Worlds in collision

[rick55]

I'm back at the water drops (everyone groans ).

I figured out that if I increase the pump volume, I can send two drops in succession. It took some fooling around, but I got that to work.

I also changed the setup, based on comments from before, and the fact that I got an extension tube. Unfortunately, I don't have my good lens: I sent it off to Canon, and they did find an issue with the auto-focus, which they're fixing under warranty. But I couldn't wait for it to come back, so this is with the old kit lens, 18-55mm, pre-IS, with a 12mm extension tube, so what you see is just barely cropped. Exposure f/13, 1/125, two flashes. 580EX RF triggered, manual mode, 1/8 power, 430EX slaved. Both flashes are pretty far forward, about 6 inches from the reflector at the back of the pan, one on each side. On the left a 1/2 orange gel, on the right a 1/2 blue gel, to try to add some color gradient.

In 1&2, you can see some blown highlights in the ball/throat. That's a problem I'm having fairly consistently: if the exposure is at reasonable levels for the main part of the shot, certain shapes of the water apparently act as a lens, and cause this kind of thing. Any ideas?

And of course, any C&C always appreciated.

Cheers,
Rick

[carregwen]

Excellent shots, Rick. I really do have to try this (wife groans stage left) What are the white blow-outs - are they reflections of the flashes? What pump did you use? I think #1 &3 will take a bit more contrast/brightness to make them pop a bit more. Great stuff!

[rick55]

Thanks, Rob. I'll try pushing the contrast a bit more. The white blow-outs seem to be where the water acts as a lens, and concentrates the light that's coming from behind the water. All the light is coming from a piece of foam-board placed behind the tray: the flashes illuminate it.

The pump is homemade, using some check valves from Amazon (here), a lever to squeeze the tube between two check valves, and a servo motor to push the lever. I figured that would be more controllable than trying to use a regular pump and just giving it brief bursts of power.

Cheers,
Rick

[Dave Humphries]

In 1&2, you can see some blown highlights in the ball/throat. That's a problem I'm having fairly consistently: if the exposure is at reasonable levels for the main part of the shot, certain shapes of the water apparently act as a lens, and cause this kind of thing. Any ideas?

Only the obvious one; don't use water, or any clear liquid - use milk or similar instead, that'll stop any 'lens effects'

Good shots though,

[carregwen]

The pump is homemade, using some check valves from Amazon (here), a lever to squeeze the tube between two check valves, and a servo motor to push the lever. I figured that would be more controllable than trying to use a regular pump and just giving it brief bursts of power.

Cheers,
Rick

Right, I think they are what you mentioned before. I'll have to order one. This could presumably become addictive? Is there a help group for therapy?

[rick55]

This could presumably become addictive? Is there a help group for therapy?

It's definitely addictive: so many shots are so close, except for some spot in the wrong place, so you just try over and over. I'll be the first member of the group.

Cheers,
Rick

[carregwen]

It's definitely addictive: so many shots are so close, except for some spot in the wrong place, so you just try over and over. I'll be the first member of the group.

Cheers,
Rick

I know absolutely nothing about electronics, but it seems to me that if you could release the water drop electronically, and knew the time of it's fall rate, you could electronically fire the camer shutter at the right moment. No? You would surely get a higher hit rate at the very least.

[rick55]

Exactly: I use an Arduino controller board to drive the servo, and it also drives a relay that fires the camera. The variable speed of the servo arm controls the spacing between the drops, and then I have a variable delay between the start of a pump cycle and when the camera fires. By playing with those, I can get it to hit pretty regularly.

Cheers,
Rick

[carregwen]

Exactly: I use an Arduino controller board to drive the servo, and it also drives a relay that fires the camera. The variable speed of the servo arm controls the spacing between the drops, and then I have a variable delay between the start of a pump cycle and when the camera fires. By playing with those, I can get it to hit pretty regularly.

Cheers,
Rick

I hate to ask this... but I don't suppose you would like to package all of that up and start selling it... I'll be your first (loyal) customer...

[rick55]

I'll have to take a picture of it: you'd run screaming. It's all on a breadboard, jumper-wired together, and the pump is built on a small piece of 2x4, with the servo mounted on a couple of angle braces. I change it around every time I get a new half-a**ed idea, like putting out two drops. Not much of a product.

Cheers,
Rick

[CNelson]

Very nice work and I love the splash on a splash....great capture. Someday I'll have to take my splash shot...such great shots inspire.

Chuck

[rick55]

Thanks, Chuck. When you're ready to try it, check Jim B.'s picture of his setup in his album: I copied it shamelessly, except I used two flashes and the cobbled-together pump.

Cheers,
Rick

[Wirefox]

No 3 is an absolute cracker. My only gripe is that the incidental water drops could be removed in PP. I think I will give the drop making machine a miss though. I am a 'stoker' (Mech Engineer) by trade and the nearest I have come to producing anything like this is designing a constant volume feeder for a reprocessing plant. Far too much flow..but the kind of thing you can take a 2lb hammer to if it doesn't work.

Steve

[rick55]

My only gripe is that the incidental water drops could be removed in PP.

Thanks, Steve - I thought about removing them, but thought maybe they added to the feeling of action. The only things I removed were some random spots: I'd say they were droplets on my lens, but they were in different places in different shots. You've answered the question, though: they're obviously more of a distraction than an addition.

Cheers,
Rick

[Jim B.]

Rick,
The collision shots are outstanding! I still haven't produced anything close to these. Your pump set up is working great. I would love to see a picture of it.

[rick55]

Thanks, Jim.

I also told Rob I'd post a picture of the pump setup: think Rube Goldberg.

You can see the tray and the right-hand flash (430EX) with gel holder. The left-hand flash is just out of the frame on the left.

There are some red labels on a few things.

At (1) is the pump. The servo is on top, and the thing right under the label is the lever, which is a piece of UHMW plastic on top of a piece of steel. The servo turns about 40 degrees and back again in each pump cycle. You can see the high-tech system I use to keep the tube in the same place: a piece of string looped around the pivot point. The 3/8 in. inside diameter (ID) tube is just about a foot right in the area of the pump, with adapters at each end. To the left of (2), you can see a bit of the 1/8 in. ID tube that continues on, with one of the check valves.

At (3) is the breadboard with the electronics. The Arduino is directly to the right of the "3". The white cable with the red warning tag on it is a USB connection from the real computer. The LCD display was used to show what the parameters were when I was adjusting them with resistors rather than directly from the computer: I just left it because it provides a good indication that everything is right. The black thing at the far left of the breadboard is a reed relay, feeding the black cable that disappears off the bottom of the picture: it goes to the camera shutter.

The junk at the top, like the RF switch, is from other projects: I'm not very good at keeping my work area clean. I always have 5 things going at once.

If anyone wants to do something similar, I'll be happy to send along the code, circuits for the shutter, that kind of stuff. I am looking at an idea that would make it all a little less silly. Although in theory, the timing of a drop is perfectly predictable with the pump driven by a servo, it has problems in practice. For example, I discovered that I need to keep topping off the bottle it draws from, because when it has to pull farther, it doesn't pump quite as much. So I ordered a slot sensor from DigiKey. The opening is large enough that the drop won't touch the sides, and I'm hoping I can note the drop falling, and use that to set the shutter time, like the ShutterBeam product you pointed out. If it works, it would still require some electronics, but it could be used with a hand pump.

Cheers,
Rick

[Jim B.]

Rick,
Thanks for the picture and the details of your set up.I could probably put something like this together reading instructions,but to think the design through from the start,not a chance!

My hat's off to you! Looking forward to your refinement of the system.

Jim