Johnny Johnson wrote:
> At 06:13 PM 7/6/2006, Moose wrote:
>
> Sorry for the late reply. I needed to run get a haircut, pick up
> some fertilizer at the farmer's exchange, eat dinner and cut the front lawn.
>
I didn't know there was a hurry.
Just to get it out of the way, I've found some lens tests where I did a
pair of shots with the 300D and 5D of a distant, real world subject at
300 mm. Set-up was a tripod with QR, the two bodies with plates. Bodies
were swapped on the tripod and the lens swapped on the bodies. So
everything was identical but the bodies and whatever light changes
occurred over 5-10 minutes.
Both shots at iso 400, f8. Both processed in DPP at camera defaults
<http://www.moosemystic.net/Gallery/tech/Wires1.htm>. I left the samples
at their original sizes so you can see visually why I resampled before.
The subject was obviously picked for artistic merit. :-) Or maybe
because it had some fine detail in the center and a wide range of
brightness, texture, etc.
It still looks to me like the 5D reveals more detail (avoiding the word
resolution) in the wires, roof and foliage and separates the wires from
the foliage more clearly. It's certainly not a big difference, but I
think it is at least clear that the 300D doesn't display more details
than the 5D. Sorry I couldn't get a bird to sit still for lengthy tests.
;-)
I suppose in theory I should have shot at iso 100, as the greater noise
of the 300D does have an effect. But again, this is a test for me and
I'm not going to be taking 300 mm shots handheld at iso 400 very often,
let alone 100. So it is a practical test for me.
> <snip>
>> That was just a quick test, and might have
>> been a preliminary to something more formal. Except the difference in
>> sharpness was so striking that I haven't bothered to go further so far.
>>
> But, you referred to it as "resolution". Aren't resolution and
> sharpness two different things? In your test aren't you looking for
> resolution first but resolution that can be sharpened to a pleasing degree?
>
All true. and to fully test that, you need MTF curves or at least
resolution line tests at both high and low contrast. On the other hand,
shooting with two cameras from the same manufacturer, with related
sensors and processors, both at their default settings, I don't
necessarily need more than a few pics to come to conclusions that seem
useful to me.
>> The 10D and 300D sensors are the same, I believe.
>>
>
> The same size and pixel pitch in any case.
>
And tested noise and resolution on dpreview.
> <snip>
> I'm still not sure why you went through all the resampling on each
> image.
It's a simple visual thing, people tend to find a bigger image better
and it's harder to compare images of different subject sizes as in the
above samples. In the original test, I did one sample in one direction
and the second in the other just to even out any resampling effects.
> I've got to admit that I skipped over your original message
> so there must be something in there that explains what you set out to
> test and it must be different than I thought. The question that I
> thought you had was: If, for example, I'm using a 300mm lens to take
> a picture of a bird and my position is fixed so that I can't zoom
> with my feet and the image size is smaller in the 5D viewfinder than
> I'd like - would I be better off switching to the 300D body with its
> 1.6 "crop factor" or would I be better off continuing with the 5D and
> cropping the image later in software?
That is indeed my question.
> If that was/is the question it
> would seem to me that all you have to do in the test is to shoot the
> same subject from the same distance with the same lens at the same
> focal length with each body and process the images using your best
> methods for each and then compare the important detail at 100% (sorry
> ag - if you're still listening) on the screen.
>
That is about what I did, except I used camera defaults and did no post
adjustment to either image in the posted samples. Everything about the
setup was identical except for a swap of camera bodies.
> (BTW, the real answer to the question is "neither of the
> above". Actually you should mount the 300mm lens on a body like the
> E1 with a 2x crop factor.) ;-)
>
Won't fit!
> <snip>
>> Theory says one
>> should be sharper and the simple test shows the opposite. Maybe I have a
>> particularly lame 300D and/or a particularly great 5D, I don't know.
>> Maybe the sharpening on the 5D is different than the 300D, but I don't
>> see how I could sharpen the 300D image to equal the 5D. I seem to recall
>> that I tried that briefly.
>>
> In this discussion you haven't talked about such things as Raw vs
> JPG, in-camera sharpening settings, sharpening settings in Raw
> processing software (if you shot RAW), any USM applied later, noise
> reduction settings in camera or other settings or work flow that
> might affect the image quality.
Everything shot in RAW and processed in DPP at the camera default settings.
> You did talk about resampling that
> could degrade the image quality. I'll assume that you used camera
> settings and work flow for each body that have proven to give you the
> best results. So, looking at the simple pixel pitch numbers would
> lead one to believe that the 300D would have a higher lines per mm
> resolution than the 5D, the DPReview tests show that this is true, my
> tests with the 10D show that this is true, other tests that I've read
> on the web over the last 6 months say that it's true yet you've found
> just the opposite. As you mentioned above - there could be a several
> reasons for this difference.
>
> a) You have an above average 5D.
> b) Your 300D is below average.
> c) You've found a superior work flow for the 5D.
> d) Your 300D work flow/camera settings didn't show the 300D to its
> full potential.
>
Same work flow for each. I had lots of experience working with the 300D
and essentially none with the 5D. Also, I have done my RAW processing
for the 300D in ACR, and I don't have the version that will do the 5D.
So for a first test, I stayed away from playing with what came out of
the camera. Hard to believe of me, no?
> e) I don't have a clue as to what I'm talking about. (This may be
> the most likely pick, especially by the "I agree with Moose" t-shirt
> wearers.) :-)
>
Seems unlikely.
>> In any case, results are better than I expected, so I'm happy so far.
>>
>
> And in the end that's what counts the most. Based on your work flow
> you found what is best for you and you're happy with it. I hope you
> realize that I my intent was not to rain on your parade.
Heaven's no! How are we going to learn if we don't challenge each others
ideas and work?
> I probably should have kept my mouth shut like I did the first time the
> subject
> came up. I'll try to do better next time. ;-)
>
Please don't.
Moose
==============================================
List usage info: http://www.zuikoholic.com
List nannies: olympusadmin@xxxxxxxxxx
==============================================
|