• Wondering which camera, gear, computer, or software to buy? Ask in our Gear Guide.

Composers - A Few Requests From The Audio Post Team

I am currently working on the animation short of a fellow IT-er. Very, very nice work. I also really like the score. However, the audio of the score track itself is highly compressed, as if it were a pop music single.

Film/video requires a different approach to audio mixing. As sound designers, sound editors, composers and rerecording mixers we do not take front and center. (None of the other crafts should either, but we'll delay that discussion for another time...) The sound track is a part of an integrated whole. The sound track - dialog, Foley, sound effects and score/source music - needs room to breathe. By overly compressing the score you don't leave any room for the rest of the sound track. This leaves the rerecording mixer with two very opposite but equally bad choices. Either the score is front and center sonically, and the dialog, Foley and sound effects remain faint sonic impressions, or the sound design takes the acoustic lead and the score is faded into the aural background.

An uncompressed score, however, has lots of "air," or room for other sounds to occupy. All of the sonic elements have a chance to work together instead of competing for sonic space.

Something else that composers should do is to record stems. For a sound team a stem means all of the members of a sound group are isolated onto individual sub-mixes, and even subsequently recorded separately. For instance, there would be a separate stem for dialog, Foley, sound effects, ambient backgrounds and score/music (yes, there will be sub-mixes inside the sub-mixes). There is also the standard "M&E" (Music and Effects) mix, which is all of the audio except for the dialog so the dialog can be dubbed in another language. The composer should record separate stems as well - strings, brass, reeds, percussion, etc. as well as synths, drums, guitars and the like. Rerecording mixers are control freaks, and they really don't want to mess with your mix, they want to use it to every best advantage to enhance the visuals, the story and the characters.

As an example... It is very important that the audience understand that the character is walking away from a situation in a very angry state of mind, although s/he (for the purpose of the story) has not been able to let it show on his/her face. However, the Foley team did a wonderful job with making angry footsteps, and the sound effects team selected and edited in a terrific sound of a door being opened and bashed against the wall. The score, however, is full of percussive elements that do not blend well with the Foley and sound effects. If the rerecording mixer has the stems s/he and either eliminate the percussion from the score track altogether, or work magic with EQ, processing and effects so that the "angry" footsteps and door are able to do their job of informing the audience of the characters state of mind.

So, to sum up...

Composers, please DO NOT compress your score. And please record stems.

Our job is to support a story and characters, not be stars in our own right. If you want your music to be the star then become a pop star or write symphonies where the music is the star.
 
+1 on mixing to stems.

As a composer, I sometimes have to step back from my desk and remember that even though the music (and audio in general) is extremely important to the film, it is still just a piece of the overall puzzle, and it has to fit in with everything else that is going on.

Also, Alcove, if you want to give a lecture on the loudness wars, the evils of submitting a cue that is -6db rms to guys like you,...you have my full support.
 
I totally agree, I don't even over compress my NORMAL music, it's bad practice picked up by idiots who had more money than brains.

I agree with everything Alcove said re: compression & film soundtracks...

But I would disagree with this statement.. I really don't think Quincy Jones is an "idiot with more money than brains"...

https://www.youtube.com/watch?v=e-H1mCEAoiA

The things they did with compression and mastering on that album are still pretty amazing.

Just saying. ;)
 
Last edited:
You obviously have no idea what you're talking about. lol
Have you not heard modern music? Thriller was released in 1982 before over-compression was common practice. Compression itself is a very handy tool and everyone should use it, but the loudness wars are more than that. There's only so much you can contain in a digital file and when you hit that wall you lose quality - distortion, volume drop outs, all that.
 
Probably a bad example, Will. I'm a really "old school" guy when it comes to music recording and engineering. I can remember being limited to eight or 16 or 24 tracks. When you did a re-take you recorded over the original take and the old take was gone forever. You would "slam" the tape to get natural tape compression. Editing was done with a razor blade. And "window punches." Quincy was a genius, as was Tom Lord Algae, Alan Parsons and quite a few others who mixed and mastered in the '70's and the '80's. Pink Floyds Dark Side Of The Moon (which turns 40 this month) is brilliantly mixed and mastered, as is Steely Dans Aja (which is my audition CD when checking out a new studio). Aja, especially, has plenty of air, despite the complexity of some passages, and you don't lose any detail or presence in the quieter passages.

Over the last ten years record execs have insisted on "steamroller," "brick wall," "radio" style compression being done during the mastering process to insure that their record is as loud as it can be. This leaves no dynamic range to the music whatsoever - the softest parts of the music (whispered vocals over nylon guitar) are as dynamically loud as the pounding drums, thunderous bass, blaring guitars and screaming keyboards. In other words, when measured on a dB meter everything registers at -0.01db below the digital distortion thresh hold. And quite often there is digital distortion within the mix itself.

Not only is this artistically unattractive to me personally (I really like dynamics in my music) it is physically/aurally tiring. On my current project the music is so overly compressed that, even at low listening levels, I experience ear fatigue very rapidly. (Yes, your ears do get tired.) I spent six or more hours every day for four days creating and auditioning sound effects, including some "loud" sounds, and had no ear fatigue problems. Within an hour of working on the mix of the entire sound track (score + sound effects) my ear were fatigued.

As MetalRenard said compression is a very useful tool. When I was doing music engineering compression was a wonderful friend that smoothed out bass parts and tamed vocals, and I used it in many other useful ways. Now, in the digital age, it is used as a crutch to cover up creative/technical incompetence and as a marketing tool.
 
Perhaps not an ideal example, and yes modern music is awful.

Agree to disagree I guess... compression has it's place. Yes it is overused now. My point is that making a blanket statement that it's "bad practice by idiots with more money than brain" is false.

And, for what it's worth, I too have cut audio by hand with a razor blade.. ;)
 
And, for what it's worth, I too have cut audio by hand with a razor blade.. ;)

Aaaahhhhhh!!!!!! The aroma of a freshly opened box of Ampex tape..........







Waiting around for the tape to rewind.....


Biasing the tape deck......


Demagnetizing the heads......




Hmmm............ I think I like digital, but I do miss that old analog sound.....
 
I'll second what Alcove has said and I'd like to add:

Compression is an essential part of music production and I'm not suggesting that film score composers never use compression. I'm saying that compression should ONLY be used if required to smooth out vocal lines, etc. The heavy limiting type compression, used to slam the levels during mastering, should NEVER be used, ever! In fact the music should NOT be mastered at all for film!

The problem at the lo/no/micro budget level is that there really are no experienced film score composers. As the numerous posts on indietalk demonstrate, there are countless people out there who are either; students or freshly graduated students of music technology courses with maybe a few shorts under their belt or others with some experience or even with a lot of professional experience of producing music for commercial release and/or the games market. Almost without exception, they seem to think that film score creation is the same as other forms of music production, with virtually no differences apart from potential earnings. It really is amazing to see so many offering their services as film score creators and even sound designers when they actually know absolutely nothing about creating film scores. Come on guys, at least do a little basic research and find out something about the job before advertising yourself as a film composer and/or sound designer. The number of people out there advertising themselves and looking for sound design work, when they don't even know what sound design is, geez!

The simple fact is that on micro/lo budget projects the first thing I do when I receive the music from the composer is import it into my mix session and set the level at -30dB before I even press play to hear the music for the first time. 9 times out of 10 -30dB is pretty close to the final mix level.

BTW Alcove; I try to call the music submixes "splits" rather than "stems". As you know, "stems" is an already used film sound term and calling the music splits "stems" can lead to some serious confusion down the road.

Gi
 
Agree to disagree I guess... compression has it's place. Yes it is overused now. My point is that making a blanket statement that it's "bad practice by idiots with more money than brain" is false.

With all due respect, you're talking about something completely different. Compression is a GREAT tool, I agree, but heavy/over-compression on the master channel is bad practice. It is used everywhere by big labels. There's nothing to disagree about, it's fact. :)

Read up about it some more before continuing to tell me what you think:
http://en.wikipedia.org/wiki/Loudness_war
Informed discussion is much more fun.
 
Last edited:
I find that stemming goes hand-in-hand with payscale. Myself, and people I know that provide stems (which is everyone I know) demand a higher rate that most producers/companies do not want to pay. Stemming takes way more time, or requires a 2nd system slave-locked to keep things moving smoothly.

Depends on the $ and the scope of the project, typically.

On compression I agree with you. Story is first, music is complimentary.
 
Probably a bad example, Will. I'm a really "old school" guy when it comes to music recording and engineering. I can remember being limited to eight or 16 or 24 tracks. When you did a re-take you recorded over the original take and the old take was gone forever. You would "slam" the tape to get natural tape compression. Editing was done with a razor blade. And "window punches." Quincy was a genius, as was Tom Lord Algae, Alan Parsons and quite a few others who mixed and mastered in the '70's and the '80's. Pink Floyds Dark Side Of The Moon (which turns 40 this month) is brilliantly mixed and mastered, as is Steely Dans Aja (which is my audition CD when checking out a new studio). Aja, especially, has plenty of air, despite the complexity of some passages, and you don't lose any detail or presence in the quieter passages.

Over the last ten years record execs have insisted on "steamroller," "brick wall," "radio" style compression being done during the mastering process to insure that their record is as loud as it can be. This leaves no dynamic range to the music whatsoever - the softest parts of the music (whispered vocals over nylon guitar) are as dynamically loud as the pounding drums, thunderous bass, blaring guitars and screaming keyboards. In other words, when measured on a dB meter everything registers at -0.01db below the digital distortion thresh hold. And quite often there is digital distortion within the mix itself.

Not only is this artistically unattractive to me personally (I really like dynamics in my music) it is physically/aurally tiring. On my current project the music is so overly compressed that, even at low listening levels, I experience ear fatigue very rapidly. (Yes, your ears do get tired.) I spent six or more hours every day for four days creating and auditioning sound effects, including some "loud" sounds, and had no ear fatigue problems. Within an hour of working on the mix of the entire sound track (score + sound effects) my ear were fatigued.

As MetalRenard said compression is a very useful tool. When I was doing music engineering compression was a wonderful friend that smoothed out bass parts and tamed vocals, and I used it in many other useful ways. Now, in the digital age, it is used as a crutch to cover up creative/technical incompetence and as a marketing tool.

Totally agree that "Dark side of the moon" is a masterpiece of recording and mixing. IT doesn't really relate to film, but the overcompression and limiting of so much modern music is crazy. There's a recent U2 song ( don't remember what it's called) that starts with a solo guitar and then when the band comes in the sound "shrinks," so that subjectively it sounds all wrong to me. The band should sound louder than a solo guitar, yet it actually sounds smaller. That's modern loudness war brickwall limiting etc for ya. But this is a digresssion :)
 
BTW Alcove; I try to call the music submixes "splits" rather than "stems". As you know, "stems" is an already used film sound term and calling the music splits "stems" can lead to some serious confusion down the road.

Probably one of the many the semantic differences between the US and UK/Europe; a sub-mix of any kind during the rerecording process is referred to as a stem here.
 
Compression is a GREAT tool, I agree, but heavy/over-compression on the master channel is bad practice.

In film/TV we are not talking about "heavy" or "over" compression on the master buss, there should be NO compression or limiting whatsoever!

I find that stemming goes hand-in-hand with payscale. Myself, and people I know that provide stems (which is everyone I know) demand a higher rate that most producers/companies do not want to pay. Stemming takes way more time, or requires a 2nd system slave-locked to keep things moving smoothly

Creating splits should take virtually no extra time and certainly does not require a slave-locked second system, especially as the track counts are generally so low for music creation anyway. Instead of bussing to the master channel you just buss to an aux master first, it should take just a minute or two of set up.

Probably one of the many the semantic differences between the US and UK/Europe; a sub-mix of any kind during the rerecording process is referred to as a stem here.

Not really, it's a film/TV thing rather than a geographical thing. Stems are Dialogue, SFX and Music, so the complete music mix is already a "stem". Spliting the music out into sub-mixes would at best be sub-stems, although I prefer to use the term "splits" as "sub-stems" is usually used to mean splits of the SFX stem into say Foley, ambiences, etc. There is no absolute universally agreed convention here though amongst audio post personnel.

G
 
Back
Top