ReviewTheLight: Imalent SA04

I kind of wish they designed this light, still using 4ƗAA, but with a single LED. Iā€™ve heard talk that this dual color emitter would be nice for photographers, but honestly, how high if a demand is there for something like this. Maybe itā€™s just cause Iā€™m not the biggest fan of that yellow tint.

Thanks for the review. This new brand of lights have really caught my eye recently and itā€™s great to see a detailed review of the pros and cons along with the UI. Keep up the good work.

Yep, doing the spectrum is a lot of fun, and can be very insightful as well!

Yeah, Iā€™m not sure yet what the practical uses of adjustable tint will be. For me, I usually set the tint to one I like and just leave it there.

Just found your review..

FYI, your warm xml is NOT an XML2, but an older XML, just like mine.

see here: https://budgetlightforum.com/t/-/27930

Hey, it looks like youā€™re right, thanks for catching that!

Nice review.

Just an FYI mine has both XM-L2ā€™s (verified by disassembling) and a metal reflector so there seems to be quite a bit of discrepancy. Mine is also completely water resistant.

Best part, the display uses a pretty standard PIC 18Fxxxx display driver so Iā€™ll be able to eventually have all custom graphicā€™s and buttons!

Could you post a few pics? Im very curious about the metal reflector

Thatā€™s the absolute joy of this light and the whole point. :smiley: You are able, with one light, to get EXACTLY the tint you prefer with no compromise at all. This light is truly perfect for ALL users no matter their tint preference (within reason :stuck_out_tongue: ).

Very nice review, quite a niche light though. I also like the spectra. Did you use one of the public Lab spectrometers?

Thanks! Yes, I use the publiclab Web app with a modified spectrometer design. :wink:

I will be adding interior picā€™s to Jmacā€™s review.

In the mean time worth noting; the inside of the light was very dirty, the only surfaces not covered in flux residue were covered in sticky tape residue. Also they couldnā€™t seem to make up their minds between using Kapton as an insulator or plain old electrical tape, there were multiple instances of both.

That's funny! So there is more to it.. Not just the few things I've noticed during my review.

Damn. Expensive!

Thanks, and I want to know more , not to criticize anything (I think it is great to do such a thing that you do), but I am curious about these kind of measurements and I am thinking of building a spectral wavelength measurement set-up myself. Do you use the piece of cd as a diffraction grating, or something else, or even a prisma (your spectra look very smooth). And what is the unit in which the different wavelengths are measured (before correcting for the human sensitivity curve), is it a representation of the optical power, or of the nr. of photons, or plainly the values of the pixels of the camera (and what are the pixel values representing at all)? And is the V(lambda) correction tailored for how and what is measured by the spectrometer? Is the grid used transparant for all measured wavelengths? Sorry for the multitude of questions .

No trouble with questions ;). I actually use a piece of DVD for the diffraction gratingā€”they have more tightly packed lines, so the spectra I get are smother than those from CDā€™s. Iā€™m working on an upgraded design, one of the features being a different diffraction grating, something more like this. A DVD has about 1350 lines/mm, a CD has about 625 lines/mm, and that one I linked to is in between with 1000. So, the resolution wonā€™t be as good as with the DVD, but it has the advantage of being straight-lined (and still pretty cheap). Because the lines on the DVD are curved, the spectrum bands are curved, so when I calibrate a certain x-coordinate on my camera readout to associate it with a certain wavelength, it only works when taking the spectrum from the exact y-coordinate I calibrated on (does that make sense? Iā€™ll upload a picture if I need to). With a straight grating, the bands wonā€™t be curved so that will actually remove more error from my measurements than what is lost by the lower lines/mm.

When the data comes in from the webcam, it just reads the relative brightness observed by each pixel on the camera (in each of itā€™s separate red, green, and sensors). The red, green, and blue values from each pixel are added up to be the brightness for that pixel. Itā€™s not calibrated for any absolute brightness measurement. I hope to incorporate the next spectrometer design into my integrating sphere so that I can make absolute measurements possible (I left an extra port available on my sphere just for this purpose :wink: ). Because itā€™s not calibrated, there isnā€™t really any unit, just the brightness value returned by the camera for that pixel.

The intensity correction for human perception isnā€™t done by me, it was done by professionals :P. The webcam itself (and itā€™s driver) automatically take the sensor readings from the CCD and correct them for human perception (thatā€™s why the picture you see when you use a webcam looks like what you see when you use your eyes). However, it makes me uncomfortable that Iā€™m not the one who calibrated that function, so Iā€™ve been researching ways for calibrating relative brightness readings on a spectrometer and Iā€™ve got a few ideas for how to get that done, but Iā€™m waiting to follow through with that until my next design has been built. And you bring up another good point, even if the webcamā€™s driver has a perfect correction curve, itā€™s unlikely that the DVD transmits all wavelengths equally. Thatā€™s another reason I want to get a new diffraction grating and calibrate the correction curve myself.

I think I got all your questions, let me know if I missed anything or if you think of any more :wink:

Thanks for that link BG79, Iā€™ve been looking for something better than DVDā€™s since I got into my RGBW stuff. Because of the curved lines it didnā€™t work very well at all with a multi emitter light and even suffers with am XML color. I had started wondering about a prism but for so cheap I grabbed a few of these things.

Did you have any issue at a with th SA04 from the dual emitters casting light slightly out of alignment? When I did mine it was a little bit off but not like my TK45 RGB, I was able to stick a piece of diffuser film on the lens and take care of it totally.

I did end up getting a usable image for the TK45 but I had to go crazy with the setup, 2 layers of diffuser film (one on each lens, one single piece suspended in front of the lght) and bounce that off a white wall with the camera/DVD around the corner (a piece of cardboard) and then I got a very dim image but still worked, I feel this way was more percise with each individual line being much narrower than when shining the light directly at the DVD.

I take it you mean the SA04 casting light out of alignment for the spectrometer reading? No, that wasnā€™t really an issue for me. I suppose if I oriented the emitters parallel to the lines on the grating then there would have been some error due to the top emitter being stronger higher up on the band and vice versa, but I oriented them perpendicular to my grating. The narrow slit takes care of the rest, combined with setting up the light pretty far from the apparatus. And youā€™re right, having some diffuse surface to bounce the light off can soften it a lot and help you get better readings, though you have to use the same wall every time if you want to compare spectra :wink:

Thanks for all the answers BM, that explains a lot :-) . The grating slide looks useful!

The only thing that I do not understand is that you say that a webcam corrects for the human eye sensitivity, so that we can see the image correctly. If it does that, the image when we look at it is 'filtered' twice, first by the software, then by our eyes. What in my understanding the camera+display should do is reproduce reality (without correction) as well as possible so that we see as little difference as possible between image and reality.

Youā€™ve brought up a good point, but unfortunately itā€™s moving into a realm I know less about. This has to do with the display properties of your video driver I believeā€”the part of your computer responsible for deciding what brightness values to send to the RGB pixels in your monitor. I canā€™t speak with certainty here, but Iā€™ll give my best understanding, and someone who knows more about computers can correct me if necessary. I believe the video driver takes into account the properties of your specific screen to take the data from whatever youā€™re doing on your computer and display it in such a way as your eyes will see the color intended by your programs. So, the webcam first has to give the computer data about what colors the user should see, then the video driver will translate that into the light that needs to be emitted by the screen to give your eyes the correct colors.

While I donā€™t understand the whole process there, I do know that it works, because when I look at a webcam picture the colors look pretty much as they should :wink:

This certainly does strike me as the simplest way to design a webcam-displaying computer, but in this case the webcam is not the only thing our computers display, and most of the color data on our computers is in terms of what colors the user should see, so the webcam data must be presented accordingly.