Perseid and the EX1

[Charles Young]

I plan on recording what I can of the event.

I do have a few questions I need to resolve before I try. I would like to point the camera at the constellation Perseus with a fairly wide field of view and focused on infinity. I want to record continuously maybe at 1 frame per second. I don't think any single event will take more than a second (maybe) but don't want a delay between one frame and the next (might miss one or catch a partial event).

In EX Slow Shutter mode the options are 16, 32 and 64.
I assume that would mean each frame would consist of 16, 32, or 64 frames and would be recorded:

At 60P and 64 added frames=1 frame every 1.07 seconds
At 60i and 64 added frames=1 frame every 1.07 seconds
At 30P and 64 added frames=1 frame every 2.14 seconds

I have to wonder the time between when the 64th frame of one image and the 1st frame of the next image.

If I go to S&Q motion and set it to 1 frame per second and turn the shutter off, will I get similar sensitivity as above? Is the shutter somewhat equal to 1 second? Is there a big gap where the shutter is closed between frames?

Briefly experimenting, the 64 EX Slow Shutter is considerably more sensitive.

Humm...

And then there is the interlaced or progressive question

Are there other setting I need to tend to or that would improve the capture?

Thanks in advanced

Chuck

http://science.nasa.gov/headlines/y2...erseiddawn.htm

[Justin Carlson]

16, 32 & 64 is a frame accumulation.
So if you're wanting to do a timelapse and record 1 frame every second, the camera will then record 16, 32 or 64 frames of 'light' and apply it to that one frame. It doesn't change the frame rate at all.

[Dave Elston]

I think the Persieds last more than a couple days and most nights from now until @20th August should be good to catch sight of some heavenly fireworks - why not spend an evening (or two) before the peak night comes to run some tests.

As far as the look you want, I think the shutter angle is a useful guide to figuring out what to expect and how to get the result you desire.

Unless I am mistaken (?...)

64 frame accumulation:
@60p = shutter speed of 1/0.935
@30p = shutter speed of 1/0.46875

So if you record using the slowest setting (64) at 1 frame per 1 sec interval, the effective shutter angle becomes just over 385° (for 60p) or 768° (for 30p) - in both cases you'll get an overlap of motion blur from each recorded 'frame' to the next. This may or may not be an effect you desire, with time-lapse you do have a relatively open license (horses for courses).

With a meteor shower, such large shutter angles would result in each event leaving a distinctive trail across the sensor spanning up to 3 captured frames (~3secs realtime). Note that upon full speed playback you would clearly see these but at the expense of detail of the meteor itself - though in this case the size of the objects and their motion-path will be pretty much 'uniform' anyway.

If you wanted to reduce the motion blur to something approaching 'normal' rate (ie 180°) then you could use the 32 frame accumulation (for 60p) or 16 frame (for 30p) - in both cases giving an effective 192° shutter angle. This would give you greater 'definition' of the meteors but reduce the trail length while also potentially introduce 'normal' shutter gaps in the motion blur for events that last more than a second.

Let us know how you get on, and please post some clips when the show(er) is over.

Good Luck!
Dave.

[Dave Elston]

BTW This is a GREAT opportunity for those lucky new EX3 owners looking for a subject to do some time-lapse trials, its free, its different every night and, providing the sky stays clear, you could get a full night of activity captured from the comfort of your own back yard!

This year's Perseid shower is shaping up well - so far, data seems to suggest you could see up to 20 per hour (tonight) peaking up to around 60 every hour by 12th/13th.

More here...

http://www.imo.net/live/perseids2008/

[Alister Chapman]

I used 16 and 32 frame shutter combined with 1 frame every seconf timelapse to capture the Northern Lights earlier in the year. Some of my footage has satellite passes and the odd shooting star, so you should be good to go. I would suggest 0db gain and 32 frame shutter as a good starting point. You will see more grain/noise with longer slow shutter speeds.

[Charles Young]

Originally Posted by Justin Carlson

16, 32 & 64 is a frame accumulation.
So if you're wanting to do a timelapse and record 1 frame every second, the camera will then record 16, 32 or 64 frames of 'light' and apply it to that one frame. It doesn't change the frame rate at all.

So what you are saying is in HQ 60i or HQ 30p the EX1 can accumulate 64 frames in 1 second? That is amazing.

[Justin Carlson]

Originally Posted by Charles Young

So what you are saying is in HQ 60i or HQ 30p the EX1 can accumulate 64 frames in 1 second? That is amazing.

... almost ... 64 frames into 1 frame. It's great that way. I've done a lot of night time lapses that way.

[Kevin Shaw]

But wouldn't capturing a shooting star image work best with limited frame accumulation? If you do a 16 frame accumulation you'll reduce your odds of "catching the moment" by a factor of 16, and with only so many shooting stars per hour that seems counter-productive. I'd say start with a two or four frame accumulation and go up from there only if that doesn't yield a bright enough image.

[Justin Carlson]

I think it's the other way around.
By accumulating 16 frames and applying it to 1 frame per second you increase your chances of getting the star (and it being brighter) as the camera is almost always taking in light to apply it to that 1 frame.
If you're recording at 30p and then do a 2 frame SLS to record to your 1 frame per second time lapse, then there's still another 28 frames per second that your camera is not recording and could 'miss the moment'.
Where-as if you're recording in a 32 frame accumulation, then there's no chance you'll miss the light of the star as it's falling.

[Kevin Shaw]

I guess it all depends what kind of look you want: with a long frame accumulation you'd get the equivalent of a still image with a single streak of light across it, versus capturing the movement of the shooting star in traditional video form. Plus with frame accumulation you'd bring out the brightness of the background stars without getting any more light from the shooting star itself, which would then look dimmer by comparison. Or am I missing something about how frame accumulation works?

[Alister Chapman]

I don't think you'll find the EX sensitive enough to capture stars without using the EX Slow shutter. If you use 32 EX slow shutter and 1 frame every second time lapse you won't have any gaps in your recording so you should not miss anything. It also means that you can playback hours of action in just minutes. The shooting stars we captured while filming the Northern Lights do appear as streaks but the longer lasting ones are over 2 or 3 frames so they appear to streak across the sky.

[Charles Young]

I am trying different settings as we speak. In 64 accumulation it is recording audio. Interesting...

[Charles Young]

Any idea what the white balance temperature should be?

[Justin Carlson]

I think 4000 is what is considered moonlight.

[Charles Young]

as it turns out, In 64 EX slow shutter every 64 pictures (@60i 1080) are the same. The noise is not acceptable, it needs noise reduction.

Did get a meteor though.