350D and 10-22 Image Quality Issues

fotomat

Seeing that there are several foks here with 350D IR conversions, I thought it would be a good place to post this question. I recently had my 350D converted for IR with an internal R72 filter. I was hoping to use it with a Canon 10-22, but at 10mm, the edges inward 1/3 of the frame are blurred/smeared looking as seen here -

www.lensmateonline.com/images/IMG_7262_10-22_10mm.jpg


These are unprocessed jpegs except for a levels adjustment to make them easier to evaluate.

I had the camera calibrated to the lens by Vitaly @ lifepixel, and he (as well as Dan L. of maxmax) say it's an IR convergence problem due to the wide lens and/or its design. If I zoom to 14mm, things improve somewhat, but not that much. Here's a test image at 14mm

www.lensmateonline.com/images/IMG_7085-10-22_14mm.jpg


I've tested 2 other copies of the 10-22 with the same result, BTW. My Canon 17-40 looks a little better, but the edges are still soft -


www.lensmateonline.com/images/IMG_7257-17-40_17mm.jpg

Could anyone using the 350D/10-22 link to a full size image so I can compare? Full wide (10mm) of a distant scene with edge detail would be best for evaluation. I would love to see someone getting better results than this - I was hoping to be able to make larger prints with this setup, but you can see the distortion in a 4x6 print. Some 17-40 samples @17mm would be welcome also.

Thanks in advance,

Tony

boogie

Hi Tony

Sorry I can't help you with your problem as I don't own the 10-22 or the Canon 17-40.
I didn't have a lens calibrated for my Rebel XT IR conversion and have been using a Sigma 18-125 lens on it and have no problems.

FWIW I thought your pictures looked good but I can understand your concern. Maybe someone can help you.

gluwater

That does look strange. The distortion is out of control on those. Have you tried PTlens on them to try to fix the distortion? I don't know how much it will fix but it may be worth trying. I have been using my 10-22 on my IR 350D and have not had any problems. Here are two shots using that combo.
Click on the pictures for full size.

@10mm


@14mm

fotomat

boogie wrote:

Hi Tony

Sorry I can't help you with your problem as I don't own the 10-22 or the Canon 17-40.
I didn't have a lens calibrated for my Rebel XT IR conversion and have been using a Sigma 18-125 lens on it and have no problems.

FWIW I thought your pictures looked good but I can understand your concern. Maybe someone can help you.

Hi Boogie,

Thanks for the response. Curious, who did your conversion? In case you didn't see the 10-22 @10mm sample at 100%, here's a left edge crop next to a center crop -



Tony

fotomat

gluwater wrote:

That does look strange. The distortion is out of control on those. Have you tried PTlens on them to try to fix the distortion? I don't know how much it will fix but it may be worth trying. I have been using my 10-22 on my IR 350D and have not had any problems. Here are two shots using that combo.
Click on the pictures for full size.

@10mm


@14mm

Nick,

I've tried the CS2 lens correction filter, and though it improves the affected area slightly, it also makes the center worse.

In your first image, if you look at the top of the tree on the right, you can see a similar distortion. The foreground looks pretty good, but I've noticed that foreground/near objects look good on mine too. At 14mm, looks better than mine.

Where did all the noise come from @iso100??? Did the camera get hot? I see virtually no noise @ iso100 on mine. Then again, we've only cracked 60 deg here (Seattle) a few times this year.

Who did your conversion?

Thanks for your help,
Tony

boogie

fotomat wrote:

Hi Boogie,

Thanks for the response. Curious, who did your conversion?

Tony

MaxMax converted my Sony P200 and LifePixel converted my Rebel XT.

The P200 is easy to use but I need some practice with the XT..

gluwater

fotomat wrote:

Nick,

I've tried the CS2 lens correction filter, and though it improves the affected area slightly, it also makes the center worse.

In your first image, if you look at the top of the tree on the right, you can see a similar distortion. The foreground looks pretty good, but I've noticed that foreground/near objects look good on mine too. At 14mm, looks better than mine.

Where did all the noise come from @iso100??? Did the camera get hot? I see virtually no noise @ iso100 on mine. Then again, we've only cracked 60 deg here (Seattle) a few times this year.

Who did your conversion?

Thanks for your help,
Tony

Using an Ultra Wide you will always have distortion, the distortion in your pictures seems to be excessive though. As for the noise in my images @ ISO100, well I botched the exposure and had to dial in some major +EC during post so I think that is where it is coming from. My XT is from Maxmax.

Andy

Bump, for Tony.

Roger_Salmon

fotomat wrote:

Seeing that there are several foks here with 350D IR conversions, I thought it would be a good place to post this question. I recently had my 350D converted for IR with an internal R72 filter. I was hoping to use it with a Canon 10-22, but at 10mm, the edges inward 1/3 of the frame are blurred/smeared looking as seen here -

www.lensmateonline.com/images/IMG_7262_10-22_10mm.jpg


These are unprocessed jpegs except for a levels adjustment to make them easier to evaluate.

I had the camera calibrated to the lens by Vitaly @ lifepixel, and he (as well as Dan L. of maxmax) say it's an IR convergence problem due to the wide lens and/or its design. If I zoom to 14mm, things improve somewhat, but not that much. Here's a test image at 14mm

www.lensmateonline.com/images/IMG_7085-10-22_14mm.jpg


I've tested 2 other copies of the 10-22 with the same result, BTW. My Canon 17-40 looks a little better, but the edges are still soft -


www.lensmateonline.com/images/IMG_7257-17-40_17mm.jpg

Could anyone using the 350D/10-22 link to a full size image so I can compare? Full wide (10mm) of a distant scene with edge detail would be best for evaluation. I would love to see someone getting better results than this - I was hoping to be able to make larger prints with this setup, but you can see the distortion in a 4x6 print. Some 17-40 samples @17mm would be welcome also.

Thanks in advance,

Tony

Tony

I've converted a significant number of Canon DSLRs for people and the lenses which give the greatest problems are the superwide angles. The problem is twofold I believe. Firstly you can see a shift in focal plane with focal length. The 16-35mm f2.8 exhibits this strongly if you shoot at f2.8. If the camera is adjusted to give good AF accuracy with say a 50mm lens then the 16-35mm zoom will show marked front focusing at 16mm which reduces significantly by 35mm. This effect can be controlled to some extent by stopping down to f8. I've had clients who wanted their camera adjusting for this lens specifically. Unfortunately it then gave significant back focusing with most other lenses. The second problem with your lens, I suspect, is that the focal plane at 10mm is no longer flat. I have no proof other than seeing similar results to yourself though not quite so marked.

The one thing to remember here is that the vast majority of lenses are optimized for visible light performance. Once in the near-IR all bets are off. The more extreme wideangles will give the greatest problems. If you want a good alround lens for your camera try the 17-85mm IS or the 28-135mm IS.

I hope this helps.

Roger

fotomat

Roger_Salmon wrote:

Tony

I've converted a significant number of Canon DSLRs for people and the lenses which give the greatest problems are the superwide angles. The problem is twofold I believe. Firstly you can see a shift in focal plane with focal length. The 16-35mm f2.8 exhibits this strongly if you shoot at f2.8. If the camera is adjusted to give good AF accuracy with say a 50mm lens then the 16-35mm zoom will show marked front focusing at 16mm which reduces significantly by 35mm. This effect can be controlled to some extent by stopping down to f8. I've had clients who wanted their camera adjusting for this lens specifically. Unfortunately it then gave significant back focusing with most other lenses. The second problem with your lens, I suspect, is that the focal plane at 10mm is no longer flat. I have no proof other than seeing similar results to yourself though not quite so marked.

The one thing to remember here is that the vast majority of lenses are optimized for visible light performance. Once in the near-IR all bets are off. The more extreme wideangles will give the greatest problems. If you want a good alround lens for your camera try the 17-85mm IS or the 28-135mm IS.

I hope this helps.

Roger

Roger,

Thanks for your comments. At this time, my 10-22 is being tested on other IR converted 350D bodies, including ones with different IR filters. I'll follow up with the results when available.

Thanks,
Tony