check as the point of that post. You just happen to be the one that used it. There are tons of things that have to do with the subject of perception. Yet no one really knows how it really works. Even the experts don't know. Of course its was a very long time ago that I was studying perception.
|
Tom F
Max Output Level: -48 dBFS
- Total Posts : 2749
- Joined: 2007/07/08 05:56:12
- Location: Vienna (the one in Europe)
- Status: offline
RE: Robert Babicz on mastering
2008/05/17 11:24:04
(permalink)
hi - btw did you like the logo?  i am not a 96 guy either but i think that its a very different thing to cut above 17k at mastering or summing 40 files all cut at 17 (not that you were saying this) but there might be several points of view about bits and herztes...as far as i believe there is is more affecting accustic perception than the purely audible part...so i am also sure that interaction of frequencies in inaudible bands reaffect the audible band...so having those effects cut only at the latest stage might be very reasonbable....cheers
|
AJ_0000
Max Output Level: -76 dBFS
- Total Posts : 738
- Joined: 2007/05/05 01:32:03
- Status: offline
RE: Robert Babicz on mastering
2008/05/17 13:46:48
(permalink)
ORIGINAL: deleter47
In the GREAT QUEST TO ADD WARMTH.......IMHO some pluggies do add a little too much distortion, or smearing, or something (not an expert but I can hear it)....you throw in a couple of compressors here, a little convolution reverb there....and all that (purposeful) smearing, and distortion starts to add up (over many tracks). Maybe my ears have gone bonkers over the years but to me it seems to get worse when you mix it down from 24 to 16 bit. Before digital came along, when shopping for outboard gear, ie compressors, effects, amps, or whatever......I was always looking at the THD (total harmonic DISTORTION, and the S/N (signal to NOISE ratio....noise, distortion, smearing.....these were the enemies back then. Now were are paying guys to put this into our tracks.......Go figure.
Yep. I am someone who thinks without a doubt that high-quality, analog, tube gear is superior (in addition to being extremely expensive). However, I think digital gear or software plug-ins that attempt to emulate analog gear sound like garbage 9 times out of 10. I don't include analog synth emulations in that statement, some of which I think are excellent and use frequently. I do include amp simulators, with apologizes to those who use them. Something like the VC-64 in Sonar definitely has its uses, but you've got to be careful with it.
|
DaveClark
Max Output Level: -71 dBFS
- Total Posts : 956
- Joined: 2006/10/21 17:02:58
- Status: offline
RE: Robert Babicz on mastering
2008/05/17 15:25:04
(permalink)
Hi guitartrek, ORIGINAL: guitartrek
DaveClark - you bring an interesting perspective to this topic. I'm curious what you think of multiband compression and mastering plug-ins like Ozone3?
Thanks for your comment. I also enjoy reading all the different points of view expressed here. On that type of plugin: Sometimes I use them and sometimes I don't. Generally I regard them as a "quick fix." Unfortunately, as I'm sure you know, the industry has made a mess for itself by raising the average volume levels over the past few years. Some people are forced to use them, even though they would much rather not. Regards, Dave Clark
|
mark s
Max Output Level: -68 dBFS
- Total Posts : 1140
- Joined: 2004/01/20 22:08:41
- Location: Kansas City, Missouri
- Status: offline
RE: Robert Babicz on mastering
2008/05/17 15:38:31
(permalink)
ORIGINAL: s_barber
75 Ohm is the impedance level of the input and output devices. Not the resistance of the cable itself. Try an Ohmeter on that cable. It should read very, very close to zero, yet not absolute zero. Only a superconductor can do that and that's only in theory. Every wire has inherent resistance which is determined by length, size of wire, braided, unbraided, even the contacts not being gold or whatever will add resistance to the cable. It will be a very small value. That combined with a small capacitance coming from the cable wires running in paralell will create some kind of very subtle filter on the audio. It will be there and you may or may not hear it but it's there.
I think you ought to get another ohmmeter since you should be getting a reading somewhere around infinity. You don't want a cable with a short in it. What you're referring to is reactance/impedance which is something an ohmmeter cannot measure since that is only exhibited under an alternating current.
|
DaveClark
Max Output Level: -71 dBFS
- Total Posts : 956
- Joined: 2006/10/21 17:02:58
- Status: offline
RE: Robert Babicz on mastering
2008/05/17 16:12:11
(permalink)
ORIGINAL: Undertow
Saying that feedback (or very long reverberation decays) affects "dynamics" is disingenuous.
This is a personal attack that is unrelated to the claims I made. To prove it, UnderTow needs to show that I don't believe what I posted. There is no need for personal attacks. What should have been said was that in my post I confused a change in dynamics with a change in volume level or something like that rather than accusing me of dishonesty. ---------------------------- No, I did not confuse issues. Raising the volume level does not change the place in time of any particular impulse (thinking of Green's function approach that utilizes integration of pulses). Convolution does --- it modifies all of them. Saying that this is in the same class as raising the volume, therefore does not change the dynamics, suggests a misunderstanding of what convolutions actually do. Because one does not know the input in advance, there will be chance enhancements and chance cancellations, therefore places in time where sounds are enhanced and places where they are diminished. But this need not be debated. Simply go measure the loudest and softest portions of a small section of audio. Feed it through various convolution reverbs. Measure again. With a mere volume change, the difference in db is the same. With convolution reverb, it may or may not be. UnderTow would be correct if and only if for all convolution reverbs there was no difference. For those who like Gedanken experiments, think about an impulse response that is composed of two positive spikes separated by a time t and with amplitude one. Convolve this with a single note. Result: two notes. This much different than raising the volume! Where before there was silence, there is now sound. Now feed in two identical notes with exactly the same separation as the IR spikes. Result: Three notes with the second one twice as loud as the first. The dynamics have changed considerably (for example, from 0 db to 6db). Now take that original IR and make the second spike minus one. Again input two identical notes with the same separation as the IR spikes. Result: Two notes, but with twice the separation in time as the very first case. There were four, but the second and third cancelled each other out. ---------------------------- The fact that changes in dynamics in a particular situation do not specifically match compression or expansion is not a warrant to conclude that dynamics don't change. Regards, Dave Clark
|
UnderTow
Max Output Level: -37 dBFS
- Total Posts : 3848
- Joined: 2004/01/06 12:13:49
- Status: offline
RE: Robert Babicz on mastering
2008/05/17 16:16:43
(permalink)
ORIGINAL: mark s
ORIGINAL: s_barber
75 Ohm is the impedance level of the input and output devices. Not the resistance of the cable itself. Try an Ohmeter on that cable. It should read very, very close to zero, yet not absolute zero. Only a superconductor can do that and that's only in theory. Every wire has inherent resistance which is determined by length, size of wire, braided, unbraided, even the contacts not being gold or whatever will add resistance to the cable. It will be a very small value. That combined with a small capacitance coming from the cable wires running in paralell will create some kind of very subtle filter on the audio. It will be there and you may or may not hear it but it's there.
I think you ought to get another ohmmeter since you should be getting a reading somewhere around infinity.
A cable should never have a resistance near infinity. That means there is a break in the cable. It should be near zero as s_barber writes. (You connect both ends of the cable to the meter terminals!) UnderTow
|
UnderTow
Max Output Level: -37 dBFS
- Total Posts : 3848
- Joined: 2004/01/06 12:13:49
- Status: offline
RE: Robert Babicz on mastering
2008/05/17 16:39:51
(permalink)
ORIGINAL: DaveClark
ORIGINAL: Undertow
Saying that feedback (or very long reverberation decays) affects "dynamics" is disingenuous.
This is a personal attack that is unrelated to the claims I made. To prove it, UnderTow needs to show that I don't believe what I posted.
There is no need for personal attacks. What should have been said was that in my post I confused a change in dynamics with a change in volume level or something like that rather than accusing me of dishonesty.
Either you are being disingenuous as to the point of my post or you are missing my point. You don't seem stupid so I guessed you were being disingenuous. If you genuinely did not see my point, I apologize. You wrote: "Because of the changing out of IR's, if one models a compressor as a nonlinear system with a collection of IR's, then yes, compression will result from dynamic convolution that uses those IR's. That's the whole idea of using dynamic convolution to model a compressor." As I explained (and is written in the manual), this is not how these devices work so I do not agree that this is the idea behind using dynamic convolution as applied in these devices. Sure anything that affects a signal can have an effect on the dynamics of the signal. That is not what I am talking about. I am saying that these devices do not use convolution (dynamic or not) to model the dynamic processing of a compressor. These devices use dynamic convolution to measure frequency response, distortion etc at different input signal levels. Just read the manual, it clearly explains to turn off all dynamic processing when IRing a compressor to measure it's non-linearities. UnderTow
post edited by UnderTow - 2008/05/17 17:02:21
|