bob,

what's a 'cheap' motorised head worth?

would love to do some time lapse round here with motion....

leslie

OK - I'm new at this, so please bear with me. So should I get up the first morning, wait for the sun to rise above the horizon, then set the exposure to auto, then remember that setting.

Then the next day, set the exposure to manual for whatever it was good at the day before, and use that manual setting for the whole video?

I filmed during the sun rise with some tips from some guys on this forum and some help from John Meyer. What I did was show up about 30 minutes before sunrise and filmed for about an hour (I think- I know it was over 30 minutes) with the camera capturing directly into the laptop via Vegas' external capture program (I would have used DV Rack but it was only SD). I then rendered it out using Ctrl+Stretch and I repeated this until I got it down to 10 seconds. I did not directly film the sun, but if I would have moved the camera over about another 2 inches it would have been in the shot.

I just used everything set to auto.

You can see the results here. It is the second video from the top and it is in DivX format.

As for the paning, you can get the add on for telescopes that will do this for you. It keeps the telescope lined up with whatever you have it focused on while taking into account the rotation of the earth.

For a cheap pan without speeded up footage, you can try a turntable or a lazy susan for unmotorized.

j razz

All up cost me AUD 220.
Have a look here:
http://mizar-optical.com/tele/parts.html
I bought the KD Motor mount.

Pan time is about 15min or 30min for 180deg depending on setting of Slow or Fast on the controller

I just need to find the time to get someone to machine an adaptor plate so I can mount it onto a Miller 75mm bowl and nut that I bought from Miller and I'm good to go. Even bough a couple of circular spirit levels.
Once I get it all together it'll be available for a loan.

Bob.

j razz - Cool video, thanks for the example.

So you filmed for 30-60 minutes or so, and sped it up to the 7 seconds? Just want to make sure I understood that correctly.

And you said you used the auto setting. Looking back do you think it would have been better with the manual setting described above?

Don't get me wrong, your sunrise was very nice, just wondering how it would have looked with the manual setting.

Of course I would have done it different had I more time to experiment. I was willing to risk using auto as I had no standard by which to set my camera. Plus, this was a "hmm, this might be neat" shot. It did not have to go into the video and it was not on the list of shots I had to get. So, no worries if it didn't come out looking descent.

I probably should have gone back and did one the next day and try out some manual settings but there's just something about getting up that early when I don't have to that beckons me to stay in bed that extra hour :)

And yes. You are correct: I filmed between 30-60 minutes and sped it up to ~7 seconds. It would have been better and longer had I of gotten there earlier. I do know that I cut away some of the footage that was on the end as the sun had risen and the hues stopped changing. So that is why I say between 30-60minutes.

j razz

Good morning Kevin. Here's another approach and it yields HD footage. I used the intervalometer in my DSLR to take a 300 photos in an hour. Then I loaded those into Vegas as a sequence. This was just a "proof of concept" so the content isn't all that interesting but here it is: http://www.vimeo.com/1117449

Jerry

If you look at the waves you'll see a common problem with timelapse. As you speed up the video you increase the effective shutter speed by the same amount. Adding a lot of motion blur in Vegas and/or a slow shutter speed will fix the problem and reduce fast motion to a blur.

Bob.

Bob, that is nice to know. Next time I'll add a neutral density filter so I can use a slow shutter speed.

Jerry

I don't see how effective shutter speed would in any way be affected. If you do the timelapse by speeding up on the Vegas timeline, all it is doing is dropping frames. This is definitely true if you turn off "resample" (something I would definitely recommend). Each individual frame of video is not altered in any way.

When you say "shutter speed" are you perhaps referring to something else besides the time that each frame of video is exposed? I actually don't know of any way to decrease that (which would eliminate motion blur) and even attempts to increase it (by blurring video frames) don't really do the same thing.

So, it is my belief that if you take a frame of video from the final timelapse, and you can find the original frame of video from the original footage and line them up above each other and then A/B between them, you will see absolutely zero difference, especially if resample is disabled.

BTW, I use a laptop and Scenalyzer for all timelapse. It gives you the ability to directly do the timelapse with no further work in post production, and you can do amazingly long timelapses, such as a building project, flower growing, etc.

John, what I took from Bob's comment was that the waves in my time-lapse would look better if there had been some motion blur in the original frames, as you would have if you used a slower shutter speed during capture. This would presumably give them a more organic look, along the lines of what you see with the clouds in my clip.

Jerry

Thanks for the information guys, and Jerry, your time-lapse video was cool, but yeah, the waves kinda took something away from it.

Now I'm trying to figure out how I can avoid the problem that J-Razz had of not getting the sun in the picture.

I know the Sun rises in the East, but East is a big place and it's not exactly due-East, and it changes with the season. So does anyone know of a chart or something that will tell me exactly where the Sun will rise at a given location at a certain date?

Edit: I know the "easy" way is to get up the first morning, make a note of where the Sun comes up, and then the next morning do the filming, but there's gotta be a chart online...

I would do 30 minutes before sunrise time, and go to 30 minutes after.

Unless you are very good at manual settings, I would go the day before at 30 min after sunrise and remember that number. The following example I didn't have it set right and kept closing the iris about 6 times throughout. Correcting helped a bit, but it is a really big pain!

This was about 40 minutes of the footage condensed down to one minute. Took a couple of sets of renders to shrink it down.




Edited: The HD version is very cool, you can see the outside elevators at a hotel go up and down during the night section.

"I know the Sun rises in the East, but East is a big place and it's not exactly due-East, and it changes with the season. So does anyone know of a chart or something that will tell me exactly where the Sun will rise at a given location at a certain date?"

Actually, yesterday (Sept 22) the sun DID rise exactly due east (equinox / first day of Autumn in northern hemisphere). It will now creep a little bit south each day until the first day of winter (Dec 21), when it reaches its southernmost rise and then begins moving to the north again.

If you are going to shoot this week, you will probably be okay to just make sure that "due east" is to the left of center in your viewfinder, as the sun will rise on an angle that takes it to the upper right of the picture (assuming northern hemisphere) The amount of angle as it rises depends on your latitude -- the further north you are, the more of an angle from vertical you will see (I think).

(But let me finish by saying I'm NOT an expert at any of this... Hopefully somebody smarter will present better details...)

Tim L

UPDATE: Here's a cool sunrise calculator link that's so simple I can understand it:
http://www.timeanddate.com/worldclock/sunrise.html
Once you've selected the city and click "See sunrise/sunset", note that you have to select Columns: "Rise/Set Time/Azimuth" and then click "Show" to get the angle info.

fldave - thanks, yours was very cool, especially the boats going down the river, that was neat. I would have started it at about 20 seconds though, but still very neat.

Tim L - thanks for the information, that's great to know. I don't care if the sun is dead-center of the video, I just didn't want to miss it completely. I'll point the camera due-East and get it in the frame.

I was afraid maybe it would be too far SE or NE and miss it either way...

I'm surprised that there's no script to create a time lapse sequence out of real time video, or is there? That would be a neat FX for Vegas The advantage would be that you could probably start on a specific frame and set a variable interval, which then you could tweak again and again until you got what you originally envisioned. The problem with doing a TLS in camera is that if something goes awry, that's the end of it, back to the drawing board.

A really cool time lapse video is http://www.nmedia.com/777-200LR_Boeing.wmv b777 built in 4 minutes[/link]

I'm surprised that there's no script to create a time lapse sequence out of real time video, or is there?

Not a script but Scenalyzer has a time-lapse capture feature.
From the site:

TIME-LAPSE CAPTURING

* With ScenalyzerLive you can use Time Lapse Capturing to create impressive effects that can be used in your final video.
* Record e.g. one hour of a Sunset and then use ScLive to capture the video into an e.g. 15 seconds .avi file which will show the sunset 240 times faster.


Mike

Only goes to prove that one should RTFM. I own a paid for copy of Scenalyzer live but was oblivious to that capability. Certainly ScL, at what? $36 US is a bargain indeed.

I have a script, posted over at VASST, which lets you take one frame from each event in order to create timelapse. This was intended for use with camcorders which have a "timelapse" feature which actually records about five seconds of video, then stops, then a minute (or some settable interval) later records another five seconds. Obviously this doesn't produce a very good result, but if you instead take just one frame from each event, you end up with the real thing.

Taking single frames from video to produce a timelapse kind of kills the effect. Consider a film camera, the shutter is locked to the frame rate. If you undercrank the camera to 1fps the shutter is open for 0.5secs.
Conversely if you shoot video at 60i your shutter speed is 1/60th. If you took 1 in 30 frames to create your timelapse then you got an effective shutter speed of 1/1800 which will render say a car travelling at 60mph very strangely. It should look like a car traveling at 1,800mph i.e. a blur.
There's some example of how it should look here:
http://www.vimeo.com/1269325
Can't find any of Phil Bloom's work but I know he's done some great timelapse with the EX1.
Now you can get exactly the same result using Vegas's Motion Blur envelope. It does exactly what the frame accumulation mode does in the EX1, it stacks frames over time. Both methods however still are not optically correct, the best results would come from a DSC and an intervalometer with the correct shutter speed.
Possibly one could use one of the fancier AE plugs to get an optical correct result in post from video.
On the other hand if there's nothing in front of the camera that moves quickly then you can ignore all of the above.

Bob.

It's a hundred times easier to film a sunset and reverse it. I did this for the opening of a movie a couple of years ago.
http://www.weatherstonstudio.com/Galleries/video/zorg/zorg.html
I was the DP but I don't think it was my idea. I think we tried to shoot it earlier in the production and completely missed the sun. But I didn't have any waves or leaves to deal with.
Good Luck!

. If you took 1 in 30 frames to create your timelapse then you got an effective shutter speed of 1/1800 I just don't see how that is right. Each frame is independent of every other frame, whether on film or on video. The time during which each frame is exposed is determined by the shutter in a film camera, or by the time a charge is accumulated in the case of most modern digital cameras or camcorder. I really don't see how the number of frames taken per second in any way changes the effective shutter speed.

I think you may be confusing yourself with what happens in high speed cameras. In that case, when a camera -- through whatever process is used -- takes hundreds or thousands of frames per second then, by definition, each frame has to be exposed for far less than 1/24 or 1/30 or 1/60 of a second. Otherwise, there wouldn't be time to advance and expose the next frame of film.

However, when you take video using normal settings and then simply decimate the result, there is absolutely no change in shutter speed, "effective" or otherwise.

Here's an example of what I am talking about.

If you film a car traveling sixty miles per hour using a normal camcorder, and then film that same car traveling thirty miles per hour using the same camera and the exact same settings, but then drop every other frame in post production, you will not be able to tell any difference whatsoever between the two shots. The car will be in exactly the same position in every other frame of the thirty mph shot as it is in each frame of the sixty mph shot, and the shutter speed of both shots will be the same because you used the same exact settings in both cases.

The two shots will be therefore 100% identical in the end, and you will not be able to tell one from the other.


[edit] Hmm... OK, the 60 mph car will be blurred more in each frame, won't it? Is that what you are talking about? I guess you will be able to tell them apart. Interesting ...

[further edit] But if I double the shutter speed of my 60 mph shot, I will then get the same effect. This says I should always use really high shutter speeds when doing timelapse, yes?