Can the 5D cope with Richo GX100 (P&S) for DOF Image-Quality ?

http://aminphoto.blogs​pot.com …at-digital-ever-have.html

It took me quite a while to get the author's drift ...

He makes good points. I had recognized the macro implications for some time. My mother takes hand held flower pictures in her garden all the time with a P&S while I'm lugging a tripod and shooting at f/11.

Razor thin DOF is not always desireable. The APS-C dSLRs are in some respects a nice compromise. They can get shallow DOF with primes when it is wanted (unlike most P&S) but they don't have to be stopped down as much going the other way. My 10-22 is focussed at infinity for the most part even when it is wide open.

I just downloaded his raw test images, the 5D shot is severly underexposed by just over 4 stops & bringing up the exposure when at iso3200 makes it pretty hard to see what is going on, what was he thinking?

(update): looking through all the noise I can see there is much more dof on the 5D shot over the other one, hmm strange, maybe ive missed the point also

Nick_C wrote in post #4430481
I just downloaded his raw test images, the 5D shot is severly underexposed by just over 4 stops & bringing up the exposure when at iso3200 makes it pretty hard to see what is going on, what was he thinking?

(update): looking through all the noise I can see there is much more dof on the 5D shot over the other one, hmm strange, maybe ive missed the point also

Hi, I posted that blog post and would be happy to address your points. Based on the format sizes of the two cameras, one can calculate the f-number needed to get the same DOF for each given a fixed subject distance, output size, diagonal angle of view (the diagonal is used here rather than height or width since the aspect ratios differ). When f/18 is used on the 5D, f/3.9 will give the same DOF using the Ricoh (assuming fixed subject distance, output size, and diagonal AOV). The DOF appears greater with the 5D mainly because the aspect ratio is 3:2 vs 4:3, thus including less vertical angle of view, and secondarily because the vantage point for the 5D is slightly higher since each camera was sitting on the ground using a self timer for the shot (the 5D is taller). You can check for yourself using an independent source such as DOF Master that the settings I used give the same DOF for the two systems. The underexposure of the 5D is a consequence of the fact that it has to be stopped down just over 4 1/3 stops to get the same DOF, and that this test required each camera to use the same shutter speed. Thus, the ISO had to be maxed (at ISO3200, which is two stops more than the Ricoh) and then pushed just over 2 1/3 stops additionally to match the apparent exposure of the Ricoh. Not sure what app you are using to push it an additional 4 stops, but that results in overexposure in my case. In fact, I found it interesting and consistent that Lightroom, on its own, chose +2.4 stops when I simply selected auto exposure.

Why did this test require the same shutter speed? And why was the ISO speed allowed to change? I don't get the point of this execise. I'm no genius, but given everything I know about photography (admittedly it's not everything...) tells me that in order to get the same comparison picture in terms of DOF and exposure, I would have to leave the ISO capped at 800 and adjust the shutter speed to get a GOOD exposure. I'd certainly need a tripod for the 5D which is to be expected... but it would be the same image in terms of exposure and DOF, but I gaurantee that image quality results would be much different.

There's no right way to do such a test. You might find more than a certain level of noise acceptable for each camera and cap the ISO accordingly. I was doing a test with low light street photography in mind. In those circumstances, I often encounter moving subjects (hence shutter speed matters if one wants the picture a certain way) and deep DOF. Clearly capping the ISO and varying the shutter speed favors the 5D, but we don't need a test to prove that! In the great majority of situations, the 5D will have better image quality than a small sensor camera. What I found interesting was that there is one type of situation - one which I happen to encounter periodically - in which a small sensor camera may prevail.

Nick_C wrote in post #4430481
I just downloaded his raw test images, the 5D shot is severly underexposed by just over 4 stops & bringing up the exposure when at iso3200 makes it pretty hard to see what is going on, what was he thinking?

How did you figure that out [4 stops underexposure] from the RAW ?

roli_bark wrote in post #4430624
How did you figure that out [4 stops underexposure] from the RAW ?

Here's what the 5D RAW looks like pushed in Lightroom.

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by FLICKR

asabet wrote in post #4430614
What I found interesting was that there is one type of situation - one which I happen to encounter periodically - in which a small sensor camera may prevail.

A "fairer" comparison to Richo's smaller sensor would be using a wider lens [thus getting higher DOF] for the 5D, setting SAME shutter/ISO, and cropping the image accordingly.

JeffreyG wrote in post #4430457
He makes good points. I had recognized the macro implications for some time. My mother takes hand held flower pictures in her garden all the time with a P&S while I'm lugging a tripod and shooting at f/11.

Razor thin DOF is not always desireable. The APS-C dSLRs are in some respects a nice compromise. They can get shallow DOF with primes when it is wanted (unlike most P&S) but they don't have to be stopped down as much going the other way. My 10-22 is focussed at infinity for the most part even when it is wide open.

That's true. Different strokes for different folks.
There are definitely applications where a small sensor camera is a better choice & fit. I for one, carry one in my shirt pocket all the times when I go out, regrdaless of weather, & circumstances. Can't say that for my DSLR ...

roli_bark wrote in post #4430644
A "fairer" comparison to Richo's smaller sensor would be using a wider lens [thus getting higher DOF] for the 5D, setting SAME shutter/ISO, and cropping the image accordingly.

That would be an interesting test, and I'm confident it would favor the 5D; however I don't see how it would be any fairer (unless you define "fairer" by which test favors the 5D ). I, for one, don't like using a 24mm lens to compose when my intended print has an angle of view corresponding to a 50mm lens. If you don't mind doing that, then the comparison done your way would make sense. Still, I like your idea, so I'll try to post that test on my blog later this week. A 24mm lens on the 5D should give only very slightly less DOF than a 10.5mm lens on the Ricoh. Although the 5D will be using just over 1/4 of its resolution, I still predict it will come out well ahead in this test.

Correction: Cropping and magnifying the 5D image would decrease the DOF, so I need to figure out how to properly do this next test that Roli proposed...

I used CS3, when I said 4 stops it was only a rough guide, even with 4 stops there are only a few areas that are blown, the carpet in front is ok..

I always assumed it was common knowledge that a P&S at say f/3.2 is more like say f/16 or so on a DSLR, its also true that you should be able to get more DOF with a P&S, but there are so many factors that are bound to make the overall image worse, no P&S in my view will ever come close to the detail of a DSLR with decent lens, also on all the ones ive owned the AF has been very slow & I remember on my Powershot Pro1 seeing the yellow LED flashing meaning no focus lock was a consistent sight.

HOSTED PHOTO
please log in to view hosted photos in full size.

[QUOTE=asabet;4430796]​That would be an interesting test, and I'm confident it would favor the 5D; however I don't see how it would be any fairer (unless you define "fairer" by which test favors the 5D ). I, for one, don't like using a 24mm lens to compose when my intended print has an angle of view corresponding to a 50mm lens. If you don't mind doing that, then the comparison done your way would make sense. Still, I like your idea, so I'll try to post that test on my blog later this week. A 24mm lens on the 5D should give only very slightly less DOF than a 10.5mm lens on the Ricoh. Although the 5D will be using just over 1/4 of its resolution, I still predict it will come out well ahead in this test.

Correction: Cropping and magnifying the 5D image would decrease the DOF, so I need to figure out how to properly do this next test that Roli proposed...[/quote]

If you were to take the 5D shot from farther back, then crop to the same view, this would give more DOF.

asabet wrote in post #4430632
Here's what the 5D RAW looks like pushed in Lightroom.

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by FLICKR

I think I know what's going on here. I'm seeing less pushing because of the brightness setting in LR (added automatically to both Ricoh and Canon files during autoexposure). Either way, what at stake here is not whether Canon and Ricoh are using ISO values the same way. When you set both cameras to the same shutter speed, framing, DOF, and output size, then whatever ISO is needed to get the same apparent exposure is what one needs for this particular test. If that is +4 in your method, then so be it. In my processing algorithm, the histograms lined up at +2.4.

Yes its too much brightness on that one, although I get your point, this isnt what you were demonstrating!

asabet wrote in post #4430796
That would be an interesting test, and I'm confident it would favor the 5D; however I don't see how it would be any fairer (unless you define "fairer" by which test favors the 5D

By "fairer" I mean same shooting conditions. Same shooting conditions are, among other things, same perspective. Same perspective can be achieved only by using a wider-angel lens for the larger sensor camera.