Pointless E-3 / 5D comparison 2

dgilder · Nov 16, 2007

If you missed the original see here: http://forums.dpreview.com/forums/readflat.asp?forum=1022&thread=25627233

Same caveats apply, FF vs 4/3 isn't comprable, its a waste of my time, primes vs zooms is useless, handcapping one cam or the other invalidates everything, blah blah blah. See the other thread for plenty more reasons. Please refer to the word "Pointless" in the subject before posting about the pointlessness of doing this.

I went to the state capitol today to do a head to head between the cameras, but in a more friendly lighting environment.

The Gear:

Oly E-3 w/ 12-60mm
Canon 5D w/ 50mm f1.8 (metal mount)
Tripod

The Rules:

E-3 at 29mm to better mimic 5D's magnification at 12MP with a 50mm (28mm would have been better though).

Both cameras in aperture priority mode set to f11, with auto metering.

Autofocus on the star, using center AF point.

E-3 at +0.5 exposure comp, and 5D at +0.3 - This is to even out the 5D's slightly hot ISO values.

ISO 100, 200, 400, 800, 1600, and 3200 compared.

Auto white balance

Photoshop CS3 with ACR used to process, with ACR default values, no tweaks etc. This means 0 Luminous noise reduction, and 25 color noise by default.

The Combined Results (you'll need to look at the full size to really see the difference:

Full size: http://davidgilder.com/misc/5DE3/E35D_Dome.jpg

Thoughts:

Using ACR to process the photos, we can really see how good of a job the E-3 is doing at ISO 1600. It is only slightly noiser than the 5D at that point. The 5D has the clear advantage at ISO 3200, the E-3 has many artifacts in its noise. No surprises here, the 5D has its reputation for low light capabilities for a reason.

What does surprise me is how well the new Oly noise reduction algorithms work, comparing this to the shots processed in Olympus Master. The fact that the camera can run these same algorithms on JPEGs on the fly really speaks to the processing power of the camera.

Side Notes:

This was inspired by a series someone shot with one of the pre production cameras of a church ceiling (or something like that). I don't recall who that was but this is just a repeat of that concept, but with both cameras.

The articulating live view came very much in handy while shooting some other shots of the dome. I used the zoom option of live view with manual focus and its really nice to be able manually focus without having to be upside down on the floor with my neck bent:

dgilder · Nov 16, 2007

Whoops, two other quick notes:

For some reason the 5D jumped from 100 to 200 on the change to ISO3200, this made that section a tiny bit darker than the rest of the 5D series. Not sure on that, might have been a flash or something from a tourist while it was exposing.

The full size is a JPEG that was saved at level 12 from photoshop direct from the combined RAW files, so its not really extreme compression JPEG artifacts causing the problem in the 3200 on the E-3 side of things.

Torsten Hoff · Nov 16, 2007

Two thumbs up on the execution and for doing all that work! Not pointless at all, and much appreciated.

dgrogers · Nov 16, 2007

The E-3 looks slightly more detailed, even at the higher ISO's.

This is actually fairly interesting. Don't think there's a big difference in IQ between either camera, but the E-3 is certainly a little punchier (it was the opposite in the first test).

dgilder said:
If you missed the original see here:
http://forums.dpreview.com/forums/readflat.asp?forum=1022&thread=25627233

Same caveats apply, FF vs 4/3 isn't comprable, its a waste of my
time, primes vs zooms is useless, handcapping one cam or the other
invalidates everything, blah blah blah. See the other thread for
plenty more reasons. Please refer to the word "Pointless" in the
subject before posting about the pointlessness of doing this.

I went to the state capitol today to do a head to head between the
cameras, but in a more friendly lighting environment.

The Gear:

Oly E-3 w/ 12-60mm
Canon 5D w/ 50mm f1.8 (metal mount)
Tripod

The Rules:

E-3 at 29mm to better mimic 5D's magnification at 12MP with a 50mm
(28mm would have been better though).

Both cameras in aperture priority mode set to f11, with auto metering.

Autofocus on the star, using center AF point.

E-3 at +0.5 exposure comp, and 5D at +0.3 - This is to even out the
5D's slightly hot ISO values.

ISO 100, 200, 400, 800, 1600, and 3200 compared.

Auto white balance

Photoshop CS3 with ACR used to process, with ACR default values, no
tweaks etc. This means 0 Luminous noise reduction, and 25 color
noise by default.

The Combined Results (you'll need to look at the full size to really
see the difference:

Full size: http://davidgilder.com/misc/5DE3/E35D_Dome.jpg

Thoughts:

Using ACR to process the photos, we can really see how good of a job
the E-3 is doing at ISO 1600. It is only slightly noiser than the 5D
at that point. The 5D has the clear advantage at ISO 3200, the E-3
has many artifacts in its noise. No surprises here, the 5D has its
reputation for low light capabilities for a reason.

What does surprise me is how well the new Oly noise reduction
algorithms work, comparing this to the shots processed in Olympus
Master. The fact that the camera can run these same algorithms on
JPEGs on the fly really speaks to the processing power of the camera.

Side Notes:

This was inspired by a series someone shot with one of the pre
production cameras of a church ceiling (or something like that). I
don't recall who that was but this is just a repeat of that concept,
but with both cameras.

The articulating live view came very much in handy while shooting
some other shots of the dome. I used the zoom option of live view
with manual focus and its really nice to be able manually focus
without having to be upside down on the floor with my neck bent:

--
dgrogers

http://www.pbase.com/drog

kuaimen · Nov 16, 2007

All thumbs up!!! This is really nice to see for an ordinary person like me - all those scientific numbers make ordinary people even more puzzled, so they just go for larger pixel count - something easy to understand and fits the "more is better" stereotype perfectly.

Well, this is not to say more is not better ...

Torsten Hoff said:
Two thumbs up on the execution and for doing all that work! Not
pointless at all, and much appreciated.

Torsten Hoff · Nov 16, 2007

One other thing I'm pleased to see is that the color shift with increasing ISO that was reported elsewhere appears to be totally absent in the release firmware.

Claus A · Nov 16, 2007

David,

Wow, this 5D is a fine camera.

The point, IMHO, is not how well th E3 compares to this cam, but: that it is second in every respect. Is it just for the lenses, weather protection and the familiar user interface with the E3 now??

One side note: Developing the RAWs of my E510, I have experienced that it's better to work on luminosity noise, not chroma. Did you try that on the E3's RAWs?

Thanks a lot for that second shooting,

best,

Claus.

--

... when the photograph annihilates itself as medium to be no longer a sign but the thing itself...

stimmer · Nov 16, 2007

They may not have any point but they are keeping me entertained. Nice comparison. I like the live view example there. That is going to be really useful for me, way more than I thought.

The high iso is just a nice little bonus.

lloyd007 · Nov 16, 2007

Huh? Maybe it's just me but the only difference I can see at iso 100-400 is that the 5D has a bit lighter exposure and I really had to stare to see any difference at iso 800. At 1600. At even iso 3200 and some cleaning up I don't see how there will be any difference unless you are really printing 30x40

dgilder · Nov 16, 2007

kuaimen said:
Well, this is not to say more is not better ...

Actually, the 5D has 2912 vertical pixels, the E-3 has 2736. 2912-2736=176, 176 * 3648 (E-3 horiz pixel count) = 642,048. 4368 (5D horiz) * 2912 = 12,719,616; 642,048 / 12,719,616 * 100 = only 5.05% of the 5D's extra pixels are on the vertical axis. Those 2MP are really mostly on the ends, and usually end up being cropped off for a lot of print sizes. So the 5D's bigger is really just longer.

dgilder · Nov 16, 2007

Claus A said:
The point, IMHO, is not how well th E3 compares to this cam, but:
that it is second in every respect. Is it just for the lenses,
weather protection and the familiar user interface with the E3 now??

I wouldn't be comfortable saying the E-3 is second to it in every respect, just a few really, extreme ISO (and even there the E-3 is doing well) and total pixel count. The E-3 is more affordable (at the moment), has higher FPS, extremely high build quality, nearly identical vertical resolution, faster access to most camera features, and the sales reps will pour water over it for you

. It does interest me though to see where the upcoming 5D's replacement ends up being priced and what features it might have. Even so, each tool has its purpose, which is why both are in my kit.

Claus A said:
One side note: Developing the RAWs of my E510, I have experienced
that it's better to work on luminosity noise, not chroma. Did you try
that on the E3's RAWs?

Well yeah, it works just fine and they clean up well. The color and luminous noise settings were just the ACR default. I was showing both cameras at ACR's default so you could properly compare them. If I wanted to show best case for each camera, I would develop the RAW files in DPP and Oly Master respectively and then compare.

Claus A · Nov 16, 2007

Hi,

You have downloaded the big image?

Look at the star in the center of the dome, for example (ISO400/800), or the stucco details (left, ISO 100).

Take into consideration that the E3 image is a little bit more zoomed in (58mm equivalent, as described in the posting, to compensate for the difference in resolution).

The E3 image is just great. That of the 5D simply better.

I'm not familiar with the Canon lenses - is this 50mm 1.8 a pro lens? Like the 12-60mm Zuiko?

Best,

Claus.

--

... when the photograph annihilates itself as medium to be no longer a sign but the thing itself...

dgilder · Nov 16, 2007

stimmer said:
They may not have any point but they are keeping me entertained.
Nice comparison. I like the live view example there. That is going
to be really useful for me, way more than I thought.

It really is, I forgot how handy that was on my old E-10 (my sister in law has had it for years), and even how useful the articulating screen can be. The live view is going to be one of my favorites for product photography.

Claus A · Nov 16, 2007

dgilder said:
The E-3 is more affordable (at the moment), has higher FPS, extremely high build quality, nearly identical vertical resolution, faster access to most camera features, and the sales reps will pour water over it for you . It does interest me though to see where the upcoming 5D's replacement ends up being priced and what features it might have. Even so, each tool has its purpose, which is why both are in my kit.

That's my perception, too.

Could you please tell me about the 50mm f/1.8 lens you used on the 5D?

Thanks again,

best,

Claus.

--

... when the photograph annihilates itself as medium to be no longer a sign but the thing itself...

dgilder · Nov 16, 2007

Claus A said:
Take into consideration that the E3 image is a little bit more zoomed
in (58mm equivalent, as described in the posting, to compensate for
the difference in resolution).

Yep, should probably be 56mm equiv to match the 5D's extra vertical pixels and the Oly not-quiet-2x FOV factor. Its hard to get it dead on too, since the 12-60 markings jump from 25 to 35.

Claus A said:
I'm not familiar with the Canon lenses - is this 50mm 1.8 a pro lens?
Like the 12-60mm Zuiko?

Heck no! This is one of the cheapest lenses Canon makes. BUT! There is a huge caveat here because optically it is a fine piece of glass. 50mm lenses are one of the easiest to manufacture, require less elements and don't need any special elements that a zoom needs. Its a prime lens, and therefore has a lot of advantages over a zoom, and its just not as expensive to produce. The version I have is also the metal mount version which isn't made anymore and is supposedly better than the current 50mm f1.8 which have a plastic mount. All in all, it is still a very good lens, even if not a "pro" lens.

Robert999 · Nov 16, 2007

Thank you for these (I think you win DPReview for the month, btw)...I have to agree with Claus, the 5D looks better than the E-3 to my eyes, on my monitor, at full size, at every ISO.

But not by very much at all, not really much at all.

To get the build quality of the E-3 in an equivalent Canon FF I would have to buy a IIId, which let's face it doesn't have a good rep right now at all. And it WOULD give me slightly better pictures (if the 5D results are extrapolatable, which they should be!). It would also cost double. For the slight difference in image quality, it is not worth it to me. It's a trade-off, but one that leads me back to the E-3 as my next camera.

Robert

dgrogers · Nov 16, 2007

Claus A said:
I'm not familiar with the Canon lenses - is this 50mm 1.8 a pro lens?
Like the 12-60mm Zuiko?

The 12-60 Zuiko is a mid grade lens. The 50mm 1.8 is cheap but comparable to much more expensive lenses (its main sacrifise is build quality). Besides, it's a prime lens which in theory should be sharper than a zoom.

--
dgrogers

http://www.pbase.com/drog

lloyd007 · Nov 16, 2007

It really isn't as simple as you make it out. When I have to squint and give myself a headache to see an absolutely minuscule difference (and I truly cannot see any difference at iso 100 sorry) from 100-800 it really doesn't matter to me. And then at 1600 and 3200 if you cleaned up the E-3 image it could be pretty comparable.

Edit: I am of course talking about details not the obvious differences in color of which makes the 5D's images look 'better' imo.

dgrogers · Nov 16, 2007

Robert999 said:
Thank you for these (I think you win DPReview for the month, btw)...I
have to agree with Claus, the 5D looks better than the E-3 to my
eyes, on my monitor, at full size, at every ISO.

But not by very much at all, not really much at all.

In the past, ACR has been pretty bad with Oly files (it never got the E-1 right), and I'm sure the E-3 results would be better processed in Studio (the difference between Studio and just about every other RAW converter for the E-1 was huge). I don't know how well ACR does with Canon either, so it may be that both could do better.

Next up should be E-3 processed in Studio and the 5D processed in whatever the best converter for Canon is. Perhaps if we ask dgilder nicely?

--
dgrogers

http://www.pbase.com/drog

dgilder · Nov 16, 2007

dgrogers said:
In the past, ACR has been pretty bad with Oly files (it never got the
E-1 right), and I'm sure the E-3 results would be better processed in
Studio (the difference between Studio and just about every other RAW
converter for the E-1 was huge). I don't know how well ACR does with
Canon either, so it may be that both could do better.

The consensus I've seen is that ACR doesn't do as well with the Canon RAW either, which is why I compared them there in neutral territory. My previous comparisons were to Oly's advantage because the only thing that could process them was Oly Master until last night.

dgrogers said:
Next up should be E-3 processed in Studio and the 5D processed in
whatever the best converter for Canon is. Perhaps if we ask dgilder
nicely?

Maybe, I'm going to be shooting a lot this weekend, which means processing a lot next week, so I don't know when I'll get a chance, but yes, it would be interesting, though it doesn't really apply to me because I use ACR in my workflow and I hate processing in order to be able to process

Pointless E-3 / 5D comparison 2

Well-known member

Well-known member

Veteran Member

Veteran Member

Forum Enthusiast

Veteran Member

Veteran Member

Veteran Member

Senior Member

Well-known member

Well-known member

Veteran Member

Well-known member

Veteran Member

Well-known member

Leading Member

Veteran Member

Senior Member

Veteran Member

Well-known member

Similar threads

Keyboard shortcuts