If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below. |
|
|
Thread Tools | Display Modes |
#1
|
|||
|
|||
Lazy people and "smartphones" continue to erode P&S sales
Trevor wrote:
"Whisky-dave" wrote in message ... I'm suprised pre-triggers aren;t used as we use them in electronics. That way you can store images that the camera sees before pressing the shutter that way you have a minus time 'lag'. You simply start shooting whenever you like at somewhere between 4 and 60 frames a second. Then delete what you don't need. A pretrigger simply does the same automatically. Maschine gun shooting's only good if you don't care much *what* point in time you hit exactly, but care only that you hit *lots* of points. With a DSLR not in movie mode, you introduce a lot of shutter wear though if you are *always* shooting. Never mind battery usage ... a DSLR can have everything but the buttons off (and they don't draw power if not pressed), a pre-shooting system has to have everything but the card writing running at full tilt. -Wolfgang |
#2
|
|||
|
|||
Lazy people and "smartphones" continue to erode P&S sales
Wolfgang Weisselberg writes:
Trevor wrote: "Whisky-dave" wrote in message ... I'm suprised pre-triggers aren;t used as we use them in electronics. That way you can store images that the camera sees before pressing the shutter that way you have a minus time 'lag'. You simply start shooting whenever you like at somewhere between 4 and 60 frames a second. Then delete what you don't need. A pretrigger simply does the same automatically. Maschine gun shooting's only good if you don't care much *what* point in time you hit exactly, but care only that you hit *lots* of points. Not lots, but you've got a frame within 1/n of any actual point in time. So, depending on the speed of the event, and whether you can trigger reliably at some early moment of it (either manually or with tech of some sort), and how accurately you need to hit your point, it may be the easiest and cheapest solution. For example, my bowling ball mortar pictures were done that way. Using a fuse to set off the mortar didn't give us a very predicatable time, and I didn't have a sound trigger (still don't, but waiting for Trigger Trap from the Kickstarter project, due this spring), but 5 or 8 frames per second got us interesting shots. With a DSLR not in movie mode, you introduce a lot of shutter wear though if you are *always* shooting. Never mind battery usage ... a DSLR can have everything but the buttons off (and they don't draw power if not pressed), a pre-shooting system has to have everything but the card writing running at full tilt. Would be a nice option to turn on -- but not yet compatible with the hardware platform, quite. -- David Dyer-Bennet, ; http://dd-b.net/ Snapshots: http://dd-b.net/dd-b/SnapshotAlbum/data/ Photos: http://dd-b.net/photography/gallery/ Dragaera: http://dragaera.info |
#3
|
|||
|
|||
Lazy people and "smartphones" continue to erode P&S sales
David Dyer-Bennet wrote:
Wolfgang Weisselberg writes: Trevor wrote: "Whisky-dave" wrote in message I'm suprised pre-triggers aren;t used as we use them in electronics. That way you can store images that the camera sees before pressing the shutter that way you have a minus time 'lag'. You simply start shooting whenever you like at somewhere between 4 and 60 frames a second. Then delete what you don't need. A pretrigger simply does the same automatically. Maschine gun shooting's only good if you don't care much *what* point in time you hit exactly, but care only that you hit *lots* of points. Not lots, but you've got a frame within 1/n of any actual point in time. So, depending on the speed of the event, and whether you can trigger reliably at some early moment of it (either manually or with tech of some sort), and how accurately you need to hit your point, it may be the easiest and cheapest solution. Of course. Slowing down high speed actions is a typical case. OTOH, a photo where the bullet just has left the (now exploding) apple probably wouldn't work, due to light and exact timing of the extremely short exposure. For example, my bowling ball mortar pictures were done that way. Using a fuse to set off the mortar didn't give us a very predicatable time, and I didn't have a sound trigger (still don't, but waiting for Trigger Trap from the Kickstarter project, due this spring), but 5 or 8 frames per second got us interesting shots. You could have used an (e.g. infrared) active electric eye, a passive electic eye (shadow of the ball) or even a very thin wire that's torn by the bowling ball, too. With a DSLR not in movie mode, you introduce a lot of shutter wear though if you are *always* shooting. Never mind battery usage ... a DSLR can have everything but the buttons off (and they don't draw power if not pressed), a pre-shooting system has to have everything but the card writing running at full tilt. Would be a nice option to turn on -- but not yet compatible with the hardware platform, quite. You want really high speed? Then you want low resolution[1]. Readout speeds are limited by image quality. -Wolfgang [1] Or sweeping over several (logical and/or physical) sensors. |
#4
|
|||
|
|||
Lazy people and "smartphones" continue to erode P&S sales
Wolfgang Weisselberg writes:
David Dyer-Bennet wrote: Wolfgang Weisselberg writes: Trevor wrote: "Whisky-dave" wrote in message I'm suprised pre-triggers aren;t used as we use them in electronics. That way you can store images that the camera sees before pressing the shutter that way you have a minus time 'lag'. You simply start shooting whenever you like at somewhere between 4 and 60 frames a second. Then delete what you don't need. A pretrigger simply does the same automatically. Maschine gun shooting's only good if you don't care much *what* point in time you hit exactly, but care only that you hit *lots* of points. Not lots, but you've got a frame within 1/n of any actual point in time. So, depending on the speed of the event, and whether you can trigger reliably at some early moment of it (either manually or with tech of some sort), and how accurately you need to hit your point, it may be the easiest and cheapest solution. Of course. Slowing down high speed actions is a typical case. OTOH, a photo where the bullet just has left the (now exploding) apple probably wouldn't work, due to light and exact timing of the extremely short exposure. Yes, 1/4000 sec. actually freezes the bowling ball, and my mechanical shutter can do that. For significantly higher speeds you end up using flash in the dark, and generally therefore only get one shot per trial. I'd think it might be worth rigging a laser trigger instead if you were doing it on film, so you wouldn't have to experiment to get the placement right. Never did this on film. For example, my bowling ball mortar pictures were done that way. Using a fuse to set off the mortar didn't give us a very predicatable time, and I didn't have a sound trigger (still don't, but waiting for Trigger Trap from the Kickstarter project, due this spring), but 5 or 8 frames per second got us interesting shots. You could have used an (e.g. infrared) active electric eye, a passive electic eye (shadow of the ball) or even a very thin wire that's torn by the bowling ball, too. Certainly, although the muzzle flash was *huge* (being enhanced with two baggies of gasoline :-) ). And some of the best shots were far enough back that rigging a light beam without equipment appearing in the shot would have been hard. But a sound trigger plus delay would be a good choice too. (As I mentioned, I supported Trigger Trap on Kickstarter, and am looking forward to receiving one this spring. It can use any of these modes. Actually, it can probably trigger *on* the muzzle flash.) (Here's one from the first year of the bowling ball mortar shoots http://dd-b.net/cgi-bin/picpage.pl/photography/gallery/fourth-2008?pic=ddb%2020080704%20010-176. Sorry the URL is so long, make sure it doesn't get broken or it won't work! Use the "up" link from there to find the rest of the shots from that session.) With a DSLR not in movie mode, you introduce a lot of shutter wear though if you are *always* shooting. Never mind battery usage ... a DSLR can have everything but the buttons off (and they don't draw power if not pressed), a pre-shooting system has to have everything but the card writing running at full tilt. Would be a nice option to turn on -- but not yet compatible with the hardware platform, quite. You want really high speed? Then you want low resolution[1]. Readout speeds are limited by image quality. -Wolfgang [1] Or sweeping over several (logical and/or physical) sensors. No, of course not; I want high speed *at high resolution*; for catching fast events lazily for fun, not for making useful scientific studies of fast events :-). Yeah, not feasible at sane cost, I understand. Moving data that fast is *hard*. -- David Dyer-Bennet, ; http://dd-b.net/ Snapshots: http://dd-b.net/dd-b/SnapshotAlbum/data/ Photos: http://dd-b.net/photography/gallery/ Dragaera: http://dragaera.info |
#5
|
|||
|
|||
Lazy people and "smartphones" continue to erode P&S sales
"Wolfgang Weisselberg" wrote in message ... I'm suprised pre-triggers aren;t used as we use them in electronics. That way you can store images that the camera sees before pressing the shutter that way you have a minus time 'lag'. You simply start shooting whenever you like at somewhere between 4 and 60 frames a second. Then delete what you don't need. A pretrigger simply does the same automatically. Maschine gun shooting's only good if you don't care much *what* point in time you hit exactly, but care only that you hit *lots* of points. Right, but as I said, taking one shot is even less likely to give you the right one for fast action because of reaction time. Will depend on predictabilty of course, which is not always guaranteed. Never mind battery usage ... a DSLR can have everything but the buttons off (and they don't draw power if not pressed), Actually the processor is on to detect the button press. If you turn it off all together, pressing the button does nothing. And even when on you introduce lag when the system is in sleep mode before a shot can be taken. Not huge for a dslr these days, but significant for anyone capturing action shots. pre-shooting system has to have everything but the card writing running at full tilt. Right, and yet pro's do demand high sequence rates these days, and even the cheapest P&S's have continuous still shooting modes (apart from movie mode). Once upon a time only pro's had high speed film winders and large capacity backs on their SLR's. You had to have a good reason to go to that expense, PLUS film cost! Trevor. |
#6
|
|||
|
|||
Lazy people and "smartphones" continue to erode P&S sales
Trevor wrote:
"Wolfgang Weisselberg" wrote in message Maschine gun shooting's only good if you don't care much *what* point in time you hit exactly, but care only that you hit *lots* of points. Right, but as I said, taking one shot is even less likely to give you the right one for fast action because of reaction time. Nope. You have to differentiate between - repeatable actions: anything that can be shot again and again - semi-repeatable actions: Things that can be repeated, but only a very limited number of times, and not identically. E.g. Raising the flag at Ivo Jima - unrepeatable actions: The moment JFK was shot. The first goal in the superbowl 2012. And between - specific predictable action (where you have a 'countdown' to the decisive moment), e.g. a runner reaching the finish, a ball entering the goal after a longer flight, ... - generally predictable action (where you can sort of frame and preshoot), but cannot predict the exact moment. E.g. a water ballon, being filled, just starts to burst. - unpredictable action (no time to frame or too long a time to wait): someone robbing a store 6 years after the camera was attached. A specific lightning bolt. - well predictable actions that are too fast between the start of the 'countdown' and the event for a human to trigger: a bullet just leaving an apple. A foot kicking a ball --- at the moment where the ball is maximally deformed. If it can be triggered by an experienced human, if it can be triggered by any technical means (available to you) and is either repeatable or maths and physics can show you how much delay to use you're better off without machine gun shooting. If it's not triggerable, you have to make do with preshooting, or better, extremely high frequency shooting. You'll miss the deceive moment, but depending on the shooting frequency you'll be close. If that difference matters depends on you. Of course, if you want a sort-of movie ... maschinegun it is! Will depend on predictabilty of course, which is not always guaranteed. If you've no chance to frame it and focus it ... Never mind battery usage ... a DSLR can have everything but the buttons off (and they don't draw power if not pressed), Actually the processor is on to detect the button press. Nope. I'll bet you that it's asleep. Even desktop computers on mains send their processor asleep nowaday till there's some work to be done --- why should a battery driven camera let it run? You don't need the processor to generate an interrupt. You might need it to handle the interrupt, but not to generate the interrupt that wakes it. If you turn it off all together, pressing the button does nothing. So your laptop does nothing if you press a key and it's just in S3? And even when on you introduce lag when the system is in sleep mode before a shot can be taken. It's absolutely negible. It doesn't need to boot again, after all. (and that's 0.25s in a 20D and again much faster in newer cameras, with 0.1s being average) Not huge for a dslr these days, but significant for anyone capturing action shots. It's vastly faster than the time between half-press and full press of the shutter, even if you press as fast as you can. When a halfway modern laptop handles 400 wakeups per second and is 93% in C3 (deep sleep, every clock signal stopped) or even deeper (down to "everything but a static RAM inside the CPU at 0V") (and 0.7-2% in C2) and is then idle 90% of the rest (with another 8% in the lowest frequency) ... you can imagine yourself how fast wakeups are. Not something you want to do 100,000 times per second. Starting up the PLLs and reloading the caches takes time. But not *much* time. 400 wakeups and 93% idle time means that wakeup must be (much) faster than 1/400 of 7%, or much faster than 1/5700s. 1/10,000s or 1/30.000s or faster. That's "significant for anyone capturing action shots"? That's much faster than the action shooter's reaction time variability, much faster than even the 300x200'ish high speed video shot-to-shot time available in digital prosumer cameras. pre-shooting system has to have everything but the card writing running at full tilt. Right, and yet pro's do demand high sequence rates these days, I just see the landscape, the portrait, the architecture, the stilllife, the macro, and several other pros[1] clamoring for 600 fps[2] There are some very few areas where 10, 11, 12, 13 fps are well usable --- remember, pros demand high resolution and high image quality! And that means usually at least a crop DSLR, usually a FF DSLR or medium format--- amongst them: - Studio where the second shot is just lit by the backlight flashes (to automate masking). They'd want 50 or even more fps, so subject movement (hair waving from the wind machine, for example) better frozen. However, reading *many* pixels *fast* means either low image quality or masses of digitizers (i.e. extremely expensive). - some type of wildlife (not all of them are happy to shoo the wildlife away by mirror and shutter maschine gun firing). Capturing the chameleon's tongue shooting out. - Expanding on the above: when the subject is non-cooperative and fast moving and has fleeting, irregular, too hard to predict moments you're trying to catch. Politicans showing their real face, for example. Pros might also want to capture events very close together: Pie approaching pop star, pie hitting in face, pie parts flying everywhere, body guards tackling pop star, pop star just about to hit the ground, pop star on ground while body guards draw weapons and start firing wildly into the masses :-) and even the cheapest P&S's have continuous still shooting modes (apart from movie mode). But usually no AF then, and usually a 1-3 fps. Once upon a time only pro's had high speed film winders and large capacity backs on their SLR's. You had to have a good reason to go to that expense, PLUS film cost! Once upin a time most pros didn't use either, and still did marvelous work. -Wolfgang [1] And that's "pros do demand", not "pro's", there's nothing belonging to "pro" that demands, there is no "pro is do demand" either. Sheesh, I'm supposed to be the non-native speaker here, making stupid mistakes ... [2] that's AFAIR available with some consumer/prosumer bridge cameras ... at very low resolution. |
#7
|
|||
|
|||
Lazy people and "smartphones" continue to erode P&S sales
David Dyer-Bennet wrote:
Wolfgang Weisselberg writes: David Dyer-Bennet wrote: Yes, 1/4000 sec. actually freezes the bowling ball, and my mechanical shutter can do that. But not for the whole image at the same time --- it's a travelling slot and that means the ball will likely not be circular in the image. (It depends of course how visible that is and if that is good, bad or neutral for the image.) For significantly higher speeds you end up using flash in the dark, and generally therefore only get one shot per trial. I'd think it might be worth rigging a laser trigger instead if you were doing it on film, so you wouldn't have to experiment to get the placement right. Never did this on film. Yup. Though you can probably get close with some math. For example, my bowling ball mortar pictures were done that way. Using a fuse to set off the mortar didn't give us a very predicatable time, and I didn't have a sound trigger (still don't, but waiting for Trigger Trap from the Kickstarter project, due this spring), but 5 or 8 frames per second got us interesting shots. You could have used an (e.g. infrared) active electric eye, a passive electic eye (shadow of the ball) or even a very thin wire that's torn by the bowling ball, too. Certainly, although the muzzle flash was *huge* (being enhanced with two baggies of gasoline :-) ). And some of the best shots were far enough back that rigging a light beam without equipment appearing in the shot would have been hard. But a sound trigger plus delay would be a good choice too. (As I mentioned, I supported Trigger Trap on Kickstarter, and am looking forward to receiving one this spring. It can use any of these modes. Actually, it can probably trigger *on* the muzzle flash.) (Here's one from the first year of the bowling ball mortar shoots http://dd-b.net/cgi-bin/picpage.pl/photography/gallery/fourth-2008?pic=ddb%2020080704%20010-176. Sorry the URL is so long, make sure it doesn't get broken or it won't work! Use the "up" link from there to find the rest of the shots from that session.) Nice. So we'll see some more interesting shots in autumn. You want really high speed? Then you want low resolution[1]. Readout speeds are limited by image quality. [1] Or sweeping over several (logical and/or physical) sensors. No, of course not; I want high speed *at high resolution*; for catching fast events lazily for fun, not for making useful scientific studies of fast events :-). Yeah, not feasible at sane cost, I understand. Oh, it's feasible enough right now ... with analog methods. :-) Moving data that fast is *hard*. Moving data isn't the main problem. The A/D converters just need some time to reach quality, AFAIUI. Some time * xx million pixels = slow. No wonder most faster cameras have multiple A/D converters. -Wolfgang |
#8
|
|||
|
|||
Lazy people and "smartphones" continue to erode P&S sales
Wolfgang Weisselberg writes:
David Dyer-Bennet wrote: Wolfgang Weisselberg writes: David Dyer-Bennet wrote: Yes, 1/4000 sec. actually freezes the bowling ball, and my mechanical shutter can do that. But not for the whole image at the same time --- it's a travelling slot and that means the ball will likely not be circular in the image. (It depends of course how visible that is and if that is good, bad or neutral for the image.) Tolerable, in any case. And it's not very warped (the link to it was in the previous message; here it is again http://dd-b.net/cgi-bin/picpage.pl/photography/gallery/fourth-2008?pic=ddb%2020080704%20010-176). For significantly higher speeds you end up using flash in the dark, and generally therefore only get one shot per trial. I'd think it might be worth rigging a laser trigger instead if you were doing it on film, so you wouldn't have to experiment to get the placement right. Never did this on film. Yup. Though you can probably get close with some math. And with more experience you'd get better at it -- as with most things. Makes using the same gun and ammo each time more important, for example. I hear that you can see the image by eye due to persistence of vision, but you'd probably be safely back from the impact area so you would be seeing it from quite a distance. For example, my bowling ball mortar pictures were done that way. Using a fuse to set off the mortar didn't give us a very predicatable time, and I didn't have a sound trigger (still don't, but waiting for Trigger Trap from the Kickstarter project, due this spring), but 5 or 8 frames per second got us interesting shots. You could have used an (e.g. infrared) active electric eye, a passive electic eye (shadow of the ball) or even a very thin wire that's torn by the bowling ball, too. Certainly, although the muzzle flash was *huge* (being enhanced with two baggies of gasoline :-) ). And some of the best shots were far enough back that rigging a light beam without equipment appearing in the shot would have been hard. But a sound trigger plus delay would be a good choice too. (As I mentioned, I supported Trigger Trap on Kickstarter, and am looking forward to receiving one this spring. It can use any of these modes. Actually, it can probably trigger *on* the muzzle flash.) (Here's one from the first year of the bowling ball mortar shoots http://dd-b.net/cgi-bin/picpage.pl/photography/gallery/fourth-2008?pic=ddb%2020080704%20010-176. Sorry the URL is so long, make sure it doesn't get broken or it won't work! Use the "up" link from there to find the rest of the shots from that session.) Nice. So we'll see some more interesting shots in autumn. We can hope. May take a while to learn how to productively employ the new trigger, which is now delayed until April. You want really high speed? Then you want low resolution[1]. Readout speeds are limited by image quality. [1] Or sweeping over several (logical and/or physical) sensors. No, of course not; I want high speed *at high resolution*; for catching fast events lazily for fun, not for making useful scientific studies of fast events :-). Yeah, not feasible at sane cost, I understand. Oh, it's feasible enough right now ... with analog methods. :-) Mmmmmmm; strip cameras! They did some wild stuff for the Manhattan project, long long ago now. But then they had a real budget. Moving data that fast is *hard*. Moving data isn't the main problem. The A/D converters just need some time to reach quality, AFAIUI. Some time * xx million pixels = slow. No wonder most faster cameras have multiple A/D converters. If you want multiple photos, data rates are a problem, or you need a reall BIG ram disk. -- David Dyer-Bennet, ; http://dd-b.net/ Snapshots: http://dd-b.net/dd-b/SnapshotAlbum/data/ Photos: http://dd-b.net/photography/gallery/ Dragaera: http://dragaera.info |
#9
|
|||
|
|||
Lazy people and "smartphones" continue to erode P&S sales
David Dyer-Bennet wrote:
Moving data that fast is *hard*. Moving data isn't the main problem. The A/D converters just need some time to reach quality, AFAIUI. Some time * xx million pixels = slow. No wonder most faster cameras have multiple A/D converters. If you want multiple photos, data rates are a problem, or you need a reall BIG ram disk. Say 5000 photos at 16 MPix, 14 bit. That's 140 GB --- not terribly much. You can probably compress it to ~80-90 GB. Sure, you won't find that in a prosumer camera today or next year ... but give it a couple of years ... -Wolfgang |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Lazy people and "smartphones" continue to erode P&S sales | Wolfgang Weisselberg | Digital SLR Cameras | 0 | January 28th 12 03:45 AM |
Lazy people and "smartphones" continue to erode P&S sales | PeterN | Digital SLR Cameras | 5 | January 22nd 12 12:08 AM |
Lazy people and "smartphones" continue to erode P&S sales | Alfred Molon[_4_] | Digital SLR Cameras | 3 | January 17th 12 06:03 AM |
Lazy people and "smartphones" continue to erode P&S sales | Alfred Molon[_4_] | Digital Photography | 2 | January 16th 12 07:41 PM |