Tuesday, 4 March 2014

Sony a7 - innovation or gradual (often sideways) progress


People around the various forums are going on about how the Sony a7 is some sort of game changer. For sure its an attractive camera for many reasons, but to me it represents more gradual progress than stunning game changing innovation. Commonly people believe that cramming a full frame into such a small package is somehow amazing. My view is that its not amazing and in fact long over due. Make no mistake I think the a7 is a desirable camera, but if its a game changer its mainly because of price for full frame.

I have argued for some years (like about when the G1 came out) that there was no real reasons why interchangeable lens cameras needed to be so big as the removal of the mirror box (you know, SLR has a moving mirror right?) makes way for allowing much smaller cameras.

In a long ago blog post (here) I presented this comparison of the G1 and my 10D and went on to discuss that while the G1 looked quite small that in reality it was not so much of an innovation in size. In fact we had been progressively getting used to obese cameras and many years before 35mm cameras were actually the same size (or smaller) than the amazingly small mirrorless.

Essentially people who grew up in the digital age somehow suckle up to the advertising nipple from the camera makers and just accept that cameras must be big because Nikon and Canon made big cameras. I presented two images taken without moving the (taking) camera (or refocusing the manual focus lens) of the G1


and then OM-10 35mm film SLR in exactly the same spot:


They look to be almost exactly the same size, because they are. This becomes clearer when you hold them in your hand.

Now since that time almost nothing has happened in the SLR market because perhaps nothing needs to happen. Economic Times are tight and if people think that things are all ok, well then don't rock the boat and keep selling them what they are already buying. Essentially my view is that Canon and Nikon (well and almost everyone really) have stopped innovating and just churned out more of the same because no one demanded it.

Now to support my view that the A7 is not really a massive innovation let me present a few pictures (to perhaps save a few thousand words).

getting rid of the mirror for starters

The new trend in mirrorless cameras has made it clear that a mirror box (if not an actual shutter) can be eliminated from in front of the sensor, making room. Decent SLR cameras have a little circle with a line through it to allow you to measure the distance from subject to film plane (although this has dropped off many cameras now days as noone actually does this anymore).  The 5D however still has such a marker, so its easy to see where on the body the sensor plane is.

Everything in the grey area can now go.

So with that identified we can look at what part of the camera bulk exists just to allow the mirror to flop up and down (and not collide with the lens) and could be removed.

Which of course in mirrorless cameras it is. I've put below 3 mirrorless cameras (2 micro4/3 and one APS-C) from even the first ones released way back in 2009 to show that indeed this is how much that in reality can be saved.

Clearly there is much less space needed with a mirrorless interchangable lens design.

The next thing is of course bulk ... why so big? I can only suggest its because its a fashion. I suspect its from the days when people seemed to expect cameras to be bulky, and perhaps a public mentality:
A big camera is a kewl proffessunal lookin camera. I wanna take proffessunal lookin pictures so I need a big camera.

So comparing the a7 and the current 5D we see that from the front too there is much extra size. I've taken these two images and scaled them so that the sensors are both aligned and the same size in pixels and as they are both 36x24 FF sensors that should be a good approximation of

Now recall that the Sony a7 actually uses the same mount as the NEX, meaning that the mount holds the lens away from the sensor by exactly the same distance. Which makes it clear that an a7 could actually be the same dimensions as the m4/3 (which gosh it is). So to me all that Sony have done is to recognize that a market actually exists and put the bigger sensor into the body that the mirrorless design enabled nearly 5 years ago now.

I recall being stunned when I first saw the D30 in Tokyo that it was in fact fatter than my EOS 630, which is a film camera and so even has all the motors needed to pull the film through (as well as two spots for the film to come from and go to). This was explained away (back 15 years ago) as being needed for the electronics and battery ... well that aside looking at the current m4/3 cameras all that space clearly hasn't been needed for some time.

So to me Sony have introduced a welcome progression in digital cameras perhaps a little later than I'd like, but none the less very welcome.

What Sony has in my opinion also done is to show up CanNikon for being the dinosaurs they are and not innovating at all. To be fair however since clearly the market (comprised in the main of clueless sheep but with quite a few KoolAid drinking Wangers proclaiming they can see the way) has not demanded it, why supply it?

In my view Sony has been quietly picking up expertise in the area of Digital Cameras starting back in 2009 with their very desirable Alpha series (which btw was still a 25Megapixel Full Frame camera not much development in size if you ask me).

Personally I'll be pleased when the camera makers stop with megapixel madness and start making sensors which actually do yeild a good 14 (or more) bits. Even this is not really innovation because back in 2004 Fujifilm made a great camera called the S3Pro with the SuperCCD in it. This essentially allowed the user to get a fantastic dynamic range. For instance in testing I did ages ago I found that the Canon sensor only gave a little over 11 bits of data (0000 1111 1010 0111) while the Fuji SuperCCD gave a good 14 (0011 1101 1111 1101). Folks that's 2 stops extra headroom. A stunningly larger amount of actual information in the photos. The DxO site has since found this out (and made the data a little more mainstream) so the message is beginning to be understood. What this means for the normal photographer is of course less blown highlights as well as a much smoother dynamic range without needing HDRI.

For example below is an image comparing what was captured in RAW on the SuperCCD and the camera JPG (same camera settings). I think anyone can see that the reds are just blown severely on the JPG.

No blown highlights and no blown red channel. Even RAW files from other makers would blow the channels in this test. Now personally I'd call that more like innovation, not just whacking in more megapixels, as to be honest there are very few lenses on the market (and perhaps fewer buyers) which will get the 36Megapixels in resolution projected on the sensor.

Such innovation in digital imaging will not really take a hold however until more people use RAW or camera makers make better RAW engines in their cameras and see the differences (even in gentle light camera JPG engines have a tendency to blow channels see here). So perhaps again such innovation will not be around for a while.

Meanwhile another glass of Corporate Kool-Aid for you?


Anonymous said...

I tried a friends a7r at work yesterday. The electronic viewfinder was much better than anticipated. You could see individual hairs on the subject, without magnification. Except for digital indications of over exposing (zebra blinking) I thought there wasn't that much difference to optical viewfinder of a digital slr. (My analog Pentax MG viewfinder beats everything.)


obakesan said...

Hey Anon/Alex

the EVF's are getting good these days aren't they. I was against EVF's (being horrid pixelated crap) until the G1 came out, since then they iteratively get better.

The think I love about an EVF is that I can quickly magnify and check focus (unlike whapping on an expensive accessory magnifying optical view finder), better yet I'm checking focus on the taking sensor and don't have to worry about focus screen - sensor alignment issues.

I loved my Pentax MX finder too, I was quite sad when a house break-in saw it stolen. While I love my OM-1 I really prefer the LED series up the side that the MX had. Wonderful compact cameras those early Pentax cameras.


Yu-Lin Chan said...

One of the reasons they can't make the SLR very small, or at least very thin, is because of the lens flange, unless they redesign the mount. Take a look at the Pentax K01, it just can not be made thin even though it has no mirror.

obakesan said...

exactly .. which is of course why Mirrorless is so able to be thinner.

That's my point ... its mirrorless which has been the "leap" not something intrinsically amazing about the A7 being super spectacular (which is what people have been saying).

Anonymous said...

Have you taken a close look at the output from the Sigma DP Merrill cameras at low iso? I have a feeling you would find that very interesting.


obakesan said...

Alex, actually, sadly I've not been able to get my hands on the Sigmas. They are one of the innovators that's for sure :-)

obakesan said...

PS Alex ... if you have a raw file from one of them I'd love to see it.

Anonymous said...

My cousin just bought a DP2, and I'm waiting for my DP2 and DP3 to arrive. The raw files are just too big to email, and the sigma raw software is only used to output to tiff for further work in lightroom (such as cropping or export to jpeg). But I have some jpegs in full size I can email you if you like. Or better still - I will put them on drop box and email you the link. What's your email?


obakesan said...


Pellicle at hotmail will get me


HF said...

Fully agree to your article. I stated at some forms, too, that I didn't find the Sony to be a real game changer or innovation. It's an evolution, albeit a good one.