Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Timelapse Tests?

S

Steve Thomson

Well-known member
Joined
Feb 16, 2007
Messages
57
Reaction score
0
Points
0
Age
56
Website
www.timelapse.co.nz
Has anyone who has received a camera experimented with Time-lapse yet?
Keen to hear of any RED time-lapse experiences/accessories/configurations...
Have been advised that for a stable tripod mount should buy ARRI baseplate?...
 
I have shot extensive timelapse.

All I can say is, "wow".

All of it has been on a Cartoni Delta. Some of my shots are in the MammothHD reel...

http://www.mammothhd.com/MHD_RED1.html

-Thor
 
Thor has 2 timelapse shots in the MHD reel (Sundance and the POV driving shot), with more in the TimeLapse Gallery... and we're currently reviewing and editing a whole set of new material that just came in...... :-)
 
Timelapse on this camera is a dream.....! I've never had to spend so little time in post getting clips out. New Zealand is the place to be with this camera Mr. Timelapse... ;)

-Eli




mrtimelapse said:
Has anyone who has received a camera experimented with Time-lapse yet?
Keen to hear of any RED time-lapse experiences/accessories/configurations...
Have been advised that for a stable tripod mount should buy ARRI baseplate?...
Click to expand...
 
Has anybody been able to do 1/2 or 1 second exposures in any build in 4K or 4K 2:1?
 
DSLRs are not exactly frame accurate. Most of them don't clock at an exact FPS even when computer controlled. My experience with the Canon EOS 10D was that it didn't shoot fast enough to give a fluid motion to whatever I was filming. It either ran out of buffer or just couldn't do it.

Shooting RED hugely benefits from easy access to accessories and lenses. Fitting a 10D to a bridgeplate and mattebox to use graduated filters is not exactly easy nor convenient... though it makes for good pub talk. You also benefit from seamless integration with the footage you already shot and don't have to deal with differing curve characteristics and grain structure. You also save a bit of post because you don't have to convert between formats.

My two cents.
 
The KoNE said:
DSLRs are not exactly frame accurate. Most of them don't clock at an exact FPS even when computer controlled. My experience with the Canon EOS 10D was that it didn't shoot fast enough to give a fluid motion to whatever I was filming. It either ran out of buffer or just couldn't do it.

My two cents.
Click to expand...

Good info. So what's your typical time interval between shots?
 
I find when shooting clouds and landscapes something between four to eight frames per second at a 180˚ shutter gives good results. Exposes each frame for 1/8th to 1/16 of a second which allows a smooth motion blur which helps keep wiggling phone lines and tree leaves at bay. When shooting the classic city nighttime and cars I feel going with frame blending in post over more or less the same FPS gives the best looking results. Takes away from the jerkiness of city motion. Those are the looks I usually go for, but then you just need to bring the shutter down and shoot a bit faster to get the opposite results. I personally like my timelapses to be slow and fluid (creamy?)... but at the end of the day, the script call for the shots, not me. :)

The 10D just couldn't cut it. Best frame rate was 3 FPS up to 9 frames in RAW mode. Too slow AND buffer ran out... when manually controlled. On the computer, the bundled app doesn't let you shoot to CF so you have to transfer all of your 9 mb frames at USB 1 speed as you shoot them. Dreadful. I couldn't get any better than a frame every five to nine seconds. The software also wouldn't let you transfer as you are exposing. Since the camera itself didn't sport an intervalometer... It was poo. (Good poo for non-motion nevertheless)

The 10D is OLD by now and just wasn't the right tool for the job. Still... I feel most DSLRs just can't deliver the goods. They make for great practise when you don't have a motion picture cam though.

With a RED, a RED drive, a mains adapter and a UPS shooting time lapse would be such a breeze.
 
The KoNE said:
DSLRs are not exactly frame accurate. Most of them don't clock at an exact FPS even when computer controlled. My experience with the Canon EOS 10D was that it didn't shoot fast enough to give a fluid motion to whatever I was filming. It either ran out of buffer or just couldn't do it.
Click to expand...

What does this mean, not frame accurate? If you set your intervalometer to 1fps and have 240 resulting frames, you just drop them into a 24p timeline and it should be fine, right? With each frame being progressive, it's up to you the speed of the timeline. Or are you talking about moco or something?

I'm surprised that you would shoot EIGHT frames per second with clouds and landscapes. I've never shot faster than 1fps. Even one shot every 3 seconds can give you silkly smooth clouds, if they are not moving too fast.

BTW... that Mammoth/Red Reel 1 was gorgeous! Some beautiful shots in there. That was amazing how friendly that manatee guy was. :biggrin:
 
cinepost35 said:
Has anybody been able to do 1/2 or 1 second exposures in any build in 4K or 4K 2:1?
Click to expand...

Yes, many of the shots I did for Mammoth were at 1/2 second, specifically the shot looking down Park City Main Street, and the freeway shots from the moving vehicle.

Some of the new shots I just sent to Mammoth were also shot with 1/2 second, or 1 second shutter speed, but they are not posted yet. (Thanks for the nod, Clark.)

Hopefully, the next reel will show some of these shots. You could sign up to be on the Mammoth email list to know when the new reel is released. Come to think of it, Mammoth's reel is one of the best ways to see what the Red can do in a variety of situations. Any you never know, it might make more sense to license a shot down the road than to shoot it yourself...and if you are a Red owner, you will know that the footage will cut back and forth with the stock shot, and you never have to worry about it "looking right" because it was also shot on a Red. :biggrin:

One quick thing to note about the 1 second shutter... the camera "lets" you shoot on 1 second shutter with a 1 second interval, but the file ends up being a continuous still shot of the first frame. It's a bug that Red is aware of, and they are working to fix. So, if you want 1 second shutter speed, you have to shoot on 2 second interva.

Also, Build 14 does not include 1/2 or 1 second shutter speed. The slowest is 1/8. Only Build 13 has the slower shutter speeds. Hopefully, Red will bring them back in build 15.

-Thor
 
joelnet said:
What's the advantage of doing timelapse with RED as opposed to a DSLR?
Click to expand...

Not all DSLRs are like this... but the Nikon D200 will ONLY shoot 999 frames using the preprogrammed interval. Red has no such limitation. That is one of the immediate advantages I can think of, at least for the D200 camera and the rest of the Nikon models that are "downstream" from it. Not sure about the higher end models.

-Thor
 
What I mean by not frame accurate is that even with an intervalometer, your DSLR is unlikely to snap its picture exactly when it's supposed to be. This results in an uneven amount of time between frames which shows up as a certain jerkiness of motion in the final picture. Think of it as shooting a non crystal-sync camera.

As for shooting 4-8 FPS, like I said I like my timelapses slow and tend to use frame blending in post. I feel it gives me a bit of extra flexibility and looks good. Would you rather slow down your timelapse in post with motion estimation or speed it up with frame blending? Well, that's how I see it.
 
The KoNE said:
What I mean by not frame accurate is that even with an intervalometer, your DSLR is unlikely to snap its picture exactly when it's supposed to be. This results in an uneven amount of time between frames which shows up as a certain jerkiness of motion in the final picture. Think of it as shooting a non crystal-sync camera.
Click to expand...

But these are single, individual frames, so you can drop them into a timeline at any speed. How is this important unless you are trying to match moco or something? I can't imagine much jerkiness resulting from tiny fractions of seconds in terms of interval times. Is this basically talking about shots that are more than 1fps?

As for shooting 4-8 FPS, like I said I like my timelapses slow and tend to use frame blending in post. I feel it gives me a bit of extra flexibility and looks good. Would you rather slow down your timelapse in post with motion estimation or speed it up with frame blending? Well, that's how I see it.
Click to expand...

I haven't shot much timelapse on video. So frame-blending kind of ads to the exposure? I'm not familiar with it.
 
Thor Wixom said:
Not all DSLRs are like this... but the Nikon D200 will ONLY shoot 999 frames using the preprogrammed interval. Red has no such limitation. That is one of the immediate advantages I can think of, at least for the D200 camera and the rest of the Nikon models that are "downstream" from it. Not sure about the higher end models.

-Thor
Click to expand...

Thats 999 frames per folder. You can set up to 9 folders at a time.
I have shot 3 x 900 frames or 2700 frames in one session.
 
Tom said:
But these are single, individual frames, so you can drop them into a timeline at any speed. How is this important unless you are trying to match moco or something?
Click to expand...

Yeah sure, and they'll play back at whatever speed you want... but that doesn't change the interval between which they were snapped. On my 10D it was quite obvious as the difference between each shot could vary from 2-3 seconds even though all the shots' exposure time was the same. Lemme put it into numbers. You have motion going on that's numbered from 1 to 18. Brackets represent what's captured in a frame "crystal sync" capture of this would result into this :

[01 02 03 04] 05 06 07 08 09 10 11 12 13 15 16 17 18
01 02 03 04 05 06 [07 08 09 10] 11 12 13 15 16 17 18
01 02 03 04 05 06 07 08 09 10 11 12 13 [15 16 17 18]

An equal amount of time is lost between frames.

Non accurate capture would result into this :

[01 02 03 04] 05 06 07 08 09 10 11 12 13 15 16 17 18
01 02 03 04 05 06 [07 08 09 10] 11 12 13 15 16 17 18
01 02 03 04 05 06 07 08 09 10 11 12 13 15 16 [17 18 19 20]

Between frames one and two two units were lost.
Between frames two and three four units were lost.

I felt that that variable gap was noticeable. I felt it produced the same effect at faster speeds too, but sure it's not as noticeable. It bugged me nonetheless.

As for frame blending, when you speed up video instead of just skipping frames you can group and fade through them.

Here's a convoluted example. Lets say you shot pictures 1 to 10 but you want them to play twice as fast. You can blend several frames into one group, then you alternate the frames in that group to create a new frame.

[01 02 03 04] 05 06 07 08 09 10
01 02 [03 04 05 06] 07 08 09 10
01 02 03 04 [05 06 07 08] 09 10
01 02 03 04 05 06 [07 08 09 10] ...

Frames in [brackets] are part of the group
Frames in yellow are equal power
Frames in pale yellow are shared power
Effectively for group #1, the frames would have opacities of [(1/6) (1/3) (1/3) (1/6)]

Frame blending when you slow down your footage does just the opposite.

[01] 02 03 04
0[1 0]2 03 04
01 [02] 03 04
01 0[2 0]3 04
01 02 [03] 04
01 02 0[3 0]4
01 02 03 [04] ...

As opposed to frame skipping which would just go

[01] 02 03 04 05
01 02 [03] 04 05
01 02 03 04 [05] ...
for speeding up

and

[01] 02 03
[01] 02 03
01 [02] 03
01 [02] 03
01 02 [03]
01 02 [03] ...
for slowing down
 
You must log in or register to reply here.
Back
Top