Hi all,
I am in the middle of a project and am looking for a way to make some audio sound like it’s coming from further away and from behind the camera. I am not too familiar with audio stuff so if anyone can explain how to do this then that would be great.
Hi Felix.
It depends on where and how the video is filmed and the output format you are exporting to.
Was some of the audio recorded with the video at the point you want to add the effect?
Was it recorded inside or outside?
If the sound was recorded with the video, was it recorded as mono, stereo, binaural stereo, or using a full surround field microphone?
Are you exporting as stereo or as a muliti-channel surround sound output?
For a convincing placement, you need to have the use of a VST reverb plug-in capable of capturing the ambient acoustics of the area you filmed in. These plug-in effects then can 'read' a section of the soundtrack and reproduce the original environment, at which point, if you can record or have access to the recording you wish to add, you can feed that to the effect unit and fine tune the result. The source file should be as sound neutral as possible and close-miked or the effcts unit will not work effectively.
If you don't have access to such a plug-in, have a look on the net to see if there are any free VSTs that will do the job. There used to be one but I can't remember it's name. It was very good though.
I'm going to assume you are mixing to stereo or you would have a dedicated audio editor.
With stereo, it is difficult to have placement at varying heights or to discern whether a sound source originates from behind or in front. Binaural stereo overcomes this with very slight, almost imperceptible timing differences which involves calculating the distances between a person's ears and then works out the time delay, but even then there is no way to determine a sound from being dead center front or dead center rear.
Sound, like light, decreases over distance, so one trick is to gently lower or increase volume and to roll-off higher frequencies as the sound increases in distance, but these are subtle changes and often are not that effective on their own. Also as it is behind and not in shot, the viewer of the video has no idea of the distance involved.
If the shot involves something speeding past and suddenly comming into view then you could try adding a Doppler effect, where the sound, if approaching from behind will increase slightly in pitch until it is directly overhead and then beginning to gradually lower in pitch as it proceeds to progress ahead.
In short, it's not easy and possibly not possible with the editor supplied with MEP.
Either way, it's a lot of work, with or without a dedicated audio mixing program.
Hi Felix.
It depends on where and how the video is filmed and the output format you are exporting to.
Was some of the audio recorded with the video at the point you want to add the effect?
Was it recorded inside or outside?
If the sound was recorded with the video, was it recorded as mono, stereo, binaural stereo, or using a full surround field microphone?
Are you exporting as stereo or as a muliti-channel surround sound output?
For a convincing placement, you need to have the use of a VST reverb plug-in capable of capturing the ambient acoustics of the area you filmed in. These plug-in effects then can 'read' a section of the soundtrack and reproduce the original environment, at which point, if you can record or have access to the recording you wish to add, you can feed that to the effect unit and fine tune the result. The source file should be as sound neutral as possible and close-miked or the effcts unit will not work effectively.
If you don't have access to such a plug-in, have a look on the net to see if there are any free VSTs that will do the job. There used to be one but I can't remember it's name. It was very good though.
I'm going to assume you are mixing to stereo or you would have a dedicated audio editor.
With stereo, it is difficult to have placement at varying heights or to discern whether a sound source originates from behind or in front. Binaural stereo overcomes this with very slight, almost imperceptible timing differences which involves calculating the distances between a person's ears and then works out the time delay, but even then there is no way to determine a sound from being dead center front or dead center rear.
Sound, like light, decreases over distance, so one trick is to gently lower or increase volume and to roll-off higher frequencies as the sound increases in distance, but these are subtle changes and often are not that effective on their own. Also as it is behind and not in shot, the viewer of the video has no idea of the distance involved.
If the shot involves something speeding past and suddenly comming into view then you could try adding a Doppler effect, where the sound, if approaching from behind will increase slightly in pitch until it is directly overhead and then beginning to gradually lower in pitch as it proceeds to progress ahead.
In short, it's not easy and possibly not possible with the editor supplied with MEP.
Either way, it's a lot of work, with or without a dedicated audio mixing program.
That VST reverb plug-in sounds interesting.
Yes, the audio was shot with the video, and the audio is in 2.0ch Stereo.
Some of the clips were recorded inside and some outside. And I will be exporting as 2.0ch Stereo.
Would using the Equalizer do any good? How can I roll off higher frequencies?
I am using MEP 2018 Premium by the way.
Hi Felix.
It the sound is static then you may be in with a chance but the graphic equalizers in MEP are limited in frequency bands so that may be difficult although one I think gives the user a control to sweep the frequencies. That may still be impossible to automate. That I don't know and haven't tried with MEP. I know some parameters can be automated in the mixer in MEP such as volume and placement within the sound stage, but I'm not sure if the frequency sliders in the graphic equalizers can be which would be needed for the Dopler effect.
The VST plug-in I used was some time back and was a free download. Whether it still is on the net or free I can't tell you.
It was something similar to the plug-in used in this video.
Hi Felix.
It the sound is static then you may be in with a chance but the graphic equalizers in MEP are limited in frequency bands so that may be difficult although one I think gives the user a control to sweep the frequencies. That may still be impossible to automate. That I don't know and haven't tried with MEP. I know some parameters can be automated in the mixer in MEP such as volume and placement within the sound stage, but I'm not sure if the frequency sliders in the graphic equalizers can be which would be needed for the Dopler effect.
The VST plug-in I used was some time back and was a free download. Whether it still is on the net or free I can't tell you.
It was something similar to the plug-in used in this video.
Do all of these free VST plug ins allow you to do what you can do in the editor demonstrated in the video above? i.e. can you record and add in your own room tone and match it with the dialogue?
No, which is why you have to search for something similar on the net if one still exists for free.
I think mine came with a computer music magazine on a DVD years ago and subsequently lost as I upgraded computers.
Other free plug-ins stopped working when I switched from a 32-bit to 64-bit operating system. I lost quite a few useful free plug-ins at that time but also you need to check out the automation capabilities of any audio editor. I only do basic audio editing in MEP as I'm more used to using another program for detailed audio editing that allows more options. Even then, I don't alter video audio that much as most of what I do with video is about recording local events.
As CubeAce has said, audio varies infinitely according to where and how it was recorded, so it will always be difficult to exactly capture an authentic soundtrack. There are a few things that it will help to bear in mind, the first of which is that the nearer you are to the sound, the more accurately you will capture it. So close recording a voice for instance, will always give a more intimate and accurate sound of what is being produced, but may not sound the same as people in the same room would hear. This is because other influences come into play as you move further from the source. In a closed quiet room, these influences will be mainly distance and reflections. Your brain automatically focuses on the sound that you want to listen to and compensates for room reflections and distance, but a recording made in the room will not filter out those sounds. If you have an idea of what you want the recording to sound like, you can often get best results by making a recording close to the source, then a second recording of the ambient background sound. The ambient can then be mixed in to the clean sound to whatever level you feel sounds best. Frequently., adding a small amount of reverb to the mix can also help. Another advantage of using the close mic is that the clean sound can be moved across the stereo spectrum to coincide with any movement in the video.
Outside recordings are usually influenced not only by ambient noise, but also reflections from nearby buildings and surfaces. Again the two recordings technique can be quite effective and rather than reverb, adding slight echo effects can help to simulate building reflections. Audio is one of those skills that take a lot of experimentation and experience to get absolutely right, but can be great fun if you have the patience.
A few years ago, I used to prepare radio and DJ jingles using made up scenarios, anywhere from warfare to jungle storms. These were made with a multi layer mix of BBC sound effects CDs, real recordings and home made sound effects. For instance, the sound of a lion roaring can be made by coughing into a mic, then slowing the recording right down. Sound is not necessarily about real time recording, it's about creating something that is believable.
Roger
No, which is why you have to search for something similar on the net if one still exists for free.
I think mine came with a computer music magazine on a DVD years ago and subsequently lost as I upgraded computers.
Other free plug-ins stopped working when I switched from a 32-bit to 64-bit operating system. I lost quite a few useful free plug-ins at that time but also you need to check out the automation capabilities of any audio editor. I only do basic audio editing in MEP as I'm more used to using another program for detailed audio editing that allows more options. Even then, I don't alter video audio that much as most of what I do with video is about recording local events.
From reading through other posts online regarding the same topic, a lot of people say that if the audio soundtrack was recorded in stereo then you should first convert it to Mono. Would you agree with that?
If you mean the track to be added, then yes, or better, just use the left or right channel and pan into position, which would be safer and avoid any phase variance between the two channels that could cause quality problems when combined.
Hi
. . . . a lot of people say that if the audio soundtrack was recorded in stereo then you should first convert it to Mono . . . .
I agree with @CubeAce comments, however if the original audio is a stereo file then the L-R pan may not be as effective in placing the sound object correctly.
Conversion to mono depends on where the audio soundtrack item of interest is located within the sound stage of the stereo audio file ie if it is not dead centre.
The method I use is to mix down to mono into one stereo channel, eg left, and then copy the left channel into the right channel.
Thel stereo audio file with both channels identical then allows correct placement on the audio stage as @CubeAce has mentioned using the Pan control.
HTH
John EB
Hi
. . . . a lot of people say that if the audio soundtrack was recorded in stereo then you should first convert it to Mono . . . .
I agree with @CubeAce comments, however if the original audio is a stereo file then the L-R pan may not be as effective in placing the sound object correctly.
Conversion to mono depends on where the audio soundtrack item of interest is located within the sound stage of the stereo audio file ie if it is not dead centre.
The method I use is to mix down to mono into one stereo channel, eg left, and then copy the left channel into the right channel.
Thel stereo audio file with both channels identical then allows correct placement on the audio stage as @CubeAce has mentioned using the Pan control.
HTH
John EB
Do you know how one can also make the audio sound like it is coming from higher up or lower down?
It's not really possible without additional speaker channels, which is why higher end cinemas went down the multi-speaker approach. Hearing and eyesight are really influenced more by the brains interpretations of signals more than the actual organs responsible for the appropriate sense. Give anyone a new sound within a scenescape with only directional left-right input and they would not know whether to look ahead, up or down. Then try an aircraft or bird noise and instinctively they will look up. Nature versus nurture.
If you are really interested in what can be done with sound positioning, try the results of this google search I did for you. The results will be from early BBC experiments on surround sound positioning with soundfield microphones allowing binaural recordings for radio broadcast. Once you come away from binaural stereo recordings and using headphones for playback to try to produce surround sound, things get more complicated rather than simpler.
https://www.google.com/search?client=firefox-b-d&q=BBC+soundfield+experiments
Hi
. . . . Do you know how one can also make the audio sound like it is coming from higher up or lower down? . . . .
In a nutshell that is not possible. You are dealing with the direction, time and phase relationship of the sound as it hits the ears and this cannot be easily replicated/created from an existing audio file (mono or stereo).
The only way to get this effect to be realistic, is to use a dummy head recording of the actual object approaching from the desired direction, height and speed.
However there is an issue with this - listeners will only perceive the same location direction using headphones.
The introduction of loudspeakers introduces more problems ie ensuring identical speaker positioning eg separation, height, angle, balance, and room acoustics eg echo, all introduce more variables to the phase/time relationship of the audio.
HTH
John EB
Hi Felix.
It depends on where and how the video is filmed and the output format you are exporting to.
Was some of the audio recorded with the video at the point you want to add the effect?
Was it recorded inside or outside?
If the sound was recorded with the video, was it recorded as mono, stereo, binaural stereo, or using a full surround field microphone?
Are you exporting as stereo or as a muliti-channel surround sound output?
For a convincing placement, you need to have the use of a VST reverb plug-in capable of capturing the ambient acoustics of the area you filmed in. These plug-in effects then can 'read' a section of the soundtrack and reproduce the original environment, at which point, if you can record or have access to the recording you wish to add, you can feed that to the effect unit and fine tune the result. The source file should be as sound neutral as possible and close-miked or the effcts unit will not work effectively.
If you don't have access to such a plug-in, have a look on the net to see if there are any free VSTs that will do the job. There used to be one but I can't remember it's name. It was very good though.
I'm going to assume you are mixing to stereo or you would have a dedicated audio editor.
With stereo, it is difficult to have placement at varying heights or to discern whether a sound source originates from behind or in front. Binaural stereo overcomes this with very slight, almost imperceptible timing differences which involves calculating the distances between a person's ears and then works out the time delay, but even then there is no way to determine a sound from being dead center front or dead center rear.
Sound, like light, decreases over distance, so one trick is to gently lower or increase volume and to roll-off higher frequencies as the sound increases in distance, but these are subtle changes and often are not that effective on their own. Also as it is behind and not in shot, the viewer of the video has no idea of the distance involved.
If the shot involves something speeding past and suddenly comming into view then you could try adding a Doppler effect, where the sound, if approaching from behind will increase slightly in pitch until it is directly overhead and then beginning to gradually lower in pitch as it proceeds to progress ahead.
In short, it's not easy and possibly not possible with the editor supplied with MEP.
Either way, it's a lot of work, with or without a dedicated audio mixing program.
How do I roll off higher frequencies?
The method I use is to mix down to mono into one stereo channel, eg left, and then copy the left channel into the right channel.
How exactly do you do this?
With a graphic equalizer. Preferably one that can be automated as the effect will not work as a static effect and has the ability to control narrow, sweeping frequency bands. Personally, I think you are setting yourself an impossible task or at least one wich will require herculean effort for little to no gain.
As both I and John have tried to explain, it's not just a recording problem but a playback problem as well.
Did you read any of the links I put up?
In my room at home, where I'm typing this and do most of my audio or video editing, I have a 500 watt, 5.1, Tannoy Reveal surround system, which came with a dedicated program to run on my PC for 5.1 surround sound to work in my room, which is acoustically treated for listening and not recording (both require different acoustic responses). I have to set up a stereo mic in my listening position and run the program which listens to the speaker outputs one at a time, then all together in set sequences, and then works out volumes and time differences. It even works out the polarity of each speaker and tells me if one of them is wrong. So when I sit in my seat I can hear 5.1 surround encoded sound as it should be heard. If I move my listening position slightly it no longer works as well. The illusion is not as good. Bear in mind that even if you get it to work for you, your audience may not hear it as you heard it. Try transferring the experience to another location and it may not work at all. Even in an ideal environment for audience participation, the sweet spot for those able to glean a favorable listening position would be a handful at best.
For people that work in this field of audio manipulation for video and film, there are two main programs used by major studios. They are Pro Tools and Nuendo. They contain most of the tools for successful basic surround sound manipulation at this level of manipulation for audio and visual work. The fact that there are only a handful of such studios across the world exist, shows how difficult this side of audio is to get right, and why such recording studios come at a premium cost.
While none of the above is absolutely necessary, it is certainly very desirable and makes such tasks for those involved somewhat easier.
This is not something I would consider doing myself and I have done some limited live theatrical work in this field in the dim and distant past.
With a graphic equalizer. Preferably one that can be automated as the effect will not work as a static effect and has the ability to control narrow, sweeping frequency bands. Personally, I think you are setting yourself an impossible task or at least one wich will require herculean effort for little to no gain.
As both I and John have tried to explain, it's not just a recording problem but a playback problem as well.
Did you read any of the links I put up?
In my room at home, where I'm typing this and do most of my audio or video editing, I have a 500 watt, 5.1, Tannoy Reveal surround system, which came with a dedicated program to run on my PC for 5.1 surround sound to work in my room, which is acoustically treated for listening and not recording (both require different acoustic responses). I have to set up a stereo mic in my listening position and run the program which listens to the speaker outputs one at a time, then all together in set sequences, and then works out volumes and time differences. It even works out the polarity of each speaker and tells me if one of them is wrong. So when I sit in my seat I can hear 5.1 surround encoded sound as it should be heard. If I move my listening position slightly it no longer works as well. The illusion is not as good. Bear in mind that even if you get it to work for you, your audience may not hear it as you heard it. Try transferring the experience to another location and it may not work at all. Even in an ideal environment for audience participation, the sweet spot for those able to glean a favorable listening position would be a handful at best.
For people that work in this field of audio manipulation for video and film, there are two main programs used by major studios. They are Pro Tools and Nuendo. They contain most of the tools for successful basic surround sound manipulation at this level of manipulation for audio and visual work. The fact that there are only a handful of such studios across the world exist, shows how difficult this side of audio is to get right, and why such recording studios come at a premium cost.
While none of the above is absolutely necessary, it is certainly very desirable and makes such tasks for those involved somewhat easier.
This is not something I would consider doing myself and I have done some limited live theatrical work in this field in the dim and distant past.
Thank you.
By the way, may I ask why it is that you should first change the soundtrack to Mono? And if the source soundtrack is recorded in 2ch Stereo, then do you still need to change sound effects from Stereo to Mono?
Also, when panning sounds either to the far left or right, I find that it sounds rather strong or overpowering in whichever ear the sound is entering (left or right earphone). Is this normal or am I missing something?
First.
You don't have to, but is often desirable to reduce possible confusing/contradicting ambient sound information within the mix. Equally, you could just pan both channels to the desired locations within the mix if the effect needs to be heard 'wider'.
Second.
What you are hearing is normal and again, you should mix to taste. If the volume needs reducing as it pans then that is what needs doing, hence the need for the mixer to be able to be automated so such actions can be recorded. Which becomes even easier if you have a physical mixing desk or midi capable controller which can be 'read' and the physical movements of controls recorded while mixing down. No matter if you can't MEPs audio mixer at least can be automated a bit. How much I haven't really expanded on but do know you can draw in curves for volume and panning at least. More than that I haven't tried. Normally because I use Cubase as my audio mixer when things get complicated.
I admire your perseverance. I can only presume this is a project of importance to you. Possibly for an educational qualification?
Hi Ray,
Just remembered that I wanted to comment on some of this.
In MEP's mixer, you can automate Volume and Panning:
On an object, go to Effects, Audio Effects, General, and you have Volume, Panorama, Aux1, Aux 2, Sharp filter, Smooth filter and distortion. Place keyframes as normal. Turn on the curve view in the Arranger by clicking on the Eye button beside the effect in the keyframe area.
Object volume curve can also be turned on by right-clicking and selecting Volume Curve.
For Objects, points can be added and modified directly on the curve in the object on the Arranger, but only deleted by using the keyframe tools.
For the Track, points can be moved, but it's a bit difficult, especially if there are a lot of points close together. They cannot be deleted. The best is to start playback and move the volume slider or panning knob to change the curve.
Here is a messy illustration. The colours of the curves can be changed in the keyframe area.
As you can see, controlling audio is fairly well covered in MEP. There are many more possibilities with effects on tracks and objects, including plugins.
John CB
Hi John.
I did mention in my reply of [ 09/05/2019, 22:55 ] that some of MEPs parameters can be automated but it is complicated compared to say Cubase, where I can correct any automaton much easier, and MEPs automation is very limited for such a complex mix as this would need. You can also delete or add extra control points on a drawn curve or even select a curve tool, whereas as you say, you may not be able to delete recorded automation points unless you undo the action within MEP, but you can move those points around in both time and relative position (volume or pan).
I can't see any way to automate plug-in controls though, beyond the gain feed of the aux sends.
In a dedicated audio editor or complete video editor that specialises in automated multitrack sound mixdowns such as Nuendo, You can even loop sections of a mix to automated and updated commands on each pass and select predetermined drop-in and out points. The shuttle action is also somewhat smoother than MEPs and more controllable with a mouse.
I don't know if Vegas or ProX is any better in this respect.
While I find MEPs sound mixdown facilities are useful for basic editing and there is nothing wrong with the quality of the sound output, and for normal use is more than adequate, I would find it rather limiting for complex sound sculpting. For instance, you can't group faders to work together within MEP, there are only two auxiliary sends and there is no facility for sub monitoring. In Cubase, I can use my midi controller sliders and rotory controls to automate more than one parameter at a time and assign different controls to different plugins and groups. That in itself is invaluable if you need to hear how two or more effects interact with each other.
Of course, on day to day normal use, this is way above what an average user would need and possibly an overcomplication of what MEP needs to be able to do.
[EDIT]
I will say that within MEPs automation I do find it strange there is no erase tool to erase control points. It would make some of the audio work within MEP easier.
Actually, I will see if it is possible in the Key Frame Editor.
The audio features in MEP/VPX are not as sophisticated as a DAW. For anything complicated, I use Samplitude Pro X3.
However, the audio features are far more extensive than most users are aware of, as far as I can tell. Remember that individual objects or mixed down objects can be sent out to an external editor, which then gives many more possibilities to the user.
For object automation, I failed to mention that you can change the mouse mode to Curve, select the audio effect in the keyframe area, and draw the automation curve on the object in the timeline using the mouse.
I have not found a way to do this on the track curve. I will check, but I think that it's the same thing in Samplitude.
John CB
Hi Ray
. . . . you may not be able to delete recorded automation points . . . .
You can add and delete nodes on the automation curve - it is all in the double click - see clip below. The tricky part for deleting is getting the pointer to the correct location on the node - AFAICS this is the left side of the node, miss it and you open the Audio Cleaning dialog!
HTH
John EB
Thanks for that John. I'll give it a try. It still looks like a lot of work compared to using an erase tool but at least it means not leaving MEP. Steady curves are really hard to achieve. What I do like is the move control where you can drag an audio automation point across the timeline and watch the movie play to the point where it has been taken to. That is useful. There does seem to be a limit of seven parameters you can automate. Unless there are others around I haven't found.
As for accidentally bringing up the audio cleaning dialogue, I not only have the T shirt but I'm in dialogue for a possible TV series.
The only audio automation I've done inside MEP so far was for a 1940s weekend video to pan the sound of a Spitfire that was going overhead and away to going from left to right. For the most part I try to keep my video audio as simple as possible.
Hi John EB,
Ah, I've got it! Make sure that the thumb end of the hand is on the node or line, then Double-click to delete a node or add a node. This is only for track automation points.
Hi Ray,
See my last post about the Curve mouse mode.
Yes, as far as I can tell, there are only the 7 parameters available for object automation. The 2 Aux parameters allow object automation on the volume that is sent from the object to one of the FX tracks in the Mixer. This allows you to have multiple effects per FX track, but the automation is per object on that track, so per group of effects if more than 1. Are you able to understand the diagram below?
I have reverb on the first FX strip of the Mixer, but I could have more, including plugins.
Hi John. Yes I understand, but my earlier point was that you then can't automate the plugin to say, change the reverb time from short to long or change the curve of a graphic equaliser over the timeline. Nor can you run that many multiple versions of the same effect using different settings.
That these extremes are seldom needed except when doing experimental foley work is a good reason not to overcomplicate software unnecessarily.
The other problem for the initial inquiry may be that even if the sound made is accurate, it may not sound correct in the video. Sometimes a sound that you need to be recognisable has to be made in an unorthodox way so that the viewer's brain accepts it as being what they expected to hear and not always what should have been produced.