Share your Coin Photography Tips & Tricks!

[Kaleun96]

8 hours ago, kirispupis said:

I also appreciate the recommendation of the beamsplitter, and I ordered three different ones to experiment. I'm still using the GND, but these look to be much better alternatives.

I've made some alterations to my setup as well. The biggest one is I switched my speedlight for an Einstein 645. I was a bit surprised that the light output is weaker, though, and I believe it's because the speedlight had a much smaller head compared to the Einstein and therefore provided more directed light even though it's much less powerful. I've attempted to address that by ordering a different reflector that adds a stop of light, along with a honeycomb grid as mentioned by @HipShot Photography. I've also thought of using an optical snoot, but since it has the same budget issues as the stand, I'll see what the honeycomb grid can do first.

Nice! Based on my experiences with the beamsplitter so far, I think you'll at least gain a lot more light from its better overall transmission (% of light that makes it from the light-->coin-->camera) than you might've lost to changing light sources. I'm just using a Godox AD100 and it has more than enough power. I think I did the tests above at 1/128 on the flash and f5.6 ISO 50 1/250s on the camera. With the UV filters I had been testing previously, I was probably at 1/32 on the flash and f5.6 ISO 200 1/120s on the camera.

[Roerbakmix]

I have been following this 10-page thread with much interest, not per se because I'll follow up on the advice here, but mainly because I enjoy the strive for perfection. Please continue!

[kirispupis]

Yesterday, I received the new reflector. Below is the first coin I photographed with it, a small AE from Carthage. I'm guessing it added about two stops of light, which is significant.

 

Then today I received the beamsplitters. I ordered three different kinds and I'm still playing with them, but so far they definitely add more light. Here's the same coin with the "medium" beamsplitter.

 

So far I'm very pleased, though I have noticed the following issues.

1. It takes quite a bit of playing to get the angle right. It's extremely easy for the reflection to hit the coin again, which basically ruins the shot.
2. I'm seeing a magenta tinge in some shots. You can see it a bit on the reverse.
3. The strongest beamsplitter provided the most light, but the reflections were so strong that I moved down one.

Going forward, I believe I've graduated to a point where I need to be more inventive about how I hold the beamsplitters (I currently use a clip). The position needs to be exact and I don't have the ability to precisely place it yet.

These shots are still handheld. I'll need to look into a proper stand eventually, though I just blew too much on coins so I'll have to wait a bit on that. 🙂 

[Kaleun96]

20 hours ago, kirispupis said:

Yesterday, I received the new reflector. Below is the first coin I photographed with it, a small AE from Carthage. I'm guessing it added about two stops of light, which is significant.

 

Then today I received the beamsplitters. I ordered three different kinds and I'm still playing with them, but so far they definitely add more light. Here's the same coin with the "medium" beamsplitter.

 

So far I'm very pleased, though I have noticed the following issues.

1. It takes quite a bit of playing to get the angle right. It's extremely easy for the reflection to hit the coin again, which basically ruins the shot.
2. I'm seeing a magenta tinge in some shots. You can see it a bit on the reverse.
3. The strongest beamsplitter provided the most light, but the reflections were so strong that I moved down one.

Going forward, I believe I've graduated to a point where I need to be more inventive about how I hold the beamsplitters (I currently use a clip). The position needs to be exact and I don't have the ability to precisely place it yet.

These shots are still handheld. I'll need to look into a proper stand eventually, though I just blew too much on coins so I'll have to wait a bit on that. 🙂 

Nice, thanks for sharing your results! The beamsplitters can definitely be a bit sensitive to minor adjustments and are a pain to hold in place. The first axial setup I tried several years ago involved making a holder from foamcore (see photo below), which should be available cheaply from hobby stores. It's easy to cut and glue, and black foamcore does a decent job at preventing additional reflections. So I can recommend something like that if you need a cheap solution.

Two things worth checking in your setup, if you haven't already, are: having some material on the opposite side of the glass to the light source that absorbs the light and prevents back-reflections, and taking a photo with the setup exactly how you would use it but minus the beamsplitter. That's a quick test you can do to see how much non-axial light you're letting into the image. You probably want some but you can also check if the non-axial light is somewhat evenly distributed across the coin or you're getting a "leak" from the main light source that just hits a particular part of the coin.

 

Here's my old foamcore setup for reference, back when I was using an 8x10 picture frame glass or something like that.

 

 

[Kaleun96]

I've been quiet here for awhile but have some updates to share for those interested. I've been growing a bit tired of my current lighting solutions for coin photography and wanted to tackle some of the problems I have with it. This will probably be a multi-part post and to adequately explain everything I have to start with an explanation of my current solution. Skip the sections on axial lighting if you're already familiar with that.

Part 1

Goals and Axial Lighting

My goals with lighting  for coin photography are mainly reduced to: (a) recreate the axial lighting effect as close as possible, and (b) modify the lighting slightly to reduce the harshness of this effect while keeping its benefits. In other words, replicate axial lighting without the bad parts.

Axial lighting is where the light hitting the object is parallel to the axis of the camera' lens, i.e. the light is hitting the object at a perpendicular (90 degree) angle, which by the law of reflection means the light reflecting back is also at 90 degrees and goes straight back to the camera. Axial lighting has several benefits:

First, light hitting the coin perfectly perpendicularly (i.e. from straight-down) creates contrast by varying how much light is reflected to the camera based on the contours of the coin. The fields of the coin are mostly flat so they reflect most of the light back to the camera, producing bright fields that bring out the toning of the coin and features like flow lines and lustre. Parts of the coin with bumpy surfaces, like a portrait, reflect different amounts of light back. Flat areas like the cheek might reflect a lot of light, while the edges of the portrait reflect most light away because they are angled. This means the edges of the portrait are darker than the main parts of the portrait, creating contrast between the portrait and fields of the coin. However, the portrait is not uniformly bright, the slight variations in surface flatness reflect different amounts of light, slightly darkening contoured areas and accurately capturing the depth of the features. You can see an example of this in this axially-lit photo.

Second, axial lighting provides consistency to your photos. You don't have to worry about getting the angle of the lighting just right or recreating the precise setup you used the previous weekend, instead it should be relatively easy to replicate the exact same lighting conditions.

Third, axial lighting doesn't use shadows to create contrast and thus you don't risk losing detail due to insufficient light in some areas. Typically people use shadows to create contrast and depth in their coin photos by having the light at an angle relative to the coin. Some light will be blocked by the features of the coin, producing shadows on the opposite side of those features. This can make for inconsistent photos due to the shadows varying between coins and also you can obscure detail or make it hard to see by covering it in shadows. Axial lighting only produces the harshest shadows in areas that are parallel to the light source, almost everything else will reflect *some* light, it will just be less bright than other areas.

Lastly, blown-out highlights are limited and easy to control. When using off-axis lighting from an angle, it's very easy to produce bright white areas on the tops of portraits. Axial lighting doesn't have this issue because it's much easier to evenly light the entire coin since each part of the coin is receiving the same amount of light. When you're lighting off-axis you often have to increase the light power to get parts of the coin furthest from the light to appear bright enough on the camera.

Downsides of Axial lighting

So, if axial lighting is so good, what's the problem? The main problem with axial lighting is getting the light to be axial to the lens. You can't put a light inside the lens itself, only around the lens. But if the light is around the lens and the coin is aligned with the center of the lens, how do you get axial light, surely the light can only hit the coin at an angle? This is exactly the problem.

As mentioned before, the idea is to place a mirror or piece of glass between the lens and the coin at a 45 degree angle. You can then reflect horizontal light off of the glass, down onto the coin at a 90 degree angle, which then reflects back up towards the camera.

There are two problems with this: (1) putting a piece of glass between the lens and coin introduces optical aberrations and defects. You lose sharpness, ability to resolve detail in the coin, and may have other optical issues like chromatic aberrations. (2) you need a lot of light. If your mirror reflects 50% light and transmits 50% light, when the light first hits the mirror you only have 50% of it being reflected to the coin, and when the light reflects off of the coin back to the camera, only 50% of that is transmitted through the mirror. This means only 25% of your original light actually makes it to the camera.

Current solution

To avoid these problems, I've been trying something called "pseudo-axial lighting", which means trying to get the light as axial as possible without using the mirror method above and avoiding the issues it introduces. My idea was to recreate a ring light except one powered by my flash. It would be mounted around the lens and try to reflect light down onto the coin at a slight angle, achieving most of the axial effect.

The screenshots below of the 3D model are of my standard flash adapter. The green bit on the left is detachable and is where the Godox TT350 camera flash head is attached and held in place through friction. The grey object has a hole in the middle about 30mm in diameter, the other side of which has threaded adapters to mount to my camera' lens. The wide opening at the top is about 65mm in diameter and faces the coin. The third photo shows my "dual flash head" version of this so you can see how it is positioned relative to the coin.

The goal is to get light from the flash down onto the coin at the highest angle possible - we want the light to hit the coin as if its coming from straight down, the idea being that the light then reflects straight up into the camera. The device below is not going to achieve this perfectly. Light comes in horizontally from the flash head and hits the grey cylinder, reflecting light in thousands of different directions, some of which will exit the cylinder and go towards the coin. One hopes that at least some of this light is hitting the coin roughly axial to the lens but a lot of light is not going to be, hence "pseudo-axial".

 

Problems with this method

The main problem with this method is the 30mm hole in the middle needed for the lens to see the coin. This hole means no light can be reflected from the flash head towards the center of the coin from this area. In practice that means more light is hitting areas around the coin then in the middle of the coin. What then happens is that you lose the axial effect in the middle of the coin and instead end up with diffused lighting that doesn't produce a lot of contrast in these areas or capture the depth of the features.

You can see this in the following picture. Look at the side of Baal's leg, his himation, and the left part of his chest and face. Similarly on the lion pay attention to the main part of the lion's body. All of these areas look flat, they all have a similar amount of brightness so our eyes interpret that as meaning these features are all the same distance from our eyes, thus are perceived as flat and lacking depth. In aesthetic terms, I find that this can make the coin in general appear flat, lacking in contrast, and reducing the aesthetic quality of the coin.

So how do you fix this? Really, there's no easy way. This solution requires a hole in the middle of the device so that the lens can see the coin but as long as that hole is there, you can't have light reflected it from it axially and the middle of the coin has a weakened axial effect. I have reduced the axial effect on purpose in this photo anyway, by using a second off-axis light source, but ideally you have full control over your lighting. The way to achieve this is by having as perfect of an axial lighting effect as possible and then dialling it back in certain areas in a controlled manner. Currently, I'm getting the axial effect in some areas of the coin but not in others (i.e. the center) and I have no control over that.

I've iterated on this general design many times to try and mitigate the issue. You can see some of these designs below, which probably represent less than half that I've tried over the past 2-3 years. The main things I've tried are: (1) adding a second flash, hoping that it can light parts of the coin the first flash couldn't, (2) adding modifiers that attach on the outside to help direct light towards the center, and (3) adding modifiers on the inside like angled surfaces or "spouts" in the middle to help direct light towards the center. Some of these worked better than others but still aren't good enough in the end.

 

Part 2 to come...

Edited January 14 by Kaleun96

[Kaleun96]

Ok continuing on from above...

Part 2

Improving "Pseudo-Axial"

As mentioned above, the issue with my pseudo-axial device was that it couldn't get approximately-axial light to hit the center of the coin because light could only come from around the coin in the flash adapter above, around where the lens sits, making it difficult to force light to cover the center of the coin ~75mm down below.

One idea I toyed with a lot is directing the light to that particular spot using angled surfaces inside the flash adapter. For example, instead of having vertical walls inside the flash adapter, what if it was a 50 degree slope? If it were 45 degrees, it'd direct light axially towards the coin but the light would miss the coin entirely since the angle surfaces in the flash adapter can't be over the coin itself - so, we need a slightly wider angle to push the light towards the coin.

The lines in blue show the light path coming horizontally from the flash head, reflecting off the 50 degree surface at 80 degrees relative to the axis of the light from the flash head, striking the nearly one whole half of a 30mm diameter coin placed 75mm away, and reflecting off of the coin at an 80 degree angle (relative to the coin's surface). 80 degrees isn't 90 degrees (i.e. axial lighting) but should be close enough! Importantly, with these distances and angles, the light reflected off of the coin still makes it back to the lens, it isn't too wide of an angle to miss the lens entirely.

 

The only problem with this approach is that only light opposite the flash head is going to be reflected in this manner, meaning only half of the coin might receive that axial light. That makes the "ring light" design a bit superfluous if only the surfaces opposite the flash head are needed. The 50 degree slope would also be sharply curved around the adapter, throwing some of the reflections not towards the coin but in random directions away from the coin.

Adapter with 50 degree curved interior surface:

 

What this means is that the 50 degree slope may as well be flat and perpendicular to the flash head, not curved, to maximise the sloped surface area that can redirect light towards the coin. One could get rid of the "ring" and make it a box shape with the wall furthest from the flash a flat sloped surface but in my tests I took a different approach. Instead of mounting the flash horizontally, I instead mounted it under the camera at an angle and used a more minor sloped surface to reflect the light onto the coin.

The New Design

So this is the design I've been experimenting with the past week or two. It's essentially the same as above in terms of angles - the light hits the coin at an 80 degree angle and reflects off of it at an 80 degree angle. Using the geometry of the design, I can also block light that would not hit the coin (since most of my coins are about 30mm diameter or less), preventing any stray unwanted reflections that might happen if I allowed this light to be reflected.

I tried various iterations of this design: one flash head, two flash heads, a box shape, a circular shape, two lights opposite each other, two lights at 45 or 60 degrees to each other, and then using black plastic for the everything except the reflecting surface I want light to bounce off. I don't know why I never thought about this before, though with my earlier designs I was encouraging *some* random light reflections so as to lessen the harsh contrast of pseudo-axial lighting. But these stray reflections were largely adding to the "flatness" problem I described earlier and I found that making the body black except the one surface I want reflections to come from made a huge difference to the flatness.

 

You can see about half way through I started using black for parts of the body (as I ran out of white filament) and when I got around to doing a close comparison of the photos, I noticed ones from that body (3rd from the left, top row) were much better than any photos from the white body devices.

Here's a comparison of how they differ in "flatness". The slightly yellow photo is from a white body, the greyer photo is from one of the black bodies with white only on the reflection surface. The area to pay attention to is around where the lion's front leg joins to its body - the shoulder joint and chest area. Note in the more yellow photo this area seems slightly flatter, while in the other photo you can get a hint of the contours of the lion's chest and shoulder joint. This is from the better control over the axial light and minimization of unwanted reflections.

It's a subtle difference when viewed from afar but makes a big difference to the end result and I think it's something your brain subconsciously picks up on when trying to picture the relief of the design and the shapes of the features.

I'll wrap up Part 2 here and come back later with a Part 3 comparing some final results on this coin. I suspect I'm not done experimenting just yet, and I'll need to test this on more coins, but so far I'm liking what I'm seeing.

 

Edited January 14 by Kaleun96

[Kaleun96]

Part 3

Tests

Time for some tests/examples. I'll start with an example shot using a proper axial light setup, i.e. with a piece of reflective glass between the lens and coin. And then I'll go through some "pseudo-axial" setups, starting with my current flash adapter, one of the new ones I made, and then a final design I've roughly settled on for now.

Just to make sure we're on the same page when I discuss some particular areas below, the parts highlighted in red are my main focus. These seem to be the flattest areas of the design and areas where I want to try and capture some of the contours that are actually there but can appear flat in some lighting situations.

 

Axial Lighting - First up is the pure axial lighting setup. There's great contrast around the designs, the toning is really brought out of the surface, and the focus areas mentioned above don't appear flat at all except for in small parts. The front leg of Baal for example is nicely rounded and the definition and depth can be seen in his chest as well as the lion's. The main negative is that the toning seems too much in places, such as the red, oranges, and blues on Baal's head, chest, and the lion's body. And as mentioned before, axial setups like this do reduce sharpness and detail, though it's practically unnoticeable unless zoomed in.

Pseudo Axial 1 - This is using my current flash adapter, the one I've used for most of my photos. It's still a nice photo with great contrast around the figures and the toning is preserved but much of Baal, and parts of the lion, seem featureless and flat. Baal's chestand front leg in particular but also his arms, the lion's limbs, baal's head etc all are very flat.

Pseudo Axial 2 - This one uses the black device in the bottom row 2nd from the left seen here. It uses two flashes positioned 180 degrees from each other, each reflecting light that should hit the coin at an 80 degree angle. For this photo they were positioned at 12 and 6 o'clock, i.e. top and bottom. Compared to Pseudo Axial 1 (PA1), we have less contrast around the figures but also less flatness. We get some definition back in Baal's legs and chest and the lion's shoulder, chest, and hind area represent the contours of the body just a bit better. We keep much of the toning but the surfaces of the obverse in particular look a bit more natural than in either the PA1 photo or the Axial Lighting photo.

Pseudo Axial 3 - Finally we have what I think is the "winner". It's the black device in the bottom row at the far right in this photo, i.e. a single flash version of PA2. I believe this one balances the pros and cons of axial lighting best. It has the least amount of flatness out of any of the Pseudo-Axial examples while maintaining the toning and contrast. Compared to PA2, the lion has extra contrast on the underside as we're no longer lighting from that direction, only the 12 o'clock direction, but this results in much better depth being perceivable across the lion's trunk and limbs. The downside is mainly on the obverse where the axial effect struggles to make the surfaces pop near the bottom of the coin, as very little direct light can reach here and what light does likely bounces away from the lens. Nonetheless it's still possible to get a relatively evenly-lit photo while the majority of the reflected light is theoretically only targetting the top half of the coin.

 

It can be difficult to see some of the nuanced differences from images side-by-side so the GIFs below alternate between two photos to make it a bit easier:

AL1 vs PA3

PA1 vs PA3

PA2 vs PA3

 

If you made it this far, let me know which one you prefer! I think I'm on the right track with the last one (PA3) and have a few ideas for some improvements but feel free to send some ideas my way if anything occurs to you.

Edited January 16 by Kaleun96

[CPK]

If you had showed me any one of those photos by itself, I would have thought it was a winner! But I think you're right, the last one seems to combine the best of all worlds. Great work!

[kirispupis]

Very interesting work. If I had to pick, I'd go with pseudo-axial 2. For pseudo-axial 3, I'm not a great fan of the edges of the coin. IMHO they should blend into the background, which is more the case with pseudo-axial 2.

[Kaleun96]

8 hours ago, kirispupis said:

Very interesting work. If I had to pick, I'd go with pseudo-axial 2. For pseudo-axial 3, I'm not a great fan of the edges of the coin. IMHO they should blend into the background, which is more the case with pseudo-axial 2.

Thanks!

I wouldn't put too much stock in the edges in these comparisons as that's something I edit in post. Since I use transparent PNGs for the photos on my website by removing the background, I use an edge lighting diffuser to capture detail in the coin's edge otherwise it'll be entirely black and that doesn't look good on bright backgrounds. I then control the edge brightness in Photoshop using inner shadows to dial the brightness back down and add some natural fall-off. I actually forgot to use the edge diffuser on PA-2, which is probably why it seems darker, though I had to manually bring the brightness up before stacking so I could capture some of the detail that was there.

The main problem is that bright edges look bad on dark backgrounds and dark edges look bad on bright backgrounds, so I have to try and strike a balance as it's not guaranteed my photos will always be against a dark background. I think this is something I can also control using CSS so I can have "dynamic edges", so to speak, depending on background, though that doesn't work for when I post my photos to other sites. For the photos above I just used a black background JPG to keep the file size smaller.

[Kaleun96]

Just a minor update this time. So with the angles involved it's difficult to balance getting enough light on the coin at a high enough angle that it will reflect back into the lens after bouncing off the coin. Higher angles tend to mean less of the light can be reflected onto the coin because the geometry simply doesn't allow it but if you use lower angles, you get more light bouncing off of the coin but lose the axial effect and most of this light is going to bounce away from, not back to, the lens making it wasted light.

In one of the diagrams above I mentioned that I designed a "blocker" to prevent some of the reflected light from exiting the device because it will never reach a coin of the diameters I commonly photograph. In the screenshot below, only the light that is allowed to reflect onto the coin is shown and is depicted by the yellow highlighted areas. Behind that space you can see an empty rectangle of sorts, this light would reflect off of the first surface and then hit that triangle shaped blocker between it and the coin.

So you can see I am wasting a fair amount of light but it's not that I necessarily need it either. It would be useful, though, if it can be used to light the coin from the other side and at a similar angle. I worked out that this should be possible with the right angles and came up with this:

 

So the original reflection surface now has a secondary angle at the back, which previously just reflected light to nowhere. It now reflects the light horizontally (shown in green) across the device to a third reflector, which reflects it towards the coin at an 80 degree angle - same as the light reflected on the other side. I'm hoping this will give me a more even axial effect across the coin instead of the majority of that effect being seen on one side of the coin. I've also mounted magnets in the reflectors so that they can be removed and swapped out, in case I don't want this secondary reflection.

One issue I had when initially testing this was that the white plastic reflectors really do not reflect a lot of light. The two reflections required for this secondary reflection (green) meant almost no light actually reached the coin from this side, only from the original side. I then tried covering all the white reflectors in aluminium tape, which has a mirror-like finish, but that reflected too much light! My flash was too powerful at its lowest setting when using the aluminium tape reflectors. So instead I used the aluminium tape on just the initial reflection surface of the secondary reflection to get enough light across to the reflector on the other side but I left that reflection surface and the original reflection surface as just the white plastic as that seems sufficient for now.

Edited January 18 by Kaleun96

[kirispupis]

This has been a fascinating read. I thought I'd mention my current thoughts on where I'll take my system next.

I'm not one for diagrams, but it's a relatively simple setup. I have one Einstein strobe that fires into an axial sheet of glass (the same you referred me to on EBay). I then manually hold the camera above the coin and trigger the whole thing.

As you can imagine, there are serious issues.

• It can be difficult shooting hand-held to get a decently sharp photo. The obvious solution is to move to a stand. You've provided some great recommendations there and I intend to follow them once I slow down enough with buying coins. 🙂
• I currently use velvet as the background. I have a small Lego contraption I attach to a chemistry stand and put the velvet over it. Velvet is an excellent background, but I need to separate the subject from it more. Right now, subject separation is a major pain in post. The solution there is just rebuilding my Lego contraption to accommodate it. The challenge is my wife recently sold all my Legos.
• One thing I've noticed, and it should be obvious, is the same light doesn't work for every coin. In fact, it differs significantly. Slight changes in the position of the glass make a huge difference. I therefore want to build a Lego contraption using gears to allow for minute changes there. That seems to be the most flexible way to handle it and the cheapest vs any specific contraption. The challenge is the same as the previous - that I need more Legos.

My goal is to tackle these shortly after I redo my web site, which is currently taking all my attention.

[Kaleun96]

49 minutes ago, kirispupis said:

This has been a fascinating read. I thought I'd mention my current thoughts on where I'll take my system next.

I'm not one for diagrams, but it's a relatively simple setup. I have one Einstein strobe that fires into an axial sheet of glass (the same you referred me to on EBay). I then manually hold the camera above the coin and trigger the whole thing.

As you can imagine, there are serious issues.

• It can be difficult shooting hand-held to get a decently sharp photo. The obvious solution is to move to a stand. You've provided some great recommendations there and I intend to follow them once I slow down enough with buying coins. 🙂
• I currently use velvet as the background. I have a small Lego contraption I attach to a chemistry stand and put the velvet over it. Velvet is an excellent background, but I need to separate the subject from it more. Right now, subject separation is a major pain in post. The solution there is just rebuilding my Lego contraption to accommodate it. The challenge is my wife recently sold all my Legos.
• One thing I've noticed, and it should be obvious, is the same light doesn't work for every coin. In fact, it differs significantly. Slight changes in the position of the glass make a huge difference. I therefore want to build a Lego contraption using gears to allow for minute changes there. That seems to be the most flexible way to handle it and the cheapest vs any specific contraption. The challenge is the same as the previous - that I need more Legos.

My goal is to tackle these shortly after I redo my web site, which is currently taking all my attention.

Interesting to hear! Great that the eBay glass worked out for you too. It's certainly the best I've yet found. 

Let me know if you come up with anything for the background separation issue. I went down a complicated rabbit hole by using an LED light from under the coin to provide separation. It ends up being similar to the axial light problem but in reverse: you need the light rays from the background to be perpendicular to the coin or else the light will bleed into the coin's edges and blend it with the background slightly. My current solution works great except for a minor focus stacking related side-effect but if you find an alternative solution I'd love to know as it may nullify that issue.

I guess an obvious solution is just distance between the coin and a static background that you light separately. That would avoid reflections onto the coin edge if the distance is sufficient. Unfortunately in my case I don't have enough room for that separation. 

[panzerman]

I have nothing to contribute/ I am as useless as a tool when it comes to photography and driving a standard tranny.

[kirispupis]

1 hour ago, Kaleun96 said:

Interesting to hear! Great that the eBay glass worked out for you too. It's certainly the best I've yet found. 

Let me know if you come up with anything for the background separation issue. I went down a complicated rabbit hole by using an LED light from under the coin to provide separation. It ends up being similar to the axial light problem but in reverse: you need the light rays from the background to be perpendicular to the coin or else the light will bleed into the coin's edges and blend it with the background slightly. My current solution works great except for a minor focus stacking related side-effect but if you find an alternative solution I'd love to know as it may nullify that issue.

I guess an obvious solution is just distance between the coin and a static background that you light separately. That would avoid reflections onto the coin edge if the distance is sufficient. Unfortunately in my case I don't have enough room for that separation. 

In my case, I don't light the background. My situation may be a bit simpler because I have no intentions to ever show my coins on something other than a black background.

You can see the issue a bit here. I had to manually cut out the edges and then "hide" it by selectively burn the most egregious parts, but you can still see where I didn't burn. There are also focus issues on the obverse where I had trouble holding the camera parallel to such a tiny coin, not to mention the lighting issues on the reverse.

Samaria, “Middle Levantine” Series
Circa 375-333 BCE
Minted by Delayah(?)
AR Obol 9mm, 0.81 g, 2h
Bridled horse stepping right; [DY (in Aramaic) above]; all within dotted square /
Winged sphinx, with head of Persian king, seated right; Y[D] (in Aramaic) above; all within dotted square.
Meshorer & Qedar 27; Sofaer 118–20

 

Here's a slightly larger coin using the old method. Back then, I didn't use axial but a twin-light flash. The lighting suffers from the absence of shadows, making it more difficult to discern - especially on the reverse. However, the edges are natural. There was no manual separation necessary, nor did I need to burn. The background between the two was the exact same piece of felt, but in this one it was about 15cm below the coin.

Mysia, Pergamon
450-400 BCE
1.51g
Laureate head of Apollo right
Bearded head right, wearing a Persian tiara within an incuse square. ""ΠΕΡΓA""
SNG BN 1546-8

[Kaleun96]

On 1/19/2024 at 2:07 AM, kirispupis said:

In my case, I don't light the background. My situation may be a bit simpler because I have no intentions to ever show my coins on something other than a black background.

You can see the issue a bit here. I had to manually cut out the edges and then "hide" it by selectively burn the most egregious parts, but you can still see where I didn't burn. There are also focus issues on the obverse where I had trouble holding the camera parallel to such a tiny coin, not to mention the lighting issues on the reverse.

Samaria, “Middle Levantine” Series
Circa 375-333 BCE
Minted by Delayah(?)
AR Obol 9mm, 0.81 g, 2h
Bridled horse stepping right; [DY (in Aramaic) above]; all within dotted square /
Winged sphinx, with head of Persian king, seated right; Y[D] (in Aramaic) above; all within dotted square.
Meshorer & Qedar 27; Sofaer 118–20

 

Here's a slightly larger coin using the old method. Back then, I didn't use axial but a twin-light flash. The lighting suffers from the absence of shadows, making it more difficult to discern - especially on the reverse. However, the edges are natural. There was no manual separation necessary, nor did I need to burn. The background between the two was the exact same piece of felt, but in this one it was about 15cm below the coin.

Mysia, Pergamon
450-400 BCE
1.51g
Laureate head of Apollo right
Bearded head right, wearing a Persian tiara within an incuse square. ""ΠΕΡΓA""
SNG BN 1546-8

I definitely notice the axial and pseudo-axial method makes it harder to get naturally-lit edges in my photos as well. Bit of a pain because it can be tricky to light them without them standing out too much. By the way, why do you have to cut out the edges from background if your photos will always be on a black background? I used to darken the blacks around the edges when I was saving them to black backgrounds but I just relied on Lightroom's auto-mask on the brush to do that.

[Kaleun96]

I've got some examples to follow-up my earlier post that talked about the second reflector to get some light hitting the coin from the opposite site. These aren't the fairest comparisons as the old photos may have been lit with additional lights and not only using my previous pseudo-axial ringlight and the coins have noticeably toned since then.

Aspendos staters are also relatively flatly struck so aren't the best for showcasing depth but these are a coin I typically struggled with for this reason and I think the new photos make some definitive improvements. The only thing I might change is making the newer photos a touch brighter but otherwise I'm quite pleased with them.

Old photos = top / black background

New photos = bottom / transparent background

 

 

 

 

Photos of the new device with the second reflector on the other side.

Edited January 21 by Kaleun96