| Author |
 Replies: 46 / Views: 7,278 |
|
Valued Member
United States
88 Posts |
I have an HP Pavilion laptop running Win 10. I also have a Plugable USB Microscope, but there are two issues with taking pics with that.
1) I would like a lower magnification where the pics don't unduly magnify minor nicks
2) I am trying to take pics of Canadian $2 coins, but the flexible gooseneck simply is not long enough to get pictures of the entire coin when the microscope is in its stand.
I also have a Cybershot digital camera, but I cannot seem to get good, clear pictures that sufficiently mute the glare while getting adequate details of the coins.
I bought a Torchstar metal swing arm desk lamp, hoping to have enough light so that the light would be consistent over the whole photo area, but with mixed results. I will be guying a Soraa 08837 - LED PAR30 Short Neck - CRI 92 bulb. It produces a diffuse wide flood beam with 1190 lumens.
I would appreciate input from the Community, sharing what has worked for you. I want good enough pics that I can create my own graded inventory with pics that accurately reflect the grade.
|
|
|
|
Pillar of the Community
United States
9395 Posts |
You might check out the coin photography subforum.
|
|
Bedrock of the Community
13014 Posts |
If you have an iPhone or same type smartphone, the pictures are basically as good as very expensive setups.
|
|
Moderator
 United States
56855 Posts |
Quote:
2) I am trying to take pics of Canadian $2 coins, but the flexible gooseneck simply is not long enough to get pictures of the entire coin when the microscope is in its stand. Stack books under the stand until you get the best distance,that's what I do and I can take a full coin pic of a large silver dollar. John1  |
|
Moderator
United States
188952 Posts |
Quote:
You might check out the coin photography subforum. This topic was moved to that forum for the proper attention.  |
|
Pillar of the Community
United States
4038 Posts |
|
|
Valued Member
 United States
88 Posts |
Thanks Ray. That's a wealth of information over there!
|
|
Pillar of the Community
Canada
2784 Posts |
Webs ray helped me get to where I am today. that link is and excellent link. follow it ray and other here will help you including myself. here is a 2019 toonie in plastic. its still in the mint plastic holder. take your time dont be afraid to ask questions. thats how we all learn  |
|
Pillar of the Community
United States
9395 Posts |
I've been using something like the "$400 setup" for some years now. It is flexible enough to do whole-coin shots with axial and conventional lighting, or extreme closeups (using a 10x Mitutoyo objective and zerene stacker). It can shoot whole-slab NGC/PCGS shots, and can even shoot translucent coins with lighting from below.
The results may not be better than what you can get from a smartphone (especially the latest ones with 60-100MP sensors), but the setup is much more flexible.
|
|
Valued Member
United States
88 Posts |
Cash flow is a factor for the moment, but you can be sure I am going to move in this direction after the holidays. I will be more than happy to draw on the collective wisdom from experience from you all.
|
|
Pillar of the Community
United States
4038 Posts |
I have yet to see a really good smart phone pic.
...of a coin, that is
Contact me for photographic equipment or visit my home page at: http://macrocoins.comEdited by rmpsrpms
11/09/2019 8:09 pm
|
|
Pillar of the Community
United States
4038 Posts |
Regarding the really high MP count sensors in new smartphones, most of those extra MP will be wasted on shooting macro subjects like coins. The pixel pitch of those sensors is ~0.8um, so diffraction starts to degrade the image at around f1.2. The very best lenses in those phones are f1.5, so the sensors are fine for landscape and portrait use with only minimal degradation. But let's say you try shooting a Cent with one of these. Depth of field at 0.5:1 magnification with f1.5 is very shallow. Typically you need to stop down to around f5.6 to f8 to get enough depth of field to avoid stacking. At f5.6 and m=0.5, the effective aperture is f8.4, which is 7 times the f1.2 DLA of the sensor. This means that the 100MP sensor is reduced to an effective 2MP. So you're paying for a 100MP sensor but only creating images that are effectively 2MP.
Edited to add: the <$400 setup also requires an aperture of f5.6 to avoid stacking. But the sensor of the Canon XS has 5.7um pixels, with DLA of f9. For the same Cent photo as with the 100MP camera phone, the effective aperture is around f10, just a little beyond the DLA, reducing the effective pixels to around 8MP. So while the 100MP sounds great, you can actually achieve 4x the effective image resolution from the lowly 10MP XS.
Edited by rmpsrpms
11/09/2019 11:27 pm
|
|
Valued Member
Russian Federation
172 Posts |
Hello Ray.
It is correct that lenses in smartphone are not good choice for macro, but can't agree with some calculation.
First of all is that no need to stop down to f5.6 with that small sensor. F2.4 is more than enough for the coin to avioid stacking.
And magnification is around X0.2 for typical 4.6mm (in high) sensor, so the effective aperture will be 2.4(1+0.2)= f2.88
According to your calculation effective size will be 17.3MP not only 2MP.
Also with cannon XS you need X0.7 (in high) for same coin and FOV. It give us effective aperture around F10.
The real bad thing will be with attempt to shot only part of coin. You loose doff very fast on large sensor to 0.0X mm.
More megapixels is always a good thing, the smaller pixels increase the sampling resolution.
Airy disk size can be computed in terms of pixel's size, but which is just a comparison magnitude scale, and the pixel size as a measurement unit is NOT the problem. The problem is the diffraction size, regardless of the pixel size. That absolutely does not mean we need larger pixels, which would just be less resolution too. Less diffraction would be the solution.
As if the diffraction ring were somehow magically centered on each pixel? LoL. It already was larger and worse. Larger diffraction size obviously must affect additional pixels. It is NOT about the pixel size. It's about diffraction size.
Edited by Justwalking
11/12/2019 7:51 pm
|
|
Pillar of the Community
United States
4038 Posts |
Hi JW,
The large MP sensors supposedly have 0.8um pixels, and are ~12k x 8k (actually a little bigger but that's close enough), so are about 9.6mm x 6.4mm. Indeed I did calculate the magnification based on width rather than height, so the mag would need to be more like 0.33 rather than 0.5.
I'm actually not sure about calculating the depth of field when operating deep into diffraction. You state that f2.4 is enough, but for normal DOF calc with COC based on actual pixel size it is not. However, since diffraction is smearing the image, is a larger COC acceptable? If so I guess you need to do simultaneous equations to solve for the optimum aperture. Did you do this and come up with f2.4? If not, how did you get f2.4?
Assuming your f2.4 is correctly based on the diffracted COC, and assuming m=0.33 is correct based on pixel size and count, then we are at f3.2 effective. This is only 2.7x the DLA, so a much better situation than I had described, but still pretty bleak. The ~100MP sensor is reduced to ~14MP. This is better than the 8MP for the XS, but only 32%.
|
|
Valued Member
Russian Federation
172 Posts |
I tried with f2.4 on coin surface and did not notice any loss of focus at X0.4 at 1600x1200 monitor. Equation with ideal mathematical lenses not available in real life was not my concern. Additionaly even with 3:1 was not needed to stack: https://www.easyzoom.com/image/123968/popularAnd finally even 8MP is to much to see not resized whole pic on any monitor )) So the 14MP is not better than 8 )) Compared different sensors with % is just funny idea, no more. Need to try real as you did with 16MP HD camera. Unfortunatelly I do not know lenses to resolve 100MP on such small sensor. Edited by Justwalking
11/12/2019 9:20 pm
|
|
Pillar of the Community
United States
4038 Posts |
I see DOF problems at m=0.4 and f4, with 4.3um pixels. I usually need to go to f5.6 or even f8 to fix it. Smaller pixels would just make the problem worse.
I agree that for publishing a full size image, 8MP and 14MP are too big. But the expectation is that large images can be cropped to show smaller details.
To fully resolve 0.8um pixels, you need a lens with at least f1.2 that is diffraction-limited at that aperture. Do these exist? I don't know. There are f1.2 lenses, even f1.0 or f0.95, but I don't think they are diffraction-limited at that aperture. Maybe at the center?
|
| |
Replies: 46 / Views: 7,278 |
Posted 11/06/2019 03:25 am
