George Paulides
Senior Member
So was there a link shared to this landscaper photographer?
--
http://www.flickr.com/photos/21702997@N03/
--
http://www.flickr.com/photos/21702997@N03/
Last edited:
-
Welcome to the new forums! Please read this first. For known issues we are working to resolve, click here.
Ran into a Great Landscape Pro with Sony a7RIII and 24-105 F4
- Thread starter Greg7579
- Start date
Sal Baker
Forum Pro
5
57even
Guest
Because at base ISO, the output medium has far fewer distinct grey and colour levels than the camera. The camera is not the limiting factor. That's why it's so easy to get posterisation on RGB displays. To avoid it, you need a 10-bit display which few people own.
This isn't true for small sensors because of noise dithering. Ironically that actually reduces posterisation.
Agreed, as I said in my other post.
Then I guess it was a flawed test, given that images on most displays are very low resolution compared to camera MTF.
The downsizing method and sharpening have a lot to do it. For instance, if you zoom by 50% or 25% on a display, the image looks much sharper than when you zoom by 39% or 53%. Similarly sharpening should be done after resizing and optimised for viewing size.
Standard resizing methods are not equally good, even the ones in Photoshop.
I saw a test posted on DPR a few years back comparing a 3MP Sony digicam with a 6MP APSC DSLR (D100? I can't remember). A significant number of people got the answer wrong.
So there are many "tests" that prove the exact opposite. So much for the internet.
I made my own comparisons with an Xpro2 before selling my D800. It wasn't just me who couldn't tell the difference. Why would I sell the D800 otherwise?
Unfortunately, I can't 'prove' this to anyone without showing them all the same prints, but it is consistent with measurements of human eyesight (20-20 is about 35 cycles/degree). It is also consistent with SQF measurements (something not often seen these days) which are based on contrast sensitivity.
So I am fairly certain my tests of the Xpro2 and D800 were not flawed in some way. The relative equality is easily explained by the AA filter on D800. The D850 is noticeably sharper - but with much more moire.
However, for my normal print size, 24MP without an AA filter is more than adequate, and close enough to 36MP with an AA filter for the difference to be irrelevant.
And my eyesight is better than 20-20.
--
Reporter: "Mr Gandhi, what do you think of Western Civilisation?"
Mahatma Gandhi: "I think it would be a very good idea!"
Last edited:
5
57even
Guest
Actually, noise limits that to fewer shades that we can actually distinguish. Check the tonal range numbers available from DXO.
Obviously, this depends on how much the image is downsized, but it's still more than 256 at low ISO on any camera with a half decent sensor size. Around 9-10 bits (500-1000 shades).
So you are certainly correct in principle.
Without noise, the human eye can distinguish about 800-1000 grey shades on a typical computer display (without adaptation). But that requires a 10-bit display. An 8-bit display is no better than an 8-bit JPEG. 256 levels.
Prints have far less contrast, so the number is even lower, but still more than 256.
There is a good reason why HDTV standards are now 10-bit, not 8-bit. On a 10-bit display you are more likely to see the difference between FF and APSC - but less so in print.
Of course, FF does allow you to increase ISO by one more stop before you notice anything.
5
57even
Guest
Nothing to do with bit-depth, more to do with shot noise. Bigger sensor, less noise, better depth. But whether this is visible in prints is debatable unless you also use a 16-bit printer.
If I was able to achieve a very high hit rate, something is going on that we're not capturing.
Good question.
Ah - it's my eyesight. That's the difference - I'm not even 20-20. Having flawed eyesight apparently makes me able to see the difference.
Hey Greg, any chance you got to ask the guy if we could check out his work?
I had a travel experience yesterday that had an impact on my thinking as a Fuji travel-shooter. Teresa and I were sitting in a farmhouse breakfast room at 7 in the morning in the pouring rain yesterday deep in the Lucanian Dolomites down in the boot instep of Southern Italy (Basilicata) just outside an amazing mountaintop town called Castelmezzano. We had stopped for the evening in one of those agriculture - farmhouse places with rooms and farm-to-table meals that are so popular in Italy now. We were seeking shelter and a place to hole-up for the day because it was storming badly. The last week of our 8-week trip here is turning very bad on the weather.
Anyway, we were sitting drinking coffee and in comes this Austrian couple our age speaking German and they saw us sitting there all quiet and they switched to fluent English. I of course immediately spotted what was in his hand. It was a landscape-shooter's 42.4 MP Sony a7RIII -- not the a7III. On it was the Sony 24-70 F2.8. I was sitting there fiddling with my XH-1 with Brick attached, deleting bad images from the crappy weather the day before.
I had my Dell XPS laptop on the table. He pulled out his Mac. We of course started talking because the two couples were alone in the kitchen area there at the farm house.
We talked for two hours and played with each other's cameras. The amazing thing to me was the size-weight. The two rigs were about the same, but the Sony seemed even a bit smaller. I have the RRS L-Bracket mounted, which makes it bigger, but I have to admit I'm not getting the much-touted size advantage with my Fuji rig vs the Sony - at least not from what I could tell. I played with that Sony for an hour and walked around and shot it. I didn't like the feel or ergo of it near as much as the Fuji, but I am of course biased.
I must say too that Helmut was impressed with the Fuji dials and aperture ring. He is an old-school pro and like me, shot film for decades. He knew about Fuji but had never really played with one as he did mine. He loved the ergo, but was extremely dismissive of APS-C, like so many FF guys are. We talked about it for an hour and he was a hard-core FF believer and threw out every anti-APSC dinger I have ever heard. He even said two stops of difference on DR and other hardball dingers about resolution and IQ. I didn't argue with him because his work was mind-blowing good and he was on a roll. He was even making fun of the new a7III because it had such "low" resolution compared to the a7Riii. He was a big believer in landscape resolution. Yes, I did opine that he should go Fuji MF!He said no way, that his a7riii had as good of IQ as the Fuji MF. I let it go....
Anyway, he fired up the Mac and showed me his website and blog. His landscape images were worthy of being among the greats. He says he studies and copies all the greats. The images were just amazing. Of course, I was probably looking at his best work. But I think I can say that they were the best landscapes I have ever seen. I was stunned actually. He looked at my Flickr at the bazillion images from this trip because they were down in Southern Italy doing the same basic trip for a few weeks, and he was extremely complimentary of my images, so I was much easier on him than I am the Sony guys and drive-by shooters on this Board. I did not resort to using my vast arsenal of FF vs Fuji information and counter-attack zingers. I just let it go.
But really -- his fantastic images had nothing to do with the camera if you ask me. This guy was an artist. I am going to contact him today and ask if I can attach his link and info here. If he says OK, I will. I need to check him out, because I think I am going to find out this guy is a well-known landscape pro in Europe. He shoots only landscape - nothing else, and all hand-held.
I must admit, as always with Sony I am tempted … really tempted. But I love the XH-1 and the Fuji ergo. I love the Fuji glass. If I got that 42 MP FF Sony, it would not improve my landscapes.
You guys are lucky! I just got good WiFi and posted 100 more images from this amazing Southern Italy trip. We just made a run back to Naples and I was almost on the street because every room in and near Naples is booked because it is one of their biggest Holidays, and I was totally unaware of that bit of key information. It s pouring rain and this is going to be another long day of no shooting. Fly out of Rome Wednesday. Been gone two months.
Greg Johnson, San Antonio, Texas
https://www.flickr.com/photos/139148982@N02/albums
5
57even
Guest
Entirely possible, but if so it was an error in the way the comparison was done. Different resolutions downsized by different factors using non-optimised algorithms, and not resharpened after the event. Would easily account for it.
Hmm, interesting. Logically, one would assume that it would have the opposite effect.
However, human acuity is not simple. For instance we are much more acutely aware of edges and lines than random detail. False detail (above Nyquist) can mess up our perception. Have to ponder that one.
Ha ha. You know just kidding.
But what was it about those images that made them easy to pick out? Was it the style maybe? I have no idea.
Maybe he ran into him with his car. That's why he's not saying.
5
57even
Guest
Just to be clear, were they identical images of the same subject reproduced at the same resolution? I ask because...
- Different subjects have very different detail construction. Distant trees vs brick walls for instance.
- We are much more sensitive to luminance contrast (detail) in certain colours which have a big variance in the green channel. Moreover, we have more green photosites.
- If they were not resized physically and displayed at their actual size, then they are resized by the display driver to fit the page. This uses very crude interpolation (which is why images in DPR galleries always look really blurred).
Otherwise, it's meaningless.
It was a bunch of random images of all physical dimensions. It didn't seem to matter the size or subject.
What made it not meaningless was the number of images. All of the things you would want in a controlled test were basically randomized by the quantity of images.
Similar threads
Keyboard shortcuts
- f
- Forum