Why I Make Chinese Films

Joshua had quite the find with those video tapes from '99. I honestly do not remember those events taking place, but it sounds like Joshua has verifiable evidence of our antics. I will take him at his word.

Some natural questions would be, "Why would you make a film in Chinese if no one you know speaks that language? Why not English? And why the choice of Chinese over anything else?" Good questions. I'm glad you asked.

First we must give this film a title, though it has never had one, since we will refer back to it many times over the course of this blog. And if Josh gets what he wants we will be referring to it a great deal in the future. For lack of a plot, and to allow for some nostalgia, we'll call it 這是正常的。

First we must find a reason to make 這是正常的 in a language other than English, since that's the only language that I speak. To figure this out we must go Wong Kar-Wai (or Kar-Wai Wong, depending upon which side of the globe you're on). Though I am unable to recall the exact inspiration for 這是正常的 or what the film would even be about, I say with much certainty that it was inspired by a then-recent watching of Chungking Express. Almost everything I do can be traced back to that film. (On a side note: 這是正常的 was probably about nothing in particular, just an excuse to make a film in Chinese.)

A key to the narrative in Chungking Express is the use of voice over. Dialogue itself is less significant. Instead, voice over and music take up equal parts with dialogue, forming a sort of WKW trinity. But does voice over really work? We have all no doubt at one time or another watched a film in English that had voice over. It almost never works. Why? Because voice over is cheesy, its obvious, its a narrative shortcut that adds no real depth, voice over performances never mesh well with the character. And yet, despite those facts, it works in Chungking Express (and all other WKW films). I had to know why.

The answer was found: because we are not really taking into account the performance so much as we are simply reading (through the miracle of subtitles) WKW's excellent personality portraits. All of a sudden performance didn't really matter, since you got very little from the actual performer.

Voice tone, inflection, pace, pitch, power, modulation, feeling, emotion. These were all now barely discernible due to the change in language. All you needed to feel were the words themselves. The need for this is heightened by the fact that we would be using unskilled actors. This would allow us to use much cheesier lines without having them sound that way. For instance, do you think any American actor could get away with using this line in a voice over...

Somehow I've become very cautious. When I put on a raincoat, I put on sunglasses too. Who knows when it will rain, or when it will turn out sunny?

I know what you're thinking: what color would your subtitles be? Of course, we would use only the highest quality of yellow to ensure that bright lights would not distract in the reading of our movie. However, I could not guarantee the highest quality of translators. Yes, the film would have to be translated by an independent interpreter to ensure that the true message of the film was obscured just enough. As any true film-maker would do.

That's one thing that I do recall about 這是正常的, I always felt that it must have a true Chinese script. It was one thing for people to speak gibberish, but they had to at least have some motivation. Of course, myself having no knowledge of this language (or any other for that matter) this was an impossibility. But again the technological age comes through for us. Recently, AltaVista has added Chinese simplified and traditional to its excellent translation utility. Now you can transform the following line of dialogue...


"Oh no! The man who runs has been shot!"

...into the following line of gibberish...


"Oh 沒有! 跑的人是shot!"

...of course, when you try to translate that back...


"Oh does not have! Runs the person is shot!"

...it really is a shame that Oh does not have. I just wish I knew who Oh was and what they didn't have. But that's a topic for another day.

Let's now shift our focus from my brilliant film to Joshua's idea for a film. Basically, Josh wants to do the old film within a film thing. Gee, that's so fresh and new (wave?)! I.E., drip, drip drip...

Why try to taint what is so pure? Why try to improve upon a Mona Lisa? Why would you want to add that guy from Married With Children to Sports Night? What possible purpose would that serve?

If I'm going to devote my time to a film it will be the original, the alpha, the ancient of days.

That's all.


Written, Directed, and Produced by Jeremy Provost


Vonnegut Lives!

Kurt Vonnegut is still kicking at age 81, and every little bit as entertaining...

Cold Turkey


Tell Me About The Film You're Making

So, I'm going through some old 8mm video tapes, looking to capture anything interesting while I still have Gabe's camcorder.  So, I've got a tape from September 1999 with me doing a walk-through tour of my first house, to send home to the peeps in Massachusetts.  At the end, I play a little guitar.  It's pretty bad.  It's embarrassing.  Then, the tapes is done, my bro takes it back to Mass, and all is well.

Not quite.  Doing a final pass last night, I let then tape run through, and, boom, there's my brother Carl, self-taping, extreme close up.  He says:
So, if you're watching this, you probably had to sit through that horrible tape of Josh playing guitar.  Also, you probably see that I need a shave.

This is followed by a number of late night shots in my bros bedroom, with Jeremy and Carl taking turns taping each other on the computer.  It's all quite fascinating.  In the last clip, Carl relates the "Putt-Putt" mini golf bit made famous by Patch C-Holm, or some such retail cinema employee.  Nice.

The most intriguing clip, however, is of Jeremy, responding to the query posed in the title of this blog entry. 
So, I want to make this film... in the style of a Wong Kar-Wai film.  You know, influenced by Godard and that whole New Wave, French thing.  And we'll use people we know... our friends... these are American Actors.  But they'll speak Chinese, but with subtitles in English... because I think that, really, any film is better, with subtitles. (Carl: But how will you get them to speak Chinese?)  Look, basically... it'll be gibberish.

It's tough to capture the pace and style of the language used, but, trust me, this is wildly funny.  And on the way back from lunch--we're talking, not ten minutes ago--and this really starts to inspire me.  Because it's difficult, for me anyway, to come up with good ideas for something to film, something plausible, feasible, something that I'd love to shoot.  So here it is.

A semi-fictionalized film about a small-time visionary young (in his own mind) Director (played by Jeremy), obsessed with Wong Kar-Wai (this is true), following him in his attempts to make the above film.  No one shares his vision (though he'll share it in abundance, ad neaseum), so this becomes a difficult task.  Like a WKW film, this film (and the film within a film) would be largely unscripted, just setting a situation and letting it go.  Jeremy can freestyle about WKW, and we can reference a lot of WKW elements of style (repetitive music, language, voice-overs) and substance (do you like pineapple?).  No doubt, this film would be a gem to other WKW aficionados.

What do you think, Jer?

En El Estudio

Fatigo is going into the studio today to record their 727 Records debut album.  They signed on about a month ago, and are jumping right into it.  I describe them as psycho mariachis.  Mad props for amping a classical guitar through a distortion pedal.  The album should be out early 2005.



I had tickets to the Braid show last night, but I missed the show to take care of Angie, who wasn't feeling well.  I cooked Sloppy Joe's.  Jimmy Eat World showed up at the last minute to play a surprise set.  I've thrown away more concert tickets this year than I've used!


Why I Make DVD's About Debue

The saga of the Debue DVD is nearing its end.  It began almost a year ago.  At that time, I accepted some video tapes from the band, agreeing to transfer them to DVD and return promptly.  Why did it take a year?
For starters, I had already accepted a similar project from my friend Gabe.  He had shot The Casket Lottery and Small Brown Bike in Chicago in the summer of 2003 on Video8 while simultaneously DAT taping the audio.  My task was to synch the two sources, create some menus, and burn to DVD.
Beyond that, I had a number of DVD projects already on my hard drive, waiting to be completed.  There was the Cleary Time Is My Crisis music video, the 727 Records showcase concert at The Center for Arts in Natick, a new The Moon Is No More music video, and a bunch of old skateboard videos that needed to be edited.
Basically, when I agreed to do it, I already had six or seven projects in the queue.  And then there were the technical issues...
Any non-DV sources, such as Gabe's Video8 and the Debue camera angle that was shot on VHS-C, were going to be captured through my ATI All-In-Wonder card.  Well, it turns out the quality wasn't that good at all.  The best quality was the lossless setting, which resulted in files roughly 30MB/minute. 
First, I purchased the ADS Tech PYRO A/V Link box (at about $250) to convert the analog sources to DV on the fly.  This little number was flawed from the start.  The chip in the box couldn't lock onto anythign but perfect analog signals, and basically couldn't capture more than a minute or so of any source.  It took me a couple months to figure that out.  ADS support was non-responsive (Finally, in the last few weeks they have agreed to replace these units with functioning ones).
Finally, I purchased the Pinnacle Studio MovieBox DV (for another $250).  This baby worked like a charm right away, and I started to catch up on my projects.  First, Gabe's DVD, since I had his camera for almost a year.  I did a sweet pre-menu video with blended, ghosted clips of the band rocking out, synched to a Casket Lottery song.
Even with that out of the way, it wasn't smooth sailing to get started on the Debue DVD.  This project entails editing together two camera angles (wide shots on VHS-C, with another doing close-ups on miniDV), on about 20 songs over nearly two hours.  My original captures of the miniDV had some nasty artifacts due to a dirty camcorder used to do the transfers, and those had to be recaptured (and I don't own a camcorder, had to borrow all the way).  I thought I had a good VHS-C camcorder to do the transfers, but realized that it didn't have any outputs.  I guess they expect you'll always have one of those convertor cartridges around.  I didn't.  So, I had to buy one of those.
Anyway, that's where we are.  I did some tests earlier to make sure the whole thing would work, but the bulk of it begins now.  I need to do at least two songs a night to get this ready for the DVD release party that's scheduled for August 6.  I'd love to run the finished edit through Magic Bullet, but it would take about four weeks to render, according to my calculations.  Maybe I'll keep the footage around and run it through when I get a faster processor.

Magic Bullet Redux

After extensive testing and non-stop rendering over the weekend, I think I have hit on a workflow to eliminate the jerkiness I was seeing with Magic Bullet.
Taking my original 60i video, I feed it into Magic Bullet and allow it to deinterlace, convert to 24p, and output the 24p file.  Then I encode that as 24p in TMPGENC with 3:2 pulldown enabled on playback.  The trick here is that the DVD player is doing the 3:2 pulldown on the fly.  For whatever reason it does this a lot better than Magic Bullet does when it converts 24p back to 60i, actually rendering the 3:2 pulldown.
This eliminated the jerkiness, where you could see the interlacing on playback, but there still is a stroboscopic effect some have discussed.  Really, I have figured out that this is really only happening in my head, as I have seen the original video numerous times.  Of course I can see the differences!  I watched a movie on TV last night and when I really looked for it, I could see the same effect and type of motion.  So that's it, I think someone else coming it not having seen the original footage probably would be impressed and wouldn't think anything was out of line.
Total time to render 2:33 video: 14 hours (but the film processing looks sweet and the letterbox is nice and crisp)


Only the Good Debue Young

What in the world could possibly be worse than a great band meeting its demise before its time? Death and taxes, my friend.
If you ever lived in the Franklin, Massachusetts, United States of America area then perhaps you were familiar with a little four-piece that went by the name of Debue. There may have been an accent mark in there, like in the word Café. But maybe not.
These four knew how to jam. That was their specialty. Put them together in a room, give them four instruments, 15 minutes, and one inspiration and they'd blow your socks off. Were they derivative of every great 70s prog-rock band? Sure, but can you blame them? Is that a problem?
Unfortunately, Debue met their Maker earlier this year, torn apart by distance. I didn't think distance could tear things apart anymore, not like it used to. Remember before the Internet? Probably not, but bear with me. Remember when you had to use a pen and paper to communicate with someone who didn't live in your hometown? That's when distance was a problem, but not anymore... right? I guess not.
Its members were of a different generation, no matter how old they may have been. They were from a generation where meeting someone face to face was what life was all about. These guys didn't have time for technology and e-mail and blogs and digital music. These guys thought that MP3 was John Michael Palermo, III without his first initial. They were too busy filling the air with sound. Good for them. Too bad it killed them.
R.I.P., Debue.
But wait! Despite their shunning of the technological world, the technological world just couldn't bring itself to shun Debue in turn. Instead, Debue is resurrected by bits and bytes. Joshua is hard at work on an amazing DVD of a live Debue concert. This was Debue in their heyday, at their peak. This was like Larry Bird in 1986 or Sports Night in its first season. This was unstoppable. And so, through the miracle of Digital Versatile Disc, Debue lives again. Even if only in our minds, hearts, and DVD players.
I say it again: R.I.P., Debue.


Already, I'm starting to see a pattern here: I post some potentially useful and/or timesaving information (see CL vs. MB, below) and you rant about how pointless what I am doing is.  Interesting.
I think blogs come back to our secret exhibitionism.  We could easily be emailing each other.  Instead, we're making a scene in public, as it were.
I will defend that hopefully some of this will be useful.  Take Brock's blog for an example.  Solid film and record reviews.  He's not writing for a zine, but the quality of the writing and analysis is better than most.  That'll get spidered, and a Google for "beastie boys five boroughs review" will bring it up, in time.  And my stuff about film look for video will come up in the same searches I exhaustively performed over the last few days.
It's all good-nood in the hood-nood.

The Over-Blogging of America

When JohnAnthony creates a blog then you know this has gone too far. What is everyone's fascination with blogs? Why would anyone want to read some stream-of-consciousness garbage written by a complete stranger? Can we commission a study please? I'm serious.
One question that you have to consider: is anyone actually reading the thousands of blogs that exist? How many people actually visit these sites on a regular basis? These results just in: not many.
Sure, we don't have scientific data on that, but it makes sense, right? And anyone who starts a blog probably realizes that this is the case. Now that leads to the question: why write a blog that no one is going to read?
I wrote a piece once, called "Why I Write About George Orwell," that dealt with this issue, though in a pre-blog context. Writing is a release. Anyone who writes will tell you that. It just so happens that thanks to the prevalence of the Internet, the world is just now waking up to this realization.
Deep down, people always wanted to write, but they just didn't know the form that it should take. Blogging just gives them excuse to do what they've been craving. Its really the same principle of a diary or a journal, except you get the added bonus of having the possibility that someone might stumble upon your writings and enjoy them.
Possibility. That's all it takes. It doesn't matter that no one actually reads them. It doesn't matter that there's 2+ billion words floating around in cyber-space unaccounted for. It just matters that someone might read a piece called "The Over-Blogging of America" and enjoy it. And that's enough for me.

CineLook vs. Magic Bullet

OK, so here are the results of the bake-off.  DigiEffects CineLook vs. Red Giant Software Magic Bullet.  They both do the "film look" for video thing.  I had to pull a lot of info together, so in the interests of saving some work for anyone else trying to figure this out, I'll dump some knowledge here.
The first thing anyone will tell you (and the documentation for both software packages reports the same) is that you can't make bad source material look good.  So, bare minimum you need to start with something like miniDV source.  Second, much of the look of film has to do with lighting, angles, composition of shots, attention to detail in sets, quality of acting, etc.  If you come at it from the get go with professional techniques and approach, the final result will be much better.
With all that said, the final step is running a plugin to achieve the technical flourish of 24 fps conversion, color correction, etc.  CineLook and Magic Bullet take different approaches to this.  Funny to note that on CineLook's web site they talk about how their plugin is not a "magic bullet."  I'm guessing that was a little inspiration for the Red Giant developers.
DigiEffects CineLook
CineLook has been around for a while, and it seems as though it hasn't gotten much development attention in recent years.  It's focus is the bundled presets that simulate a variety of film stocks from Agfa, Kodak, Fuji, and more, in 8mm, 16mm, 35mm, color and black and white.  This was intruiging, since I was hoping to simulate the look of certain films for which I knew the film stock used.  So, in addition to the 24 fps conversion, the preset has the gamma and color correction, and simulates the grain of the stock.
I ran my test video (Cleary's "Time Is My Crisis", directed by Jeremy Provost) using a number of the presets.  The frame rate conversion looked very good, it was very smooth, not at all jerky as I was told poor frame rate conversion could be.  The presets gamma and color correction seemed over the top (more on that later).  I can't imagine any film giving the extreme look I was getting.  The grain really detracted from the look, it added a lot of noise, and looked very stylized.  Really, the grain of 35mm film is so fine, that in DV resolution, it should not be visible at all.  Finally, CL adds a ghosting effect to simulate the typical 1/48 shutter speed of a film camera and the supposed slight motion blur this would create.  Well, this effect was also very over the top.  My clean cuts became quick fades due to the persistence of the image.
Overall, the result looked pretty bad.  I'd say it really changed the video, where it should really just enhance it.
Red Giant Software Magic Bullet
MB is a newer package, which has supposedly been used in a number of major films.  It takes a different approach, with modular plugins that can be mixed and matched.  The Magic Bullet plugin proper performs the 60i->24p-60i frame rate conversion, Look Suite has the gamma/color correction presets, Letterbox does what you would expect, and Broadcast does a final color correction for broadcast colors.
So, the Magic Bullet frame rate conversion produced superior results in terms of deinterlacing and deartifcating to create some very slick looking stills.  However, it's conversion back to 60i, the 3:2 pull down process, created obvious jerkiness on playback.  This could be a show-stopper.
The Look Suite presets take a different approach than CL.  Instead of simulating specific film stocks, they simulate certain film processing and development techniques, such as leaving the skipping the bleaching process and leaving silver on the film to create a desaturated look.  They have presets that reference a number of major films, such The Matrix, Amelie, and more.  The effects are much more subtle, and when you see them in action, they are immediately recognizable as film.  The look is very high quality, and combine a lot of film concepts, such as grads, black and white diffusion, and more.
Look Suite actually helps me to think a different way about the whole process, and made me feel more comfortable about tweaking presets to create something is unique.  Since I'm simulating looks, not film stock, I'm free to mix and match within a project, and give certain scenes a unique fell, such as when a certain character is present.  And since it's modular, I could conceivably use Look Suite on all my raw footage first, then Magic Bullet and the others later.  I'm free!
Letterbox allowed me to clean up the letterboxing on the project, which could be blurry, or could be inconsistent.  Or, I could go with a narrower letterbox altogether, going from 16:9 to 1.85:1.
Broadcast worked as advertised.
Overall, MB skips grain and motion blur, giving it a slick look that on PC/DVD looks more like real film and less like an effect.
Final Analysis
After my tests, I found another article where the creator of CineLook admitted that the presets were intentionally over-the-top so that average users could see a dramatic result.  Subtlety had been thrown out the window.  That bugs me, but I'm not averse to using CL where it works well, the smooth frame rate conversion, if I can't get Magic Bullet working right.  I could use MB for looks, CL for conversion, and finish with MB for letterbox, etc.
Another option is to let MB convert to 24p, and have the DVD player perform 3:2 pulldown in realtime.  Will have to test that.
I think what I have learned is that the real-world techniques of filmmaking are going to make or break a project, not plugins.  My sample video was reasonably well made, even though it was shot guerilla-style.  When the shots are steady, framed properly, and we lucked into decent lighting, there is a certain illusion.  But the shots that are poorly lit and where lip synching wasn't perfect detract from the professionalism.
All in all, after watching any of the processed clips and then going back to the original finished video, the difference between film and video becomes very obvious.  The video is overly crisp and clean and the motion is ultra-real.

Attempt Quitting

Hey, that highlights a depth in your songwriting even I hadn't seen before.  The more you dig, the more you find.  I like the strugle implied in that simple phrase: action and inaction.  Good name.
To show just how crazy this is, even JohnAnthony started a blog.  Unbelievable.
The results from Magic Bullet's overnight render are in and they are good.  I'll get to that a little later.

Why "Try Avoidance"?

What I can't understand is, why call this blog Try Avoidance? I mean, what does that phrase really mean anyway? I did not name the blog myself (that honor goes to Joshua), but let's try to explain anyway.
First we'll begin with a definition of the two words that make up the title of this blog.
Try: To make an effort to do or accomplish (something); attempt: tried to ski. But a definition that I think is more appropriate in the context of this phrase would be the following: To test or use experimentally.
Avoidance: A dismissing or a quitting; removal; withdrawal. Deliberately avoiding. 
I like the thought of testing or using experimentally a withdrawal or deliberate avoiding of something. But what are we avoiding? And doesn't the thought of trying something really leave a lot of room for failing at something? Certainly.
But here's the thing: there is no avoidance involved in the creation and maintenance of this blog. If anything, the exact opposite situation is going on. To blog is to embrace, to share, to open yourself up to the scrutiny of every single stranger that may stumble across this acre of CS. It's the anti-avoidance: confrontation.
So I vote for a name change.

And, Joshua, you hijacked my Courier font. I want it back.


Video as Film

I've been experimenting with techniques to make miniDV video look more like film. I've got Cinelook and I'm going to try out Magic Bullet. The process takes a long time, though. It makes MPEG2 encoding seem fast in comparison. I've got something encoding almost 24 hours a day it seems.

The big trick seems to be the temporal difference between 30fps (60 fields interlaced) and 24 fps (whole progressive frames). 24fps seems to be just on the borderline of perception of still frames versus motion, giving it a not-quite-real look. Slower sampling equals more motion blur, and then there is the grain and color response of the respective film stocks.

I spent a lot of time trying to find out what type of film stocks were used in Wong Kar-Wai films. Best I can work out from various interview with DP Christopher Doyle:

  • As Tears Go By - Agfa?
  • Days of Being Wild - Agfa? 320T pushed 2 stops/4 filters
  • Ashes of Time - Agfa?
  • Chungking Express - Agfa XT 320
  • Fallen Angels - Fuji 250T?
  • Happy Together - Fuji 250T
  • In The Mood For Love - Kodak?

And that's to say nothing of the other tricks WKW employs. For now, though, I'd settle on giving the Time Is My Crisis video a little polish.

hello, world

I was inspired by Brock and Gabe, so I started my own blog.