The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

Hassy H4D/40 Release

David K

Workshop Member
I also cannot wait to try it with the 100/2.2 which I off-center focus wide open with a lot ... usually really close.
I would think that's the acid test for many of us, it sure would be for me. From my non-technical perspective I would have thought that the ability to put the focus point anywhere in the frame would have been an easier problem to solve, but apparently not.
 

Guy Mancuso

Administrator, Instructor
So just reading some comments it really only works say if you are holding the camera straight the sensor pivots left and right and on vertical up and down BUT No vertical adjustment when held in a landscape mode. So if I had a subject say man in front of me focused on his eye's than tilted camera up for building it really would not work in this case. It is more the pivot going right and left.

Do I have that correctly in my thoughts?

Also Marc and David that illustration really shows the wide angle effects now compared to a longer lens. I was trying to wrap my head around it at first now I see it is more effective with a wide angle. So when testing this out maybe try a 28mm, 80mm and like a 150mm to see how that works in practice. Very interesting concept for sure
 

gogopix

Subscriber
At first this didn't seem much different from autofocus lock (that even dead systems like Contax have! :)

BUT, it now appears TF will adjust the focal point for a point (eyes) by using one or two pieces of info;
the angle of tilt; this is enough if the subject and camera do not change planes (it would adjust a 45degree tilt by 1.4 setting the lens 40% farther. Actually, this would ONLY work if you moved camera to have BB in focus.

Using tilt angle and the new focus point (belly button! :) it would add 40% to the NOW focus distance.

So which is it? one or both?

or is there a third method? (I assume the purpose is nothing more than adjusting for the tilt. Since Phase already have horizon, it seems a similar inclinometer could be used and TF could be done in the Phase/Mamyia set as well.)

Hard to see how this is an advantage over AF lock.

Victor
 
@Guy - Imagine your H4D on a tripod, ball head. What movement do have on the camera? 360° Rotational - yes? So your scenario you mention would work.

I said.. As long as it is rotational it doesn't matter what plane it is in. i.e Landscape/Portrait... anything in between!

@David K - Correct. You will probably find the tilt sensor in the Phase/Mamiya isn't accurate enough, as was the sensor in the H3D.

Hence a new kind of sensor in the H4D.
 

David Klepacki

New member
David G,
Thank you very much for posting the information from Hasselblad. I do think TF is innovative and will work very well for situations within its constraints. I also appreciate Hasseblad's full disclosure here when pointing out such a limitation when the DOF is shallow, " ... A camera movement closer or further away from the camera even as small as 1 cm will change the result and True Focus might not fully correct the focus...", which will most likely occur with telephoto lenses at close range.

I think this technology will raise the bar in the industry regarding focus accuracy, which benefits everyone. Furthermore, it will force higher accuracy of the base autofocus for each and every lens, since any slight error in the initial autofocus accuracy would then translate into errors for the TF adjustments as well. Actually, it is this need and emphasis of base autofocus accuracy that has my attention more than anything else.
 

Guy Mancuso

Administrator, Instructor
@Guy - Imagine your H4D on a tripod, ball head. What movement do have on the camera? 360° Rotational - yes? So your scenario you mention would work.

I said.. As long as it is rotational it doesn't matter what plane it is in. i.e Landscape/Portrait... anything in between!

@David K - Correct. You will probably find the tilt sensor in the Phase/Mamiya isn't accurate enough, as was the sensor in the H3D.

Hence a new kind of sensor in the H4D.
Thanks David that helped. Sounds pretty cool
 
Hi David (A lot of Davids!),

Yes, base focus accuracy is very important. We have been working on this for years, with the sensors, processor, processor speed, mathematics...

Tiny incremental changes every year. Also in lens design we can do other things to improve AF performance.

Additionally the H4D has a bright white LED AF assist which really helps in lower light.
 

gogopix

Subscriber
David K, (lots of David's here!)
I understand AF lock doesnt compensate. I was just saying I couldn't think of the situation where it wasn't what I wanted. Also sounds like you HAVE had the situation. With degrees in both math and physics i can see mechanically what they are doing - what i don't understand is WHY. Sounds like the image setup, aperture, speed, DOF distance etc need to be right on the hairy edge. Can somebody show an example with, without?
Seems as Guy says pretty cool, so are margaritas; but I drink red wine! :)

David G,
Understand the tilt for horizon may not be accurate enough.

With this technology, maybe this is just the first step on making some real improvements in basic photography. Ultimately, with a
reference image and a holographic transform you won't need to focus (except roughly) at all. True! (almost) UCLA is working on I think.

Victor
 

Nick-T

New member
T
I also cannot wait to try it with the 100/2.2 which I off-center focus wide open with a lot ... usually really close.
Marc
This is exactly how I have been testing the system, ie with the 100 @2.2 focussed at around 0.9m (closest focus sans extn tube). What I found was an improvement handheld that was slightly offset by ME moving inconsistently. On a tripod the effect (of True Focus) was dramatic and repeatable.

And to re-iterate what D.Grover has said, think of the Tilt/yaw sensor like a Nintendo Wii. In Orlando the product manager for the H4D was hand holding the camera and twisting it this way and that while on screen a 3D model mimicked exactly his movements.
Nick-T
 

gogopix

Subscriber
So it can correct in the two dimensions yaw and pitch (tilt)?

That's pretty impressive. But soundslike real value is wide swings at close focus.. yes?

Victor
 

David Klepacki

New member
Hi David (A lot of Davids!),

Yes, base focus accuracy is very important. We have been working on this for years, with the sensors, processor, processor speed, mathematics...

Tiny incremental changes every year. Also in lens design we can do other things to improve AF performance.

Additionally the H4D has a bright white LED AF assist which really helps in lower light.
Can you explain better what Hasselblad does to ensure consistent focus accuracy for each and every lens? For example, Leica's approach is to quality check each individual lens at their factory in Solms. However, even with these careful quality checks, we have often found slight front or back focusing with such brand new lenses. So, maybe Hasselblad has found a more efficient and cost-effective way to ensure the accuracy of all its lenses (as compared to Leica)?

Also, Rodenstock and Schneider seem to distinguish their digital lenses from their film lenses. Their digital lenses supposedly take into account the IR glass of the sensor. Since the HC lenses can be used with both film and digital, it would seem that the H3D and H4D compensate for this IR glass within the camera body to achieve the same compensation for digital use, is this correct? This is probably a more practical approach than designing separate lens lines.

Again, I do appreciate all the technical information that you (and Hasselblad in general) provide. It really makes me wonder now how often Alpa shooters achieve optimal plane of focus with their wide angle lenses, if they only rely on scale focusing.
 

David Klepacki

New member
Thanks Nick. That is what I expected.

In the end, the entire issue of focusing accuracy comes down to the consistent production of lenses with dead on focusing ability that neither are back-focused nor front-focused and in which every lens element is centered. Otherwise, the entire focusing chain is thrown off, as the True Focus mechanism will just propagate any such errors in the lens itself.

Leica seems to have to go to a lot of trouble to ensure their lens consistency. I am still curious as to how Hasselblad achieves the same thing, and without the same costly effort as Leica.
 

fotografz

Well-known member
Thanks Nick. That is what I expected.

In the end, the entire issue of focusing accuracy comes down to the consistent production of lenses with dead on focusing ability that neither are back-focused nor front-focused and in which every lens element is centered. Otherwise, the entire focusing chain is thrown off, as the True Focus mechanism will just propagate any such errors in the lens itself.

Leica seems to have to go to a lot of trouble to ensure their lens consistency. I am still curious as to how Hasselblad achieves the same thing, and without the same costly effort as Leica.
Hasselblad aside ... I'll let people more expert in tech minutia answer that ... (I'll just assume the long implemented "Micro Adjust" technology works in concert with True Focus until told differently).

My question is why does it seem that so many Leica lenses are delivered that need to be sent back due to focusing issues? That "costly effort" doesn't appear to correlate directly with actual results. I'll forego the litany of Leica lens experiences ... suffice it to say that the theory of high production costs doesn't necessarily mean anything.

-Marc
 
Top