Ghosting in DSLR frames of the Moon

[JamesF]

My usual practice for lunar imaging with my 450D and 127 Mak is to take frames of 1/1000th @ ISO800, but one or two people have been suggesting I should try longer exposures and lower ISO settings so last night I intended to do three capture runs of 1/1000th @ ISO800, 1/500th @ ISO400 and 1/250th @ ISO200. There was some cloud drifting around so I started with the second set and completed the third before I had to give up, meaning I had no "standard" image to compare against. When I set out to do the processing this morning the 1/500th images looked good and I'm happy with the image they produced.

The 1/250th subs though are a completely different hovercraft of eels. Every single frame has ghosting on the top half of the image and I don't understand why. If it were due to shutter vibration then I'd either expect it over the entire image or to alternate from top to bottom as my understanding of the shutter mechanism is that it operates alternately from top to bottom and then from bottom to top. Here's an example from the first frame. First a crop from the lower half of the frame (not the best one, but it serves to illustrate the problem):

and from the top of the same frame:

I've done no processing on these other than cropping and a bit of a histogram stretch to make the problem easier to see.

If anyone has any ideas as to the cause, I'm all ears.

James

[nytecam]

Is it shutter bounce in pic #2 ?

[JamesF]

Is it shutter bounce in pic #2 ?

Well that's what confuses me. If the shutter operates in alternating directions (which I believe it does), why would the problem only appear on the top half of all the frames. I'd expect it to alternate as well (or possibly only appear on half of the frames, when the shutter was travelling on one direction but not the other). This problem affects every one of 120 frames shot sequentially in exactly the same way.

James

[cgarry]

I have to say I disagree with the advice of slowing the shutter speed down and and lowering the ISO. I personally go for maximum ISO (1600), high shutter speed and leave the stacking to sort out the noise.

I think you need to go for the standard debugging techniques for these types of problem such as rotating the camera 180 degrees (and possibly 90 degrees) to see if the problem follows the scope or the camera.

I do wonder if it is down to tube currents in the top half of the tube. You could move the top of the moon to the bottom of the sensor and see if it still affected and a good old fashioned defocused star test should tell you lots.

Cheers,

Chris

[JamesF]

Rotating the camera 90 degrees or 180 degrees could well be a plan. If the clouds allow me time I shall try that next time.

I don't think it's air currents in the tube though. The scope had been out on the mount for eleven or twelve hours at this point.

James

[cgarry]

That should be enough time to reach thermal equilibrium!

Though I wouldn't rule out tube currents without a star test. I found my C14 still had tube currents after 7 hours outside, caused by the top of the scope being cooler than the bottom because the top was facing the open sky. Though I think this type of tube current issue only really happens to bigger scopes so your Mak should be fine.

Still, this is an interesting little mystery for you to get to the bottom of!

Chris

[IanL]

The near-perfect doubling of the features at the top of the frame looks more like a vibration issue than a seeing or tube current induced one, where I'd expect it to be a more random/blurry type of effect.

I'd definitely echo the suggestion of going for the shortest exposure you can manage, since you will freeze the seeing (and tube currents) better, and with a reasonably large image stack you will end up with a better result.

Are you shooting direct on the camera or using a laptop for the job? If you are using software (e.g. BackYardEOS or APT) make sure you put in a reasonable ant-vibration pause for each single exposure. It is the mirror lifting up that causes the vibration, and what the pause does is lift the mirror and then wait for vibration to stop before triggering the shutter. Try longer pauses (5s, 10s, 15s , etc.) to see whether you can get better results. (You can't do anything about vibration from the shutter going off in a single exposure, since the shutter has to go off!)

You should be able to use live-view recording in APT or BYE with a 450D. You won't get the nice, big crispy images you get with single shots, but with an AVI of hundreds of frames, you may still get better results when stacking than with a small number of still frames.

[JamesF]

Ah, that's worth a try. I use APT and take a hundred frames or more to stack, so I can certainly experiment with the anti-vibration pause. Not sure about live-view recording. I can't say as I've ever tried it. Does it record at full resolution? That might be a bit too much for my poor laptop's hard disk.

James

[IanL]

I haven't used liveview recording in APT, only in BYE. I don't recall if the full frame at 100% resolution or not, but I think it is. You also have the option to do a 5x zoom which is more useful for planetary. It is in fixed 'chunks' on the chip as the camera firmware does the zooming not the software. You get a higher frame rate and smaller files, but you have to frame appropriately and avoid drifting out of shot.

The only downside to liveview recording (i.e. 'Planetary Mode') is that it is at a fixed ISO; BYE can simulate other ISOs but I have found that some targets (e.g. Saturn) are too dim to work well.

[fwm891]

James, This the problem is vibration with the scope caused by the shutter action. To start with when the shutter begins to open the mass of the scope resists the motion of the shutter which is why you get sharp one end/side (depending on shutter direction). But the vibration is transmitted throught the OTA causing the blur at the end of the travel. To every action there is a reaction.

The amount and size of the blur will depend on the resonance of the shutter speed/scope combo.

Thats one reason the mirror should be raised when shooting with long lenses (and a scope is a long lens).

If you rotate through 180 then the effect should appear to be from the other end but is actually the same reversed.

[JamesF]

I'm struggling to understand why that should be, Francis. The double image is always at the top of the frame, but my understanding is that the shutter alternates in direction. That being the case I'd expect the double image to alternate from top to bottom. I'm not saying you're wrong. I just don't understand how DSLRs work well enough.

The same applies to mirror-induced vibration to be honest. If the mirror is causing vibration then I'd expect the part of the sensor that the shutter uncovers first to have the double image and that it should therefore also alternate between the top and bottom of the frame. Quite possibly I'm just not understanding the problem properly, but as APT allows me to set a pause before taking the image that's an easy thing to try straight away.

James

[Rhys Hardwick]

Hi James

In focal plane shutters, a first veil moves down from the top to open the shutter, then another follows down to close the shutter. This allows each area of the film/sensor to receive equal amounts of light.

This may answer why the top is more affected. 1/250 s is very quick for vibration to settle though...

[Rhys Hardwick]

And it resets to the top between shots, I should mention

[JamesF]

Ah, now that would make sense. I was certain I'd seen some high-speed film of the shutter operating in opposite directions each shot. Perhaps I imagined it

James

[Rhys Hardwick]

I'm sure you didn't. The sutter resets to it's rest position immediately, which is probably what you saw. I tried to find a link on it, can't find it now. It's particularly interesting when shotting moving objects, at high rate, and the top of the frame can look 'behind' the bottom.

[JamesF]

To close out on this, two nights ago I shot a number of frames with no mirror lockup delay and then a full set with two seconds delay, all at 1/200th exposure.

There is no apparent problem with the frames taken after a two second delay. About half of the frames taken without the delay show a ghosting problem.

I think I'm happy that the cause of this issue is vibration due to the mirror movement. Perhaps it's slightly marginal or made more obvious by periods of poor seeing. I shall stick with the delay in future.

James

[cgarry]

Well done on getting to the bottom of this. I am going to try some short exposures when I get the chance to see if my 450D shows the same issues. It is nice to have the solution available before I have had the problem though.

Cheers,

Chris