This is another note to 'Electronic front curtain shutter severely spoils bokeh'. This time the bokeh is spoiled by the ND variable density filter.
A variable density ND filter distorts the out-of-focus area. Artifacts in bokeh. Enlarge Image.
Photos in the gallery below were taken with a cheap Chinese Cuely Fader ND2-400 52mm (S) variable density ND filter and lens Viltrox AF 56mm 1: 1.4 STM ED IF CE on camera Sony a3500, JPEG straight from the camera (sources). In the bokeh, you can clearly see the 'chewiness' and 'wear' of the disks of blur:
UPDATED: How to avoid? Use better quality filters, conventional ND filters, gradient filters and polarizing filters.
UPDATE 2: here a simple and illustrative article on the same topic was published.
My experienceA: I've never had any luck with a good ND variable density filter. When choosing such a filter, it is better not to save money.
Comments on this post do not require registration. Anyone can leave a comment.
Material prepared Arkady Shapoval.
A very useful remark, I will know if I wanted a variable NDshnik for myself in the future! 👍
Thank you for the article! I've read a lot, but for the first time I see a mention of this problem!
variable ND are two polarics. and a polarik is a film of a cunning composition. from time to time it collapses and becomes covered with small cracks. which is clearly visible in reflected light.
What you wrote about does not apply to these artifacts in any way. They have nothing to do with “cracks”. They appear due to the use of linear and circular polarics to create ND
no. a working polarik does not give any artifacts. checked repeatedly. If you don't believe it, write to Dima Evtifeev and ask. he will say the same.
You just do not confuse a polaric and a variable density ND, which consists of two polarizers. You can write to him :)
boyar, I never confuse anything, because my mistake will be very expensive :) and for the third time, what you see in your photos is a defect of a particular specimen, and not a feature of any polar and / or ND filter.
if i am wrong i will apologize and correct the information
take a normal filter and see for yourself. why do you need someone's opinion))
but the fact is that I once used a similar nd-shnik from hoya 77mm, it was the same, and it was new, straight from the box (like this Chinese). And also, I have repeatedly encountered similar problems from other users who also had quite expensive solutions. Otherwise, I wouldn't be so sure. At the same time, I could be wrong
excuse me, but where does it say in the article that this effect is “a feature of any polar and / or ND filter”? I read: “How to avoid? Use conventional ND filters, gradient filters and polarizing filters. ” Those. the author says that with “any polaric and / or ND filter”, this effect will not be.
No, I write "A similar effect appears with any variable density ND filter, including very expensive ones."
do you really need to clarify the obvious? the author wrote about vario nd, my comment also applies to vario nd
I admit I'm wrong. Apparently someone wrote Dmitry Evtifeev and he made an interesting comparison, described here.
The material has been updated. Good information is more important than personal opinions.
Well, the identity also speaks about the uselessness of purchasing expensive varioND. If over time the film becomes covered with cracks. It's like a photoelectric colorimeter that you throw away after 5 years, but it's expensive.
and what is wrong with the colorimeter? more precisely, with which colorimeter?
Shadrin once had a detailed explanation. He strongly discourages colorimeters, preferring spectrophotometers.
I may be imprecise in the definitions, but something like this.
A colorimeter (like all inexpensive Spyders) works like a single camera cell or a set of three retinal cones. It registers the emitted light, passes it through three filters with known spectral characteristics, receives a certain value in the form of three responses, then applies a known transformation matrix to a given color space and receives a specific RGB value in it. Over time, filters degrade, their characteristics change, and accuracy floats. Plus, only monitors can be calibrated.
The spectrophotometer is designed differently, it irradiates the surface with its own light source (some kind of light bulb), then measures the reflected light, passing it, for example, through a prism. Reflected light - you can get the radiation spectrum of paper, for example, or some material - this is how printers are calibrated or make sure that the dye meets a certain standard. Plus, you can identify metamers - that is, colors with a different spectrum that the eye sees in the same way. But when the lighting conditions change, these previously identical colors become different. Cameras also have this kind of bullshit when the camera does not distinguish between shades of blue and purple, for example, because it has its own spectral characteristics that do not match the characteristics of the eye.
Since the light bulb can also degrade, the spectrophotometer itself is also calibrated before starting work. It is covered with a plug covered with a special white paint, which has a clear spectrum that does not degrade over time. As a result, even if the light bulbs have slightly changed their spectrum, the spectrophotometer measures the emission spectrum of the lamp and takes it into account, compensating for the measurement spectrum from a monitor or paper. Something like this.
you have mixed horses, people….
0. The spectrophotometer is generally irrelevant here. it is for the other.
1. Normal interference filters do not degrade. the problem may be with the photodiodes behind them, but this is also programmatically compensated for without any problems. Perhaps we are talking about something super-cheap, but this is already outside my competence, I have no business with junk. and for myself I made color analyzers, as with one photodiode and a "carousel" of filters above it, there and with three independent photodiodes, each with its own temperature-compensated logarithmic amplifier. everything works exactly. calibrated using proprietary instruments and monochromatic radiation sources.
2. in general, my question was to the author of the commentary, with which device he had problems, it just became interesting.
For what for another? In the context of photography, everything is for the same thing - for calibrating monitors and printers.
I already understood that your interest comes down to telling once again how you personally collected something from a colorimeter to a nuclear reactor and what idiots are all that use ready-made solutions.
calibration of the monitor or something self-luminous - either a colorimeter based on photodiodes plus filters, or a spectrometer, that is, a prism plus a CCD ruler.
calibration of something not self-luminous - a spectrophotometer, that is, a spectrometer plus an external light source.
I repeat the question to the AUTHOR of the comment, not you. which device was out of order and why.
in general, I developed a lot of things, both personally and as a team. For example, the first SOVIET digital video recorder was developed by us, at the SKTB “Video” of the “Tantal” software. unfortunately, by the time the research was completed, the country collapsed and further our developments were not used.
and already in the post-Soviet period, working in one of the joint ventures, we developed an acousto-optical spectrophotometer, this development proved to be successful.
And, well, having received instead of three values, for example, 30 or there 60, we do not have the response of some device - what it considers blue there, but what is green with its filters, but a more or less objective spectrum, from which the software will already count the exact XYZ values or Lab there, which are based on a person's color vision.
Could a respected author give similar photos from some proprietary variable plane polarizer to make sure of the global nature of the problem?
It would be nice to try a combination of two polarics from the outdated “linear” technology, plus “circular” plus “linear”, etc.
Dmitry Evtifeev wrote in his blog about the difference in the picture depending on the linear and circular sequence (but there the difference concerned the UV and IR regions).
Now I can’t, but it’s not a secret that these artifacts are associated with the construction of a variable nd filter, and not related to “quality / manufacturer”. Typically these filters are used for landscapes where there is no bokeh.
It's not new to me. I know the price of showing up in the sale of such filters, such as cheap Chinese and expensive solid virobniks. For example, MARUMI DHG vari. ND2-400, yakim had a chance to koristuvat, mayzhe so zosiraє bokeshki. With a lot of yaskravish mugs, they have bigger artifacts.
Inspired by this, the filters were invented not for photos, but for video recording with mirrorless cameras.
I noticed a cross-shaped shadow in the center of the frame. I was so upset that I did not save an example. Filter - Fujimi ND2-ND400
Even the manufacturers themselves write about this cross.
http://evtifeev.com/78976-iskazhaet-li-svetofiltr-peremennoj-plotnosti-boke.html
Thank you, I updated the information, I admitted mistakes (see comments above)