Blackberry screen pixels

Images taken in a controlled environment or with a posed subject. All subject types.

Moderators: ChrisR, Chris S., Pau, rjlittlefield

nanometer
Posts: 317
Joined: Sat Apr 30, 2016 10:14 pm
Location: Tucson, AZ

Blackberry screen pixels

Post by nanometer »

Mitty 20x, 2um steps, zerene DMAP

2 layers of screen assembly facing each other. They had an adhesive connection which probably accounts for the bubbles in the colored image as I had to separate them. ~60um width so pixels about 20um long.

Image

Image

rjlittlefield
Site Admin
Posts: 20885
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Re: Blackberry screen pixels

Post by rjlittlefield »

Nice images!
nanometer wrote:Mitty 20x
...
~60um width so pixels about 20um long
I'm having a little trouble with these numbers. 60 µm times 20X would be only 1200 µm = 1.2 mm on camera sensor. And 20 µm pixels would be 50 pixels per mm = 1270 pixels/inch. I don't see anything in the list of Blackberry screen resolutions that is nearly so large.

Is there a conversion factor missing someplace?

--Rik

nanometer
Posts: 317
Joined: Sat Apr 30, 2016 10:14 pm
Location: Tucson, AZ

Post by nanometer »

I meant that the image was ~60um wide, and there are 3 'pixels' in the image width. I guess 20um might be the pixel spacing if they are only the small blocks which it looks like they are.

rjlittlefield
Site Admin
Posts: 20885
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Sorry, apparently I was not clear.

I took your words as meaning that the horizontal distance between white circles is about 20 um on the original subject.

Then I'm questioning whether that number is correct.

Again, if the image as shown is about 60 um wide on subject, and it was shot at 20X, then the image area as shown was only about 1.2 mm wide on sensor.

Did you really use such a small sensor, or such a small part of a larger sensor? Or are you running at a lot more than 20X optical magnification? Or have I misunderstood and/or botched my own calculations? Or something else?

--Rik

nanometer
Posts: 317
Joined: Sat Apr 30, 2016 10:14 pm
Location: Tucson, AZ

Post by nanometer »

It's a crop. Technically, it's greater than 20x--all I meant was it is 20x over the full frame of the sensor, but I cropped out a 60um width.

Like stating zooms vs. sensor size in photography, the lens is what the lens is--only the crop factor changes. Maybe that's not how we are supposed to state things here?

rjlittlefield
Site Admin
Posts: 20885
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

I hate to sound either thick or pushy, but this one's bothering me. Let me "think out loud".

Exif data inside your images says that they were shot with a Sony SLT-A99V camera using a 200 mm lens. In combination with a 20X Mitutoyo objective, the 200 mm lens would give 20X optical magnification on sensor. The web tells me that an SLT-A99 has a 35.8 x 23.8 mm sensor, with pixel pitch of 5.93 microns. At 20X optical magnification, each camera pixel would correspond to 5.93/20 = 0.2965 microns on subject. So then a field width of 60 microns on subject would be 60/0.2965 = 202 pixels on sensor. But the image that you're showing us is 1024 pixels wide, about 5 times bigger than even an actual-pixels crop would be. My conclusion is that the 60 um number is wrong.

Let me try it from a different angle. Sensor dimensions of SLT-A99 are listed as 6038 x 4025 pixels. (Crosscheck: 35.8mm/6038 = 0.005929 mm/pixel, consistent with the 5.93 um spec.) Assuming that the posted image is just an actual-pixels crop, 1024 pixels wide, then it represents an area on sensor that is 35.8*1024/6038 = 6.07 mm wide. Then 6.07mm/20X = 0.304 mm = 304 um on subject, again, about 5X larger than your 60 um number. 304 um for the entire image width turns out to be right at 90 um between white circles. That would be 282 pixels per inch, much more in line with the numbers posted at https://developer.blackberry.com/design ... sizes.html for Blackberry screens.

Can you walk me through your process that concluded the image width is 60 um?

--Rik

nanometer
Posts: 317
Joined: Sat Apr 30, 2016 10:14 pm
Location: Tucson, AZ

Post by nanometer »

Ok. I simply looked at my imaged micrometer scale wrong. It's ~500 um across. The width of the sensor is ~6000pix/1.8mm, and this image is 1700 pix wide which was re-sampled to 1024 to meet your max image size requirements. My confusion came from the scale ticks for my microscope which are 10um, and this scale has 100um gradations.

rjlittlefield
Site Admin
Posts: 20885
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Aha, the old tick mark problem! That explains everything.

So, working forward from 1700 pixels, 20X, and sensor size, I calculate 504 µm wide for the whole image, 148 µm per pixel = 172 ppi. That makes sense to me for an early Blackberry.

Thanks for resolving my confusion.

--Rik

nanometer
Posts: 317
Joined: Sat Apr 30, 2016 10:14 pm
Location: Tucson, AZ

Post by nanometer »

Sorry to waste your time on the issue! Anyway, I learned to be more careful.

Post Reply Previous topicNext topic