I was referencing the coin, not the wager but your point is probably fair. Just seemed to have LoCA from the depth of focus… prob my imagination. Cool regardless!
Jman13 wrote:
It’s not LoCa because this is all in focus. However, the lens does have a small amount of fringing on bright specular highlights, and it’s possible that’s some of what you see here, but it’s also important to note that this is etched polished silicon, which is iridescent, with different colors appearing depending on the angle of light and reflection, which I think is also a large part of what this look is.
AdaptedLenses wrote:
I was referencing the coin, not the wager but your point is probably fair. Just seemed to have LoCA from the depth of focus… prob my imagination. Cool regardless!
Ah. It does have a little. It’s not quite as well corrected as their 100mm macro, or the Nikon Z 105. But it also goes way way closer than those lenses too. Given the price I think it performs very well, and it’s very sharp across the frame.
I’ve been very impressed overall, though the biggest issue is that between about 4.2-5x, it can get internal reflections that wash out contrast in two blobs on either side, depending on the light. Incoming light can reflect off the grease that lines the inner barrel and it can make it past the rear baffle. Sometimes it’s no issue at all, but other times it can be a real problem. Dropping down closer to 4x in these situations eliminates that though.
So I redid the stack of the wafer at an angle (also used a different section of the wafer, though this test cell had some abrasion damage). 298 image focus stack at 4:1 magnification - about 10 microns per step. Can't wait for my automated rail to get here...because that was a pain. Worked though, as there's no more banding. (Note that the texture in the flat parts isn't noise...it's just the many reflections of the silicon crystal.)
More nerd stuff. I had purchased an IC die of an Intel 486 DX2, which arrived today. The seller was kind enough to include two 'bonus' dies, having noted I was a photographer. These two include the SuperSPARC CPU, which was a 32-bit RISC processor used in a series of Sun Microsystems workstation computers from 1992. Performance was very similar to the (later) Pentium processors. He also sent me a super tiny little chip which is the Intel 83C196LD...a 16 bit microcontroller that would have been used in a wide range of components such as hard drives, modems, printers, etc. This was first manufactured in 1986, but the one I have was made at least in 1996 or later. Apparently, this chip was still being made well into the 2000s, and may even still be in production now.
The 486 DX2 was a chip that 12 year old super-nerd me lusted after in the early 90s. I just couldn't imagine the raw power. The 486 die is 12mm x 7mm. By this time the manufacturing process had shrunk considerably from the Rockwell processor I shot earlier, such that the details are considerably finer.
486 DX2: Z8 with Laowa 25mm Ultra-Macro @ ~2.8x magnification:
Here's the SuperSPARC CPU, which is actually quite large at around 16mm square. I did some shifted stitching, but messed up one of the quadrants, so this shot is just at 1:1 with the Z 105. Texas Instruments did the manufacturing for this CPU.
SuperSPARC TMS390, Z8 with Nikkor Z MC 105mm f/2.8 VR S @ 1:1:
This is a 100% crop of the central information, with the Laowa 25 @ 2.5x
And finally the Intel 196LD - about 6mm in width:
Z8 with Laowa 25mm at 4.5x:
Nice work!
I feel there might be a Scanning Electron Microscope on your gear wish-list.
Those different items are indeed test circuits. They are basic components (resistors, capacitors, transistors, etc) of various widths and lengths that are used at wafer probe to measure/test the parametric limits of the silicon process. Basic testing of the uProcessor will have been performed at the wafer probe stage too.
If the WAT (Wafer Acceptence Test) data from those 3 drop-ins doesn't pass muster then the complete wafer will be scrapped before the individual working chips get sliced, diced and assembled into a package for final testing.
EDIT:
That red mark on the die indicates that it did not pass test at the wafer level.
It has been 'inked' such that the die will not get picked from the wafer for the costly assembly and final test stage.
Colin
Jman13 wrote:
So more photos of technology. I picked up a silicon wafer that contains 145 Rockwell 5300 Series microprocessors from 1986 (plus 3...somethings..perhaps test circuits?). The wafer is 4" in diameter.
With the Laowa 25mm, here's a shot of one of the processors - this is at about 3.8x magnification. An individual processor is 6.82mm (according to my best measurement with a micrometer) in width. Measuring pixels, the "Rockwell International Corp" lettering is approximately 27 microns in height. https://www.jordansteele.com/2024/rockwell_cpu.jpg
And a nice easter egg left by the designers in the center of the chip: "Aloha"! The text is about 40 microns tall, but the lines that make up the letters are about 2 microns thick. From the research I did on this chip, apparently, the Aloha text was put there because the chip designers were promised trips to Hawaii if they met the aggressive deadline to finish the design. Also, in the full image above you can see the initials of the chip designers. (you can also see here that at 100%, all the etched features on the CPU are discernible) https://www.jordansteele.com/2024/aloha.jpg...Show more →
Nice work!
I feel there might be a Scanning Electron Microscope on your gear wish-list.
Those different items are indeed test circuits. They are basic components (resistors, capacitors, transistors, etc) of various widths and lengths that are used at wafer probe to measure/test the parametric limits of the silicon process. Basic testing of the uProcessor will have been performed at the wafer probe stage too.
If the WAT (Wafer Acceptence Test) data from those 3 drop-ins doesn't pass muster then the complete wafer will be scrapped before the individual working chips get sliced, diced and assembled into a package for final testing.
EDIT:
That red mark on the die indicates that it did not pass test at the wafer level.
It has been 'inked' such that the die will not get picked from the wafer for the costly assembly and final test stage.
Thanks for the info. I figured the dots were failed chips (the Rockwell processor posted earlier has black dots for failed chips...on the wafer I have, just 18 of the 148 chips passed, which is likely why that wafer ended up uncut and not in production. The 486 DX2 was presumably a chip that passed tests, but it had a cutting error, where the cutter didn't stay on the scribe lines properly.
I just find integrated circuits fascinating...even though I generally know the process, it's just incredible to be able to build these and have them work, with multiple layers of masking, photolithography, vapor deposition, etc, with perfect alignment, all at the nanometer scale, and have it come out the other side working. I'm an electrical engineer, and a computer nerd to boot, and while I don't do computer or semiconductor design in my job (I'm a consulting engineer who designs power and lighting systems for buildings and infrastructure), I still find this incredibly interesting.
akul wrote:
Beautiful shots with gorgeous rendering. Shots like these are the reason to have such a great lens. Clearly, you can’t get this with f4 lens..
Luka
Thanks Luka. While I usually use prime lenses most of the time, when traveling, 24-70 is more convenient and I am always
pleasantly surprise by its rendering.
- Z9 with 500pf
Z9 with 400 f4.5 plus 1.4 TC