October 22, 2014, 11:11:43 PM

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 - helpful

Pages: 1 ... 9 10 [11] 12 13 ... 15
151
EOS Bodies - For Stills / Re: 5D3 Dynamic Range
« on: May 04, 2012, 07:16:33 PM »
Shooting JPEG you will never see a difference

False.
How many stops of dynamic range do you get in jpeg that is different from antone else? Jpeg is 8 bit no matter how many stops you get in raw

For a linear DR it is a DR of 8 - which is why specialist printers are needed to get more.

I hate to say anything critical to anyone, and you are both correct. JPEG has 8 bits of data per channel. And a statement involving the word "it" can always be assumed true--we have no idea about the object to which "it" is referring.

However, you are saying something correct and then jumping to a completely unrelated conclusion. Here's the missing information:

* JPEG is not encoded with linear gamma, and never has been with any digital camera.

"If images are not gamma encoded, they allocate too many bits or too much bandwidth to highlights that humans cannot differentiate, and too few bits/bandwidth to shadow values that humans are sensitive to and would require more bits/bandwidth to maintain the same visual quality."

* One bit and one stop have nothing to do with each other. It's also completely ridiculous to define dynamic range in terms of bits--it is only meaningful if defined in terms of stops.

If you have ever taken a class about encoding mechanisms, one of the first things one learns is, "The first thing to remember is that bit depth and dynamic range are NOT the same thing. It is going to sound much the same, but it's not."

The terminology used on this forum is sometimes as silly as someone saying, "I am going to itch my mosquito bite" when they mean "scratch my mosquito bite."

If you have ever taken a picture with part of it dark or part of it bright, you are seeing the effect of dynamic range. It has absolutely nothing to do with the file format. RAW extends the dynamic range provided that the same gamma curve is applied in the image encoding.

An increased dynamic range in the camera has a proportional effect on the dynamic range of both the JPEG and the RAW image. You can understand this if you will be patient enough to consider the example of an interval of real numbers.

The RAW image corresponds to the interval [-x, x]. The JPEG image corresponds to the interval [-cx, cx], where c is between 0 and 1. A change in the dynamic range corresponds to a change in x. The effects on both the RAW and JPEG images are proportional to one another.

One good article to read is this one:

http://theory.uchicago.edu/~ejm/pix/20d/tests/noise/

It also talks about noise in addition to dynamic range. There are books to read about this subject as well.

152
EOS Bodies - For Stills / Re: 5D3 Dynamic Range
« on: May 04, 2012, 05:40:21 PM »
Shooting JPEG you will never see a difference

False.

153
EOS Bodies - For Stills / Re: 5D3 Dynamic Range
« on: May 04, 2012, 05:39:39 PM »
A really high DR better than the 5D3 doesn't really help. As I have explained in previous posts, a lower DR actually stores more data and detail from a scene than a camera with high DR. Ideally the dynamic range would match the scene's DR. Canon's DR probably fits more scenes better than Nikon's. If the dynamic range is higher than the dynamic range intrinsic to the scene, then it actually makes the picture worse (less fine variations in detail of recorded luminosity).

The only part of the 14-bit RAW or 8-bit JPEG data that is really worthwhile is the part where the histogram shows data has been recorded. In a low dynamic range image (like a frame filled with nothing but green grass), the histogram of a high DR camera like both the 5D3 and the D800 show nothing but a thin peak of data that was recorded. This means lots of detail is being lost because not all 14-bits are being used.

Don't get me wrong. I'm not saying that there is "empty" data in most of the picture. There is a color and brightness value at every pixel. But the variations in brightness that could be recorded if all 14-bits were adjusted to a 2-stop DR are simply lost, because the camera is always trying to record a 9 stop or 10 stop DR, for example. The variations in brightness can be recorded much finer, just like with slide film, with a good sensor that has LESS dynamic range.

High or low dynamic range can't be held against a camera, any more than someone can say negative film is "better" than slide film, or vice versa.

154
Lenses / Re: Canon 85 1.8 vs. Sigma 85 1.4
« on: May 03, 2012, 06:31:06 PM »
The Canon is good for small prints and focuses like the wind--in the highest class of fast focus. But it's not useful as a picture-making machine below f/2.8 due to the low sharpness, low contrast, not to mention the purple fringing (I don't really care about that, since it is for the out-of-focus areas--I just care that the in focus areas are extremely sharp, and they are not sharp below f/2.8 ).

There are a lot of inaccurate reviews of the 85mm f/1.8 saying it is sharp, but they're using it in a variety of circumstances shooting on automatic and most of their review photos are f/2.8 or above, lots of f/5.6 shots, etc. For those the 85mm lens is world class. For this reason I own two copies, and two copies of the similar 100mm f/2.0 as well. The same comments apply to the 100mm f/2.0 lens, despite the fact that Ken Rockwell says the 100mm f/2.0 has 5.0/5.0 perfect optics. He is talking about photos that are not near f/2.0.

But when I need those apertures below f/2.8, then the Canon f/1.8 is simply not an option. Occasionally a picture is good enough at f/2.0 that I think there is hope, but it just isn't possible to call it sharp. The fast focusing sometimes makes up for it. To be fair, I made a 20x30" enlargement for someone two weeks ago that was shot at f/2.0 with the Canon f/1.8 lens. It can be done, but it's not pretty.

The Sigma on the other hand is three steps backwards in terms of focusing speed, but that still puts it far ahead of the 85mm f/1.2L in that area. The one problem that I hate about my copy of the Sigma (I've had it for 1.5 years and shot hundreds of events with it) is that it just doesn't do focus tracking properly. It focuses plenty fast enough even to take basketball photos, but it doesn't track movement after focusing, but waits for a moment. So I have to keep my finger off the shutter, and then push it at just the right instant so that when it is locking on it will actually track the motion in order to lock on, and then I get a perfectly focused shot.

It has no trouble locking on to a moving object, but it would be a lot easier if it would keep on tracking it so I didn't keep having to lift my finger up and down. I think there is something messed up with the algorithm in my first lens. I have another one arriving this week (20% off from Amazon with the purchase of a Canon body that I ordered) and I am hoping that it will be different.

Image quality from the Sigma f/1.4 is unbelievably good. That's all there is to say. Some tests have shown that it is better at f/1.4 than the Canon 50mm f/1.4 lens is at f/5.6.

In theory that's possible since the effect of diffraction is much more negligible at f/1.4, but in lens manufacturing it's pretty hard to make a diffraction limited lens faster than f/5.6. (Diffraction limited means that the lens design is so good that the only factor limiting resolution is diffraction. If this were the case with an f/1.4 lens, then it would have resolution 16 times higher than an f/5.6 lens.)

155
Lenses / Re: Trouble with my Canon 24.1.4 ii lens at 1.4
« on: May 02, 2012, 06:40:00 PM »
Agreed. Focus and recompose is a good technique for point-and-shoots or cameras with a single focus point AND a relatively small aperture (big f/number, to clarify for anyone who thinks aperture means f/number and is confused).

But it won't work with a shallow depth of field lens because the plane of focus is designed to be a plane so that the lens will focus perfectly on a flat wall (the famous "brick wall" lens test).

* When you focus and recompose the distance remains the same, so an uncorrected lens would still be in focus.
* But lenses are "corrected" to keep a flat subject in focus, and so when you recompose the shot, the plane of focus is rotated at the same time. The plane of focus doesn't intersect the subject anymore after recomposing the shot. Rotation of a plane shifts one half-plane behind its original distance, and the other half-plane closer than its original distance. Since the subject is at the same distance, things clearly won't line up anymore.

156
Lenses / Re: Trouble with my Canon 24.1.4 ii lens at 1.4
« on: May 02, 2012, 02:17:37 PM »
Great photos on your website, Louis. This one is strikes me as better than many famous versions of Yosemite Falls:

http://www.louisamore.com/landscape/2.jpg

It looks like the ground at the bottom edge is "earth" and the falls are pouring from "heaven."

157
Lenses / Re: Trouble with my Canon 24.1.4 ii lens at 1.4
« on: May 02, 2012, 02:06:17 PM »
Louis, I know exactly what you are talking about. Autofocus microadjustment will not reliably fix your problem, although that would be a great option for the 5D Mark III.

Even using the center point will not. The 5D Mark II has a problem with this particular lens in my experience. What will work is using lifeview autofocus (very slow) but that will nail the focus on still objects that the focus selector is centered on.

The lens itself has extremely sensitive and accurate autofocus, but the 5D2's phase-detection autofocus just rushes to an approximate focus  point and then leaves it there despite a slight mis-focus. Focusing on infinity and trying again will result in another slightly random focus shift. That's why you are needing to take a whole bunch of pictures in hopes of probably getting one that is sharp on the desired focal point.

Update: By the way, the same lens works fine with the 5D3 and my 7D cameras.

158
Sounds to me like the emperor who had no clothes! If such a kind of glass really existed, would it be able to refract light? Anti-fogging, self-cleaning, etc., seems like it is just a mist with no refractive properties. Really, I don't know why I'm responding to this post... just happened to pop up in front of me at the wrong time. :)

Like the second poster mentioned, it may be impossible to make a lens out of it. If it really exists, then perhaps it would be a great window glass, to shoot photos through. But it might not refract sufficiently to actually make a lens out of it.

159
Lenses / Re: Seeking lens in the 85-100mm range
« on: April 30, 2012, 07:55:36 PM »
There is only one choice for you in my opinion, given the prime lens consideration and the focal length range. That is the Sigma f/1.4 85mm. It is astonishing, really--more useful than the Canon f/1.2L. I actually ordered a second copy today. It's fast enough to shoot basketball, and by fast, I mean both the f-number and the focusing speed and accuracy. My first copy did need considerable microfocus adjustment, however. Afterwards it is super sharp across the board; er., I mean, across the frame.


160
Lenses / Re: Help with event lens ( weddings, anniversary's, etc)
« on: April 30, 2012, 03:17:41 PM »
I agree that the 16-35mm f/2.8 II is a great option. At first the wide end of it is too tempting, and the shots one tends to take are not as good as someone with more maturity using a super wide angle 16mm focal length (just speaking from my own mistakes). Now I tend not to use the 16mm end of it as much, but for close spaces and uses for which wide angle is appropriate, it's a life-saver. An important group of people might happen to line up where there is no space and then 16mm is needed. But I prefer the range from 24-35mm.

For that matter I have been finding tremendous success with the 24mm f/1.4L II. I believe that using prime lenses helps me correlate the photo's perspective better with the human experience at the event. Humans don't have zoom lenses for their eyes; rather, they make small adjustments in position or distance to perceive things. A prime lens leads the photographer's style to make the same adjustments unconsciously, while a zoom lens does not.

So if I have the ability to set up the shots, like the pre-wedding photos, then I like the 24mm f/1.4L II, and I am beginning to be completely comfortable leaving it on for the entire wedding.

Note that I would never risk taking a critical shot at f/1.4 as mentioned earlier in this thread, unless I had time to carefully use contrast detection autofocus. At f/1.4 and as soon as a picture is cropped or enlarged, it does matter exactly whose face is being focused on, and it's hard to tell which face the camera chose at the instant the picture is snapped.

161
EOS Bodies / Re: Shoot JPEG again with 5D3
« on: April 26, 2012, 11:56:43 AM »
The RAW v JPEG debate really should be nothing more than a workflow question, and workflow questions are entirely related to the type of photography you're doing.

If you're shooting landscapes in the spirit of Ansel Adams, you'd be a fool to not shoot in RAW. You're going to be twisting and stretching that image in so many ways that you'll need all the data you can get, and the time it takes the converter to do its thing isn't even a blip on the radar.

If you're shooting the Super Bowl for Sports Illustrated, you'll be unemployed in a heartbeat if you don't shoot JPEG. You're not going to be doing any post-processing at all. If your shots need post-processing (other than cropping), it's because you messed up the exposure; make that newbie mistrake again and you're fired. And the editors don't want RAW; they want JPEGs, and they don't have time to run everything through a batch converter -- they need those JPEGs NOW.

Almost makes you wonder if, maybe, perhaps, possibly, there's a reason why the cameras support both formats and have so many options for picking which to record....

Cheers,

b&

I like your viewpoint!

162
EOS Bodies / Re: Shoot JPEG again with 5D3
« on: April 26, 2012, 11:47:15 AM »
Thanks for the multiple replies!
@helpful: I really appreciate your post! I like strong convictions and find your tone very convincing. Unfortunately, I don't entirely understand your advice: how comes that you change white balance in seconds? Or tenths of seconds, exposure settings? Do you manually adjust K on white balance? How exactly do you proceed?

I use a pure white target (white is really a reflection of an equal mixture of the three wavelengths of light) which is not pointed directly at the light source, but between a 45 to 90 degree angle. I try to duplicate the angle of someone's face relative to the light source. I can just use a white paper for simplicity. The side of the paper in the shade usually has a different color cast to it, so go with the side that is most representative of the lighting on your subjects (whether they are people in a room, people on a basketball court, etc.)

I make sure that the RGB histogram is enabled and change the exposure settings to properly expose the paper at a shutter speed 1/10th or below if I am indoors, except under incandescent lighting which does not have color shifts during the alternating current cycle. In natural light (e.g., daytime outdoors) or in 100% incandescent lighting, the shutter speed is irrelevant. I set the custom white balance to the paper, and check the RGB histogram to make sure all color channels are getting the same exposure. I repeat if necessary.

Adjusting Kelvin lets you change the receptivity to the wavelengths of light all at once, but it is not exact as taking a photo of a white subject. White contains an equal amount of all three color channels, red/blue/green, and so it allows all three color channels to be perfectly calibrated to the light source, with a single exposure.

I agree that most of the Av shots have strange exposures. How can I learn the correct exposure / aperture ? I'd be still relying on the Ettl metering to check manual settings. Well, I'll try harder to practice manual mode! Now I am tempted.

During the time when you were setting white balance, you can also determine the proper exposure value for the lighting conditions at the same time. If you know the brightness of the paper, you can use it in place of the neutral gray card (18% usually, but there are arguments about that.)

Your LCD can be a very bad idea of the exposure, especially if it is set to automatic brightness adjustment. (If you get your finger over the sensor below the LCD on the 5D3, for instance, the LCD will go very dim even though your picture's exposure is unchanged). But your histogram is an accurate indication of exposure.

You can look at the RGB histogram of the white paper shot (after setting the white balance) in order to set your exposure. (And this time, just use the same shutter speed and ISO that you expect to use for the action you plan to be shooting). Take a picture of the paper when it is directly facing the light source and you should see a thin high peak for each color channel. Change the exposure (I am assuming you are using manual exposure to be able to control aperture, shutter speed, and ISO) so that the peak for each color channel is about 10% below the right of the histogram. Your paper should be directly reflecting ("aimed at") the light source under the same lighting conditions that your subjects will be photographed in.

Note that there are often different types of light in big stadiums (i.e., some pinker lights, some bluer) and if you are at the wrong spot, sometimes only a row of blue lights are hitting the subjects, for example. And some lights in bad high schools are actually missing parts of the color spectrum. In those cases, proper colors are just about impossible, but it doesn't have anything to do with what format of image you are recording.

I am not trying to bash RAWs. They are a good way to save all the camera's information that was used when the shot was recorded and processed to JPEG, just in case any of that is needed. I will "turn them on" whenever I feel like it. But they only give 6 more bits of data, and a tiny exposure error (just less than one stop) will neutralize any benefit that the RAW may provide.


The reason why I started this thread is that with this camera just like other posters found my jpg look just like postprocessed RAWs already ! I use Aperture plus NIK suite. Could not sharpen better, could not impove the colours, could not significantly improve contrast, so I started putting RAW on CF and jpg on SD but ended up scratching my head. Well, try yourself. It's still early days with the mark III, and I don't state that it's only jpg worth, just that the jpg machine is good and tweaking the incamera settings on contrast, sharpness etc yields a good time saving for non-critical work.

163
EOS Bodies / Re: Shoot JPEG again with 5D3
« on: April 26, 2012, 11:23:08 AM »
With the 5DIII I have been shooting RAW to the CF and Large JPG to the SD, a luxury I really appreciate.  In nearly all of the sports and people captures I have taken, the JPGs look terrific.  So much so that I have ended up spending a bunch of hours in LR 4.1 just trying to get the RAWs to look as good.  Then staring at my computer screen, I am in disbelief that I can't get them to look noticeably better.   I am the first to admit I am low on the LR learning curve having used Aperture for some time and just converting about a month ago.   Frankly, I have begun to wonder if the extra effort on RAW is worthwhile for action shots that mostly end up on website albums, Facebook pages and rarely see large print.  Even a lot of the candid and tight shots of faces look fantastic 1x1 from images that originated as JPG.  There is still plenty of room to push sharpening and noise reduction even in really high ISO images.

For fast moving sports events with lots of high speed bursts, I am seriously considering going with JPG only for in-game shots.  I can use a custom function selection to quickly pop back to CF-RAW/SD-JPG for group photos and unusual situations where I want the security of a good digital negative to fall back on.   I am certainly not in the JPG only camp but there are clearly situations that suit it. 

edit - please don't refer to me as thinking I am superior at taking photos than the average enthusiast. I don't believe anything other than I am seriously struggling to bring out the benefit that RAW offers.

That's what I am trying to get across, too. RAW processing makes us duplicate what the camera is already doing, very, very well.

The ultimate output medium of any photography is essentially at the 24-bit color level of a JPEG image.

Why don't we at least try to learn how to use the camera as an effective RAW processor, simply because it is?

164
Lenses / Re: Are primes really more sharp?
« on: April 26, 2012, 11:18:47 AM »
Some of you have been talking about the words "environmental feeling" and wondering what that means and why primes are better because of it.

I have no idea what the meaning of those words is as applied to a lens, but let me suggest a possible interpretation.

People with zooms are prone to zoom when they should be moving. It's not the focal length of zooming that is wrong, but just the perspective. People with a prime lens tend to become sensitive to adding the "environmental feeling" (whatever that is) to the photograph, because they move to where a person would actually move in order to see a human perspective of the scene.

If someone with a zoom lens would move and zoom, then that could be circumvented, but the very act of zooming makes everyone, including me, forget how to properly move. It's just too much for the brain to process. A prime lens takes this confounding factor of zooming out, and let's one more naturally take pictures, and capture the "environmental feeling."

I am kind of liking those words even though they have no meaning except what we choose to give to them.

165
EOS Bodies / Re: Shoot JPEG again with 5D3
« on: April 26, 2012, 12:00:07 AM »

But why get yourself a 5D3 to shoot JPEG? There must have been a bazillion words written in the RAW vs JPEG debate and the pro-RAW conclusions remain totally valid.

Frankly it freaks me out to shoot JPEG on any camera other than my phone...the potential for post-pro grief makes it a non-starter. If you know for 100% certain your output requirements are modest, check out mRAW.


I bought 2 5D3s to shoot JPEG only.  I have shot RAW exclusively for years and edited probably over 100,000 RAW and JPEG images.  RAW is really only a benefit to me when I miss the exposure or WB.  Sure RAW captures a lot more information but if you don't need that information then it is a waste.  I shoot 20-30 weddings a year and probably shoot 3000-5000 pictures per wedding and a good JPEG is just as good as a RAW image unless your settings are off or you plan on doing extensive dodging or burning.  If you shoot manual and dial in the WB using the Kelvin color temperature and the WB shift to properly balance the color of the light source your JPEG is going to be as good if not better than if it were taken in RAW.       

Way to go. In the real world a camera that can deliver JPEGs with quality and style is a lifesaver. In most of my work there simply isn't enough time to shoot and process RAW. I do test shots to determine the proper exposure value of everything in my field of view, and then use manual exposure to shoot all types of sports. The JPEGs turn out just as good as the best quality RAW files that I could ever process. Yes, I have the latest lightroom 4.1, the latest (Pre version 6, but that hasn't shipped yet) Photoshop / Creative Suite, and even Matlab numerical imaging processing expertise. RAW gives people 6 more bits to handle exposure errors (including exposure errors for certain color channels, aka white balance). People talk about RAW vs. JPEG as if there is some sort of emotion involved rather than just mathematics.

JPEG does not deserve all the criticism it gets; it is a valid choice if you can set your camera up to the proper exposure and you don't have time to make errors in taking pictures. RAW is if you want to dodge and burn, correct exposure, or generally fool around all day with pictures you have taken, like Ken Rockwell does. (Bummer, though, he shoots JPEG, which doesn't make any sense to me since he loves Photoshopping everything. He also seems to have no artistic sense if one looks his latest photo contest winners.)

I am totally neutral on the emotional RAW vs. JPEG debate, but I can tell you one thing for sure: most newbie photographers are not getting good advice by thinking they have to shoot RAW all the time. RAW doesn't make a picture better. It just gives you the ability to do more math to it afterwards.

That's why I don't like hearing all the nonsense about RAW, and why I like to chime in a positive word whenever someone has the bravery to say they sometimes shoot JPEG.

Here's an idea that would actually help people become better photographers, rather than the suggestion to shoot RAW:

Regardless of whether you shoot RAW or JPEG:

* Learn to use manual exposure and how to meter the actual value of the light that is illuminating the subject.
* Expose to this value of light (exposure value).
* Bright subjects will automatically have the proper brightness.
* Dark subjects will automatically be rendered with the proper amount of darkness.
* If shooting JPEG this can even be controlled by changing the Contrast of the picture style.
* Set the white balance carefully with a shutter speed well under the refresh rate of any man-made lighting that may be contributing to the scene. (Usually under 1/10th of a second is a safe speed for setting white balance.). This will avoid any under or over exposure in any of the color channels making the final image.

This will result in every case with fantastically beautiful pictures that make every image pop and every exposure perfect, plus consistent colors. Photography will become super easy.

The only remaining challenge is nailing focus and learning to handle a camera to optimize its AF system quirks, which is a lifelong job as complicated as playing tennis.

But if all the photos turned out perfect, then taking them would be boring, so missing focus once in a while has the benefit of making photography interesting.

No matter how complicated the auto exposure system, shooting in Av, T, or other modes where the camera changes the exposure, results in damage to the exposures far more significant than the 6 bits of freedom that RAW gives.

You can look back through a set of photos taken in Av mode, for instance, and the photos will all have different exposures--dark, light, dark, dark, light, dark, light, light--and on and on. Why in the world??? It's because they are all the wrong exposure. A photo should only be lighter if there is a lighter subject in it, and the exposure should actually stay the same relative to the light source.

Learn to expose correctly. That's way more important.


Pages: 1 ... 9 10 [11] 12 13 ... 15