So, for those of you interested, my "clip editing process" goes something like this:
1. Copy raw footage off of ContourHD. This is when I know if I got the data or if I forgot to charge the battery or clear the memory. Although the Contour beeps loudly when it runs out of juice or memory I sometimes don't hear it in the heat of the action.
I missed at least three really good races because I forgot to clear the memory before the race. One was the Mystic Velo Crit in 2010. The camera's memory lasted 15 minutes. I got second in the race. Another was one of the first Bethels in 2010. In a terrible rainy day I managed third in the sprint, and I raced with virtually unusable brakes. The last was the 2010 New London Crit. A new course, lots of turns, and a good finish after the guy in front of me fell over in the last turn.
I now have a spare ContourHD so I hope to eliminate problems like "I forgot to charge it" or "I forgot to clear its memory" or "I forgot the camera at home" kind of things.
I missed at least three really good races because I forgot to clear the memory before the race. One was the Mystic Velo Crit in 2010. The camera's memory lasted 15 minutes. I got second in the race. Another was one of the first Bethels in 2010. In a terrible rainy day I managed third in the sprint, and I raced with virtually unusable brakes. The last was the 2010 New London Crit. A new course, lots of turns, and a good finish after the guy in front of me fell over in the last turn.
I now have a spare ContourHD so I hope to eliminate problems like "I forgot to charge it" or "I forgot to clear its memory" or "I forgot the camera at home" kind of things.
2. If the clip is good enough to work with I import it into iMovie. This is a few hour process so I start it and let it run, usually overnight. My first computer was virtually useless while running this in the background. Now, with 16GB of RAM and a quad core processor, I can do basic browsing really slowly or work on writing.
This step put a huge dent in my video clip production. At some point in 2012 I realized that the raw footage I wanted import weren't showing up in iMovie. After a lot of research, after trying all sorts of stuff (using different cars, renaming files, etc), I finally resorted to using the old MacBook, the one that is sort of obsolete.
I could upload the clips right away. This started a spasm of video production - I've finished two, almost finished a third, and I have a few more to work on.
3. My priorities in picking which clip to do breaks down to three things. The first is whether or not it's fun. For me fun is when I can substantially influence, in a good way, the outcome of a race. That means I do a good sprint, I help a teammate, stuff like that. In 2012 I was very unfit so I almost never had the legs to help a teammate so I tended to choose the races where I had a decent finish. To give an example of the opposite of fun - there's a race where I attacked at the gun. No one went with me, I pushed on when I should have sat up, and I soloed for about 3 laps. I got caught and dropped in about 30 seconds. Would this make an exciting clip? No. Maybe to laugh at me but there's nothing productive I can find in that race.
The next thing is if there's a new course. Although I like Bethel I don't want to have 80 clips of Bethel and 15 of other courses, split evenly between TuesdayTheRent and New Britain. Seriously though I don't get to do new courses too often so if I can get to a virgin helmet cam course then all the better.
The last factor is a lesson one. In some races the racers are so good at tactics and bike handling that I feel like a novice in them. In other races... well it's sort of the opposite. Watching errors on tape makes it much more real and allows analysis so that the racer can work on fixing their errors. This is what happens in football but in amateur cycling it's quite rare.
This step put a huge dent in my video clip production. At some point in 2012 I realized that the raw footage I wanted import weren't showing up in iMovie. After a lot of research, after trying all sorts of stuff (using different cars, renaming files, etc), I finally resorted to using the old MacBook, the one that is sort of obsolete.
I could upload the clips right away. This started a spasm of video production - I've finished two, almost finished a third, and I have a few more to work on.
3. My priorities in picking which clip to do breaks down to three things. The first is whether or not it's fun. For me fun is when I can substantially influence, in a good way, the outcome of a race. That means I do a good sprint, I help a teammate, stuff like that. In 2012 I was very unfit so I almost never had the legs to help a teammate so I tended to choose the races where I had a decent finish. To give an example of the opposite of fun - there's a race where I attacked at the gun. No one went with me, I pushed on when I should have sat up, and I soloed for about 3 laps. I got caught and dropped in about 30 seconds. Would this make an exciting clip? No. Maybe to laugh at me but there's nothing productive I can find in that race.
The next thing is if there's a new course. Although I like Bethel I don't want to have 80 clips of Bethel and 15 of other courses, split evenly between TuesdayTheRent and New Britain. Seriously though I don't get to do new courses too often so if I can get to a virgin helmet cam course then all the better.
The last factor is a lesson one. In some races the racers are so good at tactics and bike handling that I feel like a novice in them. In other races... well it's sort of the opposite. Watching errors on tape makes it much more real and allows analysis so that the racer can work on fixing their errors. This is what happens in football but in amateur cycling it's quite rare.
4. Skim whole clip in iMovie, a stock application that comes with Macs. I like iMovie because you can see thumbnails of the whole clip in very small increments, and if you drag the cursor around the video plays at that speed. Sliding the cursor forward means you get a good fast forward. It makes it easier to find things on the clip.
I try to remember significant tactical moves and include them in the clip. Sometimes I'm at the back when a break goes (okay, I'm almost always at the back) and so the clip can't illustrate the break going away. If the move isn't visible I usually skip it but I may allude to it in the clip's text.
I try to remember significant tactical moves and include them in the clip. Sometimes I'm at the back when a break goes (okay, I'm almost always at the back) and so the clip can't illustrate the break going away. If the move isn't visible I usually skip it but I may allude to it in the clip's text.
5. I also see if I can find any memorable but independent incidents that I remember. These include crashes, weird moves, elemental errors, and close (but safe) situations. They may not affect the tactics but they're interesting. They include things like a bottle pass between two teammates while a third rider, not on the same team, drops his own bottle. Another race, the last non-HD one, featured a newspaper stuck in a rider's derailleur that I ended up removing.
Many of these interesting sections are 30-45 seconds, and some of them are as short at 8-12 seconds. I don't go shorter than 8 seconds since it's hard to comprehend what's happening in 4 or 5 seconds. Most of my text boxes are 4 seconds long so to have an "intro text" and a "conclusion text" takes, at best, 8 or 9 seconds.
Many of these interesting sections are 30-45 seconds, and some of them are as short at 8-12 seconds. I don't go shorter than 8 seconds since it's hard to comprehend what's happening in 4 or 5 seconds. Most of my text boxes are 4 seconds long so to have an "intro text" and a "conclusion text" takes, at best, 8 or 9 seconds.
6. I get the last lap in its entirety. Since I usually contest the sprints that's the most interesting part of the race. If applicable I get the prior lap and even a lap before the final lap, but since laps take up minutes at a time, it's "expensive" time.
In the past I was limited to 10 minutes. I had to cut a second here and a second there just to make the 10 minute limit. Then YouTube "upgraded" me to 15 minutes. Before I could get a couple clips out they upgraded me to "unlimited". Remember my clip selection process - fun, course, and memorable moments. Most of a race doesn't have fun stuff, the course is the same, and there are few memorable moments. Therefore I don't upload a whole race.
With my new unlimited time I find myself cutting less. My clips seem to end up at about 10-18 minutes long before I cut them.
In the past I was limited to 10 minutes. I had to cut a second here and a second there just to make the 10 minute limit. Then YouTube "upgraded" me to 15 minutes. Before I could get a couple clips out they upgraded me to "unlimited". Remember my clip selection process - fun, course, and memorable moments. Most of a race doesn't have fun stuff, the course is the same, and there are few memorable moments. Therefore I don't upload a whole race.
With my new unlimited time I find myself cutting less. My clips seem to end up at about 10-18 minutes long before I cut them.
7. When it's close to a usable length then start adding music. The music drives the mood so that's important. Sometimes I have a song in mind so I put that in wherever I wanted to put it then I work around that.
The music tends to have a cadence-type beat, something faster than 70 or 80 beats per minute. Something closer to 100 bpm is better, otherwise the clip seems to drag. The exception is if I'm trying to relay a sense of fatigue or something like that. Then I'll select more mellow songs.
One important thing is that I listen to the songs when I'm not working on clips - the Missus will verify that, yes, we listen to those songs over and over when we're in the car. I listen to them when I'm on the trainer. It gives me an idea of the song's "phrasing". I don't know what it's called but some songs are sort of beginning songs that kick things off with a flourish, some are middle songs that build tension or have an offbeat first note (so it sounds like it's a continuation of a thought, not the start of one), some are quiet, and some are great ending songs. I use songs that my brothers played in their bands and I try to use songs that certain people wrote and/or sang because I've been in touch with them.
One singer, Derek, of Linus, URT, and Zen Men, is a local and I see him pretty much once a year at my brother's house. He refers to me as the one man fan club.
Luckily the music my brothers played was mainly fast beat stuff.
There are some songs that just aren't appropriate for a bike race. They tend to be more emotional or very slow. My brothers' bands played a lot of music, more than the 20 or so songs I've used, but unfortunately the music isn't right for a bike racing clip.
Since I'm publishing to YouTube I don't use other artists' music. I don't want to run into copyright violations and such. I've asked some local racer/musicians for original music but the cadence, the bpm, has been too low in everything I've checked out. Nothing personal, it's just the music is too slow.
The music tends to have a cadence-type beat, something faster than 70 or 80 beats per minute. Something closer to 100 bpm is better, otherwise the clip seems to drag. The exception is if I'm trying to relay a sense of fatigue or something like that. Then I'll select more mellow songs.
One important thing is that I listen to the songs when I'm not working on clips - the Missus will verify that, yes, we listen to those songs over and over when we're in the car. I listen to them when I'm on the trainer. It gives me an idea of the song's "phrasing". I don't know what it's called but some songs are sort of beginning songs that kick things off with a flourish, some are middle songs that build tension or have an offbeat first note (so it sounds like it's a continuation of a thought, not the start of one), some are quiet, and some are great ending songs. I use songs that my brothers played in their bands and I try to use songs that certain people wrote and/or sang because I've been in touch with them.
One singer, Derek, of Linus, URT, and Zen Men, is a local and I see him pretty much once a year at my brother's house. He refers to me as the one man fan club.
Luckily the music my brothers played was mainly fast beat stuff.
There are some songs that just aren't appropriate for a bike race. They tend to be more emotional or very slow. My brothers' bands played a lot of music, more than the 20 or so songs I've used, but unfortunately the music isn't right for a bike racing clip.
Since I'm publishing to YouTube I don't use other artists' music. I don't want to run into copyright violations and such. I've asked some local racer/musicians for original music but the cadence, the bpm, has been too low in everything I've checked out. Nothing personal, it's just the music is too slow.
8. After roughing out the music I start adding text. This emphasizes the mood, gives me a chance to educate riders, and I spend a lot of time here. I'll adjust text after adding music because the two don't quite match and I can't rewrite the music or change its timing. I like some of the disjointed phrases in my brother's music, like being "so far away, at least one hundred miles" when I'm not on a wheel or "spread out like satin" when the field fans out for a turn.
- If applicable I fill in the non-music sections with music.
- Once I have the thing roughed out I add the credits. This takes a while since I try to list everyone I mention in order of appearance (after teammates or significant players), I have to get all the songs, and I try to remember all the people I need to thank at the end.
9. With a rough clip in hand I start the hard work - the polishing and honing. I probably refine a dozen times or more. This includes the time intensive step of reviewing the whole clip, at normal speed. It's here that I often catch the real errors. Nonetheless virtually every clip went out with an error, either mismatched text (I edited out a few blocks but a prior or following block relied on the old text so that one block seems out of place), mismatched font (when iMovie buffers heavily it drops font selection), or outright errors (bad spelling or whatever).
Since most of my clips are 10 or so minutes long, reviewing them takes a few hours. At first I find so many errors that I can't get through a single viewing without editing stuff. I find it helps to restart the MacBook since iMovie buffers a lot of memory, slowing things down to a crawl. Now I find that things work most reliably if I restart the machine after every session of iMovie. Even during a "session I'll restart if iMovie takes up more than 2 GB of RAM I restart the machine. When I go to sleep, if I've been working on iMovie, I just restart the machine. This way I know there will be plenty of memory for iMovie the next time I use it.
The slowness isn't the only issue - when iMovie buffers it won't save things, even things that I spent a lot of time doing. The problem is that I end up with random errors - in my standard two line text one line may have the default text in it, or the original font, or no outline, or something weird. Because I don't know what caused the error I have to assume that it could have happened anywhere in the clip. This means a long and thorough review of the clip.
Since most of my clips are 10 or so minutes long, reviewing them takes a few hours. At first I find so many errors that I can't get through a single viewing without editing stuff. I find it helps to restart the MacBook since iMovie buffers a lot of memory, slowing things down to a crawl. Now I find that things work most reliably if I restart the machine after every session of iMovie. Even during a "session I'll restart if iMovie takes up more than 2 GB of RAM I restart the machine. When I go to sleep, if I've been working on iMovie, I just restart the machine. This way I know there will be plenty of memory for iMovie the next time I use it.
The slowness isn't the only issue - when iMovie buffers it won't save things, even things that I spent a lot of time doing. The problem is that I end up with random errors - in my standard two line text one line may have the default text in it, or the original font, or no outline, or something weird. Because I don't know what caused the error I have to assume that it could have happened anywhere in the clip. This means a long and thorough review of the clip.
10. Once I think it's done I export into a movie format. Then I watch the whole clip 2-6 times. I usually catch even more errors here and have to re-edit the "master", re-export, and then review.
The export takes about 2-3 hours on my machine so I usually leave this to cook while I go run errands or go to sleep. Once I export then I have a much smaller, much more manageable file, something in the 1 GB size range. I can watch this using Quicktime and I do, reviewing it at least two times, sometimes as many as 10 or 15 times.
All too often I catch errors. Many early errors, that ended up in the published clip, were due to iMovie buffering heavily. When iMovie is heavily buffered it fails to save edits and such. At least three times I've published a clip and found, to my horror, that the credits text had reverted to the default text. This happened in the yet-to-be-published 2012 M40+ New England Crit Championships and it's happened in earlier clips. At first I thought I just forgot to put the stuff in but later I realized that I had painstakingly put everything in and yet it was all gone.
The problem is that if the credits failed to save then I have no idea what other stuff didn't save. This means really checking things over, at least a few times. Since I'm watching a .mov file I can't edit things right there and then. I don't want to open iMovie and start editing because I get confused between the two and inevitably make a mistake after an hour or two. Therefore I make handwritten notes while watching the now-rejected .mov file, then open iMovie to make the changes.
11. I note errors in the exported file and fix them. I often export a clip into movie format at least three times, representing 6-9 hours of computing time just for the export. This doesn't include the approximately 1-3 hours of editing and checking for each export, so 3-9 hours of time where I'm working on the clip. Some clips take more exports and very few take less.
Since creating a new .mov file means editing in iMovie and exporting again, it means, at the very least, an hour to three hours of editing (edit errors, check whole iMovie draft to make sure no other random errors popped up), 3-6 hours of export to .mov, then maybe an hour of checking the new clip.
Finally, when I think the clip is good, I let it simmer. This means I don't look at it for a while, maybe a day. Then I check it again, with a fresh mind. Yes, I usually find some glaring error so I have to redo it.
12. I upload the file to YouTube. After final reviews of my simmered product I'll upload. With a wireless network I found it that the upload would get interrupted or get really slow, especially if the MacBook was far away from the router/modem. I'd move the MacBook close to the router and find that connecting it via a network cable works best.
The uploads usually take about 300-500 minutes. The problem is that I can't edit the YouTube information while it's uploading, at least not so I can tell. Therefore I want to be around when the upload finishes so I can put in the right title, the right description, stuff like that. I find that even when I am totally aware of when the upload finishes, by the time I've updated the title and description I see that there's already been a few views.
The main problem is that our internet throughput goes down the tubes when I'm uploading a clip. It's a huge task just checking email. Therefore I try to upload at night, timing the end of the upload so that it finishes when we wake up. Or, in the last upload, I uploaded during the day while watching Junior.
So that's what it takes to do a clip.
I can accelerate it a bit with a faster machine - with the new MacBook, before iMovie stopped importing my ContourHD files, I rarely had a editing/memory error, rarely had weird fonts, rarely had those memory related errors. The 64 bit operating system allowed me to install (and use) 16 GB of RAM. iMovie will quickly grow to 4-8 GB on that machine. The old MacBook, with a 32 bit operating system, is limited to under 4 GB of RAM, and when iMovie hits about 2.5 GB things start going south. For now, with the iMovie problems on the new MacBook, at least I know what I need to do to get a clip up.
Coming up at some point will be the 2012 White Plains Cat 3-4 crit and the 2012 New England Crit Champs M40+.
The export takes about 2-3 hours on my machine so I usually leave this to cook while I go run errands or go to sleep. Once I export then I have a much smaller, much more manageable file, something in the 1 GB size range. I can watch this using Quicktime and I do, reviewing it at least two times, sometimes as many as 10 or 15 times.
All too often I catch errors. Many early errors, that ended up in the published clip, were due to iMovie buffering heavily. When iMovie is heavily buffered it fails to save edits and such. At least three times I've published a clip and found, to my horror, that the credits text had reverted to the default text. This happened in the yet-to-be-published 2012 M40+ New England Crit Championships and it's happened in earlier clips. At first I thought I just forgot to put the stuff in but later I realized that I had painstakingly put everything in and yet it was all gone.
The problem is that if the credits failed to save then I have no idea what other stuff didn't save. This means really checking things over, at least a few times. Since I'm watching a .mov file I can't edit things right there and then. I don't want to open iMovie and start editing because I get confused between the two and inevitably make a mistake after an hour or two. Therefore I make handwritten notes while watching the now-rejected .mov file, then open iMovie to make the changes.
11. I note errors in the exported file and fix them. I often export a clip into movie format at least three times, representing 6-9 hours of computing time just for the export. This doesn't include the approximately 1-3 hours of editing and checking for each export, so 3-9 hours of time where I'm working on the clip. Some clips take more exports and very few take less.
Since creating a new .mov file means editing in iMovie and exporting again, it means, at the very least, an hour to three hours of editing (edit errors, check whole iMovie draft to make sure no other random errors popped up), 3-6 hours of export to .mov, then maybe an hour of checking the new clip.
Finally, when I think the clip is good, I let it simmer. This means I don't look at it for a while, maybe a day. Then I check it again, with a fresh mind. Yes, I usually find some glaring error so I have to redo it.
12. I upload the file to YouTube. After final reviews of my simmered product I'll upload. With a wireless network I found it that the upload would get interrupted or get really slow, especially if the MacBook was far away from the router/modem. I'd move the MacBook close to the router and find that connecting it via a network cable works best.
The uploads usually take about 300-500 minutes. The problem is that I can't edit the YouTube information while it's uploading, at least not so I can tell. Therefore I want to be around when the upload finishes so I can put in the right title, the right description, stuff like that. I find that even when I am totally aware of when the upload finishes, by the time I've updated the title and description I see that there's already been a few views.
The main problem is that our internet throughput goes down the tubes when I'm uploading a clip. It's a huge task just checking email. Therefore I try to upload at night, timing the end of the upload so that it finishes when we wake up. Or, in the last upload, I uploaded during the day while watching Junior.
So that's what it takes to do a clip.
I can accelerate it a bit with a faster machine - with the new MacBook, before iMovie stopped importing my ContourHD files, I rarely had a editing/memory error, rarely had weird fonts, rarely had those memory related errors. The 64 bit operating system allowed me to install (and use) 16 GB of RAM. iMovie will quickly grow to 4-8 GB on that machine. The old MacBook, with a 32 bit operating system, is limited to under 4 GB of RAM, and when iMovie hits about 2.5 GB things start going south. For now, with the iMovie problems on the new MacBook, at least I know what I need to do to get a clip up.
Coming up at some point will be the 2012 White Plains Cat 3-4 crit and the 2012 New England Crit Champs M40+.
Posts
No comments:
Post a Comment