Here I perform some experimentation with the Nikon 28-300 adapted to my Nikon Z 50 mirrorless. https://edruthphoto.com/post-production-ai/
Here I perform some experimentation with the Nikon 28-300 adapted to my Nikon Z 50 mirrorless. https://edruthphoto.com/post-production-ai/
I assume you are using the FTZ adapter. It sounds like you are happy with the setup and results.
P.S. Just looked at your link. The images certainly look sharp.
Not sure that you should become a florist....
The flowers look sharp and in fact I would suggest they are too sharp. Th moment you throw in Topaz Sharpen AI, I can guess that this is where that problem arises from.
I agree; some are too sharp for my taste. however, I don't think the solution is to soften them again with a blur filter. I think the solution is to sharpen them less.
This may be a case where all sharpening methods aren't the same. I don't use Topaz, so I have zero understanding of it's underlying method(s). However, the three methods I most often use--Lightroom's sharpening, Photoshop's Smart Sharpen, and a high-pass filter--can yield noticeably different results, particularly when the issue is sharpening edges vs. enhancing texture.
The Topaz product is meant to be used on images that have a slight degree of softness due to camera movement or poor focus. It is not a generic sharpening technique and if you use it on a subject that is already sharp, it will end up over-sharpening the image.
In my experience, for output sharpening, there is no better tool than experience and a standard sharpening tool like Photoshop's Unsharp Mask or Adobe's Smart Sharpen, that Dan has mentioned. Sharpening is very much an opinion; what is too much for one person could be too little for another.
The Output sharpening should always be done AFTER you have resized the image to output size. Some people suggest that you output sharpen while viewing the image at 100% on your screen and I know others that go with 50%. All I will say is that you pick one and stick with it as consistency is important for you workflow.
As Dan has mentioned, applying a Gaussian blur after sharpening, is not a strong approach. Get it right the first time.
Just to complicate matters, when I output sharpen for printing, I often do not sharpen soft areas like water or sky or even high frequency parts of the image (a rocky or sandy beach, grass or leaves of trees in a forest). Not everything in every image needs to be sharpened.
I agree with Manfred--It's often essential to apply sharpening to only parts of an image, or more to some parts than to others. There are a number of ways to do this. The Llightroom/ACR sharpening tools have a very effective masking slider that will progressively protect low-frequency (smoother) areas like the sky. A more flexible approach is to put a black mask on a sharpening layer and then gradually paint on sharpening where you want it (or conversely use a white mask and black brush to remove it). Yet another approach is to use a high-pass filter, as that tends to bring out strong borders and edges and doesn't much (if at all) affect other areas.
I sometimes create two sharpening layers, one with a high pass filter and the other with Smart Sharpen, and then go back and forth between them to see which seems better if I don't have a clear idea in advance. Sometimes I do; if you want fine detail (high frequency) sharpened, e.g., feathers on a bird, Smart Sharpen is usually better.
Whatever the method, if things start looking crunchy, it's time to back off.
I have stopped using unsharp mask except for local contrast adjustments, but that is a matter of personal preference.
Would that include compensating for softness from diffraction?The Topaz product is meant to be used on images that have a slight degree of softness due to camera movement or poor focus.
Last edited by DanK; 21st October 2022 at 12:47 AM.
Of course, my original intent was to demonstrate that the Nikon 28-300mm can be used quite effectively on the Z 7II or Z 50 and does not suffer from softness, especially for such a wide zoom range. This may be due to the Z 7II's 45.7 megapixels and the sophistication of recent Raw file post-production software. So, while some may see the images as too sharp, others may wonder how such a wide ranging zoom lens could be sharp at all. Other lens factors, such as light transmission, contrast, color accuracy etc. can all be remedied in post-production as well. We might ask why manufacturers put such effort into lenses at all if camera firmware and AI (whatever that is really) can remedy all lens sins. But, in reality, we are not yet at a point where firmware and software can remedy all resolution shortcomings. So an image may be made sharp but suffer a loss of tones or some other element not noticeable online but painfully obvious at 24" x 36." But who has the wampum to truly test this comprehensively and by what standards? Printers, paper, even human vision add to the mix of perceptions. The real question is, Are my Zinnia's happy?
If your purpose was to evaluate the sharpness of the lens, I don't think you've succeeded.
An image downsized this much isn’t a good test of lens sharpness, and it's not a test of what 47 MPX produces. The image you posted has been downsized to 2.7 MPX, which hides all manner of sins.
To test the sharpness of the lens, I wouldn’t do any sharpening, or perhaps any sharpening beyond capture sharpening. Otherwise, you are obscuring what you are trying to measure. And I would post a link to the full-size image so that one can see details that are obscured when downsizing to screen size.
Also, the appearance of sharpening created by some sharpening routines isn't the same thing as the actual sharpness of the capture. There is a good explanation of this in the sharpening tutorial on this site. On that page, go down to the illustration under the heading "concept".
Last edited by DanK; 22nd October 2022 at 01:09 PM.
Modern software such as DxO PhotoLab 6 can go a long way toward bringing out detail (Microcontrast and Fine contrast for example) and specialized software tools for sharpening often exceed my expectations. A lens may be evaluated and the evaluation described in a Modulation Transfer Function chart or MTF chart that I describe here: https://edruthphoto.com/z-50-and-z-50mm-lens/ But in the "real world" it is our expectations and perceptions that govern success. So, once again, I assert that some strain at gnats as the big picture escapes them.
I have a friend who has one and based on what I've seen directly and online "a great value lens" is probably more accurate than "a great lens".
I agree in part. I have three Z lenses (50mm, 85mm, and 24-70mm f/2.8) that all distinguish themselves at their available focal lengths beyond what I expect from the 28-300mm. DxO rates the 28-300mm at: Transsmision 5.6 TStop and Sharpness 10 P-Mpix whereas the Nikon 24-70mm f/2.8 rates: 3 TStop and 26 P-Mpix. Unfair comparison considering the price difference but such is life. My contention is that given the aptly stated "a great value," that value is enhanced quite favorably by the attentive application of time and attention in post-production.
Ed - Like Dan, I am going to have to question your photographic knowledge as you are mixing up properties inherent in the lens design / construction and what can be produced in post-processing after the image has been taken.
The lens resolution tells us what the maximum resolution of the lens is when used under ideal conditions (no movement at all through the use of a very heavy duty tripod) and based on the shot is 100% in focus. Screw up either of those and the inherent quality of the lens becomes meaningless.
You do not seem to understand the concept of a T-stop (which is used with cinematography lenses). It measures the light loss from the selections that the lens designer made to produce it. More glass elements (typical of a zoom lens) and the quality of the anti-reflective coatings will impact the T-stop. It will not have any impact on depth of field, but will impact the exposure. If you use through-the-lens metering, the camera will automatically adjust, but if you shoot in manual mode and use an external light meter, it will have an impact.
The P-Mpix is a "bogus" number that DxOMark uses to try to give a generalized performance number on resolution. I can understand why they would want a simple single value to rate a particular aspect of lens performance, but I would argue that it oversimplifies a very complex optical performance issue.
In terms of your web site; if you are writing about lens performance, you should not have (over) sharpened images to demonstrate what you are trying to say. It says more about your retouching skills than anything to do with lens performance; they are two completely different subjects and you have mixed them all up and are confusing your readers / viewers with some rather misleading information. I would not consider buying any of the lenses you have mentioned because your images are a bit "crunchy" as show oversharpening. I would not want a lens that performs to give the image quality you show.
As Dan has mentioned, the 2MP downsampled images do not do justice to your camera's resolution.
I do shot the Nikkor 28 - 300mm and it is now my "go to" travel lens, because it is so versatile. But I know that I have to forego the shallow DoF that my fast lenses give me.
I think it might clarify things to move away from photography to start. The linked post refers to "experimentation". In common speech, that sometimes means just "fiddling around with something". But technically, an experiment is a method of isolating the effect of one variable of interest. If you want to find out whether one medication is effective, you don't compare people getting nothing to people getting the medicine of interest along with three other medications. You design an study to isolate the effect of the one medication, either by design (a true experiment) or by analysis.
This case is similar, IMHO. If one wants to test and then communicate the properties of a lens, one has to isolate those from everything else affecting the image.
On the other hand, if the point is something along the lines of "here's what I can get in actual practice by using this lens, this camera, and a variety of different software packages", there's no need to isolate the contribution of the lens. But then one should be clear what's being shown.
Agree.
A general comment, which bears upon this conversation: I interpret the thrust of the Author’s website defines his pedagogy as being confined within a relaxed guideline system of teaching method.
Whilst recognizing the existence of the science of photography, the author’s writing style displays throughout, a relaxed approach to any strict scientific method and instructional teaching method, which would be applied to the technical disciplines of Photography.
It occurs to me this is articulated in an introductory at the Home Page: The science of photography isn’t particularly difficult but there is a lot of it. The instructor’s job is to teach you what you need to know to be successful.[op. cit.]
WW
My students ask me, "Where is the shutter button?" I now know to make a more studied reply.
To provide irritation and a retort with sarcasm, (if that was the meaning), was not the intent of post #16.
The intent of post #16 was to highlight the teaching style, as it occurs to me, in which the website is encased: that teaching style is neither good, nor bad; and no comment was made neither implied to its worth.
What bears on this conversation is, that the teaching style applies a relaxed approach; the text is conversational and at times humorous; at times lacking strict discipline to technical terms and thus not adhering to strict theory and axiom based scientific instructional teaching methods.
There are many ways to teach and to learn: importantly, teaching to the audiences' needs is an important factor. A
'more studied reply' is not necessarily necessary in all circumstances - 'press this' may be the best and only instruction necessary to use for some situations.
The point is - this conversation - in this thread - discussed "experimentation" and two of the possible definitions and interpretations of that word - each being worlds apart: when reading and commenting it is important to understand context. Post #16 sought to apply an overarching context to the text provided in the link in Post #1, and thus provide context for future commentary here, in this thread.
Quality forum conversations are about forwarding the conversation on the subject and neither personal matters, nor sarcasm. CiC is know for quality conversations.
If E James' feathers were ruffled by post #16, then clearly stated here that was not the intent. If E James chooses to change his teaching methodology, then he should do so after considering valid reasons why that should be done: if on the other hand after considering valid reasons not to change his teaching method, then that's fine too: they are His teaching methods.
What is unfortunate is when professional choices are made as a result of an emotive reaction.
WW
My emotive is general not specific. Take DxO's lens analysis for example. I believe it is a remarkable achievement. Yet, as with everything, if we look for flaw we can find it. Clearly, the quest for a more perfect understanding of everything is a continuing journey.