Friday, January 31, 2014

Blue Jays home opener … rip off …

Yeah … turns out they sold out in 48m and anyone who tried probably faced an unending stream of “high demand, no tickets available” …

But “dreamtix” has already listed dozens and dozens of tickets on eBay for 6x the cost …

Nice …

Tuesday, January 28, 2014

Ender’s Game


Impressive, sort of. I enjoyed the movie and was glad that they did their best to deal with the major formative points.

But so much was left out that I cannot help but recommend that you read the book as well. Watch the movie first, else it will be mildly disappointing. Once you’ve seen the movie, the book will blow your mind …

Rogers Internet … slooooooowwwwwww …

I am normally thrilled with my Internet service, which just makes times like this a bazillion times worse!

I pay for 45Mbps and regularly get 65Mbps … so 3.4Mbps is really, really horrible …

Some hours later, things are back to normal ….

Thursday, January 16, 2014

And yet one more L vs M test …. F770EXR, F900EXR, HS50EXR

EXR cameras have an unusual matrix that allows them to pair pixels up to either bin them (this is called SN mode and double the effective pixel area in hardware, yielding better noise handling) or blend them on a curve for extended dynamic range (which is called DR mode, when the bright pixels lean towards mid and shadow tones and the dark pixels lean towards mid and high tones, creating a flat-toned image that retains detail everywhere.)

This is a wonderful trick, but carries a few penalties. One is that the anti-aliasing filter must be tuned for either 1 or 2 pixels (which actually means 1 or 4 since it will be a circle of confusion that is affected) and in choosing which way to go, they will actually soften one mode more than the other (I think.)

What has always been true is that both modes can create a fairly sharp image, but one mode was always better. That mode was M size. Shooting M mode allowed you to retain low ISO (100) while blending two exposures in camera to achieve 2 stops of dynamic range extension. This makes EXR cameras better than any other technology at holding skies and opening shadows at the same time. Theoretically, at least. Some manufacturers have very good software dynamic range extension (basically digital fill) that is hard to distinguish from the Fuji solution. But doing it in hardware has worked well for Fuji.

What happens when you shoot L size, though, is that you are using all the pixels for one exposure, and hence you don’t get your DR extension. And that blows highlights and blocks shadows with abandon. Fuji does, however, have their own version of digital fill for software based dynamic range extension, so all is well. Except that it’s not. Fuji’s version of digital fill boost ISO by 1 stop for every extra stop of dynamic range. So the images look right, but the get noisy fast, which triggers significant damage to low contrast textures and edges from noise reduction. Even when you shoot DR100 at ISO100, you still end up with serious edge issues because the pixels are smaller and the shadows are deeper, thus forcing more noise reduction.

This is especially damaging on the 1/2” sensor that is present in all of Fuji’s EXR cameras with long zooms --- the F series and the HS series. The X series uses shorter zooms and the 2/3” sensor, which works fantastic in JPEG and not so well in Adobe raw processing software like Lightroom and ACR. Silkypix fares better, but I dislike its interface and would choose JPEG over using that tool.

Anyway … the HS50 and F900EXR have had some tuning to improve their handling in the raw converters and to improve the L size handling. Moon shots, for example, look quite good in L size. The craters are high contrast and the extra detail from 16mp does not go amiss – provided that the obvious protocols are followed by shooting many images and selecting the one with the best “seeing.” As for their abilities in shooting fine, low-contrast textures, they are a wash, at least in really good light.

On the other hand, when the details are one smooth curved surfaces and these surfaces are not in direct sunlight, things tend to be less clear. Some parts of the following crops are better on the Msized image and a few are better on the L sized image. What this tells me is that shooting L size continues to be very risky for no gain whatsoever. (Note that there are tests out there that show spectacular differences in favour of L size. These make no sense to me so I cannot put any faith in them, especially given that the two people pushing L size are also prone to extreme rhetoric.)

So I thought I would run one more test with my F770EXR, which is featured in the link above the previous paragraph as well. It tends to be pretty even in well balanced light and the edge destruction in shadow is there, but not rampant. Still, I would never shoot it in L size for the simple reason that there is never any real advantage and there is often palpable disadvantages.

But what about bad light? Orange, indoor light from halogen bulbs for example? What does that do? Well, I would predict that L size would take a major beating from the combination of pushing the blue channel very hard and the shrinking of pixel sizes into the range of noise grain, triggering a beating from the noise reduction. That’s pretty much what always happens in shadows, and indoor light is low enough to be like mid to shadow all the time.

So I shot JPEG+RAW at 100 ISO with M4:3 at DR400 and L4:3 at DR100. Since both are at their best (no raised ISO) this is a fair fight. I processed the JPEG just to resize the M with a hint of sharpening and add watermarks. The names reflect which is which, but you will be able to tell easily … the M version has a flatter tone curve as it should. The textures on the L size are largely destroyed, as is normal on any but the latest EXR cameras (and I would dearly love to get access to those cameras one more time to run this test, as bad light seems to aggravate the already poor handling of textures.)

It is easy to think that maybe the images are not focused the same, but since they were shot on tripod and released by timer, that’s not actually an issue. Further, the hard edged lines are almost identical on most of the image (the exception being the cross on the right, which is in shadow and is crushed by NR I believe in the L sized image.) I have created two animated GIFs to illustrate the brightly lit area top left corner where they are essentially identical after upsizing the M, and the main, which is in shadow and very dark … here, the texture destruction is absolutely rampant in the L sized image and almost non-existent in the M sized image.

So … the four images … download as you wish, but I am exerting my copyright on these that you must give credit if you repost the images and you may not post derivative works that are designed to twist the results. That’s why you are getting originals that are only resized to match and rendered in Lightroom to fit inside the 4MB limit of my Gallery host.


M4:3 Upsized

Crops Top Left Pattern

The edges are a tad smoother on the L sized image, but there is no difference in detail. The contrast is better on the M image, likely owing to the less harsh tone curve (not intuitive, I know … but I think it has to do with the overall image contrast and not the local contrast.)

Crops Donkey’s Main

Here, the difference is night and day. I have always found that the noise reduction can get completely out of control in areas where there is little light. It just cannot tell detail from noise and the smoothing is applied with a sledgehammer.


The newer models handle L size much better than this model. But there are still issues as shown in the two links above. And if you shoot anything but the HS50EXR and F900EXR, you really don’t want to mess with L size very often. It does very bad things to the files … in JPEG and in RAW.

Even with the newest models … I would not bother except in very limited circumstances. Very close macro might work, if there is no texture at play. And the moon tends to work because we don’t really care about the part of the surface that has no craters. And even then, you have to carefully process that part to avoid seeing pure mush.

So … enjoy. Shoot however you like. You have been warned Smile

Sunday, January 12, 2014

Extreme (as in insane) ISO – 12,800 and 25,600 – Panasonic G5, G6, GH3, GX7, GM1 vs Olympus E-M1 vs Nikon D4 – **UPDATED July 2016**

Update 3: Allow me to open with a caveat. This article is critical of some great cameras over a truly esoteric shooting scenario. Most people will *never* encounter such a scenario with one of these bodies. To all intents and purposes, every one of these bodies would satisfy all but the most demanding photographer. The trick is to choose the features and budget you want, not pick apart trivial IQ differences. All that said, let’s pick apart some trivial IQ differences, shall we? :-)

So why in the world would anyone want to shoot a Panasonic micro 4/3 body at such rarefied ISOs? Well, I might actually consider it, but the real reason I want to compare these bodies is that I am looking for a new body to assuage a case of GAS (as in gear-acquisition syndrome.) I want the next one to have full manual video to match my GH2 and I want it to shoot at the highest ISOs with clean shadows. I don’t mind grain as I’ve spent the better part of a decade learning how to deal with noise.

So this article is my way of – to use a politically incorrect expression – “separating the men from the boys” ….

I’ve run these comparisons many times and am able to get a good result from any and all the 16mp sensors at 3200 and even 6400 ISO. They pretty much hold their own against one another and against APS-C cameras nowadays, so what is left to test is how they have progressed in shooting extreme ISOs. In other words, where do they break down?

[Update 2&5: Clarification: I really wanted to know which body handles high ISO with the fewest blue channel issues in shadow and the best grain. The E-M1 has the best as in finest grain. The GX7 has the fewest blue channel issues without caveats, and the GM1 has the fewest blue channel issues overall – but you have to watch the eshutter limitations that can lead to banding under artificial light.]

[Update 5: In the end, I bought the GM1 and have not regretted it. Stunning image quality and a fun little camera to carry about. I rarely see the intrusion of the banding.]

I own the G5, which has the GH2 sensor with some modifications. It has better dynamic range as one example of the improvements Panasonic were able to make. It also handles shadows better than the GH2, which turns shadows blue when shooting in bad light (as in orange light that forces significant pushing of the blue channel to achieve accurate white balance once again.)

[Update 5: I sold the GH2 and the G5 and acquired the G6. I love this thing. Absolutely great image quality and ergonomics. The video is better than a stock GH2 and of course it has manual control so is better than the G5 in that way. My one regret is not keeping the G5 so I could carry both the 14-140 and 100-300 when walking in the woods. Sigh … now I am waiting for the G7 to appear to complete that pairing once again. Live and learn …]

Blue channel noise explained: I’ve been writing about this issue for years – where the blue channel goes out of control at high ISO when shooting in yellow / orange light such as incandescent and then adjusting white balance back to neutral. As pixel size falls (physical pixel size, which is caused by increasing pixel density or shrinking sensor size or both), there is a “threshold effect” where you suddenly see shadows turning blue in your raw image editor. JPEG engines sometimes handle this well and sometimes do not.

I first noted this issue on the Fuji F50fd, which doubled pixel count from 6 to 12mp on a 1/1.7” sensor. At 1600 iso the sensor went nuts and the JPEG engine could not compensate. Now, we have 16mp on a 1/2” sensor from Fuji and the JPEG engine has little problem with it. I see this “blue channel poison” in raw, though, but it is usually easy enough to deal with (tinting shadows is one approach for example.) So this article answers the question “where is the threshold for m4/3 bodies?

So how do we explain this “blue channel poison?” Well, I don’t know the physics, but I know what is happening so I’ll lay out my theory and perhaps someone with deep knowledge can help me correct it for this article. (Deep knowledge means really deep. Not another lay-opinion :-)

A sensor has a base sensitivity. Some have the equivalent sensitivity of 100 ISO and some have it at 200 ISO. The Sony sensors used by Nikon, Olympus and now Panasonic, for example, have 200 ISO as their base. What this means is that when you shoot at a higher ISO setting, all you are really doing is applying analog amplification to increase the charge that is based on the photon count. When you raise ISO, what you are really doing is shortening the shutter speed and then compensating with amplification. That’s the whole point – to allow you to shoot people in low light. Else, just put the camera on a tripod and shoot long shutter speeds at low ISO to get clean images. But that just does not work for people, and certainly not at an event.

So … you have a capture and it is amplified. Shadows are by definition places with few real photons. Now increase the shutter speed by 5 stops and you have 2 to the fifth times less light, as in 1/32 as much light. And under incandescent lighting, there is little information in the blue channel, although there is always some noise. Now apply the 5 stops of amplification and what do you get? Well, the noisier the blue channel, the more blue channel poison shows up in shadows. In other words, the noisier the sensor and the smaller the physical pixels, the worse the effect.

Now … someone tell me that this is wrong and explain what is really going on … please :-)

I have always liked the G5 at 3200 and tolerated it at 6400. It does shoot 12800, so I am adding it to this comparison as a benchmark for the G6, which is the latest camera to carry the GH2’s sensor, presumably with the same or better tweaks (but read on to find some nasty surprises lurking …)

[Update 5: I must say that the G6 does not test as well as the G5 at the highest ISOs … at least on the samples available on … but I have to say that using the G6 is actually a more pleasant experience than the G5 … at least slightly. Perhaps it is because it feels that little bit more like a very small dSLR.]

I went to one of my favourite sources for low light comparison images. That being, which is one of those sites that does its best to equalize its images across cameras. It also shoots raw images at all ISOs so that we can run our own comparisons. And that is exactly what I am doing here. I am publishing only derivative works, so please go to the original site to see the JPEGs. They all look universally disgusting to me, so I don’t even bother. But raw is a very different story when processing for high ISO. Especially for those who are not grain-adverse (in my opinion, grain-aversion is a crippling disorder for a photographer.)

So the methodology is pretty simple -- process these two extreme ISO images from raw in Lightroom 5 for all the cameras to make the camera’s individual images look as good as possible. This requires that noise reduction be balanced against grain. The goal here is to publish web pages that look good and not smeared. For the transition from 12800 to 25600, I decided to use identical settings so that we can see the exact nature of the degradation at 25600. Is there a threshold in between or not? I equalized the white balance across the Panasonic bodies (2800, +5) and the E-M1 set its own white balance which was 2750 … almost identical. I added contrast and lifted shadows a lot for the G5 and G6, but chose to not do that for the newer sensors as they looked great out of the box, so to speak.

I will not show you the actual images here, as that would be pushing the boundaries of “fair use for educational and commentary purposes” but the images are easily found at Imaging Resource. To start with, I will show you one small thumbnail of a JPEG linked to the D4’s ISO 12,800 image. This is what that image looks like, click on it to proceed to the original on the Imaging Resource site:

D4INBI012800_NIKON D4_70 mm_ISO 12800_1-400 sec at f - 4.0

So the test is to download and process the raw images at 12,800 and 25,600 ISO and to use identical settings for both images from the same camera. Each camera is tweaked to get the best from it, and an animated gif shows you the differences. This tells us whether we have reached the blue channel poison threshold or not. You will be a bit surprised in some cases, and not so much in others. I will also show head shots as cropped from the images, which is something people do a lot of. This simulated a medium sized enlargement and will give you a good idea of how well they hold up at extreme ISO when printed larger than Facebook requires.

So let’s begin …

Nikon D4

This is a legendary full frame camera that is the follow-on to the wildly successful D3s. It is able to shoot as high as 204000 ISO or thereabouts, which is truly insane. I include it as the control for this group. This is what they should all look like if there was no degradation as sensor area and / or pixel size falls.

Starting with the head shots … 12,800 ISO then 25,600 ISO

That’s shockingly good when you consider where we were a decade ago in digital photography. If you printed this head shot at 8x10, I’m betting that ink bleed would take care of the grain and create a very nice looking print.

Let’s look at the animation and see how much blue channel poison might be present. The place to look is under the arm, where there are some fairly deep shadows.

There is no blue channel poison to speak of. A slight drop in brightness is all that shows up. Remember, the processing is identical between 12800 and 25600 for each camera.

Panasonic G5 and G6

As mentioned already, I own the G5. I love this camera, as it is small and light and yet has Nikon-like ergonomics but with even better controls than Nikon’s entry dSLRs owing to the liberal use of the touch screen and the quick menu. It also has the electronic shutter, a really useful innovation for those who want the crispest of images when shooting still subjects.

So when I saw the specs of the G6, a almost fainted. The same (I thought at the time) sensor updates as the G5 plus the addition of manual video, wireless control and focus peaking. What is not to love? Well, we’ll find out in a moment …

The G5 cannot shoot at 25,600, so I will show three head shots for the pair and two animations. The first animation will be the G5 and G6 juxtaposed at 12,800 ISO. The second will be the standard pair for the G6.

Head shots first. The G5 at 12,800 followed by the G6 as 12,800 and then 25,600.


A bit grainy, but mainly because I opened the shadows more than I should have. Still, on paper it would look decent. Very slight blue channel poison at the bottom of the hair, but easily addressed I think. And now the G6 pair.

Whaaaaaaaaat? They are processed at the same white balance, yet I am seeing a colder presentation and more blue channel poison than the G5 shows. This is a surprise to me, as I was sure that Panasonic would improve the sensor yet once more. The 25,600 tells the tale …

Oh … my …. god …….

Ok, the G6 has (very painfully) been eliminated from contention for my next body. Note: This is not meant to be a general rule for anyone else. It is still a wonderful camera with amazing features for the money. Seriously. I bought one. But it is not going to take me to the desired high ISO nirvana any more than my G5 already can. In fact, I suspect that I would be annoyed the very first time I tried to process indoor images from my mother’s house, which has some of the most orange lighting I have ever seen (old compact fluorescent bulbs in a dark room.)

[Update 2: People have suggested I should just adopt full frame. Well, been there and done that. Not interested at this time in buying back into the back breaking cameras just for high ISO. I am very happy with m4/3 and plan on staying that way. Sensors get better every day and I’ve been building my noise handling skills for most of a decade, so m4/3 it will be.]

So let’s seal the deal on the G6 … painfully, I might add.

Yup … the G5 has its threshold right around 12,800 ISO … and the G6 seems to be lower than that. By the time the G6 reaches 25,600 it is throwing blue channel poison wildly about like a drunken sensor. Needless to say, Panasonic added the extra ISO setting for marketing purposes … (duh.)

[Update 5: I own the GM1 and the G6 and I shoot both in low light. But I would happily shoot the GM1 at 12800 while I prefer to keep the G6 at 5000 or lower. I will shoot 6400 if there is no other choice and it is still pretty competent up there if you are careful.]

So … moving on to newer sensors.

Panasonic GH3

The GH3 has appealed to me from day 1 as a true dSLR replacement. If any camera can be “the one” this one should be it. After all, Panasonic apparently caved in from the pressure from Olympus and adopted a variation of the Sony sensor that debuted in the OM-D E-M5. So let’s see how they have done with this sensor.

Well, the 12,800 ISO image is quite usable. Nicer than the G5’s image for sure. But the 25,600 is unfortunately quite poor. The grain might be tolerated, but the blue hair most assuredly would not. Still, 12,800 is quite good and I suspect you could easily get away with shooting at around 10,000 ISO in bad light. So a good cam for stills. And we know how spectacular it is for video.

[Update 3: I should note that I processed the E-PM2 and E-PL5 after the fact because those bodies are at excellent prices right now. With their sensors being newer derivatives of the E-M5 sensor, they should be amazing, yes? Well, no. Lots of blue, just as much as the GH3. I think it is obvious that they all come from the same generation of Sony sensor and it simply does that. The improvements we see in the GX7 and the GM1 come from whatever it is that Panny has done since. There is much less improvement with the blue channel issues in the E-M1, which frankly surprises me. I smell a potential Nikon / Canon situation developing … I hope Oly does not try the Canon trick of resting on its laurels. It took the lead with the E-M5 and should definitely try to push that in its sensors, not just build and features. Just my $0.02 and worth about that much, I know :-)]

Completing the picture …

Well, for Facebook and the like it might suffice. The blue channel noise is less obtrusive when not enlarged. And don’t forget that shadows can be individually tinted in Lightroom / ACR so this might be salvageable … but I think I would restrict my shooting to 10,000 and under as I already mentioned.

Panasonic GX7

So how about the GX7? This is the latest fully featured camera … with one tragic exception: while the G6 was given a microphone input, the GX7 was not. What were Panasonic thinking? Answer: thinking was apparently no involved in that decision.

Anyway … let’s have a look.

Hmmm … I see nothing to really worry about. There is a lot of grain at 25,600 ISO, but what do you expect? I have not even tried to process these through something like the brilliant Topaz Denoise 5 yet. That would look more like …


GX7INBI25600-Edit_DMC-GX7_50 mm_ISO 25600_1-640 sec at f - 4.0

On paper, I think you could get away with that. And putting some real effort could very well make a go of it … but the animation does show that there is a very slight issue in shadows. Not bad, mind you, just something to keep an eye on.

So it is quite clear to me that this camera can shoot pretty high up there. It is less prone to the blue channel poison than the GH3 by a long way …

Panasonic GM1

Ok … there is one newer sensor in the Panasonic lineup and it has become the darling of the video crowd (or at least some of it) and the stills crowd. It shoots magnificent images. The head shots are very promising … I’ve even left the shadows more open than on any of the others except the G5/G6 … and it looks mostly marvelous.

And at 25,600 …

Wow … lips still pure red, virtually no blue channel effects … this might be the winner. Certainly, cropped out you cannot see any effects of the use of the eshutter here.

But when you juxtapose the two exposures full size, you do see what amounts to fairly strong banding. This basically means that you have to manage your shutter speeds to stay with the mechanical shutter when shooting under any artificial light.

[Update 5: The eshutter is sensitive to the strobing of artificial lighting at high shutter speeds (this is the only eshutter than can even operate at ISOs above 3200) and so we have a collision of limitations here. Note that all eshutters have this issue, so it is not really limited to the GM1, it’s just that the GM1 forces eshutter after 1/500s while the others do not.]

[Update 4: This collision of limitations will probably affect few people, if any. I want the next body to have a lot of high ISO flexibility and so I would like to avoid such limitations. But for most people, I suspect that the GM1 will be an extremely satisfying camera when shooting in almost any light. Just don’t leave the eshutter engaged all the time.]

Olympus OM-D E-M1

No test would be complete without representation from the Empire, er, flagship of the entire m4/3 fleet. That being the Olympus OM-D E-M1. This is a beautiful camera with an incredible feature set and a jaw dropping price. But you get a lot for your money and I would love to own one. But for now I will not pay that kind of money for anything but full frame, and I will not spend that kind of money for a body with mediocre video.

So … the head shots.

Yup … it’s the winner. No problem. Although there may be a caveat to that … hold on.

Oops … there is more blue channel poison that I expected. Under the arm, the focus target, the red flowers. Hmmm … the threshold is actually lower for this camera than for the GX7 and GM1. Of course, the GM1 has a fatal flaw, so that leaves the GX7 as the overall winner. Grain I can deal with, but this funky blue channel issue is a huge pain.


The GX7 and GM1 are the most resistant to the blue channel effect, followed by the E-M1 which does show the effect. The GM1 is flawed for my desired use, but I still expect to pick one up eventually. I dislike the compromises they have made, but all that really means is that I will use another body for the more serious work or at least for any work that requires a lot of use in dim artificial light.

Meanwhile, the E-M1 is the easiest to process for grain – very similar to the best APS-C cameras with their finer grain – and should be on anyone’s list if money is no object and video is not all that important. But if video really matters, the Panasonics are the obvious choice.

My official pecking order for indoor shooting (incandescent light) … for now:

For stills only up to 12,800 ISO: E-M1, GX7, GM1**, GH3
For stills up to 12,800 ISO and some video: GX7, GH3, GM1**
For video and stills at any ISO: GH3, GX7, GM1**
For stills above 12,800: GM1**, GX7, E-M1, GH3
    **GM1 requires careful management of the eshutter issue to avoid engaging it under artificial light
And there you have it. I hope you enjoyed this trip into lala land Smile

[Update 5: Since writing this article, I have purchased the GM1 and the G6. They are both a joy to shoot and the limitations each has – eshutter after 1/500s and blue channel issues at extreme ISO respectively – do not affect either very much at all in day to day shooting. So do not shut them out just because they have these issues. Shoot around them. If you really want to shoot good video, then these two are very good as well-balanced shooting machines.]

Update: 10 December 2015

Here we are almost 2 years later and I can say that the G6, GM1 and E-PM2 are still a joy to shoot. I carry the GM1 and E-PM2 in a tiny Tamrac case when I want ultra-portability. I carry the G6 with the 14-140 when I want a good all-rounder in a portable package and with the 100-300 when I walk in the woods. I picked up a used D90 in order to better shoot tethered capture to Lightroom (Panasonic and Olympus apparently don't get it for some reason) as I like to shoot studio for charity these days, but I often tuck the GM1 and E-PM2 into a corner of the bag. The state of the art in sensors really has not changed much lately. The OM series is no better than it was for stills, although video sucks a bit less these days. The G7 has wonderful 4K video, but that does not exactly float my boat. All in all, the universe has stabilized in a good place.

Update: 07 July 2016

Well, I've gone and done it. I bought the G7 and then sold off the GM1, GF3 and E-PM2. Two of those had magnificent sensors, but I found that I was no longer all that interested in carrying them around as multiple bodies with attached lenses is more of a hassle than I thought it would be. I have also come to realise that I trust and still lve the output from the 14-140 and the 100-300. And the DFD auto focus of the G7 is something to behold. So I will be adding the GX85 to the G6 and G7 and I might even sell off the D90 and its lenses to finance the 35-100, which would complete my studio setup perfectly. With DFD in two of my three bodies, I really don't need a dSLR for AF performance. Of course, every time I pick up that camera, I remember how much I like the feel and performance of a dSLR, so I might be blowing smoke right now :-) ... Since the sensors have improved slightly, I am pretty much satisfied with anything that the G7 and soon the GX85 can do. If I want to do extreme low light in the future, I will grab a used D610 or equivalent. Meanwhile, these Panasonic bodies handle everything I could ask for.