cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

EOS 70D autofocus tracking???

Bazsl
Rising Star

Will the phase detect 19 point AF system in the 70D track subjects moving toward the camera at 100 miles per hour in AI Server mode? In other words, is the 70D usable for still photos of aircraft in flight, horse racing and automobile racing? Any references to articles on the 70D AF sustem that discuss its ability to track subjects moving toward the camera would be appreciated. Thanks.

 

Bill

1 ACCEPTED SOLUTION

If it's better than the 7D it won't be by much from what I've read but I haven't put any effort in that either. I have shot Radio Control events with a 7D with very good results & that's harder than full scale. I've also shot car racing with a 40D & had a pretty good keeper rate.

 

"A skill is developed through constant practice with a passion to improve, not bought."

View solution in original post

80 REPLIES 80

Talking your telephotos, I suspose you have a 'good' tripod?

I imagine you do but do you have a gimbal head?  That is the next step to better tele photos.

 

EOS15087.jpg

EB
EOS 1DX and 1D Mk IV and less lenses then before!

I've done the photos for the FOV test & will get at them tonight. This is an old test I did years ago re IQ of the longer lenses I had then. I realy didn't find any real IQ differences between the Sigma's & Canon's but the fact that they didn't have focus limiter switches really slowed down the AF so once I started shooting R/C they got sold off.

 

http://picasaweb.google.com/116179596240613012497/300500Samples

"A skill is developed through constant practice with a passion to improve, not bought."

In my opinion artificial reach as we have been discussing it here is real based on my tests. It is however infuenced by the smaller photosites on the crop sensor too. The helicopter isn't very big & the tail boom's diameter is just 1/16 inch (2 mm) and I didn't do any editing so that no extra info added to file sizes. I used the outline of the panel on the door to do my crops so there is a minor difference file to file, and the difference in lighting is due to a rain shower interrupting the shoot.

 

http://picasaweb.google.com/116179596240613012497/FOVComparisonSet03?authkey=Gv1sRgCL7TyMbc-LHgxwE#

 

I have used the older Picasa style album because it shows larger images on screen than the new Google + style does.

 

Larry

 

PS the Nikon lenses are soft & I have spent considerable time doing the AF adjustments as per Nikon's instructions.

"A skill is developed through constant practice with a passion to improve, not bought."

With out side by side comparisons, I don't know that I can draw a definitive decision.

I can draw the conclusion that I am glad I choose Canon so many years ago!

 

What is your conclusion of the two series of tests?

 

This is extremely difficult to make conclusive findings. Because you would have to have very similar cameras except for sensor size to make a absolute conclusion.

A certain two cameras can and could show a difference. It is probably where the crop factor tele benefit came from in the first place.

 

If a similar test was between a 70D, a 1D Mk IV and a 5D Mk III, I still maintain there would be little to no difference in the crop factor. Or, how about pitting the 1D Mk IV against the 1Dx?  At any rate I believe you will be in the pixel-peeper area. This is a huge debate with the birders crowd.

 

I shot Rebel XTi's (five of them) for years along with my 1 series. Usually giving them (XTi) to my assistants.  When I got a 5D Mk II, I started to notice how much better just random shots and photos 'generally' were.  But with the 5D Mk III, there simply is no crop advantage to using a Xti for tele work.

 

Again this is predicated on how the work is to be used. If you goal is to have a Photoshop pixel-peeping display that is one thing.  If it is to be art prints it is another and if you are going to do regular prints another.

Of course if the client wants to go to Walmart and print 4x6's or use Facebook, almost any camera will do. Crop or not.

Thank you for takling the time to do this.  I know it is time consuming but it is very interesting.

 

 

EB
EOS 1DX and 1D Mk IV and less lenses then before!

Dear Cicopo,

 

Thanks for resuurecting these data comapring FF with APS-C.  I ahve a hard time seeing the differences from the web-sized images.  Can you please summarize your findings?

 

Also, I have done some more research addressing the theoretical issues surrounding the cropped FF vs APS-C issue.

 

I have been assuming that the pixel number is synonomus with resolution.  I have since learned that this is not so.  In comparing analog (film) with digital, one must take into account sampling science.  There is something called the Nyquest Theorem that states that the resolution of a digital sensor is 50.8% of the nominal pixel size.  For a full theoretical review see wikipedia, Nyquest theorem. 

 

How does that affect our considerations regarding the roles of lens resolution vs pixel count?

 

In an earlier post I compared the practical resolution limit of excellent quality modern lenses (i.e. 130 line pairs / mm) to sensor resolution.  The lens-based resolving power of a FF sensor has an effective megapixel count of 14.5.  APS-C sensor has 5.5 megapixels.  A 6D has a FF sensor of 20.2 mp, 7D APS-C has 18 mp.  These must be adjusted in accord with the Nyquest Theorem, reducing them to 7.25 mp and 9 mp respectively.

 

Summary table

 

Effective Megapixel Resolution

 

                 Lens                       Sensor             Effective Max 

FF             14.5                           7.25                      7.25       <- Controlled by Sensor

APS-C       5.5                            9.0                        5.5         <- Controlled by Lens 

 

So it's interesting to note from this table that the sensor controls resolution for FF cameras, while lens resolution rules for APS-C cameras.  

 

An interesting aside is that FF sensors can increase in size to 29 mp before the lens begins to control resolution.  So some improvement is possible.   For APS-C sensors, we have already out-stripped the lens capabilities with the sensor mp count. No new resolution gains can be expected by increasing APS-C pixel count.  

 

Whether this bears out in practice is the question...

 

 

Correction to previously posted table:

 

Summary table

 

Effective Megapixel Resolution

 

                                 Effective Max         Nominal               Nyquest                    Effective Max

                                 Lens                        Sensor                 Adjusted                    System

                                                                                                Sensor

             

FF                            14.5                          20.2                      10.1                            10.1                           <- Controlled by Sensor

FF 1.6X Crop           5.5                          12.6                        6.3                             5.5                            <- Controlled by Lens 

APS-C                      5.5                           18.0                        9.0                             5.5                            <- Controlled by Lens 

 

If these results are to be taken at face value,

 

1)  a FF Sensor camera is about twice as good as an APS-C sensor camera.

2)  FF cropped to APS-C size is about the same as an APS-C sensor camera.  This implies that it's just as good to crop a FF down to APS-C size and not buy a separate body to " extend reach."

 

Some other considerations I'd like to raise.  

 

I read about the increased reflectivity of C-MOS sensors as compared with film.  There seems to be emphasis on eliminating reflections from the inside of the camera.  A full frame lens, dumps an enormous amount of stray light into the body of the camera.  Is there any concern about this?

 

Aside from the theoretical considerations of cropped FF vs APS-C, there should be some practical things that favor one approach over another.

 

Advantages of APS-C:

Easier to frame the image (assuming one is far enough from the subject)

Easier to manually focus as the subject is bigger

Can use cheaper lenses

APS-C lenses have less stray light (advantage?)

 

Advantages of cropped FF:

Can use professional lenses already in arsenal

More peripheral vision enhances framing of moving objects, planes, birds, etc.

Larger pixel size gives better all-round lower noise 

Larger pixel size is better in low light

 

 

 

 

 

 

 

Gary James has hit the nail on the head.

Don't get caught up in the MP count.

It actually boils down to what the eye sees.  A few larger pixels stores more quality light than a lot of smaller ones do.

 

We have not even got into noise.  Probably the most important spec.  Larger pixels are less effected by noise than smaller ones.  We won't address resolving power of the lens as we assume the lens is the same for each sensor.  But of course the lens has a great deal more to do with the quality of the photo than anything else.

I am, obiviously, not good at explaining this, however it does not seem to stop me from trying.

EB
EOS 1DX and 1D Mk IV and less lenses then before!

Garry, no offence but you're over thinking this. The reality of it is that there are only a few camera bodies & a few lenses to choose from that can fit the needs of bird / aviation / or easily spooked small wildlife, so we learn to work with what we can afford. If you're going to shoot hand held which I do for at least 99.5% of my stuff lab results aren't that important compared to proper technique. IS is a great helper so I've gone that route where ever available & in general know whether to turn it off which is pretty rare with what I shoot but one exception is the 24-105 which won't do well at panning unless the IS is OFF. 

If you intend to shoot hand held for any length of time lenses like the 400 f5.6 L & 100-400 L IS are the longest most of us can manage & you can't add a TC to either for action without a pro body or 5D3 which will tollerate a 1.4 TC but with slower AF response times (based on my reading & not actual use).

Crop bodies do provide REAL artificial reach, as can be demonstrated by my samples, but as I and others have noted at the risk of marginally lower IQ in good light & more noticibly in poor light or deep crops. No 1 camera does it all perfectly but at least there are several to choose from & each of us goes different ways thanks to what we hope to achieve.

My samples prove a few things & the only one I was concentrating on was comparing the FOV at specific settings. The relationship between the number of pixels in the crops proves the artificial reach & can be used to calculate results from other bodies such as the 7D or 70D by working out the ratio of 1 sensor to the other & using that to convert the crop info.

Although not exact the 7D has 18 Mpix, the D7100 has 24 mpix so the 7D should provide roughly 75% as many pixels in the crop that I got from the D7100. In reality they have a different crop factor (1.5 vs 1.6) but for a quick comparison 75% should be close enough. If you look at the full frame samples you can see that the 1.5 crop body at 300 mm had a similar FOV to the 1.3 crop body at 400 mm & a tighter FOV than the full frame body. This really does mean that a 400 mm lens on a crop body can render a similar magnification factor to having a longer lens on a full frame body & more importantly may be the deciding factor in being able to shoot while hand holding due to weight & physical size.

Finally my crops were just for comparison & I used the panel outline to keep things relatively close without making it complicated & I didn't take printing into consideration. Again using the full image size relative to the crop size will give a ratio of just how deep I cropped each image. If you want to work that out there are the pixel measurements for the bodies used plus what a 7D supplies

 

 

1Ds2 images are 4992 X 3328

1D mark lV images are 4896 X 3264

7D images are 5184 X 3456

D7100 images are 6000 X 4000

Any others that you may be considering will have those numbers in their specs.

"A skill is developed through constant practice with a passion to improve, not bought."

garyjames has hit the nail on the head..

It is not necessarily the number of pixels as it is the number of 'quality' pixels.

Lager pixels can store more light than smaller ones. This leads us to a subject that has not been fully discussed. And that is noise. Probably more important to cropping than number of pixels. Larger pixels because of more capacity have less noise. This will make a sharper more appealing photo.  It does eventually all boil down to what the eye can see.

 

Also one must compare lens to lenses. As the glass has more to do with ultimate sharpness than anything and the ability to crop.  If the lens can not resolve it, the sensor can't record it.

 

Except for maybe a certain combination of cameras and lenses, I remain there is little to no crop-factor telephoto advantage. Remember I am coming from a time when the best cameras made had 4 MP sensors. Yeah, you can't make big blow-ups but the quality of the prints, 8x10, was still extremely good. You still need to get into pixel-peeping arena to see the difference.

I still will grab my 1D and shoot some flowers and pretty girls with it.  Even today it is an amazing camera and has only 4 ½ MP to work with.  But they are huge!

 

EB
EOS 1DX and 1D Mk IV and less lenses then before!

Unfortunately pixel size is only part of a great image. Technique plays a very important part, especially as the lens gets longer, even when using a tripod because it needs to be rock solid & without any play in any of it's parts when set & adjusted to height etc. AF accuracy is also very important, both in the body & the lens and for action work it also must be capable of very fast re focusing. There are countless great photos from just about every digital body produced by Canon & Nikon over the last few years when coupled with quality lenses. No matter whether it's a crop body or full frame body you still need a certain number of pixels on target to make a certain sized print, but if they are soft before cropping they will be even softer after cropping.

"A skill is developed through constant practice with a passion to improve, not bought."
Announcements