Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
How important is high ISO?
- Thread starter Lovesong
- Start date
- Sort by reaction score
So, which one did you choose?
I'm going to go for the 30d with a 10-22, 24-70L, and a 50 1.4 as a starter. I think I'll get the 70-200 in the fall, and work on a couple of other primes at that time. Thanks for all your help.
Excellent choice What made you get the 50mm 1.4 instead of the 1.8, if you don't mind me asking?
What made you get the 50mm 1.4 instead of the 1.8, if you don't mind me asking?
Haha thanks.
Ok, I found this.
Blah, now I want to sell this thing for an 18-55 2.8 (NIKKOR!) I'm starting to buy into the 3rd party lens cliche.
It may be sensor reflection, not lens imperfection though, try passing it through NC's CAC function and see how it looks in any case.
I've read the 18-70 is a much nicer lens, and the 18-200 looks like it performs better...
I know, I know. I probably have something similar coming from compuwar.Blah.
Excellent choice
I think it was the plastiky-feel of the 1.8. My frind has it. Also, I've read that the 1.4 is slightly sharper... oh boy, I don't mean to start another one of these.
If you look at post #21, you'll see I'm not all that religious about it
I'd still urge you to look at the Fuji S3Pro though, the DR on that sensor is really interesting for landscapes and reduces the times you'll need to bracket or use an ND grad filter. It's not a fast camera though, so if you're not primarily shooting landscapes and brides & grooms (or if you shoot action,) it can be a bit frustrating. I'd still be looking at FF due to DoF, but that's just me.
In any case, good luck with your choice, I hope you get nothing but keepers.
It may be sensor reflection, not lens imperfection though, try passing it through NC's CAC function and see how it looks in any case.
I've read the 18-70 is a much nicer lens, and the 18-200 looks like it performs better...
Yeah yeah..
Actually I don't have NC on the mac yet, I'll see.
Well any of the 3 Nikkor wide/standard zooms are on my list, the 17-35 2.8 was stunning when it was lent to me at the Washington International Horse Show, very good performance. 18-55 2.8 and the 28-70 2.8 are the other options, latter of which would be the best replacement (and the most expensive - as much as my 70-200 2.8 VR!
)
Yeah yeah..
Actually I don't have NC on the mac yet, I'll see.
Well any of the 3 Nikkor wide/standard zooms are on my list, the 17-35 2.8 was stunning when it was lent to me at the Washington International Horse Show, very good performance. 18-55 2.8 and the 28-70 2.8 are the other options, latter of which would be the best replacement (and the most expensive - as much as my 70-200 2.8 VR!
NC NX is well worth the money. I'd go with either the 17-35 or the 28-70, from what I've read the 18-55 just isn't there. Good to see a strong case of NAS creeping up though, I was feeling lonely!
CA is lens induced, not sensor reflection in action. Some colours get bent less or more through the lens system, making part of the light spectrum show a bigger or smaller image. In the case of this Sigma you see towards the corner of the lens that to the left of the light part you see a red line, to the right you see a green line. This means that (probably) the red image is smaller. The red from the light area gets drawn into dark areas, and the green appears where red it lacking.It may be sensor reflection, not lens imperfection though, try passing it through NC's CAC function and see how it looks in any case.
I've read the 18-70 is a much nicer lens, and the 18-200 looks like it performs better...
Sensor reflection shows up as the dreaded purple fringing. It can appear everywhere in the image, not just more and more to the corners like CA, and it can be in all directions, and does not have one colour on one side and another colour on the other. Purple fringing can be easily combatted by stopping down, because that decreases the light intensity, and the reflection lessens.
Lens fringing tends to be magenta or purple in my experience, though I've read of green as well. This is the first time I've seen it seemingly without any blue.
Everything I've read says that sensor fringing is red, for example:
http://www.ing.iac.es/Engineering/CCDgroup.html
While I know that light coming in above the Nyquist frequency can cause moire anywhere in the image, all I know about sensor fringing is that it's red and not worse in the corners than elsewhere. Since the levels of light that will cause reflections in the image presented is there it's not a given for me.
(That is, I'm willing to accept it's not sensor reflection, but not that sensor reflection is purple fringing.)
CA always comes in 2 colours. One that is the colour that has a different size image because of the different diffraction, the other colur is where that coulr is missing. That is what CA is about. Most of the time it is red/green, sometimes it is more blue/yellow. That depends on what part of the spectrum is diffracting differently through the lens system. CA always gets worse toward edges, and does not appear in the center.
The purple fringing is really what is caused by sensor reflections/glass filter in front of the sensor reflections/microlens reflections.
It is called purple fringing because the UV end of the spectrum tends to bleed furter into the neighbouring pixels, making the fringing appear purple to blue. Sometimes the purple fringing actually goes more to red.
Purple fringing can appear anywhere in the frame where there is bright enough light, reducing light (smaller aperture) always lessens purple fringing.
Purple fringing can be more or less successfully be combatted by special coatings against the back element of the lens, which prohibits light boucing back to the sensor (this is what the "DG" and "Di" from tamron and sigma are mostly about).
CA you see whether you use film or digital, the way the light gets bent in the lens makes no difference to what medium captures the image.
You do not really see purple fringing with film though, since there is no sensor/glass/microlens that is so reflective.
The purple fringing is really what is caused by sensor reflections/glass filter in front of the sensor reflections/microlens reflections.
It is called purple fringing because the UV end of the spectrum tends to bleed furter into the neighbouring pixels, making the fringing appear purple to blue. Sometimes the purple fringing actually goes more to red.
Purple fringing can appear anywhere in the frame where there is bright enough light, reducing light (smaller aperture) always lessens purple fringing.
Purple fringing can be more or less successfully be combatted by special coatings against the back element of the lens, which prohibits light boucing back to the sensor (this is what the "DG" and "Di" from tamron and sigma are mostly about).
CA you see whether you use film or digital, the way the light gets bent in the lens makes no difference to what medium captures the image.
You do not really see purple fringing with film though, since there is no sensor/glass/microlens that is so reflective.
CA that I see is purple fringing. CA that others see is purple fringing. Sensor reflection that Canon protects against in the 1DS MkII is red- it's in their press releases. Sensor reflection that other CMOS sensor companies protect against is red.
I've got good into-the-sun purple fringing in shots in the corner of my WA lens where I don't have it under less adverse conditions, others see the same.
Purple fringing is due to defocus on the shortest wavelengths, while some have attributed it to microlenses, it's also possible that it's from blooming and lens CA. In the cases I've personally experienced, it's been into-the-sun shooting and clearly from the lens.
All the sensor folks try to protect against sensor fringe = red.
Are you saying that Canon is wrong?
I am saying you are wrong, and you do not understand what Chromatic Aberation means. If you did know that, you would know that purple fringing is nothing like CA and what CA looks like. I have no idea what your semi quote from whatever press release you are talking about is supposed to prove. It explains to you why there is an infrared filter in front of just about every digital camera sensor... and does not mention purple fringing.
You seem to be quoting out of context
Into the sun shooting and whatever resulting flares and reflections are not purple fringing. If however you shoot into the sun and there are areas of light/dark contrast, of course you can see purple fringing.
You can easily see that it is NOT the sensor itself that causes purple fringing, because at the same aperture and shutter speed, some lenses show it, others don't. This has to do with the quality of the coatings on the back elements, whether they reflects back or not. It always goes away with closing the aperture.
One note: not every purple looking defect is "purple fringing". CA can also give a purple edge (and a green at the other side of light areas). And like I said before, purple fringing mostly is violet to blue-ish, but can also appear a bit red-ish.
Biggest and most apparent difference between CA and PF is that CA will show up all over the place, because one colour of the image is just smaller, and PF only shows up around bright areas next to dark areas. CA is not fringing blooming in the same sense as PF is or some blooming effects on some cameras can be... CA is just that the colours do not project in the same size onto the image, making one colour smaller in size than the other. This is why for instance Nikon Capture NX can fix CA by resizing the colour channels (I think it can anyway... you can do it by hand in photoshop, although it is not totally precise due to the big steps of photoshops resizing).
It is not totally strange you do not know which is which, because if on one area of digital photography you can find a lot of nonsense on internet, it is on the subject of CA, PF and other artifacts.
This is what purple fringing looks like:
Here an explanation from Nikon about how to correct CA:
http://www.microscopyu.com/tutorials/java/aberrations/chromatic/index.html
This is what typical chromatic aberration actually looks like (we have seen that in EstorliM's examples already though:
You can see blue one side, yellow/orange on the other.
Here another example of CA, with my Tokina 12-24 f4 at 12mm.
You can see clearly red on one side, green on the other side of light areas.
Attachments
Cameras have two filters the first is a low-pass anti-aliasing filter, that's to stop aliasing, which is when the light hitting the sensor excdeeds the Nyquist limit for that sensor and gets represented as a different color (aliased.) That's also a form of CA, it's just not lens CA. The second is, as is quoted in the DP Review's 1DS mkII news release page (and a few hundred other places) :
http://www.dpreview.com/news/0401/04012904canoneos1dmkii.asp
http://en.wikipedia.org/wiki/Purple_fringing
So, as you can see, purple fringing == axial chromatic aberration == CA.
I've only gotten CA in quite extreme situations, and in those cases I knew it was going to happen. It appears a bit more when I use spot metering.
Anyways, with good glass it's not that big a deal
Anyways, with good glass it's not that big a deal
Anyways, with good glass it's not that big a deal
Fine I get the point!
I think it's harder for the 3rd party wide/standard zooms to stay on-par with the big guys, at least with the big resolution jump that happened after most of these lenses came out (with the exception of the 12-24s and stuff, which seem to be surprisingly good.)
I think this is a very big misconception. While it's true that third parties do make some junk, so do the big guys. Canon, Nikon, Pentax, etc. are all equally capable of producing Coke bottles.
My Tamron 28-75 f/2.8 is the finest zoom lens I have ever owned. I've owned a few in my day, and this one is the best. Sharp wide open, good contrast, and hardly any CA.
My Tamron 28-75 f/2.8 is the finest zoom lens I have ever owned. I've owned a few in my day, and this one is the best. Sharp wide open, good contrast, and hardly any CA.
1st point: the filter of some canon camera you keep mentioning has linke NOTHING to do with purple fringing. I have no idea why you keep bringing it up. Infrared filters are in almost all digital cameras before the sensor, including in yours.
And bringing up a wikipedia article up is quite funny. And the article actually states that what Brewster called purple fringing in 1833 is actually just "a lateral chromatic aberration".
Fact is... CA (chromatic aberration by the optical elements of the lens) are what I have shown you they are, and what you say in the examples EstorilM has posted. And purple fringing again is what I showed you.
Youy can jump high or low and not want to acknowledge what purple fringing actually means, and keep denying the CA samples from me and EstorilM are CA samples, but it does not make you right.
Why did you fail to read the Nikon page that explains CA? Less reliable source than wikipedia? Especially when wikipedia inexplicitly explains that the term purple fringing most of the time is used incorrectly:
"In optics, chromatic aberration is caused by a lens having a different refractive index for different wavelengths of light (the dispersion of the lens). The term "purple fringing" is commonly used in photography, although not all purple fringing can be attributed to chromatic aberration."
http://en.wikipedia.org/wiki/Chromatic_Aberration
The article then goes on explaining what CA is. If you after the article about chromatic aberration on wikipedia and Nikon's microscopy site still do not get that what is CA is not the same as what is actual purple fringing (it is not like I did not show you what purple fringing DOES look like), you are just stubbornly digging your heals in the sand.
Better articles about purple fringing and the actual cause of them can be found than wikipedia's confusion.
http://www.pictureline.com/newsletter/2004/july/purplefringe.html
The university of Hamburg, Germany, has a very clear and understandable explanation about what purple fringing actually looks like:
"The so-called purple fringing image artifact compromises the image quality of many digital cameras. It occurs when dark objects are in front of bright areas, especially in wide-angle lens settings and wide aperture. As you can see in the demo image (a crop of a larger image of Cologne cathedral), the effect manifests itself with bright purple/blue artifacts around the high-contrast zones that completely dominate and destroy the picture."
http://tams-www.informatik.uni-hamb...demos/00-intro/02-imageprocessing/purple.html
As can be seen on the examples I posted in above post.
Now a real referral to purple fringing on dpreview (not the strange infra red stuff you come up with everytime):
http://www.dpreview.com/learn/?/key=chromatic+aberration
Notice how the article first explains CA, then goes on to explain what purple fringing is:
"Although the above chromatic aberrations can be purple in color under certain circumstances, "Purple Fringing" usually refers to a typical digital camera phenomenon that is caused by the microlenses. In simplified terms purple fringing is "chromatic aberration at microlens level". As a consequence, purple fringing is visible throughout the image frame, unlike normal chromatic aberration. Edges of contrasty subjects suffer most, especially if the light comes from behind them, as shown in the example below. Blooming tends to increase the visibility of purple fringing."
Although I agree with the distinction they make between CA and PF (and the effects are very different, as you can see from posted samples), I do not agree with the microlens theory, which was thought of when digital cameras were mostly compact digital cameras. DSLRs show that PF is very lens dependant, and that it therefore is not a pure microlens phenomenon.
That is why there is a certain consensus that PF is actually bright light reflecting off the sensor/glass filter/microlenses, to the back element of the lens, reflecting back to the sensor and causing these purple ghosts.
Again, you are not the only person on internet not understanding CA = NOT PF. That is why you can find so many bad information about it. But the actual examples of the two phenomenas do show just how different they are.
I understand your argument here - however you need to put things into context a little bit. The Sigma lens I'm having problems with is an EX (professional) model with their best features, coatings, etc. If you were to purchase a similarly-credited Nikon or Canon lens, it would be simply unacceptable to have performance downfalls such as the ones I've encountered with the Sigma. Not to mention the fact that this is one of the HSM lenses that has the faulty chipset when used with the D200 - and incorrectly reports exposure to the camera (resulting in a -1 or greater underexposure at 70mm.)
Spend your money on faster professional-quality lenses instead of a more expensive body. For the amount of time you'll spend shooting at ISO 800-1600, you're MUCH better off with a really fast prime lens (f/1.4 or something like that) for your low light shooting and get a cheaper body (30D is awesome for less than half of a 5D - yea yea, I know it's not full frame. Boo hoo.)
As far as I know, lenses have NO idea about exposure. The lens lets light through and has no sensors. Metering is done by the camera, and the camera tells the lens to shut its diafragma (aperture) during exposure. That is all teh lens can do. So if you get underexposures, either the camera does not meter right, you do not meter right, the aperture for some reason closes a stop too much at 70mm, or you see the underexposure only with flash photography, which would mean the lens reports a wrong focal length/distance of focus to the camera.
Further more I had the impression you were talking about the Sigma 24-70mm F2.8 EX DG MACRO... but the 24-70 f2.8 does not have a HSM motor?
Well any of the 3 Nikkor wide/standard zooms are on my list, the 17-35 2.8 was stunning when it was lent to me at the Washington International Horse Show, very good performance. 18-55 2.8 and the 28-70 2.8 are the other options, latter of which would be the best replacement (and the most expensive - as much as my 70-200 2.8 VR!
The Nkkor 28-70mm is one of my favorite lenses; whenever you can get your hands on "the Beast," do so -- you won't regret it.
Both the Nikkor 17-35mm and the 17-55mm are outstanding lenses. Many landscape photographers prefer the 17-35mm, but both produce wonderful detail and sharpeness, with great IQ.
Well, say what you want - I can be fine at ~60mm and the second I hit 70 my shutter/aperture changes and it'll underexpose - it's not a little known observation of mine either, Sigma issued a press release regarding it, and has offered to replace the firmware in all appropriate lenses to be compatible with the D200.
"Combination of new Nikon 200 digital SLR camera with our following products can cause improper exposure. To overcome this we will be supplying a free upgrade to our customers. We deeply apologize for any inconvenience this may have caused to our customers.
List of lenses require update for D200 camera:
Fixed Focal Length Lenses
28mm F1.8 DG ASPERICAL RF
Zoom Lenses
20-40mm F2.8 EX DG ASPERICAL
24-70mm F2.8 EX DG MACRO
24-70mm F2.8 EX ASPHERICAL DF
28-70mm F2.8 EX DG
28-105mm F2.8-4 DG
28-105mm F2.8-4 ASPHERICAL "
..and yeah it's not HSM, not sure why I said that (it's loud as hell) still fast/accurate however.