Posts Tagged ‘sensor dust’

Wednesday, February 18th 2009

Sucessfully Cleaned My 30D Sensor

My sensor swabs and cleaning liquid arrived extremely quickly, considering they shipped across the Atlantic. I bought them last Saturday and I picked them up yesterday, only three days later. That’s impressive! The weird thing is that I only paid for slow postage …

Anyway, today when I got home from work I decided to have a go at cleaning the sensor of my beloved Canon 30D.

I took a test shot at f/22 before I started. Quite a few specks of dust as you can see, after almost three years of use, only having cleaned it with a hand blower:

Before cleaning

I also used my hand blower thing to clean the outside of the camera, as well as the sensor before I did the real thing. As you can see, this didn’t do much good at all. Only one tiny speck of dust disappeared in the bottom left corner:

After blowing

Next I cut open the plastic container for a swab and unscrewed the top of the bottle. Then I set my camera in cleaning mode to expose the sensor. I put five drops of Eclipse on the swab and stuck it in the camera.

I did a first swipe, flipped the swab over and did a second swipe in the same direction. I’m quite impressed by how effective this was. All the major dust specks are gone and only a few tiny specks are left on the right side, which is where I lifted the swab. With some practice I will hopefully get even better at this.

After swiping

It struck me as being tighter than I expected to get the swab into the camera. And I was surprised that I didn’t see the cleaning liquid (methanol) drying up in the light reflecting off the sensor. (I didn’t see the liquid at all.)

All in all I’m glad I gave this a go. Finally I can use small apertures without having to do a load of spec-hunting in Lightroom.

Monday, June 5th 2006

Aperture and Sensor Dust

One of the main problems with digital SLRs is their tendency to collect dust on the sensor. Every time you change lenses you risk getting more dust on the sensor.

I noticed dust on my sensor the other day when taking pictures of lightning, since I was using a very small aperture – f/22. (The dust only shows up in pictures taken with a small aperture.)

I thought I’d investigate how small apertures you can use without the dust becoming visible.

I simply shot a series of photos of my white wardrobe door, out of focus, at different aperture sizes.

I decided to make a little animation, running from f/4.5 to f/36:

Series of shots taken to illustrate how visible sensor dust is at different apertures.

It seems as if you won’t see the dust at all if you stick to f/8 or lower, and in most scenes you would probably not be able to find the dust at f/11 either. At f/16 you might see the dust in large single-coloured areas. Apertures f/22 and up should be used with care, since you risk getting very sharp black dots in the picture.

The good thing is that you seldom need more than f/8 or f/11 to get the depth of field you want, at least on crop SLRs such as the 30D and 350D. In other words, you shouldn’t miss those very small apertures much.

Sunday, June 4th 2006

Taking Photos of Lightning

Yesterday we had a thunderstorm, so I decided to try and catch a flash of lightning with my new 30D. I had been thinking of this previously, so I already had a plan:

  • Rig the camera on my tripod, overlooking the active parts of the sky.
  • Set it to the smallest aperture, or something like f/16 depending on how dark it is.
  • Set the camera to ISO 100.
  • Set it to JPEG quality for improved buffer size.
  • Set the drive mode to high frame rate.
  • Use the shutter-release (RS-80N3) to lock the camera in drive mode, clicking away like mad.

Having the low ISO and small aperture made each exposure ~0.8 seconds, which was almost enough for the CF card to keep up. (A Sandisk 2GB Ultra II, if you’re wondering.) I later adjusted the camera to over-expose by 2/3 of a stop to get even longer exposures, ~1.3 secs. At this rate the JPEGs were written to the card as fast as they were being shot.

Having the camera set in drive mode with long shutter speeds meant that the camera was taking in light perhaps 2/3 of the time, or more. That means that the chances of a lightning strike ending up in a picture were fairly large.

About 20 minutes later I had 1,000 pictures to sift through. I caught four flashes, but none of the pictures are very awe-invoking. In fact, the last one you can barely see at all…

Flash 1

Well, this one is very small, but still one of the clearer sparks.

Flash 2

A very weak one. Didn’t find it when looking through the photos the first time…

Flash 3

The best one. Click and zoom into this one, it’s quite nice up close.

Flash 4

This one is very weak, but if you look at the high-res version you can see that it extends quite far over to the left. Pretty cool.

Obviously, these shots are pretty lousy as lightning photos go… But I had a good time anyway, and I now know that the technique works fairly well. Although, having used up 1% of my shutter’s expected life-span I wish I’d got some better shots… 🙂

If you’ve got a digital SLR I suggest you try this sometime. (Perhaps when you’ve got a thunderstorm.) Do stand somewhere safe though.

An ND filter would allow you to get even longer exposures, so as to reduce the number of frames somewhat and increase the chances of catching a flash on film … um … CMOS. It might also allow you to use a slightly larger aperture, to reduce visible sensor dust.

It would probably have been easier to do this if it had been a little darker, too. A darker scene would also have made for more impressive-looking shots I guess.

I did notice, after 700 odd shots, that the shutter had slowed noticeably. The exposures were still 1.3 secs, so the write speed shouldn’t have been the problem. In fact, the buffer was empty, so write speed can’t have been the problem. So I’m not sure why the camera did this. It was perhaps getting over-heated or something. So if you do try this, keep an eye on your camera and abort the experiment if you see smoke coming out from between the seals or the lens mount… 🙂