Sony A77

[das_schlechte_gewissen]

Re: Sony A77 - It is working for me.

With all the expected raging enthusiasm of early adopters, it is a bit difficult to sort out the true functionality and ability of this camera

The notion that "who cares about high ISO", and "most pros and semi-pros shoot at 100 or 200 anyway", is something of red flag, since most anything on the market does really well at low ISOs. Sounds suspiciously like it isn't quite advancing the cause, and meeting expectations.

I get the appeal of a smaller camera, etc. etc. ... that interest me also, if for no other reason than as second camera to the A900. But it has to do something the A900 can't. In the opinion of those exploring this camera in hand, what might that be?

I'm not talking about video, or circus tricks to entertain the bored ... real world stuff. Does it have less lag? Shoot at relatively higher ISOs with exceptional IQ (so more available light can be done with lenses like the ZA24/2), does it focus faster in low light, and is it as accurate as the A900? (one of the A900's strengths over camera's like the Canon 5D-II). How's the out -of-camera color at ISO 400, 800 and 1,000 compared to the A900? Can it actually produce a usable image at ISO 3200/6400? ... yes, as a Pro, I would use 3200/6400, and I have pushed my A900 to do so in certain situations.

Not trying to be a wet blanket here, and no offense to buyers, just asking what this pup can do?

-Marc

I expect that the pros elaborate such questions and give the answers to us amateurs. :sleep006:

 

[LizaWitz]

One of the problems that I see all the time with ISO is comparisons of higher megapixel cameras shooting at higher effective ISO than lower megapixel cameras because the people making the comparison do not understand what ISO is. Even though the "ISO" setting on the camera in both cases may be 6400, if you don't account for pixel density you're not comparing the same ISO.

For instance, I'd be pretty surprised if you shot two full frame images with the same lens, one on a Canon 5D mk II and one on the A77, at the same ISO rating, and then reduced the images down to a 1920x1080 resolution, that you'd be able to see any difference.

The 5D has the advantage of a larger sensor, so it is actually getting more light (at the same aperture and lens) than the A77, but it is also an older sensor, and so the advance in technology for the A77 sensor will likely make up for the difference, even though at the "same ISO setting" on the camera the A77 is getting less light on the sensor.

The issue is even more dramatic when you compare a 24Megapixel image to a 14Megapixel image and don't account for the difference in the number of pixels. ITs easy to do, though, just shoot the same image, and reduce the full frame down to something manageable and look at the images.

If you're comparing "%100 crops" of two sensors with different pixel densities, to do an ISO comparison, you're not comparing the same ISOs.

This may sound like a theoretical quibble, but it is anything but. When you shoot, you shoot at full resolution, but the final image is always at the delivered resolution. For the same size sensor, and the same level of sensor quality (e.g.: made the same year), the higher megapixel sensor is going to give you better ISO performance across the whole frame at the final resolution, than the lower megapixel sensor. (Say comparing a 5N or an NEX-5 to the output of an A77) However, if you do a "%100 crop" the higher megapixel image will look worse because you're measuring the resolving power of a smaller part of the sensor, and wasting noise canceling effect if dithering a higher megapixel image down to the working resolution.

 

[Mark K]

My A77 was locked up after around 1300 shots....very annoying because I　was in the middle of my life time diving trip in Bali...The only other camera was a 5D with 17-40L. I dropped all other Canon glass and flash gun for luggage weight.
I am quite annoyed for this accident. A77 should be able to be unlocked by hard resetting the camera.

 

[jonoslack]

My A77 was locked up after around 1300 shots....very annoying because I　was in the middle of my life time diving trip in Bali...The only other camera was a 5D with 17-40L. I dropped all other Canon glass and flash gun for luggage weight.
I am quite annoyed for this accident. A77 should be able to be unlocked by hard resetting the camera.

HI Mark
Is it still locked up? Did replacing the battery cure the problem?
I have to say, I don't think I've ever had a camera which didn't lock up at some time or other, but it isn't okay if removing the battery doesn't sort it out!

 

[Jim DE]

jono I have read about s process that is being used to reboot a frozen a77..... involves removing the battery and then putting it back on while depressing 3 buttons at the same time while turning the camera on. Seems to work 100% of the time from what I am reading.

Good thing this process doesn't involve doing the chicken dance at the same time as I am a bit old for that kind of activity

 

[Goldencode]

Jono, how is the A77 during panning with the EVF ? Thanks !

 

[philip_pj]

fotografz, not many impartial users of both systems believe Canon's 5DII produces colour results that get anywhere near that of the A900. Not just users, but also several in the raw converter developer community. Dense CFAs etc.

Canon almost certainly compromised low ISO colour performance for high ISO noise performance. Interesting that the market did not care about this move after the highly regarded 'classic' 5D, colour perception and tonality is of course highly personal...nor apparently that high ISO also compromises so many other aspects of image quality.

That's marketing for you! As Marshall McLuhan once opined: 'we are suckers for a hard sell, it's the Achilles Heel of literate man.' He lived in pre-PC speech times.

I have the impression that Sony's designers are behaving as though they feel that the high ISO 'war' is quite silly when taken to extremes; I say this because the SLT designs deliberately give away a quantum of high ISO performance, yet the company fits them to all the latest crop sensor DSLRs. They appear to believe medium-high ISO performance is good enough. Sony is widely regarded as the premier sensor producer, of course.

To your question: geotagging, ground-breaking video, 24Mp on APS-C, 50 ISO, EVF confirmation of exposure and other settings for non-chipped MF lenses (improved pre-shot decision-making), and excellent manual focus assist in low light. Perhaps not your real world, but most definitely ours. Not keen on losing CF though.

LizaWitz, the issue is still befuddling so many users, sad after the accumulation of all these digital years, and the new high res monitors don't help, for the 100% view obsessives ;-)

 

[LizaWitz]

Part of the reason its befuddling is that there are too many variables. I even flubbed my explanation of it, I see now.

The key error is thinking that "%100 crops" are covering the same amount of the sensor, when a higher pixel density sensor will have more pixels per inch, and thus a given number of pixels for your crop represents a smaller part of the sensor (and thus a higher effective ISO.)

Do full frame comparisons at high ISOs, and reduce the images down to HD (or lower) resolution.... if you can't tell the difference in your final output resolution, is there really a difference?

 

[dhsimmonds]

A real life sample of A77 image taken with the CZ 24mm F2 lens at ISO 3200 converted from RAW file with ACR here: http://www.photoclubalpha.com/forum/viewtopic.php?f=32&t=5786

Also some interesting comments made by the working freelance pro photographer. Scroll down to see the full RAW data.

 

[fotografz]

Re: Sony A77 - It is working for me.

I expect that the pros elaborate such questions and give the answers to us amateurs. :sleep006:

Not really. If you can't answer the questions, that's fine, but please don't question the motivations of others asking qusetions. I don't have the camera and can't get my hands on one, so the only thing to do is ask folks I tend to trust that are using the camera ... Jono being one of them, among others.

Marc

 

[fotografz]

fotografz, not many impartial users of both systems believe Canon's 5DII produces colour results that get anywhere near that of the A900. Not just users, but also several in the raw converter developer community. Dense CFAs etc.

Canon almost certainly compromised low ISO colour performance for high ISO noise performance. Interesting that the market did not care about this move after the highly regarded 'classic' 5D, colour perception and tonality is of course highly personal...nor apparently that high ISO also compromises so many other aspects of image quality.

That's marketing for you! As Marshall McLuhan once opined: 'we are suckers for a hard sell, it's the Achilles Heel of literate man.' He lived in pre-PC speech times.

I have the impression that Sony's designers are behaving as though they feel that the high ISO 'war' is quite silly when taken to extremes; I say this because the SLT designs deliberately give away a quantum of high ISO performance, yet the company fits them to all the latest crop sensor DSLRs. They appear to believe medium-high ISO performance is good enough. Sony is widely regarded as the premier sensor producer, of course.

To your question: geotagging, ground-breaking video, 24Mp on APS-C, 50 ISO, EVF confirmation of exposure and other settings for non-chipped MF lenses (improved pre-shot decision-making), and excellent manual focus assist in low light. Perhaps not your real world, but most definitely ours. Not keen on losing CF though.

LizaWitz, the issue is still befuddling so many users, sad after the accumulation of all these digital years, and the new high res monitors don't help, for the 100% view obsessives ;-)

Thanks for the intelligent and well thought out answers Philip.

Yes, how one applies a DSLR is of course critical to the analysis ... so "real world" varies from one photographer to another. However, I would say that DSLRs do fall into a certain "parameter of performance" where expectations have been established to some common degree based on application.

I would hazard a guess that speed of operation, reasonable higher ISO performance, quick/accurate AF and of course reasonable levels of resolution to some degree or another, are chief amongst those expectations of DSLRs. Weaken or remove any one of those, and there are other alternative solutions to consider ... one of which is to keep one's current camera.

For my applications, the A900 has met those expectations quite well, save perhaps one slightly weak area: ISOs beyond 1000. If an A77 or A99 can meet them equally well (or better) plus add other less needed functional attributes, then it retains my whole-hearted interest.

Thus the specific basic questions beyond all the new attributes this camera adds (which admittedly may be of premiere value to others).

Does this camera operationally work as fast as the A900? For example, did Sony retain the highly useful one camera button Speed-light compensation available as a Custom Function on the A900? This is the best one on the market, and of high value to wedding and event shooters, to name but one large group of users.

Is the AF at least as fast as the A900, and more importantly, is it as accurate? Related to that, can various lenses be micro-adjusted and registered in the camera?

Resolution is less of a concern ... the A900 is plenty, and it seems this camera doesn't fall short. Focus peaking with MF lenses is a nice touch, but not a priority since that is more of a speciality application beyond the general scope of DSLR expectations.

Admittedly, the primary interest I personally have for anything beyond the A900 is ISO performance. Available light shooting with some of the faster aperture lenses like the 24/2, 50/1.4, 85/1.4 and 135/1.8, has only two real purposes ... shallow DOF and/or extended low available light use. Since the lenses aren't going to get any faster, ISO 1600, 3200, and to a lesser degree 6400 is the domain to be explored while retaining enough IQ to produce usable print quality (read: salable prints in my case). To date, Sony has not produced a competitive camera for this sort of application, and I had hoped the A77-A99 would step up and provide this much needed attribute in the Sony mix. The question is ... does it?

Marc

 

[Ben Rubinstein]

I downloaded the RAW files from DPReview and played around with LR3.5. These are shot with studio lighting which is hardly the caves we usually shoot in Marc, studio lighting is hugely forgiving in general and without pushing them at all they are at best a full stop worse for noise than my 5D classics from iso 400 upwards TO MY EYES. They also fall apart faster when pushing. My benchmark is the 5Dc, lots of people need less, lots need more. I need my benchmark, a 6 year old benchmark for noise per pixel (who on earth buys a 24 megapixel camera to downsize?). I haven't any side by side comparisons but the noise looked worse than my 1Ds3 at all ISO's up to 1600 where it was similar but uglier and without the saving grace of the incredible tonality of that particular camera.

Bottom line, as always, you and only you can set your own personal benchmark, never let others decide it for you and your needs. The NEX 7 which I was very excited about, if it looks like this, I'm afraid I'd prefer to pass on...

 

[jonoslack]

I downloaded the RAW files from DPReview and played around with LR3.5. These are shot with studio lighting which is hardly the caves we usually shoot in Marc, studio lighting is hugely forgiving in general and without pushing them at all they are at best a full stop worse for noise than my 5D classics from iso 400 upwards TO MY EYES. They also fall apart faster when pushing. My benchmark is the 5Dc, lots of people need less, lots need more. I need my benchmark, a 6 year old benchmark for noise per pixel (who on earth buys a 24 megapixel camera to downsize?).

HI Ben
Whilst I agree that one wouldn't buy a 24mp camera to downsize, I still think that looking at the results 'per pixel' is very unrealistic . . For many users (myself included) I'm interested in the printed image (or possibly on screen), not the pixels.

Another point about the dpreview images is that they were focused right at the back of the setup, meaning that comparing specific areas doesn't work.

However, I don't doubt for a second that the files are a stop worse than the 5D files at 100% - but that pans out to on a 12x15" print for example will not be the same thing.

I can't make a judgement on the A77 files until there is Aperture support, but what I'm hoping to find is that they maintain the good colour and resilience of the A900 files at low iso (where, let's face it, most of our shots are taken). At higher ISO I'd like to see them printing up to an equivalent level of the D700 or 5D.

@Marc
Hope you're well - I haven't a lot to add to Philip's comments. We have 3 of the cameras in the family, and we're all really enjoying using them - the results seem fine so far, but as above - I can't tell until there's full Aperture support.

One thing I can say, is that the camera is a delight to use from an ergonomic point of view - great button layout, not too heavy and the EVF at least is a good evf! Focusing seems to be excellent.

 

[Mark K]

HI Mark
Is it still locked up? Did replacing the battery cure the problem?
I have to say, I don't think I've ever had a camera which didn't lock up at some time or other, but it isn't okay if removing the battery doesn't sort it out!

It is only semi-locked up. When the power on button still works. I tried removing the battery, soft and hard reset the camera but still, it remains the same. I am looking for a period during office hours to send it back to Sony. Hong Kong Sony offers resetting the camera and upgrade its firmware to 1.03.

 

[Ben Rubinstein]

Jono, when I'm shooting at high iso I'm shooting in bad light, I'm also usually shooting fast without time to manipulate my light. That usually translates as noise. What I saw was that those files looked worse per pixel than my 6 year old 5D at every iso from 400 up and that was when they had optimal lighting and I didn't have to touch the fill or brightness slider.

Thing is, I'm already uprezzing most of the time to print from my 5Dc for my fine art work. So I'm downrezzing to kill the noise the uprezzing again and at that point it does get rather silly does it not? That's what I meant about comparing at pixel level. I need to know that a camera is as good at doing 24 megapixels as my 5D is at doing 13 megapixels and if it isn't then why wouldn't I just continue using my 5D or a D700 or a Nex 5N or whatever? What am I gaining by having more megapixels if I have to negate them at a minimum to regain the IQ I lost by having them in the first place?

Again this is my own personal take for the kind of work I do, the kind of light I shoot in, etc. For a landscape shooter or someone who rarely goes over iso 400 they probably think I'm crazy. I shoot over half every wedding at iso 1600 though and when I'm shooting fine art I'm often using iso 800/1600 for various reasons (urban landscapes in low light and freezing people movement).

 

[fotografz]

I'd say that it is yet to be determined by some extensive real world shooting and some time to get the specifics of post processing down pat.

It is a new camera and new sensor. So in all fairness, it'll take some time.

As I mentioned, the key for me is whether it can shoot high ISO in the 1600- 6400 range ... if it can't do that well, it doesn't matter how well it handles or what other tricks it can do. I have plenty of high IQ solutions including the A900 (which isn't bad at 1600-6400 ... but that took some time to achieve).

I grinned when I read that "most of us shoot at ISO 100-200" ... to me those are tripod bound studio ISOs using strobes, or for places where a neutron bomb went off and there are no living things moving about. Most of the time I start shooting available light at 640 and go up not down. I actually don't think I personally know a single photographer that shoots ISO 100 or 200 on any sort of a regular basis ... maybe some MFD guys, but no 35mm DSLR shooters. But this place is obviously different.

Best of luck with the A77, and do keep posting impressions and pics for those of us that can't get one, or have more patience

-Marc

 

[Ben Rubinstein]

There were rumours of a new firmware to solve the high iso thingy, something about the software being written badly for the new processor?

 

[dhsimmonds]

I agree with Marc about ISO settings. My A900 is permanently set at ISO 400 And I am fully prepared to use 800 or even 1600 if I need it, knowing that I can get a reasonably clean image.

Since getting my little Fuji X100, it is permanently set at ISO800 and I have no qualms about using 1600 or even 3200 if I really need it. These latest sensor designs from Sony are really incredible and experts are tearing up existing text books on digital photography as fast as we can write all this stuff!

I am still waiting for my A77, there does seem to be some delay in getting the second batch into the UK but I am assured by Sony UK that it won't be too long now. I can then report on the latest Sony sensor as part of a Sony camera system, hopefully providing the ergonomics and beautiful colours that I love so much from Sony.

 

[fotografz]

One other note on use of high ISO in lower light: (this isn't specific to the A77, but is relevant when examining A77 higher ISO results IMHO.)

There really are only a few reasons to jack up the ISO; to increase the shutter speed for action photography or longer lens use, or to increase sensitivity to gain DOF with a smaller aperture, or to gain sensitivity in very low ambient light in order to make any photo at all using the slowest shutter speed and widest aperture you can get away with.

IMO, the hand-held acid test is low ambient light ... which is almost always artificial light in my case (and most photographers I know.)

The issue I have with many "test shots" touting decent high ISOs is that they are often heavily Tungsten contaminated, which IMO is a false result in terms of evaluating high ISO performance and attendant noise. (this is ignoring those images where color temp contamination was deliberate for creative reasons.)

I find that it is increasing difficult to use any auto or even manual color temp setting other than custom WB, as artificial lighting has been migrating to energy efficient bulbs which are warmer than anything before ... moody directional lighting scenarios being the worst.

So, I see ISO 3200 and 6400 shots that are quite warm red/yellow that appear to the eye to be properly exposed, but when accurately color corrected become under-exposed looking with shadow density visually blocking up. Attempts to open those shadow and darker end mid-tone areas then reveals the "real" noise issues.

BTW, regarding this as it relates to the A77 EVF, it is of great interest that the adjustments of color temp are real time in the viewfinder ... please correct me if am I wrong on this. If true, coupled with a decent high ISO in that 1600 to 6400 range, the A77 (or A99) would be exactly what I personally need for ambient candid work at events and on the street.

-Marc

 

[jonoslack]

BTW, regarding this as it relates to the A77 EVF, it is of great interest that the adjustments of color temp are real time in the viewfinder ... please correct me if am I wrong on this. If true, coupled with a decent high ISO in that 1600 to 6400 range, the A77 (or A99) would be exactly what I personally need for ambient candid work at events and on the street.

-Marc

HI Marc
There's no doubt that there are benefits in the EVF - I quite agree with what you say about WB and low energy lighting.

But I think the jury is still very definitely out on the A77 image quality - not because I have any particular doubts / worries, but because the current firmware (1.02) was clearly optimised for the wrong processor (!!). RAW support is available for lightroom now, and also for C1, but this may also change with the new firmware.

The obviously splendid IQ on the new NEX 5n suggests to me that we can expect something similar from the A77, but I guess we'll just have to wait.

In the meantime I'm enjoying using a delightful camera.

all the best