Using Canon EOS 1000 as webcam imager

[MishMich]

I have posted my first results using the EOS as a webcam, using a GNU-licenced package called EOS Camera Movie Recorder. These are done through an SPX 200 f6, with 2x barlow lens in the T-adapter, with video magnification set to x5.

I've been struggling to get an image out of the s/h Opticstar PL-150C or the Trust HD Sonix/CMOS webcams I have - but this hack produced much better images than I have managed with those.

This is for info more than anything, to show that if you have a DSLR already, but not a webcam or planetary imager, this might be worth considering. Not sure how trustworthy the source is, but I downloaded via SourceForge, and I tend to feel more comfortable with GNU-licenced software. Be sure to check for viruses etc. if you do try it.

I found that for Jupiter, I could get down to ISO200 set at 0.1-1 second exposures and it reported 10 fps. On Saturn I had to go about ISO400 set at 1-10 second exposures, which reported 20 fps. Don't understand how the exposure and fps relates, because what i see on the AVI relates to what I saw bubbling along in the viewfinder as well. Stacks OK in Registax on Windows 7 if you set it to Vista Mode and Extended Mode for all AVIs. Resulting planetary image is slightly larger than what I achieve through 1280x1040 webcam, when set at 100% - so pixel-for-pixel, the EOS seems to record slightly more than an HD webcam (or VGA). Have to have the 5x zoom selected - either on camera or software, otherwise a the whole sensor frame is recorded in a smaller image than the 5x zoom image. Don't quite understand it myself - but seems to work.

I'm new to this - so apologies for the poor quality, but believe me, this is better than anything else I've managed with planets so far.

All images processed in Gimp - using sharpening, curves, plus 20% dodge and burn to enhance where necessary.

M.

M.

[PH-R]

Great images, And you have an eclipse as well!

Best wishes,

Paul

[resonator77]

I could not take anything similar with my Canoon 1000D

that is very impressive images for DSLR

[MishMich]

Nor could I, but with this piece of software you effectively turn the EOS into a webcam with a large sensor and wide field of view.

The details are here:

Canon-40D-webcam

(he uses a 40D, but this works the same with any EOS with live-view, such as the 1000D)

M.

[themos]

I've been playing with the same software and my 450D. It can certainly record. What would be great is a way of turning this frame stream into something that other packages can use the way they use a webcam (guiding, PEC, drift analysis, focusing automation, collimation automation etc)

[themos]

Maybe this will do...

VH Screen Capture Driver

[MishMich]

I've posed the question at wxAstroCapture. Problem is, this being a hack, I doubt Canon would support it (let's face it - the 500 is £200 more than the 1000, with AV recording built in). So, while I can see that GNU-licenced packages might develop on this, I don't know that commercial developers would touch it.

How do you envsion the VH Capture Driver working to facilitate guiding, out of interest?

Have you tested this out on DSO's, by any chance? I've only tested on planets, as I figured the exposure would be too limited for DSOs, as they are too dark to show up on the viewfinder,

M.

[themos]

VHCD can provide a WDM-style video source for other applications.

[Taff]

Hi' nice capture with a moon in transit.Mark

[MishMich]

Thanks.

I converted 3 AVIs to 1950 JPEGs, ran them through Registax, manually selecting the best, which at a 75% limit resulted in about 250 stacked frames. I then used the wavelets (having now read up on how to use this), and finished off in the Gimp, using only minimal dodge & burn, sharpen and curves. I am amazed how much detail was hidden away in what looked like three over-exposed AVIs - and the red spot can be seen pretty clearly now:

[Clayton]

This is coming along very nicely and showing a lot of promise. There is lot's of detail visible. I would say you are on a winner with this system

[beamish]

Very impressive guys!!!

Is this based on the Russian hack that appeared a while back ??

Either way cracking results !

[MishMich]

It is Russian, but has a GNU license, and is available via SourceForge. Whether it is the same hack I don't know. I didn't touch the original one, because it involved flashing the EOS firmware to facilitate the hack, potentially invalidating the warranty. This one doesn't overtly involve doing this (although it is possible that happens in a covert way) - what it appears to do is interface with the camera in the same way as the Canon-supplied EOS utility.

So, you get a display window on screen, and you can adjust the ISO and the exposure from within the interface, as well as engaging the 5x digital zoom. Whatever is being captured appears to relate to the video signal recieved via live-view. If I record with no magnification, I end up with a 768x512 pixel image of a tiny planet. If I record with 5x or 10x magnification selected either on the interface or from the camera, I get a 768 x 792 image of reasonable size (as seen above) planet (above was done with 2x barlow as well). So, the resulting frame is not that much different from a standard webcam - but is better than anything I've managed so far using either the Opticstar (1280x1040) imager, or a Trust CMOS (1600x1280) webcam I tried. I've not yet managed to get hold of a CCD webcam - as I am too mean to shell out £50-70 on a s/h Philips or new MS or Logitech CCD webcam, only to pull it apart.

This looks like a reasonable alternative, as I bought the Canon cheap for DSO imaging (and avoid messing up my Sony Alpha) - and having one imager for both functions seems like a good scenario. I had considered getting a EOS 500, as that has movie capability - but the 'wisdom' is that DSLRs are no good for planetary work. I don't see that is correct.

M.

[MishMich]

Had another view of the sky last night. Very poor seeing early on, and Jupiter was too low to capture, I think - I couldn't get good visual, and not enough time to let the OTA cool off. However, I did manage to push the SPX+EOS 1000 up to max mag using a revelation 5x barlow. Hard to say whether the quality was down to conditions, or the barlow is just pushing a bit too far.

I did manage to get a better of image of Mars than before, but the clouds kept rolling over, and I think that even when they were not obvious it was hazy; the wind was pretty strong last night, so the tube was buffeted quite a bit.

Anyway, these are my first cut of the EOS transmitting 5x digital zoom using a 5x barlow on the SPX (which maxes out at 400x, which this just about hits I reckon). The polar cap is clearer, but still not getting any other detail from the planet. This was processed within Registax, uding wavelets, and then minimal PSE7 curves and sharpening (rather than Gimp to retain the 16-bit colour-depth).

Any hints of how to get the planetary detail out of either planet would be appreciated - I didn't use any filters. Would a filter help? The only ones I have are a fringe-killer, a contast booster, an LPR, a moon filter, & nos. 47, 82A & 8.

M.

movie_139 x.tif

movie_117_x.tif

[MishMich]

I attached 2 files yesterday - I meant to post the link to the images I I uploaded to my album yesterday. This is Mars,

I have tried to enhance the images of Jupiter I took, but cannot get any of the detail to come out. Had another go capturing today. Maybe going through them frame by frame is the only way to do this? Or, it could be that 5x barlow is pushing the SPX past its limit (or the Revelation barlow may just be not good enough; however, Jupiter was low, and it was hard to see any detail visually with the 2x barlow either. It has been cloudy and hazy. Or, it could be me not managing to focus properly.

[Scott Tilley]

Hi, had a go at this last night with my Canon 50d. Recorded through software as suggested but couldn't seem to adjust the exposure (mars was coming out as bright white disc) Any ideas what I'm doing wrong, should I have the camera set in a certain mode? Was in full manual last night. Thanks in advance.

Scott

[George]

Yes its the Russian hack software, been around for about 12 months or so, here's the original site put in babelfish for translation.

Translation result for http://valexvir.narod.ru/

Or if you can read Russian

Çàïèñü âèäåî çåðêàëüíûìè ôîòîàïïàðàòàìè

[MishMich]

If you are using the EOS Video Recorder software, then you can adjust the ISO exposure in the software, as well as the exposure (aperture is fixed to scope, obviously). I tend to start at ISO 1600, about 10 second exposure. Then work down to ISO 400 & 1 second exposure, depending on how bright the object is. I don't understand how altering the exposure this way works - but it just does. It obviously doesn't relate to actual exposure, although longer exposure settings do result in a slower frame rate - which can vary between 5 & 20 fps depending on the exposure selected.

However, if you are using the native AV mode on the EOS 500, I'm not sure I can help - if the image is too bright, try decreasing the ISO setting down to 4-200. If there is an exposure setting, try decreasing that? I'm pretty sure the AV mode works at night - I have seen night scenes (non-astro) in a photo mag done using AV which were stacked to enhance the detail and contrast.

P.S., if that doesn't help, try sticking a lunar filter in the path. I have been experimenting with a CCTV camera, and having trouble reducing the gain sufficiently, I did find a lunar filter helped darken the image enough to capture something. I had no such problem with the EOS 1000D, though, but did use a barlow, which does reduce the brightness.

M.

[Kerrp30]

Nice what does your setup look like

Paul

[MishMich]

I have no pictures of it.

I have put more recent shots of Mars (& Saturn) in my album.

Stargazers Lounge - MishMich's Album: imaging

M.