2 stops better low light performance from Canon Rebel to 7D and from 7d to 5D Mk. II?

[Chris7]

Nikon D500 vs. D300 at high ISO?

(see link I posted)

Edge: Thanks. Sorry, but I’m not sure I’m finding it. Are you referring to the Dxomark link? Are these measurements at the pixel level or at image level?

I assumed these measurements were at pixel level, so I put in the a few camas of with a similar amount of pixels: 12.2 MP Canon XSi/450D, and the 12.3 MP Nixon D500 and D300s . Comparing the high ISO performance of a $630 and $1700 camera seems absurd, but I don’t see the difference of the two Nikon's at high ISO. What am I missing?

Thanks,
Chris

 

[Phrasikleia]

Edge: Thanks. Sorry, but I’m not sure I’m finding it. Are you referring to the Dxomark link? Are these measurements at the pixel level or at image level?

I assumed these measurements were at pixel level, so I put in the a few camas of with a similar amount of pixels: 12.2 MP Canon XSi/450D, and the 12.3 MP Nixon D500 and D300s . Comparing the high ISO performance of a $630 and $1700 camera seems absurd, but I don’t see the difference of the two Nikon's at high ISO. What am I missing?

Thanks,
Chris

High ISO isn't everything. Those cameras have a lot separating them in build quality, performance, and features. The sensor is only one part of the cost.

 

[Phrasikleia]

Since this thread is still going and DxO has again reared its ugly head, I'm going to return to the chart OreoCookie posted on Page 1. It occurred to me shortly after my last response that his chart was probably the "screen" comparison, which completely brackets out the issue of resolution. Indeed it is. The "screen" summary is an abstract and artificial construct with no basis in any practical application of interest to photographers. The tab you want to look at on the SNR (signal-to-noise ratio) option is the "print" tab, which takes resolution into account; that is, it normalizes the output from each camera.

When you look at this more informative tab, you'll see that the 40D doesn't fare as well and that the 7D is the frontrunner at higher ISO sensitivities.

Here's that chart:

It just goes to show that you can't separate one little factor in a camera from the others and get anything meaningful out of it. If the OP is shopping for a camera, he needs to have a wider scope than "high ISO" in order to make a good decision.

 

[OreoCookie]

Comparing the high ISO performance of a $630 and $1700 camera seems absurd, but I don’t see the difference of the two Nikon's at high ISO. What am I missing?

Sometimes, cheaper and more expensive cameras share a lot of components. E. g. the D80 shares the sensors, most of the AF system and such with the D200. Hence, the two cameras should be (and are) very close in terms of performance.

But I agree, you should compare cameras that are comparable to each other.

 

[OreoCookie]

When you look at this more informative tab, you'll see that the 40D doesn't fare as well and that the 7D is the frontrunner at higher ISO sensitivities.

I haven't seen this one, it could have saved me quite a bit of time

In all data points but the last two, the 40D is slightly better. In the last two, the 7D beats the 40D by 0.5 db and 0.9 db, respectively. This corresponds to an improvement of 6-12 % at these high ISO settings. At lower ISO settings, the 40D is better by the same amount (0.5~0.9 db, i. e. 6~12 %). (I'm sorry to be so nitpicky, but I'm a scientist, I like numbers )

The 7D's improvement is that you can use your camera at much higher settings, although if the 40D were designed today, the graph suggests that this were easily possible as well.

It just goes to show that you can't separate one little factor in a camera from the others and get anything meaningful out of it. If the OP is shopping for a camera, he needs to have a wider scope than "high ISO" in order to make a good decision.

Seconded.

 

[Edge100]

Edge: Thanks. Sorry, but I’m not sure I’m finding it. Are you referring to the Dxomark link? Are these measurements at the pixel level or at image level?

No. I'm referring to the Ti1 review at the-digital-picture.com. The T1i outperforms the 50D in terms of high ISO performance, despite have the same sensor.

That said, and as others have mentioned numerous times, high ISO performance is not everything. The 50D is in many ways a superior camera to the T1i. You need to pick a camera based on the totality of its features, not just one metric, which, arguably, isn't even all that important to begin with.

The noise performance of virtually any current DLSR at ISO3200 will be acceptable for web-sized prints, with the possible exception of the very lowest MP models. As I've shown, a 6 year old 8MP DSLR can produce very useable results at ISO3200. Granted, my 1DmkII tops out at ISO3200, so if I need ISO6400 I'm out of luck. But to be honest, in the last 2-3 years, I can think of maybe one instance in which I really needed ISO6400 (that is, I didn't have anything faster than f/2.8 and ISO3200 gave unacceptably slow shutter speeds). f/2.8, 1/50, ISO6400 is VERY dim.

There are so many other things to worry about. Get fast glass. When you're at ISO6400, 1/50, f/1.2 on your $2000 85/1.2L, THEN you can start worrying about noise performance. But if you're shooting with a slow-as-hell Canon kit zoom with sketchy IQ and also worrying about pixel-level noise on a 15MP camera, I think your concern is, perhaps, just ever so slightly misplaced .

 

[toxic]

the 500D and 50D do not share the same sensor. the only Canon SLRs that do are the 20D and 30D.

 

[Edge100]

the 500D and 50D do not share the same sensor. the only Canon SLRs that do are the 20D and 30D.

Was that really all you took from the previous post?

But fair enough; they're likely not the same exact sensor. But...

T1i = 15.1MP
50D = 15.1MP

Might not be the same sensor, but same pixel pitch. T1i has lower noise than the 50D, which was in direct response to the OPs question of whether Rebel Series < xxD series <xD series in terms of noise. This is an example that shows that at the same pixel pitch, a Rebel series camera can outperform an XXD camera.

But again, this is really nitpicking at this point. I will say this one more time; unless you're shooting the fastest glass you can and STILL bumping into noise issues, then concerning yourself with which camera has marginally better high ISO noise performance is misguided. Buy the camera you need in terms of resolution, performance, IQ, handling, etc (all of which make more difference in real world, everyday use than does ISO performance), and then buy fast glass.

Let's say you have a 17-55 f/3.5-5.6 kit lens on your T1i, and need to shoot something moving at 50mm (so IS doesn't help you), in very low light. At this focal length, you're looking at f/5.6 as max aperture.

So...

1/50, f/5.6 @ ISO6400
1/100, f/2.8 @ ISO3200
1/200, f/1.4 @ ISO1600

All of these give the same exposure. But because I invested in a fast 50/1.4 prime, I'm able to shoot at ISO1600 and I get very useable images (with a motion-stopping 1/200 shutter speed). If my DOF is too small at f/1.4, I can move to 1/200, f/2, ISO3200, and I'm still getting decent noise performance.

On the other hand, your kit lens is limited to f/5.6 and you've got motion blur @ 1/50 AND your images are noisy. With your slow zoom, you can get 1/200, but you're going to have to use ISO 25600; good luck with that.

Forget about noise performance. All modern DSLRs are good enough unless you are earning six figures as a photographer. Buy the cheapest body that serves your purposes and spend the bulk of your money on fast lenses, and preferably fast primes. That's the way people dealt with low light in the film era, and the same should hold true for digital. High ISO does not replace fast glass (nor, it should be said, does it replace proper flash photography).

Forget about noise.

 

[toxic]

Was that really all you took from the previous post?

no, I'm just clarifying that they aren't the same sensor.

 

[Edge100]

no, I'm just clarifying that they aren't the same sensor.

Thanks for adding to the discussion, then. Glad I didn't make any spelling or grammar mistakes.

 

[toxic]

Thanks for adding to the discussion, then. Glad I didn't make any spelling or grammar mistakes.

quit being spiteful. are you ok with spreading misinformation? you even wondered yourself why the 50D and 500D would have different performance with the same sensor. well, they don't have the same sensor.

 

[Edge100]

quit being spiteful. are you ok with spreading misinformation? you even wondered yourself why the 50D and 500D would have different performance with the same sensor. well, they don't have the same sensor.

I take the time to give a proper answer to the OP; to address questions, to respond to other inquiries, and to give a well-thought out opinion.

Your contribution is to point out (correctly) a slight error in my post, which doesn't affect the conclusion one little bit. Thanks.

When the error was noted, I owned up to it. The sensor is not the same; the pixel pitch is. This still doesn't affect the conclusion one little bit, and that is that a Rebel series camera can have lower noise than an XXD-series camera.

 

[Chris7]

Phrasikleia, OreoCookie, Edge, and toxic,

A sincere thanks for your help.

-Chris