Audio Test & Post Workflow

[trueindie]

I'm working on a feature film, and I'm doing the Audio Post myself. I'm brand new to Audio Post and this is the first time I'm attempting audio post at this level. I've only been using pro-tools for about a month, and Izotope RX for about a week, and still don't know any short cuts in either software.

Here is the audio (and video) test.
https://vimeo.com/117028317
password: audiotest1

It's obviously not good enough. Just my first tepid attempt.
Please excuse the changing color schemes. I'm still experimenting.

For you audio experts:
I could use some advice on post production workflow. As you can see above, I took a little bit of my edit to practice the audio process today as I continue the edit process, and this is what I did:

1. Exported OMF
2. Exported Video
3. In Pro Tools:
- Separated Dialogs in two tracks for different characters
- Separated ambient noise such as chair creaks and footsteps into two separate FX tracks
- Separated music to music tracks

Then when I was going to start noise reduction, I realized that I have a bunch of broken up tracks, so I would have to do NR on each separate portion....

. So Now I feel stuck.

So now I'm beginning to think that I should change my workflow. I should
1. First do NR/DeClick/Dehum
2. Separate DIAL/FX/MUS
3. Create DIAL / M&E
4. Create Combined Track
5. Level Combined Track
6. Export back to NLE


Some other general inquiries from a baby sound person.

A. When I export to OMF am I supposed to do Broadcast Wav, or AIFF
B. Why Interleaved option?
C. Around minute 00:29 and 1:29 (among other places), you can hear the microphone away from the guy as his pitch shifts. How can I correct this? Also at 1:46, suddenly the audio sounds different. Will just lowering the volume minimize the apparent difference? Or is this the right time to learn about EQ Match? Seems like overkill in this case?
D. Around minute 2:02 (if you can ignore the crackle), how can I make the train noise more menacing? Exactly what should I do with the EQ? or is there some other method?
E. When I level dialog, after NR, it raises the noise floor in those areas. Am I supposed to do a second NR run after leveling? How to solve this problem?
F. Is there a theory on how loud the music should be in a movie with respect to other sounds?
G. Any other general advice for an Audio Post Baby

I will appreciate all comments, of any nature.

Much thanks for your time.
aveek

 

[ryanjf]

Hi Aveek,

Nice film!

Typically, for live action films, the dialogue tracks are usually separated by angle, which means same mic tonality/quality. That way the mixer can just process similar regions/clips on the same track. It's less work for the person that is going to mix the film for you, and if you're mixing it, you'll find out SOON enough why you want to group similar mics on the same track. Splitting it out by character is possible but not necessary, it just leads to extra work. However, if two characters are captured on the same mic and are on the same region/clip and they have different sonic signatures, then it might be worth it to split the characters onto different tracks so they can be EQ'd/Compressed/Noise Reduced differently. It all depends on what works for that scene, and what makes the mixer's life easier.

With that said, if it's animation you're working on... then it's obvious that splitting the tracks by character makes the most sense.

You've gotta read this: https://www.gearslutz.com/board/pos...9-dialogue-pre-dub-splits-each-character.html

I would separate your chair squeaks and footsteps into PFX (production fx) tracks, which can be used for the MnE. For low budget stuff, it doesn't really matter where you put this (it can live in the dialogue tracks). It's just nice to get into the habit of doing so.

On to noise reduction... assuming you took my advice above (by sorting the dialogue tracks by same angle/mic tonality) ... you are now left with tracks that have similar noise profiles! You can then just highlight the entire track and do a "batch" NR (whatever that might mean to you) on the entire track! Or you can put Inserts in Pro Tools that can process that track in real time, and since all of those clips/regions have the same noise "profile" or tonality, it will be applied across all of those regions.

On the other hand, now imagine if you had one track that had many angles/mic tonalities on it. Imagine trying to apply a realtime noise reduction insert on this track.. You'd have to automate each region differently. MUCH MORE WORK.

I guess my overall point is: make the mixer's life easier, even if that's you. Organize and edit the tracks in a way that leads to minimal confusion. The mixer has a 1000 other tracks to deal with, the last thing he wants to do is jump between tracks and copy and paste automation between them all because of poor organization and editing.

A) OMF/AAF exports, I prefer .wav.

B) For me, I never do interleaved though that might change in the near future..

E) that's tough, if the noise is really bad, then ADR. I do multiple passes of NR, and if I can't get the dialogue smooth enough, I request to ADR that line.

hope this helps!

Ryan

 

[Alex H.]

3. In Pro Tools:
- Separated Dialogs in two tracks for different characters
- Separated ambient noise such as chair creaks and footsteps into two separate FX tracks
- Separated music to music tracks

Most of this can (and should) be done in the edit. The bane of my existance is receiving a session from an editor who lets audio tracks fall wherever they will, paying no mind to target tracks. I can waste hours just trying to reorganize tracks to a point that I can finally start organizing them the way I need them.

Save yourslf some time in audio post (or if you hire a sound editor, save him/her some time and you some money). Drop VO/narration on A1. Use A2, A3, and A4 to checkerboard your dialog. A5 and A6 can be used for mono FX. A7-8 and A9-10 for stereo FX and ambient sound beds, and A11-12 and A13-14 for music. You can use more tracks for each section as needed, but you get the idea. In ProTools, the track count will most certainly increase, but at least you're starting with something that makes sense.

As for "Separated ambient noise such as chair creaks and footsteps into two separate FX tracks," two things: one, chair creaks and footsteps are not ambient sounds but are instead FX; two, do you mean (as it sounds that you did) that you took these out of the original production sound? These are sounds that are replaced in movies.

A. When I export to OMF am I supposed to do Broadcast Wav, or AIFF
B. Why Interleaved option?

BWF and AIFF are both uncompressed files, but BWF can preserve metadata. Interleaved tags 2-channel sound sources as audio pairs. I tend not to used interleaved in OMF/AAF export, though I also use only one channel of dialog recording per source.

C. Around minute 00:29 and 1:29 (among other places), you can hear the microphone away from the guy as his pitch shifts. How can I correct this? Also at 1:46, suddenly the audio sounds different. Will just lowering the volume minimize the apparent difference? Or is this the right time to learn about EQ Match? Seems like overkill in this case?

This is bad boom work, and the differences in room reflections are there to stay. Very difficult to fix without ADR or a reshoot. Only one of those is generally possible, and niether are without expense and time investment.

D. Around minute 2:02 (if you can ignore the crackle), how can I make the train noise more menacing? Exactly what should I do with the EQ? or is there some other method?

It needs more creaking/screeching metal and more low-end rumble. These are not things that are fixed with EQ. Again, this is a sound that is replaced, not used from production sound. Rarely is something that you hear in a film comprised of one, real-world recording of that sound. It requires creating that sound from other sources that may not even originate with the sme kind of object. You may be looking at 6 tracks layered together to create that sound. In reality, on a big budget, it may be a couple dozen tracks (or more). Think about the different elements that exist there, and then figure out how to create each one individually. There's the sound of the subway station (stereo room ambience), the blast of air as the train breaches the tunnel, the wheels on the track, possibly some sublte crackle and spark from the third rail, the screech of the wheels and brakes, the creaking of the train car chassis... what else do you hear, or want to hear?

E. When I level dialog, after NR, it raises the noise floor in those areas. Am I supposed to do a second NR run after leveling? How to solve this problem?

What do you mean by "level dialog"? What process are you using for this?

F. Is there a theory on how loud the music should be in a movie with respect to other sounds?

Season to taste. If it's under dialog, it shouldn't step on the dialog. If it is the featured sound element, it should be up front in the mix.

G. Any other general advice for an Audio Post Baby

You have to create the sound of your film's world, and a lot of this happens with replaced audio both in ambient sound beds and in FX. There's a lot missing in your sample. For instance, in the tavern scene at 4:12, where is the rest of the world? I hear only the dialog as they talk about the guy sitting in the corner. Where's the ambient sound? Where's the music playing on the jukebox? Where's the distant sound of the street outside (we even see reflected light from a passing car)? From an audio standpoint, that scene is completely naked.

I suggest you read THIS post from the Audio 101 sticky to get a bit more perspective on the audio-post process. It's an intense process to do it properly, and not for the faint of heart.

 

[trueindie]

Thank you guys for the replies and the links. Truly appreciated.

To further clarify, in the test session you guys saw:
I was using boom audio only. But I do have access to lav audios as well. So maybe in the areas where boom audio isn't perfect, and there are noticeable pitch shifts, maybe I'll try to mix in the lav audio as well see how it works.

When I was breaking up the audio, I wasn't really thinking about mic angles or mic sources. I watched a tutorial where this guy was checkerboarding the dialog tracks by subject, and so I was doing the same just to get into the habit. But in the future, I'll try to be more purposeful in my checkerboarding.

I did noise reduction through RX. My understanding was that RX does NR by sampling surrounding audio, so even if I'm using different mic sources with different noise profiles or tonality, it shouldn't matter, as it's not using the same noise image to perform NR for the whole region, but using some sort of adaptive surrounding noise image to do NR. Am I wrong in my understanding? But yeah, as I go along and experiment, and learn, I'm thinking of pushing NR into a later stage of the process.

And to answer your question Alex, yes, I am using the production sound to create the PFX track, for M&E purposes.

For leveling, I used the leveler in RX4. I just set the leveling to a target RMS and a target dynamic range. And then I went through the audio again and adjusted the levels as I felt was necessary in the dialog. So since I had done NR first, and then one of the subjects speaks really softly, when I raise the gain for her, I begin to hear more noise when she speaks.

I'm beginning to realize that I need to create the sound of the film's world as you put it Alex. But I'm petrified, as I get into audio. I spent a lot of time using the spectral repair in RX taking out the outside beeps of cars backing up, and I was happy that I couldn't hear the outside noise. But of course you're right, that without the outside noise, the audio seems naked. But I lack experience in this and I'm more scared that if I add ambient noise, I'll do it wrong. Another Audio person mentioned how flat and boring the audio was, and I agree completely. I have to get a bit more adventurous with audio.

Any recommendations on a relatively inexpensive sound library, for filmmakers that will help me with the chair noises or footsteps or production sound replacements?

Thank you guys so much for your time
Aveek
ps. Ryan, thanks for saying 'nice film,' I'm glad you liked the visuals at least

 

[Noiz2]

It could be a nice film, no zombies (yet) ;~)

Your work flow is OK, I would stick with WAV. Pro Tools just started to be able to use interleaved stereo and OMF is a mono format so I would generally stick with mono for sending to Pro Tools.

Big budget films separate out dialog tracks and pull out all the PFX and then fill the holes with Room Tone. I find that a waste of time on 90% of low budget projects. It's done on big budget films for a number of reasons that makes sense for them but not you.

Also why are you doing an M&E? The only reason to do an M&E is for doing a foreign language dub and that is close to never done for a low budget feature. So unless you have a distributor that is insisting on a M&E AND they are willing to pay for it I would skip the pain.

If you do an M&E then you also need to do a Foley pass to cover all of those things that also go away with the production tracks. It's a lot of work for nothing generally on low budget projects.

So I separate production tracks by problem. That will often be by character IF you are using lavs. If it's boom audio it is more likely to be different takes, which again is often characters when you are doing one shots and reverses. I will checker board cuts on two tracks and then pull out problems as they come up. I don't over clean the production. NR I only use when there is a problem, I don't do it to all tracks always.
You don't want it to come out too sterile.

Once you add the "world" a lot of the small NR issues will get eaten up by the BG's (background sounds). I cut out anything that shouldn't be in my "world" or is stepping on a line (if that can be done), but I don't do a lot after it sound decent till I have laid in some sound. I also don't necessarily fill holes in production with RT because in a lot of cases it will just be added noise that I will mix out later.

The BIG difference between a low budget workflow and a big budget workflow is that they have a lot of money and people but very little time and you have the opposite. So in a big budget film you cover everything and Foley will cover a bunch of stuff and ADR will shoot a LOT of the script and the composer will score it all to death and in the end the mixers get to sort it all out. Because it's faster that way and they are paying $$$$$ in completion bonds every day the film is not finished.

On the other hand with the exception of Sundance fever (where the film was shot late and the picture edit pushed even farther and sound NEEDS to do it all in a few days because the director just KNOWS he is going to will the Lotto at Sundance) you have as much time as you need. That means you can be MUCH more efficient than a big budget film can.

So I rough in the dialog and figure out if there is anything that can't be saved and needs ADR, and then I go to the SFX side till that is filled in. IF I can get the score by the time I start SFX I note where the music is going to be the star and I seriously back off on BG's and such that I will need to loose anyway. Hard FX and Foley generally work a lot better with the music.

Then I go back to the production and see what now needs fixing, usually not much.

I also generally separate PFX when they are in the clear and not when they overlap dialog. On a big budget you would be going through all the alt takes looking for something to swap in or mark it for ADR since they want completely clean dialog tracks.


RE: Rx. I don't use it as a track plugin, though plenty do. I started doing something a friend mentioned and he heard it from some other editor... I dup all the tracks that come in on the OMF, disable the tracks and hide them. At that point I can do destructive edits as in NR and always have the option to go back easily to the original.

Unless there was a big change in the latest version Rx does noise print based noise reduction. So you need to identify what you consider noise and have RX sample that and then have it work on the larger clip. Even as a track plug in you need to make a noise print so it knows what to remove.

NR works by subtracting what you have defined as noise from the rest of the signal, it doesn't fill in anything. Spectral Repair uses sounds from around what you want to remove to generate replacement sound, nut not noise reduction.

On the film.

You have two spots when our lead is doing things and there is only music and no other sound. Since we start out with the "you have no song" thing this seems weird. The first one I was O' this is her song, but then later there is a different song. IF this isn't intentionally trying to confuse the no song issue then you need to not have "music video" moments like that. If those are both score then they should not be the sonic stars of the scene.

Low budget filmmakers have a tendency to slather music over scenes as a way to hide that they didn't do any sound work on the scene. But that is really a mistake on a few levels. First unless it's a dream sequence or some other "out of body" experience it screams "no budget film". And second you lose a great opportunity to tell some story. The "world" and what's in it, and what isn't, tells us a LOT about where we are where the character is and adds all kinds of story info that doesn't come from the dialog or visuals.

The bar for instance.
We don't know anything about the bar, but we could. You could make it a classy place or a dive bar. Are there a bunch of customers or is it dead. Is she working in a good job in a popular place or is she scraping by in a dump that has so few customers that she is constantly worried about loosing her job when the place closes???

Her apartment.

Where is it? Is her place OK because she cleaned it up but it is actually in a low rent building with neighbors yelling at each other, or is she in a swank place that costs a lot?

Mix that with the Bar and we could know a LOT about our character and her possible problems with out any new dialog and no inserts.

From your post I'm making an assumption that you are the picture and sound editor and probably also the director and maybe the writer. If anf of that is true then you have a big hurtle to get over and that is you need to see the film as your audience will. You have a world that the film takes place in because you know it so well. Your audience on the other hand is coming in cold. They have no idea if it's going to be a love story, a slasher film or maybe she gets abducted by aliens in the next shot??? So you have to help us. Tell us what we need to know so we can feel the world the film takes place in. Knowing the world we will have expectations and you can then use that to lead us along or surprise us. Get some pretty birds in that park and quiet city roar in the BG and it's all bucolic and... We are ready for that to continue so it's easy to lead us that way or we will be really surprised when the slasher runs out and hacks her to pieces. ;~)

And don't be too scared. Basically stick in what you know is in the world of the film. What ever you want us to hear or know at any given point, stick it in. Lots of specifics tend to not get too cluttery but lots of BG's do. After you have gone crazy with it you can start subtracting things that are not quite right or doesn't work. It will feel right when you get there. Other than marking where big music moments are I don't edit with music playing. It'a very easy to miss stuff or get lazy if the music is playing. Really your film should be able to play and feel finished with the music off.

 

[Noiz2]

...

Any recommendations on a relatively inexpensive sound library, for filmmakers that will help me with the chair noises or footsteps or production sound replacements?

There are a couple that have been mentioned in any number of threads. You can do a search on Sound FX. There is a current thread that lists a couple and I bet there is a sticky.

Read the sticky Alex linked. You are generally not looking for "a sound" but a number of sounds you will cut up, layer and build the sound that you need with.

Often the best and cheapest sounds are the ones you record yourself. Certainly anything like chair sounds will fit better. Some sounds you can't record because there is nothing around you can affordably get to to record but most stuff in your film will be available.

 

[DPStewart]

Wow.
what a thread.
hey, trueindie - this is all spectacular advice. It's incredible what these pros are willing to share with you for free right here!
it took me years of working all aspects of production to learn what these guys just mapped out here in this thread.
i can assure you that it is all solid advice.
this forum blows my mind sometimes.

~cheers gents

 

[ryanjf]

For leveling, I used the leveler in RX4. I just set the leveling to a target RMS and a target dynamic range. And then I went through the audio again and adjusted the levels as I felt was necessary in the dialog. So since I had done NR first, and then one of the subjects speaks really softly, when I raise the gain for her, I begin to hear more noise when she speaks.

I know it's very tempting to use the Leveler module in RX4, but I'd advise against that since it "bakes" the clip gain permanently into the audio when you render it back into the Pro Tools timeline, especially if you're dealing with lots of noise to begin with. You're much better off "riding the faders" for a more natural feel and approach. If your project has a quick schedule, try to use the Dialogue Denoiser as an Insert in PT. I've gotten great results with it and it is more tuned to dialogue versus the regular Denoiser module. If you have a longer schedule, then I'd spend the time to AudioSuite the NR, if needed. Once your audio is as clean and smooth as possible (by dialogue editing techniques such as building room tone bridges, noise reduction, etc), then you can start mixing with faders (again, not with the Leveler, in my opinion).

When I've used the Leveler module, the noise floor does tend to rise after words and sentences (depending on the "Speed" parameter), and so do breaths. Yes, I can put a "Debreath" plugin to eliminate that, but that's just more things to deal with. I think it's much easier to ride faders instead of having a program tell you how dynamic your dialogue should be.

As for RX's noise reduction process, it has to "Learn" what the noise is in order for RX to reduce the noise (for the Denoiser modules at least). When I use the Denoiser, i copy the audio region that needs NR onto a "Work" track and I pull out all the handles so I can see all the material before and after the region. There SHOULD be some "silence" or longer passages of "noise" which contains the noise profile. I highlight this "silence" ... open Denoiser as an AudioSuite plugin and click "Learn". This will take in what you just highlighted and subtract it from the region you're trying to NR. Remember to always keep copies of the original, non-NR'd region on an X Y or Z track before noise reducing anything.

Spectral repair - i've used this to fill in or replace pauses, or unwanted beeps in the audio, or even remove bird chirps (from sound effects recordings) and dog barks. The spectral repair module "samples" surrounding audio and fills in the things that you're trying to take out.

In regards to splitting or breaking up the audio in your timeline, it really depends on the scene and what works best for the mix.

 

[trueindie]

Hey guys,
Thanks very, very, very, very much for the additional info and perspectives.

I just wanted to put my situation a bit in context, since you guys have been so generous with your time and advice.

I normally make short films. low budget. for youtube. I pay minimial attention to sound. I worked with one pretty good and one relatively good sound guy in the past and they'd generally send me a noise reduced file that I'd work on and then not worry about, and would do no post audio work beyond whatever gain adjustments I had to do in my NLE.

This movie is a feature film, and I spent a bit of my hard earned money. The original plan was to release it online, and not even bother with the festivals (as I hate them as a matter of principle). But I had decided that audio on this was going to be markedly better than any of my other products so I had purchased Pro Tools and RX.

Recently, after talking to an employee for a Broadcast company in Canada, I realized that they're looking for Canadian content (they're specified by law to show a certain amount of Canadian content, and most content on TV is American), and she wanted me to send over a trailer when I got the chance. The thing about Canadian TV also is that if you can somehow make the product both for French and English languages, it is a better product, as it then has a wider market. That is why I decided to do M&E, to make it a superior product for Canadian TV. But Scott, I understand that that may be difficult to achieve without proper Foley and other resources.

So my goal with this film is not a theatrical release, or even an online release, but to have a Canadian broadcaster purchase it. As a result, I am now more interested that the audio adheres to the CALM act, than has the dynamic range nuanced for a theatrical release, and so I purchased Izotope Insight as well.

This is where I'm coming from. And in many ways, I'm blind. Actually I'm completely blind. Because I've never done anything quite like this before where it is so easy to make mistakes. I'm watching tutorials, asking questions everywhere I can, and trying to figure out what to do by myself, as wherever I've asked, post audio costs seem prohibitive for someone like me. The only thing I've got going for me is time. There is no producer breathing down my neck. I did write, direct and finance the film also. I'm giving myself until the end of March to finish this project.

Scott: It is absolutely tragic (and I mean it with absolutely no disrespect but with a heavy heart) that you felt that Bach music montage sequence was a low budget music slathering to hide the lack of sound work. If you go to page 11 of the script, which can be found here: http://thegirlwithoutasong.com/wp/w.../TheGirlWithoutASong-fd8-BLUE-Script-Full.pdf, you will find that the montage sequence is written into the script, with the music that is actually playing in the current edit, also specified in the script. In other words, the sequence was planned, and not a mistake corrected. I'm mortified that the scene is coming off that way . Because I know exactly what you're talking about, and those things, typically with audio, that happens in every low budget indie film that I watch, are the very things that I really strived very hard to avoid on this project.

Most of your other questions on the film are answered in the film, but not in those 6 minutes you saw unfortunately .

Ryan: Thanks for the insight into the Leveler module. I do use the dialog denoiser only. I haven't had the need, on this film anyway, to use the other tab. None of the audio I have requires any heavy denoising. I almost feel, after denoising that I want a little noise sometimes . I've debreathed, with Spectral Repair, and it always leaves me feeling ambivalent. I like breaths because they seem natural, but once I remove them, I like the debreathed audio also, and then I don't know what to do

The reason I'm interested in the Leveler module, in learning about leveling in general, is because I'm interested in making sure that my film adheres to the CALM act for North American broadcast standards. I don't know much about it, and am doing some reading on it, and playing around with the leveler module just to see what happens when I play around with the variables.

This is where I am currently as I approach audio for this film.

 

[ryanjf]

Ah! Now that all makes perfect sense now.... I'm glad you mentioned you're trying to hit a broadcast standard.

Since that's the case:

Do NOT use Leveler on the dialogue regions. Instead, you will use Loudness (in RX) for the entire composite mix (music, dialog, fx, Foley).

I do something similar to the following every single week at work:

- Import OMF/AAF
- Split tracks into their appropriate departments (dialog, music, fx)
- Edit dialogue, music, fx (do NOT use Leveler on any of this)
- Call the broadcaster and see what their audio specs are (they will most likely provide a True Peak and Integrated Loundness value)
- MIX. When you are mixing, put iZotope Insight on your MASTER bus (or an AUX bus... more on this in a bit). As you are mixing, LOOK at iZotope Insight and see if your mix is NEARING the requested broadcast standards for audio. (a typical spec for the entire mix (dialog, music, fx) is to hit -24LKFS) You DO NOT have to hit your spec. If you are off by a few dB, it is OK.
- Assuming you are mixing STEREO -- when you are satisfied with your mix, export a stereo .wav file.
- Open iZotope RX4 standalone and open your newly exported mix .wav file
- Open the "Loudness" module/tab
- Adjust the sliders on the right to match your broadcaster's audio specs (see screen grab)
- Click Process
- Now your entire mix is compliant with the broadcaster's audio specs!

This prevents you from using Leveler on the dialogue audio regions. The above results in a more natural sounding mix because your dialogue levels will vary instead of iZotope forcing all dialogue to hit a certain RMS level - that is a BORING mix. We love dynamics, so create them by using the Pro Tools faders!

When mixing, I have iZotope Insight on an Aux track, which is then routed to another audio track to "print" my final mix in real time, that is "watching" the levels of my entire mix, whether it be stereo or 5.1. However, I do not let Insight "control" the way I mix. I just use it as a guide. Of course, in order to make a mix exciting to listen to, you're going to sometimes exceed your broadcast's audio specs in terms of levels and true peak values. But that's ok. Do NOT mix by numbers. Use your ears. Make your mix as exciting/cool/dynamic as you want. Use Insight as a general guide.

Remember, the Loudness module in RX (stereo mixes only) or Dolby's DP600 or Nugen's LM Correct (http://www.nugenaudio.com/lm-correct-loudness-correction-automatic-quick-fix-plugin-aax-au-vst_19) will automatically meet your broadcast spec. You put in the true peak and "Integrated" loudness in the mentioned loudness correction tools and it'll do all the work for you. You should still be mindful of where the levels of your mix are at when you are mixing though. I have a spec of -24LKFS (plus or minus 2dB LKFS) to hit, but sometimes I go above -21LKFS for the louder action sequences, of course! How boring would a mix be if your action/battle sequences are at the same level as a conversation between two people sitting in a quiet room.

When you mix, run pink noise through your speakers. Pink noise should be playing out of one speaker at a time. Get a SPL meter and turn up the volume on the back of the speaker (or amp, etc.) so that the measure level is between 79-81dBSPL on each speaker, whatever is comfortable to you. I set each of my speakers at 80dBSPL as I find 79dBSPL sort of quiet at times. I mentioned this in another thread, but I know my room so well that when I mix (without looking at Insight) my dialogue at 80dBSPL, it actually is right around -24LKFS on Insight! You'll get to this point when you get more accustomed to your room.

 

[trueindie]

Oh my God Ryan. You're the best. The absolute greatest, my friend.

I just got a couple of M-Audio BX5s, and an SPL meter. Just to give you an example of how green I am with this whole audio thing, it was only after I got them that I realized that I couldn't connect them directly to my computer. So now I just ordered a Fast Track Pro and a couple of TRS cables.:grin:

But Ryan, thank you, thank you, thank you. I'm floored. I can't express how grateful I am.
Aveek

 

[ryanjf]

That's great! Have fun with it! Oh and here's a download link to the pink noise file:

http://www.thedubstage.com/pink-noise/

NEVER use the pro tools pink noise in the Signal Generator plugin!!! Never.

 

[trueindie]

Ok, I won't!!
Thanks again.

 

[Noiz2]

Don't get too worried about the "tragedy", I didn't mean to say that I thought you were "slathering" music to cover lack of post work, you said it was an early pass, but that it was a common low budget thing to do and to watch out about doing it. I kind of assumed that you were going to add some sounds in later. Structurally though I would do some test screenings to see if others are confused about the "no song" and then song montage (assuming that the montage is very music forward). If it's not just me you may want to think of a way to fix that.

There is the old, but true, saying that there are three films, the one you wrote, the one you shot, and the one you end up with. They are related but almost never the same.

Two short cases related. A feature I worked on had these black screens that separated "chapters" in the story. I had put in lead in sounds to kind of start the next section with out so much of a lag. The picture editor really wanted them to be silent and the director and I went along. I didn't think it worked as well but on a small screen it was OK. Then they screened it in a theatre... You felt like you could have gone out and gotten popcorn in the gaps.

The relation is carful about sticking with something, check it out and see if it works for the audience/ format if at all possible.

That was low budget, but on a MUCH bigger budget film there was a similar issue where familiarity really weakened a moment.
In the big ship disaster film there is a scene where there are all these frozen bodies floating in the Atlantic. That shot started a A/B reel so in editing and mixing it always started from a dark quiet place, and it was devastating. The picture editor wouldn't even stay in the theatre when it was playing. When I saw a screening with an audience I was waiting for that moment. And it was not a big deal. Why? Well we hadn't really screened it adequately from reel to reel, we knew how impactful it was so you discount the zillionth time you see it that you are not quite as hit as you were before, and there would have been a fight if they had added time back in to a very long film. The scene before is the final sinking and it's loud and dramatic and the audience is worked up. By the time they calm down and are settled for the next beat most of the panorama of bodies has passed by and we are back in to dialog. If that had been given maybe 30 seconds of breathing space they would have been selling Kleenex at concessions.

Both are really cases of we knowing a lot more than the audience and making decisions based on how WE perceive the film as opposed to how the audience will see the film.

Again not saying you need to change anything, just a reminder that once your not doing films just for you that it becomes important how others perceive the film, and sometimes you need to change things so that they get the perception you want them to have even if you don't need that change.

Anyway great news on the broadcast possibility. And BIG points for taking the time to polish things up and to do a real sound post.

eMail me if you need some sounds or?

Ryan is absolutely right about the built in pink. I believe the pink he pointed to is the ones from Blue Sky ,they and Dolby both have accurate Pink files.

 

[Noiz2]

Also You should probably get on the DUC (Digi Users Conference) there is a postproduction section. The one thing about ProTools is at some point you are going to hit a roadblock, some weird thing that just won't ...? Or strange crash or need to do something Digi is not happy about or??? And the DUC is the place to go. Someone there has already had exactly your problem and has already gone down the rabbit hole and either has an answer or a work around.

 

[trueindie]

Don't get too worried about the "tragedy",

Haha!. I appreciate the critique actually. I can see what you're talking about, and still do something about it. Those are the perfect critiques. As opposed to comments like "maybe you should have shot your film in the desert, instead of in the Arctic." Those are comments not much can be done about, once the film has been shot and no remaining funds for a reshoot. Your comment is perfect. It leaves me room to consider changes.

The film is about a world where everybody has a soul song, but this girl doesn't (she does also, but it will be revealed later). The movie however has songs everywhere. There are about 12 to 13 songs and two pieces of classical music in a 90 minute film. There will be songs before and after, and so that piece by itself shouldn't be jarring. At least I don't think so. But I will most definitely screen it for a test audience and keep your comment in close consideration as I move forward.


Thanks for the anecdotes. I'm looking at the DUC site right now.

There is the old, but true, saying that there are three films, the one you wrote, the one you shot, and the one you end up with. They are related but almost never the same.

Honestly, is there any other truth?

Again not saying you need to change anything, just a reminder that once your not doing films just for you that it becomes important how others perceive the film, and sometimes you need to change things so that they get the perception you want them to have even if you don't need that change.

Absolutely agree, 100%.

eMail me if you need some sounds or?

Thank you. I'm just loving this site.

Ryan is absolutely right about the built in pink. I believe the pink he pointed to is the ones from Blue Sky ,they and Dolby both have accurate Pink files.

I heard it in a tutorial to use the pink noise from Blue Sky as some Audio Post people do not believe the Pro Tools Pink Noise is accurate. But he was a lot less emphatic than Ryan. Sounds like Ryan got burned... haha.

Thanks again Scott. The insight and the commentary on the footage were very helpful.

Also, just wanted to mention how my brain is evolving on this issue. I wasn't thinking about how bare my audio was before. It's only now after reading all of this that I'm beginning to realize how rich the sounds are in a normal movie, because of the all various FX tracks, and ambience and door noises and chair creeks that are added later. So now I've started to think about that also. I checked out Sonniss, as per your other post. Great site.

Best,
Aveek

 

[Noiz2]

The pink that PT generates has a couple of issues and I may not be remembering exactly what they are but I believe they are outlined in the stickie on the DUC post section on calibrating your system. If I remember correctly it's not a true pink (equal energy per octave). So your calibrations will be off by a few dB.