Jack Flesher wrote:
I think a more significant reality is the changing world...
Digital is growing, with high-def displays replacing paper and ink for a lot of what people view. For the most part, video is growing more popular daily while still imagery wanes, mainly due to ease and convenience of creation and editing digitally. And current digital video can only be seen on a digital display of some form -- but then so can stills, and I suspect this is where the digital art world's "viewing medium" is headed...
Will paper and ink printing survive? Sure, but more for cards and wrapping paper and as a niche artistic medium much like silver, wet-plate or platinum photography is now; and any actual art paper will probably be used more for paint or pen than digitally printed artwork.
20 years ago after getting my first digital, 4mp DSLR, I predicted a similar death for silver film, stating Kodak would finally stop producing it's last remaining film, Tri-X, in 2022. Clearly I was wrong, but not by much -- you can still get Tri-X ...Show more →
This remind me of this article I wrote for LL in 2008... that was mistakenly attributed to Michael during a site migration... still need to have that corrected.
On the topic of DoF with APS-C offering more... yes I was having this argument in 2005 when I shot with my D2x vs 1Ds users... In the meantime we have Helicon focus and cameras do a fine job at automating the capture process, but a manual 5 shot DoF often works well enough for those images that benefit from a lot of DoF.
bernardl wrote:
This remind me of this article I wrote for LL in 2008... that was mistakenly attributed to Michael during a site migration... still need to have that corrected.
On the topic of DoF with APS-C offering more... yes I was having this argument in 2005 when I shot with my D2x vs 1Ds users... In the meantime we have Helicon focus and cameras do a fine job at automating the capture process, but a manual 5 shot DoF often works well enough for those images that benefit from a lot of DoF.
Cheers,
Bernard
Focus stacking is almost a guaranteed necessity for most 44x33 shots if one is after perfect foreground to background sharpness. That means IBIS can't save us from needing a tripod. That also means compositional limitations since the focus stacking software can't always merge complex foreground objects with a background at or near infinity. For example, try shooting a scene with a cactus or tree in the foreground that overlaps the mountains and sky in the background – it's just too complex and the end result is messy like early iPhone bokeh.
highdesertmesa wrote:
Focus stacking is almost a guaranteed necessity for most 44x33 shots if one is after perfect foreground to background sharpness. That means IBIS can't save us from needing a tripod. That also means compositional limitations since the focus stacking software can't always merge complex foreground objects with a background at or near infinity. For example, try shooting a scene with a cactus or tree in the foreground that overlaps the mountains and sky in the background – it's just too complex and the end result is messy like early iPhone bokeh.
Yes... that can happen indeed. Now, not all landscape images benefit from a lot of DoF. This recent example I find works better with limited one.
highdesertmesa wrote:
Focus stacking is almost a guaranteed necessity for most 44x33 shots if one is after perfect foreground to background sharpness. That means IBIS can't save us from needing a tripod. That also means compositional limitations since the focus stacking software can't always merge complex foreground objects with a background at or near infinity. For example, try shooting a scene with a cactus or tree in the foreground that overlaps the mountains and sky in the background – it's just too complex and the end result is messy like early iPhone bokeh.
And more or less precludes anything moving in thr subject?
Although an early experiment I did with stitching 3 16mpx shots of a panorama including sheep did include the same sheep twice in the stitch. 😀
highdesertmesa wrote:
Focus stacking is almost a guaranteed necessity for most 44x33 shots if one is after perfect foreground to background sharpness. That means IBIS can't save us from needing a tripod. That also means compositional limitations since the focus stacking software can't always merge complex foreground objects with a background at or near infinity. For example, try shooting a scene with a cactus or tree in the foreground that overlaps the mountains and sky in the background – it's just too complex and the end result is messy like early iPhone bokeh.
gyoung143 wrote:
And more or less precludes anything moving in thr subject?
Although an early experiment I did with stitching 3 16mpx shots of a panorama including sheep did include the same sheep twice in the stitch. 😀
Gerry
I had the same issue once, what was the sheep’s name?
bernardl wrote:
I had the same issue once, what was the sheep’s name?
Cheers,
Bernard
Tilt will certainly NOT cope with a cactus in front of a distant view if the cactus is verical and fills anything like a good proportion of the frame height.
But otherwise you can try PTgui’s layered output and modify the layer masks in Photoshop. If you have a reasonably recent GPU (I use a 2019 Mac Pro whose GPU are probably 5 times slower than the latest), it’s 10 times faster than pS stitching also btw. It computes and outputs to a Thunderbolt 3 Raid6 DAS unit a 300 megapixels .psb file in less than a minute.
It’s typically faster than the time it takes for PS to just save a file of such a size. 😁
bernardl wrote:
T/S can only deal with a single plane of sharpness, typically the ground. DoF stacking is a more generic solution.
Cheers,
Bernard
I still occasionally use a TS lens. (A Pentax MF zoom on the Mirex TS adapter on a 5DsR.)
As you point out, while it can solve (or at least help) with some situations, it has some serious limits. Consider a photograph of the ground leading off into the distance, say a flat desert landscape. Now imagine a cactus a few feet into the frame that extends from the bottom of the frame to the top. :-)
bernardl wrote:
understand Jack, I stitch less than I used to back in the days.
Two things that we used to have to do all the time that I virtually never have to do any more: stitching and exposure blending.
When I worked at a School of Art & Design we had an Artist in Residence for a short while who produced a whole exhibition of photos taken in a dense wood of young silver birch trees, using Tilt on a Sinar and largish aperture to put a 'slice' of sharp plane of focus through the trees at various angles. 20 prints 20x16 all similar.
He thought it a marvelous idea, gave me a headache!
gyoung143 wrote:
When I worked at a School of Art & Design we had an Artist in Residence for a short while who produced a whole exhibition of photos taken in a dense wood of young silver birch trees, using Tilt on a Sinar and largish aperture to put a 'slice' of sharp plane of focus through the trees at various angles. 20 prints 20x16 all similar.
He thought it a marvelous idea, gave me a headache!
Gerry
That is a fascinating idea on several levels, including its reference to a classic photography technique, but in an obvious and atypical way.
gdanmitchell wrote:
I still occasionally use a TS lens. (A Pentax MF zoom on the Mirex TS adapter on a 5DsR.)
As you point out, while it can solve (or at least help) with some situations, it has some serious limits. Consider a photograph of the ground leading off into the distance, say a flat desert landscape. Now imagine a cactus a few feet into the frame that extends from the bottom of the frame to the top. :-)
Two things that we used to have to do all the time that I virtually never have to do any more: stitching and exposure blending.
Do you mean that you stopped stitching when you started to use your 5DSr? What caused you to stop doing exposure blending?
I personally stitch less for the following reasons:
- I do less landscape than I used to
- I find 100mp to get close to the level needed for A0 prints that are not too compromised, it’s not quite there but getting close
- I am not as fit as I used to be, or I should say was because I am back to a reasonable level of fitness leaving aside my current broken foot, and therefore pack my pano head less often
- the GFX system lacks an outstanding standard prime akin to the Otus 55mm or 50mm f1.4 ML. I could use a DSLR Otus 55mm but it doesn’t cover the frame perfectly and I sold mine years ago anyways. The 55mm f1.7 is a bit too wide, the 80mm f1.7 too long and heavy. The 63mm is ok but not great. None of them match a Voigtlander 50mm f2.0 APO or Zeiss Otus 50mm f1.4 ML
- the 3:2 aspect ratio of the 35mm cameras makes a bit more sense when stitching than the 4:3 one of the GFX
- I had a phase during which I got a bit lazzy
But stitching is easier now than it ever was and a 80mp Z7III would be the ideal stitching and therefore landscape camera IMHO.
bernardl wrote:
Do you mean that you stopped stitching when you started to use your 5DSr? What caused you to stop doing exposure blending?
I personally stitch less for the following reasons:
- I do less landscape than I used to
- I find 100mp to get close to the level needed for A0 prints that are not too compromised, it’s not quite there but getting close
- I am not as fit as I used to be, or I should say was because I am back to a reasonable level of fitness leaving aside my current broken foot, and therefore pack my pano head less often
- the GFX system lacks an outstanding standard prime akin to the Otus 55mm or 50mm f1.4 ML. I could use a DSLR Otus 55mm but it doesn’t cover the frame perfectly and I sold mine years ago anyways. The 55mm f1.7 is a bit too wide, the 80mm f1.7 too long and heavy. The 63mm is ok but not great. None of them match a Voigtlander 50mm f2.0 APO or Zeiss Otus 50mm f1.4 ML
- the 3:2 aspect ratio of the 35mm cameras makes a bit more sense when stitching than the 4:3 one of the GFX
- I had a phase during which I got a bit lazzy
But stitching is easier now than it ever was and a 80mp Z7III would be the ideal stitching and therefore landscape camera IMHO.
I can produce quite large prints of excellent quality from the 5DsR files on my Epson P9000. Software improvements have made this even more viable — for example I can recover ever more shadow detail when necessary now that the Adobe AI Denoise feature does such an excellent job. (I actually have old photographs made in poor light — for example pre-dawn bird-in-flight photographs — that were formerly unusable, but which can now be used after applying this tool.)
As to stitching, unless I am doing something extraordinarily large — let's say some sort of really big pano shot — it simply isn;t necessary. (I have licensed some stitched images at huge sizes, like between 18' and 30' wide, for commercial use.)
Back when I used the plain old 5D and to some extend with the old 5DII, I did do exposure blending, but it is extremely rare that I would need to do that now.
gdanmitchell wrote:
I’m pretty clear about what gear I do have do not hae experience with and about the nature and depth of that experience, as I was in this post.
My point wasn’t to assert that I have experience proving that the 70-300 is great or not but to assert that a presumption that Fujifilm’s less expensive lenses in general are not good performers is open to question — based on my experience with several of them, two of which I mentioned in the post.
I think I was clear about that.
My post wasn't intended to be directed at you, though I see how it might be seen as such. It was more a sarcastic take on the reliability of internet advice.
I'm confused about some of the discussion about DoF and diffraction above. I thought the equivalence argument implied there was no free lunch one way or another between different formats and the tradeoff between DoF and diffraction. For example, M4/3 has more DoF but also becomes diffraction limited at about f5.6, meaning if you need more DoF you start suffering from diffraction softening at f8. Meanwhile APSc has about one stop shallower DoF but diffraction softening doesn't start until around f11.
Now I know the above doesn't quite hold if viewing at 100 per cent, because then pixel level diffraction matters, but surely for landscape work actual viewing size rather than viewing at 100 per cent should be the benchmark?
In the meantime we have Helicon focus and cameras do a fine job at automating the capture process, but a manual 5 shot DoF often works well enough for those images that benefit from a lot of DoF.