CHDK Forum
Using CHDK => Creative Uses of CHDK => Topic started by: matt314159 on 08 / May / 2012, 15:46:43
-
*ETA* if this is not the best forum, would the mods please stick it where it's most appropriate? I apologies if I picked the wrong place.
Hi Folks, I've been doing some reading and I have found several time lapse threads, just perhaps none quite so ambitions as this. I'd like to pick your brain and see if I can get some help 1) determining that this will, in fact, work with my goals in mind and 2) optimal choice of camera for this project.
I work for a small private college in the IT department, and we are looking to make a time-lapse video of a new $14M learning commons that is being built. Construction started about a month ago (they haven't even poured the footings yet) and it is anticipated to run until next august (2013). The powers that be have allocated absolutely ZERO funds for this project, so we are working with just whatever money we can spare out of our own department budget, hopefully just a couple-hundred bucks (max).
I would like to do maybe one photo every 5 minutes or so. We plan to render incremental videos every month or so, maybe a little more frequently, so mini-time lapses with what we have "so far" as we go along. I think over the 15 months of the project, a 5-min interval would give me plenty to work with. I understand an EyeFi SD cards can offload content on the fly and therefore we would not have to change cards out. This is good, and I would hope to incoproate this feature into our set up.
We are in Northwest Iowa. Summers can be 100degrees with 80% humidity, and, conversely, the winters can be -20F with crazy windchills and blizzards.
We have one of these (http://i.imgur.com/zpV5S.jpg) in our boneyard downstairs. It was used for another construction project about ten years ago. The seals seem decent still, and I think it could work decently as an enclosure. It has a sealed power cord. The interior dimensions are roughly 4.5" wide and 3.25" tall. So might have to look at maybe an elph style camera to run CHDK? But would that survive 15 months of brutalization?
I would genuinely appreciate any and all feedback you might have for me in this regard. I really wanted PR to allocate $1k or so to put together a full DSLR+Intervalometer+ weatherization kit type setup, but it was a complete no-go. Frankly it's ridiculous considering the massive cost of this new building going up by comparison. But such is the life of working at a non-profit.
Thanks in advance for any comments you may have for me!
Matt
-
First thing - CHDK is not for DSLR - for them there is a Magic Lantern. CHDK is for P&S cameras.
Second - humidity and outdoor timelapses - look here: http://chdk.setepontos.com/index.php?topic=7743.0 (http://chdk.setepontos.com/index.php?topic=7743.0) - cheap and easy ;)
Third - EyeFi cards allows you to download photos from camera and delete the downloaded ones, but I'm not sure, how camera behaves when the photos are removed - it might not update the info about that photos. Just not sure, but I don't think there should be problems
Fourth - photo each 5 minutes - to be done with almost any intervalometer. Just my notes:
- use kind of auto mode in camera due to changing light conditions during the day or calculate the apropriate Tv and/or Av values for each photo (for example like in my intervalometer: http://chdk.setepontos.com/index.php?topic=7757.0 (http://chdk.setepontos.com/index.php?topic=7757.0) )
- in 'Debug parameters' set 'Restart Lua on error' to ON. Sometimes something goes wrong and script can stop yielding an error. With this option it will be restarted. I have this option always on - I have lost some nice sunrises due to errors (Note - CHDK is the program that runs on completely undocumented systems using the hacks and it's well known that it can cause some interferences with the camera firmware - then some things might fail - script can stop, CHDK can lose the control or even the Canon safety code can turn off the camera - we call this a 'crash' - that's why you HAVE TO check whether the camera still works from time to time).
15 months x 30 days x 24 hours x 12 photos each hour = 129 600 photos. I'm not sure whether the camera is able to survive such continous work. Even without photos at night (not interesting for this purpouse) it is still around 50 000 photos. This might be a problem. However if you'd have a new camera it can be under a varranty - and varranty doesn't say you cant get 50 000 photos in one year ;)
-
Thanks, outslider. I am thinking a small P&S for this project, to fit inside the weatherproof housing we already have. Heck, some of these models of camera can be had for like $40 each these days; All I really need is 1080p which any 4-5MP camera would be able to give me. Ostensibly I could buy two...configure them identically, and if one goes out, swap the backup one in without fuss. And still get off pretty cheaply. I will look into the links you provided. I would probably like to avoid shooting at night, waste of shutter actuations, etc. I was thinking I'd probably just set aperture priority and let it figure out the shutter speed; but then flicker could become an issue; In previous daytime time lapses I've set full manual exposure, but I don't think this would work, even if I was only shooting during the day, because cloudy days would then be severely underexposed, etcetra. Unless anyone has a solid argument for manual mode, I'd probably just go aperture-priority. The enclosure has AC power so power's not a concern.
Ordinarily I'd shoot in raw, edit in LR4 and then export a time lapse, but I think like 3.2MP JPG fine mode or equivalent would probably be fine for this project.
Sorry I'm all over the place, never done anything quite so long before...lots of different things bouncing around in my mind. The main thing being, can the camera handle the torture for an acceptible length of time. I don't care if I have to go through 2-3 cameras over the course of this project, I don't think the final product would be hurt much if I had a day or so go missing from the shoot. I would go for more time in between intervals, except we might be doing updates once or twice a month. shooting 10hrs a day (say 7AM to 5PM) @ 5 min intervals is about 5 seconds a day. In a "final" render, I can probably just pull every X frame as needed to shorten the whole thing up, as the video would be like 40 minutes long if I left all the frames in.
-
Unless anyone has a solid argument for manual mode, I'd probably just go aperture-priority.
Many of the smaller / cheaper Canon's don't actually have a variable aperature so you actually don't have a lot of choice there.
Ordinarily I'd shoot in raw, edit in LR4 and then export a time lapse, but I think like 3.2MP JPG fine mode or equivalent would probably be fine for this project.
If the goal is a time lapse video, then there doesn't seem to be much point in shooting RAW and dealing with all the potential issue with image correction. Don't forget that CHDK RAW gets you exactly what the sensor sees - which can be pretty rough as the Canon JPEG conversion takes much of that out automatically.
The main thing being, can the camera handle the torture for an acceptible length of time. I don't care if I have to go through 2-3 cameras over the course of this project, I don't think the final product would be hurt much if I had a day or so go missing from the shoot. I would go for more time in between intervals, except we might be doing updates once or twice a month. shooting 10hrs a day (say 7AM to 5PM) @ 5 min intervals is about 5 seconds a day. In a "final" render, I can probably just pull every X frame as needed to shorten the whole thing up, as the video would be like 40 minutes long if I left all the frames in.
That looks like a pretty nice CCTV enclosure. I'd worry more about the -20 deg F weather than the +100 - is there anyway you can install a small heater ?
-
Unless anyone has a solid argument for manual mode, I'd probably just go aperture-priority.
Many of the smaller / cheaper Canon's don't actually have a variable aperature so you actually don't have a lot of choice there.
Ordinarily I'd shoot in raw, edit in LR4 and then export a time lapse, but I think like 3.2MP JPG fine mode or equivalent would probably be fine for this project.
If the goal is a time lapse video, then there doesn't seem to be much point in shooting RAW and dealing with all the potential issue with image correction. Don't forget that CHDK RAW gets you exactly what the sensor sees - which can be pretty rough as the Canon JPEG conversion takes much of that out automatically.
The main thing being, can the camera handle the torture for an acceptible length of time. I don't care if I have to go through 2-3 cameras over the course of this project, I don't think the final product would be hurt much if I had a day or so go missing from the shoot. I would go for more time in between intervals, except we might be doing updates once or twice a month. shooting 10hrs a day (say 7AM to 5PM) @ 5 min intervals is about 5 seconds a day. In a "final" render, I can probably just pull every X frame as needed to shorten the whole thing up, as the video would be like 40 minutes long if I left all the frames in.
That looks like a pretty nice CCTV enclosure. I'd worry more about the -20 deg F weather than the +100 - is there anyway you can install a small heater ?
Thanks for the feedback. My usual DSLR workflow for a timelapse is to shoot manual, raw, then in LR I can make the adjustments and send them across the whole set. Need to get out of the dslr mindset ;-)
I will look at what there are for heaters. I know for Dish satellite subscribers you can get a little peel-n-stick heater thing that warms the dish enough that ice melts off it. maybe something like that? I dunno.
-
I will look at what there are for heaters. I know for Dish satellite subscribers you can get a little peel-n-stick heater thing that warms the dish enough that ice melts off it. maybe something like that? I dunno.
I suppose the worst part of it being cold is actually the effect on battery life. Which is not your worry. Yuo might be okay as long as the lens doesn't feel the need to focus...
-
Follow-up question:
Is there any one specific model that stands out as being a good candidate to accomplish this? can't be more than 4.5" wide, including any power connections that may come off the side, so needs to be a smaller unit. But if there are any that run CHDK better than others, or that tend to be more reliable than others, or the best "bang for your buck" etc, I'm all ears. Again, just needs to be maybe 3.2MP max, as anything else is overkill because it will get downsized to 1080p max anyhow.
I will have power in that box, which is a good thing, I think I mentioned that before as well but just in case I didn't.
-
Is there any one specific model that stands out as being a good candidate to accomplish this? ca
I'm not sure anyone will have good answer to that question beyond "whatever you can get cheaply".
-
Fair enough, I will look around and see if anything catches my eye. Since I'm buying it to fit in a pre-existing enclosure, I think that will be a factor as well. i.e I will need the tripod mount to be right in the middle, max width of the camera body, etc, etc.
*eta* ...is it a safe assumption that ANY camera running CHDK will have what it takes to run the various timelapse scripts? Are certain features enabled/disabled depending on hardware limitations? Any guidelines at all? I'm all ears for even the most general guidance. There are so many compatible cameras out there, and many under $50, that it's kind of overwhelming...a good problem to have, don't get me wrong!
-
is it a safe assumption that ANY camera running CHDK will have what it takes to run the various timelapse scripts?
Yes
Are certain features enabled/disabled depending on hardware limitations?
Nothing worth worrying about.
Any guidelines at all?
You will need an A/C adapter of some sort. Most models have a Canon version for $100 and a version shipped out of China for $8.
I'm all ears for even the most general guidance. There are so many compatible cameras out there, and many under $50, that it's kind of overwhelming...a good problem to have, don't get me wrong!
Where are you getting sub-$50 Canon P&S cameras ? eBay ? The only caution is that not all cameras have a CHDK version and sometimes a camera might be supported but a particular firmware version might not.
-
Some recent cameras require you to open a rubber flap in the camera base and insert a dummy battery connected to an AC adapter.
It is more convenient to use cameras that have a DC socket.
For them, you can often use a standard multi-voltage AC adapter as long as the current it can source is about 1A or more.
Such as Ebay # 190674176721
(interesting, has a USB port, handy for charging my EVO 3D stereo camera/phone battery).
David
-
Thanks folks for the feedback. Waterwingz, ebay #160795334159 would be an example of a sub-$50 unit. okay, after shipping it would be like $53...but for our project, heck, 7.1MP should be fine.
Here's what I'm thinking so far.
CCTV Heated/Cooled outdoor enclosure (http://www.bhphotovideo.com/c/product/594461-REG/Lorex_ACC1721HB_ACC1721HB_Weatherproof_Camera_Enclosure.html): $45 shipped
Silica Gel to help keep the enclosure dry inside (http://www.bhphotovideo.com/c/product/242571-REG/Pelican_1500_500_000_Silica_Gel.html): $9
Eye-Fi Mobile 8GB SDHC card (http://www.ebay.com/itm/Eye-Fi-8GB-Wi-Fi-Mobile-X2-SDHC-Class-6-Wireless-Memory-Card-EYE-FI-8MD-FS-BIN-/270972265948?pt=Digital_Camera_Memory_Cards&hash=item3f1734a5dc#ht_3909wt_1198): $60
AC Adapter: $10 (I'm leery of the chinese knockoffs...I've tried the cheap laptop power supplies and they die after a few months...might have to buy a couple and keep a backup on hand.
As far as what kind of camera, I was thinking an A series powershot but newer--DigicIII model probably, they explicitly state they work with SDHC cards. That can still be had for around $60 or so, used, on eBay.
In the CCTV enclosure the mounting post inside is the size of a tripod screw but it's permanently affixed to the center of the enclosure--you can move it forward and backward only. The Canon's mostly have their tripod mount off to one side on the bottom, so I may just forego using the internal mount and get one of those peel-n-stick dash-mount camera holders, I could shear off half a side and epoxy it right where I want it.
With this setup, once working, I should be able to just walk out there with my phone or an iPad or laptop and suck all the photos down off of it every now and then. an 8GB card, saving 3.2MP photos (or whatever a 'medium' type resolution is, that's still larger than 1920x1280) should be able to hold 3-4 weeks of photos, at any rate. We should be able to just stick it up on a pole.
I was thinking of using this (http://chdk.wikia.com/wiki/UBASIC/Scripts:_Ultra_Intervalometer) script, but if there's a better one, I'm all ears. If there's one that's designed to minimize flicker somehow, I'd look into that, but above all, I want reliability. I don't want to have to take it down and reset/restart the script every other week, I'd like it to just snap away in perpetuity if possible. I understand that there may be some bugs to work out in the beginning, but I'd like it to last for awhile at least.
I could still see this setup killing cameras just because of the weather, but if I could get 6 months each of of them or so, I'd be happy with that. The enclosure has temperature control, it has a blower to keep air moving inside of it, and a heater that kicks on at like 55 degrees F, so I think that could help prolong the life of things as well.
What do you think, worth a try? Anything else to add?
-
I was thinking of using this script (Ultra Intervalometer)...I want reliability...I don't want to have to take it down and reset/restart the script every other week, I'd like it to just snap away in perpetuity if possible...
15 months x 30 days x 24 hours x 12 photos each hour = 129 600 photos
Since that many photos will be hard on most cameras, I'd recommend using the Selective Intervalometer script (http://chdk.setepontos.com/index.php?topic=6105.0 (http://chdk.setepontos.com/index.php?topic=6105.0)). It will allow you to reduce needless shots. Since there are several mods available in the posting, recommend reading all of the postings to see which would suite your situation the best. How would the Selective Intervalometer script help in this case? Here are a couple of examples:
- If a typical construction day is 10 hours instead of 24 hours, you could reduce the number of shots by 58% (15 months x 30 days x 10 hours x 12 photos each hour = 54,000)
- And if you knew that there would be no construction on the weekends as well, you could reduce the number of days to about 22 and reduce the number of shots by almost 70% (15 months x 22 days x 10 hours x 12 photos each hour = 39,600)
Good luck on your project!
-
Thanks Skywalker, I appreciate that. I was thinking it would be nice to have it not shoot at night (7am-5pm or something like that only) but wasn't aware there was a script for that. So long as that script is reliable, I think I'd like to plan on using that...needless to say, I have some reading cut out for me ;-)
I'm actually not that worried about 129,000 actuations, it seems P&S cameras are more tolerable to stuff like that than a DSLR (in fact, in the ultra intervalometer section of the wiki, they mentioned testing A series PowerShots for 4 months straight @ 2 shots per minute, over 300K actuations, and the camera was still functional afterwords), I have a feeling what will kill the cameras is being in the outdoors 24/7...in spite of being in an enclosure. I may be wrong about this, who knows. I think a weak link will be the chinese power supply as well, I can see that failing early. I was thinking I'd prefer a camera with a 1/1.7" sensor as opposed to the smaller 1/2.5" sensors on the A series cameras; To me there is a noticeable improvement in sharpness, but considering 1) I'll be shooting through the glass window on the enclosure and 2) I'll be sizing the final product down to 1280x1920, I don't think it's worth the extra cost, especially when factoring in the fact that i'm almost using these as disposables. I think I'll get two identical units and configure them the same, and have one waiting in the wings as a backup for if/when the first one fails.
-
Thought about using a UPS but then that's just one more thing I have to try to keep weatherized :'(. This is going to be out in the middle of the campus quad up on a light pole (all the poles have AC outlets, thankfully) so right out in the thick of the weather.
Something like this might fit in the back half our your CCTV enclosure though : http://accessories.us.dell.com/sna/productdetail.aspx?c=us&l=en&s=soho&cs=ussoho1&sku=A1716823&SMCID=702&CAWELAID=286007877&dgc=SS&cid=27722&lid=628335 (http://accessories.us.dell.com/sna/productdetail.aspx?c=us&l=en&s=soho&cs=ussoho1&sku=A1716823&SMCID=702&CAWELAID=286007877&dgc=SS&cid=27722&lid=628335)
The CCTV enclosure we're buying has about the interior capacity to hold a 12" subway sandwich. 4.5" x 3.5 x15 or some such. Also included in that is a small heater and exhaust fan at the back.
This (http://www.bhphotovideo.com/c/product/594461-REG/Lorex_ACC1721HB_ACC1721HB_Weatherproof_Camera_Enclosure.html) is the enclosure we're planning to order. We have one similar to it already but that one is 10 years old; this is new (so the seals will be new) and includes the heater/blower.
-
@matt314159
I've updated the Selective Intervalometer script. The latest modificaton will make it easier for you and others to use the script (Days of Week selection). I edited this posting http://chdk.setepontos.com/index.php?topic=6105.msg85318#msg85318 (http://chdk.setepontos.com/index.php?topic=6105.msg85318#msg85318)
Here is the first of two postings with the newest version of Selective Intervalometer: http://chdk.setepontos.com/index.php?topic=6105.msg85376#msg85376 (http://chdk.setepontos.com/index.php?topic=6105.msg85376#msg85376)
If you have the CHDK autorun option selected, the DOW parameter will enable anyone to power up the camera and automatically start running. This should help in the case of a power outage.
-
I did see that, Skywalker, thank you! Sorry--I have been out of town last week.
We demoed the A570IS running your script (the updated version) today as a matter of fact, with the demolition of a building: Not a valid vimeo URL
Both youtube and vimeo added a nasty moire on the brick building when they transcoded it, the original, rendered out of lightroom 4, doesn't have that.
At any rate, some things I learned from today:
1) your script works, is easy to use, and runs GREAT! for today's shoot, I set it up the night before, left it in place, and just told it to start snapping at 6:30 AM. When I finally got to work at 8, I checked on it, and it was snapping away happily.
2) the eye-fi card doesn't seem to download the photos while the script is running. not sure if I can get around this to make a truly 'endless' card or not.
This was shot indoors through a window. It worked so well, actually, that we changed the design to bring the whole camera system INDOORS! There is an out-of-the-way window with a clear view of the construction site right above where we were going to put the external enclosure. Bringing it indoors lets lots of amazing things happen. 1) I can use a UPS. 2) I don't even NEED the eye-fi card; I can just run up there and swap SD cards every month or so. 3) I can probably go with a bit nicer camera since I don't feel so much like I'm going to ultimately kill it with the weather.
The demo video has glare because I simply shot with a tripod and didn't do anything to attenuate glare. I plan to use a suction-cup window mount, blacken the camera with gaffe tape, and drape a black shade behind the camera to get rid of most of the glare we saw in this video.
Without an external awning/shade of some sort, I expect glare to be more of a concern than it was in the enclosure, which had a 3" bill on the top and sides to shade the lens from glare, but it should still be manageable.
I think I'd like to find a camera with at least a 27-28mm (35mm equiv) field of view--the A270IS is only 35mm, and it seems all the A series powershots are that way. When some funds become available next week, I might do like a Powershot S90. Camera is black already, compact, sharp, wide angle lens, uses the 1/1.7" sensor, etc, etc. I think it will work as a great replacement to a true DSLR setup, with nearly the same image quality when all is said and done.
----------
LASTLY
----------
What would be the best exposure settings to help minimize flicker in the final video? I don't want it to vary too much if possible between slightly differently lighted scenes (if some clouds roll by, etc) and if it auto-meters each shot, flickr could become an issue. On the other hand, I can't really do full manual exposure either, since it would change on a bright sunny day versus a day we're having a massive storm. I'm all ears as far as feedback on what some good settings are for exposure. until then I'll just use program mode and let the camera meter things every shot.
I do plan on doing manual focus so it's not having to refocus every time it takes a shot (and if there happens to get a spot on the window, I don't want it to AF to that by mistake) but anything else I'd love some feedback from the gallery. I don't think the test video came out too bad using standard settings (even autofocus) but I want to optimise this as best I can.
-
Hi Matt.
Aliasing artifacts often happen when resizing video with a low quality scaling algorithm. If you output your video from your editing package at 1280x720 or 1920x1080 your video will be crisper then if Youtube or Vimeo does the scaling for you.
...Both youtube and vimeo added a nasty moire on the brick building when they transcoded it, the original, rendered out of lightroom 4, doesn't have that.
For the final output of your timelapse you will probably want to use a deflicker plugin in your video editing / compositing package as well.
What would be the best exposure settings to help minimize flicker in the final video? I don't want it to vary too much if possible between slightly differently lighted scenes (if some clouds roll by, etc) and if it auto-meters each shot, flickr could become an issue. On the other hand, I can't really do full manual exposure either, since it would change on a bright sunny day versus a day we're having a massive storm. I'm all ears as far as feedback on what some good settings are for exposure. until then I'll just use program mode and let the camera meter things every shot.
-
Aliasing artifacts often happen when resizing video with a low quality scaling algorithm. If you output your video from your editing package at 1280x720 or 1920x1080 your video will be crisper then if Youtube or Vimeo does the scaling for you.
Yeah, it's funny because the MP4 Lightroom 4 rendered out for me was 1280x720 (720p), so it seems youtube and vimeo both played with it somehow. Same file uploaded to facebook plays back flawlessly.
-
Well the S90 came in, and I've got it installed in the window with it happily shooting during the day on weekdays and uploading files directly to a nearby PC via direct-mode wifi. The card is set to endless, so when it reaches 90% capacity, it will delete all downloaded photos automatically. The PC is on our wired network, so it's as simple as navigating to its shared folder and pulling the files to edit in LR with the help of LRTimelapse to render out incremental videos. Theoretically, so long as it never loses power (A UPS will be installed soon) it should just be completely hands-off for the duration of the project.
TO help minimize glare, I actually blacked out the whole window pane with gaffer's tape (white on the sticky side and black on the topside). The entire window is taped off except for a 2" hole where the camera lens is looking out of.
I think I will need to fashion some sort of lens hood for it though because once the ball of sun starts to directly get into the frame, you a flare that moves across the frame. But I don't think there's a really good solution for that, so might just deal with it.
Here's a partial day's worth of test-shooting:
http://youtu.be/YNr-37HhxsI (http://youtu.be/YNr-37HhxsI)
That's from 7AM to about 3PM when I went in and installed the eye-fi card. I just had it shooting with a regular SD card for most of the day.
This is a 3 second video @ 5min intervals.
With 5 min intervals, I will end up with 4 seconds of video per day, 20s per week. 20 minutes by the end of the 60 week project.
I think we will render out incremental videos every 2 weeks, which would make each video 40 seconds long. When we go to render the final build start to finish, I think I'll be shooting for 5-7 minutes max, otherwise I think it would get boring, so at that point I'll jstu select every 'nth' frame and render it that wya.
Playing with LRTimelapse, it's crazy good, I love it. You select keyframes throughout the photo set to edit in lightroom for each different scene, then when you import that back into LRtimelapse, it will apply gradual changes to transition from each set of edits to the next. Then you can apply a deflicker algorithm that smooths out changes in luminosity. Very powerful (and free) program. As we get further into this project, I'll probably shoot the author some money Amazing software for free.
-
Nice !
-
I'm hoping that this will hold out for the duration of my project. (96 photos a day, 5 days a week, 60 weeks = 28,800 snaps). The ultra-long duration of this project is why I want everything on the camera powered down as much as possible so it minimizes not only wear and tear on individual components (LCD for example) but also minimizes heat and things like that.
You are putting a lot of confidence - in a suction cup. Changing weather can cause condensation on windows - frost and or ice during cold periods and sweat condensation due to cooler air inside and warm air outside during warmer months.
My thoughts on this - LOL!! - if it were me I think I'd opt for a tripod - anchored (the legs might be set into 1 gal paint buckets filled with sand AFTER the leg is in) so that it can be adjusted and set. Also the camera could be fitted with a quick release plate - and using a file - that plate could be "keyed" for exact repositioning each time. This would allow a very minor interruption - say every 1st of a month - to dump the photos for processing and allow a quick return to service. Also - should the camera have any type of problem - another camera could be easily substituted in EXACTLY the same position without too much fuss. But that's just how I am. 60 weeks sounds interesting - long term project of over a years "work" for 16 minutes of video (30 fps). I am impressed!! LOL!! And I mean that in a sincere way!!
Hillbille
You are putting a lot of confidence - in a suction cup.
My thoughts exactly !
It's a fair criticism, but I had to make a judgment call and pick what I felt to be the lesser of two evils.
The location of the camera isn't tenable for a tripod, and for a couple of reasons. 1) it's in the back of a large chapel with clerestory windows. These windows have electric shades that draw down when they need to use the projector and such. This is in the furthest clerestory window toward the end of the chapel, on a ledge. The shade CAN be set to draw down and unhooked to make it STAY down for the duration of the project. They did NOT, however, want it to remain up the entire time. It's also a ledge that can be easily accessible by students if they want to screw with things. Even given this factor, I'd say that a suction cup has less chance of getting moved than a tripod would, just because of student shenanigans. Additionally, if it were on a tripod, some nasty reflections from the other windows in the auditorium can be seen throughout the day. The only solution for this setup was to get the camera as close to the glass as possible, block out all the light from that pane but a 2" hole (I used gaffer's tape to black out the entire window which not only helps with potential reflections, but also heat...the tape is white on the sticky side and black on the backside), put a hood over the camera, draw down the shade BEHIND the camera, hiding it better from the students, and cross my fingers, say a prayer, and hope for the best.
Is it ideal? nope, not at all. But considering there are virtually no quality expectations from anybody, even if I have to re-mount the camera I'm not going to consider the project "ruined" ...That being said, however, you seriously haven't seen this suction cup! It's damn hefty, with a lever that holds it really freaking tight. I picked the best one I could find, and I have a fair amount of confidence it will work okay. And, like I said, if not, I had to make the best decision I could given the limitations of our budget and location. To properly mount an external weatherproof camera setup was going to shoot the cost up and present further problems with using CHDK...like, if power were lost, I'd have to get on a scissor lift to get things running again. again, I come back to the word "untenable"
When you work for a nonprofit who is penny-wise and pound foolish, there's nothing you can do but try your best and hope for the same.
FWIW, here's a quick-n-dirty preview I threw together with shots taken between wednesday mid-morning to friday at 5:30PM. During the three days this was shot I did a lot of changes, dealing with attenuating some reflections, re-framing the shot (I consulted with somebody who was planning it and they told me I was too far to the left--it was going to have extra space on the left, while cutting off part of the builidng on the right), etc, etc. This scrap of footage probably won't go into the final rendering, but hopefully now, after three days of testing, its' ready to go on its long run starting monday. Crossing my fingers.
Oh, yeah, the link: Week1Test (http://www.youtube.com/watch?v=eVLroBLqodk#ws)
Here are some photos of where this camera is located.
Inside view, from the stage looking toward the back of the auditorium (camera location marked in red)
(https://chdk.setepontos.com/proxy.php?request=http%3A%2F%2Fi.imgur.com%2FTAWRN.jpg&hash=a2faa432d50ee6f7faaacf1ccf68d32e)
Exterior view of the side of the chapel, camera location marked:
(https://chdk.setepontos.com/proxy.php?request=http%3A%2F%2Fi.imgur.com%2FRHNHA.jpg&hash=21344c49cd79efd4648cc193a3eaeaea)
So, a too-long answer to a simple, and fair, criticism, lol. I'm flying on a wing and a prayer here, but given limitations that are prohibiting me from "doing it right" I am having to do it in as "least wrong" a manner as possible, if that makes sense.
-
So I'd suggest cleaning the window really well where you intend to mount the suction cup with IPA. Add some fiducial marks with tape so that you can remove and relocate the suction cup precisely. Then maybe "reload" the suction cup every month or two ?
-
I cleaned it really well with our denatured alcohol that we use for cleaning electronics in the department, so that part is done. I love the idea of marking it though, that is a really good idea. Will do that on monday. I'm not sure how comfortable I am "resetting" the cup periodically just for the heck of it, but I think I will check now and then to make sure it still feels as tight as day one, then re-[admin: avoid swearing please] the lever if need be. I would almost think that any condensate, if the cup is creating a good vacuum, would actually form on the outside of the suction cup instead of underneath, but we'll see.
fun stuff, yikes! lol.
*eta*....I love automatic swear filters. Re-C.o.c.k ...
-
I would almost think that any condensate, if the cup is creating a good vacuum, would actually form on the outside of the suction cup instead of underneath, but we'll see.
There is a small amount of air under the cup which will have moisture in it. But as I normally "wet" suctions cups before sticking them to my windshield for my GPS, I would not worry about that. The failure mode in what you are doing is slow air infiltration - no seal is perfect and "nature abhors a vacuum".
Update : I was trying to figure out what brit-slang caught the attention of the swear filter - didn't guess re-C.o.c.k
-
You've been thinking about this! Your suction cup sounds like one Panavise sells as an optional base for their vises. Should hold solidly on glass for a long time, but I'd still want to leave a pillow under it. Of course, I'm influenced by having only one camera to lose. I couldn't bear to give it up for a year.
Whatever you clean the window with for better seal, rinse thoroughly with water. Alcohol, oils, or detergents caught under the suction cup may harden the rubber or turn it to goop over time.
It's fun to see someone with a real project, as opposed to the random experimenting that I do with CHDK. Good luck.
-
The suction cup design is very similar to the Panavise, withe the lever that you flip down to engage the suction. But it has a longer, 8" neck that can be adjusted, and the head is better articulating than most. The neck is really difficult to bend, and I am confident it's not going to change positions or sag over time, etc. (especially when the entire camera weighs only 6oz).
I'm cautiously hopeful that this will work well for the duration of the project, but it's a little scary having the whole thing hinge on such compromises, you know? but when we approached administration with the proposal that would have cost $1500 (DSLR, weatherproof heated/cooled enclosure, solar-powered, the works, totally self-sustaining) they kicked it down without a moment's thought. So we had to go cheap. This whole thing, including the S90, power adapter, eye-fi card, and camera mount, has cost just shy of $300...the bulk of that being because I decided I wanted a wider FOV than the 570IS was giving me and bought the $225 S90 (camera + power adapter). cost could have been under $150 if we'd used my cheap-o camera I bought at first.
I'm still unsure of some settings; I have the camera set with MF right now, which is fine, but it's just in program mode at present. I was thinking I should probably be shooting in aperture priority, but then wasn't sure what aperture to set it at. My first instinct was to go with something like f/8, but then reading on some other forums, they say that can introduce more flicker than if you shoot wide open. Well, wide open is f/2 on this camera, and even at ISO 80, i would think that would just be too wide for daytime shooting. So I may go back and tweak those as I find more information, but in the meantime, it's shooting.
I'm also not sure if I should be shooting in higher resolution than I currently am. Right now I"m using M2 (2,272 x 1,704) which would easily let me crop a 1920x1080 frame out of it, but wonder if I'd get better results by shooting in higher resolution and then letting lightroom scale down the photos. I'm guessing the difference would be miniscule, at best.
The thought of shooting raw mode was rather appealing at first, since we have an essentially endless memory card, but the thought of processing 30,000 raw files sickened me. even at the M2 fine jpg setting, the photos are a little over 1MB each, so I'll be looking at 30GB of photos when the project is done. RAW would be 10x that, easily.
I've got to say I'm really loving this eye-fi card. it's an 8GB card, so using the M2 resolution, it can hold about a month's worth of photos, but I don't even need that, it turns out. I was able to stick an old GX620 PC (it's the model we retired last year so surplus inventory) up in the sound booth, and equip it with wifi in addition to its wired ethernet. It's sitting up there right now, headless (just a tower, power cord, and ethernet connection, no monitor, kb, or mouse, all administration is done remotely) whose sole job it is to download the photos from the eye-fi card via direct-mode wifi after each shot is taken. The folder that those download to is shared out to me in my office, and so I see each photo within a minute of being taken.
Our webadmin is going to write a script that constantly pulls the most recent photo from the folder and displays it on our webpage as a sort of "construction cam" for the PR department. two birds, one stone, i suppose.
Last thing I'm still not sure of is the interval. I'm using 5 minutes right now, but I'm tempted to knock that down to 3 minutes to slow down the biweekly videos a little bit. But that adds a lot to the total photo count in the end, and I'm not sure I want to do that or not.