Edac2: Why can't a 60 x 60mm sensor be made? 43.8 x 32.9mm is nowhere near "full frame" and there will be a significant loss of focal length using this back (43.8 is 73% of 60, and 32.9 is only 55% of 60). Perhaps having the sensor plane further from the lens than a film plane compensates for this; then you would just lose some of the square image at the top and bottom of the frame. There would be other compromises, though, such as not being able to use the camera's viewfinder. Would one keep the mirror flipped up to use this back?
6x6 is more like 54mm x 54mm.
But I do agree the sensor should cover much more for this to be interesting compared to the 36x24mm format.
Holger Drallmeyer: I didn't know low ISO is such a big deal. After all my 5 year old original Canon 5D has ISO 50. Just wondering.
Sure, but you didn't understand what I wrote.
My comment was about the 50 ISO setting in current Canon cameras (esp. 5D series). Shooting at 50 ISO makes no difference to sensor amplification so setting 5D series camera to 50 ISO doesn't result in darker RAW image.
No :) For the same exposure an ISO 50 image should receive twice as much light. Because with Canon it is only a "fake" setting, an image shot with ISO 50 is just ISO 100 image with exposure shifted towards shadows by 1 stop.
This means that highlight range is clipped but shadows are somewhat cleaner.
Deano255: 1) Flat profile should be a firmware update2) Adding Zebra for exposure is fine, not adding focus peaking at the same time makes no sense. Video pros use both as we're in manual mode most of the time.3) 60fps is adequate for slow motion. Nikon is basically catching up with everyone else here. 96fps or even 120fps would have turned heads.4) LiveView is still poorly implemented. Why can't we get a more sophisticated implementation of this feature? Are there too many technical limits due to locking up the mirror? This design has been baked in for way too many years now.
As a D800 user there us nothing here that will convince me to upgrade. For new users the Sony A7R and A7S are compelling alternatives.
Yeah sure...IMHO Nikon needs to fix their coke bottle lenses 50/1.4G and 58/1.4G asap.
The latter lens should never have been released with such mediocre wide aperture performance. Zeiss and Sigma really are much better choices if one is after better IQ.
A7R is still nice alternative because it can adapt almost every other lens in the market. Nikon is more limited and lens line up lacks some must have lenses like alternative to TS-E 17mm.
D800 was is not too good with LV shooting and consumes battery very fast in that mode. I hope D810 improves 1:1 zoom quality, refresh rate and shooting speed with LV mode activated.
You really should know that 5D series doesn't have true ISO 50. It is just ISO 100 overexposed by one stop.
True ISO 64 might have even better DR than D800 @ ISO 100, which was already better by over 2 stops compared to 5D mkIII.
Plus the IQ of original 5D looses to even MFT cameras of today.
ZhanMInG12: Pretty pointless article. E-mount mirrorless cameras have been having true ISO 100 since 2012, and Canon has offered a true ISO 50 on their high-end bodies since the early 2000s. ISO 64 is less than 2/3 stops of gain over ISO 100, so you save maybe one stop of neutral density filtration - not very useful.
And I seriously doubt how many people are going to use the D810 for video. This sensor cannot be used without line skipping, and the small photosites don't make for very good high iso performance in video. You'll be much better served by an A7S or GH4 or, if you absolutely need RAW video, a hacked 5d mk3.
I don't think Canon that has had true ISO 50. Rather they're ISO 100 images overexposed by 1 stop. This also means that DR is reduced by one stop in highlights.
What is interesting is the potentially higher DR of D810 vs. D800.
Otherwise I do agree about the video.
Just another Canon shooter: The only camera that cannot resolve the whole text here, at ISO 6400, is ... the A7S!
So after some NR, the A7S is the worse low light camera of the four?
That sample region is from a better exposed spot. A7S gives much cleaner shadow areas with more detail.
Fe. my D800 has ugly striped noise after at higher ISO settings, which is much harder to clean than usualy chroma noise.
If 12MP is enough for resolution in a specific situation, there's no reason to use high resolution sensor IMHO. Larger pixels gives better IQ.
Naturally higher resolution sensor can record more detail if noise levels are in control.
How about this selection? Clearly A7S gives much better high ISO dynamic range and shadow noise than any other current full frame camera.
Zeisschen: oh oooh...
I see a lot if Canikon users getting very nervous here. Is it because they paid 3 times more for the same performance and have to carry 3 times the weight ;-)Kudos to Sony for making these incredible cameras available for us customers at more affordable prizes is all I have to say!
In this reality the 7D looses to even OM-D EM-5 in sensor performance.
At 12.800 ISO the 7D has ca. 3 stops lower dynamic range than the Alpha 7S. That is a very large difference.
Aspenz: Don't like it, don't buy it, simple as.
The IQ is pretty decent and high iso has improved. I've never seen a difference between the M43 and N1 images anyway apart from iso, and it's more or less equal now. If anything, the Olympus cameras have sharpening haloes and M43 in general has poor rendition of highlights. It's funny when the M43 crowd equate themselves with APS-C or worse still FF, when there clearly is a difference.
And as good as these are they're still jpegs. If you know what you're doing with the raw files there's lots of details to be gotten.
I tend to trust DXOmark values for sensor performance. I've found them quite accurate compared to my own experience with several cameras. For example, EOS 5D mkII has worse dynamic range and shadow noise than OM-D EM-5 at base ISO. And V3 is quite far behind of the best m4/3 sensors in terms of SNR and DR.
Right... have you looked at DXOmark scores of V3 vs. EM-5?
OM-D wins hands down in every aspect. Almost two stops better DR btw.
No, I would never use a 12" touch screen for anything as complex piece of software as Photoshop is. Plus the screen would get too greasy already after 30 minutes use.
beenthere: Canon/Nikon/Sony be afraid, be very afraid..
Toy cameras as seen in smartphones are for holiday use at most. Meanwhile professional and amateur photographer will continue to use real cameras with much better shooting performance.
mosc: The dynamic range page here is one of the best pieces of technical analysis DPR's done. Nicely done guys!
These are measurements of JPEG dynamic range, and so are largely irrelevant for RAW performance which is much much more interesting.
Nikon D4S, D800 and Sony A7/7R beat the heck out of 5D mkIII in RAW dynamic range as reported by DXOMark. By more than 2 stops.
This is easily confirmed by field tests while shooting high contrast scenes.
iae aa eia: I think it was supposed to be already mirroless, with a nice EVF, and its sensor 56 x 41.5mm, as it was the 645 film frame area. 33 x 44mm looks like what the APS format is to the 135 full-frame, a cropped sensor to cut their investment some slack. Not a true medium-medium. Just medium.
Yes, I can and do for landscapes. But rotating camera usually changes perspective vs. using shift. I prefer latter method more for architecture or interior.
Phase One and Hasselblad uses ca. 54x40mm and 49x37mm sensor sizes in their top models. So there is no standard size. These bigger sensors make more sense for wide angle needs.
44x33mm is only 68% bigger than 36x24mm sensor. About 2/3 stops.
By using the Canon TS-E 24mm/3.5L II on Sony A7R to make 3 image stitch along shorter edge produces ca. 70MP image with nice resolution ("sensor size" ca. 36x48mm). Same with the TS-E 17mm/4 L.
Dave Luttmann: Sony, Nikon and Pentax have pulled ahead of Canon so much it is rather sad.
Canon has advantage in lenses. It has choices in its lens line up that no competitor has and AF system works very well.
It is NOT standard size as there is various sensor sizes in MF world. Some of them are more close to 645 format film area.
And those focal lengths available look about the same as in film era of 645 format. Which means not very wide lenses available. Also there is no tilt shift option.
pancromat: going to their website i don't see spotting scopes and binoculars. i see a manufacturer of high precision industrial and CCTV lenses. so they chose a few lenses off their portfolio which are capable of 4/3" and affordable, put on a MFT mount - thats it. no big invention, no "joining the MFT waggon".
And your logic is?
FT is not the same mount as MFT and also the focus & aperture controls look very different to me. So obviously more than just "putting on a MFT" was needed for producing these.
Especially those wide angle versions would be very promising if the lenses actually deliver high resolution and low distortion as promised.