DOF Comparison

[Dennis Wood]

Cool, Andy, your challenge now is to make it work.

[Andy Gordon]

You obviously don't want to change the diffuser Dennis, too bad eh?

Here's some more apples, Brevis compared to Brevis plus a piece of poc20 stuck onto the diffuser. This is the full bifta diffusion. I even stopped the poc20 down to give the brevis a head start in the dof haha.

http://www.savefile.com/files/5768677

The apparent dof is much shallower. Light loss is obviously worse. Probably a couple of stops. Probably no good unless you have daylight or properly lit shots. The grain is visible under certain conditions, it's a bit like an optosigma 1500 grit static grain. Not sure you'd see it on a TV or if the audience would notice it.

I actually like both of the looks, there are pros and cons. I think the brevis + poc20 is more film-like (let's see that comparison!). In poor lighting or if you want deeper dof you'd want to go with the Brevis.

Dennis if you were willing to do it, I'd buy a diffuser made of poc20 plus the same fresnel as on the brevis diffuser. The light loss would be a bit better than my current set up, and bokeh would be about the same. How about it?

[Dennis Wood]

Andy, I have no problem with changing the diffusor at all. What makes sense is to use the same diffussor, but allow the user to choose their preference. This is exactly the same thing as my spinner setup with a lower and higher diffusion set of discs that can be swapped.

That way, I could use the existing standard diffussor, with the user deciding if they want the extra diffussion (with the accompanying softer image, and light loss), by attaching a supplementary plate, similar to the POC. Keep in mind too, that HD compatability is very much an issue here. Dan D's (MPIC) goal was also HD compatability to film out...and he's a very, very experienced soul in the world of film.

The basic issue with engineered diffusors is that they do it with random structure, with larger structures giving less diffusion. Now larger structures give more grain, which in turn, can be much more difficult to hide. In the case of POC20..the structure is very large. The POC50 has super small structure (no grain) but is 4 f/stops of loss. Now if a shooter is very aware of the light/softness hit of a higher diffusion setup, a dual diffusor setup makes very good sense. Now I could revert to mechanical diffusors, but then the halo (which I can't stand) shows up around specular highlights. I did some tests with coated spinning imaging elements..and could not escape the halo. You'll see the same halo with M2/M3 footage.

Not to worry, the film test should be done this week sometime...I have some catching up to do.

BTW, got some feedback from a non beta HVX user...he loves the adapter.

[Andy Gordon]

I've been playing around with it a bit more and the light loss/softness/grain is too much. So I've gone back to straight Brevis haha. I'm actually really starting to like it... I can fully appreciate the choices you've made now that I've satisfied my own curiosity. So apologies for kicking up such a fuss over the bokeh.

I think when you stop down the lens the difference between high diffusion/low diffusion becomes a lot less significant.

Anyway apart from the niggly difficult to align lens mount, I think it's an awesome adapter. It's pretty amazing what you've achieved in the space of about six months. Look forward to the film comparison.

Cheers
Andy

[Dennis Wood]

You have no idea how glad I am that you've come to this conclusion The lens mount issue is already resolved.

BTW, I had measured the Nikon flange distance at 13mm above the Brevis outside ring. Is this working for everyone?

[DJ Lewis]

.... The lens mount issue is already resolved.

What changes have you made to the mount?

[Dennis Wood]

just changed the config so the set screws don't fall into a ridge area...easy fix.

[Moises Crespo]

Speaking about screws my 72mm step ring does not fit on to the Brevis, so I have to use the 58 then the 72 and the 4 screws dont tighten when I use the 72 step ring..

[Dennis Wood]

Moises, I'm not getting this. Is the step ring one that I provided or your own? If so, a couple of pics via email would help. Maybe I can figure something out, or send you one of the rings I've tested. The 58mm mount (4 screws attached to the brevis) should be tightened before you put the 72 to 58 stepdown ring on.

I'm working on a way to omit the 58mm mount altogether and go with a "native" 72mm mount. 90% of current customers/prospects have cams with 72 or larger threads.

[Steve Strickland]

A native 72mm mount would be awesome. How about different mount sizes that are specified upon order?

[Dennis Wood]

Steve, different mount sizes are too hard to implement in design inexpensively. It is completely do-able, but would add a fair bit to the cost of the unit. Step up/down rings are very inexpensive, and readily available/replaceable in case of stripped threads.

I looked at actually threading the Brevis directly, but damaging those threads would render the adapter useless. Having a replaceable threaded mount solves the rotation adjustment issue, and makes the threads non-critical.

[Moises Crespo]

It's my own 72mm ring. With the 58mm ring and the 4srews I cant tighting the 72 to the 58. The screws don't let the threads connect.If you could send me one that would be OK.

[DJ Lewis]

Search for seller "heavystar" on ebay (13000+ rating)

I ordered some step downs rings from him and they work perfectly.

http://cgi.ebay.com/New-72-58mm-Metal-Step-down-Ring-72mm-58mm-72-58_W0QQitemZ7616433704QQcategoryZ30040QQcmdZViewItem

[Moises Crespo]

Thanks I order one..

[Rafael Lopes]

Here's a quick sample you can freeze frame to see how much deeper the Brevis DOF is than it should be. The difference is huge. My Rode videomic was on the cam as well. Noise is pretty obvious.

http://www.savefile.com/files/1897965

Hope you can see the dramatic difference in bokeh. Much more film like if you ask me with the POC20. The Brevis is nice and sharp, but the DOF is way too deep for an f1.4 lens...

I've nearly got my own adapter working, just need to fix the oscillation so it's circular, the grain is pretty obvious at the mo.

Light loss is the same since you have to stop down the Brevis to f4 to remove the haze.

I'll post some daylight comparisons at the weekend if anyone's interested.

Totaly agree. The POC20 is the way to go by far. The footage without the POC20 looks like it was digitaly blured with 2 layers. Too bad about the light loss and grain. I shoot HD and I cannot have it. There's gotta be a way to achieve that look without such a sacrifice.

[Andy Gordon]

Not sure you can avoid the light loss but the grain can be solved. I'm hoping the 350grit gg I've ordered will do the trick, without too much light loss.

Have a look half way down the page on this site:

http://www.digitalartform.com/lenses.htm

To me the Brevis at open is doing more gaussian blur than bokeh.

[Rafael Lopes]

That is exactly my thought. Right now the Brevis has what? Half a stop light loss? If it was 1 stop or maybe even one stop and a half with a ncie bokeh it would be acceptable. Both the G35 and the M2 have 2 versions that tackle this
very same issue. You may choose which one is the best for your needs. In the case of the M2 they made a system that you can switch between both options. This would be an exelent idea.

[Dennis Wood]

Well here's a look at 100 ASA film vs the Brevis35. As far as I know, no one has done a test like this with their adapter. My SLR choked on it's max of 1/1000s shutter speed (required for exposure at F2), so none of the film shots at F2 worked. The same 50mm lens used for the film shots was used on the Brevis shots.

Note that the octagonal specular highlight (film shot at F4, bokeh specullar highlight) is something I've yet to see from any adapter at low f/stops. All I've seen from other adapter is hazy mush with a halo that grows as you rack focus. If there's an example out there (really shot with an adapter) I'd like to see it. I tried many spinning elements, including AR coated ones, to try and get that typical bokeh spec highlight look...to no avail. One of the GS400 guys in England has an M3, and demonstrated the same thing..mush with a halo. what I like about the Brevis version is that there is halo. At +F8 the look is very, very close.

What's very interesting is the increase (compression really) in dynamic range using the adapter! The pics were shot on 100 ASA film, then scanned at 400dpi, then resized to ~740x480. The GS400 Brevis shots show a lot more range...something noted by Dan D. when using an adapter. The shots were all taken within about 10 minutes of each other, with a clear sky, so lighting was very consistent.

www.cinevate.com/images/bokeh100asa.jpg

[Andy Gordon]

The Brevis compares very favourably in those shots... looked like you were on minimum focus distance for the lens. Where I see more dramatic influence of the diffusion is in the mid range of the focus. Looking at the two glasses, would you agree the apparent dof is shallower on the slr shots? IMO when you're down at f8 the dof is so deep it's hardly worth using an adapter.

Regarding an adapter showing solid specular highlights at low f number, I thought that's exactly what I showed in my comparison. But I agree I've never seen it on any other adapter at low f stop apart from my stuff haha.

Does high shutter on the slr reduce the dynamic range? maybe you need to put an ND on there?

Thanks for doing the comparison.

[Dennis Wood]

Andy, no question that as the F/stop drops below F4, the level of apparent diffusion diverges. I had an ND.6 on the lens for both samples... I used approx 1/1000s to 1/60s going from F2 to F16. For whatever reason the 1/1000s shots seem to be photographing the SLR shutter itself. As a rule, I never use shutter speeds this high, so I have no idea what's going on there.

I also did a test using 400ASA, and using focus distance of approx 4 feet. Because part of the test was under cloud cover, and part not, combined with using 16:9 mode on the GS400 instead of 4:3, I tossed the test. Too many variables out of whack. What I would say is that the look at F4 and above is very, very close. I can't speak for F2, cuz none of the shots at 1/1000s turned out. For the 400 I had both an ND and CP attached.

I am taking advantage of the optical engineer at my "disposal" to look at the possibility of tweaking some diffusor variables.

Btw, I watched your DVD again. The more I pay attention to your work, the more I'm impressed. It's a keeper for sure.