Anyone have any idea how much low light performance we're sacrificing with the 7D versus the 5DMkII? The sensor is a lot smaller and the handful of clips I've seen so far just don't seem to have that contrasty, dramatic "bite" that 5DMkII clips seem to have. The image seems subtly "noisier" as well.

Thrilled about 24p and the price, but I guess I'm hoping to see some genuinely inspiring footage soon, like this minor classic shot on a 5DMkII:

1st 1080p from my Canon 5D MKII on Vimeo (http://www.vimeo.com/2403682)

The smaller the chips, the more light you need. Same for any camera.

You won't notice a difference - or it might actually be better than the 5D.

The issue with sensitivity and noise is mitigated by the fact that the 7D is based off an a newer generation system than the 5D2 itself. There are a lot of other features here at stake besides a cursory generalization about a "smaller" sensor. That doesn't even begin to scratch the surface.

True gapless micro-lenses on the 7D like the 50D - so that the sensor can pull more light and increase density of photosites on the sensor without increasing associated noise. When the 50D came out, this was a big new feature, but because the development cycle of the 5D Mark 2 is closely linked to the 1Ds Mark 3, it was impossible to include this feature in the final release.

The 5D2 has gaps between photosites and the microlenses so it is unable to fully utilize the surface area of the sensor chip itself.

Did you know that the photosite density of the 7D compared to the 5D2 is over two times as dense over the same area of space? If you made a Full Frame Sensor with the pixel density of the 7D, it'd be somewhere around 44 megapixels.

Another neat feature is the actual data processing scheme of the 7D and the 5D2 for video. While the sensor is 18.1 megapixels, full 1080p is not even close with a scant 2 megapixels. What do you think Canon does with all these spare pixels? It uses them to calculate noise data and how to remove it without creating a blurred, flat, and lifeless image.

The 7D is pretty exciting stuff. I know I'm excited. :)

My cursory read of the spec sheet tells me that the expanded ISO range is somewhat reduced (to H1, 12800 effective ISO). I wonder if we'll be able to access the full range for video, or if we will be limited to an upper ISO of 6400.

I took some grey and black sections from the 7D sample photos on dpreview and tried to compare it to the noise levels produced in the dpreview noise test. See attached bmp.

In my opinion the ISO 800 of the 7D matches best the 1600 of the 5D and the 3200 matches 6400 respectively.

This is not a precise test, merely a rough indication that the 7D is about half as sensitive as the 5D.


Material taken from here:

Canon EOS 5D Mark II Review: 21. Photographic tests (Noise): Digital Photography Review (http://www.dpreview.com/reviews/CanonEOS5DMarkII/page21.asp)

and here:

Canon EOS 7D Hands-on Preview: 14. Samples: Digital Photography Review (http://www.dpreview.com/previews/canoneos7d/page14.asp)

Thomas

I was wondering about the dynamic range, it looked like there were more blown highlights in 7D vids, but maybe we haven't seen it in the most capable hands yet.

Anyone have any idea how much low light performance we're sacrificing with the 7D versus the 5DMkII? The sensor is a lot smaller and the handful of clips I've seen so far just don't seem to have that contrasty, dramatic "bite" that 5DMkII clips seem to have. The image seems subtly "noisier" as well.

Thrilled about 24p and the price, but I guess I'm hoping to see some genuinely inspiring footage soon, like this minor classic shot on a 5DMkII:

1st 1080p from my Canon 5D MKII on Vimeo (http://www.vimeo.com/2403682)

you'll have to wait. Per pixel noise coming off the chip will be worse, but the 7D may use more sophisticated processing. I wouldn't extrapolate noise in stills to noise in video.

<snip> the 7D is about half as sensitive as the 5D.</snip>

Half, or twice? If 800 on the 7D is 1600 on the 5D2, I'd say the new camera is more sensitive (requiring less amplification).

Any conclusions drawn from that comparison are baseless.

When DPReview does their comparison, they do so under repeatable circumstances.

The samples from the 7D posted in that comparison could have been taken under any number of circumstances. All of which absolutely will not match DPR's studio comparisons.

It's possible that the blacks look so dreadful because it's in a situation where the photo is extremely bright and the photo is pulling detail out of the blacks where there is none. We have no idea of what the settings were and which photos these comparisons were taken from and whether or not the lighting was the same color temperature and in the same volume as well.

This is nowhere near an even comparison of what real world performance is going to be for the 7D vs the 5D2.

Yes. I see what you mean.. I was just going to buy a 5D too..
Problem is wide angle lenses. The Sigma 30mm f1.4 is the only option for a lowlight standard lens..
The full frame has the excellent 24mm f1.4 ..
With APS-C there is no equivalent..
What I mean is, that its nearly impossible to get the f1.2 or f1 lenses on the APS-C sensor for wide angles..So if you try to shoot low light, APS-C is a handicap..

However, can anyone try to use Super 35mm lenses on the APS-C sensor ? There are some nice lenses there..
Half, or twice? If 800 on the 7D is 1600 on the 5D2, I'd say the new camera is more sensitive (requiring less amplification).

Half the sensitivity, in my opinion the noise in the greys on the 7D at ISO 800 looks as strong as the 5D at 1600. The same is true for 3200 and 6400. The noise in the greys looks comparable to the 40D at same ISO, which is incredible for cramming twice the photodiodes onto the same chipsize.

I sense that some people felt offended at stating something like that based on very patchy data. My intent has been only to contribute something factual to the discussion. Please believe me that it was not my intent to offend. And, as indicated, very imprecise data and only a rough indication.

No need to speculate, look no further - Rob Galbarith has high ISO pics comparing the 7D with 5D Mark 2. The 5D Mark 2 is noticably cleaner at high ISO..

Rob Galbraith DPI: Canon announces 17.92 million image pixel, 8fps EOS 7D (http://www.robgalbraith.com/bins/multi_page.asp?cid=7-10042-10239)

Just looking at the samples again, even the 400ISO pics show a LOT more grain/noise than the 5D2 or even my old 20D for that matter... Yikes! http://www.robgalbraith.com/public_files/Canon_EOS_7D_ISO400_Thunder.jpg

Good for a small sensor with such higher resolution but it doesn't compare against a larger sensor.

You won't notice a difference - or it might actually be better than the 5D.

Another neat feature is the actual data processing scheme of the 7D and the 5D2 for video. While the sensor is 18.1 megapixels, full 1080p is not even close with a scant 2 megapixels. What do you think Canon does with all these spare pixels? It uses them to calculate noise data and how to remove it without creating a blurred, flat, and lifeless image.

The 7D is pretty exciting stuff. I know I'm excited. :)

Nope, they won't use another part of the sensor to calculate noise data and apply that elsewhere. It's not a uniform noise pattern across the entire sensor.

IMHO, the noise levels on this camera will make or break it.

The photo community thought the 50D had too many megapixels and if the 7D has noise issues then 18 megapixels actually looks like a step backward.

Although waiting to pass judgement is the best solution as the 5D mkII was supposed to have "too many megapixels" to improve over the 5D, and it delivered.

I don't think Canon would release a noisy camera at this pricepoint or market segment.

I think comparing the cameras, based on noise levels between the 5d2 and the 7d is perhaps a bit of a folly. These are two cameras intended for different users with different sensor sizes and architectures. The correct comparison is with the 50d, or really with the D300s.

Imaging Resource has in fact done this with a pre-production camera...and I think the 7d holds up just fine in the noise department (the color seems to be tuned differently than the other cameras...which may be a pre-production unit issue...or not).

Imaging Resource "Comparometer" ™ Digital Camera Image Comparison Page (http://www.imaging-resource.com/IMCOMP/COMPS01.HTM)

and for those that insist that the 7d is somehow related to the 5d and thus should be put on the same pedestal..I think it's quite surprising how WELL the 7d does in comparison to the 5d in IR's comparator...I don't think it bests the 5d (it shouldn't)...but it's not blown away, either. Certainly much closer than I would have thought.

Barry, you are correct.

But, I use the 5D MkII and need a second camera/backup for still work.

Right now I have a 30D which is a fine camera but does not run with the 5D MKII.

So I am viewing the 7D as a sister camera to the 5D MKII.

It is really useful to have a full frame and a crop with compareable image quality so you don't feel like you are stepping down using the other camera.

I hope they use the same battery system as well.

So I think many people are comparing them because they want them to work as a duo.

I hope they use the same battery system as well.


It uses the same battery as the 5D Mark II. I'm pretty sure I saw that in one of the articles/reviews of the 7D. maybe dpreviews?

Thanks.

I just checked it at DPReview and you are correct.

This is good news for me.

The smaller the chips, the more light you need. Same for any camera.

My understanding is that it is the size of the individual pixels rather than size of the sensor that matters most (assuming a 1/3" sensor size or higher).

I don't think we'll really know until real DPs get their claws on this thing. And as many have said, the speed of the glass is going to be a big factor.

All that said, as a maker of penny dreadfuls, I sure hope it's close to the Mark II.

jdv