rscheffler
Offline
• • • • • •
Upload & Sell: Off
|
 p.988 #20 · Leica M/X/T/S/Q/CL/SL Picture Thread | |
Interesting discussion about exposure. I'm kind of old school in this respect and prefer to set everything manually if time and conditions allow. Sometimes I'll do aperture priority, but so far haven't ventured into auto ISO. This is for both my DSLR and Leica systems.
Touching on what Richard mentioned, if it's a work situation, then I'm thinking 'production efficiencies' and want to find the correct manual settings as quickly as possible and keep them locked in for faster processing later. For example, when I shoot football, like the game yesterday, where it was full sun most of the time, I have a standardized exposure for the sunny side and one for the backlit side. I try to avoid shooting in cross light if possible. When it's sunny, or other situations with consistent light, my opinion is there is zero need for auto exposure modes because the light does not change, therefore why should the exposure? But of course, if moving from sunny side to backlit side, of course it changes. On a walking tour such as what Charlie has been posting recently, I'd leave the camera at ISO 200 and any combination of 1/500 f/8 or f/9.5 (I find I have to give the 28 Cron about a half stop more exposure than the 21 SEM or 50 Lux ASPH, or pretty much any other lens).
With the M9 I kept ISO at 160 or 640 and pushed in post as needed. With the M240 I try to keep it at 200 whenever possible and at 1000 for low light situations. I'll go higher than 1000 as the situation warrants and if I feel I can live with the slower fps rate and buffer depth. But usually I opt for 1000 for the faster shooting ability.
Richard, the M240 tolerates over exposure better than the M9 did and underexposure somewhat less gracefully. Shadows of ISO 200 files pushed 3+ stops will look pretty green, which is what that Lightroom plug-in fixes. As the ISO increases, it takes less push in post to reveal the green shift, though shooting at ISO 1250 and higher eliminates it too.
28 Cron Test
So with the most absolutely perfect summer day of the summer thus far, IMO (my sympathy to the folks out in snowy Alberta), I put off many client obligations and went out to shoot a 'quick' 28 Cron on M9 vs. M240 test that seemed to be a topic of interest in the "M240 purchase" thread.
When I get the time, I'll organize the images, but here's what I noticed so far:
- My $5 vented off-brand hood probably causes some vignetting and probably screwed up this aspect of the results, but it's on all of the test images, therefore at least it's consistent.
- Cameras were set to DNG + Jpeg Fine, all manual settings and base ISO for each and manual WB (5200K). I shot the Jpegs in addition to the DNGs to see what Leica's in-camera profile is, rather than seeing Adobe's interpretation of Leica's profile in Lightroom.
- At the identical WB setting, the M9 is a lot cooler, or one could say the M240 is a lot warmer.
- All image sets were shot coded and uncoded on both cameras. The M240 shows less difference between coded and uncoded settings. There is much less color shift and color quality is nearly identical. Vignetting is a bit stronger if uncoded. On the M9, there's a bigger difference in all these qualities. There's greater color shift uncoded and WB/color quality also changes more if uncoded, to a slight magenta tinge. Vignetting also increases.
- The M240 images seem to be every so slightly wider in view, showing slightly more content, hinting at a slightly larger sensor size.
- Looking at the Jpegs, ignoring color/WB differences, overall tonality isn't all that different. Shadow values are quite similar, mid tones, highlight.... at least when shot at f/4 and more stopped down. At f/2 the differences between the M9 and M240 images are significant. Again, this is based on the OOC Jpegs. I haven't yet looked closely at the DNGs. The f/2 M9 images have great contrast and 'sparkle' in the mid tones. The M240 looks quite drab and dull, especially in mid tone and highlight values, as if they were pulled back with excessive highlight recovery. This can be clearly seen in a 'white wall' test where the M240's f/2 and f/2.8 images, to a lesser degree, show the highlight values pulled farther to the left in the histogram compared to the M9. Stop down past f/4 and the histograms from each camera look very similar, as do the images.
- The f/2 white wall test shows an interesting vignetting 'reversal' between the M9 and M240. In the M9's f/2 images, vignetting seems pretty uniform in all directions, from side to side, top to bottom, though the upper part of the image (shot horizontally) seems to have slightly stronger vignetting than the bottom. Normally this would be where the sky is in a landscape... With the M240, the bottom of the image is noticeably darker and the top lighter than the M9. As the lens is stopped down, this difference between the M9 and M240 decreases pretty quickly and is pretty difficult to see after f/4. At f/11 the M9 seems to have a touch stronger vignetting all around. Note that the difference in odd M240 vignetting quality is consistent whether coding is on or off.
- M240 and 28 Cron shot without coding doesn't seem to suffer much from noticeable color shift in most situations and it could be an option if desiring slightly stronger vignetting character. I wouldn't advise it with the M9 due to stronger color shift. That said I was never really crazy about the 28 Cron's coding on the M9, which for me often resulted in green/cyan shift in the lower left and lower right corners.
I might reshoot the test again without a hood, just to eliminate it as a variable for the wider open scenes, though I don't think it will make a significant difference. And I still have to look at the DNGs.
But thus far, if shooting the lens wide open for 'character' images, the M9 will give the aesthetically more pleasing result SOOC. The M240's Jpegs at f/2 look dull in the mid tones and highlights. I put a couple of the DNGs in Lightroom and can tweak the M240 f/2 files to result in similar mid tones and highlight pop, but the odd vignetting 'reversal' remains. I also think color/luminosity mapping is different between the cameras. Blue sky is deeper in M9 images vs. M240. I have to pull back blue luminance in LR ~20 points and add a vignette to the top of the image to get a similar feel.
Bottom line, from initial impressions, it seems the M240's 28 Cron profile is wonky for f/2-2.8. Beyond f/4 the Cron's look on both cameras evens up very closely beyond basic WB differences.
|
. I'll see though, coding might not be essential...
