Lowlight and Sharpness standards

L

Lenilenapi

Carbonite Member
OK everyone is talking about the Letus extreme here and it certainly sounds promising, but a lot of claims are being made about its low-light & sharpness performance that are not always being backed up by careful standards.

RE: Low Light

Showing me a shot of a scene in a house at night illuminated by a 250 W lamp tells me almost nothing at all. A 250W Lowell Pro can be bright as hell especially if its not diffused. How close is it, is it spotted or flooded - all of this can make for many stops difference in performance.

Please - Foot candles is what i want to know and measure the results with a Waveform.

Can I suggest if you are demonstrating low light performance, you tell us

1 - THE FOOT-CANDLES on the scene,
2 - the f-stop of the HVX or other video camera
3 - the shutter speed and frame rate
4 - What is the 35mm lens and its f-stop.
(If you're wide open at f1.4 - check to see how different it is at f2.8)
5. Is it Academy or Vistavision type of framing
6.- How far is the HVX or other video camera zoomed in.

Considerations to keep in mind when judging this are:

1. - Adapters from what I see do not always have a linear response to f-stop increments on the 35mm lens. Some adapters will show little or even no increase in exposure as the 35mm lens is opened from f2.8 to f1.4. - Its weird but its true. This is something to note and it may be different with different cameras, lenses and diffusing screens.

2 - I've found that on the HVX f-stop numbers don't always seem like they produce a linear response as you change exposure in the low vs higher f stops.

3. The f stop number readouts in the viewfinders of the video cameras are very sloppy often by 1/3 of a stop or more.

4. Different Lenses may not always transmit the same amount of light at a given f-stop. Especially primes vs. Zooms.

5. Vistavision vs Academy may make a difference on a given adapter as will how far in the video lens is zoomed.

6. On an IRE chart response isn't linear on most video cameras so you have to careful interpreting waveform changes.

7. The most reliably linear way to vary exposure on most video cameras is probably shutter speed. IF you can use this, do so to judge how much you have changed the exposure. i.e 1/60 tp 1/30 is absolutely a full stop - there are no variables. On the HVX you can vary this very precisely.

Frankly what I want to see the most is a face at say 60 to 70 units both with and without the adapter keeping the footcandles the same.
Second best is a chart or else just a card. KItchen scenes don't tell me enough.



Re: Sharpness

I don't know how anyone can make claims of how sharp one adapter is against another unless you've got them side by side or have shot the same chart under identical conditions. At the very least let us know what all the parameters are.

I suspect that the achromats used in these adapters all have their sweet spots for good corner to center focusing. Its makes the whole business pretty complicated.

Its certainly a combination of:

1. distance from achromat to the screen
2. Zoom focal length on the video lens
3. The size of the image be it Academy or SLR full frame (Vistavision)- (this is related to the zoom length)
3. The Focus distance on the video lens (also a function of 2 & 3)
4. The video camera itself and its lens. This can alter many parts of the equation and its why Dennis has different screens for different video cameras.
5. Of course the f-stops of both video camera and 35mm lens

I wonder if an adapter that has great sharpness at Vistavision can maintain that while zooming into Academy and vice versa. If not, is an adapter that looks great with say Vistavision Nikkors going to look as good with Zeiss Academy cine lenses (and the reverse)?

Could an adapter that's great on one camera not be as sharp or vignette on another?

If you are shooting sharpness tests, if you can, check both Academy and Vistavision and I prefer to see tests with the video camera's f-stop wide open.

RE: Grain & DOF

Oh yeah these are the other performance characteristics I would want to know about the Brevis vs. the Letus. Are the DOF & Bokeh different and importantly what f-stops can you get to before you start seeing grain from the vibrating adapter.

The Brevis CF1L gives you a lot more latitude than the CF3L say but the CF3L is lower DOF and needs more light. Sounds like the Letus is closer to the CF1L in light needs and grain problems, but I wonder how the DOF characterisics differ. Unfortunately you probably need to put them side by side to tell. Plus soon we will have some diffuser announcements from Dennis as well as his Flip system. Hard to keep up with all this. No time to shoot.

BTW - If anybody wants to send me a LetusExtreme by the way I can compare it to a Brevis (with 3 diffusers) and an SGPro here in San Rafael CA.


Lenny Levy
 
Last edited:
Lenny, while I appreciate your requests, I simply don't have time to do such detailed tests and what I am showing is the best I can in the limited time available to me. If the guys from Letus were paying me to demo this unit, that would be a different story.

Perhaps Ryan Walters would have the time to do the kind of test you are interested in although I wouldn't presume on his behalf. He is certainly more scientifically inclined than I am.
 
I realize its a long list (it took along time to even write it) and I don't mean to diss anyone who is taking the time to post images - sorry if it came across that way. Obviously I'm not bothering myself and I appreciate the effort you guys have made.

It's obvious the Letus extreme is a very nice unit, but people seem to be assuming these images prove one adapter is better than another, and I think its important to realize how many variables underlie these shots.

Just some footcandle readings alone would go a long way.

Maybe even a "here is the shot with the adapter and here it is with the naked camera and the difference in shutter speed, f stop or whatever is "x".

- Lenny
 
Last edited:
Seriously? Taormina has the most valid point. You can do res charts all day. Fact is, i'm not shooting resolution charts.

The only low light test I care about is the one with only house lights on. Lamps, overheads, etc. Granted, Disjecta had some Lowells okay. But in the kitchen, it didn't seem like there was much added at all.

At that, sharpness? If it's sharp it's sharp. Again, Resolution Charts aren't my actors, so that's really nothing I'm concerned with; just another way of counting pixels.

I've got an M2 and I can vouch for the softness of the image. I do see that the SGpro is sharper (albeit not that much in my case) and fronm what Disjecta's footage shows, the Letus seems to be even sharper than that. I'm not claiming to know HOW... but that's what my eyes see.

Also gotta say that it might just be the XH-A1 + Letus combo. Which is why we need more HVX footage with the Letus EX...
 
Last edited:
I agree completely with Lenny. I am the firm belief that any adapter can "look" good on compressed web footag when used by a professional, and edited by an experienced hand. Very few folks have compared footage in a clinical comparison, Phil Bloom making the best attempt I've seen, with 3 adapters in hand.

Product recommendations have been made here before, that later turned up issues after further testing. I have recommended a method for testing light loss that doesn't require any equipment, and makes sense to most users. Most users just want really to know how much light they need to add beyond what is being used with the bare camera on an identical scene.

Lenny is correct in adding a few factors that need to considered:

1. Light loss from our adapter increases from CF1 to CF2 and CF3. The relationship between camera aperture and video aperture is not linear. This means that the tested difference between adapters at f5.6 can be drastically different than when the lens is wide open. By drastically, I mean 2-3 stops. Some 35mm lenses are terrible wide open, so users do dial in a few stops, and light changes under these conditions is very important.

2. Light loss increases as a video camera is zoomed in. Therefore field of view is critical as it determines the level of zoom in required.

3. There are various issues contributing to sharpness, the largest being edge softness from 35mm lens, or camera lens used wide open. Testing with the video camera at f4, and the taking lens wide open and at f5.6 makes sense.

Testing a system without a lens attached introduces a lot of variables. For example, an EOS lens mount opening has 50% more area than a Nikon lens opening. How long the is the lens tube opening? How close is it to focussing plane? Was the camera zoomed identically?
 
Last edited:
While I agree with what Dennis and Lenny are saying, there is no substitute for just using equipment in the way that one specifically needs to use it.

I also agree that there are a lot of impressionable people on this board that can be swayed very easily by good looking footage. The problem is, I can't afford to buy 3 more adapters to show their differences. If others want to provide me with a unit for testing side by side, I will do that but until that happens, I will continue to post footage that I think shows the Extreme at its best and I will try to be a little more diligent in providing f-stop info so that it can be gauged more accurately.

Also let it be known that on no occasion have I compared the other adapters unfavorably with the Extreme only to say that what it has is a built-in flip...and at the moment that's one of its unique features.
 
And Disjecta, I think I speak for many here when I say thanks for your efforts. Considering you took it upon yourself to do all these tests, it's kinda chintzy to ask you to do even more.

Someone needs to step up to the plate and do a real test - someone with all the adapters.

Still - I don't care what the results are. They all look pretty darn good. It all boils down to what features you need and what works for you.
 
Agreed and I'm not at all complaining by what is being asked. There are some on this board that are very technically and scientifcally inclined but I am not one of them. I rely on necessary knowledge but mostly my eye. What looks good to me tends to look good to others too. It's not for everyone but it works for me.
 
Okay but still... can you find an HVX and shoot some stuff? =P Just kidding.


No but really.
 
At some point in the near future Barry will have a complete test rig to play with for a week or so. First we need to get a pile of stuff back from Canon :-)​
 
disjecta said:
Perhaps Ryan Walters would have the time to do the kind of test you are interested in although I wouldn't presume on his behalf. He is certainly more scientifically inclined than I am.
Click to expand...

Thanks much appreciated. :)

As for those tests, I would have to get a meter that displays foot candles, as mine does not. (I know, I should have one that does....) I'm also working on arranging some other testing, so that has got to come first.

BTW- Lenny, I just sent you a PM...
 
Ryan, Sorry I guess I just assume everyone has a footcandle meter - left over from the film days.

Actually even without a meter that reads out footcandles, you can figure it out backwords from the old basic rule of thumb :
@ 1/50 sec ( common cine shutter speed @ 24fps) and 100ASA
then 100FC = f2.8
so @ 200 ASA then 100FC = f4

If you set your meter at 1/50th and 200ASA then whatever f-stop it tells you can be translated to footcandles back doing the math from f4.
Thus f2 (means less light )= 25FC etc
Actually you can figure it out from any setting on the light meter if you have that starting point you just have to adjust for all three variables.
( 1/60 @ ASA 200 & f2 = 20FC)

That may be a little arcane and I wouldn't expect anyone to run around doing that on tests here but for the mathematically minded maybe it might be helpful or maybe, possibly interesting. It will give you a standard for measuring light. Don't use that setting to set your exposures though because video is not linear like film.

By the way I'm talking about an incident reading here or a reflected reading from a grey card.

I have a little transposition chart taped to my spot meter for this but I never shoot film anymore. It still comes in handy sometimes.
 
Last edited:
Lenny you are the man.

Finally someone into actual testing. I just tested the poop out of a series of Xenon 35mm Primes for an older than Dirt Arri IIc, and Arri BL4. The lenses are sharp, but the DP im working with was showing me how the tests can help prepare you for your shoot. Measured sharpenss vs Eye sharps, or how the CA can affect your image when shooting landscape vs, Faces.....

Fascinating stuff. i'm s glad im 1st-ing for her.

Also, inorder to do these kinds of tests, you need, THE SAME LENSES. i hate when people put up comarisons, and one sytem is on Nikons, and the other on Canons...
 
Lenny, thanks for the formula above. I can try this with my light meter sometime.

Timur, would you mind sharing with us a some more information regarding the tests you did, what you used to measure, and what to look for?

Thanks.
 
i will do a write up, send me a PM on Sunday.... im friggin swamped right now. i will forget.
 
Timur, I'd like to see that as well.
What do you mean by CA? (this is probably obvious huh).

BTW - I used to work in film rental houses sending out lenses and cameras, and I used to test the utter s---- out of film stock & filters on stills etc when I was shooting film. As a result I'm really conscious of how easy it is to make a mistake or to not be aware of variables that throw it all off.

I'm also keenly aware that we have a bunch of very hard working adapter designers, who are busting their nuts trying solve all kinds of technical problems for us, and are depending on this business to recoop their considerable R&D and maintain their families and homes. I would hate to see someone's business seriously hurt by inaccurate tests that end up all over the web. That's why I've been kind of obsessive in pointing out all the variables these adapters introduce. I've found those in my own testing at home.

Remember how many people were dissing the HVX because in the initial shootouts they mistakenly set the details too low?

That said - go for it test and compare away.

- Lenny
 
I would hate to see someone's business seriously hurt by inaccurate tests that end up all over the web. That's why I've been kind of obsessive in pointing out all the variables these adapters introduce.
Click to expand...
Exactly, precisely, exactly. Any testing should be done to controlled parameters so that only the adapter differs; that way any change in performance can be solely attributed to the adapter.

Any comparison test should use the exact same lens and the exact same camera, properly white-balanced and exposed to the exact same exposure, etc.
 
You must log in or register to reply here.
Back
Top