26 Megapixel 40D Compared to 35mm Film

[mauro_franic]

Some back ground on the origin of this article:

 

I shoot mainly 6x7 TMX and Velvia 50-100. For quick studio work I use a Canon 40D.

 

This month with the introduction of Ektar 100 I set out to use 35mm film which I had not used in a looong time to

test the new film. The results were excellent in detail, saturation and fine grain like I don't remember seeing

in color negative. It brought back my interest in 35mm - since I had stop considering the 40D an option for

landscape work, could I bring my 35mm SLR instead of my Mamiya when I needed portability....? It looked worth it....

 

Well, before sharing results, there is more to the story. Like always, we need new toys, so I went on B&H

looking for yet another 35mm SLR. To my demise, only two Canon models seem to be in production the EOS 1V and the

Elan 7ne. Without hesitation I ordered a 7ne seizing the moment in case soon they are no longer available new.

The camera is a joy, great build, IR remote built in, eye control, metal body, etc. Feels similar to the 40D

but a tad better with raised metal symbols and letters on the knobs instead of paint.

 

This morning I set out the make a quick comparison between the new Ektar 100, TMX and Velvia 50. I also shot the

40D from the same tripod, making it an equivalent to a 25.8 megapixel FF DSLR of great quality. I just developed

the TMX and will send the color film to the lab. TO MY SURPRISE, the TMX film under the microscope far out

resolves a 26 megapixel camera that still doesn't even exist. Even a simple scan of the film outresolves a 26

MP DSLR of the future in areas of high contrast.

 

The equipment:

 

Lens: Canon 50mm 1.4 at f8 was used for all shots for film and digital.

 

Digital camera: Canon 40D set on the same tripod with the same lens as the film camera. Making it a 25.8

megapixel equivalent of the 40D.

 

Film Camera: Canon Elan 7NE - new portable toy for landscapes.

 

Film: Ektar 100, TMX and Velvia 50.

 

Scanner: Coolscan 9000

 

Manfrotto tripod and head.

 

Lighting: A softbox from each side and a low power unbrella next to the camera for fill.

 

Here is the comparison of the digital shot and the TMX scan:

 

http://shutterclick.smugmug.com/gallery/6560295_hraSq#417496914_LLccY-O-LB

 

Conclusions are yours. I can email raw files to whomever emails me.

 

 

I will post the Velvia and Ektar scans after they are developed.

[sattler123]

How do you get to 26 MPs for the 40D? Based on the attached comparison, I'd say the 40D wins hands down. I don't see how you came to the conclusion that TMX wins in this comparison - maybe the pictures are just too small, but the 40D shots are way cleared and seem to have more detail (and certainly much less noise) than the TMX shots. What am I missing?

[mauro_franic]

As an additional reference, the negative seems to have aprox 30-40% more resolution under the microscope than in the scan.

 

e.g. All the small marks on d measuring tape are clearly defined. "Nutrition Facts" can be clearly read on the bottle label. The resolution chart resolves in between the 6th and the 7th mark.

[mauro_franic]

The 25.8 megapixels of the 40D are the result of shooting it on the same tripod as if it were full frame.

 

You are missing that this is not a win-lose article. Just an observation on detail.

[mauro_franic]

Also remember you are just looking at the scan. The film outresolves the scan I posted.

[rainer_t]

-- "TO MY SURPRISE, the TMX film under the microscope far out resolves a 26 megapixel camera that still doesn't even exist"

 

So, you'll be using TMX from now on and watch your results under the microscope rather than using a camera that doesn't yet exist? ... If that is what you're meaning, I would say "clever move".

[mauro_franic]

Gotta give you that sounds funny.

[frank_skomial]

According to dpreview :

 

http://www.dpreview.com/reviews/canoneos40d/

 

the 40D was only 10 MP camera few days ago...

 

what do the dpreview knows anyhow ?

 

Boy... the technology changes that fast.

[eric merrill]

"Even a simple scan of the film outresolves a 26 MP DSLR of the future in areas of high contrast."

 

To my eye, your samples from the 40D are "better" than the scanned tmax. I don't care about absolute

resolution of high contrast targets. My subjects rarely walk around with test charts taped to their

heads. :) Comparing your medium contrast pictures looks like a hands-down winner for the 40D just

based on noise alone. Same with low contrast.

 

So while technically true, your quoted statement is misleading.

 

 

Eric

[mauro_franic]

Another note: the grain on the scans is barely visible. The reason why you see it is because I applied sharpening off-the-charts (on both film and digital) to make the comparison on detail quick and easy - This sharpening is not to make the picture look good by any means.

 

This is what it looks like with milder sharpening:

http://shutterclick.smugmug.com/gallery/6560295_hraSq#417558785_ZecZh-A-LB

 

Obviously the TMX scan resolution extinguishes at the Coolscan's limit. That is the only reason why it stops at around "5". On the film it is around "6" or "7". Still, on the 40D-26MP moire starts after "4".

[kelly_flanigan1]

ERIC; come on now.:) One's subjects might be Football, Basketball or Hockey REFS! :)

[mauro_franic]

I also agree the 40D-26MP looks very very good.

[stwrtertbsratbs5]

I, too, nave abandoned cameras that do not exist.

[sattler123]

Call me stupid, but I still don't understand why shooting from the same tripod as the film camera makes the 40D a 26MP camera?

[steve_parrott]

I have read through this thread over and over. Maybe I am just dense, but for the life of me I don't understand how just putting the 40D on the

same tripod and using the same lens as the 35mm camera makes it 26 MP ???????

 

I guess I need to buy a tripod that has been used by someone with a Hasselblad H3D and then I could put a point and shoot digital camera

on it to equal it. Yeah... that makes no sense, but you stating the 40D is now 26 MP because you are using the same tripod and lens makes

about as much sense to me.

 

Again, maybe this is just over my head, but I DON'T GET IT!

[eric merrill]

The 40D is 3888 x 2592 with a sensor that is 22.2 x 14.8mm.

 

If that were extrapolated to a 36x24 sensor (aka "full frame") while keeping the sensor density the same,

the resolution would be a little over 6000 by a little over 4000 pixels. This would yield 26 megapixels.

 

At least, that's my interpretation of what Mauro is saying.

 

This is using a pixel size of about 6 µm, which is approximately the theoretical limit of resolving power

for current sensors. Much smaller than this (like the 50D's 5 µm) and you're just resolving diffraction.

Of if you use an f-stop smaller than f/8, you're also going to be seeing diffraction results at this pixel

size.

 

 

Eric

 

 

Eric

[mauro_franic]

FOR CLARIFICATION:

 

I modified the post so you can see the area captured by the 40D vs the 35mm camera.

 

http://shutterclick.smugmug.com/gallery/6560295_hraSq#417587578_MAq5N-X3-LB

[daniele_chiesa]

It's not that hard really: if you put the 40d with the same lens on the same position, you get a CROP of what would be a 26MP full frame image. So the 40d have the big disadvantage of missing part of the picture :)

At first i didn't get it either; took me about 2 minutes to figure it out.

I really look forward to see what the Ektar turns out to be; please post some samples when you can, Mauro.

[mauro_franic]

I too can't wait to compare TMX, Velvia and Ektar next to each other. Obviously TMX will be king in high contrast resolution, Velvia in low contrast resolution, and Ektar in D-range - price - convenience. But how much? I'll post when I receive them.

[mauro_franic]

Use this link to see at 100%. The above is at 75% -by mistake.

 

 

http://shutterclick.smugmug.com/gallery/6560295_hraSq#417587578_MAq5N-O-LB

[sandy_labana]

I do not know what good is looking under a microscope but I like the 40D images much better than the film. I must admit, I do not have

good eyesight. Sandy

[martyphotoarts]

Being somewhat new to DSLRs myself, I'm going to have to put mine on a tripod more often because without it, it's only a 12.3 megapixel engine. I simply didn't know there were all those other cylinders it could fire on.

[graybrick]

the 40d gives essentially no noise, where the grain at this magnification even in unsharpened scans will certainly be a factor, and at some point beyond the resolution of your Nikon scanner will become resolution-limiting (which seems evident here anyway). Whatever the maximum resolving capability of these films, they will not match the smooth clarity of the digital image and given the inherent lack of grain in the digital image are not comparable "apples to apples". That said, I still use film and digital, and I tend to prefer film for most subjects I shoot. At these magnifications, however, the digital gives a more pleasing image which is essentially the same evident resolution as the film.

[philip_maus]

Hm. I'm a big proponent of film, and I still think the 40D samples are far sharper and better resolved then the film. I don't quite get the 25MP assumption, but if I just take your word for it, I still don't see how you can say that the film outresolves the digital in this comparison. Maybe it's just the image quality of the JPEG you posted for comparison and you can't really see the difference on a photo displayed on a computer screen. As for the microscope, me neither. The closest I'll ever come to that is looking through a loup to decide what I want to print.

 

I, for one, do believe that one can take a film and a digital image, which have been 'captured' under the same conditions, and compare them side by side and it could be a close call as to which has the greater resolution. Especially if both images are made under close to ideal conditions. I like the idea of this test, but I don't think you're making your point with this image presented in this way.

[bill_keane2]

Amazing how the 40D blows the film out of the water.