Has anyone tried this - sync beyond max sync speed

[george013]

By shooting at full power the flash outputs ~ 1ms pulse of light, so as long as the shutter speed is fast enough (greater than 1/1000th s or so), for all intents and purposes we have a short burst of "continuous" light, from the sensor's point of view. In fact, the faster the shutter speed, the light less drop off will come into play (the front end of the trace that Dave has shown).

.

Manfred,
Read this carefully over and try to explain it to yourself.

George

[dje]

From your earlier post here. I just can't explain why the flash duration is depending on the shutterspeed. If it's set to manual, it shouldn't change.


What I remember of HSS, is that the flash starts just before the first curtain opens and ends when secund curtain is closed. And the flash is stroboscopic. But in this video no HSS was used.

George

George it's manual power level selection as distinct from power selection using iTTL metering. Once the power level has been determined (either way), it appears that the camera talks to the flash over the hot shoe data line and sorts out the flash duration with it.

I agree with you on the duration of HSS.

Dave

[Manfred M]

Manfred,
Read this carefully over and try to explain it to yourself.

George

I have, that's why I have written it the way I did. If your flash is outputting light for 1ms and your camera shutter is set for 1 ms (or less), then you will effectively have a continuous, but non-linear, light source, albeit quite a short one, based on what the "sensor" sees.

If I get some time over the next few days, I plan to do some testing with both an SB-600 and and SB-900 (and possibly with some other lights as well). I think I understand in principle what is being done, but I'm not completely convinced it will work as the light burst must start just as (or before) the first blade of the shutter starts its travel and must continue throughout the period of when the second shutter blade travels (or beyond). If the timing of the flash is not right, this technique cannot work.

[george013]

George it's manual power level selection as distinct from power selection using iTTL metering. Once the power level has been determined (either way), it appears that the camera talks to the flash over the hot shoe data line and sorts out the flash duration with it.

I agree with you on the duration of HSS.

Dave

In manual YOU determines the power. And if full power takes x ms than it can't be changed to less ms. Otherwise it's not full power anymore. That's what I don't understand. Not now and not in the former thread.

To be clear, we are not talking about HSS or something like that. Just pure flash.

George

[dje]

In manual YOU determines the power. And if full power takes x ms than it can't be changed to less ms. Otherwise it's not full power anymore. That's what I don't understand. Not now and not in the former thread.

To be clear, we are not talking about HSS or something like that. Just pure flash.

George

George the duration was the same for 1/60 and 1/100 and then it started to drop as the ss headed towards sync speed. I think it is all about the shutter fully open time getting less as you get close to sync speed. This does mean that the total energy in the flash gets smaller closer to sync speed ie less exposure. If you like, the max power level gets less at the upper end of the range.

Dave

[dje]

I have, that's why I have written it the way I did. If your flash is outputting light for 1ms and your camera shutter is set for 1 ms (or less), then you will effectively have a continuous, but non-linear, light source, albeit quite a short one, based on what the "sensor" sees.

If I get some time over the next few days, I plan to do some testing with both an SB-600 and and SB-900 (and possibly with some other lights as well). I think I understand in principle what is being done, but I'm not completely convinced it will work as the light burst must start just as (or before) the first blade of the shutter starts its travel and must continue throughout the period of when the second shutter blade travels (or beyond). If the timing of the flash is not right, this technique cannot work.

Manfred I would certainly be interested to see your results.

Dave

[george013]

I have, that's why I have written it the way I did. If your flash is outputting light for 1ms and your camera shutter is set for 1 ms (or less), then you will effectively have a continuous, but non-linear, light source, albeit quite a short one, based on what the "sensor" sees.

With a sync speed of 1/250 the curtain speed is 1/250. A shutter speed of 1/1000 is created by closing the secund curtain when the first curtain has passed 1/4 of its traject. The higher your shutter speed, the smaller that chink of light. But everything within the 1/250 time.
So for what this guy is doing, you need a flash duration time of 1/250 or 4ms, in this example. He is using a SB900 with a flash duration of T0.5 being 1/1000 or T0.1 being 1/300 or something like that.

I still don't believe him.

George

[Manfred M]

Okay - I gave it a quick try.

On the D800 (manual exposure), on camera SB-600 set to manual an pointed in opposite direction of subject (Master mode and full power).

On light stand. SB-900 at full power (Manual - slave mode) directed at subject.

Camera ISO 100, 1/5000th at f/2.8

[george013]

Okay - I gave it a quick try.

On the D800 (manual exposure), on camera SB-600 set to manual an pointed in opposite direction of subject (Master mode and full power).

On light stand. SB-900 at full power (Manual - slave mode) directed at subject.

Camera ISO 100, 1/5000th at f/2.8

Compulsary flash firing it says. Or in Nikon language FP.

I don't think you will get any further with the flash on camera. Until now I didn't succeed with SB700 on camera, flash and camera manual and going pass the sync time.

George

[dje]

Okay - I gave it a quick try.

On the D800 (manual exposure), on camera SB-600 set to manual an pointed in opposite direction of subject (Master mode and full power).

On light stand. SB-900 at full power (Manual - slave mode) directed at subject.

Camera ISO 100, 1/5000th at f/2.8

Thanks for posting Manfred. Did you by any chance test it at shutter speeds of say 1/1000 or 1/2000 ?

Also I assume your camera sync speed setting would have been one of those that include Auto FP ? I think with these settings, HSS would be automatically enabled in your hot shoe flash for shutter speeds above the max sync speed.

Dave

[Manfred M]

Compulsary flash firing it says. Or in Nikon language FP.

I get that whenever I have a Speedlight go off, George, not just FP / HSS mode. And yes, I just confirmed that by running tests. It gives that message when the unit is set on manual. If I start using the automated functions, then I will also get messages regarding "Strobe return light detected", etc.. If I put a non-Speedlight on the camera (I used a Godox Wistro), and because it is a "dumb" flash, the camera doesn't see it at all.

[Manfred M]

Okay - I have repeated the test using a number of scenarios.

1. Using a Speedlight (SB-600) as the master and a second Speedlight (SB-900) as a slave, both on manual and on full power, with the the camera set to shoot at a shutter speed of 1/5000th. I got results similar to the ones in the video (usable picture). The SB-900 was mounted on a light stand; i.e. this was an off-camera flash scenario, like in the video.

2. I used a hot shoe mounted Speedlight on the camera hot shoe, set to manual and at full power. I set the camera with a shutter speed of 1/5000th and got a useable picture.

3. I used a hot shoe mounted Speedlight on the camera, but turned the Speedlight to i-TTL mode and the camera to shutter priority and set it to 1/5000th sec exposure. Again, I got a usable picture.

In all of these scenarios, the camera and flash were set in ways that the communications channels were operating.




I then tried several scenarios where I isolated the camera from the flash so no direct communication was possible. All tests were done on full power on the flashes.

4. I used "dumb" radio triggers; a PocketWizard Plus III on the flash. The flash to camera connection was via synch cord. I used a PocketWizard Plus II on the camera. Shutter speed was set to 1/5000th sec. I did not get a usable image.. I then reduced the shutter speed towards synch speed and went through a range of partially exposed images and when I reached synch speed, I got a properly exposed image. The SB-900 was mounted on a light stand (off camera flash).

5. I used a Godox Wistro (manual flash) and mounted it on the camera hot shoe. Again I started at full power and at 1/5000th sec, and did not get a usable result. As I reduced the shutter speed, I got the same type of results as in the previous test scenario (scenario 4), with a properly exposed image coming out at synch speed (1/250th sec).

6. I used the Godox Wistro in the same scenario as in Scenario 4 (off camera flash) using radio triggers. Again at full power and 1/5000th sec, I did not get a usable image. I got partial exposures as I approached the camera's synch speed and a useable on at synch speed (1/250th sec).


Conclusion - I think it is likely that the camera software automatically switched to FP / HSS mode and the scenario identified in the video is due to this feature; rather than being a hyper-synch effect. For hyper-sych to work, one needs the appropriate radio triggers; like the PocketWizard miniTT1 or the FlexTT5.

If I get a chance over the next few days, I will see about testing hyper-synch, as I have the relevent equipment (miniTT1 and FlexTT5 as well as the Speedlights). This equipment does communicate with the camera and the Speedlights.

[dje]

Okay - I have repeated the test using a number of scenarios.

1. Using a Speedlight (SB-600) as the master and a second Speedlight (SB-900) as a slave, both on manual and on full power, with the the camera set to shoot at a shutter speed of 1/5000th. I got results similar to the ones in the video (usable picture). The SB-900 was mounted on a light stand; i.e. this was an off-camera flash scenario, like in the video.

2. I used a hot shoe mounted Speedlight on the camera hot shoe, set to manual and at full power. I set the camera with a shutter speed of 1/5000th and got a useable picture.

3. I used a hot shoe mounted Speedlight on the camera, but turned the Speedlight to i-TTL mode and the camera to shutter priority and set it to 1/5000th sec exposure. Again, I got a usable picture.

In all of these scenarios, the camera and flash were set in ways that the communications channels were operating.




I then tried several scenarios where I isolated the camera from the flash so no direct communication was possible. All tests were done on full power on the flashes.

4. I used "dumb" radio triggers; a PocketWizard Plus III on the flash. The flash to camera connection was via synch cord. I used a PocketWizard Plus II on the camera. Shutter speed was set to 1/5000th sec. I did not get a usable image.. I then reduced the shutter speed towards synch speed and went through a range of partially exposed images and when I reached synch speed, I got a properly exposed image. The SB-900 was mounted on a light stand (off camera flash).

5. I used a Godox Wistro (manual flash) and mounted it on the camera hot shoe. Again I started at full power and at 1/5000th sec, and did not get a usable result. As I reduced the shutter speed, I got the same type of results as in the previous test scenario (scenario 4), with a properly exposed image coming out at synch speed (1/250th sec).

6. I used the Godox Wistro in the same scenario as in Scenario 4 (off camera flash) using radio triggers. Again at full power and 1/5000th sec, I did not get a usable image. I got partial exposures as I approached the camera's synch speed and a useable on at synch speed (1/250th sec).


Conclusion - I think it is likely that the camera software automatically switched to FP / HSS mode and the scenario identified in the video is due to this feature; rather than being a hyper-synch effect. For hyper-sych to work, one needs the appropriate radio triggers; like the PocketWizard miniTT1 or the FlexTT5.

If I get a chance over the next few days, I will see about testing hyper-synch, as I have the relevent equipment (miniTT1 and FlexTT5 as well as the Speedlights). This equipment does communicate with the camera and the Speedlights.

Manfred thanks for the update, I agree with your conclusion.

The reason I asked whether you tested the first scenario at lower speeds such as 1/1000 was that I thought it might give some insight into what was happening with the slave flash ie whether it was in HSS mode or not. If it were not in HSS mode, you might get to a slower shutter speed where the flash duration was not long enough to get proper exposure.

Dave

[Manfred M]

With a sync speed of 1/250 the curtain speed is 1/250. A shutter speed of 1/1000 is created by closing the secund curtain when the first curtain has passed 1/4 of its traject. The higher your shutter speed, the smaller that chink of light. But everything within the 1/250 time.
So for what this guy is doing, you need a flash duration time of 1/250 or 4ms, in this example. He is using a SB900 with a flash duration of T0.5 being 1/1000 or T0.1 being 1/300 or something like that.

I still don't believe him.

George

George - that was my going in thought as well. And your description as to how the shutter works is the same as mine.

Once you drop below synch speed, the shutter speed is fixed, but the amount of sensor that gets exposed is reduced. Assuming 1/250th sec synch speed, at 1/500th, the second blade travel starts as soon as the first blade has uncovered 1/2 the sensor. At 1/1000th, the second blade starts when the first blade is 1/4 way through the travel, etc.

The only way that hyper-synch can work in principle is if the light source is emitting for at least 4ms and there would be some level of fall off on either side of t0.5. PocketWizard's paper on the subject does warn of uneven lighting.

[Yucel]

Ever since seeing profoto b1's demoed.. in a studio... I've been working w hss and speedlights...reading tutorials... experimenting.


Can't seem to upload photos... you can see samples at my website yucelphoto.com, they are in the headshot section.

I'm using RT 600s and Yongnuo 600 in radio mode...

[george013]

George - that was my going in thought as well. And your description as to how the shutter works is the same as mine.

Once you drop below synch speed, the shutter speed is fixed, but the amount of sensor that gets exposed is reduced. Assuming 1/250th sec synch speed, at 1/500th, the second blade travel starts as soon as the first blade has uncovered 1/2 the sensor. At 1/1000th, the second blade starts when the first blade is 1/4 way through the travel, etc.

The only way that hyper-synch can work in principle is if the light source is emitting for at least 4ms and there would be some level of fall off on either side of t0.5. PocketWizard's paper on the subject does warn of uneven lighting.

That leaves me with the question what is that guy on the video really doing using a SB600 and a SB900? The speedlights are going automatic over on FP. And pay attention to the fact that that part of his picture that makes use of the flash is only in the middle. Some fall off wouldn't be visible in his picture.
And with that other question from that other thread. The influence of the 2nd curtain was completely different between Nikon and non-Nikon. I can't find it back, maybe David still knows it.

George

[Manfred M]

I did a bit more testing today. No matter what I did other than using "dumb" radio triggers, the Nikon camera (D800) and Speedlights (SB-600 master and SB-900 slave) always recognized the fact that I was trying to shoot faster than the synch speed. I can only assume that the switch over to FP / HSS was automatic. So, based on my experimentation, I'll have to confirm what I wrote before. The guy in the video is incorrect in suggesting he was working around the synch speed issues in a new and interesting way. Rather, he was using FP / HSS and did not realize that he was.

That brings me to the second question regarding Hyper-synch, as outlined in the links that Kathy posted.

My first attempt was to use this technique with an SB-900 triggered by a Pocket Wizard FlexTT5 on my camera. I set the flash to manual and set full power. I was able to use Hyper-Synch to up to 1/400th before I started getting artifacts from the shutter blades, and this got progressively worse as the shutter speed increased. I would say that the images were marginally usable up to a shutter speed of 1/1000th.

I then switched over to the Godox Wistro 360 at full power; nominally over 300 W-s. As mentioned before this is a "dumb flash" with no communications between it and the camera. I've posted the frames and EXIF data below. With this unit I was able to get shots with the entire frame up to 1/1250th. At 1/1600th through to 1/8000th, I was getting a tiny bit of clipping at the edge of the image, as shown in the second image.

NOTE: The exposures are off. I was looking at the limitations of the equipment / technique.


Fastest shutter speed without clipping (Godox Wistro 360)




Similar results from 1/1600th through 1/8000th on the Godox Wistro 360




Conclusion. Hyper-Synch does work, but there are limitations that vary by camera model and by flash used. So, the guy in the video is right in terms of this working in principle, it's just that he was not actually doing it.

[dje]

I did a bit more testing today. No matter what I did other than using "dumb" radio triggers, the Nikon camera (D800) and Speedlights (SB-600 master and SB-900 slave) always recognized the fact that I was trying to shoot faster than the synch speed. I can only assume that the switch over to FP / HSS was automatic. So, based on my experimentation, I'll have to confirm what I wrote before. The guy in the video is incorrect in suggesting he was working around the synch speed issues in a new and interesting way. Rather, he was using FP / HSS and did not realize that he was.

Manfred I did some quick tests myself this morning after realizing that even though I only had one (SB-500) Speedlight, I could do a similar test using the pop-up camera flash on my D610 as the Master unit (mode set to "__" which means it signals to the remote but doesn't fire it's main flash). The SB-500 was set up as the remote flash.

This allowed me to use any shutter speed I wanted and get uniform exposure across the frame every time. I therefore come to the same conclusion as you, viz the remote flash is automatically going into HSS/FP mode once the sync speed is exceeded. In other words, the remote SB-500 is behaving in exactly the same way it would if it were mounted on the hot shoe of the camera.

I will be in a position to borrow an oscilloscope in a week or so and will capture some flash waveforms then. I'll also see if I can capture the signaling pre-flash pulses from the master to the remote.

Dave

[george013]

I did a bit more testing today. No matter what I did other than using "dumb" radio triggers, the Nikon camera (D800) and Speedlights (SB-600 master and SB-900 slave) always recognized the fact that I was trying to shoot faster than the synch speed. I can only assume that the switch over to FP / HSS was automatic. So, based on my experimentation, I'll have to confirm what I wrote before. The guy in the video is incorrect in suggesting he was working around the synch speed issues in a new and interesting way. Rather, he was using FP / HSS and did not realize that he was.

That brings me to the second question regarding Hyper-synch, as outlined in the links that Kathy posted.

My first attempt was to use this technique with an SB-900 triggered by a Pocket Wizard FlexTT5 on my camera. I set the flash to manual and set full power. I was able to use Hyper-Synch to up to 1/400th before I started getting artifacts from the shutter blades, and this got progressively worse as the shutter speed increased. I would say that the images were marginally usable up to a shutter speed of 1/1000th.

I then switched over to the Godox Wistro 360 at full power; nominally over 300 W-s. As mentioned before this is a "dumb flash" with no communications between it and the camera. I've posted the frames and EXIF data below. With this unit I was able to get shots with the entire frame up to 1/1250th. At 1/1600th through to 1/8000th, I was getting a tiny bit of clipping at the edge of the image, as shown in the second image.

NOTE: The exposures are off. I was looking at the limitations of the equipment / technique.


Fastest shutter speed without clipping (Godox Wistro 360)




Similar results from 1/1600th through 1/8000th on the Godox Wistro 360




Conclusion. Hyper-Synch does work, but there are limitations that vary by camera model and by flash used. So, the guy in the video is right in terms of this working in principle, it's just that he was not actually doing it.

The exif still states "compulsery flash". I can't find a good explanation for that phrase but to me it's FP flash.
If so, you're looking for the limits of FP, not of manual. Lower the shutterspeed and find out where that "compulsery flash" comes up.

George

[pnodrog]

The exif still states "compulsery flash". I can't find a good explanation for that phrase but to me it's FP flash.
If so, you're looking for the limits of FP, not of manual. Lower the shutterspeed and find out where that "compulsery flash" comes up.

George

I think compulsory flash just indicates that the flash has been set to fire by the photographer rather than displaying the operating mode. Easily checked by setting to rear curtain sync and see if or where the EXIF wording changes.