Hi to all pixel-nuts.
Few times people were asking me how to take good hardware photos. So I decided to bodge up quick article about this topic, instead of writing paragraphs of text to each and everyone personally.
It took a while to make everything described, but
article is out now.
Some common bad practices are busted, as well as examples of few photos and used setups provided.
Try to read IC labels on SDRAM and PCB on photo above, or what is exact voltage displayed on DMM?
Same photos after reading article and NOT using flash:
Good Keithley 2002 ADC pic as example
What gear you are using? What tricks you think useful for EE-related photography?
Show your stuff, let's improve on content we produce in 2016!
If the PCB is reasonably flat, use a flatbed scanner. Works very well for smaller stuff like for example boards out of point & shoot cameras (my guilty teardown pleasure).
Great thread !
Yeah, flash is a big no no
Also imo, most bad ones especially close up shots were made because of unsteady camera, as common using cellphone camera as it can not be secured to a steady base like a tripod. Heck, even el cheapo $10 tripod and cheap P&S camera will make huge differences compared to a handheld, high end cellphone camera.
Sharing my much simpler setup with a little trick to get crt screen like below as I don't have the special hood/shade for shooting it ->
Shooting CRT Based Scope Screen/?action=dlattach;attach=22223;image)
IMHO it's not just about how the pics are taken, how they are presented is important too.
In a thread by
c4757p how he's done the imagery is the best that I can remember:
https://www.eevblog.com/forum/repair/tek-485-repair-and-restoration-prologue/Image files have been uploaded in the normal way to EEVblog's server leaving the normal thumbnails at the foot of each post.
But instead of inserting the images throughout his post by way of the "Insert Image" tool, he's used a "Insert Hyperlink" leaving a link where an image might be. This method offers 3 distinct advantages:
1. The page loads FAST. I have slow internet and threads like this are a joy.
2. It allows an easy work-around for the forum restrictions of pic size, by using an external pic host and with a corresponding URL great pic detail can be offered when/if needed.
3. By using a URL, a click opens a new browser page yielding a FULL page pic of far higher detail than be attached within a post by using the SMF "Insert Image" tool.
Trying to get everything in focus on a larger PCB pisses me off, any tips? X(
Here's the first notch filter board on my Panasonic 7722 audio analyser, gonna have to upgrade one or two of those ancient opamps, most likely the NE5534, it likes to have a 2nd harmonic peak at the -120 or -130db range when used in active twin-t or bridged t notch filters. Swapping it for something like an OPA1611 should suppress the 2nd harmonic by 7-10db.
Not sure how much I can improve the unit though, lots more work to be done on the amplifier stage before the notch....but that's for another thread, or not, folks here don't really like talking about audiofoolery stuffs.
BravoVNice, but I'd reduce lamp backlight a notch or two.
deadlylover, is that a solder blob between R140 and R143?
I think you have decent pic, just a bit non-parallel to camera sensor, causing minor blur on left edge. I covered this in article
tautechWhile c4757p's method good for threads with few just few pics, it will become unfriendly if you have 10+ pics, as you will not see easily which photo relate to what text.
2. Storing images on external hosts, especially free pic hosts is a huge no-no in my vocabulary. Why? Because after week/month/year these photos will be gonsky! Exception - own server, which is backed up on regular manner
3. 90% of pics I post are thumbnail-ed (clickable for hi-res version).
I love to pixel hunt.
TiN,
Yep sure looks like it, I'll investigate the next time I take apart the unit. Bodge city in there from the factory, I love the triple stacked relays on the right. XD
I feel like I should get a small spinny table "lazy susan" thing so I can rotate the board easier, as I have the camera pointed horizontally and not vertically (don't have a nice tripod).
Thank you! Honestly, it is very hard to get the light conditions right. I have to take several photos, work related, for documetation, and it is impossible to get the light conditions right. Neon lamps above, shiny soldermask, and the office doesnt spend a cent on camera.
Even at home, I had zero sucess making photos.
I am considering to build two DIY light box (not the curved amazon type)
Anyone tried making photos of electronics with theese kind of lighting?
It will work fine, but I found using generic LED flashlights easier and way more portable. Also they dont take space in lab, unlike softboxes, and such.
tautech
While c4757p's method good for threads with few just few pics, it will become unfriendly if you have 10+ pics, as you will not see easily which photo relate to what text
Agreed but one can manage that by simply adding more posts.
2. Storing images on external hosts, especially free pic hosts is a huge no-no in my vocabulary. Why? Because after week/month/year these photos will be gonsky! Exception - own server, which is backed up on regular manner
+1
There's examples of this all through the forum
and for that reason I only upload to Dave's server.
3. 90% of pics I post are thumbnail-ed (clickable for hi-res version). I love to pixel hunt.
Please describe the process/syntax used.
Is this possible for pics greater than the 1000Kb EEVblog limit?
Edit
Seems like you link from your own server, well that puts you in another league to the bulk of us that are constrained by Dave's limits and the abilities of SMF forum software.
c4757p's method stands out IMO as even a quite small size (Kb) pic is opened as a full sized browser tab and I've seen nothing displaying better quality images when hosting on Dave's server.
I understand this thread is about quality imagery but show us a better way for the bulk of our members to display that imagery is just as important too.
That's a very fair point.
I think serious enthusiasts (which have lots of content/photos) may consider or already have static IP address or machine which is accessible from the web, which makes possibility of using "own" server as well.
It's not hard to setup simple webserver to serve pics and files, there are hundreds of guides over the web showing how to do that.
Alternative and more common way, is to have webhoster to serve your data. Today getting domain name, web-hosting and few gigs of space is not that expensive, and can be done just for 10-20$/year. It's way less than most of us spend for electronic toys anyway. It does not take long to set everything up
Still, if one just posts from time to time, backup pics and don't count on safety of free hostings. In worst case you could just reupload stuff again.
It will work fine, but I found using generic LED flashlights easier and way more portable. Also they dont take space in lab, unlike softboxes, and such.
They produce light which is from one source. So you will have shadows and reflections. Or are you somehow using them indirect?
The space is not an issue. I plan to use them as general lighting, and I do it at home.
See the video in linked article
. I use them to point at object from all around, with camera on tripod with 1-5 second exposure time. Of course there should be no bright lights around to have this work.
I've done some photography for internet and I will offer the following suggestions. You may not be able to control all of these, but any of them you CAN control are worth investigating. It will be up to you to develop a setup that works for your equipment and situation.
1. Lens. Is it clean? If not, you will be taking foggy photos from the outset - but before you reach for your favourite glass cleaner, please check the lens cleaning instructions for your camera. If you can't find any then ask. If you are still unsure, then I would suggest you get some good quality camera lens cleaning tissue and follow the directions. If you have a half-decent camera shop near you, they will not only be very likely to carry this, but they will be happy to show you how to use it.
2. Camera mounting. Don't hold your camera freehand. Have it fixed to something so that it does not move during the exposure. Camera shake is impossible to remove in post-production. Tripods are good (except Dave's). If you must hold your camera in hand, try to have your hands resting against something solid.
3. Lighting. This is the single most important aspect of photography. As has already been demonstrated, on camera flash is a bad idea. In fact any kind of flash can be problematic as you won't know what shadows or reflections you are going to get until after the shot is taken. If you can, have set illumination - from the side or at least off-axis from the camera lens. Diffused light from something like a softbox can work well. If you must use a flash, then do it off-camera unless the subject is a completely matt and non-reflective finish. To get an idea of what reflections a flash might give, get a bright torch and hold it in front of the flash pointing in the same direction as the flash. Move the torch until any reflections are the least problematic and then set the flash in the same position and orientation. Repeat for each flash. Do a test shot to check for flash flare and for overall lighting of the subject. Adjust if necessary. Rinse and repeat.
4. Direct viewing. Being able to see what you are about to photograph will aid setting up immensely. If you are using a smartphone, that's pretty much a given - as it is with many digital cameras - UNLESS the camera is placed in such a position where the viewfinder is not readily accessible. In such cases, having an external monitor will make your life easier.
5. Reflections. If you are taking photos which include any specular surfaces or even just shiny ones, just be mindful of what reflections you might pick up. You might not provide x-rated material (that featured in some ebay listings a while back) but you might end up with some distracting bright spots that have nothing to do with the subject. Once I had a dark subject with a shiny surface that showed some coloured patches that simply didn't exist on the subject. The patches turned out to be from a coloured sticker that was on the front of my camera body, reflected off the subject. Once I identified that, I was able to locate other bright spots which were from other shiny parts of my camera. My solution involved a couple of old black woolen socks being sacrificed for the cause.
6. Focus. If you can, it is best to have the ability to focus manually or to be able to 'lock' the focus to a set point. This way, YOU can make the judgement as to where the best focus point is.
7. Depth of field. This is getting into the nitty gritty of photography, but it is simply a term which describes the range of distances where a subject appears to be in focus. While only one distance is actually in focus, some objects that are a bit closer to the camera as well as some a bit further away from the camera can still look like they are in focus because their 'blurriness' is not enough for our eyes to notice. What is interesting about this is, the smaller the lens aperture, the greater this range (or depth-of-field) becomes. There is a trade-off, however. For the same amount of light to fall onto the sensor, as you reduce the aperture, you will have to increase the exposure time. This gives you increased susceptibility to movement of the camera. With a hand-held camera, this can become a problem very quickly, but if the camera is fixed - as in being mounted on a tripod for example - then this becomes almost a non-issue. If your subject is flat then all this isn't a big issue, but if you have a PCB with a couple of tall caps on it, then you might want to investigate this a bit further. Please note that the depth-of-field is a function of the physical aperture size of your lens. An electronic 'equivalent' will not give you any benefit.
8. Resolution. Remember that you can always reduce the resolution of an image in post production, but you can never increase it - so seriously consider the resolution you want for your original image.
9. Zoom. There are 3 ways to get a subject to fill the image space.
(a) Move the camera and/or the subject.
(b) Use Optical zoom
(c) Use Digital zoom
Moving the camera/subject is a rather obvious option, but it is one which applies to every camera. Sometimes this is either awkward or not practical, but there are times when it might be worth the effort.
With the camera and subject in set positions, using zoom is a natural choice. But, if you have both optical and digital available, which one should you use? While you can use both, optical zoom should be your first choice because it changes the size of the image as it falls onto the sensor, giving the maximum resolution for the resultant picture. If your camera has the ability, changing lenses is another option - which can provide better quality. [Fixed focal length lenses are designed to provide the best quality at a single focal length whereas (optical) zoom lenses are designed to operate over a range of focal lengths, which can mean some compromise on image quality. This compromise is not something most people would be concerned about - in fact, I would venture to say that most people wouldn't even notice, especially as today's optics are generally pretty good. (Professional photographers certainly know - and it's one of the reasons they carry a range of lenses with them.)]
With digital zoom, some of the pixels from the sensor are discarded and the remaining ones scaled up** to fill the frame, resulting in lower quality. Thus, it is pretty much safe to say that just about any optical zoom will result in a better quality image than a digital zoom.
** Digital zoom can be performed in a camera with that ability or the equivalent can be done post-production in image processing software. The result is effectively the same, but there can be differences depending on the upscaling algorithm used (there are several).
10. Distortion - barrel. When doing close-up photography, you are more prone to barrel distortion. When it comes to subjects where there are long straight lines and/or rectangular shapes, this can be most disconcerting. While some photo editing software has corrective functions, it's always better to minimise this at the time of the shoot. The best answer to this is to have your camera as far away from the subject as possible and use your optical zoom to fill the frame.
11. Distortion - trapezoidal. If your camera is not square on to the subject, you can end up with rectangles that look more like trapezoids. In photographs of buildings (for example) it is called 'perspective'. The best solution is work out the plane of your subject, find the middle point and draw an imaginary line straight out from it at right angles. Place your camera so this imaginary line goes right up the middle of your lens and with the back plane of the camera parallel to the subject. For example, if you are photographing a PCB laying on a table using a smartphone, then hold the smartphone level (ie parallel to the table) with the lens directly over the middle of the PCB.
12. Shutter release. Try to have a means where you can release the shutter without touching the camera. The action of pressing (or even touching) the release button will cause some camera movement that may end up affecting your final shot. Use a cable release or other remote means if you have it - or if you have a smartphone like my Galaxy S4, you can set it up to take the shot when you say 'smile!'.
13. Colour accuracy. If this is not critical, then skip this subject and be thankful. However, if you want absolutely correct colours presented on the internet, I have only two words for you: FORGET IT! Firstly, the average camera is not going to capture colours with Pantone accuracy. Certainly, there are better cameras and there are colour profiles which can help you with it all, but there is a limiting factor over which you have absolutely NO control: the displays upon which your images will be presented. There are several factors involved here. The first is the display technology - CRT, plasma, LCD, OLED to name a few. The next is the specifics of the technologies - such as the phosphors for CRT and plasma and TN / IPS for LCD panels. Then there are the specific settings - you know - those people who's TV's have every caucasian looking like sunburned Oompa-Loompa's. Add to that the issues of ageing displays and unknown ambient lighting conditions and you must soon resign yourself to the fact that once your image hits the internet, its fate is unknown.
There are a couple of things I have found that have allowed me to produce reasonable results without breaking the bank or my spirit. The first is to set your White Balance. This is where your camera is calibrated to understand what a white subject looks like under the current lighting conditions. The process for setting the white balance on your camera may be automatic, manual, from a selection of pre-sets or non existent. If you're not sure about setting the white balance on your camera, then you can photograph something that is white and, using your photo editing software afterwards in post-production, work out what adjustments you need to make for it to become white on your RGB scale and then make those same adjustments on each photo.
The second suggestion I can make is to do any post production image processing on an IPS monitor. I have found it to result in images that seem to present with a more consistent colour rendition across a variety of displays.
14. Image editing. Aside from the basics such as cropping and resizing, be very careful about any 'enhancement' processing. It's far, far better to get the original image captured as close to the finished product as possible than to make even the simplest of tweaks. This is most definitely one of those areas where 'less' is 'more'. It speeds up post production significantly and reduces stress.
These are just some observations I have made in my efforts. There will, no doubt, be some details that some might not agree with, but the topics are valid and my solutions have been good enough for my current needs.
Edit: Spelling.
Edit: Added section on Zoom
Edit: Added upscaling comment
Stongly agree on color aspect. It's whole pandora box. Luckily, it doesn't usually matter if our PCB's look a bit more greenish than they are.
Also one thing I forgot - taking photograps of monitors. Usually it's nasty, so people, use screenshots. Easiest way - hit Printscreen button and save resulted image from copybuffer into file. Yes, even mspaint works. Use JPEG 70-90% for fullcolor bitmaps and PNG for stuff like graphs, charts, tables, schematics.
Also for DSLR it's usually does not matter if lens is clean or not. Don't believe?
Check this then
. On contrary, sensor cleaning is highly important when using apertures f/8 and above.
Also one thing I forgot - taking photograps of monitors. Usually it's nasty, so people, use screenshots. Easiest way - hit Printscreen button and save resulted image from copybuffer into file. Yes, even mspaint works. Use JPEG 70-90% for fullcolor bitmaps and PNG for stuff like graphs, charts, tables, schematics.
Certainly screenshots are the best way to capture an image that
can be captured that way, but if you want to capture a screen that does not offer that feature then you have to consider a few things. The main ones are the ambient lighting conditions and the shutter speed. Rule number one is - don't use a flash. Next is, if you are photographing a CRO trace and it gets washed out by the ambient light, try reducing the ambient light before turning up the brightness of the trace. The shutter speed used should allow for at least one full scan of the target image. In the case of the old Australian analogue TV standard of a 625 line 50Hz interlaced signal, then you would not shoot at a shutter speed any faster than 1/25th of a second. This ensures the whole screen has been completely painted during the period the shutter was open. Translate this into whatever display you wish to photograph.
Also for DSLR it's usually does not matter if lens is clean or not. Don't believe? Check this then . On contrary, sensor cleaning is highly important when using apertures f/8 and above.
I did smile at that article - and I understand the point that it is making.
The fact that a good quality image was able to be captured in the earlier examples is because there was a significant area of glass that was devoid of contaminants and those areas that were 'contaminated' were highly obstructed. The end result is that much of the light that made it through was mostly unaffected, allowing a sharp image to be formed. All the 'contaminants' did was reduce the amount of light getting through. These earlier examples were much like looking through a flyscreen - when you get close enough, the flyscreen is just a vaguely fuzzy as you look through it to see your subject - reasonably clearly.
I should have, perhaps, been a bit more specific about the particular lens contamination I had in mind. It wasn't the discrete granular type from a single dust incident, but more the milky cloudiness built up over time or the whacking great fingerprint on the smartphone lens. These won't produce the same results as the article. Instead of looking through a flyscreen, it will be more like looking through a sheet of baking paper.
One of the most important things in my opinion is having a stable camera - a tripod is your friend!! I'm using a DSLR, and the tripod lets me use apertures smaller than f/8 which often results in exposure times of a half second or more. The tripod and a cable release allows this technique to give a nice sharp image with a great deal of depth of field. It also lets you take your time positioning lights and subject to reduce or eliminate glare (sometimes moving something only a few degrees can make or break you in this regard), and allows for repeatability when taking a sequence of photos - provided that you keep the subject in the same spot on the bench, the images will not jump around a lot with reframing by hand, and the lighting will be consistent.
Smooth, even lighting is also important; I have an articulated lamp that I can reposition as needed, and also keep a piece of foam core board handy to use as a fill reflector to help reduce shadows.
-Pat
If you are having trouble photographing reflective surfaces try using linearly polarised filters on the illumination source and camera.
I found this out while taking close up photos of laser marked PTFE/ETFE wire for a character recognition project that never happened. Without the filters there was too much glare or reflection from the insulation and it took a lot of post processing to clean up the image so it could be used for character recognition. Linearly polarised sheet in front of the illumination leds and camera fixed the problem.
Circularly polarized filters, used in some 3D cinema glasses, might work but I havn't tried it.
While post production editing is not optimum, I have had quite a bit of success eliminating reflection problems by orienting the camera off perpendicular from the subject, and then correcting the trapezoidal errors with software. Just a couple of mouse clicks. If the original image is of sufficiently high resolution the correction is not noticeable. To make this work well the camera distance has to be large with respect to the height of the tallest objects on the board, and the off axis angles have to be kept as small as possible consistent with eliminating whatever reflections are present.
Bump .... as a reminder for making a better photoshoot.
Great thread! For PCBs I use our office scanner - it's hard to beat with flat-ish PCBs.
I would add my piece of advice on publications: When you convert to PDF (or any other format) make sure you check the output. The default settings on PDF converters can mangle the images in your publication beyond recognition. The considerable effort put into making very nice pictures like those presented will be laid to waste due to the fact that the PDF converter was set to "Fit it into 100bytes or less, chop of the extra bits".
The other side of this (what's done to vector stuff) is also disturbing. See:
http://www.daqq.eu/?p=1135Here are four examples of different compression settings on a PDF generator:
Another example, especially for those who still making excuses can't make a decent photos ...
This below was taken using cheap 7 years old
point & shoot camera (yes, NOT those expensive DSLR) , with built in self timer, a crappy 3 dollars tripod and just a simple table lamp, thats it.
Oh yeah, no need for gazillion bytes sized
either, this photo was cropped, resized & compressed down to 117 KB, and still has enough details to convey the message and meaning, imo.
PS : It was made to discuss certain old and nasty rotting plastic connector that oozed out flakes.
My approach to reduce reflections: Using off-camera flash(es). This already reduces the direct reflections of surfaces which are parallel to the cam. Avoid direct flash the target: Use a white wall as bouncer to soften the flash. And optional: For beauty-shots using an additional reflector (e.g white paper) to reduce further hard shadows.
The down side: It takes a bit to set up the rig (including the right position and direction for the flash. And TiN is right even with a single lightbulb you can get good results.
But the most helpful thing is a Tripod! This way you can correct reflections reliable as the cam does not move between the shots and it prevents unsharp pictures from potentially longer exposure times.
@BravoV: Good point: Point&Shoot cams can do great macros. Remember to enable the macro mode first. They have smaller sensors. This results in bigger depth of field which helps. With a (D)SLR you might need macro-lenses or extension tubes to get close enough. And you still face shallow DoF.
Even a half-crappy cellphone (in my case a 3 years old Nexus 5) can take decent pictures.
I don't say that this image have studio/DSLR quality, but I'd say it is good enough for a forum or website posting.