hyalinejim

Colour is a tricky one to talk about for sure. Nevertheless, there's been some interesting discussion in this thread.

Thickness of an image is a term that originated in reference to the density of a film negative. It's still used by some people when talking about digital images, where its meaning is unclear (qualitative rather than quantitative).

Perhaps a more useful avenue is to ask about characteristics of digital colour that are desirable. These could be

- Subjectively pleasing colour

- Objectively accurate to reality colour

- Objectively accurate to a film stock colour 

- The ability to push colour in a grade

On the topic of the first three above, it's interesting to note that Kodak designed their colour film stocks on the basis of customer feedback. They walked around shopping malls in the US and showed people photos with different colour renderings and asked them which they prefer.  More info here:

https://www.canopycanopycanopy.com/contents/color-goes-electric/?fbclid=IwAR0mWacShDadxrKexLSPLJ5tatHtZm4t7yBSUMEGQArQ4eSwSJCU8icX3WQ#six-one

Actually from reading that link it turns out that, for good or ill, DXO doesn't take colour accuracy into account in their "portrait" rating.

They say all DSLRs score similarly for colour accuracy.

There's a lot on that page I don't understand. But I have the sense that their portrait rating has a lot more to do with quantitative than qualitative issues.

In short, numbers are compelling but in reality I'd be hesitant to use their rating in deciding which camera to buy to take photos of people in a studio.

Wow, no love for IBIS today?

I couldn't go back to a non-IBIS camera. I shoot lots of B roll of people who've never been on camera before and will be gone away in a minute, doing things. IBIS means I can shoot handheld and get three different steady shots from different angles while directing the person(s) in the same amount of time it would take me to get one shot on a tripod.

I don't use it as a substitute for a dolly, slider or gimbal. I still remember my horror at the shaky jello of my first handheld shots on the 5D Mk 2!

1 hour ago, noone said:

Maximum color sensitivity reports in bits the number of colors that the sensor is able to distinguish.

I've also checked out DXO's portrait colour rating. The last thing you'd want for a nice portrait is accurate colours. Perhaps it's better to call it a colour accuracy rating then (and who knows whether their methodology is sound or not). But accurate colour is not nice colour.

Reality looks kind of shitty, in terms of colour, compared to how its colours can be represented in photographic images.

Colour accuracy should not be the goal, IMO, unless you're reproducing artwork or products. I would like my images to look better than reality, when it comes to colour. 

So I would absolutely not expect the top rated DXO cameras for portraits to make nicer portraits SOOC than those lower on the list. 

Yes, density of a negative increases with exposure. It's measured with a densitometer. 

That's where the dense/thick terminology comes from, where it is an observable quantity. 

For digital, it's a qualitative assessment and its meaning might vary from one person to the next.

3 hours ago, herein2020 said:

R6's video quality blows the GH5 out of the water....

In what way is it superior?

The classic illustration of that is an explosion of confetti: IPB falls apart. 

Punch in while recording added for 700d

https://www.magiclantern.fm/forum/index.php?topic=5951.msg231434;topicseen#msg231434

So I'd expect 2 stops less highlight headroom at 160 versus 640.

14 minutes ago, MrSMW said:

It looks good but it’s a real struggle outdoors in even moderately bright light.

Any option to set light or dark, users preference, like *cough* DPreview?

At bottom of page click Theme and you can select light mode.

1 hour ago, Andrew Reid said:

Ah when phone screen brightness is on medium, it does seem to look too dark. Have tweaked the text so it's brighter. Cheers for the feedback, keep it coming

I think for the forum that the text in each post should be the brightest element as it's the most important.

I think dark mode needs more contrast, at least for mobile devices, as it's v difficult to make things out unless you're in a dark environment.

This is what a Reddit app looks like in dark mode. Everything is legible

17 hours ago, IronFilm said:

"Tricky" cameras also get more discussion.

Hence the presence of the XC10. Love/hate is not the right term. Joy/disappointment is closer.

I was actually thinking about it last night when browsing C70 threads and I remembered its ingenious twisty grip. But the temporal noise reduction was pants 😭

The pristine 1080p is one of the things I love about the GH5. I shoot a lot of 1080 as it keeps file sizes down and is easier to edit. 

So crappy 1080 would be a deal breaker, as they say, for me.

32 minutes ago, Oliver Daniel said:

My camera lets me do it. I am baffled. 

I'd guess that if you check the highlight clipping point at 100 in SLog3 there'll be fewer stops above middle grey.

I sometimes use GH5 ETC and definitely notice a drop in image quality. I did not know there was a 2x digital zoom so thank you for that!

2 hours ago, maxmizer said:

how much time wasted in useless chatter...

On the contrary, the last few pages of discussion gets right to the heart of the question raised. And I think many people would agree that the answer is:

• The necessity or desirability of a format size depend on whether the lenses available for that format will give you the visual qualities that you value.

Keep us posted! ML needs interpreters for us plebs as it's a bit esoteric 😂

30 minutes ago, tupp said:

I don't have any conclusions in regards to larger formats vs. smaller formats other than the ones I have mentioned in this thread and in other threads.

Can you link to those posts?

I'm still no clearer on what the differences in DOF rendering due to large/small format lenses actually looks like in an image.

Actually, that's not quite the point I was making. I don't know much about large format lenses.

My point was that any observed differences might be due to the difference in aperture required to maintain equivalence between formats. But I'm certainly open to the idea that there are other factors that also contribute. 

Are there any conclusions you would draw on the pros and cons  of large formats versus small in terms of the qualities of images afforded by the glass associated with each? For example, are large format lenses well suited to narrow DOF pics that maintain sharpness and small format lenses well suited to deep DOF without suffering as much from diffraction?

Well, I think that DOF as defined by circle of confusion etc can be matched because equivalence theory states that you can, if the lens for the smaller format is bright enough. However, I would expect to see a considerably softer image with lots of vignetting as you'd need a very fast lens to replicate the narrow DOF of this shot, and that's how lenses behave wide open. So although the DOF might be technically the same, the images will look different. But this is caused by the glass, not sensor size.

And yes, the selection of lenses available for different formats is different. So your choice of format will have an impact on the look of the image. But I think a lot of people are making the point that those differences are derived from the glass and are not inherent to the sensor size.

So theoretically, sensor size makes no difference to DOF. But in practice DOF is rendered qualitatively differently because the lenses are different / behave differently / must be set differently for different formats.

If true, it's an interesting dichotomy. But it does suggest that any equivalence test is really just a comparison of two different lenses. In the same way that one of my 50mm lenses looks different from the other 50mm lenses I have for the same format. So would you agree that once you match focal length and aperture for the same shot on different formats, you're comparing lenses?

2 hours ago, seanzzxx said:

any perceived differences are more likely to be due to individual lens characteristiscs or other uncontrolled variables which are not related to the film back size.

I would say that this is very likely.

2 hours ago, tupp said:

He does not show how the limits of DOF are delineated. 

It still sounds intriguing though 🙂 You misunderstood my proposed comparison though. The simulated crop on the 50 would simulate a notional sensor 1/4 the size of full frame. But I suspect that any differences observed would have more to do with the glass involved (and the necessary apertures) than the sensor size. Perhaps this is what accounts for your observation that the differences in rendering of DOF are greater when the disparity of sensor size is increased: to maintain equivalence, one lens is quite wide open and/or the other is quite stopped down.

Well, you've convinced me that there's something worth considering there for sure. I had always conceived of the equivalence debates as being along the lines of flat-earthers etc 🤣

But now I see that for some it's just about a level of complexity and possibly even bokeh connosseurship.

For my purposes, these differences are so rarefied as to be irrelevant. However, the idea of testing for them is interesting. How about a test where you simulate a small sensor by using a center crop from a full frame stills camera? Like this:

Full frame
200mm, f7.1

Simulated 4x crop sensor
50mm, f1.8

These would be different lenses. I could do this (when I have time) with a Canon 50 1.8 and Sigma 70-200 f2.8 and use ACR to correct for lens aberrations. Or instead with an OM Zuiko 50 1.8 and OM Zuiko 200mm f4 (possibly similar primes? But can't correct for aberrations)

Would 4x be enough to show a difference?

I realise you're dealing with a lower megapixel image from the central portion of a wide open lens for the simulated 4x so the image will be softer.  Would these variables make the test invalid?

23 minutes ago, tupp said:

I wouldn't say that the rolloff is "shorter" or more rapid on larger formats nor "longer" nor slower on smaller formats.  Nor would I say that the rate of the falloff is smooth/constant in larger or smaller formats.

 

26 minutes ago, tupp said:

However, there is more involved in "format specific looks"

This implies that you maintain that there is a look inherent to a format, independent of variations between lenses. If so, it should be consistent as format size changes and it should be describable. How does the look of small format compare to the look of a larger format, at equivalent focal lengths and apertures?

I'm just interested here, as I use DOF calculators to help my understanding when moving between FF, micro 4/3 and speedboosted micro 4/3. But I also see a huge difference in the images posted, which I would not have expected.

On 9/15/2020 at 10:28 AM, tupp said:

Do you not see in the area outlined in red how the distant white building exhibits sharp edges in the 1-inch image, while it is much softer in the full frame image?

I certainly can see what you're talking about in the areas you've highlighted. It's very clear.

But if you plug equivalent settings into a DOF calculator you get the same amount of DOF for both, according to the calculator. This is the theory, and it depends on certain assumptions regarding circle of confusion that I won't pretend I understand in depth.

So is it fair to say that your position is something like this:

"Even though DOF calculators show that the theoretical DOF of equivalent shots is the same, in practice there is an observable difference in how DOF is rendered between equivalent shots"?

If so, and if I understand you correctly, that would mean that a DOF calculator is showing us 2 points on the DOF continuum, the point of near focus and the point of far focus, and these are the same for both formats if the focal length and aperture is equivalent. But it's not telling us anything about the DOF characteristics elsewhere on the continuum, which is noticeably different.

That's certainly very interesting and worth investigating. I think for most people it's enough to know that they can match the field of view precisely using equivalence theory, and that the DOF is "the same" according to its assumptions. But others may notice and be very interested in differences in DOF behaviour that's not described by equivalence theory... if what you say is true! So we would be talking about a kind of DOF rolloff, which is shorter on larger formats and longer on small formats, according to equivalence sceptics.

Is this understanding correct @tupp?