Luminous Landscape Forum
Equipment & Techniques => Cameras, Lenses and Shooting gear => Topic started by: BernardLanguillier on August 24, 2011, 08:49:02 pm
-
Tough days for the Canon and Nikon shooters focused on the specs of their cameras! ;D
The R&D power of Sony is impressive. It will be interesting to see how the big guys will manage to answer to the beating they are taking specwise (and possibly also qualitywise if we extrapolate based on recent sensors from Sony). That is probably going to translate soon in sales.
The only question for me remains whether an EVF is the right option or not.
Other than that the new Sony line up is very interesting! The ability to use a given set of lenses for stills and for a dedicated video camera sounds a lot for relevants than attempts to use DSLR form factors devices for video applications. Canon is IMHO going to loose big if they don't react soon.
Cheers,
Bernard
-
Nex-7 with Leica glass.....I'll take one sight-unseen! They are going to sell as many of those as they can make if they are any good. My 24 Elmar and 35 Cron are chomping at the bit.
- N.
-
What I want is an upgraded D700 with D7000 pixels, but I guess that will be a long time coming. ;D
-
Nex-7 with Leica glass.....I'll take one sight-unseen!
I would do the above with a NEX9 (Full-frame sensor), assuming they release a couple of key FF lenses that cover some of the oft-used FLs.
-
Nex-7 with Leica glass.....I'll take one sight-unseen! They are going to sell as many of those as they can make if they are any good. My 24 Elmar and 35 Cron are chomping at the bit.
- N.
24 megapixels in an APSc size sensor....it begs the question: What does this density convert to in a full frame 24x36mm sensor? I already have a A900---it is magnificent!
Dave
-
24 megapixels in an APSc size sensor....it begs the question: What does this density convert to in a full frame 24x36mm sensor? I already have a A900---it is magnificent!
Dave
Taking into account:
1) dimensions of this new Sony sensor: 23.4mm x 15.6mm = 365.04 sq.mm area;
2) FF sensor dimensions: 24mm x 36mm = 864 sq.mm area;
3) FF sensor area is 2.367 times bigger than this new Sony aps-c sensor (864sq.mm / 364.04sq.mm = 2,366863905325444);
4) Sony new sensor is 24.3 Mpix,
FF sensor with the same pixel density would be around 57.5 Mpix (24.3 Mpix x 2,366863905325444 = 57,51479289940828)
BR,
Roberts
-
I'm unsure about all of this. I have a Sony 850. Going back to a crop sensor camera is not very appealing but I do want to have video capability.
-
I'm unsure about all of this. I have a Sony 850. Going back to a crop sensor camera is not very appealing but I do want to have video capability.
Craig: I agree with you. I have a A900, full frame camera, and also use a Pentax 645D, so I hesitate to go back to a small sensor. BUT the small, light weight 24 Mp camera that accepts my existing lenses is so tempting I asked my dealer to notify me when they receive one.
Dave
-
Hey Dave,
If you ever want to sell the 645D can you contact me??
Mike
-
Hey Dave,
If you ever want to sell the 645D can you contact me??
Mike
I'm taking it with me to my grave. It is the most straight forward camera I have ever used. I am at this minute printing out a 24"x32" image. It looks like this:
-
I'm unsure about all of this. I have a Sony 850. Going back to a crop sensor camera is not very appealing but I do want to have video capability.
I've gone down this road, and I haven't missed it THAT much. After using the NEX-5 and rangefinder lenses for a few months, I sold my A900 system that I'd been shooting for a couple of years. Sure, there are slight differences in the print size that I make, but it isn't drastic, and I expect the NEX-7 to make up most of the difference. The only thing to get used to is that lost stop of equivalent DOF, but I'm not a super shallow DOF kind of guy all of the time, so the transition has been ok for me. Being able to have such a small gear footprint has been a godsend.
-
It's awful when a new camera comes out and all of a sudden the pictures from my current camera start looking like crap. Hate when that happens.
-
The nex-7 seems really attractive to me but I wonder if you will really be able to take advantage of the 24mp on the sensor. With the 5dII I find that I have to shoot at x2 the focal length to get sharp images handheld -- I imagine it would be a much bigger problem given the greater pixel density.
-
The nex-7 seems really attractive to me but I wonder if you will really be able to take advantage of the 24mp on the sensor. With the 5dII I find that I have to shoot at x2 the focal length to get sharp images handheld -- I imagine it would be a much bigger problem given the greater pixel density.
Some advantages of the NEX (over a dSLR) are the lack of vibration from mirror-slap and also lack of vibration from the shutter (due to the electronic first curtain), until after the exposure,, all of which conspire to make it a lot more hand-holdable than an equivalent SLR.
However, your point is true that with such high pixel densities, one will have to use perfect technique (preferably on a solid tripod) to get the most benefit from it.
-
The nex-7 seems really attractive to me but I wonder if you will really be able to take advantage of the 24mp on the sensor. With the 5dII I find that I have to shoot at x2 the focal length to get sharp images handheld -- I imagine it would be a much bigger problem given the greater pixel density.
I'm not sure there should be a bigger problem if both cameras (full frame and cropped-format) have similar pixel count. Not much difference between 21mp and 24mp.
If you find that you need a shutter speed of 1/2FL with the 5D2, then the same requirements should apply to a 24mp cropped format, provided the FL figure used is the FF 35mm equivalent, ie. 1/160th with an 80mm lens on the 5D2 becomes 1/160th with a 50mm lens on the cropped format, both situations providing the same FoV and the same resolution of image (approximately).
Nevertheless, as pixel count increases, faster shutter speeds are required for the larger images or prints that are possible. A base ISO of 200 is really nice. Both my Nikon D700 and Canon 50D have a base ISO of 200.
However, there's no free lunch. My D7000 at its slower base ISO of 100 provides significantly greater dynamic range than the D700 and 50D at ISO 200. Even at ISO 200, the D7000 provides 0.75 stops greater DR than my D700 and a whopping 1.5 stops greater DR than the 50D.
Since the D7000 uses a Sony sensor, I'd be surprised if the DR at ISO 200 in these new cameras were not better than the DR of the 5D2 at its base ISO of 100, although the 5D2 should still have lower noise at 18% grey because it's a full frame format.
-
Nice picture. What if they bring out a 60mp version next year?
Mike
-
I've gone down this road, and I haven't missed it THAT much. After using the NEX-5 and rangefinder lenses for a few months, I sold my A900 system that I'd been shooting for a couple of years. Sure, there are slight differences in the print size that I make, but it isn't drastic, and I expect the NEX-7 to make up most of the difference. The only thing to get used to is that lost stop of equivalent DOF, but I'm not a super shallow DOF kind of guy all of the time, so the transition has been ok for me. Being able to have such a small gear footprint has been a godsend.
My concern is I shoot some interiors and am not sure what lenses I would be using to replace a 24mm focal length for instance. Some rooms require even wider FL than 24mm with full frame. So for me to get back to a 24mm view I'd wind up with a 16mm. Things start to look a little weird with super wide angle lenses.
-
I'm not sure there should be a bigger problem if both cameras (full frame and cropped-format) have similar pixel count. Not much difference between 21mp and 24mp.
If you find that you need a shutter speed of 1/2FL with the 5D2, then the same requirements should apply to a 24mp cropped format, provided the FL figure used is the FF 35mm equivalent, ie. 1/160th with an 80mm lens on the 5D2 becomes 1/160th with a 50mm lens on the cropped format, both situations providing the same FoV and the same resolution of image (approximately).
Nevertheless, as pixel count increases, faster shutter speeds are required for the larger images or prints that are possible. A base ISO of 200 is really nice. Both my Nikon D700 and Canon 50D have a base ISO of 200.
However, there's no free lunch. My D7000 at its slower base ISO of 100 provides significantly greater dynamic range than the D700 and 50D at ISO 200. Even at ISO 200, the D7000 provides 0.75 stops greater DR than my D700 and a whopping 1.5 stops greater DR than the 50D.
Since the D7000 uses a Sony sensor, I'd be surprised if the DR at ISO 200 in these new cameras were not better than the DR of the 5D2 at its base ISO of 100, although the 5D2 should still have lower noise at 18% grey because it's a full frame format.
I believe the A77 and NEX-7 have a base ISO of 100.
-
I believe the A77 and NEX-7 have a base ISO of 100.
I remember the time they stated that a Full-frame version of the Leica M-rangefinder was an impossibility and that at best it will need to be a 1.3x.....well, we have had a FF Leica-MR for a while now. As long as the mount is large enough (and in fact the Sony E-mount is larger than the Full-frame Leica M-mount) a FF will arrive. My prediction.
-
I believe the A77 and NEX-7 have a base ISO of 100.
Yes. It seems to be the case, which could be considered a disadvantage for such a high pixel-count sensor. But it also depends on the degree of IQ difference between ISO 100 and ISO 200.
On the D7k, there's a significant difference in DR and noise between ISOs 100 and 200, whereas on the 50D there isn't. In fact there isn't any difference at all. The only purpose of ISO 100 on the 50D, as far as I can tell, is to allow one to use a slower shutter speed to create blurring effects.
I wasn't aware of this fact when I fist started using the camera. If I had been there would have been many occasions when I would have used ISO 200 at F4 instead of ISO 100 at F2.8.
-
First impressions of the A77, compared with the D7000, don't look promising regarding deep-shadow noise. There seems to be some magenta-cast problem. However this may not be an issue in the production models. I understand in the following shot from the A77, Imaging Resources used a pre-production model.
The black cup and napkins have been raised with the middle slider in levels.
-
I'm taking it with me to my grave. .......
Hi Dave, LOL, quite an endorsement. How are you doing with Irene?
Tom
-
"Some advantages of the NEX (over a dSLR) are the lack of vibration from mirror-slap and also lack of vibration from the shutter (due to the electronic first curtain), until after the exposure"
Super :D but how does it fare outside the camera store?
Does EVF effect reliability in extreme conditions and battery drain that maybe depends as much on the design, efficiency & cost of the new Sony line-ups power system compared with existing range of DSLRs than if any given model has EVF or not, and while I am at it what about weather proofing?
-
"Some advantages of the NEX (over a dSLR) are the lack of vibration from mirror-slap and also lack of vibration from the shutter (due to the electronic first curtain), until after the exposure"
Super :D but how does it fare outside the camera store?
Does EVF effect reliability in extreme conditions and battery drain that maybe depends as much on the design, efficiency & cost of the new Sony line-ups power system compared with existing range of DSLRs than if any given model has EVF or not, and while I am at it what about weather proofing?
lowep, I think you're the kind of person that can dream up disadvantages of any system if you put your mind to it.
Any system you know that has no disadvantages? ;)
-
I guess it depends on who you ask ;D
actually what I was hoping was that somebody with experience with one of these things would be able to confirm that yes the EVF is more reliable than the old-fashioned SLR mirror systems and yes the latest generation Sony systems do use power more efficiently and yes the weather proofing is as good as with comparable DSLRs
but this by no means diminishes the acuity of your observation.
-
lowep, I thought you were asking a retorical question, since you asked how these held up "outside the store".
None of them has been "outside the store" (for the general public I mean) since release is still a few weeks away.
From the press events the announced weather sealing of the A77 and new 16-50 has been tested a bit in a power boat and held up nicely.
Officially the Nex7 and A65 are not weathersealed, but no word on any experience what that really means.
I haven't found anything on battery life in actual use conditions yet.
-
Here (http://www.sonynex7.net/) is what Sony says:
Sony is taking professional quality imaging and artistic shooting options to new levels with the introduction of the new α NEX-7 camera...the viewfinder provides a bright, detail-packed, high contrast view of still images and video, with approximately 100 percent frame coverage and a wide field of view that rivals pro-grade optical viewfinders...Two control dials on the top of the camera are complemented by a control wheel, giving users three customizable, manual controls and the ability to create images with the flexibility and precision of an advanced DSLR camera.
-
Here (http://www.sonynex7.net/) is what Sony says:
and......... ???
-
Sorry for not making my point more clearly that I try to do by adding bold:
Sony is taking professional quality imaging and artistic shooting options to new levels with the introduction of the new α NEX-7 camera...the viewfinder provides a bright, detail-packed, high contrast view of still images and video, with approximately 100 percent frame coverage and a wide field of view that rivals pro-grade optical viewfinders...Two control dials on the top of the camera are complemented by a control wheel, giving users three customizable, manual controls and the ability to create images with the flexibility and precision of an advanced DSLR camera.
what I am really trying to figure out is could this be a viable alternative to a full frame DSLR for my particular needs? I guess the only way to find out is as you say wait and see what users who buy one have to say so best for me to withdraw for now with thanks for the feedback.
-
OK, I see what you're geting at
Well, for me it's just marketing speak. Let me give you my 5 c how I read these terms:
professional quality imaging: without a better definition that doesn't tell us anything, there's professionals that shoot with a holga ;D
rivals pro-grade optical viewfinders: what does "rival" mean: better, as good as, almost as good, just as useful,..... for me it's horses for courses, OVF's will still trump in clarity, brightness, immediate feedback but are lousy for judging dof of the final picture and can't zoom in. EVF's will do much better at judging dof and can zoom in instantly for a precise focus check but will still (and some say always) lack the direct feedback and clarity of an image projected on a ground glass.
flexibility and precision of an advanced DSLR camera: I would read this as "advanced DSLR camera with the same sensor size", but maybe I'm too forgiving for the marketing dept here.
Obviously they want you to read more into it than I just wrote above and rush out to buy the camera unseen, but if you really analyse it their use of words it's just marketing speak that needs to be validated by real life test of people we trust (which is why we come to LL :))
-
Sorry for not making my point more clearly that I try to do by adding bold:
Sony is taking professional quality imaging and artistic shooting options to new levels with the introduction of the new α NEX-7 camera...the viewfinder provides a bright, detail-packed, high contrast view of still images and video, with approximately 100 percent frame coverage and a wide field of view that rivals pro-grade optical viewfinders...Two control dials on the top of the camera are complemented by a control wheel, giving users three customizable, manual controls and the ability to create images with the flexibility and precision of an advanced DSLR camera.
what I am really trying to figure out is could this be a viable alternative to a full frame DSLR for my particular needs? I guess the only way to find out is as you say wait and see what users who buy one have to say so best for me to withdraw for now with thanks for the feedback.
Well, the Full-frame Sony A900 has one of the largest/brightest 100% optical viewfinders among all makes/models, regardless of pricepoint (larger than the Nikon D3s/D3x). The NEX7 and A77 viewfinders have been compared favorably against the A900, for size/brightness under most shooting conditions. One additional advantage is the additional information that is present on the top and bottom of the EVF, that is not available in the OVF. Sure, being an EVF, it will of course consume some power and display some lag when it comes to panning fast and so on, but for most shooting situations it should do just fine.
As with any such thing, there are pluses and minuses from all such systems, whether it be an OVF-based one or whether it be an EVF based one and you need to pick the one that works best for you. What I would not do, is to bad-mouth something that you have never used or worked with.
-
Well, the Full-frame Sony A900 has one of the largest/brightest 100% optical viewfinders among all makes/models, regardless of pricepoint (larger than the Nikon D3s/D3x). The NEX7 and A77 viewfinders have been compared favorably against the A900, for size/brightness under most shooting conditions. ..........[some text deleted for space considerations].......
As with any such thing, there are pluses and minuses from all such systems, whether it be an OVF-based one or whether it be an EVF based one and you need to pick the one that works best for you. ............
The EVF on the A77 is the largest, brightest, highest resolution and most contrasty electronic viewfinder yet, from anyone. Incidentely, it is the same 2.4 MP OLED device as the one on the A65, the NEX-7, and the FDA-EV1 accessory viewfinder for the NEX-5N.
Is it good? Yes, it's very good – for an EVF. I would go so far as to say that it is much preferable to the dim and small reflex viewfinders found on most reduced frame DSLRs. This then begs the question – how does it compare to the viewfinder on a full frame DSLR like the Sony A900.
Not so well. The viewfinder on the A900 is one of the brightest and largest found on any 35mm DSLR – ever. Sony's new OLED EVF, as good as it is, is up against what may the gold standard, and it doesn't quite deliver.
While very good as EVF's come, Michael doesn't seem to agree regarding your statement on how it compares to the quality of the A900 OVF. Since I have an A850 (OVF nearly as good as the A900) I can't wait to actually hold and test one of these new OLED EVF's myself to further make up my mind.
On the other hand I fully agree with you that EVF's have some inherent advantages over OVF's in different aspects (dof preview and zooming, see my post above).
-
I remember the time they stated that a Full-frame version of the Leica M-rangefinder was an impossibility and that at best it will need to be a 1.3x.....well, we have had a FF Leica-MR for a while now. As long as the mount is large enough (and in fact the Sony E-mount is larger than the Full-frame Leica M-mount) a FF will arrive. My prediction.
I'm not sure what that has to do with ISO 100, but the big difference there (and with 135 digital cameras in general) is that Leica has decades worth of legacy lens users to make happy. Nex is an entirely new system with lenses that are designed from the ground up for Aps-c. While Sony could fit a 135 sensor in the Nex mount, if they moved the electrical contacts, why would they? To sell Leica lenses?
-
56 mpx..... of bad quality!
-
No it won't! The 24mpx sensor of the Nex-7 will give to the camera much better video quality and worst photo quality than the Nex-5, same will apply to the a77 vs. a55 comparison! Hence the decision of Sony to cut the real viewfinder from the a-77! The 24 mpx sensor results in exactly 3x1920pix and hence 3x1080 vertically (after we throw away the top and bottom parts to create 16:9 from 3:2). This means, that there will be exactly 9 pixels recording for each of the imaginary 1920x1080 pixel, so...., there will be improved video sharpness because there is no interpolation in translating the 24mpx info to video2k, there will also be better video color, because again, there is exact color info for each of the 1080x1920pix info from the 9 pixels that are an RGB poled mixture!!! There will be worst photo performance because: a) the sensor has solutions to improve video (e.g. stronger AA) but damage stills b) the bloody sensor is too small to avoid noise, which will result in better per pixel definition of the Nex-5/a55!!!
Why no FF? Simply because FF is wrong for filming...(!), too shallow depth of film....(!), in cinema 35mm film is shot vertically (in a width of 22-24mm) which is identical to APS-C sensor! Even the REDs have an APS-c sensor. No pro videocamera has an FF sensor and since Sony traditionally is a major video market leader and is (like everybody else) going hybrid, ......NO FF!!!!
-
Hi,
I guess that the major issue is that a stills camera has an OLP (Optical Low Pass) filter that has been optimized for small sensor pixels. Whatever sampling method is used for video it will not be supported by an optimized OLP. On the other hand there will be a significant oversampling and that may be used for antialiasing in software but I'm not sure in camera electronics are fast enough.
If downsampling is done by binning there would be little negative effect on noise characteristics weather done in hardware or software. If unused pixels are simply discarded noise will significantly increase. But the difference between 24MP and 16MP would be small, anyway. Barely measurable?
Best regards
Erik
No it won't! The 24mpx sensor of the Nex-7 will give to the camera much better video quality and worst photo quality than the Nex-5, same will apply to the a77 vs, a55 comparison! Hence the decision of Sony to cut the real viewfinder from the a-77! The 24 mpx sensor results in exactly 3x1920pix and hence 3x1080 vertically (after we throw away the top and bottom parts). This means, that there will be exactly 9 pixels recording for each of the imaginary 1920x1080 pixel, so...., there will be improved video sharpness because there is no interpolation in translating the 24mpx info to video2k, there will also be better video color, because again, there is exact color info for each of the 1080x1920pix info from the 9 pixels that are an RGB mixture!!! There will be worst photo performance because: a) the sensor has solutions to improve video (e.g. stronger AA) but damage stills b) the bloody sensor is too small to avoid noise, which will result in better per pixel definition of the Nex-5/a55!!!
Why no FF? Simply because FF is wrong for filming...(!), too shallow depth of film....(!), in cinema 35mm film is shot vertically (in a width of 22-24mm) which is identical to APS-C sensor! Even the REDs have an APS-c sensor. No pro videocamera has an FF sensor and since Sony who is traditionally a major video market leader is (like everybody else) going hybrid, ......NO FF!!!!
-
Hi,
I guess it may depend on viewing conditions. Electronic viewfinders show actual image and are more flexible.
The APS-C cameras using EVF should be compared to the Sony Alpha 700 which had similar sensor size.
I use both Alpha 55 and Alpha 900 and there are advantages to both.
My view is that you would go with different lens systems for APS-C and full frame. A 24-70/2.8 is in my view not very practical on an APS-C camera for instance, even if it would be very sharp.
Best regards
Erik
While very good as EVF's come, Michael doesn't seem to agree regarding your statement on how it compares to the quality of the A900 OVF. Since I have an A850 (OVF nearly as good as the A900) I can't wait to actually hold and test one of these new OLED EVF's myself to further make up my mind.
On the other hand I fully agree with you that EVF's have some inherent advantages over OVF's in different aspects (dof preview and zooming, see my post above).
-
Hi,
I guess that the major issue is that a stills camera has an OLP (Optical Low Pass) filter that has been optimized for small sensor pixels. Whatever sampling method is used for video it will not be supported by an optimized OLP. On the other hand there will be a significant oversampling and that may be used for antialiasing in software but I'm not sure in camera electronics are fast enough.
If downsampling is done by binning there would be little negative effect on noise characteristics weather done in hardware or software. If unused pixels are simply discarded noise will significantly increase. But the difference between 24MP and 16MP would be small, anyway. Barely measurable?
Best regards
Erik
It's more important than you think Erik, as you mention, "unused pixels will significantly increase noise" and this happens on the NEX-5 (or any other photo camera that does video). On the other hand there is also much more simplicity in the camera logic on the NEX-7/a77, less electronics and hence more reliability. I think that Sony's intention with the NEX-7 (or the a77 if one chooses) is to give the a900/850 professional user that needs video, a cost/space effective way to do so, either with the a77 or with the NEX-7+adapter. I also think that people have not realize yet that a77 and (the same performance) NEX-7 will prove to be (with the proper lens) the best "interchangeable lens cameras" in the market for video quality! I bet they will prove to be much better than Canon Eos7 or 5mkii, but much worst for stills, even from the NEX-5. Of course you are right on the importance of OLP in stills performance. Best regards Theodoros, www.fotometria.gr
-
Some advantages of the NEX (over a dSLR) are the lack of vibration from mirror-slap and also lack of vibration from the shutter (due to the electronic first curtain), until after the exposure,, all of which conspire to make it a lot more hand-holdable than an equivalent SLR.
???? What does mirror-slap and shutter vibration have to do with handholding? For me anyway, hand tremor would obliterate any blurring you might see from these sources.
-
I bet they will prove to be much better than Canon Eos7 or 5mkii, but much worst for stills, even from the NEX-5.
You'd lose the bet on the NEX-5 comparison.
Michael
-
I bet they will prove to be much better than Canon Eos7 or 5mkii, but much worst for stills, even from the NEX-5.
You'd lose the bet on the NEX-5 comparison.
Michael
Obviously "bet" is just an expression, but I won't loose it Mike!!! The NEX-5 with the same lens used on it (via the Sony adapter of course) is the same as the a55 which is almost the same performer as Nikon d7k and Pentax k5 (trust me I've done the test using the same lens, a P-6 CZJ biometar!). My conclusion is that this 16.2 mpx sensor, will be on the top of APS-C STILLS performance for at least a few years! In the first report on the web of the a99 that the reporter used a pre production model, he did mention that it didn't do as good as the a55. Believe me, if it was any good for photography, Nikon or Pentax would have grabbed it already and Sony would have let the pentaprism on the camera to satisfy photographers as they did with the a55, but they removed it because it's useless for video and they want to maximize video performance! I believe that Nikon will use this 24mpx sensor in a near future model that will also have video priorities having stills as a supplement. Cheers Theodoros, www.fotometria.gr
-
Hi,
For stills the most important factor is sensor size not pixel size. Smaller pixels will have better resolution if the lens can keep up. Sharpness will be lost due to the OLP filter and that loss will be less with smaller pixels. Once a picture is scaled to a certain size the pixels are effectively binned regarding shot noise.
Interestigly, Michael Reichmann has found that image quality on the Alpha 55 is similar to the Alpha 900 and surpasses it at high ISO and he also found this to be the case with the A77. I have both cameras and although I have not made formal tests I'd say that the Alpha 55 performs very well in comparison. A great advantage of the Alpha 55 is that critical focus can be achieved using LV-focusing at actual pixels.
The major reason for Sony dropping the mirror and pentaprism/pentamirror is probably cost. It's an expensive assembly that needs to be adjusted within 10 microns. Live view can also show all kind of useful information in the finder like live histogram.
You may check this article to get some additional insight on the effect of pixel size on image quality:
http://www.luminous-landscape.com/essays/dxomark_sensor_for_benchmarking_cameras.shtml
My guess is that if Canons or Nikons surpass Sony in image quality it is because they may have better lenses.
Best regards
Erik
It's more important than you think Erik, as you mention, "unused pixels will significantly increase noise" and this happens on the NEX-5 (or any other photo camera that does video). On the other hand there is also much more simplicity in the camera logic on the NEX-7/a77, less electronics and hence more reliability. I think that Sony's intention with the NEX-7 (or the a77 if one chooses) is to give the a900/850 professional user that needs video, a cost/space effective way to do so, either with the a77 or with the NEX-7+adapter. I also think that people have not realize yet that a77 and (the same performance) NEX-7 will prove to be (with the proper lens) the best "interchangeable lens cameras" in the market for video quality! I bet they will prove to be much better than Canon Eos7 or 5mkii, but much worst for stills, even from the NEX-5. Of course you are right on the importance of OLP in stills performance. Best regards Theodoros, www.fotometria.gr
-
Hi,
For stills the most important factor is sensor size not pixel size. Smaller pixels will have better resolution if the lens can keep up. Sharpness will be lost due to the OLP filter and that loss will be less with smaller pixels. Once a picture is scaled to a certain size the pixels are effectively binned regarding shot noise.
Interestigly, Michael Reichmann has found that image quality on the Alpha 55 is similar to the Alpha 900 and surpasses it at high ISO and he also found this to be the case with the A77. I have both cameras and although I have not made formal tests I'd say that the Alpha 55 performs very well in comparison. A great advantage of the Alpha 55 is that critical focus can be achieved using LV-focusing at actual pixels.
The major reason for Sony dropping the mirror and pentaprism/pentamirror is probably cost. It's an expensive assembly that needs to be adjusted within 10 microns. Live view can also show all kind of useful information in the finder like live histogram.
You may check this article to get some additional insight on the effect of pixel size on image quality:
http://www.luminous-landscape.com/essays/dxomark_sensor_for_benchmarking_cameras.shtml
My guess is that if Canons or Nikons surpass Sony in image quality it is because they may have better lenses.
Best regards
Erik
Hmmm.... not always and not exactly. In stills resolution is one thing and sharpness another. If you print a large size print from my D7000 and then print THE SAME SIZE from my D700, you will find that the FF is considerably better. Interestingly, this is not so evident if you do the comparison on a monitor, but in prints its clear and since a photograph is the printed thing ONLY, the conclusion can only be that you print better from better pixel definition! The above test will lead to the same conclusion if you print same size with my imacon 528c DB and compare it with a D3X, the DB wins easily. BUT..., the smaller the pixel the worst the light reception, especially as we move away from the sensors exact center and of course problem is bigger if we use a wide angle and even more so on the wider FF sensor, thats why ALL manufacturers make their FF sensors with much larger pixel sizes than their APS-c sensors and in relevance, APS-C sensors have much wider pixels than smaller sensors. After having test a lot and taking advantage of the fact that I own a top class printing lab, (www.fotometria.gr) I have come to the conclusion that sensor resolution and quality must be balanced to an optimum and that with tech advancement we can increase analysis but we have to be sensible to this approach. Don't forget that: a) Physics says that the smaller the sensor the larger the N/S ratio b) Manufacturers can built an APS-c sensor with hundrents of MPX, (look at your cell phone and multiply up to the area). I'm not familiar with mr. Reichmann's findings, but I do remember that when I tested an a900 with my D700 I found the Sony considerably better on all aspects of photography below iso500 and only worst in noise above that, but clearly this had to do with Sony's decision to use as little NR as possible to maximize sharpness and of course to the smaller pixel than the D700. Hence, it would surprise me if the a55 is so much better than the D7000, to surpass the D700 and even the a900, clearly he must have done something wrong in the procedure of the test. However, I've read so many nonsense on web, that nothing will surprise me anymore! Cheers, Theodoros.
P.S.By the way, large sensor resolution can't affect video performance, because the recording area for 2K is huge anyway and YES a77 is a very promising videocamera that will be able to shoot acceptable stills! In fact I will be waiting for Nikon's version (no other reason than lenses and gear) to use as a my prime videocamera(s).
-
If you print a large size print from my D7000 and then print THE SAME SIZE from my D700, you will find that the FF is considerably better.
This does seem a little odd, unless you forgot to stipulate that this only applies above base ISO.
The D700 base ISO is 200. If one compares the performance parameters as measured by DXO, the D7000 at ISO 100 either outperforms the D700 (at ISO 200) or has equal performance in every category.
SNR at 18% grey is virtually the same, with a 0.1dB edge to the D700.
Tonal range is virtually the same, with a 0.06 bits edge to the D7000.
Color sensitivity is identical for both cameras at their respective base ISOs.
However, the Dynamic Range of the D7k is a whopping 1.72 EV better than the FF D700, at equal image size, ie. downsampling the 16mp D7k image to the 12mp of the D700.
Nevertheless, I have to admit that I've never compared equal size prints of the same scene taken with both cameras. If you have, making sure that both images are of identical scenes under identical lighting conditions, both equally exposed according to the clear and simple principles of ETTR, as expounded in the many threads on this site, both taken with lenses of comparable quality using different focal lengths and f/stops to ensure equal DoF and equal FoV, then this could be a revelatory observation you've made that deserves considerable investigation and explanation.
-
This does seem a little odd, unless you forgot to stipulate that this only applies above base ISO.
The D700 base ISO is 200. If one compares the performance parameters as measured by DXO, the D7000 at ISO 100 either outperforms the D700 (at ISO 200) or has equal performance in every category.
SNR at 18% grey is virtually the same, with a 0.1dB edge to the D700.
Tonal range is virtually the same, with a 0.06 bits edge to the D7000.
Color sensitivity is identical for both cameras at their respective base ISOs.
However, the Dynamic Range of the D7k is a whopping 1.72 EV better than the FF D700, at equal image size, ie. downsampling the 16mp D7k image to the 12mp of the D700.
Nevertheless, I have to admit that I've never compared equal size prints of the same scene taken with both cameras. If you have, making sure that both images are of identical scenes under identical lighting conditions, both equally exposed according to the clear and simple principles of ETTR, as expounded in the many threads on this site, both taken with lenses of comparable quality using different focal lengths and f/stops to ensure equal DoF and equal FoV, then this could be a revelatory observation you've made that deserves considerable investigation and explanation.
NO! It applies to base ISO as well, although you wouldn't mind the difference there! DXO measures sensors converted with their own RAW converter, not with the latest NX2.(...) that is optimized for both cameras, the mistake is the same as it happens with Dpreview, you just don't compare cameras because you use the same (third party) RAW converters in both of them, because you don't know if the converter is more compatible with one camera or the other! This is common sense isn't it? In other words you wouldn't compare two different films by developing them with third party chemicals and assume that the films have the same reaction to those chemicals as with the original chemistry that they where designed to be used with!!! Your impression of DR is not correct, again highlight DR is much more important than shadow DR and the total DR range is a personal matter that depends on the acceptance of the noise included in the printed info. IMO, a good approach would be to measure the linear part of the S-slope only(!), (yes, I mean 4-5 stops), IMO its a good indication of the camera capabilities, especially in the highlights. This doesn't mean that the curved parts of the S-slope are of no importance (please don't let me be misunderstood), I have find the D700 to be about half a stop better than the D7k in highlight DR (very important) and about a stop better in shadows (not so important, but welcome) for my noise standards in the shadows, both of them are more than a stop behind the Fuji S5pro in the highlights (the later with Hyper Utility and nothing else as a converter of course), in fact the fuji is the only DSLR that can compete (not beat) with a MF-DB in DR. If you print at 72dpi with the D700 (the limit), the D7k will be at 83dpi for the same size (100x150cm), believe me the difference is not so important. The method of comparing FF and APS-c performance by equalizing DOF is ridiculous, photographers don't do that to take a picture! They put their cameras to perform at their best, if for instance you shoot a landscape with a 16mm on APS-c and you need infinity in focus and 1/125 to freeze leafs from blur and the rest is f5.6 at 100ISO, a photographer doesn't use his FF with the equivalent 24mm at f11 because at 1/30 the leaves will blur nor he drives the lens towards diffraction at f11, neither he pushes ISO at 400 on daylight!!! He simply uses the same values (1/125, ISO100, f5.6) and refocuses his lens to include infinity in the frame!!!!!!!!!!!!!!!! He will only loose a meter (yard) or so from DOF in front of the camera where he has nothing, on the other hand how will you take with an APS-c the portrait that with FF is 70-200mm at 100mm, f2.8, iso100? Are you going to invent 65mm f1.4 and IS025? NO! Of course not! You move the lens to 80mm and move a little backwards to compensate for DOF by focusing at the nose than the eye and keep the bockeh! Anything else is new methodology that proves that the world is now full of "Ainsteins" and "Gurus" that have nothing to do with photography but are trying to find a life to live on web by attracting some more ignorants to new theories. ;) Cheers Theodoros www.fotometria.gr
-
NO! It applies to base ISO as well, although you wouldn't mind the difference there!
Ah! Slightly different from your remark in your previous post which was:
If you print a large size print from my D7000 and then print THE SAME SIZE from my D700, you will find that the FF is CONSIDERABLY BETTER. Interestingly, this is not so evident if you do the comparison on a monitor,
Nevertheless, whilst I would NOT think it odd that there would be very minor and insignificant differences between the D700 and D7k at base ISO, I do find it odd that such insignificant differences would be more apparent on a print than on the monitor.
My experience is the opposite. I find that subtle differences which are apparent on the monitor, such as slightly greater noise in a particular part of the tonal range than the other image exhibits, tend to get lost in the conversion to print. I presume this is because the degree of enlargement possible on the monitor allows one to pixel peep in great clarity every defect however minor.
As regards some of your other points, you appear to be under a few misapprehensions. (I hope you don't mind my saying that. ;) )
DXO measures sensors converted with their own RAW converter, not with the latest NX2.(...)
This is simply not true. However, it is true that DXO produce their own RAW converter, presumably based upon the measurements they make from the RAW data from various sensors.
Here's what DXO have to say on the issue.
We deliberately chose to perform all DxOMark measurements on RAW images in order to evaluate the intrinsic image quality of lenses and camera sensors.......
Only RAW-based measurements report on the image quality of the photographic hardware irrespective of the RAW converter. Evaluating RAW converters is a significant task in itself that we currently do not address on DxOMark.
In other words, in case there is still confusion, the measurements are based on the RAW data. When the RAW data is converted to a viewable image, it is no longer RAW.
The method of comparing FF and APS-c performance by equalizing DOF is ridiculous, photographers don't do that to take a picture! They put their cameras to perform at their best, if for instance you shoot a landscape with a 16mm on APS-c and you need infinity in focus and 1/125 to freeze leafs from blur and the rest is f5.6 at 100ISO, a photographer doesn't use his FF with the equivalent 24mm at f11 because at 1/30 the leaves will blur nor he drives the lens towards diffraction at f11, neither he pushes ISO at 400 on daylight!!! He simply uses the same values (1/125, ISO100, f5.6) and refocuses his lens to include infinity in the frame!!!!!!!!!!!!!!!!
Not this photographer, but each to his own. Shutter speed, DoF, and ISO are parameters I'm frequently juggling to get the best effect.
The differences in the DoF between FF and the cropped format, at any specific F/stop and shutter speed, can be seen as either an advantage or disadvantage between the two formats.
At shallow DoFs, the larger format has the resolution advantage because the lens at the 'less narrow' aperture is likely to be sharper. (For example, F4 with FF compared with F2.5 with the cropped format.)
However, when a large or extensive DoF is desired, the larger format tends to lose that resolution advantage because we may be comparing lens resolution at F5.6 with F9, or F8 with F12 etc. But notice I wrote tends. There are other factors which affect resolution.
I have find the D700 to be about half a stop better than the D7k in highlight DR (very important) and about a stop better in shadows (not so important, but welcome) for my noise standards in the shadows, both of them are more than a stop behind the Fuji S5pro in the highlights (the later with Hyper Utility and nothing else as a converter of course), in fact the fuji is the only DSLR that can compete (not beat) with a MF-DB in DR.
Sorry! This doesn't make sense to me at all. Both the D7k and the Pentax K5 will outperform many MFDBs in Dynamic Range, but not in terms of noise at 18% grey.
For example, the Nikon D7k has almost one stop greater DR than the P65+ (0.9 EV to be exact). However (and it's a big 'however'), at 18% grey, roughly the luminosity of skin tones, the P65+ has a very significant 7.9dB better SNR, and that would definitely be very noticeable on screen and print.
I'm not sure what you mean by 'highlight dynamic range', but I'm assuming you are referring to SNR in the highlights.
In the attached image from DXOMark you will see the full SNR plots for the D700 and D7k, covering the entire tonal range from highlights to deepest shadows.
At base ISO you should observe that the D700 has a marginal advantage in SNR in parts of the midtones. At 10% grey it's approximately 1.5dB better than the D7k (bearing in mind that 10% grey is a lot lighter than you might think because this is a log scale.)
However, I should add, in case there is some confusion about the significance of 1.5dB in SNR, this could be equivalent to about half an f/stop or EV in the midtones, but not in the highlights, nor in the shadows where such differences would go unnoticed, as I understand. But correct me if I'm wrong.
-
As i clearly stated A PHOTO IS ONLY THE PRINT, nothing else! DXO does RAW conversion, that's why they supply 2 results "screen and print". There is no comparison that can be made on a monitor, monitors and graphic cards have DR and colour accuracy differences. When you compare prints, it must be done in 16bit tiffs! When I compare prints, I open the RAW files on a fully calibrated EIZO GM242, and print on an equally calibrated and linearized EPSON 9900, controlled by a full power, special order, MacPro. Is that good enough to now what I'm talking about? ;) My major income is from doing copies of Byzantine Icons that I shoot in microstep (16x), through my contax 645 or Sinar P2 on my Imacon 528c tethered to my macbook (again special order, full power and fully -as possible- calibrated), please have a look at the latest images under "studio" in my page (www.fotometria.gr). D7K and Pentax K5 can beat what.....? ;D A DB.....? ;D According to who? Is it DXO again....? ;) I can assure you they are no where near a DB, especially in highlight DR. The latitude in highlights of any DB (new or old) is as much from the D700, as the D700 from any APS-c (S5pro excluded, but again this has no latitude, its a natural champ)! I guess that "latitude" is the key word here, because full DR appreciation, depends on anybody's allowance/acceptance of shadow noise, which is much more evident than highlights! Cheers, Theodoros.
-
I repeat, DXO test-results for sensors are based on an analysis of the unconverted RAW data. To quote again from their website:
DxOMark measurements are performed on unprocessed RAW files (the only image format that reflects the intrinsic performance of digital cameras and lenses)
D7K and Pentax K5 can beat what.....? ;D A DB.....? ;D According to who? Is it DXO again....? ;) I can assure you they are no where near a DB, especially in highlight DR. The latitude in highlights of any DB (new or old) is as much from the D700, as the D700 from any APS-c (S5pro excluded, but again this has no latitude, its a natural champ)! I guess that "latitude" is the key word here, because full DR appreciation, depends on anybody's allowance/acceptance of shadow noise, which is much more evident than highlights! Cheers, Theodoros.
Could you please explain what you mean by highlight DR in this digital age? You sound as though you are comparing color negative film with slide film.
Dynamic Range, as I understand it, is the ratio between the brightest highlight the camera can capture, without blowing detail, and the darkest shadow it can simultaneously capture without clipping detail. Most DBs seem to have a DR of around 12 stops. This little toy, the Pentax K5, has about 14 stops of DR.
If you were attempting to capture a scene of high dynamic range (or high subject-brightness-range) of 12 stops or more, and you were not able to bracket exposure for merging to HDR because of subject movement, leaves or waves etc, and you did not have studio lighting or flash to lighten the shadows containing a Byzantine Icon poking its head out of the undergrowth, you might get a better shot using a K5 rather than a P65, provided of course you also wanted to retain full detail in those magnificent fluffy clouds overhead. ;D
-
especially in highlight DR.
Apply -1 stop of EC, and you will gain 1 stop of highlight DR (and loose one stop of lowlight DR). Atleast that is my understanding.
When comparing sensor specs of different cameras, manufacturer policy wrgt auto exposure is of little relevance. Therefore, it is usually the entire DR that is measured/debated.
-h
-
I repeat, DXO test-results for sensors are based on an analysis of the unconverted RAW data. To quote again from their website:
Could you please explain what you mean by highlight DR in this digital age? You sound as though you are comparing color negative film with slide film.
Dynamic Range, as I understand it, is the ratio between the brightest highlight the camera can capture, without blowing detail, and the darkest shadow it can simultaneously capture without clipping detail. Most DBs seem to have a DR of around 12 stops. This little toy, the Pentax K5, has about 14 stops of DR.
If you were attempting to capture a scene of high dynamic range (or high subject-brightness-range) of 12 stops or more, and you were not able to bracket exposure for merging to HDR because of subject movement, leaves or waves etc, and you did not have studio lighting or flash to lighten the shadows containing a Byzantine Icon poking its head out of the undergrowth, you might get a better shot using a K5 rather than a P65, provided of course you also wanted to retain full detail in those magnificent fluffy clouds overhead. ;D
1. It has been answered already, sorry you don't understand.... :'( 2. I suggest you do it your way (with DSLR & studio flash ???), its really good money in art reproduction and a few people that can do it!! :) Cheers, Theodoros.
P.S. Among others, I think (correct me if I'm wrong, please don't if I'm not) you totally confused "microstep" 16x (or "multishot 4x), with HDR shooting, where did this HDR came in the conversation?
-
Apply -1 stop of EC, and you will gain 1 stop of highlight DR (and loose one stop of lowlight DR). Atleast that is my understanding.
When comparing sensor specs of different cameras, manufacturer policy wrgt auto exposure is of little relevance. Therefore, it is usually the entire DR that is measured/debated.
-h
1.Your understanding is wrong, -1 stop of exposure, will give you -1 stop ONLY IN THE LINEAR PART OF the S-slope, not in highlights or shadows where the info is compressed! ;) 2. Yeah! The problem though is how they measure it and if its usable! ;) Cheers, Theodoros www.fotometria.gr
-
1.Your understanding is wrong, -1 stop of exposure, will give you -1 stop ONLY IN THE LINEAR PART OF the S-slope, not in highlights or shadows where the info is compressed!
I politely suggest that you discuss the topic instead of declearing that everyone else is wrong all of the time and using exclamation marks in every sentence.
Where is there an s-curve in your camera? I am aware of no such thing in mine, it is highly linear from clipping and into the noise-floor.
-h
-
I remember the time they stated that a Full-frame version of the Leica M-rangefinder was an impossibility and that at best it will need to be a 1.3x.....well, we have had a FF Leica-MR for a while now. As long as the mount is large enough (and in fact the Sony E-mount is larger than the Full-frame Leica M-mount) a FF will arrive. My prediction.
Could be, I don't think so for the near future though and certainly not with an electronic finder, I don't thing that there are many photographers that will excuse that and thus justify enough market share! But it can be done with an extra finder that can carrie focus and FOV issues through the dedicated hot shoe if it has a build in motor..... 8) I guess they may be looking at it...., you never know... :) Cheers, Theodoros www.fotometria.gr
-
I politely suggest that you discuss the topic instead of declearing that everyone else is wrong all of the time and using exclamation marks in every sentence.
Where is there an s-curve in your camera? I am aware of no such thing in mine, it is highly linear from clipping and into the noise-floor.
-h
Sorry for my use of grammar!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ;) I suggest you take some photo lessons, you obviously don't need grammar ones!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ??? If your camera's data is not recorded with compressed highlights or shadows, I suggest you change your camera!!!!!!!!!!!!!!!!!!!!!!! ::) But I'm not aware of such a camera!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! :-X Cheers, Theodoros www.fotometria.gr
P.S. Thanks for "politely" it really gave your quote a touch of class. S-slope is not in your camera, it's a fundamental in photography and it tries to equalize the compressed highlights and shadows that your (and my) brain records. If your camera recording was linear, you wouldn't be able to see more than 4-5 (at best) stops of DR! Data recording with compressed highlights and shadows is not new to digital photography, it's with us from the beginning of film, even earlier than that!!!!!! :o
-
I suggest you take some photo lessons
I suggest we keep personal insults out of the discussion.
If your camera's data is not recorded with compressed highlights or shadows, I suggest you change your camera
My 7D is working very fine. If it had a non-linear response, doing HDR bracketing would be very difficult. Luckily, I can report that it works...
S-slope is not in your camera, it's a fundamental in photography and it tries to equalize the compressed highlights and shadows that your (and my) brain records. If your camera recording was linear, you wouldn't be able to see more than 4-5 (at best) stops of DR! Data recording with compressed highlights and shadows is not new to digital photography, it's with us from the beginning of film, even earlier than that!!!!!! :o
Do you use raw files when shooting? I suggest that when doing anything exposure-related in post-processing, you do it before applying non-linear curves. That is how it should be implemented in most raw developers.
Of course, when you adjust the exposure in-camera, that will be linear because the camera sensor knows of no such thing as s-curves, gamma or tone-mapping.
1.Your understanding is wrong, -1 stop of exposure, will give you -1 stop ONLY IN THE LINEAR PART OF the S-slope, not in highlights or shadows where the info is compressed!
To summarize: your statement is bogous.
The clipping-point and noise-floor of your camera sensor provides the basic lilmits for Dynamic Range, and the sensor is linear. Subjective and practical constraints cause your signal to be placed somewhere along that range, and anything you loose on the top, you will gain at the bottom, DR-wise.
-h