Discussion in 'Other Genres' started by Speedliner, Mar 28, 2016.
Using the Sony a6300 for Sports in Low Light - MirrorLessons
This is the benefit of having a camera and image sensor business under the same roof. Sony has a reason to develop the best-in-class APS-C sensor: they sell APS-C cameras. Sure, it's important to keep paying customers happy (those like Nikon, who also use Sony APS-C sensors in their DX cameras), but when you can develop the sensors that go in to your own cameras, that's when you can really drive development to be the best in your class.
I've said this before about Olympus. Olympus uses Sony m43 sensors in their cameras, but the problem is that Sony has no real reason to dedicate a ton of resources to developing a revolutionary m43 sensor for Olympus, especially since Olympus cameras compete with Sony's APS-C E-mount cameras. Instead, Sony can make evolutionary changes to the sensor, modestly increasing things like resolution, dynamic range, and ISO performance, while dedicating more resources towards their own APS-C sensors that will go in their cameras, as well as others like Nikon.
If Olympus is serious about really driving m43, why not buy and use Panasonic sensors in their m43 cameras? At least Olympus and Panasonic have skin in the same game: driving the m43 system forward. By purchasing sensors from Panasonic, it will give Panasonic more money to put towards developing the next revolutionary m43 sensor, helping keep m43 on a more level playing field with companies like Sony and Fuji.
I know there's more to it than that (possible purchasing agreements and what not), but I don't see why Olympus continues to use a sensor made by a direct competitor, rather than driving that business towards a "partner" who actually wants to create a sensor that can outperform the APS-C sensors in the most recent cameras.
I think the E-M1 sensor is made by Panasonic. Maybe the GX8 and PenF too?
The E-M1 did but I believe that is the only Olympus camera to do so. I believe the Pen-F uses one of the 20MP Sony sensors that was mentioned in a patent filing a couple months back.
Unfortunately the Pen-F and GX8 are too new for them to be in the database for what cameras use what sensors.
Panasonic is more of a competitor to Olympus than Sony - Sony owns a piece of Olympus' camera business, while Panasonic only shares m43 standards with Olympus and competes directly against them in camera body and lens sales apart from the E-M1 sensor.
While Sony does own a partial stake in Olympus, I would consider Sony/Olympus to be competitors, while Panasonic/Olympus to be substitutes.
By definition, a substitute is a product that consumers view as the same or similar. Within the camera market, this holds true for Panasonic/Olympus, because m43 owners can buy camera bodies or lenses from either manufacturer and continue to use the same gear.
Meanwhile, Sony/Olympus are competitors because consumers are forced to choose one or the other, as they share nothing between them.
Also, with what you're recommending, Sony can continue to product evolutionary sensors for Olympus' m43 cameras and say "take it or leave it". In doing this, Sony has no reason to make the best m43 sensor they can, especially if it can equal or beat the performance from their APS-C sensors. Take their 1" sensors for example. Sony has a stake in the 1" camera market with their RX10 & RX100 cameras, and therefore have a more compelling reason to dedicate more resources in the development of those sensors. That's why the 1" sensor in the RX10 II and RX100 IV perform nearly as good as the m43 sensors, despite being much smaller in area.
With Olympus taking their business to Sony, Panasonic is left to foot the bill on their own for m43 sensor development. So, Panasonic can't devote more money towards developing better sensors that will compete with the next-gen APS-C sensors, and Olympus continues to get mediocre sensors from Sony, helping fund the development of the next generation APS-C and 1" sensors from Sony which will only further outperform the m43 sensors.
It's a downward spiral. The old saying "If you want something done right, do it yourself" holds true in many situations, and that's the case with m43 sensor development as well IMO.
Olympus did use Panasonic sensors (12MP) up till the E-M5 & they weren't getting a great sensor when other manufacturers were turning out much better performers. Sony offered something better & Olympus took it. The only Panasonic sensor since has gone into the E-M1 & with limited video use for two years (while the GH4 excelled with video using the same or similar sensor). That contract finished last year in October (2015) when we got FW4.0 in the E-M1 with improved video selections.
I guess it all depends on what can be available at what cost & limitations that might be applied to its use (at least for a limited time).
sadly there are only 2 or 3 (depending on whether Samsung is still making sensors or not) companies producing camera sensors. Panasonic and Sony being the big 2. Spinning up the R&D facility to get into that game is going to be really expensive. With the tight and diminishing margins on digital cameras, I don't think it would make huge sense to develop a sensor business for consumer cameras.
Sony just announced a restructuring around their camera business. They may spin it off into a separate division. Will the sensor division go with them or will they sell sensors to the spun-off company as they do to other corporations? It's an interesting question, and I'm sure they're eager to keep their existing customers, Nikon, Olympus, et.al, and at least give the appearance of providing them with excellent service.
When a camera company like Nikon or Olympus deals with the sensor division at Sony, there's a fair bit of back and forth with their engineering teams. Nikon has direct influence on the sensors being produced and I expect Olympus has some pull as well. That arm's length dealing probably has an impact on time to delivery however.
I remain patient and optimistic in the eventual release of an amazing EM1mk2.
The sensor division is a separate Sony company now & it just means the imaging division will become an independent Sony company.
Here's the quote from DPR: Sony is considering splitting off its imaging products business, according to a new statement from the company. Additionally, its previously announced semiconductor business split will take place next month, placing its image sensors under a wholly owned subsidiary.
I wonder if Olympus' switch to Sony sensors for the E-M5 was due to Panasonic suspending sensor development in 2011? The E-M5 was released in early 2012, so could it have been that Panasonic notified Olympus that the sensor development would be halted, at which point Olympus went to Sony for their sensors?
One thing to look at is the comparison between Sony APS-C, m43, and 1" sensors in 2012 and 2015/16. Referencing DXO's sensor database and comparing the E-M5 to Sony's NEX F3 and RX100, all of which were announced/released in early to mid 2012, we see the following performance (link to comparison):
Resolution: 16 MP
Color Depth: 22.8 bit
Dynamic Range: 12.3 EV
Low-Light ISO: 826
Sony NEX F3
Resolution: 16 MP
Color Depth: 22.7 bit
Dynamic Range: 12.3 EV
Low-Light ISO: 1114
Resolution: 20 MP
Color Depth: 22.6 bit
Dynamic Range: 12.4 EV
Low-Light ISO: 390
So comparing the Sony sensors from 2012, the m43 sensor was basically equivalent to Sony's APS-C and 1" sensors of the same time, with the exception of ISO noise where the APS-C sensor performed better by about 1/3 a stop (not a big deal), and the 1" sensor performed worse by about 1 stop. Coincidentally, it was the E-M5 that really put m43 on the map as a serious camera system. It was extremely popular, sold very well, and likely took business away from a competitor like Sony.
Now, let's look at 2015/16 Sony sensors for APS-C, m43, and 1" to see how each has progressed (link to comparison, note the a6000 and E-M5 II were used as the a6300 and Pen-F haven't been tested by DXO yet):
Olympus E-M5 II
Resolution: 16 MP
Color Depth: 23 bit (0.9% improvement over 2012 m43 sensor)
Dynamic Range: 12.4 EV (0.8% improvement over 2012 m43 sensor)
Low-Light ISO: 896 (8.5% improvement over 2012 m43 sensor)
Resolution: 20 MP
Color Depth: 24.1 bit (6% improvement over 2012 APS-C sensor)
Dynamic Range: 13.1 EV (6.5% improvement over 2012 APS-C sensor)
Low-Light ISO: (21% improvement over 2012 APS-C sensor)
Sony RX100 IV
Resolution: 20 MP
Color Depth: 22.9 bit (1.3% improvement over 2012 1" sensor)
Dynamic Range: 12.6 EV (1.6% improvement over 2012 1" sensor)
Low-Light ISO: 562 (44% improvement over 2012 1" sensor)
So, comparing Sony-made sensors from 2012 to 2015, it's pretty evident that where Sony makes and sells cameras, the sensors in those cameras see a larger improvement compared to cameras where Sony simply makes and sells the sensor, which then goes in to a camera that directly competes with Sony's 1" or APS-C cameras.
Couple this fact with the announcement from late last year where Panasonic is restarting their sensor development, and I seriously question why Olympus continues to funnel money to Sony, who is apparently putting more development efforts in to their APS-C and 1" sensors compared to their m43 sensors. If Olympus gave their business to Panasonic, it might help Panasonic develop next-gen, revolutionary sensor technology for the m43 cameras, rather than the basic evolutionary development that Sony is providing Olympus with.
It's AF performance that concerns me more at the moment. The PEN-F's performance is disappointing. That as much a software issue as H/W and inexcusable.
For e-m1 and phase detect, the lack of phase detect technology elsewhere in m43 is a concern. R&D dollars and the economics of a subset of a relatively small volume platform is a worry.
I've stopped buying m43 lenses and such until I know where e-m1,II is headed. Hoping for bees soon. I'm sure the design and direction were determined quite awhile ago.
I need to know!
Calm down people!!
Did anyone noticed that the latest Nikon D5, Nikon's flagship, has base ISO photographic dynamic range the "SAME" as an E-M5 Mark II or an E-M1? In fact, the D5 has gone backwards in its base ISO and is Nikon's current LOWEST base ISO in the offering. And it's a very expensive camera.
Bayer sensors have basically reached its limits. What we are seeing now is simply sensors being made more efficient to squeeze out any performance left or in the case of the D5, trade off low base ISO with more high ISO performance. I also disagree that the RX100 series is close to the IQ of the recent m43 sensors. Certainly, that's not what I'm seeing from real life images and also confirmed by Bill Claff's analysis from Photonstophotos.net. I think we need to understand that at the end of the day, we are all taking pictures and seeing them on prints. If you pixel peep; YES you can see a significant difference in big prints which separates m43 vs 1" Sony sensors. If you are printing big, like 24x36 to begin with up to 40x60 and beyond, that is when you will need the 36MP Nikon D810 or the 42MP Sony A7RII. These 2 cameras will provide enough resolution, tonal gradation and base ISO dynamic range to make fantastic huge prints. If you don't, then why fret about dynamic range? People who shoot with a Nikon D5 isn't going to make big big prints like others who own a Nikon D810. Olympus makes cameras with certain compromises for a specific market just like Nikon does with its D5.
Having said that, I dislike when people compare the dynamic range of cameras based on "LOW ISO" without taking account the shooting discipline. Yes, while the latest full frame cameras and the Sony A6300 have 1.5 to 2 stops more dynamic range, but that's at base ISO. As the ISO rises, the dynamic range drops. So if you plan on shooting with base ISO, your full frame shutter speed will almost always be 2 stops slower than an E-M5 Mark II. With the A6300; about 1 stop slower than the E-M5 Mark II. However, the E-M5 Mark II's latest IBIS has 1 extra stop of added stability without the need of a tripod. There is no IBIS with the A6300 or with your D810, you will always need to carry a tripod if you insist in shooting at base ISO only. In fact, that's exactly what those full frame guys do in my meetup groups. They carry a heavy carbon fibre tripod with a beefy tripod ALL THE TIME with them during daylight and at night whereas I, with my E-P5, is FREE from tripods unless I want to get the best sharpest image using Base ISO 200. Having said that, that is my shooting style. No tripod. Guys with tripods are slower and limited in their composition, whereas with me, I am free and nimble and ALMOST ALWAYS gets the best shot.
But if I am to go head to head with a Sony A6300 or a Sony A7RII, then of course my E-P5 will loose at base ISO or higher ISO. But you can mitigate that somewhat with IBIS and fast primes if you shoot within Bill Claff's photographic dynamic range. Incidentally, most professionals I know DO NOT SHOOT at insane high-ISO just to prove a point that they can. They are shooting with Ansel Adam's Zone System in mind which actually happens to mirror Bill Claff's photographic dynamic range chart. At the end of the day as I keep harping here; it's all about getting the shot with pleasing colors, tones and sharpness for presentation. We are not shooting for 12.6 ev dynamic range and ISO 890 etc.. Those numbers don't appear on a photo. If it does, it'll be a boring photo to say the least..
So be happy with your gear and be rest assured the E-M1 Mark II will be competitive AF wise with whatever same level of cameras you see out there with mirrorless.
In regards to autofocus and tracking; keep in mind that the E-M1 is a 2013 camera with a technology dated back further back. And it took Sony a very long time to come out with the A6300 probably to solve the low light performance and accuracy with the A6000. My friend's wife A6000 had to be sent in 4 times to fix the continuous focus because while it was fast in tracking, it kept front or back focusing a lot. It even does that in single AF, but mostly in low light.
What good is fast autofocus if what you get is front or back focus and even sending to the factory 4 times and replacing it with a refurb camera couldn't fix the problem?
Don't be too impressed with specs. Sometimes, real life performance don't always mirror specs..
You make some good points about "measurbating" @bikerhiker, but I do want to mention a few things:
I do regularly print at 30"x20" or 36"x24", most of them landscape-style shots, so improved dynamic range and noise performance at base ISO is really important to. It would allow me more latitude in post-processing my images with regards to pushing shadows without introducing excessive noise, or pulling highlights while retaining detail and information. I can do it with my GH4's, but it requires me to closely monitor my exposure, and a lot of times shoot HDR for really challenging scenes.
As for talking about the shutter speeds @ base ISO, that's not a real concern for me when I'm shooting landscapes, because I typically am using a longer shutter speed anyways. I think a situation where you have to look at that is say for sports. Say you need an ISO of 200 with a shutter speed of 1/500, and the aperture required for this exposure is f/4. So, how does a FF or APS-C camera perform at those settings, which aren't base ISO for their sensors, against a m43 sensor where ISO is the base? My guess would be that the larger sensors would still perform better in terms of noise and dynamic range.
Honestly, I don't think additional resolution is required. The bump from 16 to 20 MP isn't that much in terms of actual pixels along the width and height of the image. If I want to print large, I end up up-sizing the image anyways, and it prints fine regarding detail when enlarged. What I think would benefit the m43 system the most would be a new sensor, 16MP, but with an actual base ISO of 100, with lower noise and better dynamic range at base ISO. If this compromises high ISO, so be it. I would be fine with a new 16MP sensor with a sensitivity ranging from 100-6400, but people get so caught up in the number of megapixels that they don't care about parameters that will actually affect image quality, things like noise and dynamic range.
I'm a bit surprised by this post. Literally the exact same pros and cons relating with respect to the A6300 vs. the E-M1 could have been posted 2 years ago with the A6000 vs. the E-M1. The relative advantages and disadvantages are largely the same.
The Sony has, and always had, more resolution, better AF tracking, and a deeper buffer than it's M4/3 competition. The Sony has, and always had, worse build quality, a worse EVF, fewer direct controls, no real weather sealing, a much worse lens line-up, no touch-screen, a clunkier user interface generally, and fewer bells & whistles.
What is it about the A6300 that's made you sit up and take notice?
The E-M1 will be refreshed eventually. Maybe the same relationship between the two bodies will exist then, who knows. The A6000 just happened to get refreshed first.
I have compared images from the Sony A7R and the A7RII as well as the Nikon D810 against my E-P5 at base ISO and if I am to dedicate myself shooting landscape, there is no doubt my favourite camera will be the A7R or the D810. It just so happens that they have 36MP so I chose them. The Nikon D7200 is not too far behind as well as the Nikon D5500 which is a steal if one wants a cheap landscape companion to your current m43. I use Perfect Resize to upsample my 12-36MP to bigger sizes (both Olympus and my Nikons) and I find that it's good enough for the job. So I agree with your assessment that having more dynamic range and printing big can help with challenging scenes.
Having said that, I also use the Nikon Imaging Space service and upload my photos to share privately with clients. The good thing about this service is that all other NPS shooters use the same free service. What's interesting is that, they have a new reporting feature and in it, it shows your most used apertures, focal lengths and shutter speeds and you can compare it against others in the database and creates a graph. Almost always my shutter speed is 2 stops faster than Nikon APS and full frame sensors. My main used aperture is f/1.4 and f/2.8 and theirs are f/4 to f/5.6. Is it ever a coincidence that they are about 2 stops apart?!? And this is real life usage; not some theoretical setup you find at DPREview image comparator or DXO. This is what real life Nikon shooters shooting parameters really are. Now; some people may fall under some rather unique circumstances like yourself who shoot with slower shutter speeds. But generally, people tend to stop down 2 stops from wide with their full frame lenses where my m43 pro zooms and primes I do not and hence, I gain the same dynamic range and noise performance if I don't stop down 2 full stops like the full frame people do. If you decide to carry a tripod and shoot at lower shutter speeds, then I simply can't compete with my E-P5. But then, I'm not carrying a tripod. If I'm after max quality, I would carry a heavy tripod with the D810. But how often does anyone here need to print big or need max quality. Not much and a very small small percentage. Certainly showed from the data I saw with Nikon's report. In fact with landscape, I need MORE depth of field not less, which is why the reason I carry a tripod with my E-P5 would be to create stack of images so I can focus stack with Helicon Focus and create better deeper sharper landscape images for prints.
@bikerhiker, yeah if you shoot hand-held and in good light most of the time, the m43 system is fantastic because the quality at base ISO is quite good, and the lenses are sharp wide open. No need to stop down 1-2 stops like with most DSLR lenses.
Like you said, my type of shooting may be a special case, but when working with wide dynamic range scenes, it's very easy to blow highlights or underexpose the shadows too much, so I'm pretty careful about exposing with the histogram, and shoot a scene with a bracketed exposure covering 5 EV's to capture detail in those shadow and highlight areas. I'm really thought about getting either an A7R or A7II to use with adapted Canon FD glass for landscapes and what not, but when I look at the cost to buy the camera, accessories, and lenses, it hardly seems worth it. I have invested money in the highest quality m43 glass and it certainly shows in a number of my images, but good glass can only do so much for you, where a good sensor can really take an image to the next step.
I too use an enlarging software, Blow-Up, and like you said, it works fine. I crop most of my images to 3:2, so I'm only working with a ~14MP file at best (if I don't do any additional cropping), but I find printing at 30 or 36 inches wide to not be a big problem. I've printed even wider when stitching images, and have had great success there as well. I just wish we could get a m43 camera that would emphasize things like a lower base ISO, better dynamic range, and lower noise over something like megapixels. Oh well...
The Sony did get a little more attractive for sports and wildlife with the launch of the 70-200 GM last month and the new 70-300mm G today. Both pro-grade telephotos.
Still nothing to compete with a PL 100-400, though.
Not so concerned with the sensor. Better is good, but it's the C-AF that needs to catch up the most for me.
Granted that it's the Archer more than the Arrow, but I want to hunt with the sharpest arrows.
But don't you enjoy the challenge of hunting with a dull arrow? It just makes it that much more exciting when you actually catch something!
Separate names with a comma.