ISO 1600: 40D vs. 5D

[yakim_peled1]

http://forums.dpreview.com/forums/readflat.asp?forum=1019&thread=24687298

 

IMHO this is VERY impressive.

 

Happy shooting,

Yakim.

[bobatkins]

Based on purely the noise levels, I never saw much difference between the 20D (or 30D) and the 5D. You can see some of my noise samples at http://www.bobatkins.com/photography/digital/canon_eos_5d_review_2.html

 

Final prints from a 5D may look better because they are enlarged less and so the noise is smaller and less visible, but looking just at noise in 100% image samples, there's really not much difference.

[robert_clark]

Why were all the previous posts deleted? Does the moderator really think his one comment is a substitute for the earlier clash of opinions?

[david_amberson1]

Its mind blowing to me that people keep posting that there isnt much difference in noise levels from the 5D and the 20D. I really think people that do these reviews and tests are biased in some way. Lets get real here. All of the real pictures that I have ever seen from the 5D at 3200 are way better that the 20D at 3200. I am talking night and day difference. No 100% crops. And in the posts here, you have these 2 cameras performing almost exactly the same. That is an insult to everyone that reads from this website. I have some high school night football photographs with a 20D at 3200 that were not even usable they were so bad. It looked like a picture from the 70's. Pictures that I have seen from weddings taken at night with a 5D were usable right out of the camera. Screw post processing. Are they usable out of the camera. Thats whats important. A good camera is supposed to make your job easier. You should be able to tweek your cameras settings to your preference, know what light you are working under, set that white balance and use JPEG and deal with the noise that your camera produces. Not create all this post processing. Try taking 450 pictures at a football game and then go home and spend 4 hours processing these stupid pictures and doing noise reduction to everyone of them. So why is it that every freeking website known to man, has the 5D praised beyond recognition for noise control. You never read in a website how the 20D is setting the bar for noise control. I swear, I think some these guys have a stock pile of 20 and 30D's somewhere they are trying to sell by posting these rediculous ISO comparisons where it looks like they used the same camera. Why dont somebody take a 40D to a night time sporting event where 90% of its buyers would use it at and shoot some 3200 ISo shots with it and post the pictures as they are. No processing or 100% crops or noise reduction. Just the pictures as they are so we can see whats actually usable or not. All of the posts are showing examples in the broad daylight at 3200. What jackass would do this. Is it too much to ask. Sorry, Ive had enough of the stupid reviews and had to vent. I need a nap after that. See ya!

[david_amberson1]

Oh ya, Robert Clark, you hit the nail on the head. I didnt get a chance to read the previous before deleted, but I can guess what they were like.

[mike sisk]

<blockquote><i>... Why dont somebody take a 40D to a night time sporting event where

90% of its buyers would use it at ...</i></blockquote>

 

<p>90% of 40D buyers take pictures at night-time sporting events? I'm not so sure about

that.</p>

<p>It's interesting that any criticism of the 5D seems to generate a lot of posts like this. I

wonder why? (seriously)</p>

[jimstrutz]

I think it partly depends on what you're talking about. If you compare the noise of a 20D image to a 5D image at 100%, the 5D is better, but not all that much better. Where the 5D shines is in the fact that the larger, and higher resolution sensor doesn't have to be enlarged as much as the 20D image. The result is, the 5D's image is considerably better in print, but comparing 100% crops, not so much.

 

David A, you said, "Screw post processing. Are they usable out of the camera. Thats whats important." That's one valid viewpoint, but certainly not the viewpoint of all others. A lot of other people want to know how good it can be made to be with a lot of tweaking. Your's isn't the only correct perspective.

 

Also you said, "...and post the pictures as they are. No processing or 100% crops or noise reduction. Just the pictures as they are so we can see whats actually usable or not." It's not that simple. The camera, by default adds noise reduction and sharpening to all JPG images. It's just how a camera's built in RAW interpreter works. The problem is, they are not all equal. Some make smoother images at the expense of sharpness, etc. Comparing JPGs will not tell you the whole story of image quality. And again, some people want to know what it looks like when tweaked to the optimal amount. Obviously that's not your cup of tea, but it is for others.

 

I do agree with the idea of shooting high ISO images in the dark though. It only makes sense, and sometimes there seems to be a significant difference.

[bobatkins]

"Its mind blowing to me that people keep posting that there isnt much difference in noise levels from the 5D and the 20D. I really think people that do these reviews and tests are biased in some way"

 

Then look at this page, look at the images and make up your own mind - http://www.bobatkins.com/photography/digital/canon_eos_5d_review_2.html

 

I'll admit "real world" tests are more important, but it's hard to do well controlled "real world" tests.

 

If the 5D images look better, they look better, but the intrinsic noise level of the sensor isn't very different from that of the 5D.

 

You should also be careful of accusing people of things for which you have no evidence. I don't think any of the reputatble test sites have any bias related to commercial gain.

[david_amberson1]

To Mike Sisk: Maybe 90% is a little much. Maybe 90% of the people I know would use it for this. Fair enough. With the critism of the 5D, it doesnt bother me. I dont own one. But was heavily impressed with noise control with all of the pictures I have seen out of the camera. Its just odd how the 5D was praised to no end for noise control and now from certain people, its no better than the 20D. Based on their test shots of charts. Why would I buy a 5D if the 20D is almost exactly the same.

 

To Jim S.: The point I was making is that, no one uses pictures at 100%, they print them as they are or some very minor cropping. So 100% crop isnt all that important when looking at noise. What does it look like as a picture. You know...3888x2592. I know JPEGS are sharpened and filtered in camera. But we cant change this in any camera. So are the JPEGS in the "40D" good or not. You can make any photo look good with enough photoshop. But if you are gonna do this, then buy a Rebel XT and spend the rest of your days processing and photoshopping. Those people dont enjoy photography, the enjoy goofing around with computer software. People come to this site for help. They want to know if they can take a picture with the 40D as a JPEG and it look better than the 20D as a JPEG. Most people dont shoot raw when taking candids around the house of a pet or something. So what it could mean is, "If I buy a 40D, shoot in my living room as JPEG at 1600 ISO, will it be better than my 20D at same settings" I dont like shooting above 800 with my camera, but often should for shutter speed. But am also not willing to do all of the processing and tweeking and filtering for a nice picture of my cat rolling around in the floor. Thats why you spend a little more for a better camera...so you dont need to do all of this. A good camera shouldnt get in your way. It should be easier and more consistent. But according to some people, Canon has 5 lines of camera, and it doesnt matter which one you choose, you'll get the same results. People want a good comparison so they make a good decision. For most people, this camera is very expensive and they want to know all they can.

 

To Bob Atkins: That page is the exact one I am refering to. These do not show any difference in noise levels, which I know to be different. There is no way for the 5D to have that much of a better image out of the camera than the 20D and then only to look this close in your shots. And as far as accusations, I was refering to other sites that say the same as you. " I dont see a difference in the 5d and 20D in noise control", "The 20D looks as good as the 5D", and then they say, "Oh buy the way, I have 15 20D's in stock, how many do you want". And, reputable test sites are trying to make money. Its why they have all the blinking advertisments on the page. If they can persuade a vonerable new buyer to want a 20D instead and they just so happen to have a bunch to unload, then they have paid themselves. They dont care if we're happy or not. I wasn't exactly accusing you, just your examples look awfully the same. Way different then what I have seen. If someone had never seen a side by side comparison in person then they wouldnt know any difference. They have to believe you. I have and I aint buying it. That camea is double the price of the 20D and there is no way they are that close. But hey, if the shoe fits, then wear it.

[david_amberson1]

I bought a 40D today. I havn't seen much difference at all in the 40D and 20D ISO wise. Everything else though is a nice upgrade. Will do a better test tonight with a tripod and just change cameras to see a side by side comparison. First I will do an XTi since alot of people want to see if its worth upgrading. So far, the 6.5fps, 3" screen, faster AF and better feel in hands is well worth the upgrade alone IMO. We will see if we can put this 5D vs 20D ISO thing to rest once and for all soon. If I am wrong, I will apologize to everyone I argued with on the subject.

 

What you are going to get is a tripod mounted 70-200 2.8L IS and I will shoot the 5D at 1600 and 3200. Then swap with 40D and do the same. I will do JPEG+RAW on both. No noise reduction, sharpening or 100% crops. Just camera VS camera. We will see which one has better noise control as a picture, just the way the camera made it. If you want to see how good your software skills are, by trying to see just how good you can make it, thats another topic. This will be Camera vs Camera. Now to schedule my buddy with the 5D at dark so we can see.

[eric merrill]

Bob:

 

"I'll admit "real world" tests are more important, but it's hard to do well controlled "real world" tests."

 

I would think that even controlled noise tests should show noise in shadows. Not in the midtones. I'm not surprised there's little difference at 18% gray.

 

Other factors come into play, too. Are the exposures exactly the same? I wouldn't think this would have been too difficult, but I was surprised to find that I got different exposures when comparing noise between 100 and 200 ISO with the same camera. I had to change to manual mode. I just figured that if I bumped the ISO up one stop and then shortened the shutter speed by a stop, that I'd be getting the same exposure since the camera was on a tripod and the light source was constant.

 

 

Eric

[david_amberson1]

Well, I did the XTi vs 40D last night. It looks to me that 3200 ISO on the 40D is vertually the same as the XTi at 1600. So far it seems to be a nice upgrade. Almost like getting a free ISO stop without sacrifice in IQ. I talked with the guy with the 5D and I might be out of luck. He thinks it might be sold. I will know tomorrow. If not, I will find a way to do the comparison.

[keith_lubow]

As Mr. Atkins, said, even if the noise is exactly the same, when prints of the same size are made, the 5D's will be cleaner looking due to less "enlargement". It's exactly the same as using a high-quality film in a Pentax Auto 110 and a K-1000, and making identically sized prints. If you were to look under a microscope, the films would look identical. But with the prints, which one do you think will show less grain?

 

Keith