blue_skies: I just read Rishi's comment and commented back to him:
Really? The A6000 outdoes the A7ii????
(continued) The trigger points could be ISO 1600 and 6400 for a FF sensor.
For the low region, I would expect sensor size & resolution to play no role, for the middle region, I would expect a noticeable role, and for the high region, I would expect sensor age, size & resolution to come into play.
With different sensor sizes, equivalent exposure and try ISO ratings begin to matter, and this is where DxO gives meaningful charts.
The relative comments from dpreview should align with such graphs. As written, they are confusing, see the discussion here and in other threads.
I think your comments are well intended, in their context, but the context is not made very clear. If it was, there wouldn't be a lot of references to DxO and user's own experiences, imho.
Also, if differences are small, call them small. Making a relative statement wrt/ a puny difference is not saying much about the product differences at all.
As written, the review does not align with DxO, nor my own, data
(continued) The high performance ISO plays a secondary role for me. Highlight clipping and protection, lens and sensor flare, microcontrast (after post-edit) are all big factors.
If I go past ISO 6400, I also tend to use the image for smaller size purposes, such as Web, half-size print, or poster collages. This gives me an extra layer of noise control without getting the dreaded watercolor effects.
Your findings tend to focus on the studio scene at high ISO using an ACR Raw flow. Although this gives me interesting info, it does not help me for my purposes, as I try to explain here.
If anything, newer sensors (eg. BSI, smaller process node) will keep continue to improve on older sensors, no surprise here. If this is visible at very high ISO, then great. But I would still not choose the A7s, as this is for ISO when details begin to disappear, and images loose their punch.
Therefore, I suggest perhaps to break the ISO into three sections: low, medium and high.
@ Rishi, low light shooting can have many different forms.
ISO can be based on spot metering, average metering, or highlight protection. One of the reasons to choose FF cameras is because of the greater dynamic range.
The resulting image is sensitive to how much of the noise remains in your subject: in a night scape, the entire scene is the subject, but a performer on stage is only a part of the image. I would automatically find a way to use a lower ISO (fast lens, tripod, longer SS) for the larger (edge to edge) subject, and would look for ways to illuminate a smaller subject (stage light, bounce flash, ambient light, etc.)
Then, the Raw work flow would reflect what I am trying to achieve in terms of detail and noise control. I use graduated filter to listen or darken part of the image if needed, thereby increasing or decreasing the need for good high ISO performance in shadow areas.
1/3EV is not a meaningful difference to me: the subject matter plays a much bigger factor.
I agree, but dpreview's findings contradict your statement.
To repeat Rishi's comment from above: "NX1 > D5300 > a7 II > a6000"
@ Rishi, RE "Please don't put false words in our mouths. We never said the a6000 outdoes the a7 II. The a6000 has worse ISO performance than the D5300/D5500 in our studio tests."
Referring to your comments in http://www.dpreview.com/forums/post/55575295
Rishi: "The a7 II, on the other hand, shows ISO performance similar to, or perhaps slightly worse than, that of top APS-C competitors (and by top, I also mean dirt cheap D5300/D5500s...)."
In other words, the D5300 outdoes the A7 II.
Then, in http://www.dpreview.com/reviews/sony-alpha-a6000/12
"... In low light, the a6000's Raw performance is a fraction behind that of the Nikon D5300..."
This basically says that the D5300 and A6000 are within a hair of one another, and, coupled with your first statement, this implies that the A7 II is slightly worse - yes - than the A6000.
@ Rishi, RE NX1 > D5300 > a7 II > a6000
please take a look at http://www.dpreview.com/forums/post/55580989
I, too, suspect something in ACR, hence I was perusing the JPG images at low light & high ISO. But, see my post, I would definitely argue that:a7 II > D5500 > D5300 > NX1
Wouldn't you agree?
Perhaps this is a judgement for ACR, rather than for the camera/sensor itself? I would still be curious as to what the sensitivity is that triggers such a difference in the results (from RAW).
Also, do you have access to e.g. DxO PRIME?
cheetah43: Marking: 82%. OK, fine. What improvements would take the mark beyond 90%? There seems to me a great deal of confusion about marking standards and consistency. In DPR's view how could a camera attain 95% in its class? It would be interesting to know. if DPR says, "It is impossible; the matter is subjective" then marking loses its sense.
Not sure where you are going with this.
In handling the two cameras (A7 and A7ii) I got a very close appreciation of just how many changes Sony made in the A7ii. I can only assume how much feedback they must have been going through and that all the changes were made to make things better, per the collective feedback.
The fact that a mirrorless camera has to look and feel like a DSLR camera is a predictable outcome in the end :)
As to the scores: as long as you use a 0-100% scale you are always in a bind if you want to rate new entries. Dpreview knocks new entries down because technology advances. This makes it very hard to understand relative ratings within a class, as an older entry gets higher scores than a newer, better entry might get, and when you look across classes it gets really confusing.
I like DxO's approach with an absolute scale. As technology improves, you simply move up on the scale. And all old data remains valid, and new improvements can be seen immediately.
You didn't open the link.
This was specifically about Rishi's comment that class-leading (D3300/D5500) APS-C sensors outdo the A7ii's FF sensor, and that this thus implies that the A6000==D5500 runs toe to toe with the A7ii and even outdoes it.
As an owner of both the A6000 and A7ii I find this odd. For daytime perhaps, but not for nighttime. My findings agree with DxO, essentially the FF sensor is a full stop faster.
My findings however are not based on OOC JPG or default RAW flows, I use DxO and LR for high ISO flows, and I don't use the defaults. In fact, using my method, the test-charts that are used in the studio setup end up looking a lot sharper, cleaner, more vivid and more saturated, with deeper blacks than the posted images.
Also, the A7/ii RAW images have simply more headroom than the A6000, and I presume the D5500, per the DxO stats.
I wasn't commenting about AF, battery life, handling, ergonomics, etc.
ikenkoin: I couldn't find in the review a reference to the sensor reflections that caused some discussions with the original A7.
Have I overlooked it ?
I was curious, if an improvement has been noticed by objective testing.
I have had my A7 II only for a few days and so far I can tell that the ergonomics are okay for my small hands and the shutter sound is awesome.More subdued than from any Nex I have owned before.
I can confirm the sensor reflection issue: I tested both cameras side by side and the A7ii is noticably better than the A7 in this regard.
It is not a major issue, imho, but with certain lenses and certain (outdoor) lighting situations, the A7 shows halos and glare that are absent in the A7ii. Even if you push the image, the A7ii remains noticably clean.
For me, this was the major reason to upgrade the camera, not the IBIS.
I just read Rishi's comment and commented back to him:
Peiasdf: Call me cynical but I suspect the reason this camera isn't as good as it can be giving SONY's track record of always going for the best is to leave space for the A9 coming out later this year. If A7II have lossless RAW, no OLPF, A6000 level AF and so on who will pay much more for A9 when the only difference is 36mp?
A7II is a nice improvement over A7 in most areas and it is going for the same price. SONY can do better but clearly A7II is not the flagship of the lineup.
(continued) your observations apply to the default results, and may not be typical for user experiences.
Conclusions therefore also miss some of the obvious.
Would I chose the A7/ii for ISO 6400 shooting? Not if I can avoid it.
Would the D750 be better for that? Perhaps, but again, I would rather avoid it.
You point out wisely that underexposing + punch in post has other benefits, and this is of interest between the cameras.
But, per my knowledge, the D750 and A7 share a 'very similar' sensor, and the D5300 and A6000 also share a 'very similar' sensor.
Test methods that show very different results should address the understanding as to why this is so. If it is just defaults, then it has little meaning for post-edit results. If it is because of sensor calibration, high/low ISO bias, etc., it is good to know what is gained AND what is lost, as it comes down to trade-offs.
The A7/ii both work for me for what they do, based on my personal experience. Plus, I prefer live EVF
Rishi, the report reads as if the A7ii is a regress from the A7.
My comments points to my experience that the A7ii is rather identical to the A7 in terms of IQ and ISO performance.
DxO suggests that the A7ii is improved.
In my own observation, the studio images posted show a very vanilla (defaults?) result.
I tend to look at images post post-edit flows. E.g., if I process your studio scene, from RAW, I tend to add more contrast and sharpness than you have, per your JPG, or JPG-from-RAW images.
I would argue that my images are 'better', i.e. more saturated blacks, more punchy colors, more sharp edges. The difference is not severe, but rather subtle, yet very noticable. (I use LR and DxO).
I also have a feeling that, when applying my same method to other cameras/images, the results will become closer, not further apart.
In post-edit, noise levels, sharpness and saturation can be controlled, and, per your comment above and in the review, I am led to believe that (continued)
blue_skies: I don't like commenting here, but I do want to point out, per the many other threads about this, - that the FW used was 1.10, and not 1.20 (which does fix a number of things),- that RAW capture was done in 'fast' mode, i.e. fewer bits, either due to bracketing, or continuous shooting mode, and - that it appears that IBIS was left on (per the EXIF) while using a tripod.
If any or all true, the captured examples are not representative.
A simple question: same sensor in A7 and A7ii? Perhaps, but the results don't seem to indicate this. The sensor topping is different, as sensor reflections are way down (I have both A7 and A7ii).
Also, the video wrt/ focus tracking being poor seems to be a user setting issue. I can make it behave like that and make it go away, by simply changing the settings and/or control the camera movements.
I don't think that this is a bad review, but the camera was not handled by an experienced Sony user - I see a number of 'operator error' mistakes.
Rishi, first, I am not criticizing the review, it is a thorough review. Rather, I am criticizing some of the findings. E.g. the comparator tool still shows FW v1.0 on A7 versus v1.1 on A7ii. I am using v1.2 on both, and have different observations from you.
In fact, the JPG engines are closer, much closer than your comparison tool suggests, and the RAW data seems identical to me. Your conclusion: the A7 has better JPG and RAW (comparo-meter) just doesn't add up for me.
As to the AF hunting: there is the AF-Lock method in AF-C, which has different sensitivity levels, there are in fact two positions which each 3 levels (L,M,H) for the spot focus.
When I try to recreate your scene, I not only do not see the hunting, when I move outside the image view, and then return, the subject snaps right back into focus. Are you half-pressing the shutter at all in your experiment?
Yes, the A6000 is still better, but the A7ii is not 'unpredictable'. Tripod and IBIS/OSS should not be combined.
PeaceKeeper: I just can't get over how absolutely huge those lenses are. That 55/1.8 is ridonkulous.
I'd own one of these were it not for the complete lack of effort on Sony's part to make lenses that match the mirrorless systems strengths. It's as though all they did was recycle SLR lens designs and add a spacer to the back.
I think that you are looking at the wrong lens.
The 55mm lens requires, well, a 55mm distance from the sensor, and the lens is pretty much like that. You can call it a recycled SLR lens, but it really is a tad better than that.
Heavy? 1.2lbs? I really think that you are mixing your facts, this lens is very light, sorry.
Oh, and the AF works rather well. Why would it not?
I don't think that Sony did anything. In fact, DxO rates the sensor/processor/FW combination higher for low light than the A7.
I think that the results are what they are because of simple operator error mistakes, which indicates a true flaw of the camera, namely it being sensitive to user settings.
Likely not a big deal in real life, but in a controlled documented studio setup, big conclusions are derived from small mistakes.
I have both A7 and 7ii, and I don't see what dpreview sees, sorry.
As to the differences, the A7r has no AA filter, no PDAF grid (CDAF only), a higher resolution, and a more advanced layout technology (per their earlier marketing). This has always shown in the benchmarks. There is no need to cripple the A7ii, just let it be (24Mp, AA filter, etc).
An A9 with superfast AF and 50Mp? Fingers crossed, I can see the increased resolution, but not the superfast AF this year. Similar perhaps.
But yes, the A7/ii is not the flagship, but it is a solid compromise.
I don't like commenting here, but I do want to point out, per the many other threads about this, - that the FW used was 1.10, and not 1.20 (which does fix a number of things),- that RAW capture was done in 'fast' mode, i.e. fewer bits, either due to bracketing, or continuous shooting mode, and - that it appears that IBIS was left on (per the EXIF) while using a tripod.
Sannaborjeson: Dpreview forgot to mention it's shutter sound.I own a7r and the shutter sound is just horrible. Can't believe Sony engineers managed to make it that loud without using gunpowder.
The Sony RX1 has a fixed lens and uses a leaf shutter design - actually a shutter that is built into the lens.
The ILC cameras use a focal-plane shutter, and the noise level is proportional to the mass that has to move. The bigger the sensor, the louder the noise. It is mechanical.
Making it quieter means making it slower - how about 1/1,000th, rather than 1/8,000th?
The EFC means that the first curtain does not get activated - only the second curtain has to move and this reduces the sound to half (one, rather than two, shutter movements).
Other than a shutter, there is no instantaneous way yet to stop the exposure and begin the readout of the pixels.
Perhaps an inverse sound can be created to mitigate the sound of the shutter? Or advanced dampening techniques may be discovered at some point?
Bigger body cameras have the same internal sound, but baffling and the sensor sitting deeper lower the sound perception.
Personally, I don't find the sound THAT loud, just louder ... :)
UnChatNoir: I didn't had my hands on an A7 or A7r, but I can imagine DPReview tells a few things that the fan crowd doesn't necessarily likes. The same happened for a long time with the Fuji X-series. Owning both a pro-DSLR and a X-Pro1, I can say, well, I like the MLIC lightweight and even IQ, but regarding reliability, speed and flexibility, nothing still beats a pro-line DSLR. However, having spend the money to MLIC, quite some people have problems to admit this, even more: they start believing the opposite. A DSLR is crap, outdated, for those 'not seeing the light'. Well, I'm one of those that returned from a MLIC to a full blown DSLR. And I must admit, everything falls back into its place. 55 years of Lens experience, 10 years of new models & firmware development have ironed out all the possible glitches while with a MLIC, every mission is a new challenge. And from what I read about the A7 and A7r, it's pretty much the same as with all the other MLIC: first gen stuff for the adventurers.
I believe that you are 100% correct, and also that your comment is also completely besides the main issue here, imho.
MILC cameras have (lots of) quirks, all of them. And Gen.1 cameras suffer from (designer’s) quirks – usually some decent customer feedback improves on things.
I see the main reason for the backlash of comments being that the vast majority of A7 (now experienced) users simply do not notice, experience or agree with DPreview's findings.
Either DPreview make a mountain out of a molehill, or they did not appreciate the camera in its own context, i.e. the way that users are actually using the camera. Either way, it makes you question the reviewer, for being prejudiced, inexperienced, biased, ignorant, new at this, or inconsistent.
If the review had been very much in line with other reviews, people would not react as strongly. But as written, the review does not align very well with past reviews, other than for the spec/details sections.
And that is noticeable.
HFLM: Another thread discussing it:http://www.dpreview.com/forums/thread/3599612#forum-post-52809694Forget the conspiracy against Sony. Even if it were a minor issue (and probably it is) it was s.th. the reviewers didn't find in the D610, for example, for similar testing conditions. Together with the minor quibbles and lack of system behind it a silver award is great. If you enjoy it continue to enjoy it.
@HFLM: "Nevertheless it's shocking, how close m43 is if you don't print large and use good glass. ... ?"
I partially disagree - in the Studio Comparison tool you can clearly see m43 (EM1), APS-C (Nex-6) - both at 16Mp - and the A7, all showing different results, especially at higher ISOs.
But I agree that, if you don’t print large, a smaller sensor camera may suffice. Why stop at m43? If your only output is HD (tablets, monitors, TV), even a much smaller sensor cameras will do just fine, at low ISO. They have the resolution and can produce impressive images. Consider the RX100-II? Or the RX10?
Not very shocking to me :)
Still, I believe in sensor size when it comes to low light (high ISO), cropping (resolution), lens choices (FF is much easier on older lenses), (shallow) DOF control, creative photography (shift, fisheye, wide angle), DR and color rendering. This where the A7 and your D610, as full FF cameras, deliver, and where smaller sensors cameras simply cannot keep up.
It is funny how DPreview's own interpretation of the Studio Comparison tool does not line up with the conclusion in the review: