| Author | Thread |
|
|
07/27/2006 01:23:02 PM · #1 |
Sorry if this has been posted before.
Someone emailed me this ISO 1600 noise comparison of most dSLRs.
ISO 1600 Noise
Draw your own conclusions!
-Chad |
|
|
|
07/27/2006 01:25:32 PM · #2 |
I conclude we are in good shape, purser ;-)
R.
|
|
|
|
07/27/2006 01:26:27 PM · #3 |
| Thanks! Very happy with the results ;-) |
|
|
|
07/27/2006 01:28:24 PM · #4 |
| looks to me like the 5D wins easily with the 20D close behind. I'm kinda surprised at the 1 series cameras not being better. |
|
|
|
07/27/2006 01:29:27 PM · #5 |
| Amazing how well the 350D held up again the others, considering it has been called a soccer mom camera. |
|
|
|
07/27/2006 01:29:47 PM · #6 |
Originally posted by kyebosh:
looks to me like the 5D wins easily with the 20D close behind. I'm kinda surprised at the 1 series cameras not being better. |
Yeah I figured it would do better. The 30D uses the same as the 20D and that looks good to me.
MattO
Edit cause apparently I shouldnt read/post while eating.
Message edited by author 2006-07-27 13:31:09.
|
|
|
|
07/27/2006 01:31:37 PM · #7 |
| I want a do-over. The KM 5D they used appears to have a dirty sensor. Look at that sensor dust spot! ;^) |
|
|
|
07/27/2006 01:33:24 PM · #8 |
I'm still amazed at how well the 5D performed. I think I need one :-)
-Chad |
|
|
|
07/27/2006 01:35:44 PM · #9 |
| Serious question (not just trying to knock the competion down) - Do the Full Frame cameras do as well with noise in the corners / edges? |
|
|
|
07/27/2006 01:42:12 PM · #10 |
Originally posted by Megatherian:
Serious question (not just trying to knock the competion down) - Do the Full Frame cameras do as well with noise in the corners / edges? |
I would say if the corner or edge is exposed as well as another camera yes. The vignetting however presents and interesting problem for this as we all know underexposing causes more noise to appear... good question. |
|
|
|
07/27/2006 01:42:25 PM · #11 |
Originally posted by Megatherian:
Serious question (not just trying to knock the competion down) - Do the Full Frame cameras do as well with noise in the corners / edges? |
Corner/edge should be irrelevant as far as noise in concerned. That's an issue of lens performance, and sensor performance ought to be nominal edge-to-edge.
R.
|
|
|
|
07/27/2006 01:43:22 PM · #12 |
| OH, Yeah! That's right. The 5d rocks!!!! Just another reason to love my baby. But there are sooo many. lol j/k ;) |
|
|
|
07/27/2006 01:47:15 PM · #13 |
Originally posted by cpurser:
Sorry if this has been posted before.
Someone emailed me this ISO 1600 noise comparison of most dSLRs.
ISO 1600 Noise
Draw your own conclusions!
-Chad |
Interesting but, if you buy a camera only based on noise levels then you buy for the wrong reason. You have to look at all points when buying. Every camera brand have there strong and weak points. You have to decide what is most important to you and the way you shoot. Noise is not an issue with me and this is one reason I didn't buy Canon. For me and the way I shoot the Olympus E-1 was a better choice. If all you are planning to shoot is very low light then Canon would be a better choice.
|
|
|
|
07/27/2006 02:00:22 PM · #14 |
Originally posted by TomH1000:
Interesting but, if you buy a camera only based on noise levels then you buy for the wrong reason. You have to look at all points when buying. |
Ahh but for pure advocacy it's far easier to just pick one point ;)
Still good to see the old 20D holding its own. |
|
|
|
07/27/2006 02:34:07 PM · #15 |
Too bad, I went to the site referenced and it is in Japanese. I was unable to find the report and see what parameters they used to do the tests.
|
|
|
|
07/27/2006 03:03:53 PM · #16 |
Interesting test. Sure wish they had shot someting with fine detail. It's often better to have a little more noise if stamping out the noise leads to smearing of the fine detail.
Nonetheless, it's amazing just how much difference in both the amount and characteristics of noise are present.
With regard to the question above about vignetting and noise, the answer is no... and yes. There won't be any difference center to edge in the actual magnitude of the noise, but if you have 1 stop of vignetting, and correct that in conversion or post, you will increase the apparent noise level in the corrected area. |
|
|
|
07/27/2006 03:05:49 PM · #17 |
| RAW or in-camera jpeg processing? |
|
|
|
07/27/2006 03:15:17 PM · #18 |
Originally posted by faidoi:
RAW or in-camera jpeg processing? |
I'd ass-u-me it was in-camera JPEG. Which begs the question of what contrast and sharpening settins were used. Certainly a camera with more aggresive default settings for contrast and sharpening would show worse noise.
Given that it looks like they took care to get the WB and exposure pretty close to the same, I'd guess that they paid attention to the other variables as well as they could.
|
|
|
|
07/27/2006 03:16:42 PM · #19 |
| WOW... I do have a decent camera :P |
|
|
|
07/27/2006 03:35:57 PM · #20 |
The fact that they didn't clean off the sensor dust o the Minolta makes me wonder just how much attention to detail they actually paid to get an even test bed so as to not produce biased results.
If I didn't know any better and based my opinion solely on this chart I'd never buy a Minolta (the fact that I can't is besides the point) because it appears Minolta put big blobs of goo on the image. Also, since we can only see the one tiny spot of the statue it's impossible (from that "study" alone) to determine how well each camera performs on the other parts of the photo. What about the darkest and lightest areas - would the results be the same?
I'm not trying to argue the actual cameras. I know canon, especially the FF sensors, has a clear advantage. The actual validity of the test may not be so clear though. |
|
|
|
07/27/2006 03:39:27 PM · #21 |
|
|
|
07/27/2006 03:42:22 PM · #22 |
Originally posted by faidoi:
Kit lenses? |
Immaterial to sensor performance.
R.
|
|
|
|
07/27/2006 04:48:03 PM · #23 |
It would be interesting to make sure that they used the same aperture/shutter speed settings for all cameras. Increasing exposure time leads to more noise.
Nothing really unexpected in these results, though. The Canon sensors lead with the Nikon sensors just a bit behind.
The most interesting part for me was the performance of the D200's sensor. It's a bit noisy, but look at the qualities of this noise. Other sensors' noise looks like colour blotches, but D200's looks very much like film grain: just the luminance noise, no colour noise! Interesting. |
|
|
|
07/28/2006 11:57:48 AM · #24 |
Here's the original (AFIK) compendium.
//www.pbase.com/miljenko/image/57722794/original
The difference between the 5D and D200 is more pronounced
There's also a comparison of ISO 800 at //www.pbase.com/miljenko/image/57722796/original |
|
|
|
07/28/2006 12:14:48 PM · #25 |
Aha, thanks for the link. I know Miljenko from the DPReview forums. He is a very reasonable person, and I trust his ability to conduct a test properly. However, as I already said, there are no surprises in the results.
It's interesting that the reviewer himself, having conducted the review, chose to buy a Nikon camera over Canon. Just another evidence to the point that noise performance is not everything.
Here's a relvant quote: Canon cameras are well known for superior high ISO noise performance. Anyone trying to prove the opposite either is a liar or performs an uncontrolled testing. Even this test contains a questionable D200 pic that is suspected to have worse noise level than shown. However, noise performance is not everything. I have chosen few other Nikon virtues over the Canon's low noise although lower noise would made me much happier. |
|
Home -
Challenges -
Community -
League -
Photos -
Cameras -
Lenses -
Learn -
Help -
Terms of Use -
Privacy -
Top ^
DPChallenge, and website content and design, Copyright © 2001-2026 Challenging Technologies, LLC.
All digital photo copyrights belong to the photographers and may not be used without permission.
Current Server Time: 01/03/2026 02:37:40 AM EST.
