The overall aim of this prototyping period was to design and manufacture multi-element lenses to provide a high image quality than my previous singlet designs. Ultimately, while I was successful in learning and designing the new lenses, flaws in the way I manufactured them have resulted in inadequate results. The lenses do not represent the design specifications, which therefore renders the entire period a failure.
At first this may seem like a damning conclusion, as there are positives to be taken from the work regardless of what went wrong. Therefore I am dedicating this post to quantifying the scope of the failure, pointing out what went wrong and what has been learnt, and also where this leaves me in terms of my research plan. Since I have shown a few people my new lenses and explained to them what happened, they have all had certain questions for me. Therefore I thought it would be easier to quantify things if I asked myself a series of questions:
How did the prototyping go overall?
Prototyping is never easy. Anyone can design something on paper, but when it comes to actually making it physically it really isn't easy. I knew the limitations of the machinery I had access to while I was designing the lenses, and I knew I was pushing the boundaries of what was possible. Over the week I encountered problems and ran into issues as you would do during any prototyping period. I had machines stall on me, parts fail, tools snap, parts snap, parts get manged by the machines, I had to make new tools, experiment with new methods, learn to program CNC machines, etc. I had to do a lot more than I was anticipating. The biggest issue was the cutting of the concave faces. I knew before I even left for England that it would go 1 of 2 ways. Either the machine would recognise the profile and cut it no problem, or it wouldn't recognise the profile and my week would be a nightmare. However in the end, having focused so much of my time on that, it was something completely different that ruined the lenses. If I needed an analogy it would be this. It's like being treated for a disease for a long time, walking out of your final therapy session, and being hit by a bus. The main issue all the way through was the cutting of the lenses, but when it came to completion the thing that ruined everything was the polishing. It was so unexpected and so late in the day that there was no easy fix and I ran out of time, left with 2 lenses which I knew didn't perform as I needed to, resulting in failure.
You managed to manufacture 2 complete lens designs, why do you still consider it to be a failure?
The lenses were designed to be an improvement over the singlet prototype which I made a few months ago. The designs which I created performed exceptionally well in ray tracing simulations, and I was expecting the same or similar results from the physical prototypes. Due to issues with time and manufacturing, the lenses that I managed to produce did not perform according to their designs, resulting in image quality that was obviously lower than the initial singlet prototype. Without a lens that performs according to it's simulations, I can't continue with my plans to test and analyse the discrete differences including advantages and pitfalls in image quality between multi-element designs and singlet designs. Therefore the purpose of the prototypes has been lost, and the period is considered a failure.
Don't worry, if you had managed to make it perfect that easily then everyone would be doing it!
Firstly I totally understand that everyone saying this is just trying to be nice, so please don't take what I am about to say the wrong way... but what on earth makes you think that what I am doing is easy?? What I am doing is bloody hard, this is 6 months of solid research and development, learning stuff that has it's own degree dedicated to it (Applied Optics). Not even people who study it at degree level get to actually work on machines and make their designs for real. I'm doing stuff which I'm not even sure has been publicly done before (I can't find any record of people making wide aperture DSLR lenses from plastics). So I resent the idea that what I am doing is 'easy', even if I am not getting it perfect. Sure I may not being doing anything new, after all you could pick up a book on optical design and read through it, learn to program a machine (providing you have access to one) and try to do it yourself. But people don't. It takes time, effort and dedication. That's just like saying, oh anyone could be a doctor, you know if they just decided to dedicate 6 years of their life to going to Med school. So yeah, rant over.
It's not like the prototypes don't work though, can you not still test them?
Yes and no. The new lenses do work in the sense that they render an image, which is an achievement in it's self as it confirms that the multi-element designs which I created do actually work. This means that the lenses are 90% correct, and given more time and machine access, I am sure the lenses could be remade to a higher standard so that they work according to their specifications. However, testing the prototypes I have made do now work well enough to give me fair test results. They cannot be used to analyse the differences in image quality between singlet and multi element designs, which was their purpose. I can however test the lenses in a different way, for an alternate motive. For example, I have friends here in Orléans who are borrowing the lenses to take sample photographs on a range of different camera bodies, helping to kick off a discussion about the qualitative features (rather than technical) of image quality. This could be useful for the second half of my research which strays away from technicalities and analyses what image quality is as a concept.
So do you consider it to have been a waste of time?
Of course not. I know for a fact multi element lenses are the way forward for my work. Every lens manufacturer in the world makes multi-element systems simply because it is the easiest way to control image aberrations. I know that when I remake these designs, they will work, and they will outshine my previous prototypes. During the work, I have encountered endless problems which had to be quickly solved. Although I obviously didn't manage to solve all of them in time, each problem solved at this point in my work will save me huge amounts of time in the future. After all, I could be discovering these things while manufacturing my final degree prototype, wouldn't that be a nightmare? It is hugely frustrating for things not to go as you expect them to go, but at least it has happened at a stage where I can afford to re-arrange my diary and not lose any time.
What positives can you draw from the prototyping period?
The biggest one for me is the knowledge of what these CNC lathes can really do. The lathe I used was a CNC training lathe from Boxford. I was really pushing the boundaries of what it could do. I ended up leaving the school with a better knowledge of how the machine worked than the people who own it. The biggest positive for me is that I can now program the lathe to do exactly what I want. I am no longer constrained to what the software will allow anymore. And combined with my own custom made tools, I can use the machine to cut profiles which a standard configuration would not be able to cut. Basically I have learnt to both alter and control the machine to cut shapes which had previously been impossible in it's default configuration. This gives me faith that I can design with freedom knowing that whatever profile my design requires can be cut using the machine and my knowledge of how to configure it.
So if you can't do the testing you were planning, what are you going to do now?
It's highly frustrating that I can't continue with the testing I had a lot planned to do with the new lenses. Until I am able to remake the lenses to a higher quality this testing will have to be postponed A lot of people wonder why I am so harsh on myself for this prototyping having not gone to plan, but there is a very valid reason why. I only get access to the machines once every few months when I am back in the UK and my old school have the kindness to let me use their machinery. Therefore, if i don't get everything I need to get done completed in the short period of time when I am back, then I effectively get put months behind. Right now I know exactly how to solve the problem of the poor surface quality. If I had the access to the machines I would be able to fix it within only a couple of days. But the reality is, I don't have access to machines and I won't have these lenses remade for a good couple of months minimum. Therefore, the whole technical side of my research has been stalled. Therefore, I am planning to do two thing.
1. I am going to use new resources I bought in the UK to continue learning about the technical side of lens design. I have a graduate level book containing an entire semesters worth of lectures focusing on Applied Optics which I am working through in order to understand every single technical aspect of basic lens design. This will fill in any gaps I have missed, help solidify what I have already learnt, and will help me build a solid understanding which will be unquestionable when I return to my home institution DMU to start my project in September.
2. I am also going to start the second half of my research. All this technical work is essential, but is all academic It can be learnt from a book, it isn't new or ground breaking. For my degree project I will have to design something new and interesting, which will leave the confounds of traditional lens design. To do this I am undertaking research into the concept of "Image Quality". I will be posting about this on blog soon.
