Sharon kindly took the laser spot tests this morning.
I was surprised at how much of the flare was captured at the 1/8000 sec. setting.
I’ve only have a quick play, but this is an animation of the shots. Please note that we are missing the 1/2 second photo, but I don’t think its that important.
All pictures were taken with the exact same setting as @RocusHalbasch used at the beginning of the thread.
In all, it was a very useful exercise, well worth the time away from the desk.
I have just followed @Ante_Vukorepa work flow in photoshop to create the log scale image of my laser flare - thank you very much for the instructions - I wouldn’t have known where to start
I honestly have to say that that looks bettter than I expected… I would have thought the red part would have been much bigger.
Any thoughts on what colors are harmful? I’d suggest that green/lightblue is acceptable, but of course you’d be better off without them. Red and yellow are probably the main concerns.
I have no idea, but if you say just the brightest part of the laser (red areas) are what cures the resin then my flare from the spot center is about 1mm towards the back & 0.8mm towards the front.
If you include the orange areas then we are looking at 2,5mm rearward & 1.4mm forward.
In a perfect world I have no idea what the optimum diameter of the spot should be.
In the FormLabs specifications they say the the minimum features size is 0.3mm so I would guess that the optimum spot is this or smaller.
What we ideally need is for someone with a good laser to do this test. At least that way we will be able to bench mark.
I am not knowledgeable on DSLR so I have the following question. I know CCD’s are not all calibrated at uniform intensity. Back in the old days we used to calibrate them against a standard gray card. How do you all know when you are each posting using different cameras that the measured intensity is the same? Since the size of the spot is dependent exposure time/energy I do not understand how this becomes truly quantitative.
Variations between cameras are no more than 1 EV (1 stop), so it’s not that significant (compared to total dynamic range of the laser spot).
Also, the absolute value is irrelevant - it’s the comparison between the center of the beam and the surrounding flare, from case to case, that’s the desired bit of information.
I.e. “how strong is the flare compared to the hotspot in the middle in case 1 vs. case 2”, not “how strong is the flare in case 1 vs. the flare in case 2”.
As I mentioned earlier this colorization suggests a far more linear change in intensity than is really the case, instead of each progressive shot getting more sensitive by the same amount each shot is twice as sensitive as the last which is an exponential change not a linear one. The actual bad part of your flare is probably smaller than the red by a little bit.
There’s also the issue of human perception of colours.
Our eyes (and brains) don’t have the same dynamic range in all parts of the spectrum. Not to mention the monitor colorspace influencing things as well.
These are an illustration more than a serious tool - but even so, they should be useful for doing comparisons. Ideally, though, some kind of isoline (isopleth) visualization would probably be ideal.
There are certainly very intelligent and knowledgeable people on this forum! I just wanted to offer a couple of insight to the ongoing work.
First, the wide variety of random flares that we’re seeing here just says to me that there’s optical alignment issues with the laser. Optic alignment can be done quite easily. But I think what’s really needed here is just better assembly of all the parts and smarter designs for mass production.
It is possible that different cameras have different spectral responses to the 405 nm laser. However, since only the 405 nm wavelength is in question, the more important aspect for a fair evaluation would be to choose smart exposure times. Wrong choices for exposure times can be error-prone. For instance, having too much exposure such that the camera can’t tell the difference between the brightest spot (at the center) and the second brightest spot (next to the center) would lead to misreporting beam intensity profile. It is important to capture just enough light so that the brightest spot is near but not at the maximum intensity that the camera can detect. At the same time, you want the exposure to be high enough to fully detect the edge of the beams. With that said, it may be better to capture two different images, one with high and one with low exposures as previously mentioned, given the limitations of most commercial cameras.
Also, photopolymerization can occur throughout the entire volume of irradiation. That means for a beam with almost a 5+ mm beam size (based on Steve_Johnstone’s picture), a single exposure would print a 5+ mm voxel. However, in reality, despite having such a broad beam (which is quite surprising to me), the dissolved oxygen, which acts as a inhibitor to the polymerization reaction, would set something of a “threshold” for irradiation. For example, the dissolved oxygen may say that x amount of photon energy is required before the solution solidifies, and that amount of photon energy may be delivered to the tight 0.3 mm spot in the center in the programmed exposure time.
What does this all mean?
If the laser is stronger, then more energy is delivered per time, and less exposure time is needed to deliver the amount of energy for printing.
If the scanning is faster, then the laser stays on the same spot for a shorter time (i.e. shorter exposure time), and less energy is delivered in that time.
If there is no scanning, energy will be continuously delivered and eventually the entire volume exposed to the beam (for the log pictures, that would be the lightblue areas, i.e. 5+ mm) would have solidified.
We actually did a large range of 14 shots at single stop increments, starting from one wher only the brightest spot was visible. Hence the title dissection. I have yet to look at the full set of @Steve_Johnstone’s shots but I’m guessing his lest sensitive exposure shows a significantly smaller dot than his red section. His red section is probably the area covered by several stops so may represent a range of up to something like 1/16 of the brightest all the way up to the brightest. His post has a download for the whole series.
So if this thread is active, why will it still close in 3 days!!! That logic is bonkers… we will end up with countless repeat threads just to continue conversations!
I still plan on doing this with my laser, just haven’t had a second to spare yet.
If you scroll at the very top, you’ll see that multiple exposures are being taken, in order to cover the entire dynamic range of the spot. The shortest exposures are not blown out anywhere, i.e. they are preserving the total range just fine.
The false color images posted later in the thread are composites assembled from multiple exposures.
So looking at your images the bright part of the point in your 1/8000 image is smaller than 1mm in diameter. Your images don’t really show all that much of a bright flare till about the 1/1000 image which is 8 times as sensitive to light so the flar in that one is less than 1/8 as bright as your center dot. Around 1/125 the flare has almost the same profile as your red area but that is 64 times as sensitive to light as the 1/8000 so even the really bright areas in that one may be as little as 1/64 as bright as the center dot. Thus the red area might be a little to broad to assume it is the problem area.
@Sam_Jacoby I know you can kill a thread. Can you keep one from dying? This thread is still alive but your new autokill feature is about to kill it. Please give us a hand.
@RocusHalbasch, took way too long to do this, but here it finally is… Not sure how to do the color gradient, I haven’t used photoshop in quite a while so I’m sure the intensities aren’t quite right.
Hmmmm site cut me off, I have 3 more pics - 1/4, 1/2, 1 sec - but I must have hit the file size cap I suspect… I am posting full sized images so I’m not surprised. In any case the pictures I did post captures my spot well enough I think.
