Author Topic: Hasselblad True Exposure tests  (Read 576 times)

Offline Buddy

  • Full Member
  • Senior Member
  • **
  • Posts: 170
  • Karma: +3/-1
Hasselblad True Exposure tests
« on: December 30, 2017, 07:36:26 AM »
I have run a test for True Exposure with X1D and 45mm XCD lens which x-sync's up till 1/2000 sec as you know, 2 series of images with same aperture/ISO settings (f16 and ISO 100) but varying shutter speed from 1/125 sec up to 1/2000sec, one set with True Exposure On and one with TE = Off.

I think I understand the concept of TE pretty well, that the preset shutter speed is altered (fastened) with TE until the fastest speed cannot be fastened anymore, in which case the aperture is being further changed (closed a bit) to avoid overexposure by the longer effective shutter speed of leaf shutter lenses without TE. 

The tests were run in ambient light combined with a studio flash on a grey card target, all images with same power setting and similar ambient light level. I would have expected that the illumination on the targets (in foreground, lit by the flash) would be equal/similar until 1/2000 shutter, but this seems not to be the case. In fact, my test series show that the changes on the aperture already kicks in at a shutter of 1/1000sec or eventually 1/800sec (don't know this). The target on this image and all images later in the series (1/1500, 1/2000) are reduced in illumination, indicating that the effective aperture was changed. But such changes of shutter speed or aperture are not displayed in the metadata in Phocus (didn't check with PS).

RGB values on the same target areas for image 1 to 5 read as follows:

shutter     RGB values

1/250       238/245/249
1/500       237/245/249
1/1000     198/209/215
1/1500     198/209/215
1/2000     195/206/213

In Hasselblad's explanatory note about TE the cut-off is described at 1/800sec shutter, but this was written before the new lenses with the faster speeds were available.

Can someone confirm/knows this?

I can imagine that this might be in connection with the mechanical shutter speeds up till 1/800 or 1/1000 sec (?) and some secret electronics’ means for shutter speeds of 1/1500 and 1/2000 sec for the new lenses for HCD with the orange dot and all XCD lenses (however, not sure about this technology on the new lenses….read somewhere about that).

I try to figure this out in order to determine under what conditions one might set True Exposure on or off when shooting in ambient light combined with flash for foreground (otherwise, i.e. when no flash is involved, I believe True Exposure should always be set to on for consistent exposures and especially for smaller apertures).


Thanks


Offline NickT

  • Administrator
  • Senior Member
  • *****
  • Posts: 1784
  • Karma: +11/-0
Re: Hasselblad True Exposure tests
« Reply #1 on: December 30, 2017, 03:04:54 PM »
I wonder if (part?) of what you are seeing is due to flash duration?
Nick-T typing at you from Flexframe's secret location under a Volcano

www.nick-t.com

Offline Buddy

  • Full Member
  • Senior Member
  • **
  • Posts: 170
  • Karma: +3/-1
Re: Hasselblad True Exposure tests
« Reply #2 on: December 31, 2017, 02:03:28 AM »
No, flash duration was constant and for each shot t 0.1 = 1/3150 second. In the series of images with TE = Off the RGB values remain constant around 245/250/255. It is definitely the compensatory effect of True Exposure to shorten the exposure time slightly at small apertures values at the shutter and from a certain shutter setting on the aperture. The images were shot at f/16, so TE kicks in here. Based on H's own measurement there is a +0.8 EV difference in the effective shutter between f2.8 and f11 for a given shutter speed of 1/800 sec (=1.25 ms)

TE does this compensation. But I was thinking that TE would do that "transition" on the new lenses at 1/2000 shutter and not at 1/800 or 1/1000 as on the old lenses.

But I am pretty sure now: it is stated as such with a “transition point” at shutter = 1/800 sec also in both the new manuals for H6D-100c and X1D (all XCD lenses go up to 1/2000 sec)

True Exposure is basically a way to compensate longer than necessary effective exposure times despite the selected shutter setting when a small aperture is in play on a leaf shutter lens and thus will shorten such effective exposure time a bit (up to 1 stop).

This is due to the fact that at smaller apertures (e.g. f11) in combination with rather fast shutter settings (e.g. 1/800sec) the sensor will be exposed over-proportionally longer because of the travel time of the shutter blades when opening and closing in relation to the aperture blades opening (travel time in addition to shutter speed).

Why is this important to know? Because in some shooting scenarios one must know whether to set True Exposure to On or Off. Generally, one can say to always enable True Exposure for correct exposure timing. However, if one is using manual exposure mode without TE, he would probably simply further adjust the shutter and/or aperture setting on the next shot, after having seen the image just taken with a wrong exposure or via the histogram (or adjust in post).

But when shooting with a flash system (with or without available light; e.g. outdoor or studio) and with rather small apertures and faster shutter speeds, then TE definitely must be disabled. Otherwise it will happen (for faster shutter speeds than 1/800 sec) that TE will do the adjustment not on the shutter speed but on the aperture. And that is definitely unwanted in flash photography, since the flash exposure controls the foreground and this only by the aperture. So this would lead to an underexposed image by up to 1 stop.
 
I had this situation already after measuring the scenery with a incident light meter and TE was enabled in available light and with flash for the foreground (the model) at fast shutter speed and rather small aperture settings, and couldn’t figure out, why the image taken with these settings was underexposed by as much as 1 stop against the calibrated Sekonic L-758D measurements…. 

Offline jerome_m

  • Full Member
  • Senior Member
  • **
  • Posts: 322
  • Karma: +5/-0
Re: Hasselblad True Exposure tests
« Reply #3 on: December 31, 2017, 06:29:25 PM »
The shutter on the new lenses does not really go up to 1/2000 mechanically, it tops at 1/1000s. To reach faster speeds, the sensor is reset after the shutter blades are full open. That effectively doubles the shutter speed. That may explain the effects you are measuring.

Offline Buddy

  • Full Member
  • Senior Member
  • **
  • Posts: 170
  • Karma: +3/-1
Re: Hasselblad True Exposure tests
« Reply #4 on: December 31, 2017, 09:46:14 PM »
yes, I thought that as well, mechanical up till 1/1000 and then electronical cut-off.

Actually I was reading one of the Hasselblad published articles about lens design, which was written already a couple years ago when only the 1/800 lenses were available. In this report H stated, that it would be easy for them to produce the same lenses with 1/1000 shutter already then, since the measured deviations of the performance criteria were already that time within the narrower tolerances given by H versus DIN. In 2016 obviously they did this with the new orange dotted lenses. And with the additional 1/1000 they were beating Phase One's Schneider lenses with max. 1/1600. Smart. The consequently introduced sharp price increases on the lenses later on was less of a welcome though.

The manual of X1D and also H6D describes the transition for enabling or disabling True Exposure at 1/800 or 1/1000, depending on the lens. At these shutters TE will already kick in as my test from above revealed, it was shot with XCD 45mm and orange dotted 100mm f2.2 with XH adapter on the X1D and shows already a drop in exposure (by closing the aperture instead of increasing shutter speed) for the 1/1000 shutter setting.

Offline ChrisL

  • Senior Member
  • ***
  • Posts: 88
  • Karma: +0/-1
Re: Hasselblad True Exposure tests
« Reply #5 on: December 31, 2017, 11:14:19 PM »
Surely the cut off must be before the maximum mechanical speed is reached as the principle is to increase the speed, if already at maximum there is nowhere to g. So on the original shutter/RGB values table 1/500th is set and, given Hasselblad quote the f stop change seen elsewhere, you could in principle calculate the actual shutter speed used to give that "correct" exposure. I see where Hasselblad quote the cut off at 1/800th or 1/1000 meaning it no longer works at those speeds but it is easy to read that as if setting 1/800th is fine that is the fastest you can use which is what most would understand by a cut off.
If you have a rev limiter on a car with a cut off at 8,000 rpm that means you can run to 8,000 not that the limit is actually 7,500.

Offline Buddy

  • Full Member
  • Senior Member
  • **
  • Posts: 170
  • Karma: +3/-1
Re: Hasselblad True Exposure tests
« Reply #6 on: January 01, 2018, 02:09:45 AM »
fully agreed.