DPChallenge: A Digital Photography Contest You are not logged in. (log in or register) 
 

DPChallenge Forums >> Individual Photograph Discussion >> Panning at night technique: why unsmooth blur?
Pages:  
Showing posts 1 - 21 of 21, (reverse)
AuthorThread
 09/26/2006 01:00:11 PM · #1
The results of panning in this picture (1/15s, f/1.4) puzzle me a bit. Why are the motion-blurred contours not blurred smoothly? Instead, they are repeated about eight times.

My best guess is that the artificial lighting (street lamps) is to blame, and those contours coinside with the frequency of the alternating current. However, the frequency is 50Hz in North America, so at 1/15s I sould catch 3.3 periods. Since each period has two vertices, then there should be, at most, 6 contours replicated, but I am seeing 8.
 09/26/2006 01:04:09 PM · #2
Dunno but it looks cool :)
09/26/2006 01:17:19 PM · #3
It would be 6 assuming that you caught it at the center of a cycle. If you started with a vertice and ended with a vertice, you would have 8 (2 ends plus the 6 in the middle).

That's how my logic goes anyway, and I'm out of my league with this topic. Becky
 09/26/2006 01:26:45 PM · #4
Originally posted by rjkstesch:
It would be 6 assuming that you caught it at the center of a cycle. If you started with a vertice and ended with a vertice, you would have 8 (2 ends plus the 6 in the middle).

I thougt about that. Suppose I caught one right at the opening of the shutter (time 0). Then there will be 3.3 more periods left to go. Since we have a peak each half period, so we have at most 6 more peaks to go. So, we should be getting at most 7 contours. I guess this is close enough for the explanation. Unless there is another explanation for the phenomenon.
 09/26/2006 01:28:13 PM · #5
In the US our power line frequency is 60 Hz, what would you calculation come out to then?
 09/26/2006 01:33:03 PM · #6
Originally posted by kawesttex:
In the US our power line frequency is 60 Hz, what would you calculation come out to then?

This would work. At 1/15s I should be catching exactly four periods of 60Hz. However, I thought that the power line was 110V/50Hz in North America, and 220V/60Hz in Europe. Am I wrong?
 09/26/2006 01:39:31 PM · #7
Originally posted by agenkin:
Originally posted by kawesttex:
In the US our power line frequency is 60 Hz, what would you calculation come out to then?

This would work. At 1/15s I should be catching exactly four periods of 60Hz. However, I thought that the power line was 110V/50Hz in North America, and 220V/60Hz in Europe. Am I wrong?


The UK (and I believe the rest of Europe also) is 220V/50 Hz (Single Phase) and the US is 115V/60Hz (Single Phase).
 09/26/2006 01:43:21 PM · #8
Yeah, it has to do with the lights actually not being on all the time, but cycling...you got it. That is a really cool photo by the way...definately deserves to be in print in a magazine.
 09/26/2006 01:45:48 PM · #9
Originally posted by kawesttex:
The UK (and I believe the rest of Europe also) is 220V/50 Hz (Single Phase) and the US is 115V/60Hz (Single Phase).

You are right. It is 60Hz in the US and Canada:
//www.voltage-converter-transformers.com/world-electricity-conversion.html
I guess the mystery is now solved. Thanks!
 09/26/2006 01:46:44 PM · #10
Originally posted by agenkin:
... Instead, they are repeated about eight times.

... Since each period has two vertices, then there should be, at most, 6 contours replicated, but I am seeing 8.

Actually, there is 8.5, almost 9. Look at the verticals just above the second rider.
 09/26/2006 01:49:13 PM · #11
Originally posted by deapee:
That is a really cool photo by the way...definately deserves to be in print in a magazine.

Thanks, but I don't think so (well, perhaps as a demonstration of the curious effect the street lighting has on panning at night). Panning worked well here, technically, but otherwise this picture is unremarkable.
 09/26/2006 01:50:15 PM · #12
thats right .. in the states it is 110/50hz ..

very interesting...
what else can we dig out of this picture...
can we find the speed at which the cyclist was moving at ??? ...
09/26/2006 01:51:58 PM · #13
if there were a ruler in the background or something that you knew the length of and if we knew there was no distortion.
 09/26/2006 01:53:11 PM · #14
Originally posted by glad2badad:
Actually, there is 8.5, almost 9. Look at the verticals just above the second rider.

Yes, 8.5. This can be still explained. The shot happened to be almost perfectly timed with the beginning of a period. Since exaclty 8 periods could fit in the 1/15s shutter speed, we could get 8 to 9 "flashes" of light.
 09/26/2006 01:54:39 PM · #15
i would say the trace of one light compared to the cyclist head looks like 1.5 his head from back to nose... thats what 30-40 mm .. 35mm...
you counted 8 peaks... so time taken for one light to flash from first position to last would be 8 x 1/60 = 0.13
speed should be distance/time = 35mm/0.13 sec = 2.7 meters/sec
mmm.. i guess thats close enought :) ... correct me if i'm wrong anyone ??
your picture would be a great riddle or a science problem pal

09/26/2006 01:56:53 PM · #16
2.7 m/sec = 9.7 km/hour ... would that make sense
 09/26/2006 02:09:51 PM · #17
Originally posted by rami:
your picture would be a great riddle or a science problem pal

:) One known measure of distance in the picture is the diameter of the wheel. To me, this looks like a hybrid bike, which has rim diameter of 28''. If I am wrong, and this is a mountain bike, then it has rim diameter of 26''. Comparing the rim diameter to the distance between the contours should tell you the distance travelled and the speed. :)
 09/26/2006 02:26:21 PM · #18
agenkin .. can you calculate speed using your method and compare it to my results.. im very curious.. and bored too
 09/26/2006 04:02:08 PM · #19
Originally posted by rami:
agenkin .. can you calculate speed using your method and compare it to my results.. im very curious.. and bored too

I will later tonight. A bit busy at the moment. If anyone wants to tackle it in the mean time, I don't mind. :)
 09/26/2006 04:09:09 PM · #20
Laurielblack is DPC's resident Math wiz, she'll solve that problem in no time! :D
 09/27/2006 07:56:57 PM · #21
Originally posted by rami:
agenkin .. can you calculate speed using your method and compare it to my results.. im very curious.. and bored too

You know what, I think that there is data missing to compute the travelled distance (and, hence, the speed). The problem is that all the multiple contours are visible on the other side of the street, but our measure of distance is much closer than that. So, we need to account for the fact that far objects project as smaller shapes on the sensor (compared to the near objects). I believe that, to calculate this, we need to know the distance from the sensor to the bike.

Another problem is that panning was done by pivoting around my own axis, rather than moving in parallel with the biker. So, the objects on the other side of the street should have travelled a longer distance on the sensor, than those on the near side of the street.

I was using a D Nikkor lens. D means that the lens communicates the focused distance to the camera (used for flash metering, I believe). I wonder if that information is recorded somewhere in the RAW image. This would give us the distance between the sensor and the biker.
Pages:  
Current Server Time: 03/13/2025 04:19:11 AM

Please log in or register to post to the forums.


Home - Challenges - Community - League - Photos - Cameras - Lenses - Learn - Help - Terms of Use - Privacy - Top ^
DPChallenge, and website content and design, Copyright © 2001-2025 Challenging Technologies, LLC.
All digital photo copyrights belong to the photographers and may not be used without permission.
Current Server Time: 03/13/2025 04:19:11 AM EDT.