Paul C. Buff, Inc. Technical Forum

Technical Discussion Forum for all Paul C. Buff, Inc. Products

Login

Post a reply
 [ 65 posts ] 

Tue Dec 22, 2009 4:15 pm

Site Admin
Joined: Tue Dec 08, 2009 10:43 am
Posts: 5266

Sync speed, as described in a separate thread, is a physical limitation of the camera's shutter. In a low light environment, the flash durations are fast enough to freeze motion, thus a faster shutter speed is not beneficial.

To circumvent the shutter limitation, and achieve entire frame exposure of flash, camera manufacturers designed their brand of flashes and cameras to communicate with each other. The camera tells the flash to fire multiple times (or one longer time), to ensure any part of the frame covered by the shutter at one part of the exposure gets evenly exposed by the flash during another part of the exposure. This works fairly well. However, with no lunch being free, there are some drawbacks. First is a reduction in effective power, since the same amount of power has to be divided up. Second is the need for more expensive and dedicated flashes, with less power than a similarly priced mono light.

Manufacturers of some newer remote controls have started tapping into the camera/flash communications. This is allowing users to alter how/when their cameras trigger studio lights. Photographers are now able to fire studio lights early, and rely on the "slow" burntime of studio lights to achieve faster flash durations (think- a short lived continuous light). This is known as "HyperSync™" technology in the Pocket Wizard line. Again, the cost of this lunch is lower effective power and potentially uneven lighting, as the power drops over the duration of the burn and different parts of the frame will be exposed to different power levels. Additionally, the effects are likely to change from one shutter speed to another. (***Edit for additional info***>>>) The effects can also vary depending on flash model to flash model (those with longer flash durations being more sucessful), power setting to power setting, as well as camera model to camera model.

None of these truly change the x-sync speed of the camera.

TS




Top Top
Profile
 

#

Sat Dec 26, 2009 3:05 pm

Joined: Sat Dec 26, 2009 12:54 pm
Posts: 10

This necessitates the use of Radio Poppers or the Pocket Wizard TT-5 units to work around the mechanics of these shutters.

The Pocket wizards TT-5 units work well with all the Buff lights using their high speed sync. mode. Tested in the sunshine by moi.

They work best with the Zeuss lights because you still need f/22 at a reasonable distance if you use a box, but even an AB 400 will work if it's close enough to the subject.




Last edited by ClaudeJ1 on Tue Jan 12, 2010 2:12 pm, edited 1 time in total.

Top Top
Profile
 

#

Thu Jan 07, 2010 3:36 pm

Joined: Thu Jan 07, 2010 2:46 pm
Posts: 53

Claude has got things a bit mixed up.

No production CMOS sensor camera from Nikon, Canon, or Sony has ever used a sensor capable of electronic shutter operation. Neither did any production Foveon sensor. I used the Foveon F7X3-C9110 sensor in an industrial camera I designed. I assure you, the leaf shutter we used (because the camera had dual sensors and a beam splitter) imposed objectionable constraints on the design operation of the camera, and if it were possible to do without it, we certainly would have.

The Foveon FX17-78-F13 CMOS sensor used in the Sigma DP-1 and DP-2 is also quite incapable of snap shutter operation. The DP-1 and SP-2 have high sync speeds because they use leaf shutters. There are numerous pictures available of torn down DP-1 and DP-2 cameras that confirm this, and the data sheets for the F7X3-C9110 and FX17-78-F13 can be downloaded from Foveon.

Nikon did use electronic shutter CCD (not CMOS) sensors in their D1, D70, D50, and D40 cameras (that is why Paul's people used a D40 for their sync speed experiments). Canon used an electronic shutter CCD sensor in their original 1D. No one currently uses electronic shutters, for two very good reasons.

  • Electronic shutter circuitry (commonly referred to as a "snap shutter") requires a "dark" (metal shielded) storage area adjacent to each light sensitive photodiode. This dark area has to be the same size as the photodiode, so incorporating it into a sensor means, literally, cutting the size of the photodiode in half. This decreases dynamic range and limits high ISO sensitivity.
  • While the snap shutter allows you to clear the photodiodes simultaneously, and read them simultaneously, it doesn't do anything about the light that accumulates in the photodiodes during the period outside exposure. The photodiodes continue converting light into electricity, and under lighting conditions requiring short exposures (down to 1/16,000 sec for the Nikon D1 and Canon 1D), the extra light during readout (1/3 second in the "good old days) can overload the photodiodes and the overload spills into adjacent circuitry, contaminating the data being read out. The end result is that blown highlights turn into vertical bands, frequently all the way from the top of the image to the bottom.

It is these very real reasons, poor high ISO and dynamic range, poor handling of blown highlights, that camera makers avoid snap shutters. There was not some shadow conspiracy of "crippling mechanical shutters of the last millenium for proprietary and planned obsolescence purposes".

The original Hasselblad/Foveon DFinity used snap shutters because it was a three sensor beam splitter camera. That is one of the reasons the design failed. The other reasons were the well documented optical constraints imposed by the beam splitter design.

  • The index of refraction of the splitters caused a dramatic increase in spherical aberrations at large apertures. The lenses are manageable at f8 or f11, but get progressively softer at larger apertures: by f2.8 they're nearly unusable. High end optical houses like Schneider and Coastal design lenses specifically to cope with 2mm of optical glass in the image path, and 3-CCD video cameras have special lens designs that compensate for 20mm of glass (their sensors are small). Nothing could help the DFinity with 50mm of glass in the optic path.
  • The depth of the beam splitters are equal to about 3x the sensor's height, over twice the sensor diagonal. So the DFinity was restricted the crop factor to a massive 2.5x.
  • Dichroic beam splitters are not colorimetric. They have spectral responses with insufficient overlap between the red, green, and blue responses to mimic the response curves of the human eye. So you have lots of problems with metameric failure. Colors that appear identical to a human eye (like mineral based makeup carefully blended to match organic based cloth dyes) do not appear identical to a camera with metameric failure. Color film camera and Bayer filter digital cameras use organic filters formulated to have as much overlap as the organics of a human eye. Dichroic folters can't match this.

There was little demand for a system where lenses had to be used at small apertures, and you could not shoot wide angle shots (the widest wide angle available for the system functioned as a longish normal), as well as the color accuracy issues of a dichroic beam splitter system, and the dynamic range, high ISO, and blooming issues of a snap shutter system. Bye bye DFinity, bye bye snap shutters in medium format gear, totally.




Top Top
Profile
 

#

Sat Jan 09, 2010 9:26 am

Joined: Sat Jan 09, 2010 9:11 am
Posts: 8

This is allowing users to alter how/when their cameras trigger studio lights. Photographers are now able to fire studio lights early, and rely on the "slow" burntime of studio lights to achieve faster flash durations (think- a short lived continuous light). This is known as "Hyper Sync".

We have always been able to trigger the flash early by tricking the camera into thinking it is attached to a flash in high sync mode. In that mode, the sync signal occurs a msec or so before the first curtain opens and continues to be low after the rear curtain closes.

Given a suitably long duration flash (approx 4msec) like a speedlight on full power, you can have the flash on during the entire duration of the sensor being exposed. yes, it will uneven, but since this technique is most often used to balance bright ambient light background with a foreground subject, it is still very useful.

So if you can control the delay from when the camera puts the sync signal low to when the flash is triggered, you can can optimize this affect.

So the real question is, will there be an option to adjust the delay in firing the flash using cybersyncs?

David




Top Top
Profile
 

#

Mon Jan 11, 2010 5:51 pm

Site Admin
Joined: Tue Dec 08, 2009 10:43 am
Posts: 5266

David,
please elaborate on the technique referenced here, as i am not aware of it. As well, is this doable for most cameras?

"We have always been able to trigger the flash early by tricking the camera into thinking it is attached to a flash in high sync mode"




Top Top
Profile
 

#

Mon Jan 11, 2010 8:26 pm

Joined: Sat Jan 09, 2010 9:11 am
Posts: 8

please elaborate on the technique referenced here, as i am not aware of it. As well, is this doable for most cameras?

I know it works for Nikon cameras and from reading on the internet, Canon. Can't speak for other brands.

Here is a nice discussion:

http://www.flickr.com/groups/strobist/d ... 362003933/

Basically if you have a camera with a focal plane shutter and it has a high sync speed (HSS) mode. You put a flash on the camera that is HSS capable and then dial the shutter speed up until it exceeds the sync speed. You trigger the real flash using the PC port.

When the camera goes into the HSS mode, the PC sync signal occurs a couple of milliseconds before the first curtain opens and continues well after the second curtain closes.

If you use this PC sync signal to trigger a flash and it has a suitably long duration (like a studio strobe or speedlight on full power) you will have light during the duration when the curtains are exposing the sensor.

This technique works best if you can introduce a delay in triggering the flash and hence RF triggers work very well. By controlling the delay and syncrhonizing the flash with the curtains, in theory, you could optimize this affect.

In what I think is a rather perverse thing, a Nikon D40, which has essentially an unlimited sync speed does not work in this mode and when you use a RF trigger it becomes limited by the delay in the RF trigger.

While a camera with a focal plan shutter like a D300 which has a max sync speed of really around 1/400, (although you can only set 1/320) actually benefits from the delay in RF triggers using this HSS hack.

And the very fast and nice cybersync system may actually be too fast to get the most out of this HSS hack.

Here is a sequence I shot with a D300, top series normal shoe mount SB800 full power
middle series SB800 triggered through the PC sync port with a skyport system
last series 1/500 and 1/1000 with the HSS sync hack and the last one 2650.jpg is full 1/2 power with flash on camera

Image




Top Top
Profile
 

#

Tue Jan 12, 2010 2:03 pm

Joined: Sat Dec 26, 2009 12:54 pm
Posts: 10

Joseph S. Wisniewski wrote:
Claude has got things a bit mixed up.Apparently so

No production CMOS sensor camera from Nikon, Canon, or Sony has ever used a sensor capable of electronic shutter operation. Neither did any production Foveon sensor. I used the Foveon F7X3-C9110 sensor in an industrial camera I designed. I assure you, the leaf shutter we used (because the camera had dual sensors and a beam splitter) imposed objectionable constraints on the design operation of the camera, and if it were possible to do without it, we certainly would have.Industrial cameras are never used in the sun on an engagment session or Wedding, but are very cool nonetheless

The Foveon FX17-78-F13 CMOS sensor used in the Sigma DP-1 and DP-2 is also quite incapable of snap shutter operation. The DP-1 and SP-2 have high sync speeds because they use leaf shutters. There are numerous pictures available of torn down DP-1 and DP-2 cameras that confirm this, and the data sheets for the F7X3-C9110 and FX17-78-F13 can be downloaded from Foveon.Well than, that's what I get for assuming they would carry the original Foveon's perfromance benefits in their entirety (I have owned 3 of them at $25,000 each and still have one. I shall not err in the future by checking the tech specs fist and not assuming anything. I make no sweeping proclamations about shutters

Nikon did use electronic shutter CCD (not CMOS) sensors in their D1, D70, D50, and D40 cameras (that is why Paul's people used a D40 for their sync speed experiments). Canon used an electronic shutter CCD sensor in their original 1D. No one currently uses electronic shutters, for two very good reasons.

  • Electronic shutter circuitry (commonly referred to as a "snap shutter") requires a "dark" (metal shielded) storage area adjacent to each light sensitive photodiode. This dark area has to be the same size as the photodiode, so incorporating it into a sensor means, literally, cutting the size of the photodiode in half. This decreases dynamic range and limits high ISO sensitivity.
  • While the snap shutter allows you to clear the photodiodes simultaneously, and read them simultaneously, it doesn't do anything about the light that accumulates in the photodiodes during the period outside exposure. The photodiodes continue converting light into electricity, and under lighting conditions requiring short exposures (down to 1/16,000 sec for the Nikon D1 and Canon 1D), the extra light during readout (1/3 second in the "good old days) can overload the photodiodes and the overload spills into adjacent circuitry, contaminating the data being read out. The end result is that blown highlights turn into vertical bands, frequently all the way from the top of the image to the bottom.
Very interesting stuff. I still say that if they had decided to evolve "snap shutter" sensors to perform, noise-wise, as well as the ones that cannot do this function, we would be better off. I would gladly give ups all of these useless MEGAPIXELS and trade for an "Easy-Sync." feature in a camera. Too bad no one has gone this route

It is these very real reasons, poor high ISO and dynamic range, poor handling of blown highlights, that camera makers avoid snap shutters. There was not some shadow conspiracy of "crippling mechanical shutters of the last millenium for proprietary and planned obsolescence purposes".

The original Hasselblad/Foveon DFinity used snap shutters because it was a three sensor beam splitter camera. That is one of the reasons the design failed. The other reasons were the well documented optical constraints imposed by the beam splitter design.

  • The index of refraction of the splitters caused a dramatic increase in spherical aberrations at large apertures. The lenses are manageable at f8 or f11, but get progressively softer at larger apertures: by f2.8 they're nearly unusable. High end optical houses like Schneider and Coastal design lenses specifically to cope with 2mm of optical glass in the image path, and 3-CCD video cameras have special lens designs that compensate for 20mm of glass (their sensors are small). Nothing could help the DFinity with 50mm of glass in the optic path.
  • The depth of the beam splitters are equal to about 3x the sensor's height, over twice the sensor diagonal. So the DFinity was restricted the crop factor to a massive 2.5x.
  • Dichroic beam splitters are not colorimetric. They have spectral responses with insufficient overlap between the red, green, and blue responses to mimic the response curves of the human eye. So you have lots of problems with metameric failure. Colors that appear identical to a human eye (like mineral based makeup carefully blended to match organic based cloth dyes) do not appear identical to a camera with metameric failure. Color film camera and Bayer filter digital cameras use organic filters formulated to have as much overlap as the organics of a human eye. Dichroic folters can't match this.
You have very accurately described some of the shortcomings of the original Foveon prism camera. Nonetheless, on the pratical side, I was the only one in town who could make 30x40 family portraits (with Kodachrome-like colors) on Canvas in the year 2000 because of this "flawed" moire-free camera

There was little demand for a system where lenses had to be used at small apertures, and you could not shoot wide angle shots (the widest wide angle available for the system functioned as a longish normal), as well as the color accuracy issues of a dichroic beam splitter system, and the dynamic range, high ISO, and blooming issues of a snap shutter system. Bye bye DFinity, bye bye snap shutters in medium format gear, totally.
I hear you on this one. It's the main reason I got an 8mm Sigma fisheye for the Foveon and it was a pain trying to correct rectilinears




Top Top
Profile
 

#

Thu Jan 14, 2010 12:35 am

Site Admin
Joined: Tue Dec 08, 2009 11:49 am
Posts: 1432

purdyd wrote:
This is allowing users to alter how/when their cameras trigger studio lights. Photographers are now able to fire studio lights early, and rely on the "slow" burntime of studio lights to achieve faster flash durations (think- a short lived continuous light). This is known as "Hyper Sync".

We have always been able to trigger the flash early by tricking the camera into thinking it is attached to a flash in high sync mode. In that mode, the sync signal occurs a msec or so before the first curtain opens and continues to be low after the rear curtain closes.

Given a suitably long duration flash (approx 4msec) like a speedlight on full power, you can have the flash on during the entire duration of the sensor being exposed. yes, it will uneven, but since this technique is most often used to balance bright ambient light background with a foreground subject, it is still very useful.

So if you can control the delay from when the camera puts the sync signal low to when the flash is triggered, you can can optimize this affect.

So the real question is, will there be an option to adjust the delay in firing the flash using cybersyncs?

David


This can be made to work, but you end up using only a tiny chunk of the available flashpower. This what Radio Popper does. True HHS does about the same thing . . . it stretches out the flash duration, prefires it, then uses a little chunk. At least a dedicated true HHS system has control over how long the flash duration must be to maximize its power at a given exposure time. Either way you end up with very little usable flashpower (a couple of WS) or a big dark to light gradient across the frame if you time it to get any significant power.

OK for shooting water drops, but IMHO, of no value for sports unless you want to run around the court three feet from the guy you're shooting :shock:

I don't personally buy this tricky method. If I want full, or nearly full flash output at up to about 1/1600 sync speed I would use my D40 or get a G11 and settle for the compromise caused by the CCD sensor.

No adjustable delay for CyberSync on the horizon - you would have to delay the camera - not the flash - very complex and you still end up with little flashpower.




Top Top
Profile
 

#

Thu Jan 14, 2010 11:26 pm

Joined: Sat Jan 09, 2010 9:11 am
Posts: 8

This can be made to work, but you end up using only a tiny chunk of the available flashpower. This what Radio Popper does. True HHS does about the same thing . . . it stretches out the flash duration, prefires it, then uses a little chunk. At least a dedicated true HHS system has control over how long the flash duration must be to maximize its power at a given exposure time. Either way you end up with very little usable flashpower (a couple of WS) or a big dark to light gradient across the frame if you time it to get any significant power.

contrast the upper right hand corner, normal high speed sync with the left hand bottom corner, hacked high speed sync

Image

OK for shooting water drops, but IMHO, of no value for sports unless you want to run around the court three feet from the guy you're shooting :shock:

Image

I don't personally buy this tricky method. If I want full, or nearly full flash output at up to about 1/1600 sync speed I would use my D40 or get a G11 and settle for the compromise caused by the CCD sensor.

no, you won't get full power flash from a D40 at 1/1600 sync speed unless your full power flash is faster than 1/1600 in duration

Image

also, notice in the above sequence you actually get a dark frame at 1/1000 when you use a regular RF trigger because the delay robs time that the flash should be firing when the curtains are open - and the cybersyncs will do the same, they will rob power from the flash

but notice the hacked 1/1000 frame dsc2649 is not dark

No adjustable delay for CyberSync on the horizon - you would have to delay the camera - not the flash - very complex and you still end up with little flashpower

the heavy lifting has already been done by the camera companies, they have a delay built in when you put a FP capable flash on them

thanks for your honesty and directness on this capability, it makes my future buying decisions easier




Top Top
Profile
 

#

Fri Jan 15, 2010 1:08 am

Site Admin
Joined: Tue Dec 08, 2009 11:49 am
Posts: 1432

http://www.paulcbuff.com/forums/viewtop ... f=2&t=6176

AB400 (or even 1600) is much faster than SB800. CyberSync has less sync delay than Skyport or most other radio trippers. Notice these shots lost only .45f of available flash power with D40 and AB 400 at 1/1600, and .9f at 1/2000, with CyberSync (1/4000 sec CS delay). Without CS you would probably see no loss at 1/1600 and little loss at 1/2500 sync.




Top Top
Profile
 
Display posts from previous:  Sort by  
Post a reply
 [ 65 posts ] 

All times are UTC - 6 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 96 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum


cron