Jump to content

Fast apertures on the GH5 = Full frame


Andrew Reid
 Share

Recommended Posts

On 5/18/2018 at 6:03 PM, KnightsFan said:

Same reason why there is less vignetting on a center crop from a full frame image. Angle of incidence becomes more oblique the farther from the center of the image.

But we're talking about a native MFT lens (Voigtlander 17.5mm f/0.95), so the relative angle of incidence across the frame will be the same as a FF sensor with a native FF lens of a similarly scaled design.

Link to comment
Share on other sites

EOSHD Pro Color 5 for Sony cameras EOSHD Z LOG for Nikon CamerasEOSHD C-LOG and Film Profiles for All Canon DSLRs
1 minute ago, horshack said:

But we're talking about a native MFT lens (Voigtlander 17.5mm f/0.95), so the relative angle of incidence across the frame will be the same as a FF sensor with a native FF lens of a similarly scaled design.

Exactly. So my intuition was that a 17.5mm f0.95 on MFT would have roughly the same angle of incidence as a 35mm f2.0 on FF, thus making pixel vignetting a non-issue when comparing the two.

@Brian Caldwell has given me a few more things to read up on, thanks! Though, if I understand you correctly, the obliquity only darkens the edges of the bokeh, and will not have a large effect on the part that is in-focus?

Link to comment
Share on other sites

I did not watch whole video but so far it's a nice explanation of what has already been said in the thread.

2 hours ago, canonlyme said:

When talking about final aperture on a lense, you have to include the fact of which format the lens was originally made for.

It kinda does not matter what format or sensor size since that is not part of the equation. It's a common misconception that m43 lenses can be smaller than full frame, and while it can be smaller and keep the same field of view but not change the f stop. say a 25mm f1.4 is a lot smaller than 50mm f1.4. If you make the two lenses equivalent they would be exactly the same width. Granted if you found a magic material to make lenses from it would only be half the length, but if you look at the lens in the article it's not exactly 18mm deep.
Now this does not include any vignetting caused by the back element edge occluding the front element so the image circle is made small. (ever wondered why the g-master lens is so large to keep the bokeh balls round? that is one of the reasons)

 

2 hours ago, canonlyme said:

Does a full frame lense of 35mm 2.8 change its aperture when you adapt it with a simple adapter (no speedbooster) to m43? No it doesn't

While that is true you can see in the video that DoF do change simply because the image is viewed in a different size. You could just put a large black border and displayed it on a small square and it would be exactly like on full frame, minus a lot of content around. (meant to be a poke at letterbox if you are old enough to remember all the butchered DVDs).

Link to comment
Share on other sites

On 5/17/2018 at 1:45 PM, Don Kotlos said:

Another important point is that while these lenses are OK for video, the quality is not great for stills.

I use them for stills too.  No complaints here.  My Oly 2.8 pro lens does render a little more detail if one is pixel peeping, but nothing I've noticed to be remarkable compared to 2.8 on the Voights.

Anyway, this is the most EOSHD'y of EOSHD topics.  The great debate among camera nerds.  As if sensor size one way or another really matters to people that actually shoot stuff that gets used for any real purpose.  

I mean, I have footage from my old XH-A1 that rivals most of the footage I shot this year.  I certainly have my own preferences for shooting FF and like it, but ultimately it's not going to make a big difference on what is shot and deliver on a job.*  

Just curious, but aside from something needed in extreme low-light situations, can anyone here provide an example where what they did on M43 or FF would ultimately matter a hulluva lot to the client?

*to the people I usually gig for anyway

Link to comment
Share on other sites

4 hours ago, no_connection said:

It kinda does not matter what format or sensor size since that is not part of the equation.

I'm sry but I don't think you are right. The idea of the equation ignores how the lens is built in order to make equivalence of the final picture more simply understandable. 

A ff 50mm 2.8 lens simply adapted to m43 will still act like the same 50mm 2.8. By logic, putting a 50mm 2.8 m43 lens on a ff camera, the lens will still act like a 50mm 2.8ff, only that you will get black borders (a narrower field of view). Saying that a 50mm m43 lens really is a 100mm ff lens instead, just so that you understand the equation better, is wrong. 
The aperture describes the ratio of aperture (entrance pupil) diameter to focal length of your lense. The focal length is physically built into your lense and you can't simply call your 50mm m43 lense a 100mm ff lense, just to make sense of the equation. 

Following the equation there are 3 ways to get equivalence in depth of field. 
A. Take a 50mm 2.8 full frame lense and simply adapt it to m43 and it will still be a 50mm 2.8 lens, but the final image is not completely equivalent because of different/cropped field of view.
B. Take a 50m 2.8 ff lense and use a 0,5* speedbooster (which I think does not exist) and put that onto the 2 times cropped m43 sensor and it will stay truly equivalent hence act like a m43 25mm 1.4 lense. 
C Take a 25mm 1.4 m43 lense and put it onto your cropped sensor to get the same equivalence to 50mm 2.8. . 

What I am still wondering is why people (as in my suggested video) are sure that a speedbooster will not change the equivalent aperture of the final picture. However I am sure that they do! A speedboosted 0,7* 50mm 2.8 ff lense must be equivalent to a 35mm 2,0 m43 lens, for example. 

Correct me if I am wrong. I'm not really interested in knowing all technical foundations of how lenses are built, since I rather like to focus on using them :)

Link to comment
Share on other sites

2 hours ago, fuzzynormal said:

 

Just curious, but aside from something needed in extreme low-light situations, can anyone here provide an example where what they did on M43 or FF would ultimately matter a hulluva lot to the client?

*to the people I usually gig for anyway

I think it depends on what you want and what lenses are AVAILABLE.

My favourite lens is the Canon 17mm F4 L TS-E.         When you need shift and wide, you have will have a hard time do that with M43 (maybe a tilt shift adapter with a 8mm lens would be the closest if you can).

There is an increasing number of solutions so the number of situations where it would be harder to do things with one format or another is shrinking but they still exist.

Another would be fast longer exotics.     400 2.8 for instance (or even 300 2.8).      To get the end results to match would be close to impossible currently (how much would a 200 1.4 cost for M43 given the costs of 400 2.8 lenses anyway).

I could never afford a modern 300 2.8 AF lens now let alone a 400 2.8 AF but I still love using my old MF Tamron adaptall 300 2.8 and they can be found fairly reasonably priced and the same with some others (old Nikons, Canon FD ETC).     If you want 300mm FF angle of view at 2.8 what would you use on M43?   (These lenses are also great ON APSC and M43 but then give a 450/600mm FF angle of view.)     Maybe something speed boosted but that would have to be something fast and long anyway if it exists.

DSC05146.jpg

Link to comment
Share on other sites

On 5/19/2018 at 2:06 AM, ectobuilder said:

The issue with using faster glass on a smaller sensor to match a FF is that lenses have more imperfections the wider the aperture (F1.4 and lower usually) like color shifting, chromatic aberation, less sharp, vignetting...etc.

I am a GH5 owner and I love it, however I do believe using a larger sensor to achieve the shallow look is better than relying on faster glass.

I've used the Voigtlander 25 F0.95 extensively (version 1) and there is color shifting and fringing at F0.95.

I personally wouldn't buy a M43 or smaller camera unless I had a specific use case.  If you need the video features of a GH5 a Canon Rebel just isn't going to cut it.

Making large aperture, super wide angle lenses for more obscure mounts (ie not Canikon) just costs more.  Financially M43 doesn't make any sense for me.  Maybe you will save some dough on a body but you get hosed on the lenses.  And if you want to "upgrade" to full frame later you are screwed.

I am not into all this adapter stuff.  I use it, but view it as a necessary evil.  If I have some free or cheap vintage glass I will use it but failing that I prefer to just go out and get good OEM glass and keep it for life.  If you buy a metal L lens you will be able to hand it down to your kids.

Going with a smaller sensor, larger aperture, and wider glass just feels like fighting physics.  Go with the flow and save some dough.

Link to comment
Share on other sites

15 minutes ago, Damphousse said:

 

Going with a smaller sensor, larger aperture, and wider glass just feels like fighting physics.

It is fighting physics.  There are manufacturing limitations when grinding and polishing glass as well as manufacturing limitations as to how perfectly you can align and space each of the elements correctly.  Optical designers only have a limited range of glass types to use.  Eventually tolerances, manufacturing limitations and material availability/costs mean it's a technical and/or economic impossibility to match the performance you get from a simple longer/slower lens on a bigger sensor.  If sensor size didn't make a difference to the image we wouldn't see Arri, Sony, and RED all moving to FF / VV and bigger as their flagship camera systems.  

Link to comment
Share on other sites

3 hours ago, canonlyme said:

I'm sry but I don't think you are right.

Then you are ether reading me wrong or understanding me wrong.
f is literally the ratio of focal length to diameter. (I called it apparent front element as it shrinks when you stop it down).

3 hours ago, canonlyme said:

Saying that a 50mm m43 lens really is a 100mm ff lens instead, just so that you understand the equation better, is wrong. 

That is why I don't like this equivalent thing at all, not even when using a speed booster. They would have the same field of view tho.

 

Link to comment
Share on other sites

4 hours ago, Damphousse said:

Go with the flow and save some dough.

That's fair enough.  I got seduced by early LUMIX cams and then Olympus IBIS.  The kit size of smaller gear suits me.  These cams work well for my jobs, but I do still have a soft spot for the FF cameras.

Link to comment
Share on other sites

7 hours ago, canonlyme said:

Correct me if I am wrong. I'm not really interested in knowing all technical foundations of how lenses are built, since I rather like to focus on using them :)

You're like 90% of the way there.

 

See if this doesn't help:

The following examples would provide the exact same image, and therefore are "equivalent" assuming the camera is always in the same spot.

  1. On a M34 sensor: A 25mm 1.4 lens  (and it doesn't matter if the lens was made for FF, APSC, or M43 - this is important)
  2. On a M43 sensor: A 50mm 2.8 lens, speedboosted (pretend it's a 0.5X speedbooster).  (This is because you're now actually putting a 25mm 1.4 lens on the M43 once you've added the speedbooster, just like #1)
  3. On a FF sensor: A 50mm 2.8 lens

 - - - For #3 the ISO will have to be multiplied by the crop factor squared to get the same exposure. - - - 

 

 

 

On 5/19/2018 at 2:39 PM, Wild Ranger said:

Yes, you have to change the lens, and that is why you use crop factor. But the camera stays in the same place and perspective wont change.

I ran out of likes for today but this is very important for new filmmakers to know

On 5/19/2018 at 5:22 AM, blondini said:

The way the lens renders perspective is still like a 17mm lens, which is a result of its focal length, not the field of view or the DoF.

I used to think this but focal length has nothing to do with perspective, only the location of the camera and the relation to the objects in frame. The easiest way to test this is to shoot a camera at a chosen focal length, say 50mm, and then shoot from the exact same place with a lens with 2X the focal length (a 100mm lens in this example). Go into an editor and compare the full shot at 100mm to a 2X crop of the shot at 50mm. They will be the exact same image. This is no different from shooting a FF camera at 100mm and a M43 camera at 50mm. This shows how the perspective never changes if you don't move the camera, no matter what lens you're using. Seriously, try it out.

Link to comment
Share on other sites

Peak of image detailness and specific traits of rendering quality on Voigt m43 lenses is around f2-2.5 - see Shane Hurblut reports. Double with factor 1.5-2, that's even more shallow dof than mostly preferred value in today's cinematography (assuming super 35 as standard). Contrary, similar peak for most FF lenses is most often at about f4 or even at f8 with older generation lenses.

Another question is actual quality of these Voigts lenses and corresponding values - these are lenses quite in league with Zeisses (of course, the same Cosina manufacturer) which are mostly priced higher.

So, I think that comparing Voigts with cheaper FF lenses just because of simple dof mathematic is not quite correct - besides, there are also cheaper m43 solutions such as Mitakon or even SLR Magic with more players just in the game - Laowa or so.

Importance of shallow dof vs easier affordability of deeper dof in real narrative experience is so often and endlessly discussed, also here in different threads. With little bit of shame for repeating it, I'd just again note that, imo and im-practice, extreme shallow dof is simply secondary effect - because totally destroying environment discernibility most often is not in the favor of story. Photography has different preferences.

About light gatherings - the must calculated condition is to compare just sensors at least of the same generation, preferably of the same manufacturer. Than it will be clear that f stops provide approximately identical effects in m43, APSC or FF when sensor light reactivity is the same for all. Factors of bigger surface (FF) and lesser need for lightening power (m43) are mostly in compensating relationship.

Link to comment
Share on other sites

17 hours ago, KnightsFan said:

Exactly. So my intuition was that a 17.5mm f0.95 on MFT would have roughly the same angle of incidence as a 35mm f2.0 on FF, thus making pixel vignetting a non-issue when comparing the two.

@Brian Caldwell has given me a few more things to read up on, thanks! Though, if I understand you correctly, the obliquity only darkens the edges of the bokeh, and will not have a large effect on the part that is in-focus?

Large marginal ray obliquity  darkens the edges of bokeh circles.

Large chief ray obliquity in image space darkens the corners of the image.

Two very different but related effects.

Link to comment
Share on other sites

4 hours ago, anonim said:

 

So, I think that comparing Voigts with cheaper FF lenses just because of simple dof mathematic is not quite correct - besides, there are also cheaper m43 solutions such as Mitakon or even SLR Magic with more players just in the game - Laowa or so.

 

The Mitakons lack wider options 17, 12, 10mm.

The SLR Magics wider lenses are not so fast. 17mm T1.6, 12mm T1.6 and 10mm T2.1.

Voigtlander is the only brand with a range of lenses that are really fast across the whole range. All are 0.95. But they are very expensive and to buy all 4 here in Spain you are talking about 4000€!!!

Link to comment
Share on other sites

2 hours ago, Thpriest said:

The Mitakons lack wider options 17, 12, 10mm.

The SLR Magics wider lenses are not so fast. 17mm T1.6, 12mm T1.6 and 10mm T2.1.

Voigtlander is the only brand with a range of lenses that are really fast across the whole range. All are 0.95. But they are very expensive and to buy all 4 here in Spain you are talking about 4000€!!!

As you wrote SLRM has "T" measurement, so even 10mm is actually F1.8, 12mm is F1.4 etc. and these are really wonderful rendering and top-build lenses ... as I wrote, with newer sensors difference in light gathering is non-existent, F values in effects are the same between all important sensors size - including even BMPCC. Once again, buying Voigt lenses you buy lenses that have - I used and compared them - in fact better building (smoothness, durability and precision in operation) quality than Cosina Zeiss Classic line (which is technically, I think, still their best one). Fortunately, I've managed to buy Voigts for significantly lesser sum.

Saying all that, it is quite true that, striving just for shallow DoF, bigger sensors and other mounts have significantly more in general, and more cheaper (older) choices. But, all the same, knowing myself, if FF would be my preferred choice, I'd never find a peace till I complete, say, Zeiss line :)

Maybe it is always wise to include different aspects of evaluating - even level and method of camera stabilization in effects to usage of cinematic shutter speed and real life usability of F stops. For me, personally - finding no reason to be ashamed because of that, and till the moment of very high-end producing pretension - excellent IBIS system really change everything. So my great sympathy for m43 system, of course with great appreciation for everything else.

Link to comment
Share on other sites

9 hours ago, Pedro said:

What about lens distortion and lens compression??
-> shooting portraits with 50mm in a FF is not the same as a  25mm in a m4/3... 

It can be the same. Put your m4/3 camera with a 25mm f1.4 next to a FF cam with a 50mm at F2.8. The framing and DOF will be identical. See the many examples above for more info.

As a side note: I would like to cheer everyone participating in this nerdy thread on equivalence for not getting into petty arguments (i.e. the way this subject normally ends up being debated haha).

Link to comment
Share on other sites

Thank you for a good article Andrew. I am surprised to see in this thread how many seem to be focus on getting as shallow a depth of field as possible. Part of my reasons for switching from a 135 format Nikon to super35 Fujifilm was that it was easier to have more things sharp in the image. I can use f8 instead of f11 and so on. 

Another good argument for a smaller sensor is that I think you get more for your money with lenses. Look at the Olympus Pro lenses or the nicest Fujifilm lenses. To get similar quality (and then I'm talking about all aspects of quality) you have to spend far more on a 135 format system. Also optically I think for example that the Fujifilm 56 1.2 is superior to any 135 format 85mm f1.8. With Nikon you have to get into the best lenses, typically the 1.4 lenses, to have a similar quality in build. And those are more expensive. 

I don't see any 135 format 50mm 1.8 compare at all with for example the Olympus 25mm f1.2 Pro in terms of quality. Same goes for bodies by the way. The A7iii is getting a lot of praise but I don't see it comparing well to an EM1ii or XH1 when it comes to the quality of the viewfinder, sealing, general build et c. You have to get up to D850/A7riii level to match, and then we're in another price bracket. 

If you work with long telephoto lenses there are huge cost and weight advantages to using M43. 

There seems to be some sort of boom around 135 happening lately that I haven't seen since the D3 and 5D. Funny how trends bounce back and forwards. 

Link to comment
Share on other sites

On 5/20/2018 at 7:34 AM, KnightsFan said:

Exactly. So my intuition was that a 17.5mm f0.95 on MFT would have roughly the same angle of incidence as a 35mm f2.0 on FF, thus making pixel vignetting a non-issue when comparing the two.

I don't see what you're basing that intuition/assumption on. There is no equivalence theory to apply here - optics/ray angles don't change based on sensor size. The only impact on sensor size would be using a larger-format lens on a smaller sensor, where some of the rays are thus cropped (don't reach the image sensor), but that's not what we're speaking to here - the Voigtlander is a native lens.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

  • EOSHD Pro Color 5 for All Sony cameras
    EOSHD C-LOG and Film Profiles for All Canon DSLRs
    EOSHD Dynamic Range Enhancer for H.264/H.265
×
×
  • Create New...