Buy a Rebel XTi or keep my S3 IS?

[coldrain]

Most places only report MTF charts at infinity, while most macro shots aren't shot there- but then that wouldn't help back up his *opinions.*

How on earth are you going to do an MTF chart at infinity?
You show me one. Just one.
MTF charts are of course not measured at infinity, what are you going to do, pin a diagram to Jupiter to photograph it to determine sharpness and contrast? (the moon is too close for infinity measurements).

MTF charts are made from shooting diagrams at normal distances and counting lines. Counting lines that still are visible. And then plotting them out. And drawing conclusions from that in regards to actual sharpness and contrast of said lens on the measured focal length.

It would help if you would educate yourself a bit, you posts to just bring down this evil Canon owner are getting long in the tooth.

And I have no idea why you find it soo hard to believe that the Nikon 105mm (VR or not) just is not the sharpest kid on the block. Because it has Nikon stamped on it? And why do you bring up this MTF chart stuff again, when I in my reply already pointed out that the only time I actually mentioned MTF charts from the manufacturers was in regards to that Nikon 70-300 G lens.

Nice flame, again.

 

[compuwar]

Have you actually ever thought about on WHY it is like that that you can shoot shorter exposure times with shorter focal lengths? Is it because you shake less with longer focal lenghts? Uhmm... No.
Is it because a longer lens makes teh camera move more? Uhmmm..... NO again! Try it yourself, with a 70-200mm lens for instance... at 70 and at 200mm... HEY! 70mm is less moved than 200mm at the same shutter speed! Figure that .

Then what is it? EXACTLY! very good. The actual field of view, or the resulting magnification factor. *clapping for compuwar*. The fact that you

"field of view or the resulting magnification factor"

Pick one- they are not synonymous.

look at something from a big distance makes a small movement of the lens show big time on the image, when you maginify the subject.
So, does that mean that you actually have to factor in a crop factor? YES, it does. That is why you also have to factor that in with a digital ultrazoom like a Panasonic FZ30, ist 430mm is of course not really 430mm (big crop factor, remember?).

While the crop factor changes how much you magnify the image to print it, which should be factored in somewhat, it's not going to have the same effect. As the sensor photosite size is different on different crop factor cameras, and the resolution changes affect the image differently. Also, the magnification from the crop isn't during the period of camera movement, so it's effect is not the same as when the camera is moving.

On a 6MP camera, you get a lot more latitude due to the larger photosites than you do on a 12MP camera. Therefore it's not a given that the relationship is simply proportional to the amount of magnification both during image capture and afterwards.

What you'll find in the real world is that on a 6MP sensor body like the D70s, the 1over rule tends to work just fine, but on a D200 or D2x you'll need to adjust half a stop to a stop- they both have the same crop factor, but the sensor's photosite size affects the image sharpness due to movement much more on the smaller photosites.

 

[EstorilM]

*sigh*. You keep on talking about the 70-200 f4 L and the 70-300 IS USM without actually knowing anything about how they perform.

Okay. You're the expert. Not much more I can do here.

You're also obviously an expert on my experience with both lenses right? And I've never used either, right? Ok. That's funny. Because I have (on a 20D this past summer for a weekend event.)

Yes, I know it is hard for you to believe, and yes, it would not harm if you would actually educate yourself about these lenses intead of boasting time and time again on what a professional you are.

I consider myself fairly educated - everything I post has been factual, unlike your opinions on choosing a cheaper, less expensive consumer lens over a professional product.
I'm not boasting anything by the way, and I'm certainly not doing it time and again - I'm saying that I've used these things in conditions you can only dream of, and know that a 1 stop advantage can make the difference between catching the shot and missing it all together in certain conditions.

The 70-300 IS USM is an amazing lens for the money, and yes, like I pointed out in above post, IS actuallly WILL help you are 480mm tele range. And yes, this lens IS a better choice when you are looking for an ultrazoom replacement with IS and big range. Ridicule me all you want, you just show you do now know the lenses in question at all, and seem to just focus on a red L printed on one of them. Oh well.

I never said it wasn't a good buy, but I WOULD never say that it beats the 70-200 performance-wise (which is an even BETTER lens for the money.)
IS doesn't stop action. End of story. Fact. Like I said, shutter speeds are the only thing that will stop motion. VR and IS are most useful for stationary objects, panning is possible (disabling the horizontal stabilization planes) however most of the time lenses consider "tracking" to be "shaking" and will add blur to the image.
In other words, there are times when you need to turn IS off, then max aperture and optics are the only thing that matter for getting the shot.

You are the one who thinks a scratch on a lens is harmfull (it will not show up in a photo) and a filter is fine (that will show up in many photos, with visible refelections and loss of contrast) to protect you "investment".
So again, it is a tool, not a jewel. And lens caps do not scratch coatings.

Ok. Be my guest and scratch away if you want to. Actually - why even bother with lens caps?

This is not a Nikon 70-300 VR we are talking about, you know.

Um, first of all there are 3 different Nikkor 70-300 models, the most recent of which (that just came out) can easily hang with the 70-300 IS (which has second generation VRII and AF-S.) It's a little hypocritical if you ask me, since it's obvious that you haven't actually used this lens yet.

And about your puzzling remark about that "There's probably a reason 300mm f/2.8's go for $4500 while the f/4s are going for $1200", of course there is a reason... Bigger glass elements are very expensive to make, the lens weighs a lot more, and price also depends somewhat on production figures. But is the point you are trying to make?

That one went totally over your head didn't it? My point is that there's a reason people pay a $3300 premium over the f/4 for "only one f stop."

 

[compuwar]

How on earth are you going to do an MTF chart at infinity?
You show me one. Just one.
MTF charts are of course not measured at infinity, what are you going to do, pin a diagram to Jupiter to photograph it to determine sharpness and contrast? (the moon is too close for infinity measurements).

MTF charts are made from shooting diagrams at normal distances and counting lines. Counting lines that still are visible. And then plotting them out. And drawing conclusions from that in regards to actual sharpness and contrast of said lens on the measured focal length.

It would help if you would educate yourself a bit, you posts to just bring down this evil Canon owner are getting long in the tooth.

And I have no idea why you find it soo hard to believe that the Nikon 105mm (VR or not) just is not the sharpest kid on the block. Because it has Nikon stamped on it? And why do you bring up this MTF chart stuff again, when I in my reply already pointed out that the only time I actually mentioned MTF charts from the manufacturers was in regards to that Nikon 70-300 G lens.

Nice flame, again.

http://contaxg.com/archive/articles/beyond_the_mtf.htm

Although an MTF chart can represent the response of a lens at a wide range of focus distances, the variation of lens performance across this range is not typically reported. Neither Carl Zeiss nor the Photodo web site (http://www.photodo.com) routinely report MTF performance at a distance other than infinity. Results at infinity, however, can be markedly different than results at the closer focusing distances at which most photographs are taken. This is the case, for example, with the 60mm C f/2.8 Makro Planar for Contax SLRs, an exception in which Carl Zeiss does report an MTF at a close focus distance. The Makro Planar's performance at infinity is fine, but it really doesn't start to shine until used at a focus corresponding to a 1:10 reproduction ratio. Aside from the special case of a macro lens, a 90mm portrait lens, for example, would also be better served by good performance in the near range compared to infinity. Using the internally consistent measurements from Photodo to compare lenses across brands, therefore, must be approached with some caution. In the case of the Carl Zeiss MTF measurements, the Performance Data Sheets sometimes include a verbal indication of a lens's close focus performance, but just as often make no mention of it

.

 

[EstorilM]

This is going nowhere.

 

[coldrain]

Actually, field of view and magnification factor ARE synonymous.

The wider the field of view, the less close the subject looks to be on the photo.
The more tele you go (less wide field of view) the closer a subject seems to be. That really is all there is to it. In effect, when you crop a wide angle lens A LOT you get the same field of view (and same "tele" perspective distortion!!!) of a tele lens that would "crop" the field of view in a similar way.

And that whole blurb about photo sites... has like NOTHING to do with the rull of thumb. It just has to do with that yes, you have more resolving power with more pixels. But on the printed photo the camera shake unsharpness will be exactly the same.

And now about the maginification factor being a factor: It is not about magnification itself (else you would have a VERY hard time handholding macro photos), but about the big distance from your subject, where you move 1mm with the camera can mean that the tight tele view will show a subject movement of many centimeters.

And that is why the rule of thumb about the relation between focal length and shutter speed hand held also is the same for compact digital cameras and compact digital ultrazooms, and why an Canon S3 IS and Panasonic FZ30 with their IS turned off should be used with shutter speeds between 1/500th and 1/400th of a second, just like full frame 35mm cameras, and why the rule of thumb would be different with big format cameras (but they are not handheld that often, and not used with tele lenses that often), and why it also is the same rule with APS-C DSLRs with a crop factor influencing the field of view and therefore the magnification factor in the exact same way.

 

[compuwar]

Actually, field of view and magnification factor ARE synonymous.

No, they aren't. You can have two lenses with the same magnification
factor and differing fields of view.

The wider the field of view, the less close the subject looks to be on the photo.

Looks, not "is."

And that whole blurb about photo sites... has like NOTHING to do with the rull of thumb. It just has to do with that yes, you have more resolving power with more pixels. But on the printed photo the camera shake unsharpness will be exactly the same.

No. The rule of thumb is about *sharpness*. On typical Bayer sensors, the size of the photosites will determine how much the movement affects the picture. If the motion is less than the size of a photosite, or more accurately, if the motion is less than the distance to the edge of the next same-color site, then it won't be apparent in the picture. If it affects only the next site in the pattern, then it'll be less blurry than if it affects the next three sites. That's got to do with (a) the amount of motion and (b) the size of the sites. This is more apparent on R/G/B scenes/subjects because you get more into representative activation rather than interpolation to produce sharpness.

Again, you can see this in the real world where people switching from D70 to D200 bodies found technique flaws that with the same lenses and same crop factors produced less sharp pictures.

 

[annk]

Ok people, deep breath and back to topic.

...I have a feeling - at least for now I'll stick with what I have.. ...I guess I'll look into educating myself first and learning how to use a high-end camer, and then buy the camera...

I think you'll be pleased you waited. Your S3 has lots of manual settings you can use to get used to a bit of what a higher-end camera will give you, and you'll have a better idea of what you want.

 

[coldrain]

Okay. You're the expert. Not much more I can do here.

You're also obviously an expert on my experience with both lenses right? And I've never used either, right? Ok. That's funny. Because I have (on a 20D this past summer for a weekend event.)

Yes, of course you used both the 70-200mm f4 L and the 70-300 IS USM for a weekend.

And photozone, ColorFoto, all the Canon users that actually have tested both lenses side by side, and my own 70-200 f4 L are all wrong in noticing just how well the 70-300 IS USM performs optically. Right.
Just to educate you, experienced EstorilM, the resolution and contrast figures on the Canon EF 300mm F4 L IS USM measured on a Canon EOS 5D full frame camera by ColorFoto, german photo magazine:
Sharpness at the edge fully open/stopped down 2 f-stops:
925/1020 Line pairs per image height
Sharpness at center:
1275/1351
Contrast 53.5 points

Figures for EF 100mm f2.8 macro:
Edge:
986/1057
Center:
1368/1382
contrast:
56.5 points

Figures for the Canon EF 70-200 f2.8 L IS USM @ 200mm:
Edge:
947/1041
Center:
1232/1357
Contrast:
53.5 points

Figures for the Canon EF 70-300 f 4-5.6 IS USM @ 300mm:
961/1046
1389/1394
Contrast:
54 points

Here a comparison made by a poster on dcresource.com, comparing the 70-200 f4L IS USM (the IS version of the lens I own and know well) and the 70-300 IS USM.
He writes "The pics were taken at ISO 100, tripod mounted, cable release, IS turned off, 200mm. Shot in RAW, opened, zoom to 100% and crop the main portion of the image, saved as jpeg. No PP."
70-200 f4 L IS USM:

70-300 IS USM:

70-200 f4 IS USM:

70-300 IS USM:

Photozone has this to say about the 70-300 IS USM they tested, that was a sample before Canon fixed the lens problems of the lens btw:
"The performance of the Canon EF 70-300mm f/4-5.6 USM IS came as a total surprise. Unlike its predecessor the lens is capable to produce a very high performance throughout the zoom range without the significant drop in quality at 300mm typical for most consumer grade lenses in this range. It seems as if the new UD element helps to lift the optical quality significantly. Distortions, CAs as well as vignetting are also very respectable. So in terms of optical quality the EF 70-300mm IS can be almost described as a hidden Canon L lens. As much as it may promise here its build quality remains in line to what you can expect from a consumer grade lens and the small max. aperture is limiting its scope specifically regarding portraits where you seek for a pronounced fore-/background blurr only possible via large apertures (f/2.8 and larger). However, if you're looking for a very good, light-weight tele zoom e.g. for travel photography this lens should be high on your shopping list."

ColorFoto magazine concluded that the 70-300 IS USM was one of the best lenses they have tested on the 5D, second best after the Sigma EX 100-300mm f4 APO DG HSM, and better than most primes they tested too.

All this really does say something about just how well this consumer zoom performs.

I never said it wasn't a good buy, but I WOULD never say that it beats the 70-200 performance-wise (which is an even BETTER lens for the money.)
IS doesn't stop action. End of story. Fact. Like I said, shutter speeds are the only thing that will stop motion. VR and IS are most useful for stationary objects, panning is possible (disabling the horizontal stabilization planes) however most of the time lenses consider "tracking" to be "shaking" and will add blur to the image.
In other words, there are times when you need to turn IS off, then max aperture and optics are the only thing that matter for getting the shot.

You really have a hard time when someone makes a good point. You do NOT have to use 1/500th of a sec to "stop action". And so, at 480mm, IS can come in handy when light is not enough. This really is NOT hard to understand.

Ok. Be my guest and scratch away if you want to. Actually - why even bother with lens caps?

The plasic is SOFT compared to other things. You need something HARDER to scratch harder surfaces. That is where the lens cap comes in. Hey, maybe that is why they are not from stainless hardened steel . Protection from finger smudges (so you dont have to clean, remember), and protection od hard things that can damage the glass and coating. Don't be so simplistic in your arguements.

Um, first of all there are 3 different Nikkor 70-300 models, the most recent of which (that just came out) can easily hang with the 70-300 IS (which has second generation VRII and AF-S.) It's a little hypocritical if you ask me, since it's obvious that you haven't actually used this lens yet.

That one went totally over your head didn't it? My point is that there's a reason people pay a $3300 premium over the f/4 for "only one f stop."

Yes, there are 3 Nikon 70-300 zooms. To are quite worthless (the G and old ED version are optically just bad) and the new ED VR. The new VR is a lot better than the ED it replaces. If you want to actually believe that the Nikon AFS 70-300 VR is comparable to the Canon 70-300 IS USM, fine, believe that. And to please all that fall over the mention of MTF charts:
MTF chart for 70-300 VR Nikkor 70mm:

300mm:

MTF chart for 70-300 IS Canon 70mm:

300mm:

And about your hobby horse in this whole debate... that "one f-stop". The f5.6 is at 300mm, compare 200mm already. And in regards to IS not being of use, I covered that above already.

 

[coldrain]

No, they aren't. You can have two lenses with the same magnification
factor and differing fields of view.

No you CAN'T unless one lens is allowed to distort (ie: circular fisheye vs. normal "corrected" wide angle of same focal length.
Lenses with the same focal length have the same magnification of the subject on teh same distance of the subject. That is just how optics are.

Looks, not "is."

Yes of course looks!!!! geez, what is wrong with you. This discussion is getting to be ridiculous. With a tele lens an object looks closer. Now doesn't it. What is your point here??? When you crop a photo, the object looks closer. When you have a cropped sensor the object looks closer. It is always the same distance from you anyway so it never is closer.

Crop factor is equal to getting a longer lens, it just is that way. Try to explain to yourself why a FZ30 can bring things closer ("looks" haha) with such a short lens. Eventually you will work out what focal length and crop factor actually do and mean.

No. The rule of thumb is about *sharpness*. On typical Bayer sensors, the size of the photosites will determine how much the movement affects the picture. If the motion is less than the size of a photosite, or more accurately, if the motion is less than the distance to the edge of the next same-color site, then it won't be apparent in the picture. If it affects only the next site in the pattern, then it'll be less blurry than if it affects the next three sites. That's got to do with (a) the amount of motion and (b) the size of the sites. This is more apparent on R/G/B scenes/subjects because you get more into representative activation rather than interpolation to produce sharpness.

Again, you can see this in the real world where people switching from D70 to D200 bodies found technique flaws that with the same lenses and same crop factors produced less sharp pictures.

You confuse looking at 100% pixel size on your screen to the amount of blur on an actual printed photo of the same size. When you zoom in, things get more blurry. Nothing I said is against that logic. You on the other hand seem to mix up things. Just because you can zoom in with ridiculous amounts on your computer nowadays does not change anything about the rule of thumb, the rule of thumb is and always was about printed photos, it for instance never was about peeking at a negative film with a microscope (equivalent of what you are talking about with your confusion about higher pixelcount makes more blurry).

 

[annk]

I said deep breath, not hyper-ventilate.

 

[compuwar]

No you CAN'T unless one lens is allowed to distort (ie: circular fisheye vs. normal "corrected" wide angle of same focal length.
Lenses with the same focal length have the same magnification of the subject on teh same distance of the subject. That is just how optics are.

Sure, the 200mm lens on my D2x has the same field of view as the 200mm lens on my 5x7, not.


Yes of course looks!!!! geez, what is wrong with you. This discussion is getting to be ridiculous. With a tele lens an object looks closer. Now doesn't it. What is your point here??? When you crop a photo, the object looks closer. When you have a cropped sensor the object looks closer. It is always the same distance from you anyway so it never is closer.

Crop factor is equal to getting a longer lens, it just is that way. Try to explain to yourself why a FZ30 can bring things closer ("looks" haha) with such a short lens. Eventually you will work out what focal length and crop factor actually do and mean.
[/quote]
It's obviously different because you can crop (what a surprise!) the wider image and get the same thing. You can also magnify more and get the same subject size.

You confuse looking at 100% pixel size on your screen to the amount of blur on an actual printed photo of the same size. When you zoom in, things get more blurry. Nothing I said is against that logic. You on the other hand seem to mix up things. Just because you can zoom in with ridiculous amounts on your computer nowadays does not change anything about the rule of thumb, the rule of thumb is and always was about printed photos, it for instance never was about peeking at a negative film with a microscope (equivalent of what you are talking about with your confusion about higher pixelcount makes more blurry).

It's not confusion, it's your lack of understanding. By the time it goes from sensor photosites to pixels, the effect has already happened. That's why I specifically said photosites, not pixels.

Bayer sensors have sharpness issues due to diffraction based upon photosite size also. Photosite size has some sharpness benefit in terms of resolving power, but also detriment based on movement simply because you're affecting more photosites with the same amount of motion.

 

[coldrain]

Sure, the 200mm lens on my D2x has the same field of view as the 200mm lens on my 5x7, not.

It's not confusion, it's your lack of understanding. By the time it goes from sensor photosites to pixels, the effect has already happened. That's why I specifically said photosites, not pixels.

Bayer sensors have sharpness issues due to diffraction based upon photosite size also. Photosite size has some sharpness benefit in terms of resolving power, but also detriment based on movement simply because you're affecting more photosites with the same amount of motion.

You have to stop this though. Think about how a smaller photo site occupies a smaller fraction of the photo. What does that mean for the photo? Right. the photo will have the same amount of camera shake blur. BRAVO. Yes, and the smaller pixel (same as the photosite) will show more blur. YAY...
Still the exact same camera shake blur on the resulting photo. Happy happy joy joy.

And about that 200mm on a large format 5x7 camera... why did you forget to apply a crop factor (ie: that is NOT 200mm on 35mm film... get it?)
Same that applies with crop factors with APS-C or campact digital cameras applies here too!
First you have to convert the 200mm to an EQUIVALENT 35mm focal length. Just like you do with APS-C... just like you do with compact cameras.
200mm is in regards to 35mm film in this discussion.

If you have a 0.5 crop factor camera (a 70mm wide film) its 200mm lens would be 200 x 0.5 = 100mm in 35mm film terms.

Was I clear enough? And if you read my posts a bit better, I already mentioned large frame cameras in this repect in this thread, so you already should have realized that you need to first convert the focal length to 35mm with a "crop factor".

 

[EstorilM]

I said deep breath, not hyper-ventilate.

Haha yeah, I'm at work and can't justify feeding into this anymore. I also can't really keep track of anything more than 542345 different quotes in one reply - so.

Coldrain - maybe you could be a little less sarcastic / insulting? The bottom line is that (in spite of the lens being a good performer which I never argued, btw) it would be worthless to be, for what I do, a f/5.6 as a max aperture.

IS or not - for panning I can't risk using stabilization as it tends to do more harm than good with moving subjects, so that doesn't "save the day."

Test shots of a stationary subject, on a tripod, at ISO 100 are great and all - not but so realistic. I know you're comparing the optics of the two lenses, but I push the envelope to the max with aperture and ISO to maintain high enough shutter speeds to stop action for sports, that's my highest priority.

 

[compuwar]

You have to stop this though. Think about how a smaller photo site occupies a smaller fraction of the photo. What does that mean for the photo? Right. the photo will have the same amount of camera shake blur. BRAVO. Yes, and the smaller pixel (same as the photosite) will show more blur. YAY...
Still the exact same camera shake blur on the resulting photo. Happy happy joy joy.

(Your posts would be more readable if you quoted correctly.)

No, it's not the same blur because the amount of movement affects more photosites, and therefore the amount of detail that is lost is higher, and you don't get a direct N for 1 recovery due to the increased resolution. Again, I point you over here to the real world where as early as last month yet another D70 shooter realized loss of sharpness with the same technique/lens/tripod on a D200 is due to photosite size.

Was I clear enough? And if you read my posts a bit better, I already mentioned large frame cameras in this repect in this thread, so you already should have realized that you need to first convert the focal length to 35mm with a "crop factor".

You said they were the same thing, which is an entirely different thing than them having equivalence under some limited set of circumstances.

IS or not - for panning I can't risk using stabilization as it tends to do more harm than good with moving subjects, so that doesn't "save the day."

That's odd. I've been given to believe that IS for in flight bird shots increases the number of keepers when panning. Do you have any relevant data, examples, etc?

Hmmm, or is it that you mean instead of the speed increase? That is, all things the same, IS is an advantage, but not a substitute?

I'm considering upgrading my 400/2.8 for Eagle flight shots late next year if Nikon indeed produces a VR version as rumored, but not if it's not worth-while.

 

[sjl]

I was joking haha, and all the white canon optics are L btw.

But not all L glass is white. eg: the 85mm f/1.2L. I'm also pretty sure that the 400mm f/4 DO isn't classified as L series, but it is white.

Once again, you're misinformed and misinforming whilst trying to spout off as authoritative.

A *crop factor* is just that. It is *not* a *magnification factor.* The "rule of thumb" is about steadyness due to magnification, not field of view, so it's still 1/200th of a second with a 200mm lens and 1/300 of a second with a 300mm lens.

You're both missing the point. With sports, you want a fast shutter speed to freeze the action, not to avoid camera shake. For that purpose, you want 1/1000 of a second or faster where possible, and that means fast glass. Period. No IS can make a difference when the subject is moving fast.

 

[coldrain]

Did that D70 to D200 owner factor in that the in camera sharpening and antialiasing filter from the D70 result in a LOT sharper photos than any other 6mp camera, with more artifacts, even in RAW?

I guess not. That is the real reason why the D200 seems to be less sharp (and everything else seemed less sharp compared to the D70, including the Pentax *ist range, the Nikon D100, the KM 7D and 5D (all a same Sony 6mp CCD if I am not mistaken) and the Canon D60, 10D, 300D.)

Know what you are comparing... DSLRs do not sharpen much and therefore seem soft, and the D70 was an exception to that rule.

If you doubt what I am saying, I gladly will provide you with resolution figures for the D70(s), D50, 7D, 5D, *ist D, DS, DL, tested with prime macro lenses. Then you can see for yourself that the D70 is unusually sharp.

 

[compuwar]

You're both missing the point. With sports, you want a fast shutter speed to freeze the action, not to avoid camera shake. For that purpose, you want 1/1000 of a second or faster where possible, and that means fast glass. Period. No IS can make a difference when the subject is moving fast.

That depends on the sport, doesn't it? For instance, in motor sports, 1/1000th tends to make wheels stop, and motion blur is important to show speed, where at around 1/500th you should get motion blur and it's up to your panning skills to get the car/bike in focus. I'd expect side shots of sprinting, speed skating, etc. to be about the same, but I don't shoot those sports, so I'm not sure how practitioners do it. For team or 1 on 1 sports with more than one person as a focus, I'd expect you'd have to go with freezing the action to get usable images.

I got mixed results with very high shutter speeds and show jumping the one time I tried it, I could have used a couple of stops of ND filter and some hoof blur to get more realistic looking pictures for the side shots IMO, but that wouldn't have worked for the head-on shots.

Did that D70 to D200 owner factor in that the in camera sharpening and antialiasing filter from the D70 result in a LOT sharper photos than any other 6mp camera, with more artifacts, even in RAW?

In camera sharpening and artifacting only applies to JPEGs, which the owner didn't shoot.

Since better technique and/or a stop faster shutter speed "fixes" the problem for most folks who have this issue, the AA filter would seem to not be the factor.

Most start with a stop faster as they refine their technique, as did the last in the series of folks. It's gotten to be a pretty routine thing on Nikonians, as it was on DPReview when the D200 was released.

I almost didn't get a D200 when they first came out because of sharpness issues in most of the samples posted. Then I saw some good frames from someone who had technique nailed. The same issue happened with the Fuji FinePix S2Pro initially, though not nearly in the same volume as far as I recall.

If technique fixes it, it's not a hardware problem, but more of a WAD issue.

 

[Abstract]

How on earth are you going to do an MTF chart at infinity?
You show me one. Just one.
MTF charts are of course not measured at infinity, what are you going to do, pin a diagram to Jupiter to photograph it to determine sharpness and contrast? (the moon is too close for infinity measurements).

"Measuring at infinity" is just a term. Just move your subject so that the subject is small enough to be considered infinity. Just take the subject far enough away from the lens. Of course, this distance will depend on the focal length of the lens that's tested. From your statement, no object is ever at infinity. Depending on the lens, even 7-8 metres (say 25 feet) is considered at infinity if the subject is the size of a basketball.

Places like DPReview just take their charts and move them as far away as required to fill the frame to whatever standard they normally use, take the photo, then just zoom in to 100% and look at the resolution chart. As a physicist, I'd still consider what they do as equivalent enough to infinity that it makes sense.

And I have no idea why you find it soo hard to believe that the Nikon 105mm (VR or not) just is not the sharpest kid on the block. Because it has Nikon stamped on it? And why do you bring up this MTF chart stuff again, when I in my reply already pointed out that the only time I actually mentioned MTF charts from the manufacturers was in regards to that Nikon 70-300 G lens.

And in a large percentage of other posts you make in this forum. Someone asks about a lens, and you start talking about how a lens did with regards to MTFs and other tests. I asked you whether you actually judge things from photos because I'd say that if a lens can take good, sharp photos, even when these photos are printed to 8" x 10" or larger, then it performs well. A Nikon 105 mm VR performs well in this regard. I guess that for you, looking at a piece of paper stuck to the wall is a better indication of how well a lens performs, but I'd rather look at photos.

If I were you, I'd be happy about owning the 70-200 mm f/4 if you have taken quality photos with it. There must be some that you wouldn't have been able to take as well if you had the 70-300 mm

What is the max. handholdable shutter speed you can realy on photos to be sharp at 200mm? 200 x 1.6 = 320mm... so 1/320th of a second. Do you think it is a problem to shoot indoor action at 1/320th of a second? And what about 300mm. 300 x 1.6 = 480mm. 1/500th of a second. And you really think IS will NOT help you there? The IS at f5.6 will easily put it past the f4 lens in this case. Since you have the room in exposure time to have the IS help you get more light in.

Well firstly, Compuwar is probably right about the 1/focal length rule on APS-C cameras. The "cropping" doesn't happen until after the photo is taken, so the effect of handshake is the same as on a full frame camera. I've read this once before, and quite frankly, it's the only way it physics makes sense even if I hadn't read it elsewhere before. Not sure about the "photosites" thing.

So anyway, at 200 mm, you'll need approximately 1/200 seconds to get a decent shot, which is achievable in some indoor sports. You say the stop doesn't make a difference. So does the 70-200 mm f/2.8's extra stop over the 70-200 mm f/4 make any difference either?

That's odd. I've been given to believe that IS for in flight bird shots increases the number of keepers when panning. Do you have any relevant data, examples, etc?

Not sure about bird shots since they fly so fast, but if your subject is moving relatively slowly, then IS/VR may not be useful at all. (think about a tennis player or even a football player). You want to freeze the motion (and do a decent (but not perfect) job on the tennis racket), but if that tennis player's body isn't moving horizontally across the "frame" all the time (and they won't be facing you side-on all the time in these sports), then their movement in your frame will be a bit slower than the speed they're actually, then IS/VR deters a lot of slow to mid-speed panning.

Shooting birds? Well, I don't know much about it, but if you're panning fast enough, I doubt VR/IS would make a difference.

 

[dllavaneras]

So, to sum up: should the OP buy an XTi or keep his/her S3?

 

[coldrain]

And in a large percentage of other posts you make in this forum. Someone asks about a lens, and you start talking about how a lens did with regards to MTFs and other tests.

Yes, "and other tests. That part is correct. Real tests, showing real resolving power by independant test organizations. No idea how anyone can object to that. Only mention of MTF charts was in the 70-300 G thread, after a LONG blahblahblah that went on and on, not believing the 70-300 G was that not good. So I offered the Nikon MTF chart to show beyond a reasonable doubt.

I asked you whether you actually judge things from photos because I'd say that if a lens can take good, sharp photos, even when these photos are printed to 8" x 10" or larger, then it performs well. A Nikon 105 mm VR performs well in this regard. I guess that for you, looking at a piece of paper stuck to the wall is a better indication of how well a lens performs, but I'd rather look at photos.

I never said the 105mm VR does not perform well, I said it does not perform as well as others in resolution and contrast. You can look at their (the other, better performing macros) photos too, you know. And also be very happy.

If I were you, I'd be happy about owning the 70-200 mm f/4 if you have taken quality photos with it. There must be some that you wouldn't have been able to take as well if you had the 70-300 mm

I already mentioned the one downside of the 70-300 IS is a rotating front end. So yes, the photos I took with AF and Polarizer would be the ones I would have had to take without polarizer. I already mentioned this in posts above. And there are MANY more photos I could not take because the 320mm versus 480mm gap... that I could have taken with the 70-300 IS. And of course, the photos I could not take because of the lack of IS in my 70-200 f4, especially missed when I use my 1.7x soligor TC to make up for the 200-300mm gap.

So..... more photos missed because I have the 70-200 f4 L than photos missed if I have the 70-300 IS, quite frankly.

Well firstly, Compuwar is probably right about the 1/focal length rule on APS-C cameras. The "cropping" doesn't happen until after the photo is taken, so the effect of handshake is the same as on a full frame camera. I've read this once before, and quite frankly, it's the only way it physics makes sense even if I hadn't read it elsewhere before. Not sure about the "photosites" thing.

Well, no he is not right. As you can yourself understand when you think about a compact digital... they really do not have 430mm long lenses, you know. And cropping doesnt happen after a photo is taken, cropping is always there just because your film is smaller. So what happens when you print the photo? Oh gosh, the person in the image is BIGGER... MORE MAGNIFIED. So... if for instance the camera shake made the image blurry over 2 mm distance next to the persons eye, in full frame (200mm lens). What happens if we crop that (APS-C)? Wow, the 2mm is not 2mm anymore. The camera shake blur all of a sudden became 2 x 1.6 = 3.2mm. Just as if we shot the same photo with a 320mm lens on a full frame.
Because the crop factor does the exact same thing as making the focal length longer does.... it magnifies, including camera shake blur.

So anyway, at 200 mm, you'll need approximately 1/200 seconds to get a decent shot, which is achievable in some indoor sports. You say the stop doesn't make a difference. So does the 70-200 mm f/2.8's extra stop over the 70-200 mm f/4 make any difference either?

So anyway, at 200mm, you'll need approximately 1.6 x 200mm = 320mm -> 1/300 seconds to get a decent shot. At 300mm that becomes 1.6x 300 = 480mm -> 1/500th of a second. Since you dont need absolute stopping power with actionshots, you can go lower than 1/500th of a second. Lets say 1/250th of a second, how does that sound? Hey, that is where the IS comes in. And the 70-200 f4 L I have does not have IS.

SO, considering all this, yes, that one stop is not a biggie even in indoor action shots.

And no, I don't say f4 and f2.8 makes no difference.

So, to sum up: should the OP buy an XTi or keep his/her S3?

Did we not already cover that?
OP keeps his S3, and saves up for XTi + 70-300 IS USM.
This way OP keeps compact camera for when DSLR is too cumbersome/forbidden, and keeps movie mode.

 

[annk]

Actually this is what the OP landed on:

...
I have a feeling - at least for now I'll stick with what I have.. I'm not so concerned with Macro photography - and for general use I think I could get by with the basic lens that comes in the kit - my primary goal is to take zoomed in shots, and would require me to get one of the more expensive recommended lenses.. add on that I am not completely sure what I'm doing and letting the S3 do the work might be the best bet. I'm actually glad I posted - I liked the S3, but the few comments on here about it made me feel better about that purchase!

I guess I'll look into educating myself first and learning how to use a high-end camer, and then buy the camera... ...

So this thread reached its natural conclusion at post 16, and lens war can finally be put to rest.

 

[Abstract]

Well, no he is not right. As you can yourself understand when you think about a compact digital... they really do not have 430mm long lenses, you know. And cropping doesnt happen after a photo is taken, cropping is always there just because your film is smaller. So what happens when you print the photo? Oh gosh, the person in the image is BIGGER... MORE MAGNIFIED. So... if for instance the camera shake made the image blurry over 2 mm distance next to the persons eye, in full frame (200mm lens). What happens if we crop that (APS-C)? Wow, the 2mm is not 2mm anymore. The camera shake blur all of a sudden became 2 x 1.6 = 3.2mm. Just as if we shot the same photo with a 320mm lens on a full frame.
Because the crop factor does the exact same thing as making the focal length longer does.... it magnifies, including camera shake blur.

Using a compact to compare to the DSLR situation isn't a fair comparison, though. The small lens on a compact was designed to work with the small sensor size used in compacts. With DSLRs, you're using lens that were designed to spread light onto a 35 mm sensor (thereabouts), but using the lenses on a smaller sensor instead. The lens and focal length is the same, and the light coming down onto the sensor covers the same amount of area (ie: the beam of light has the same cross-section), but the sensor itself is only getting the middle portion of the light "beam". The entire physical process of the light hitting the sensor is the same, while nothing is physically magnified or different because of the smaller sensor.

So if you took a Canon 5D and a Nikon D200, used a 50 mm lens, and introduced identical camera shake to both cameras using the same shutter speed, there's no physical difference in how the light hits the sensor. The camera shake will spread the light over the 5D's full-frame sensor in the same way that it'll spread the light onto the D200's APS-C sized sensor, which is why the recommended shutter speed is still 1/(the lenses focal length), even on a DSLR with smaller sensor.

However, when looking at a photo, the blur will appear 1.5x greater on a photo produced by the D200, but we're talking about good shutter speed, which is really related to optics and geometry, not the photo produced. This is why I said the "crop factor" is introduced only after you take the photo.

The word "crop factor" in itself is deceiving, and is why people don't really understand it.

 

[coldrain]

However, when looking at a photo, the blur will appear 1.5x greater on a photo produced by the D200, but we're talking about good shutter speed, which is really related to optics and geometry, not the photo produced. This is why I said the "crop factor" is introduced only after you take the photo.

The word "crop factor" in itself is deceiving, and is why people don't really understand it.

It is you though, that does not fully understand it...
"Good shutter speed" is in DIRECT relation to the size of the actual captured image, in the "rule of thumb" we are discussing.
This rule of thumb is about focal lengths as we understand them on 35mm film. Not on large format cameras, not an APS-C format film.

Image circle has totally nothing to do with it. You can make the image circle as big as you want, it will not make a difference to the photo. Not in camera shake, not in depth of field, not in magnification, the only influence it can have is vignetting.

Only 2 things are involved here. Size of the imager (sensor or film) and focal length. Nothing else (besides of course the weight of the lens, but the rule of thumb does not take that into account).

Hmmm... size of the imager... HEY.... that is our "crop factor".
Size of the thing that records light from the projection made by the lens.
And the projection the lens makes is directly caused by.... the focal lenght.
This is what focal point means. The point where the rays of light all cross in the same point in front of the projected image (sensor/film).
And how wide the resulting photo is depends on how wide the imager is.
A 35mm lens on a compact digital camera will not produce a wide image at all, it will be very "tele" in fact.
A 35mm lens on a 35mm film camera will produce a slight wide angle photo.
A 35mm lens on a APS-C digital SLR will produce a "standard focal length" photo (like a 50mm lens will do on a 35mm film camera).
A 35mm lens on a large format camera will produce a very wide photo.

All you need to do is to adjust the image circle, because if you don't the large format camera would vignet something fierce, and the compact digital camera would be needlessly heavy and big. And you see this happening with APS-C SLR lenses too.... the Canon EF-S, Nikon DX, Tokina DX, Tamron Di II and Sigma DC lenses all have reduced image circles, and will produce heavy vignetting on a full frame 35mm camera.

Now back to camera shake!
If I look through my camera (an APS-C crop factor DSLR) at my Sony Playstation 2 over there (pointing at the playstation 3 meters away), with a wide angle lens(lets say 20mm), I will see a tiny PS2 in the "image". Now if I stay put, and put on a 200mm lens on my camera, and look at it again through the camera, I will see the PS2 fill the entire image.

If I happen to move the camera slightly, the exact center in the lens system that is pointed at the PS logo may be moved a tiny bit... camera shake. Lets say it will move 0.5mm to the left (measured on the logo itself) during the exposure time.

On the wide angle photo... the 0.5mm movement of the logo will hardly be noticable. On the 200mm tele photo however, the 0.5mm movement on the logo, that gets totally enlarged, seems huge, and gives the photo a soft and blurry character.

Now THIS is what the rule of thumb is about. The movement of the logo stayed the same (0.5mm to the right, because my camera moved to the left), but what makes it appear on the 200mm photo and stay unnoticed on the 20mm photo is the fact that the actual resulting image enlarges the subject (and hence, the movement of the subject in relation to the sensor) at 200mm, compared to the 20mm photo.

The rule of thumb is about how much the movement of the subject in relation to the captured image shows up in the enlargement factor. And that is why the crop factor is very important and needs to be taken into consideration. So yes, 200mm becomes 320mm with a 1.6x crop factor camera. And yes, the rule of thumb tells you to then not go under 1/320 sec for best results. And yes, this also applies to cameras with even wayyyy smaller sensors.

Take for instance a Canon PowerShot S3 IS. What lens does it have?
A 6-72mm f2.7-3.5 zoom lens. Wow. On a 35mm SLR... that would be ultra mega turbo wide to portrait range. Pretty excentric and specialized!
But when you actually USE the camera... HUH? what happens. You don't GET ultra mege turbo wide at all. And the long end does not really behave as portrait range either... its magnification is wayyy too big for that, and its tele distortion is also very big (shortening of distances).
So, how can that be? Because... it has a 6x crop factor sensor, that is how.
And that makes the lens into a 36-432mm lens in 35mm equivalence (6x6 = 36, 6x72 = 432). And what about the rule of thumb about camera shake and focal length... can I shoot it at full tele at 1/70th of a second and hope for sharp photos? Not really... actually 1/400th of a second is a better idea. (This is disregarding the fact that this camera has IS to help you with lower shutter speeds, of course).

And the same goes for cameras with bigger than 35mm imagers. Since the rule of thumb is about 35mm terms and measures... you have to calculate the 35mm equivalent first, before you can use it in refereence to its lens' focal lenght.

 

[wmmk]

sorry I missed the lens war, but you may consider the G7 or a lumix SLR-like model, as they're closer to an XTi, but have all the glass you need built in and will not hit so hard on your pocket.

EDIT: this thread just made me think of an interesting question. when using a "35mm" lens on an APS-C DSLR, is the image in the viewfinder what is being taken by the sensor, or does the viewfinder show more than the final taken photo? hope I phrased that well...