August 28, 2014, 01:31:25 AM

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - sarangiman

Pages: 1 [2] 3 4 ... 16
16
Interesting that the actual improvements seem so minimal. My comments below refer to the Mark III.

One thing that really bugs me - and I brought this up with them - is that magnified view isn't activated automatically by turning the MF ring (in Green mode). It's nice to have the option of AF *and* MF without having to switch modes. I was told this has something to do with the modes the adapter 'simulates' with the Green vs. Advanced settings -- essentially a native E-mount lens and one attached via Sony's own A-to-E mount adapters, respectively. So I guess it's tricky to change the behavior of the adapter when the adapter simply emulates these modes.

It's a shame though - especially when you also factor in that 'Advanced Mode' doesn't play well with electronic 1st curtain (you get a neutral density gradient from top to bottom at certain shutter speeds), and AF *always* fails in this mode (which I also don't understand).

Also, the adapter routinely 'crashes' on me, which I think someone else brought up above. The F-number will start flashing & I'll have to reset by either un-mounting & re-mounting, or sometimes even by pulling the battery out. Previously I've wondered if this has to do with loose contacts due to the 'squishy' lens mount - no idea if that's it though. At least this mount is vastly tightened on the A7S.

On the bright side, I'm incredibly glad these adapters exist. Canon is making very nice new glass, & it's great to be able to pair them with the A7R. Especially given the dearth of native lenses. For example, the new Canon 16-35 f/4L IS plays quite well with the A7R - assuming your lens mount & adapter are perfectly aligned (mine aren't, sadly, so I'm trying to make a custom shim).

17
EOS Bodies / Re: DxOMark Sensor Performance: Nikon vs. Canon
« on: May 09, 2014, 09:09:37 PM »

You are still missing the point of argument here, though. Yes, it does work well to underexpose those cameras by four stops, then lift the shadows in post. But doing so is an unrealistic test from a real-world standpoint. All it tells you is that IF someone were to accidentally expose their scene incorrectly by a HUGE amount (some 16x incorrectly), then they would have a greater ability to recover. Purposely underexposing by four stops for the purpose of comparing cameras is also an unrealistic real-world comparison.


Tell that to Sony, who in their A7S is introducing S-Log2 gamma curve application that raises shadows 3 stops and rolls off amplification in highlights. In other words, ISO 800 in shadows, but ISO 100 in highlights. All software-based, since actual ISO amplification off the sensor remains at ISO 100 levels.

If that's a reasonable push for videographers, why not for still photographers? Why is +4EV suddenly entirely outlandish, when a camera manufacturer is now allowing you a +3EV push as an option in-camera?

The levels to which my post was misconstrued was amazing, which is why I chose not to come back to this thread (until now). I mean, I even admitted I still shoot Canon and work around its limitations... but boy have I been so much happier with the Sony A7R sensor (barring shutter shock problems) over the past 6 months. On a boat ride in Cambodia at sunset I set my exposure for the sky and shot away without a care for my exposure - raising shadows 5 EV in post to balance out the shadows of detail & people's faces in the boat (all while holding back the exposure in the sky).

The point stands: +4 EV shifts are not unreasonable. It's the difference btwn ISO 100 and ISO 1600. If such pushes were unreasonable, why would Magic Lantern have implemented a dual ISO feature that people use to shoot ISO 100 and ISO 1600 simultaneously?

Oh, incidentally, Magic Lantern's dual ISO feature wouldn't even be needed on a Sony sensor - because you could perform the ISO 1600 push digitally! Which was the whole point of my original post.

I do still shoot Canon professionally for weddings/engagements, but that doesn't mean I don't wish it had the Sony A7R sensor. It's common knowledge now that some aspects of Canon's sensor tech are severely outdated. ChipWorks stated:

Quote
Of the Canon DSLRs analyzed, the imaging chip has remained analog, with Analog Devices’ analog front end (AFE) chips handling A/D conversion en route to the Digic-branded ISPs. Perhaps the column-parallel ADCs favored by others can’t be implemented using 0.5 µm design rules, but more likely Canon is satisfied with its system design and performance.

http://www.chipworks.com/en/technical-competitive-analysis/resources/blog/full-frame-dslr-cameras-canon-stays-the-course/?lang=en&Itemid=815

There are even other, far reaching implications of low base-ISO DR. It limits your ability to retain DR while shooting low light scenes (since you have to raise the ISO amplification in camera, as opposed to selectively raising it in post-processing from an underexposed, lower ISO file).

As for people who don't run into these problems much - YMMV based on the DR of the scenes you shoot. Posting an example of a scene you think has high DR, but where you didn't see offensive noise in shadows, is not a way to prove a camera has just as much DR as another. It's just a way of showing that it had enough DR for that particular scene. Which is why we have controlled tests, like the ones DXO perform.

By the way, thanks to those that stepped in to point out why the misinterpretations of my presentation were flawed. The truth is, low read noise sensor technology will enable new features for photographers only dreamt of in the past. To sit here complaining that some tests push (pun intended) technology too much to show unflattering limitations is, in my opinion, incredibly short-sighted. Which is especially apparent when you begin to see manufacturers and software developers doing +3 and +4 EV pushes of shadows in firmware/software (Sony, Magic Lantern, respectively). That's pretty telling.

18
EOS Bodies / Re: DxOMark Sensor Performance: Nikon vs. Canon
« on: May 09, 2014, 08:08:27 PM »

I see the difference in noise and it sure is there! But, this example is flawed in my opinion since the exposure in these photo's was just wrong. I have seen that current Sony sensors offer better shadow recovery, no doubt about that. But in this particular situation it was not necessary to have such an underexposed image and the same photograph could be achieved by both Canon and Nikon camera when properly exposed.


But in another scenario, this exposure could have been needed to avoid clipped highlights. So what you're really complaining about is that I didn't - in my infinite free time during which I happen to have both a D800 & 5DIII on my hands - find a scene with enough DR. That's a fair point, sure, but also - as you can hopefully see - a bit unfair. Because the end result would be the same.

And, frankly, that's what I'd say to everyone here who, in all the comments that followed my original post, indicated that my shot was underexposed. Meanwhile, just a few minutes before this scene, the sun was high enough in the sky that the sky would have blown. So are you complaining that I didn't shoot just a few minutes earlier when there was more scene DR? Would that have changed the end result? Are you arguing that there do not exist scenes that require more DR than what Canon sensors can provide?

Furthermore, according to my histogram when I was shooting, my channels were clipped in the sky! Therefore, the camera led me to believe that I did accurately 'expose to the right and for highlights'. So this is the result an informed photographer would have gotten anyway. Now, if you're going to complain that I didn't have a computer with me to actually check the RAW values as I was shooting, then...

There are many valid suggestions you could make. *Perfectly* nailing exposure within 1/3 EV to get your highlights just short of clipping is not one of them (though it is a noble effort). Not the least because most cameras don't accurately tell you when you're just short of clipping in RAW - in my experience, they tell you you're clipped well before you're actually clipped.

Now, you could have told me to bracket, or bracket for HDR. Valid points - but sometimes that's not an option for certain types of photography. My point is: the conclusion doesn't change. D800/Sony sensor performance gives you much more DR and much more processing latitude. That's useful to some people in certain scenarios.

Also from http://www.luminous-landscape.com/essays/on_safari.shtml#suckout :

Quote
The bottom line is that the vast majority of the tonal information that a sensor can capture lies on the far right side.  In fact, usually 75% of the total tonal information a sensor can capture lies in the small right hand area of the two top F/stops just below pure white.  And yet, almost all cameras leave the factory calibrated to center the histogram instead of moving it as far to the right as possible.  To make matters worse, the screens in the backs of cameras are also calibrated to show a good exposure with the histogram centered. This is nonsense!

So what are the consequences of Tonal Suckout?  As the name implies, it is as if someone grabbed your image and sucked the tonality out of it,  If there are millions of shades of green in a landscape, you may end up with a few dozen.  If there is a terrific richness of tonalities in a face, you will end up with a small sub-set that makes the face look bland.

Maybe useful to keep in mind the next time you want to get an underexposed image!

Just FYI, that's actually a fallacy. Read Emil Martinec's extremely informative treatise on noise: http://theory.uchicago.edu/~ejm/pix/20d/tests/noise/noise-p3.html#bitdepth

Essentially, any extra 'tones' will oversample shot (statistical) noise. Consider a signal of 200 photons generated ~100 e- of signal (QE = 0.5). At ISO 100, say we bin 4 electrons 1 DN (digital number, or increment, in the digital file). So that signal is recorded as 25 in the RAW file - pretty darn low (dark). Now, b/c of shot noise, that signal already has a stdev of 10 e- (sqrt of 100), or 2.5 DN. So there you go, you have enough bit depth to still sample the noise in the signal. So it's not limited bit depth/tonality that's the problem - it's the amount of light itself. You expose to the right (ETTR) to minimize statistical/shot noise.

Or to avoid running into the noise floor of Canon sensors... make sense?

19
Lenses / Re: EF 24mm f/1.4L II USM
« on: July 21, 2013, 05:02:04 PM »
I never used my 24 1.4 II all that much below f/2.8 or even f/4 so I actually ended up selling it for the 24-70 II (which, remarkably, more or less matches it for 24mm quality, other than more distortion and bit worse corners near wide open, but with actually LESS purple fringing/LoCA, many shots stopped down were almost identical between the two other than more distortion and LESS LoCA from the zoom; when I compared the 24 1.4 II to 24-105 OTOH, the prime blew that zoom out of the water, yes even at f/8 or f/10, believe it is a fallacy that all lenses perform the same once well stopped down, not even close.) Here and there I miss the 1.4, but not much and love the 24-70 II.

Hmm. I actually shoot the 24/1.4 at f/1.4 on the 5DIII & center AF precision is quite good after AFMA. I do remember corner/side compositions being quite out of focus though at anything wider than f/2.8 or so during the last wedding shoot, so I had to switch to live view AF. Part of this was also just side/corner softness at wide apertures. Actually, it's quite appalling how bad some of these primes are on the sides once you go to a higher resolution sensor -- plop the 24/1.4 or 35/1.4 on a NEX-7 (no optics in between, just a Metabones Smart Adapter) & you'll see flaws of the lens you'd barely see on a 5D II/III or what-have-you FF body with larger pixels. And that's not even using the full image circle of the lens. Use a SpeedBooster to use the full image circle of the lens & some copies of these primes don't sharpen up on the sides until f/8-f/11 (whereas they're sharp by f/4 on my 5DIII). But of course, in the latter scenario I'm adding extra optics in the way... anyway, I'm getting OT now.

20
Lenses / Re: EF 24mm f/1.4L II USM
« on: July 21, 2013, 04:53:03 PM »

Maybe some copies of the 24 1.4 II have sloppier breaking on the AF and combine that with the older measuring device that didn't measure slop as precisely and then combine that with a less precise AF system (for all the talk about the advanced 7D AF, it's mostly it's speed and having so many points and so many cross points where it is advanced, the one shot AF precision of it is still basically a good deal less than the 1 series or 5D3 and not even quite a match for 5D/5D2 either for one shot precision) and maybe that explains it?


Very good point. Roger Cicala hypothesizes that the extra feedback from the AF motors (or sensors that can measure movement of the focus element) help the more advanced AF systems be more precise. This extra information is present with some of the newer lenses & work in conjunction with newer bodies, apparently. Not sure about the 24/1.4 II since that's a bit older. But who knows? The best you can do is treat the system like a black box & just test it.

21
Lenses / Re: EF 24mm f/1.4L II USM
« on: July 21, 2013, 04:49:01 PM »
Saringiman, I utilized only the central focus point throughout the entirety of the testing process, but thanks for making sure it remained stringently scientific; that is of course of the utmost importance.  I am glad to hear that you've experienced similar fickleness with this lens, as I know many others have as well, but I just really wonder how much of it is the result of the BODY's AF performance in low light, as opposed to simply a lens issue.  This obviously can't be completely the case, because when I pair up the 7D with the EF 100mm f/2.8L IS USM Macro, the issue is completely gone, but I find it interesting that in better lighting conditions, the 24L performs flawlessly.  What design flaw in the lens would cause this?  Is the 85 1.2L a beast in the good sense (performs well) or in the bad sense (is VERY moody), in the context of your comment about it...?  That's certainly another lens I'm interested in, but I need a wide angle solution right now, and this lens is certainly the best horse for the job if it would only behave...

Interesting -- there really are quite a few people here talking about completely random, inconsistent focus. I wonder if the AF motor in the lens is very inaccurate on some copies (one hypothesis for why the nifty fifty is so poor at focus is just that: possible a very inaccurate motor/focus element placement).

I forgot to mention that I'm shooting on a 5D III. I stopped even bothering with primes on my 5D II. You could only use the center AF point, properly microadjusted. And focus & recompose at f/1.4 with wide angle primes where you have to move the camera quite a bit to recompose (compared to, say, >85mm focal lengths) just doesn't work. So my higher levels of consistency may be due to the rather outstanding AF performance of the 5D III.

I should also add that in low light I've typically used the 24/1.4 with a 600EX-RT flash attached to my body; the pattern the flash fires most certainly helped the lens autofocus.

By 'beast' I meant it's very hard to tame the 85/1.2. The optimal AFMA even seemed to change on a day-to-day basis; however, it's hard to decouple variables like subject distance & AF point used, so take that comment with a grain of salt (remember: AFMA can change based on subject distance & AF point; Sigma's new USB dock addresses the former by allowing 4 different AFMA values per focal length). The 85/1.2's particular sensitivity to AFMA is not surprising: at f/1.2, the plane of focus is extremely shallow; furthermore, any uncorrected spherical aberration will have a large effect on focus/sharpness at such large apertures. OTOH, given how sharp the lens *can* be wide open, I would think Canon went a long way to correcting spherical aberration. On a good day, the 85/1.2 can focus tack sharp right at f/1.2. But given that often I don't want to gamble, I find myself shooting at f/1.8-f/2.8. Hence I just picked up the 85/1.8, & may be looking to get rid of the 85/1.2 once I do a controlled AF precision test comparing the two (also against a Sigma 85/1.4).

On a related note, I hope Sigma comes out with a 85/1.4 'Art' lens where we can adjust AFMA for 4 different subject distances. OTOH, if Canon's 70D Dual-Pixel AF system makes AFMA irrelevant... that'd be revolutionary. It's an exciting time in photography (always is :) )!

22
Lenses / Re: EF 24mm f/1.4L II USM
« on: July 21, 2013, 04:44:39 AM »
Apebrains,

Sounds like you're doing a good job of some controlled testing. Apologies if I missed it, but, are you choosing the focus point? In order to eliminate yet another variable in your testing, you need to choose a focus point & stick with it. For simplicity, say, the center focus point (different AF points can show different performance &, actually, can require different AFMA values).

Low light definitely stresses AF systems, & can result in lower precision. This lower precision can, in turn, make it hard to choose an appropriate AFMA value. Since AFMA can also depend on subject distance (and, as I already mentioned, AF point), things can get complicated very fast.

In short, though, I'm not surprised with what you're seeing. My f/1.4 & f/1.2 lenses are extremely moody (the 85/1.2 is a beast in this sense), but in general I can find a somewhat optimal AFMA value per lens.

I do wish someone would test AF precision of various lens/body combos, including at different EVs, though I realize it's rather difficult & potentially resource-intensive/prohibitive. I recently was surprised by the poor low-light AF precision of my Sony NEX-6 + Sony 10-18mm f/4 OSS lens (which otherwise performs great in daylight), and that's CDAF, not even PDAF! The Zeiss Touit 12mm, OTOH, focuses with near 100% precision (repeatability) at f/2.8. This is the sort of information that, I feel, could be very valuable to consumers investing in a lens/system (as long as the tests are performed carefully/appropriately).

23
Third Party Manufacturers / Re: Sigma 35 f/1.4 DG HSM First Impressions
« on: December 21, 2012, 03:35:32 AM »
I don't recall anyone nitpicking a lens so much for it's bokeh when everything else is fab, generally the consideration for the quality of bokeh of a lens, especially at these price points and class of lenses is always there, but this must mean though that Sigma just has no other faults to look at, so everyone wants to point at something it's not the absolute best at. And talk about onion highlights? The 35L can produce those onions too!
http://www.flickr.com/photos/pong0814/5596735626/#

And, unlike the Sigma, the 35L doesn't keep rounded circular highlights when stopped down if that matters for anyone.

+1

Very, very well said.

I've always complained about Canon's 8-blade apertures... you start seeing octagons in OOF highlights when you stop down even 2/3 of a stop on many of Canon's primes (e.g. the 'venerable' 85/1.2). What I hate even more is that 8-blade aperture lenses produce 8-point sunstars.

Just adding one blade gives you 18-point sunstars. AND circular OOF highlights.

Nikon's had 9-blade apertures for a while now.

What took Canon so long?

24
Third Party Manufacturers / Re: Sigma 35 f/1.4 DG HSM First Impressions
« on: November 30, 2012, 04:46:36 PM »
Compared to 35L:
http://lcap.tistory.com/entry/Sigma-35mm-f14-vs-Canon-35mm-f14-L

Yes thanks, that's a great review/comparison.

Look at those gorgeous 18-point sunstars & the circular bokeh even at f/2. Canon's 8-blade apertures has been one of my biggest gripes. Nikon has been putting out 9-blade aperture lenses for some time now; Canon's only playing catch-up now. And who knows when they'll replace their 24L, 35L, & 85L lenses...

Also, if you look at the OOF bokeh with pictures of the leaves, it appears to me that either the Canon lens has more contrast or has a lower exposure. Does anyone else see this?

25
Third Party Manufacturers / Re: At least we have Canon quality control
« on: November 29, 2012, 01:06:01 PM »
Zv,

You'll likely be disappointed if you start rigorously testing expensive equipment. You'll start finding flaws that you thought shouldn't exist for such high quality/price products. But I find it's great to know about these limitations so I know how to work around them.

For example, 3 copies of the highly venerated 70-200 f/4L IS I tested all showed erratic softness on one side of the frame relative to the other, even at f/8 sometimes. By f/11 most of the time every shot was acceptable, though not always on at least one copy. Contrast that to the 70-200 f/2.8L II I finally bought b/c I was tired of the poor performance of the f/4L for landscapes. Typically, at 200mm | f/2.8 it's at least as sharp across the entire field as the f4L ever was. At 70mm though, the extreme left side doesn't sharpen up until f/5.6. Which isn't a big deal, and now I store that tidbit of info in the back of my head so when it becomes relevant I'll know how to set my aperture.

If you start testing the AF precision (repeatability) of AF points on any of these high-end dSLRs with primes, you may find yourself pretty surprised. Or not, if you've ever tried shooting a 5D Mark II + 85/1.2 combo anywhere below f/2.8... :)

26
Third Party Manufacturers / Re: Sigma 35 f/1.4 DG HSM First Impressions
« on: November 24, 2012, 06:40:04 PM »
jukka,

Not sure if you saw my previous post, but do you know the answer to this question:

Since AFMA for a particular body/lens combo changes as a function of subject distance, it's not possible to have an AFMA value for all subject distances. Is this any different for a lens that appears to require 0 AFMA? I.e. will it also have 0 AFMA for all subject distances?
hmm My Finnish English do not understand what you're asking, can you simplify the question and I try to answer

Ok, let me try to rephrase.

Let's say you get a 85/1.2 prime lens that needs no microadjustment for perfect focus. Will it focus perfectly for both close & far (infinity) subjects (at f/1.2)?

My 85/1.2 needs a microadjustment of +12 for a subject distance of 25x focal length (~85 inches distance). But now infinity is no longer perfectly in focus.

So what I'm asking is: since the AFMA value you enter into the camera is only really valid for the subject distance you used to determine that AFMA value, is it better to get a lens that appears to need no microadjustment whatsoever with your body?

Or will even those lenses show different back/front-focusing for subjects at various differences?

Of course, I'm only speaking of wide-aperture primes here... Above f/2.8 or f/4 you won't even notice these sorts of variations (unless your lens is really, really far off).

27
Third Party Manufacturers / Re: At least we have Canon quality control
« on: November 23, 2012, 11:35:11 PM »
And speaking of the D800 left AF problem... it evolved into a center AF front-focusing problem (relative to left & right AF points, which at least then agreed w/ one another) for some units sent back for repair.

I wonder if they finally fixed everything. That whole saga was a mess.

My leftmost AF point on my 5D Mark III slightly backfocuses compared to other points, but not drastically so... it's something I can live with.

These are very complex products, with tighter & tighter requirements for tolerance as people put higher demands on these systems. I think proper QC will always be important, but quick & easy testing methodologies for the consumer may be equally or even more beneficial.

28
Third Party Manufacturers / Re: Sigma 35 f/1.4 DG HSM First Impressions
« on: November 23, 2012, 11:27:00 PM »
jukka,

Not sure if you saw my previous post, but do you know the answer to this question:

Since AFMA for a particular body/lens combo changes as a function of subject distance, it's not possible to have an AFMA value for all subject distances. Is this any different for a lens that appears to require 0 AFMA? I.e. will it also have 0 AFMA for all subject distances?

29
Third Party Manufacturers / Re: Sigma 35 f/1.4 DG HSM First Impressions
« on: November 21, 2012, 11:42:34 PM »
Oh, also: 9 blade aperture primes for Canon FTW!! I can finally get my 18-point sunstars :)

30
EOS Bodies / Re: !!!FIRST!!! - Full Frame Mirrorless Camera
« on: November 21, 2012, 11:31:34 PM »
I work at Sony Electronics. :)

Yea we've got FF Mirrorless already.

We're often times the first to market or very early to market. I remember announcing the world's lightest laptop 6 months before the first macbook air. Even after the launch of the AIR our's was still the thinnest and lightest but did anyone know haha? I remember carrying the world's first OLED TV to a convention in Las Vegas. We never advertised it though, the images were stunning and the design was sleek. 2 years later Samsung made a huge TV ad campaign and without saying it made it sound like they were the first to make the OLED TVs. And their TV was IDENTICAL to our aesthetically, they basically took our tv and rebadged it as a samsung. I think that's my biggest frustration with this company. We don't advertise anything and unless you're really looking for a product you would never know we made it. I'm so tired of watching great products get launch with zero visibility.

Just wanted to release some steam about the biggest issue that irks me about Sony. Other than that the company is awesome, they treat us real well, we get good discounts and the work culture here is great (I knew a few people who worked at Samsung that we're very happy).

Hi Minh,

I must say I agree with you re: your advertisement comments. For example:

  • The Sony RX100 has a bounce flash built into it. Really sets it apart from other P&S cameras (save for an Olympus one; I can't remember the model). Why wasn't this advertised? It's not even listed as a feature at all on the official features/specs page!
  • Sony makes the world's best CMOS sensors. Why not flaunt this for the Sony cameras that have this high dynamic range/low noise? Some of Sony's sensors are practically ready for the ISO-less revolution (if it ever comes)!
  • Whatever happened to that really cool 84" projection screen that was matched to reflect the RGB primaries of a paired projector while rejecting all other wavelengths? Such a cool idea... a prototype was demonstrated years ago... never heard anything else about it.

I love the fact that you guys make & push cool new technology horizons. It can often be high risk (RX1, e.g., albeit that's probably a stepping stone to another product coming soon...), but can be highly rewarding. For those of us that care about quality, I must say I, for one, really appreciate the work Sony's doing. Be it Blu-Ray, 4k, SXRD, high DR/low read noise CMOS sensors, 1" sensor pocketable P&S (RX100 was one of Time's top 25 inventions of 2012!!), the list goes on.

Never considered myself a fanboy of any company... but Sony comes close.

Just don't get me started on the subpar lenses Sony's been putting on its HW30/HW50 line of projectors... good grief. Why take a great tech like SXRD & mar it like that!

Pages: 1 [2] 3 4 ... 16