Since I'm doing it dry, in a closest full of hung clothes, as I was told to do, does it matter if I use a shotgun or a hypercardioid? Whichever is best.
editing Which mic should I use for dry ADR?
- Thread starter harmonica44
- Start date
There SHOULD NOT be any static. You either have a bad cable, a bad interface, a bad mic, poor sound isolation, poor RF/EM isolation, a clumsy actor, or you're just a crappy engineer or a combination thereof - take your pick.
These things should be taken care of long before your session.
These things should be taken care of long before your session.
True. I'm using these 'domestic methods' in several tests and got no static, harmonica44. And the blanket really worked very well. Which interface are you using?
Well normally I do well at recording the actors voices or at least it seems so. But this time there was just static. Thought maybe it was the clothes but I guess not.
But I still seem to have enough usable material so far. So which reverbs in Audition do I use then, to make it sound she is in that room? I need to simulate a living room, a hallway, and a bathroom.
But I still seem to have enough usable material so far. So which reverbs in Audition do I use then, to make it sound she is in that room? I need to simulate a living room, a hallway, and a bathroom.
I don't use Audition, I use Pro Tools and Altiverb.
This is one of those things that you'll have to do on your own; that's what your ears are for. Even using Altiverb it can take me a while. Choose a reverb preset and adjust the EQ, reverb time, predelay, density, damping, mix, positioning, etc. until it matches the production dialog. Keep it in MONO. If the scene is all ADR then you just create a sound that is "authentic," whatever that means. DON'T DROWN IT!!!!!!!!
This is where an IR or Convolution reverb does a superior job as they are digital representations of real ambient spaces. If you don't have an IR or Convolution reverb you may have to put an EQ plug-in, a delay plug-in and a reverb plug-in on one effects return. This will make your job MUCH harder, but you do what you have to do.
I'll second what Alcove said regarding your "static" problem. The description "static" potentially could be a wide range of issues, maybe you can post a sample of the ADR you've recorded for us to hear and then we maybe able to identify more accurately what the problem is and what the cure (if there is one) might be. Chances are though that re-recording the ADR would be the best solution.
Like Alcove and virtually all audio post pros, I use ProTools and therefore can't give any advice with regard to Audition. However, now might be a good time to provide a general, basic introduction to reverb. I'm assuming H44 you've read my post #31 and are at the correct stage in the process to be adding the reverb?
Reverb: - Is essentially a series of echos. The delay (time) between the original sound and the first few echos (early reflections), the delay between subsequent echos (the reverb) and the number, direction, balance and frequency spectrum of the echos are the variables used by the brain to determine the physical space in which a sound occurred. To artificially recreate reverb is therefore a complex process and there are two main approaches to achieving this, Convolution reverb and Algorithmic reverb.
Algorithmic Reverb: An algorithmic reverb essentially samples the sound it is being fed and using computer algorithms, generates an appropriate number of echos, each with the correct timing, direction, balance and frequency spectrum (reverb parameters) to recreate the effect of various types of spaces/rooms. Algorithmic reverbs always have a number of presets for the different room types: Hall, Large Hall, Small Room, Large Room, etc. On a good algorithmic reverb, all the parameters which make up the room type are available for adjustment, giving a lot of fine control to vary the effect, but to do so convincingly requires very considerable knowledge and experience.
Convolution Reverb: Is based on the idea of "recording" the reverb of an actual real space. Speakers and microphones are positioned in a particular room/space and a series of special sounds (impulses) are played out of speakers and recorded through the mics. The mics have now recorded both the impulse and the reverb (impulse response) of our room/space. Software can remove this impulse, leaving just the impulse response (IR) and the convolution reverb generates its reverb effect from this IR. A convolution reverb therefore recreates specific rooms/spaces rather than general classes of spaces. So, providing you have the IR installed, instead of just a general setting like "Large Hall" you might have the IR for say "Carnegie Hall" or the "Royal Albert Hall", etc. Because the reverb is based on a specific IR, convolution reverbs allow very little control over the reverb parameters compared to an algorithmic reverb.
Worldizing: There is one other method of adding reverb which is worth mentioning, even though it's rarely used professionally these days because of time/budget constraints. Worldizing is where you take your recordings of ADR, Foley and SFX back to the original filming location, play them out of a speaker and re-record them. In the process of course you are also recording the original reverb.
Which is best? That depends on the time you have available, your skills and knowledge of how to adjust the parameters of an algorithmic reverb and the availability of appropriate IRs for a convolution reverb. Some mixers use just convolution reverb because it's quick and easy to get a realistic sound but the best mixers can expertly use whichever they feel works best artistically.
To get good quality reverb is quite expensive both in terms of cost and computer resources. A good algorithmic reverb may need to generate 3,000 echos per second, each with it's own specific processing, this is quite CPU intensive. Good convolution reverbs are usually even more CPU intensive.
I've just described what reverb is and how it's artificially created but not the techniques of using it, which is way beyond the scope of a forum post. Hope this basic info is useful to some here though?
G
Like Alcove and virtually all audio post pros, I use ProTools and therefore can't give any advice with regard to Audition. However, now might be a good time to provide a general, basic introduction to reverb. I'm assuming H44 you've read my post #31 and are at the correct stage in the process to be adding the reverb?
Reverb: - Is essentially a series of echos. The delay (time) between the original sound and the first few echos (early reflections), the delay between subsequent echos (the reverb) and the number, direction, balance and frequency spectrum of the echos are the variables used by the brain to determine the physical space in which a sound occurred. To artificially recreate reverb is therefore a complex process and there are two main approaches to achieving this, Convolution reverb and Algorithmic reverb.
Algorithmic Reverb: An algorithmic reverb essentially samples the sound it is being fed and using computer algorithms, generates an appropriate number of echos, each with the correct timing, direction, balance and frequency spectrum (reverb parameters) to recreate the effect of various types of spaces/rooms. Algorithmic reverbs always have a number of presets for the different room types: Hall, Large Hall, Small Room, Large Room, etc. On a good algorithmic reverb, all the parameters which make up the room type are available for adjustment, giving a lot of fine control to vary the effect, but to do so convincingly requires very considerable knowledge and experience.
Convolution Reverb: Is based on the idea of "recording" the reverb of an actual real space. Speakers and microphones are positioned in a particular room/space and a series of special sounds (impulses) are played out of speakers and recorded through the mics. The mics have now recorded both the impulse and the reverb (impulse response) of our room/space. Software can remove this impulse, leaving just the impulse response (IR) and the convolution reverb generates its reverb effect from this IR. A convolution reverb therefore recreates specific rooms/spaces rather than general classes of spaces. So, providing you have the IR installed, instead of just a general setting like "Large Hall" you might have the IR for say "Carnegie Hall" or the "Royal Albert Hall", etc. Because the reverb is based on a specific IR, convolution reverbs allow very little control over the reverb parameters compared to an algorithmic reverb.
Worldizing: There is one other method of adding reverb which is worth mentioning, even though it's rarely used professionally these days because of time/budget constraints. Worldizing is where you take your recordings of ADR, Foley and SFX back to the original filming location, play them out of a speaker and re-record them. In the process of course you are also recording the original reverb.
Which is best? That depends on the time you have available, your skills and knowledge of how to adjust the parameters of an algorithmic reverb and the availability of appropriate IRs for a convolution reverb. Some mixers use just convolution reverb because it's quick and easy to get a realistic sound but the best mixers can expertly use whichever they feel works best artistically.
To get good quality reverb is quite expensive both in terms of cost and computer resources. A good algorithmic reverb may need to generate 3,000 echos per second, each with it's own specific processing, this is quite CPU intensive. Good convolution reverbs are usually even more CPU intensive.
I've just described what reverb is and how it's artificially created but not the techniques of using it, which is way beyond the scope of a forum post. Hope this basic info is useful to some here though?
G
Last edited:
Awesome post... I learned alot from it, and I've been working with audio since the mid 80s 
I do, however, disagree with your earlier post saying the Spectrum Analyzer is a tool to avoid.
When teaching people who can't abstract sounds into whatever your head interprets those into (I was happy as hell to find spectrum analyzers as that's EXACTLY what I see in my head when I hear sounds, and always have), you need to get them to relate the new information to something they are familiar with. On a filmmaking forum, most of the people here know visuals. It's a great starting point for teaching how to listen... you can even add the fact that once this process is internalized, the spectrum analyzer should be weaned away. That initial step of recognizing that sound is made up of little component parts of differential amplitude waves across a full spectrum of distinct frequencies is REALLY INCREDIBLY DIFFICULT! Like trying to teach blind students to write in cursive before introducing the alphabet to them through braille.
Nothing is unlearnable online, just depends on the dedication of both the students... and the teachers.
http://www.physicsforums.com/forumdisplay.php?f=62
I do, however, disagree with your earlier post saying the Spectrum Analyzer is a tool to avoid.When teaching people who can't abstract sounds into whatever your head interprets those into (I was happy as hell to find spectrum analyzers as that's EXACTLY what I see in my head when I hear sounds, and always have), you need to get them to relate the new information to something they are familiar with. On a filmmaking forum, most of the people here know visuals. It's a great starting point for teaching how to listen... you can even add the fact that once this process is internalized, the spectrum analyzer should be weaned away. That initial step of recognizing that sound is made up of little component parts of differential amplitude waves across a full spectrum of distinct frequencies is REALLY INCREDIBLY DIFFICULT! Like trying to teach blind students to write in cursive before introducing the alphabet to them through braille.
Nothing is unlearnable online, just depends on the dedication of both the students... and the teachers.
http://www.physicsforums.com/forumdisplay.php?f=62
He didn't say it was unlearnable, he said it was
There are plenty of YouTube videos and lots of audio forums that cater to beginners.
When it comes to H44 it sometimes seems that he's so afraid of making a mistake that he doesn't learn anything. We've covered all of this ground in numerous other threads, some of them in specific response to H44s questions. And some things you need to just spend hours and hours experimenting. Reverb is one of those things - you take a familiar sound or a piece of music, apply a reverb and start altering parameters to see what they do. The same applies to other plug-ins like EQ and dynamics processing.
Experiment, experiment, experiment. Edison found out 2,000 ways not to make a light bulb. However, the results of those tests gave him huge amounts of other data that led in many other useful directions. Don't be afraid to fail. In baseball, if you fail less than 70% of the time as a batter, you can be a candidate for the Baseball Hall of Fame.
Last edited:
Glad you found it useful Knightly.
The problem with spectrum analysers and graphical interfaces is the very point you've mentioned. Sound is by definition an aural sense rather than a visual one. Translating sound into vision is at best a huge approximation and at worse, completely misleading. There really is no choice, except to loose the reliance on vision and to develop aural skills. Film is after all an audio/visual medium, not just a visual medium.
I didn't say it was impossible to teach or learn reverb over the net, just that it would take huge effort and time and is therefore beyond the scope of a forum thread!
G
EDIT: Looks like alcove and I were posting at the same time. We've both said before, there really is no option to creating decent quality audio post but years of experience/experimentation.
The problem with spectrum analysers and graphical interfaces is the very point you've mentioned. Sound is by definition an aural sense rather than a visual one. Translating sound into vision is at best a huge approximation and at worse, completely misleading. There really is no choice, except to loose the reliance on vision and to develop aural skills. Film is after all an audio/visual medium, not just a visual medium.
I didn't say it was impossible to teach or learn reverb over the net, just that it would take huge effort and time and is therefore beyond the scope of a forum thread!
G
EDIT: Looks like alcove and I were posting at the same time. We've both said before, there really is no option to creating decent quality audio post but years of experience/experimentation.
Last edited:
I absolutely agree that experience and experimentation are the only way... but what do we (H44, other google friends -- hello google friends) experiment with? When we hit the roadblocks as indie filmmakers (not audio pros), we have difficulty finding folks like you guys to answer the questions in terms we can understand / relate to... and budgets don't allow us to do anything other than attempt to figure it out for ourselves up to a certain point (so far, a decade in my case), well enough to teach it to other people form an experientially substandard beginning point. I've been where H44 was... just not in Saskatchewan, where early on in the posts, he stated that there is no local scene where he's at... no mentors for him to approach without a plane ticket or convention ticket ($$$).
What I'm trying to get away from is the "Don't bother trying, you'll never get it because you don't have access to the right people/tools/knowledge" and on to the point of this board (not just for H4, but also for our google friends -- hello google friends) who will come here exactly the same way H44 did... and I did. Lost, confused, hopeless, trying to figure out where we'd screwed up.
I had a college professor once tell me not to bother trying to replicate a soundtrack even remotely the scale of a hollywood picture (or New Zealand in this case)... the exercise was to replace the entire sound track to the opening scene of LOTR without using music or dialog... I was one of the only students who didn't use music or dialog... I got a B, but I got so pissed at the dismissive attitude of that professor that I turned my project in late just to prove her wrong... I spent about 2 weeks polishing a fully structured soundtrack for that clip that ended up sounding absolutely epic.
She pissed me off, and convinced the other, younger students that they weren't capable of doing the task... in a sound production class taught by someone who went to LA to get a job out there as an editor. She let her lack of knowledge about that aspect bleed into her teaching... I don't want to have folks coming here and encountering "don't bother trying" when they search for "How do I use reverb in my ADR?" on google (Hello google friends). That's why I (as one of those searchers who came here 6500 posts ago with the typical questions you're now encountering) am trying to get answers other than "you'll shoot your eye out kid."
This is an educational sharing community for me and for many others. H44 is very effusive with questions, they may get frustrating at times, but our answers will lead to technique, that will lead to resharing of experience that will lead to more knowledge being built within this community. This isn't just a forum to me, it's my film school... not the one from whence I received my degree, but the one where I learned most of what I know about filmmaking... including that specializing makes a better film -- but that's not where we start here in "not-LA". We start doing every job on set, and figuring out desperately how to get our project across the finish line.
What I'm trying to get away from is the "Don't bother trying, you'll never get it because you don't have access to the right people/tools/knowledge" and on to the point of this board (not just for H4, but also for our google friends -- hello google friends) who will come here exactly the same way H44 did... and I did. Lost, confused, hopeless, trying to figure out where we'd screwed up.
I had a college professor once tell me not to bother trying to replicate a soundtrack even remotely the scale of a hollywood picture (or New Zealand in this case)... the exercise was to replace the entire sound track to the opening scene of LOTR without using music or dialog... I was one of the only students who didn't use music or dialog... I got a B, but I got so pissed at the dismissive attitude of that professor that I turned my project in late just to prove her wrong... I spent about 2 weeks polishing a fully structured soundtrack for that clip that ended up sounding absolutely epic.
She pissed me off, and convinced the other, younger students that they weren't capable of doing the task... in a sound production class taught by someone who went to LA to get a job out there as an editor. She let her lack of knowledge about that aspect bleed into her teaching... I don't want to have folks coming here and encountering "don't bother trying" when they search for "How do I use reverb in my ADR?" on google (Hello google friends). That's why I (as one of those searchers who came here 6500 posts ago with the typical questions you're now encountering) am trying to get answers other than "you'll shoot your eye out kid."
This is an educational sharing community for me and for many others. H44 is very effusive with questions, they may get frustrating at times, but our answers will lead to technique, that will lead to resharing of experience that will lead to more knowledge being built within this community. This isn't just a forum to me, it's my film school... not the one from whence I received my degree, but the one where I learned most of what I know about filmmaking... including that specializing makes a better film -- but that's not where we start here in "not-LA". We start doing every job on set, and figuring out desperately how to get our project across the finish line.
Okay thanks. I see that we have covered this on in other forums, but I didn't realize that some of what was covered before applies here. That's another thing, as a beginner, I don't know that one thing works for another, and so on. Well that book by Purcell finally arrived, so hopefully it can teach me a lot. A lot of the recorded new ADR sounds good, just need to reverb it now, and see what happens.
I understand the problem. To a large extent Alcove and I have tried to answer it but I can see that our answers can easily appear to be disingenuous. If I did not know much about sound my impression of some of Alcove's and my answers would be: "They are saying that because they want film makers to hire them, rather than to do it themselves". I don't deny that I feel film makers should hire me to do the sound rather than doing it themselves but I do deny that this is affecting or motivating my answers. There's a vicious circle type problem which crops up time and time again at every level of the audio world (sound or music, audiophile, professional or beginner). The application of reverb is just one example but it works well, so I'll stick with it.
Alcove and I have advocated experimentation, which in the case of algorithmic reverb means choosing a preset, changing the parameters and listening to what effect this has on the sound. In theory this seems simple enough but in practice it's incredibly difficult, because unknowingly our experiment is probably fatally flawed and we are not listening to what we think we are listening to. Let me illustrate with an example:
Let's take Early Reflections (ER), the initial few echos created by reverb and particularly important to the brain's interpretation of room size. Let's say we experiment with changing the balance (level) of the ER. As we change the level of the ER, we hear the difference it makes to the sound ... Or do we? There are in fact a whole bunch of other variables we need to consider. Let's say you're doing this experiment on a sound system in your bedroom, the sound is coming out of the speakers and hitting your ears but some of the sound is reflecting off the various surfaces in your bedroom (the walls, ceiling, windows, desk, etc).. In other words, your bedroom has it's own reverb and ER. So when you are changing the level of the ER on your reverb what you are really listening to is the interaction between your changes and the ER and reverb in your bedroom. So depending on your speakers, exactly where they are positioned and the acoustic properties of your bedroom, what you are learning from your experiment maybe completely wrong. This is also true of EQ (every room has it's own resonant frequencies, called modes) and in fact is true of the vast majority of all audio processing.
The only way to learn the effect of changing parameters is to eliminate (or at least minimise) the other variables. This means a very accurate playback system correctly placed in a room which has been designed and acoustically treated to respond as neutrally as possible, a very expensive undertaking. This is the vicious circle problem I mentioned in the first paragraph. The vast majority of novices will not have access to the type of listening environment necessary to actually learn how to use many of the audio tools required. It's like a beginner without an instructor, learning to drive a car by driving around a muddy field. Providing they only ever drive in a muddy field they'll eventually be fine but what if they ever have to drive on asphalt, with traffic, pedestrians and traffic signals? No matter how much experience they have of driving around the muddy field, they won't have a fraction of the knowledge or experience they will need to drive competently in a city and much of what they've learnt will probably have to be un-learnt. Same with our example above, providing our novice sound guy only ever mixes in his/her bedroom and everyone who ever listens to the mixes does so in our novice's bedroom, no problem but listen in the neighbour's sitting room or heaven forbid a cinema ...
So what is the solution? Traditionally it would be to either apprentice or intern at an audio post facility which has the required listening environment or go and study at a school which runs audio courses and therefore also has the required listening environments. Without this though I simply do not know what to suggest, beyond the obvious of just experiment changing gears and driving around the muddy field!
G
I disagree... with none of what you've said! I'd like to hire you so I don't have to keep teaching folks to do good enough, only to have to fix the problems afterwards because I wasn't the one slinging the boom (we're finally getting to the point where we mostly have one person on boom)... but it's still not consistent -- because we're not able to pay folks yet.
So I know that I have to either boom the whole project (which makes it hard to light and shoot, my primary role on set)... or when I do boom, I have to have my settings match the incorrect settings everyone else has used so the recordings match to make post easier... PITA!
I want budget, I want to hire pro sound everybody (or at least great attention to detail folks who learn really quickly and have no other work obligations) and not have to worry about this ever again... until then...
H44 - post clips of the specific problems... and we'll try to help you figure them out. You have access to experienced indie folks in exactly the same boat you are... and INDUSTRY PROS! Take the advice, abstract the information and apply it everywhere you think it goes. DO IT WRONG... figure out what's wrong (learn how to figure out what's wrong -- a fantastic skill to have, IMHO). Determining the problem will let you explore the options for fixing them. Give yourself time to fix them.
Or hire pros who know exactly what needs be done and have the equipment and experience to do it for you. I'd prefer this one, but I can't afford it yet. I'm a decade in with no ROI on any project I've done and little to no recognition/exposure from it -- welcome to the indie scene.
On the upshot, I'm on my way to shoot a soccer game as a paid gig... I've made at lease one good connection throughout the years in this endeavor... keep at it. It's REALLY addictive!
So I know that I have to either boom the whole project (which makes it hard to light and shoot, my primary role on set)... or when I do boom, I have to have my settings match the incorrect settings everyone else has used so the recordings match to make post easier... PITA!
I want budget, I want to hire pro sound everybody (or at least great attention to detail folks who learn really quickly and have no other work obligations) and not have to worry about this ever again... until then...
H44 - post clips of the specific problems... and we'll try to help you figure them out. You have access to experienced indie folks in exactly the same boat you are... and INDUSTRY PROS! Take the advice, abstract the information and apply it everywhere you think it goes. DO IT WRONG... figure out what's wrong (learn how to figure out what's wrong -- a fantastic skill to have, IMHO). Determining the problem will let you explore the options for fixing them. Give yourself time to fix them.
Or hire pros who know exactly what needs be done and have the equipment and experience to do it for you. I'd prefer this one, but I can't afford it yet. I'm a decade in with no ROI on any project I've done and little to no recognition/exposure from it -- welcome to the indie scene.
On the upshot, I'm on my way to shoot a soccer game as a paid gig... I've made at lease one good connection throughout the years in this endeavor... keep at it. It's REALLY addictive!