Filter for direct sun photography

[PeterJ]

In a few months they'll be a partial eclipse visible here in Tasmania and I'd like to take a few snaps of it. I don't want to fork out too much for the excercise so will be using my existing 7D + 70-200 + 2x TC but don't mind buying a filter which I'm sure I'll find other uses for. While I don't want to fry my 7D sensor I'm even less inclined to fry my eyes so after some Googling so far my thoughts are:

Purchase a Hoya NDx400 77mm filter which is a 9 stop / 1/500 filter and shoot at around f/11. I've got no intention of looking through the viewfinder so thought I'd quickly compose and focus in live view so the sensor isn't exposed a long time and then switch back to normal mode, lock the focus and take the shots using an intervalometer for a time lapse sequence. I haven't worked it out properly yet but guess at 640mm effective length the sun will be small enough and won't move enough in the 5 minute period to need me to recompose the shot.

Anyway I just thought I'd post and see if anyone sees any holes in my plan or better filters and/or techniques I can use? One possible thing I could do is 10 mins before starting is focus the lens on something at infinity and lock it, but seem to remember reading that somewhere that apparently large objects in the sky may not require infinity focus, but can't really remember the theory and if it was a reputable source.

[Haydn1971]

I suspect the sun will move quite a fair amount in 5 minutes !  I have a B&W ND1000 which at f22, a 25-30 second exposure is about correctly exposed on overcast and blue sky days and I've had numerous empty roads pictures, a novelty I still can't stop smiling at, but my point being that I suspect with the sun bleaching past the moon, a shorter exposure may work.

[PeterJ]

Hi Haydn, thanks for the info. Do you know if the B&W ND1000 is still a current product? While having a look around it looked like it may be a better alternative but I got the impression they no longer made them.

I forgot when I posted I had a little app on my droid phone that calculates FOV and it comes up with 2 degrees vertically for that combination. My astronomy / trig is not so good but think at 360 degrees per day which is 1440 minutes it should move 360 / 1440 * 5 = 1.25 degrees which should be OK if I anticipate the direction and get the first shot near the edge, or could go portrait for 3 degrees and crop later. Then again that's not taking into account the size of the sun, so might be something to try in advance once I have a filter.

[lol]

If you want to photograph the sun directly, try this stuff: http://www.baader-planetarium.com/sofifolie/sofi_start_e.htm

Example output with that filter:

100-400L x2 for 800mm. f/11 1/1000s ISO800 cropped and resized.

[Haydn1971]

Hi Haydn, thanks for the info. Do you know if the B&W ND1000 is still a current product? While having a look around it looked like it may be a better alternative but I got the impression they no longer made them.

Bought mine about 3 months ago, about £85 for a 72mm one - it does give a slight yellow brown tint to images on my 450D, but playing with colour levels is easy.

[aldvan]

Solar photography is quite a serious affair, something to take with great care and prudence.
The solar filter I have in my Mead LX-200 telescope seems a dark mirror and the manufacturer suggesta to check it very carefully for any small hole in the treatment, since the consequences for eyes and sensors can be dramatic. Follow a very well learned procedure before starting to aim and checking, since a small mistake can be irreparable.
By the way, the result is really exciting... I'm attaching an example from  my Meade...
Don't forget that a very dark filter requires slow exposition times, something in the 1/10-1/50 range and the sun moves very fast. For that reason a telescope follows the object by computer controlled microengines.

[PeterJ]

If you want to photograph the sun directly, try this stuff: http://www.baader-planetarium.com/sofifolie/sofi_start_e.htm

Nice photo and that sounds like a good product, from what I could see it's just a film layer so how did you go about mounting to the front of your lens? I saw their suggestion of rolled cardboard for a telescope but can't see that working too well with the small ridge on an DSLR lens, maybe same idea near the base of the lens hood might be OK? The price versus the large size of the film certainly makes it look like an attractive option to experiment with.

Solar photography is quite a serious affair, something to take with great care and prudence.
The solar filter I have in my Mead LX-200 telescope seems a dark mirror and the manufacturer suggesta to check it very carefully for any small hole in the treatment, since the consequences for eyes and sensors can be dramatic. Follow a very well learned procedure before starting to aim and checking, since a small mistake can be irreparable.

Another impressive piece of work, thanks for the advice on pinholes I see the site lol linked to mentions the importance of that also. Don't worry I'm taking it seriously, reason for my post because I knew I'd get some great tips on things I hadn't thought of .

[lol]

The film has instructions on mounting it to a cardboard sandwich. I then fixed that cardboard holder onto a spare filter so I can add/remove it from the lens easily.

[TexPhoto]

I shoot the sun all the time, even video thrugh a 400mm f2.8 and have not effected my sensor.  You could add 1 or more polarizing filters to knock down some stops. or a 3-4 stop ND filter might be more usefill for other shooting.

Of course you can use live view to avoid exposing your eyes.

A quick search of ebay for "welding glass" finds $3 sections of glass that should do what you want.

[epsiloneri]

I shoot the sun all the time, even video thrugh a 400mm f2.8 and have not effected my sensor.

I can't believe I'm reading this advice from you! It's very dangerous to do what you suggest. Please note that you should always place any filter in front of the front lens. The 400/2.8 has a rear mounted filter which is not suitable for solar filters. The reason is that the light will get concentrated on the rear filter, which is bad if you point it towards the Sun - the filter will heat up and possibly crack/be destroyed with disastrous consequences for the sensor (or worse, your eye). Ever used a magnifying glass to light a fire? Then you know what I'm talking about.

For continuum light solar filters I've tried and recommend those from Thousand Oaks Optical, they even have camera threaded filters. For a 400/2.8 I would recommend purchasing a mylar sheet (or black polymer) and produce a filter yourself for placement in front of the lens.

For more tips about observing total and partial solar eclipses I recommend NASA's eclipse bulletin which are published in connection with major eclipses. There is none for the Nov 25 one, but you can have a look at older bulletins as the practical information is the same: http://eclipse.gsfc.nasa.gov/SEpubs/2010/rp.html.

Since you're in the neighbourhood, you may also want to check out the total solar eclipse on Nov 13 next year in NE Australia and New Zealand. If you haven't seen a total eclipse I would highly recommend it - a partial eclipse does not even come close in comparison.

[epsiloneri]

My astronomy / trig is not so good but think at 360 degrees per day which is 1440 minutes it should move 360 / 1440 * 5 = 1.25 degrees which should be OK if I anticipate the direction and get the first shot near the edge, or could go portrait for 3 degrees and crop later. Then again that's not taking into account the size of the sun, so might be something to try in advance once I have a filter.

I checked your numbers BTW, they are OK. The Sun is only 0.5 deg in diameter. But where do the 5 min come from? The eclipse should start at least 30 min before sunset. I haven't figured out the exact circumstances, but I imagine there being an opportunity here for a nice partially eclipsed sunset image. Without a filter when the sun is close enough to the horizon, just like you normally would shoot a sunset. Trying in advance is a good idea, in particular if you want to find the best spot for the sunset.

One idea is to find the exact spot where the sun will set (e.g. by checking the day before - remember, this location moves with time so don't attempt to find it too many days in advance) then go back and compose the image well before sunset on the eclipse day and make a time series that shows the sun entering the frame and then setting partially eclipsed. For best result, use manual mode and an exposure that you determined on the previous "trial" sunset. Focus on something distant using live view and manual focus. Start your time-series 10-15 min before sunset. Don't forget to post your results here

I have the NDx400 filter, and it's not really dark enough for day time photography of the sun - you can do it, but only with the shortest exposure time and smallest aperture (resulting in resolution loss due to refraction). As a comparison, the Thousand's Oak solar filter I previously referred to attenuates light by a factor of 100,000, or 16.6 stops - enough to safely look through. Because your particular partial eclipse is at sunset, it shouldn't be a problem for you though.

[PeterJ]

For continuum light solar filters I've tried and recommend those from Thousand Oaks Optical, they even have camera threaded filters. For a 400/2.8 I would recommend purchasing a mylar sheet (or black polymer) and produce a filter yourself for placement in front of the lens.

Thanks for that and all the other tips, I've just e-mailed to get a shipping quote on a 77mm threaded filter, they sound convenient and I don't imagine I'll be buying any big primes in the near future.

I checked your numbers BTW, they are OK. The Sun is only 0.5 deg in diameter. But where do the 5 min come from? The eclipse should start at least 30 min before sunset.

I got that number from http://en.wikipedia.org/wiki/List_of_solar_eclipses_in_the_21st_century although I see now it's the central duration so I'll follow you advice of checking the position the day before for framing so it enters the frame during the sequence. Thanks once again for all the advice, I'm just about to take a better look at the NASA info and I'll be sure to post the results  .

[epsiloneri]

But where do the 5 min come from? The eclipse should start at least 30 min before sunset.

I got that number from http://en.wikipedia.org/wiki/List_of_solar_eclipses_in_the_21st_century although I see now it's the central duration so I'll follow you advice of checking the position the day before for framing so it enters the frame during the sequence.

Since the eclipse is partial, there is no central duration. Perhaps you looked at the wrong entry? Partial eclipses are regularly more than an hour in length, but because your starts just before sunset, you'll only have half an hour or so before the sun sets (partially eclipsed). Good luck anyway, and I look forward to see your results!

[PeterJ]

Ahh yes I must have looked at the wrong column. Just had a check and the time of greatest eclipse is about 3 hours before sunset for that time of year. That's probably good though because from my location I don't get very good sunsets because of a mountain range that probably covers at least the last half hour of sunset.

[epsiloneri]

Sorry, somehow I thought Tasmania was close to New Zealand, but I now realise it's much closer to main land Australia. That means the sun will not set partially eclipsed, in fact the moon will only nudge the Sun as seen from Tasmania and the eclipse will only last some 30-40 min, starting at ~07:30-07.40 UT, corresponding to 18:30-18:40 local DST, depending on where your are.