Like color correction, I need to make all my audio sound the same in a scene, but not sure how to do that. When I put different sentences together from different takes, it sounds different. Possibly cause I had to turn down the gain or trim as the actors got louder or something like that. But that's just a guess. How do I get it to all match or blend if that makes sense. Thanks.
-
Wondering which camera, gear, computer, or software to buy? Ask in our Gear Guide.
Question about audio correction.
- Thread starter harmonica44
- Start date
Hire someone who knows what they are doing. You also need the proper tools - a quality EQ, good Noise Reduction, flexible dynamic processing and an IR reverb.
This is not a snide or flippant remark. Hire someone to do audio post on one of your projects and sit with them for every single painfully boring minute of editing and mixing. If you manage to stay awake and attentive - and you have a decent set of ears - you will observe and hear the process that turns disparate audio puzzle pieces into a coherent whole. I learned more about mixing in two days working with a master than 18 months of screwing around on my own.
BTW, did you ever get "Dialog Editing" by Purcell? That's precisely what the book is about.
The fundamental process is to trim each piece of audio to ONLY the dialog... edit that together -- then listen to it, it'll sound clunky. Now add your room tone to it (you recorded room tone, right?)... if you don't have room tone, build a clip of it from the silence between lines during the takes you've got.
Laying this underneath makes all the choppiness of the cut dialog disappear and smooths the edges. You can smooth them even further by leaving a hint of head and tail on each piece of dialog and fading in/out of them.
Everything else is experience based. You'll know on set that you can and can't do stuff based on what's screwed up your sound in the past. Experience is king here (which is why Alcove keeps insisting you get people who have it -- although they most likely learned by screwing up too, and having to solve/avoid those problems).
Time = Money; If you don't have money, spend time... if you don't have time, spend money!
Laying this underneath makes all the choppiness of the cut dialog disappear and smooths the edges. You can smooth them even further by leaving a hint of head and tail on each piece of dialog and fading in/out of them.
Everything else is experience based. You'll know on set that you can and can't do stuff based on what's screwed up your sound in the past. Experience is king here (which is why Alcove keeps insisting you get people who have it -- although they most likely learned by screwing up too, and having to solve/avoid those problems).
Time = Money; If you don't have money, spend time... if you don't have time, spend money!
Okay thanks, and yes I have room tone, but haven't added it yet. I hope this smooths it out. Which type of fade should I use?
The beginnings and endings of each line are different, so each fade is going to have to be different.
BTW, we have covered all this before in numerous posts answering questions that you asked; you should go back and read them.
BTW, we have covered all this before in numerous posts answering questions that you asked; you should go back and read them.
There is another, very important consideration that hasn't been mentioned, who is going to be watching your film and where are they going to be watching it?
The reason this is so important is that whatever speakers you are using and whatever room those speakers are in, your mix will sound different when played back on different systems. This difference might not be much of a problem for some web releases but if you are looking to get into a festival and have your film played in a cinema or you expect people to watch your film who have a decent home cinema system then you are likely to run into serious problems.
It's not just about experience, the equipment and monitoring environment you have to work with also come into play. If the speakers you are working with do not have a reasonably flat and full range frequency response or if the acoustics of the room your speakers are in has not been treated to be reasonably level and dead, a decade of experience and the best ears on the planet are not going to help. The likely problems are:
I'm trying to think of an analogy for the more visually oriented film maker: Imagine editing and colour correcting HD visuals on an old 18" NTSC CRT TV. It would probably look OK on another 18" NTSC CRT TV but play it on a decent quality 42" HD LCD TV and it almost certainly won't look how you expected, even had it been colour corrected by someone skilled and experienced. Potentially it's more serious with sound though as not being able to see the detail in the shadows or having an incorrect skin tone is not as serious a problem as not being able to hear the dialogue clearly.
G
The reason this is so important is that whatever speakers you are using and whatever room those speakers are in, your mix will sound different when played back on different systems. This difference might not be much of a problem for some web releases but if you are looking to get into a festival and have your film played in a cinema or you expect people to watch your film who have a decent home cinema system then you are likely to run into serious problems.
It's not just about experience, the equipment and monitoring environment you have to work with also come into play. If the speakers you are working with do not have a reasonably flat and full range frequency response or if the acoustics of the room your speakers are in has not been treated to be reasonably level and dead, a decade of experience and the best ears on the planet are not going to help. The likely problems are:
- Low frequency pops or rumbles, outside the effective frequency response of your speakers and therefore "invisible" in the edit suite suddenly becomes enormously apparent on cinema or home cinema systems.
- EQ between different lines/takes of Dialogue which appeared correct in the edit suite suddenly sounds completely incorrect.
- The balance between dialogue and music and sound FX changes, making the dialogue difficult to hear.
- Edits and crossfades "invisible" in the edit suite suddenly become annoyingly obvious.
- Differences in the levels and balance of ambiances and room tones, invisible in the edit suite become glaringly obvious and sound very amateurish.
I'm trying to think of an analogy for the more visually oriented film maker: Imagine editing and colour correcting HD visuals on an old 18" NTSC CRT TV. It would probably look OK on another 18" NTSC CRT TV but play it on a decent quality 42" HD LCD TV and it almost certainly won't look how you expected, even had it been colour corrected by someone skilled and experienced. Potentially it's more serious with sound though as not being able to see the detail in the shadows or having an incorrect skin tone is not as serious a problem as not being able to hear the dialogue clearly.
G
yep, and the solution for no-budget filmmakers isn't to hire the experience, it's to screw up, learn, gain the experience. H44 is that filmmaker. I'm sure he (and I) would love to be able to afford you or Alcove, but it's simply not a fiscal reality.
Yes, those are both real situations you've brought up in audio and video correction... ones I've run into and learned to overcome. I've had to, I couldn't afford someone to do it for me. I will absolutely say, I'm certain you could do a better job than I could, but I can make it passable on multiple systems and have had to do so. I've overcome those problems with limiters and carefully cutting super low and super high frequencies... and looking at my spectrum analyzers very closely.
H44, you have to dig into the problem solving process, you've made the first step time and time again... "I have a problem", "Is this a problem?" Those realizations need to be followed up by a problem solving regimen... your plethora of posts are great for bringing out information for stuffing this forum full of great searchable stuff and you're probably responsible for many of the new members here
... but you have to start shifting to figuring out how to solve these problems on your own... and having the confidence to accept that you're not going to get it perfect, but you can get it acceptable at your budget level.
The tools are on your computer already. For audio (as a visual person), you'll spend alot of time looking at waveforms (to see when sounds happen) and the spectrum analyzer (to see what sounds are happening at that point in time). In your Video work, there are the video scopes to analyze the color in the same way... learn them and love them.
From there, the solutions become: determine the problem, then open your toolbox to see if any of your tools fit as a solution. Your tool box is getting really full now with most of the stuff you'll need, you just need to put the different tools together in new and novel ways to solve problems you haven't encountered before. I've followed every single post you've made on this project... you have the tools to solve 99% of your own problems now... you just have to apply what you've learned to the new problems you're encountering.
In terms of the audio stuff... room tone has been covered in a ton of posts, dialog editing has been covered in a ton of posts. Try the solution first, then if that doesn't work and you don't seem to have the solution in your toolbox, ask the question. I was forced to learn these things with no access to mentors and it pulled my DIY mindset from my youth (dad, scouting, and geekery) into my film work... uphill both ways
but when you're on set and you're the captain of the ship (director), you need the confidence to say "I know we're behind schedule, but we need another take" or "I know it doesn't feel like it, but we really DO have what we need for the edit, let's move on and get ahead of the schedule" or (which I had to do on a shoot) "There will be a zeppelin parked behind the barn and you'll have seen it coming in as it shadowed your house... it's scary and you know it's the scavengers coming to loot your farmhouse." -- I didn't know how to do it when I made the statement, but I knew I could figure it out... so I was forced to figure it out.
Force yourself to really dig in and think through the problem/solution pair ... everything is possible, and you can do it -- we all can if we apply ourselves. If you fall down while trying, come here and we'll GLADLY help you move forward, but you have to move forward, get it close enough, move to the next bit, then if you still have time, you'll have a laundry list of the parts that need the most work... start with the worst and work down the list, when you run out of time, release it... move on. Build a body of work and a huge toolbox based on solutions that work for you, you'll start to know what will work and what won't.
Trust your eyes when you look at the monitor on set. Build crew that you use every time - you'll gain their trust and they yours. You'll have the opportunity to learn together, and you'll know that they will get the exposure correct, or ask for more light, or adjust the white balance to fit the lighting you have... or... or... This only works if you have hired experience... or fostered it. You can only foster that experience by creating more projects... you'll only get more projects if you finish this one.
Do the whole edit and the audio work, then look for the problems (watching it with people you trust to be BRUTALLY HONEST about it -- not just negative people and not family who will tell you it's great) Take notes on the comments and note the time in the edit. You may find that once you've done the editing and audio work that the problems you perceive don't exist after you lay in the room tone and ambience. Visual problems with a single shot may disappear in the edit when surrounded by other things that catch the attention... finish the edit and rough audio work first... concentrate on flow and timing... then move on to specific detaily things.
Always keep driving yourself forward... hard!
Sorry, long rant.
Yes, those are both real situations you've brought up in audio and video correction... ones I've run into and learned to overcome. I've had to, I couldn't afford someone to do it for me. I will absolutely say, I'm certain you could do a better job than I could, but I can make it passable on multiple systems and have had to do so. I've overcome those problems with limiters and carefully cutting super low and super high frequencies... and looking at my spectrum analyzers very closely.
H44, you have to dig into the problem solving process, you've made the first step time and time again... "I have a problem", "Is this a problem?" Those realizations need to be followed up by a problem solving regimen... your plethora of posts are great for bringing out information for stuffing this forum full of great searchable stuff and you're probably responsible for many of the new members here
... but you have to start shifting to figuring out how to solve these problems on your own... and having the confidence to accept that you're not going to get it perfect, but you can get it acceptable at your budget level.The tools are on your computer already. For audio (as a visual person), you'll spend alot of time looking at waveforms (to see when sounds happen) and the spectrum analyzer (to see what sounds are happening at that point in time). In your Video work, there are the video scopes to analyze the color in the same way... learn them and love them.
From there, the solutions become: determine the problem, then open your toolbox to see if any of your tools fit as a solution. Your tool box is getting really full now with most of the stuff you'll need, you just need to put the different tools together in new and novel ways to solve problems you haven't encountered before. I've followed every single post you've made on this project... you have the tools to solve 99% of your own problems now... you just have to apply what you've learned to the new problems you're encountering.
In terms of the audio stuff... room tone has been covered in a ton of posts, dialog editing has been covered in a ton of posts. Try the solution first, then if that doesn't work and you don't seem to have the solution in your toolbox, ask the question. I was forced to learn these things with no access to mentors and it pulled my DIY mindset from my youth (dad, scouting, and geekery) into my film work... uphill both ways
but when you're on set and you're the captain of the ship (director), you need the confidence to say "I know we're behind schedule, but we need another take" or "I know it doesn't feel like it, but we really DO have what we need for the edit, let's move on and get ahead of the schedule" or (which I had to do on a shoot) "There will be a zeppelin parked behind the barn and you'll have seen it coming in as it shadowed your house... it's scary and you know it's the scavengers coming to loot your farmhouse." -- I didn't know how to do it when I made the statement, but I knew I could figure it out... so I was forced to figure it out.
Force yourself to really dig in and think through the problem/solution pair ... everything is possible, and you can do it -- we all can if we apply ourselves. If you fall down while trying, come here and we'll GLADLY help you move forward, but you have to move forward, get it close enough, move to the next bit, then if you still have time, you'll have a laundry list of the parts that need the most work... start with the worst and work down the list, when you run out of time, release it... move on. Build a body of work and a huge toolbox based on solutions that work for you, you'll start to know what will work and what won't.
Trust your eyes when you look at the monitor on set. Build crew that you use every time - you'll gain their trust and they yours. You'll have the opportunity to learn together, and you'll know that they will get the exposure correct, or ask for more light, or adjust the white balance to fit the lighting you have... or... or... This only works if you have hired experience... or fostered it. You can only foster that experience by creating more projects... you'll only get more projects if you finish this one.
Do the whole edit and the audio work, then look for the problems (watching it with people you trust to be BRUTALLY HONEST about it -- not just negative people and not family who will tell you it's great) Take notes on the comments and note the time in the edit. You may find that once you've done the editing and audio work that the problems you perceive don't exist after you lay in the room tone and ambience. Visual problems with a single shot may disappear in the edit when surrounded by other things that catch the attention... finish the edit and rough audio work first... concentrate on flow and timing... then move on to specific detaily things.
Always keep driving yourself forward... hard!
Sorry, long rant.
A compressor could be useful to even the levels of the dialogs. It will make quieter parts louder while the parts that are already loud will remain the same. Before you apply a compressor you should first EQ and get rid of low frequencies or the compressor will boost them.
You can do this to individual dialogs and/or to all dialogs.
You can do this to individual dialogs and/or to all dialogs.
Using a compressor increases the amount of background noise in the audio, especially in indie projects. The way a real rerecording mixer works is to automate the volume. Yes, there may be a tiny little bit of compression on the dialog buss to add a bit of punch, but using it to level dialog is not a good idea.
This is, of course, contrary to the practice in the music industry to heavily compress vocals. This is one of the many things that folks coming from a music background need to relearn - sound for picture is very different than music, even mixing score/music for film is approached differently.
This is, of course, contrary to the practice in the music industry to heavily compress vocals. This is one of the many things that folks coming from a music background need to relearn - sound for picture is very different than music, even mixing score/music for film is approached differently.
Band-limiting the output would help for the first bullet point and limiting the dialogue would partially help the 3rd bullet point but it wouldn't help with the other points and would additionally introduce other problems.
For playback on TV, laptop or computers with average to weak sound systems, your approach could possibly give listen-able results but on a decent quality home cinema system or decent quality headphones it's much less likely to be passable and in a cinema (say at a festival), it's almost guaranteed to be unacceptable. Just so you know, limiters are never used professionally on dialogue in films destined for playback in a cinema and of course band-limiting not only removes the rumbles and pops you don't want but also removes any low or high frequency content you do want and there's no way of telling one from the other with a spectrum analyser.
I understand that H44 has zero budget and that he must learn to do the best he can with what he's got but conversely it's also important that he and other indy film makers are objective and realistic when it comes to what is possible and what is going to be acceptable with sound when you don't have the correct tools, the knowledge, the experience or the monitoring capabilities.
G
Here, we couldn't agree more! I'd love to hire you guys on every project we do (seriously, I've slung boom and operated camera simultaneously on a few, and the trade off sucks! So much missed in both the audio and the video)... but for most of the dozen + productions I've done... I've been the post person (edit, compositing, grade, dialog, sound design, master export). My time was the only currency I had to offer. I get the feeling that H44 is in the same boat... and the detail is becoming overwhelming, especially to someone who hasn't really had to focus on it as heavily as the promise of exposure has made necessary.
My goal is to help get him to the finish line... then the dissection can happen
That's where we learn how to overcome problems in our next productions.Oh, but I also come from a very competitive standpoint. In an audio class at school, our assignment was to redo the sound design of the intro to LOTR (where they're going over the history of the rings). We were told; no music; no dialog; sound design only... but don't bother trying to get it anywhere near the Hollywood standard as we'd only fail.
I took it as a personal challenge. I got a B on the project because I turned it in late, but I was the only one who used pure sound design to add life to the image -- and it was one of the class favorites due to the richness of the soundbed.
Don't tell H44 he can't... tell him how he can, that's what we're here for. The mistakes have been made, it's time for solutions (not: get a time machine, go back and redo it), not problems... we've already got those. Are you a solutions person, or a problems person?
If these things are so difficult to learn, how did you do it? Share that process... make competition for yourself! Competition makes us stronger!
I took it as a personal challenge. I got a B on the project because I turned it in late, but I was the only one who used pure sound design to add life to the image -- and it was one of the class favorites due to the richness of the soundbed.
Don't tell H44 he can't... tell him how he can, that's what we're here for. The mistakes have been made, it's time for solutions (not: get a time machine, go back and redo it), not problems... we've already got those. Are you a solutions person, or a problems person?
If these things are so difficult to learn, how did you do it? Share that process... make competition for yourself! Competition makes us stronger!
Okay thanks guys for all the info. I know how it can look on different systems. On my old computer a lot of the footage looked great. Now I got the new computer and monitor, and quite a few of the shots look somewhat blurry. I'll just have to live that and put them and make them the best I can. But I will play them through different speakers and plan on getting a better pair of headphones to hear it all. One thing I've noticed is that some of the dialogue I decided to use clips a little, but it's so subtle that I didn't even notice the clipping till about listening to it after a few times. Probably will be more noticeable on bigger speakers.
I am doing my best to find an expert to polish it all, but no one has responded to my adds so far.
I am doing my best to find an expert to polish it all, but no one has responded to my adds so far.
You're right. My advice only applies to quiet environments.
Using "Audacity", change the view to "Spectrum Analyzer" to see if the clipped audio occurs at a specific frequency, or if it's across the whole spectrum (top to bottom white at the time on the timeline when it clips). If it's just in a frequency, there's a possibility that a notch filter will alleviate the problem effectively (search for the yafi screencasts, I did some tutorials on this a while back).
If it's full spectrum, perhaps pulling back on some of it with a wider notch filter (lower "Q" value) will help hide the problem... you'll have to sweep the filter across the spectrum to find the worst offenders (my tutorials cover that as well). Alot of the repair side of audio editing is training your ear to hear the frequencies separately... it's hard to do, but the spectrum analyzer REALLY helps visual people get a grasp on how to "See" audio. I can't recommend this tool enough for video/film folks having to dabble in sound.
If it's full spectrum, perhaps pulling back on some of it with a wider notch filter (lower "Q" value) will help hide the problem... you'll have to sweep the filter across the spectrum to find the worst offenders (my tutorials cover that as well). Alot of the repair side of audio editing is training your ear to hear the frequencies separately... it's hard to do, but the spectrum analyzer REALLY helps visual people get a grasp on how to "See" audio. I can't recommend this tool enough for video/film folks having to dabble in sound.
If you knew my personal history and experience, you would be embarrassed to tell me that! I believe H44 can do anything he wants with his life, with enough dedication and self belief. There's no reason for example, why he can't be an airline pilot. But he won't learn to be an airline pilot in a couple of weeks just by trial and error and getting advice from an online forum.
I'll share it but I'm not sure it will be of much use: I trained as and was a professional orchestral musician for a number of years. I then studied recording technology, mostly on my own but with help from some top people in the business I'd met as a musician, then I started composing for low budget TV and doing some music production, working my way up the TV and film ladder for a few years. Over the course of a number of high budget projects I spent hundreds of hours watching and learning about audio post in quite a few of the higher end audio posts houses in London, gradually taking on sound FX, Foley, then dialogue editing, sound design and eventually re-recording and mixing. More recently I've been a course designer and course leader in "Sound & Music for TV and Film" for a higher education establishment in the UK. I taught the basic theory, tools and techniques used in audio post and hopefully, by the end of the two year full-time course, the students were capable of a basic level of competency. During my 20 years in audio post I've never stopped learning or studying. Quite a few people consider me an expert but not infrequently, I meet or communicate with others in my profession who know so much more than me that I feel almost like a beginner! That's understandable though as quite a few professionals spend their whole career specialising in just one of the audio post disciplines.
G
Harmonica44:-
Do the best you can and try it out on some various systems. When you're as happy with the sound as you can be, put it online somewhere and give me a link to it. I'll have a watch/listen and make some suggestions.
How long is the film?
Leave yourself plenty of time though!!! I'm not going to make any promises, I've got a music mastering project to do and audio post work on a Dutch film but if your film is short, if I have the time and if you haven't managed to achieve a reasonable mix, then I'll try to help you. That's a lot of if's and I'm a professional, not a charity, so I'm not going to make a habit of offering this but if it needs it and I have the time, I'll sort it out and mix it for you!
BTW, just so you know what I'm potentially offering, here is a pic of me in my mix room:
Do the best you can and try it out on some various systems. When you're as happy with the sound as you can be, put it online somewhere and give me a link to it. I'll have a watch/listen and make some suggestions.
How long is the film?
Leave yourself plenty of time though!!! I'm not going to make any promises, I've got a music mastering project to do and audio post work on a Dutch film but if your film is short, if I have the time and if you haven't managed to achieve a reasonable mix, then I'll try to help you. That's a lot of if's and I'm a professional, not a charity, so I'm not going to make a habit of offering this but if it needs it and I have the time, I'll sort it out and mix it for you!
BTW, just so you know what I'm potentially offering, here is a pic of me in my mix room:
Last edited:
I'm never embarrassed to speak my mind, I do apologize if you've felt attacked by me.
I have a dozen years of classical violin, and 3 decades of guitar and vocals. I've been forced to be the audio engineer in every project we've recorded (most lost to time)... with no mentorship whatsoever.
As stated, I'd love to be able to hire you for my next project as your posts speak of more specific knowledge than I bring to the table. The problem/solution dichotomy is a personal mantra every morning for me. It's a daily self analysis about how I'm going to approach the world and how I deal with the people in it.
I have no doubt in my mind that you have the solutions for H44's problems (some of which may involve a time machine -- that'd be cool). But you aren't conveying them to him in a digestible way. I spent 10 years as a professional IT person... my defining moment was when I had to teach someone to use a mouse for the first time and make them realize that the motion of the mouse had a direct correlation to the movement of the pointer on the screen. From that moment on, I realized that I had a specific set of base expectations of skill that was unrealistic outside of the world I'd grown up in and had come to understand was "normal."
H44's next project will definitely benefit from the errors made in this one... but for this project, there haven't been step by step solutions given... that's a problem considering the audience. My goal is not to bring you to task at all. My goal is to encourage you to address the problem as if you'd been handed a job by a client... how would you approach it, break it down, avoid audio engineering jargon until you've defined it. I know you KNOW the solution... I have no doubt that you could SHOW the solution directly -- the hard part about a forum is that you have to be able to TELL the person on the other end how to get to the solution as a progressive thing. Teach them to fish as it were.
I know how I would approach the problem... I want to know how you would approach it, as I'm sure you'd have a much better workflow that has been developed over years of professional work and training. You're the expert that makes me feel like a beginner! I want to hear how you would solve this problem (and I'd venture a guess, so would H44)... in a digestible way.
As always, it would help if H44 would provide a sample for us to evaluate
I have a dozen years of classical violin, and 3 decades of guitar and vocals. I've been forced to be the audio engineer in every project we've recorded (most lost to time)... with no mentorship whatsoever.
As stated, I'd love to be able to hire you for my next project as your posts speak of more specific knowledge than I bring to the table. The problem/solution dichotomy is a personal mantra every morning for me. It's a daily self analysis about how I'm going to approach the world and how I deal with the people in it.
I have no doubt in my mind that you have the solutions for H44's problems (some of which may involve a time machine -- that'd be cool). But you aren't conveying them to him in a digestible way. I spent 10 years as a professional IT person... my defining moment was when I had to teach someone to use a mouse for the first time and make them realize that the motion of the mouse had a direct correlation to the movement of the pointer on the screen. From that moment on, I realized that I had a specific set of base expectations of skill that was unrealistic outside of the world I'd grown up in and had come to understand was "normal."
H44's next project will definitely benefit from the errors made in this one... but for this project, there haven't been step by step solutions given... that's a problem considering the audience. My goal is not to bring you to task at all. My goal is to encourage you to address the problem as if you'd been handed a job by a client... how would you approach it, break it down, avoid audio engineering jargon until you've defined it. I know you KNOW the solution... I have no doubt that you could SHOW the solution directly -- the hard part about a forum is that you have to be able to TELL the person on the other end how to get to the solution as a progressive thing. Teach them to fish as it were.
I know how I would approach the problem... I want to know how you would approach it, as I'm sure you'd have a much better workflow that has been developed over years of professional work and training. You're the expert that makes me feel like a beginner! I want to hear how you would solve this problem (and I'd venture a guess, so would H44)... in a digestible way.
As always, it would help if H44 would provide a sample for us to evaluate