5D Issues?

Sp00ks

A little background. I recently purchased a 5D used from a friend. He gave me a pretty good deal so I took it. I had a week with the camera prior to purchase and I played around with all my lenses.

I purchased the camera anyway but this is what I am noticing. I have a 50mm f/1.4 and a 17-40L and the sharpness of these on the 5D is rather remarkable. Now, this is where I am running into problems and need some opinions. I am less than impressed with the image quality of the 5D coupled with my 70-200 f/2.8 and my 100-400. It seems my 30D is much sharper. I have noticed through the years there there is definitely a sweet spot with lenses on my 30D and each lens is different. Either I can't find this sweet spot or I am having focusing issues with my longer lenses.

My 100-400 is still under warranty and I have thought about sending it to canon with the 5D and having them calibrate them together. I am not sure what this costs or how long it takes but I am highly considering it.

All of this might just be me not used to the new camera body and I'm open to that. However, I did not think the learning curve would be this steep. In defense of the 5D some of this was taken in low light situations, late afternoon and some shadow. (those are the times I have available to shoot) I don't have this kind of trouble with the 30D however.

What do you guys think? I could post some examples but I just want some preliminary thoughts from those that have experience jumping from crop to full frame and or experience with the 5D.

Thanks in advance for your replies.

ziggy53

The Canon 20D/30D/5D all share a very similar autofocus section. The main difference is that the 5D autofocus is more condensed towards the center of the image. (The autofocus module is about the same size on each of these cameras, but the image capture size is larger on the 5D, so the autofocus area within the image size is smaller as a result.)

Please walk us through your autofocus setup and procedure as that could impact the results.

Sp00ks

ziggy53 wrote:

The Canon 20D/30D/5D all share a very similar autofocus section. The main difference is that the 5D autofocus is more condensed towards the center of the image. (The autofocus module is about the same size on each of these cameras, but the image capture size is larger on the 5D, so the autofocus area within the image size is smaller as a result.)

Please walk us through your autofocus setup and procedure as that could impact the results.

Center focus point always, then recompose if necessary or compose in post processing. I assume that is what your asking for?

ziggy53

Sp00ks wrote:

Center focus point always, then recompose if necessary or compose in post processing. I assume that is what your asking for?

Center focus dot should be fine for the longer lenses.

Please do take a few images, in RAW, using a tripod and a single solitary subject and switch between the 30D and the 5D. Process both sets of images using the same RAW processor and technique. In PhotoShop, or another competent image editor, resample the 5D images down to the same image width, in pixels, as the 30D image width.

Now compare the two images, remembering that the 5D will have a much wider FOV than the 30D at the same lens millimeter settings.

This should give you an excellent comparative sampling of images.

If you still see a difference then yes, you might need to send the camera and lenses in for calibration. Remember that recalibrating the lenses may shift the lenses so that they may no longer be matched to the 30D as well, so ideally you would send all of your lenses and bodies together for recalibration.

I would expect to pay $500-$600 for multiple bodies and lenses for recalibration. It could be more if they find an actual defect somewhere. In-warranty products might cost less.

nightshadow

Won't the resampling introduce differences that a straight crop would not, or am I missing something?

Sp00ks

Ok, I'll try and set up a test bed. I'm an engineer so I should be able to come up with something that is somewhat accurate.

I did a shot of a spreadsheet with the 50mm stopped down a little, on a 45 degree angle and captured what I would expect. I couple of lines in the middle of the spreadsheet are sharp and in focus then above and below those lines were blurred. The entire width of the lines were in focus. I can't post it here because it has names and addresses on it.

nightshadow

First, I would like to propose a selfless solution to your problem: Send me the 5d and I will send you another 30d body. Seriously, somewhere (I think it is on here) there is a focus target in a post that deals with under/over focus and Sigma lenses. I will look to see if I still have a copy of the target on one of the computers here.

nightshadow

This is, or is very similar to the chart I was talking about. The link goes to a 19 page document which contains and explains the use of the chart.
http://focustestchart.com/focus21.pdf

Sp00ks

Ok, I guess a nevermind is in order. It has to be me and not the camera or the 5D just doesn't like low light coupled with a zoom. More than likely it is me.

5D 100-400L


100% Crop of the above image.

nightshadow

very nice

Sp00ks

nightshadow wrote:

very nice

Thank you. That is what I was expecting to see from the 5D.

ziggy53

nightshadow wrote:

Won't the resampling introduce differences that a straight crop would not, or am I missing something?

I was trying to accomodate the different photo site sizes between the two cameras.

The 5D has larger photo sites compared to the 30D (8.2 microns vs 6.4 microns). Resampling from the 4368 x 2912 of the 5D to the 3504 x 2336 of the 30D should make viewing at 100 percent on a computer monitor more equal in terms of image detail, but it will never be identical of course.

Viewing a Canon 5D at 100% on a monitor and comparing that to a 30D image, also at 100%, will always be difficult because the two cameras are dissimilar in the area of the imagers, the photo site size of the 2 imagers and the gross pixel count of the imagers (which also affects the pixel density).

nightshadow

I forgot about the comparative size of the photo sites Ziggy. I figured that I had missed something.

Photoskipper

I have been using the 5D for more than 18 months and so far happy with the picture it delivered.

Many of my friends compares the 5D FF with other crop sensor bodies. One thing we may need to look into the effective number of pixal per mm on the sensor:
1. 5D has 12.8 MP and resolution at 4368 X 2592 on the sensor size of 35.8 mm X 23.9 mm it means there is 122 pixal /mm horizontally and 108 pixal / mm vertically.
2. 30D is 8.2 MP with 3504 X 2336 on the sensor 22.5 X 15 mm it means there are 111 pixal /mm horizontally and 155 / mm vertically.
3. 400D is 10.1 MP with 3888 X 2592 on a sensor 22.2 X 14.8 it means there are 175 pixal / mm on both side.

If we compare the picture of the same lens at the same focal lenght and distance. 5D may loss out to 30D by resolution.

122 pixal x 22.5 = 2745 horizontally and 108 X 15 = 1620 vertically, if you crop it according to the same size as the picture of 30D.

The advantage of the FF in 5D is the senor of pixal is bigger, the microlens on the senor are bigger, so that it is more sensitive to light (the bigger microlens can capture more light energy) so the color is more dynamic and lower noise level.

ziggy53

Photoskipper wrote:

...
If we compare the picture of the same lens at the same focal lenght and distance. 5D may loss out to 30D by resolution.
...

Correct, and that's why I suggested the software interpolation. A different set of sensors would have required a somewhat different correction.

Photoskipper wrote:

...
The advantage of the FF in 5D is the senor of pixal is bigger, the microlens on the senor are bigger, so that it is more sensitive to light (the bigger microlens can capture more light energy) so the color is more dynamic and lower noise level.

Additional advantages to the full-frame sensors are outlined here by LiquidAir:

http://www.dgrin.com/showpost.php?p=538634&postcount=3

... with the full discussion here:

http://www.dgrin.com/showthread.php?t=59953