JensR

Lives in United Kingdom Bath, United Kingdom
Works as a Mechanical Engineer
Joined on Nov 23, 2003
About me:

Hi,

thanks for stopping by!
If you want to see what I'm up to, send me a message :)

My 'plan':
Talk Pentax into a digital Electro-Spotmatic! (This needs some work...)

-------------

Older Signatures:

'LBA knows no bounds, and seeks no justification...' (Jim King, 2005)
http://www.jr-worldwi.de/photo/index.html - Photography, Tech and Geek stuff :}

'Why is everyone answering rhetorical questions?' (Me, 2005)

'Well, 'Zooming with your feet' is usually a stupid thing as zoom rings are designed for hands.' (Me, 2006)

'I only trust those photos I have faked myself.' (Me, 2007)
http://www.jensroesner.de/

--=! Condemning proprietary batteries since 1976 !=--

'I don't want them to believe me, I just want them to think.'
Marshall McLuhan

Comments

Total: 486, showing: 21 – 40
« First‹ Previous12345Next ›Last »
In reply to:

rbach44: Correct me if I’m wrong, but wouldn’t a RAW histogram be impossible? Wouldn’t the tonal values have to go through some kind of profile in order to be mapped like that? And if it is going to be mapped to a profile, even a sort of “neutral” one, couldn’t that potentially differ from how one would open in their choice of raw converter?

I guess my point is is that this search for a “true” histogram may not be as useful as it sounds, there are plenty of variables. In which case, couldn’t we simply use the JPEG histogram we already have and be familiar with our gear enough that we can predict if we have enough information to feed our editing process? Thats how I do and I often nail exposure to what I want at any time. And with how good modern sensors are it seems like overkill to try and squeeze every last drop of exposure out of your sensor to increase STN ratio…

A RAW histogram is entirely possible, but for humans to quickly use it, we need a logarithmic y-axis when displaying it. That is not the same as data processing or a profile, and it has been done before:
https://www.rawdigger.com/howtouse/rawdigger-histograms-what-is-the-raw-histogram
DPreview did a great job explaining why this would be a good idea, so I am not going to go over it again. Basically, JPEG histogram is "good enough" for many, but it is seriously flawed.

> And with how good modern sensors are it seems like overkill to try and squeeze every last drop of exposure out of your sensor to increase STN ratio…

Humanity does not have a history of being satisfied. Getting the best out of something and improving it seems to me to be human nature.

Link | Posted on Aug 7, 2018 at 12:16 UTC
In reply to:

Baenwort: I'm interested in seeing this on a m4/3 body as the combo of extra DOF and great IBIS sounds good to me and it seems like enough will get made I might find one for a reasonable price on the used market, eventually.

Why? You can stop down to diffraction territory on 24x36 and Sony A7/9 offer IBIS as well, with all benefits of the larger sensor.
Small sensor makes sense if you need/want to crop, though.

Link | Posted on Aug 3, 2018 at 12:01 UTC
On article Nikon counts down to new mirrorless (1057 comments in total)
In reply to:

jazja: I do not trust too much work with the adapter.
Too many possible problems.
- speed and accuracy of focus
- picture quality (same quality? I very doubt)
- Compatibility with old Nikon and of course Zeiss and Sigma lenses
- and of course price of adapter.
It could be very high.

It's probably worth not forgetting that this whole thread started with someone who does not trust a Nikon adapter adapting Nikon lenses to a Nikon camera.

Link | Posted on Aug 2, 2018 at 15:34 UTC
On article Nikon counts down to new mirrorless (1057 comments in total)
In reply to:

LadislavCZ: Its clear message from Nikon about the move from DSLR to mirrorless. They had no choice anyway as well Canon. Its also their financial managers wet dream because they will save a lot on some parts like mirror, mechanism for mirror, dedicated AF sensor, Pentaprism. All these costs in million units is huge saving.

Also its clear message the old mount will become obsolate in few ears from now. Same happend when transitioned from film era to digital. I know a lot of people dont like to hear it but its fact.

Also the adapter will only work with G lenses. Nothing else so this will make more Nikon users upset. I will not reveal the source ;) as I would expect straight question how you know...I know ;)

It is entirely possible to create an adapter that drives a mechanical aperture, look at Sony's adapter for Minolta/Sony A-mount lenses.
Whether or not that will be worth the developmental effort for Nikon is another question.

Link | Posted on Aug 2, 2018 at 15:30 UTC
On article Nikon counts down to new mirrorless (1057 comments in total)
In reply to:

jazja: I do not trust too much work with the adapter.
Too many possible problems.
- speed and accuracy of focus
- picture quality (same quality? I very doubt)
- Compatibility with old Nikon and of course Zeiss and Sigma lenses
- and of course price of adapter.
It could be very high.

Why so skeptical. Sony has a Minolta/Sony adapter that works okay (actually, several). Sure, Nikon mount is a bit different than Minolta, but I am sure they can do it.
And why would image quality be worse? It won't have glass and just be a light-tight tube with electronics and maybe a tiny motor.

Link | Posted on Jul 30, 2018 at 14:12 UTC
In reply to:

Herp Photos: What I really dont understand is how this lens from fuji is almost twice as much as their medium format GF250 F4 which is pretty equivalent. https://m.dpreview.com/products/fujifilm/lenses/fujifilm_gf_250_4_r_lm_ois_wr

I am so disapointed that Fuji opted for this mid length lens instead of something that wasnt already covered adequately by existing lenses. A 500mm F4 would have been a much better expansion to the system for wildlife shooters. The fact their new roadmap shows no hope of such a long lens but more duplicates of existing lenses or ranges is crazy.

> I am starting to repeat myself, so... point is, stop using argument that a lens should be certain price due to "the equivalency" against some other system.

I haven't done that as far as I remember. I just pointed out that the lenses that the OP claimed were equivalent (FOV/DOF) aren't.

I do think it is generally okay to compare different systems with equivalent set-ups and see which is more expensive and maybe find out why.

And yes, a 200/2 will always have at least 100mm of front group and while subsequent lens elements/groups can be smaller on a smaller sensor, there is no price-equivalence as such.

But the point is of course, and now I am repeating myself ;) you would have to compare the 200/2 with a 300/3 or a 400/4 on larger formats. And these, too, have 100mm front elements.

The closer we get to a 70/0.7 lens (the absolute physical limit), the more expensive the lens would become, but there's no clear rule as to how exactly that would translate into cost or price.

Link | Posted on Jul 23, 2018 at 21:21 UTC
In reply to:

Herp Photos: What I really dont understand is how this lens from fuji is almost twice as much as their medium format GF250 F4 which is pretty equivalent. https://m.dpreview.com/products/fujifilm/lenses/fujifilm_gf_250_4_r_lm_ois_wr

I am so disapointed that Fuji opted for this mid length lens instead of something that wasnt already covered adequately by existing lenses. A 500mm F4 would have been a much better expansion to the system for wildlife shooters. The fact their new roadmap shows no hope of such a long lens but more duplicates of existing lenses or ranges is crazy.

I am not sure what your point is then, umeet.
Of course a 200/2 is a 200/2, but a 200/2 also gives vastly different results when attached to a tiny or a huge sensor (assuming it fully covers the respective sensor) - equivalence helps us to compare that effect and put it into perspective (bad pun intended).

Link | Posted on Jul 23, 2018 at 14:32 UTC
In reply to:

falconeyes: Tested it out.
(1) Works relatively fast (MacOS 10.13.6 and .NEF, didn't update my rather old ACR)
(2) Crashes with filesystem problems (boost::filesystem::create_directories: Invalid argument). I didn't specify an output directory ...
(3) Results are as expected, i.e., aligning seems to work
(4) It runs ads during rendering
(4) There is a risk to install this (small Chinese company). But it is established and runs an Australia research center. I'd rather install KDRaw than a free tool from a single US or EU citizen ...

Summary: A viable alternative to PhotoAcute which is discontinued.

Thank you Falk,
I currently push individual exposures in Lightroom (to make use of RAW latitude), then export as JPEGs and stack them in my own simple Matlab code. Maybe I should just see if I can revise it to stacking TIFFs (I don't have Photoshop).
But staying DNG->DNG would be great.
Do you know of an "API" that would allow me to access the DNG RGB data from Matlab (or another language I could learn)?
Edit: Might be worth me trying this approach:
https://blogs.mathworks.com/steve/2011/03/08/tips-for-reading-a-camera-raw-file-into-matlab/

I haven't found a way to use the Pentax multi-exposure feature for my style of photos, I need as little time delay between individual exposures and you cannot use burst mode with multi-exposure, as far as I know.

Link | Posted on Jul 23, 2018 at 08:29 UTC
In reply to:

Herp Photos: What I really dont understand is how this lens from fuji is almost twice as much as their medium format GF250 F4 which is pretty equivalent. https://m.dpreview.com/products/fujifilm/lenses/fujifilm_gf_250_4_r_lm_ois_wr

I am so disapointed that Fuji opted for this mid length lens instead of something that wasnt already covered adequately by existing lenses. A 500mm F4 would have been a much better expansion to the system for wildlife shooters. The fact their new roadmap shows no hope of such a long lens but more duplicates of existing lenses or ranges is crazy.

> Only differences between MFT, APS-C, FF, or MF 200/2 lens are
> 1) the image circle they project
> 2) the flange distance

I have an issue with your use of "only", as it implies that these are small things. Especially the importance of image circle is huge, because it tells us on what size imager (sensor/film) it can be used and knowing that size tells us so much about the result.

If you have a specific concern with my (or anybody's) use of equivalence, please address it directly to facilitate a discussion.

It is important to remember that no one takes a photo with just a sensor or just a lens, we take photos using a combination of the two.

Link | Posted on Jul 23, 2018 at 08:22 UTC
In reply to:

Lettermanian: Hmm, no filter thread. I wasn't in the running for this lens anyway, but still a strange omission. Fuji must have known that some customers would want to use this lens for landscapes.

That lens is equivalent to a 16-36 and hence considerably less wide. Plenty of equivalent 16/17-Xmm zoom lenses offer filter threads.
12mm FF equivalent is seriously wide.
Closest I can think off is the Sony 10-18 which is at least equivalent to a 15-28, but even that is much less wide than 12-24.
I need filters, too and I share your despair, but these modern ultrawide zoom lenses tend to have such bulbous front elements it makes filters quite unfeasible (and those square filter adapters, brrr....)
There are some Leica-mount Voigtländer alternatives for Sony Nex, if you don't mind manual focus primes.

Link | Posted on Jul 22, 2018 at 17:13 UTC
In reply to:

falconeyes: Tested it out.
(1) Works relatively fast (MacOS 10.13.6 and .NEF, didn't update my rather old ACR)
(2) Crashes with filesystem problems (boost::filesystem::create_directories: Invalid argument). I didn't specify an output directory ...
(3) Results are as expected, i.e., aligning seems to work
(4) It runs ads during rendering
(4) There is a risk to install this (small Chinese company). But it is established and runs an Australia research center. I'd rather install KDRaw than a free tool from a single US or EU citizen ...

Summary: A viable alternative to PhotoAcute which is discontinued.

Hi Falk,

Do you know of any software that combines DNGs into DNGs that does no auto-align and performs a standard averaging?

I seem to get weird ghosting when aligning my long exposure images, I think it catches the moving clouds as "fixed" whereas the landscape gets adjusted, but it is a bit unpredictable.

I just tried with a bridge at night, where the main lights are fixed and pointy - a few lit boats are moving but sky and water are very low detail and the algorithm does not create any ghosting or loss of sharpness. Surprisingly effective - it eliminates the light trails from the moving boats.

Works on Pentax K-1 and K-01, but 16 images is sometimes a bit limited. But I guess I could combine batches of 16 images each?

Best wishes
Jens

Link | Posted on Jul 22, 2018 at 13:17 UTC
In reply to:

Lettermanian: Hmm, no filter thread. I wasn't in the running for this lens anyway, but still a strange omission. Fuji must have known that some customers would want to use this lens for landscapes.

AFAIK, no competing zoom lens offers a filter thread - can you name one?
Certainly not in the dSLR world, not Sigma's 8-16, not Sigmas 12-24, not even Nikon's 14-24 despite lesser zoom range and less wide angle.

Link | Posted on Jul 22, 2018 at 10:38 UTC
In reply to:

Herp Photos: What I really dont understand is how this lens from fuji is almost twice as much as their medium format GF250 F4 which is pretty equivalent. https://m.dpreview.com/products/fujifilm/lenses/fujifilm_gf_250_4_r_lm_ois_wr

I am so disapointed that Fuji opted for this mid length lens instead of something that wasnt already covered adequately by existing lenses. A 500mm F4 would have been a much better expansion to the system for wildlife shooters. The fact their new roadmap shows no hope of such a long lens but more duplicates of existing lenses or ranges is crazy.

"pretty equivalent"
One lens is equivalent in 135 terms, to a 300/3 the other to a 200/3.2.

Link | Posted on Jul 22, 2018 at 10:31 UTC
In reply to:

PowerG9atBlackForest: Apart from forensic sciences, what is wrong with taking (tampered) photographs as they are?
Good fun/marvel at the picture at best. Does one have to know the "truth"?

> is pretty much close in my book.

Not in mine and hence we won't reach an agreement.

Link | Posted on Jun 29, 2018 at 11:31 UTC
In reply to:

Alex Sarbu: The K-50 also has KAF4 support via a firmware update. Is it missing from the list a mistake, or there is some other issue?

If you give a list of compatible cameras, the K50 should probably be included, that's all that matters. The leave it to K50 owners to make the decision.

Link | Posted on Jun 29, 2018 at 11:29 UTC
In reply to:

Richard in UK: Went in to my local 'Jessops' hoping to handle the Pentax K1ii, but was told they no longer stock it; they're only available in the London store.

My first camera was a Pentax Spotmatic which was brilliant. Whilst I certainly have a 'soft spot' for Pentax, this does feel like £1000 of my money well spent.

We all (even Pentax fans) have to move on; If you're likely to be taking photographs in ten year's time; then this isn't a sound investment.

I am guessing that SRS Microsystems sells 80% of high-level Pentax gear in the UK. It is sad but not surprising that Jessops is not stocking K1 cameras in satellite stores.

Link | Posted on Jun 29, 2018 at 11:19 UTC
In reply to:

PowerG9atBlackForest: Apart from forensic sciences, what is wrong with taking (tampered) photographs as they are?
Good fun/marvel at the picture at best. Does one have to know the "truth"?

Photography is not just an art, it is also documentary. Which mixture a photo should have depends on the context. I appreciate a good dose of documentary in my advertisements.
I am baffled that badi would even compare two photos from different locations at different times to manipulating advertisement photos. Everybody knows that when they look at a photo, that clouds can be there, that it can be dark or bright or winter or summer. That is NOT the same as removing ugly buildings or cloning in palm trees.

Link | Posted on Jun 29, 2018 at 11:02 UTC
In reply to:

JensR: Not sure why you picked the 58mm, when multiple lenses are covered.
The patent is very interesting from a technical point of view and it seems this would be the first application in an AF lens?
Would be nice to link to the source/patent.
And combining this rather complicated focus arrangement with the zoom function is probably of even more technical interest than in the prime lenses.

good point! I couldn't find a detailed description of the Minolta process. Looking at the sample photos, it looks as if it might be a similar way to increase spherical aberrations. I'm sure the Canon patent does something new, which just furthers my point that I would like a more detailed technical discussion of this technology.

Link | Posted on Jun 29, 2018 at 10:55 UTC
In reply to:

Tungsten Nordstein: Great, but please can we stop calling this stuff AI? It isn't.

Tungsten, falconeyes has literally stated that right at the beginning:
https://www.dpreview.com/news/5039410137/adobe-ai-spots-tampered-images-by-focusing-on-noise-and-artifacts?comment=1541600109
There is nothing more to it: "Artificial Intelligence" is an established term by now and the work here meets that definition.

> "as I understand the term AI"
If you disagree with how that term is defined, that's another discussion. But if you don't understand its definition, you can read up on it.

And whether or not you are a "victim" or feel like one, your objection is exactly what is described in the Wikipedia effect on the "AI effect".

Link | Posted on Jun 28, 2018 at 19:19 UTC
In reply to:

HeyItsJoel: I read somewhere that 58mm more accurately depicts a 'normal' human eye FOV than the standard 50mm. Is this true?

dpreview linked a good article a while ago:
https://www.theatlantic.com/technology/archive/2018/05/how-the-50-mm-lens-became-normal/560276/

I personally like a variation of a phrase I read years ago: Human vision is a very good tele lens merged with a decent "normal" and a poor wideangle.
If you are standing at some scenic lookout and gaze out into the wide open landscape, brain&eye create a very different feel than when you try to decipher tiny print...

Link | Posted on Jun 28, 2018 at 19:13 UTC
Total: 486, showing: 21 – 40
« First‹ Previous12345Next ›Last »