Jump to content

CL vs Medium Format


Recommended Posts

Advertisement (gone after registration)

I don't think for a second this is a fair comparison, so please don't point that out to me.  I fully understand that the CL has completely different uses than a slow, medium format camera.  Since I recently purchased a Hasselblad X1D, though, I was curious to see just how big the gap is in pure image quality between a 50 megapixel medium format camera with a good lens and a 24 megapixel APS-C camera with a good lens. I a set of pictures in my back garden.  Here is the background:

 

- Static subject with essentially no motion

- Tripod mounted with self timer to reduce vibration

- Mechanical rather than electronic shutter (shouldn't pose an issue for either given the solid tripod and short focal length lens)

- Dynamic range was well within the capabilities of either camera, so no advantage to the Hasselblad there

- Base ISO

- Optimum aperture for each camera/lens combo (f/4 for the 23mm and CL, f/8 for the 45mm and X1D)

- Equivalent (approximately) depth of field for each camera/lens combo

- Focus stacking to gain the desired depth of field for the image (from foreground leaves to the important areas of the subject, but enough blur to minimize the distraction of the background)

- X1D cropped very slightly due to the different focal lengths that resulted from focus stacking (47 megapixels vs. 51)

- CL cropped much more significantly to get to the 4:3 aspect ratio of the larger camera

- Both cameras had very good lenses, but not the best that either system can offer.  I used the 45mm XCD on the Hasselblad and the 23mm TL on the Leica.  Both give similar fields of view (35mm equivalent)

 

I processed the X1D images to my own taste and to the best of my ability.  Raw files were imported into Lightroom.  White balance was set manually.  Small adjustments to exposure, contrast, highlights, shadows, and black point.  A bit of local contrast through the clarity slider to make the details in the rust "pop" a bit.  No adjustments to individual colors.  Embedded color profile.  Embedded lens profile for vignetting and distortion (though distortion is a non-issue on this type of shot).  I took three separate exposures for each final image for focus stacking.  The first image used the eye of the pig as the focus point.  The second image used the bucket (slightly foreground) as the focus point.  The third image used foreground leaves in the bottom left as the focus point.  These three images were then merged in Photoshop to provide a depth off field that covered foreground to about the pig's shoulders (depending on your personal standards for "sharp").  I originally had a fourth shot for background leaves, but found that this made the wall in the back a little too distracting due to surface texture.  I left that out.  Sharpening was applied in Lightroom after the Photoshop focus stacking.

 

The CL image was processed in essentially the same manner.  Manual white balance, ordinary adjustments to tone, etc.  I had to bump up the saturation significantly to get the output from the CL to roughly match that from the X1D.  It's not a perfect match, but it's pretty close.  The X1D definitely produces more vibrant colors than the CL.  Both cameras are quite accurate in terms of hue, but the X1D by default is more saturated.  That could be good or bad, depending on your taste, and it's easy to adjust either to match the other.  So the CL got a saturation bump.  It also needed a bit of a curves adjustment to get the same contrast as the X1D without significantly clipping shadows.  To keep the image from the CL from looking a bit washed out compared to the Hasselblad I had to slightly darken the deepest shadows, so that's one difference between the images.  Finally, the CL is a much lower megapixel camera, so I cropped it to match in terms of aspect ratio and up-sampled it to the same 47 megapixels as the X1D image just so it would be easy to compare the two images.  Not fair to the CL, I know, but it's the only way to do the comparison.  If I had down-sampled the X1D image I would have been throwing out details.

 

I'm providing three images for each... An overall image sized for the web, a crop from the central portion of the image, and a crop from the lower left corner.  I was actually really pleased with how the CL performed.  While I would never use it for this type of photography given the other cameras I own, it's nice to see just what sort of differences one gets as the megapixel count climbs.  The CL did just fine in my view, despite the 23mm not being its strongest lens.

 

Here is the overall shot from the CL...

 

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

 

 

  • Like 1
Link to post
Share on other sites

And here is the overall shot from the X1D...

 

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

  • Like 1
Link to post
Share on other sites

A crop from the CL...

 

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

Link to post
Share on other sites

The same crop from the X1D...

 

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

  • Like 1
Link to post
Share on other sites

A corner crop from the CL...

 

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

  • Like 2
Link to post
Share on other sites

And a corner crop from the X1D...

 

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

  • Like 2
Link to post
Share on other sites

Advertisement (gone after registration)

Here is what I noticed:

 

I was surprised to find I could see some differences in "pop" even with the 1024 x 768 full frame images.  I attribute this to the additional microcontrast in the medium format image.  I'm not certain one couldn't get the same result out of either, but using essentially the same processing on each aside from saturation and format differences, I wasn't expecting to see anything at all at this resolution.  Differences in a 1024 size image are not huge, but they are also not absent.The X1D has an advantage even at this size unless I take some steps in processing to equalize the two.  That was my one real surprise.  I thought I literally wouldn't be able to tell them apart.  And It's not like the gap is huge or that I'm disappointed with the CL.  It's great, and this is a very solid result, in my opinion, but some combination of larger format, higher megapixel count, and lenses that aren't pushed as hard by the magnification resulted in extra pop in the X1D image.  And that's after I made some mild attempt to equalize things by not pulling up shadows as much in the CL image as I did in the X1D.  It was actually even more noticeable beforehand.

 

The differences in fine detail are about what I expected.  24 megapixels vs. 51 megapixels is a sizable gap.  Would it show in a print?  Based on printing samples last night, no unless you print pretty large.  Certainly through 11x14 there is nothing I can see.  Haven't tried larger than that.  Anyway, no surprises in this area.  Plenty of pure resolution in either one for most routine uses.

 

Any differences in dynamic range are quite small in these images and could easily be attributable to processing as much as anything else.  In the deepest shadows, the X1D has a touch more information, but I didn't pull up the shadows as much in the CL images as I did in the X1D, and I expect that's at the root of the minor variation.  Basically, if you are within the dynamic range of either camera you can probably get the same details.

 

Obviously, this didn't test high ISO performance at all.  No comment on that area.

 

For noise performance at base ISO, both images are essentially noise free even in the darkest shadows.  Again, if I were pushing the dynamic range harder there might have been more differences, but this scene was chosen to minimize the challenges thrown to either camera--to make sure they each showed at their best.

 

The last difference I will point out is in microcontrast.  If I look at the transitions from one hue to the next or one tone to the next in areas of fine detail, the X1D simply has more gradations.  It's not resolution per se that I'm referring to.  It's how much color detail, in particular, is maintained down to the contrast limit.  Even if you look at the CL images without the up-sampling and compare them to X1D images that are down-sampled, there is just more variation in hue and tone at the level of the finest details.  This one I was expecting, but it was a bit more obvious than I was expecting.  Again, I can't see it in smaller prints once I put in appropriate levels of output sharpening for each, but the 100% (150% for the CL) crops definitely show it.

 

I'm curious to know whether the rest of you are surprised by any of this.  Do you see the same differences or something else?  Any surprises?

 

Please, don't anyone take this as a criticism of the CL.  It's not intended to be a resolution monster.  I'd take it any day over the X1D for things like:

- Street photography

- Travel photography

- Family pics

- Sports photography (by comparison to the X1D only, I really wouldn't want either for sports)

- Event photography (though, again, it wouldn't be my first choice; would rather have my SL)

- Hiking/Cycling

 

Where I would want the larger format camera would be:

- Any type of studio photography (for more reasons than just resolution)

- Landscape photography

- Environmental portraits

 

Nothing I posted changes the above.  The CL is a great camera for its intended uses, so don't feel like it needs defending as a result of my posting.

  • Like 8
Link to post
Share on other sites

 

Interesting comparison. From a normal viewing distance, i'd say the CL does quite nicely. In fact, I like more contrast of the CL image over the X1D. Perhaps a wee bit of sharpening would get the CL to show better in the corner crop, but otherwise this bodes nicely for the CL. I've recently printed CL images using the 23mm TL at 20 inches on the long side, but will also print a few at 30 inches. The CL images are amazing up to these sizes. So, it's really is about the final output that one should consider a MFD, imo. If you need to print wall murals or billboards, then an advantage would be to the larger sensor, but for the majority of those who don't the CL is an amazing camera! Thanks.

  • Like 1
Link to post
Share on other sites

This might be a big part of the differences that you show:

 

41055336451_a7dbb70235_o.pngScreen Shot 2018-03-27 at 9.32.44 PM (2) by scott kirkpatrick, on Flickr

 

and they are quite pleasantly small.  For the sizes of prints or images I am currently working with I am happy to have more then 8 bits of DR to work with.  The X1D seems to have a win on the pig jowl, but I think they are tied in the shadows.  So it seems there is more highlight reach with the X1D.  Some years back I compared my M9 with a Phase One P45+ 39 MPx back on a Hasselblad V with the Zeiss 120, and saw something similar.  With good eyes you could see some extra life in the big pixel/more pixels image.

 

Further thought -- I think the relevant contrast today is both approaches to an image displayed as best you can on a 4K Retina screen (which I don't have).  Or on a good, calibrated UHD size monitor.  That works out to be what you did when downsampling the larger file, doesn't it?

Edited by scott kirkpatrick
  • Like 1
Link to post
Share on other sites

Thanks for taking the time. Even if I agree it is not a fair comparison I find it still interesting.

What I would like to add that - IMO - there is also a better transition from focused plane to background with the larger sensor.

And I believe large sensor also usually show very good skin tones. 2 factors which can not be seen in your subjects, but one of the reasons I also like to sometimes use the x1d (or S) for family portraits/images of my kids.

  • Like 1
Link to post
Share on other sites

This might be a big part of the differences that you show:

 

41055336451_a7dbb70235_o.pngScreen Shot 2018-03-27 at 9.32.44 PM (2) by scott kirkpatrick, on Flickr

 

and they are quite pleasantly small.  For the sizes of prints or images I am currently working with I am happy to have more then 8 bits of DR to work with.  The X1D seems to have a win on the pig jowl, but I think they are tied in the shadows.  So it seems there is more highlight reach with the X1D.  Some years back I compared my M9 with a Phase One P45+ 39 MPx back on a Hasselblad V with the Zeiss 120, and saw something similar.  With good eyes you could see some extra life in the big pixel/more pixels image.

 

Further thought -- I think the relevant contrast today is both approaches to an image displayed as best you can on a 4K Retina screen (which I don't have).  Or on a good, calibrated UHD size monitor.  That works out to be what you did when downsampling the larger file, doesn't it?

 

 

I did think of the difference in dynamic range, which was actually why I chose the scene I did.  The dynamic range of the scene itself was well within what either camera could capture.  Skies were overcast, it was late evening, and the shots were taken in shade.  Very diffuse light. I don't expect the difference in dynamic range had much to do with the visible differences in the two images.  Neither highlights nor shadows were clipped in either set of raw files.  With a different subject, certainly I would expect to see more dynamic range from the X1D.  I was trying to minimize that effect.  

 

The editing decisions I made on each of the files were with a profiled 27" 5K display.  All my comments are based on viewing the images on the same display.  Obviously, things may look a little different on different monitors.  I'm not sure what that has to do with the down-sampled image?  What were you getting at there?  Maybe you could explain again in different words.

Link to post
Share on other sites

Thanks for taking the time. Even if I agree it is not a fair comparison I find it still interesting.

What I would like to add that - IMO - there is also a better transition from focused plane to background with the larger sensor.

And I believe large sensor also usually show very good skin tones. 2 factors which can not be seen in your subjects, but one of the reasons I also like to sometimes use the x1d (or S) for family portraits/images of my kids.

 

While you may be right about the better transition from focused plane to background, this probably isn't the best sample to use in order to judge that since I used focus stacking on both images.  The only area that was intentionally left out-of-focus was the background wall and the leaves closest to that wall.  Originally, I had intended to incorporate a focus layer for those components as well, but it just made things more distracting and "fake" looking since your brain (or at least my brain) interpreted that sharpness as a lack of depth to the image.  So I left out that fourth layer.  

 

Honestly, while I have hard that the transitions from focused plane to background are better with medium format, it is not something I have yet observed.  Not saying there isn't something to it, but I haven't tested for it and differences in lens design would make it pretty hard to distinguish what's inherent in the format from what is inherent in the particular lens.  For example, the Hasselblad XCD lenses all have diaphragms with eight straight blades which tends to degrade bokeh a bit unless you are shooting wide open.  It's not an absolute, of course, some lenses with straight blades produce very pleasing bokeh.  The XCD lenses were roundly criticized when the camera was first released because even wide open shots were actually slightly stopped down creating hard edged bokeh balls.  Hasselblad has addressed that with a firmware change.

 

You are right that the skin tones out of the X1D are absolutely fantastic, especially if you tone down the saturation just a touch, but I had attributed that to Hasselblad embedding individual sensor profiles into each camera rather than to any inherent advantage to medium format.  I suspect it's more a matter of knowing one's customers than it is a real advantage to medium format.  You'd have a tough time selling any studio camera that couldn't get skin tones correct in a portrait session.  I expect Fujifilm, Leica, Hasselblad, and Phase all work hard to make sure out of camera skin tones are top notch as long as the white balance is correct.

Link to post
Share on other sites

The editing decisions I made on each of the files were with a profiled 27" 5K display.  All my comments are based on viewing the images on the same display.  Obviously, things may look a little different on different monitors.  I'm not sure what that has to do with the down-sampled image?  What were you getting at there?  Maybe you could explain again in different words.

Well, a 4K display has 8 MPx to show you, so editing a 24 MPx image or editing a 48 MPx image, both are downsampled, the X1D by a factor of two more.  In rendering an image, I tend to look at it on my 23" monitor, enlarged to 100% while checking for good focus and highlights in which I may have lost color information, then put it out at 50% linear reduction for saving, sharing, Flickr, etc. If I find the desk space for a 5K system, I might rethink that.  Anyway, my thought is that your monitor sets a common standard for comparison.

  • Like 1
Link to post
Share on other sites

I don't have a CL, but comparing images from the X1D with my M10 show the following in large prints(I just don't trust comparisons on the monitors).

 

X1D files show greatly improved colors and color gradations.

X1D according to my wife produces prints with more three dimensional character (not sure I see this as much)

X1D is somewhat bulkier, especially as compared to the M10

I personally like the OVF of the M10 much better than any LCD. 

  • Like 2
Link to post
Share on other sites

Thanks for all the hard work. Makes me appreciate my CL even more. Also convinced me to order a 23mm TL lens.

Cheers, Dan

 

 

No problem on the "hard work".  Let me avoid my teenage daughter during college acceptance/rejection season.  The 23 is a great lens.  Optically it's not at the level of the 35, but it is small, light, inconspicuous, useful, and produces really satisfying results.  I expect you'll be happy.  And, though I said it's not up to the level of the 35mm, it probably is in the corners (where the 35mm is not quite as "great".)

Link to post
Share on other sites

Well, a 4K display has 8 MPx to show you, so editing a 24 MPx image or editing a 48 MPx image, both are downsampled, the X1D by a factor of two more.  In rendering an image, I tend to look at it on my 23" monitor, enlarged to 100% while checking for good focus and highlights in which I may have lost color information, then put it out at 50% linear reduction for saving, sharing, Flickr, etc. If I find the desk space for a 5K system, I might rethink that.  Anyway, my thought is that your monitor sets a common standard for comparison.

 

Got it.  Yes, it's at least a "common standard" for both images, though there's no way to ensure your monitor will look anything like mine.  That's at least not as big a problem as it used to be since most monitors these days are reasonably close to covering sRGB space if not Adobe 1998, and there is less variation in color response than their used to be.  Contrast levels still vary pretty dramatically, though, as do brightness levels.

Link to post
Share on other sites

I don't have a CL, but comparing images from the X1D with my M10 show the following in large prints(I just don't trust comparisons on the monitors).

 

X1D files show greatly improved colors and color gradations.

X1D according to my wife produces prints with more three dimensional character (not sure I see this as much)

X1D is somewhat bulkier, especially as compared to the M10

I personally like the OVF of the M10 much better than any LCD. 

 

Yes, the improved colors and color gradations would likely be quite obvious in larger prints.  I've never quite understood the three dimensional character bit, but it's clear that there is a bit more "punch" in the X1D images, and I don't just mean color saturation which can be equalized.  I've just never seen it as three dimensionality, though I know a lot of people do.  Bulkier, yes, though the weight difference isn't that big.  And while I am beginning to really like EVF's, the X1D doesn't exactly have a nice one.  I find the CL, by contrast, much more pleasant to use despite its comparatively low magnification.  Plus, it has a live-view histogram that is really, really handy for images like this one.  I really hope Hasselblad adds one with a firmware update.  

 

I found I just wasn't using the M10 any more once I had spent some time with the CL due to the differences in weight, the fairly comparable image quality, and the feature advantages in the CL such as autofocus, availability of zooms when desired, 10fps frame rate (used occasionally), and program mode (also used occasionally).  The M10 just wasn't feeling the love any more.  As a result, I traded it in for the X1D that serves a very different purpose from the CL.  With the M10 and the CL there was just too much overlap in uses, at least for my photography.

  • Like 1
Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...