Speedboosters EXPLAINED - The truth about focal length reducers

[@danielzare6766]

This deserves a million views, well done!!

[@ronsunsinger483]

You've made a very concise and very clear video that was easy to follow, was very focused, and spot on the money. Thank you for the time and effort it took you to make this.

[@alex-stone]

Hi Ron, thanks for your kind words. It means a lot!

[@PitNeex]

Thanks for the video, it's nice to have all the information on one video, bookmarked and probably re watch this one in the future 👍

[@camerond8176]

Using a speed booster does NOT change my APSc camera into a full frame camera........................However the perceived FOV of the lense will be much closer to that of a full frame camera with the same lens. The speedbooster will allow to APSc sensor to view closer to it's full frame equivalent, when the crop factor is used. Canon APSc is 1.6 crop, 0.71 Speedbooster - Means a 50mm FF lense will be seen as a 56.8mm lens on the APSc camera. 50mm x 0.71 = 35.5mm x 1.6 crop factor = 56.8mm FOV through the view finder. 50mm x1.6 Crop factor (without speedbooster) = 80mm FOV through the lense.

[@vladimirtalijan]

People who don't understand how that works will be even more confused after this video :) The main problem everyone are forgetting is the noise/ISO values. We are all used to smaller sensors having more noise and that is exactly why manufacturers of all crop cameras are lying to their customers. Just compare f/2.8 on a Canon R5 to a f/2.8 on a phone, are they similar in any way? Of course not, that phone image has no characreristics of a f/2.8 lens, so why are we calling it like that? Why would we even care about same exposure settings when image properties will be vastly different between different crop factors, because they faked the ISO values. Cameras having the same settings means nothing if sensors are not the same size, any camera with a cropped sensor should show the FF equivalent settings and then the confusion would be over. The same goes for focal reducers, that f/1.0 C70 shows in camera is actually in relation to a super35, and that's why when you compare it to a full frame camera with a f/1.4 lens it has exactly the same exposure, dof, fov, bokeh, and noise. Of course sensors will not actually transform like Optimus Prime, nor the lenses will, but when you put a 50/f2 on a 1.5x, you are not only getting a 75mm fov, but you're getting a full frame equivalent of the f/3 image in every way. Everything about the dof, bokeh and noise will look exactly like f/3 on a full frame, and that's what most people don't understand. Because when you understand that, the result of putting a speedbooster is much more clear, it's like reverse cropping, you gain what you've lost when using a crop sensor. That's what people want to know, what is the actual result in the final image, not how to calculete aperture opening which tells you nothing. Bottom line, if you want to know what kind of image fov, dof, bokeh, noise and image compression you will get compared to a full frame sensor, you just need to know what is your crop factor. If it's 1.0x, your image will be exactly the same as a full frame, again in all the image properties, fov, dof, bokeh, noise and image compression. That is why camera manufacturers should show real settings of crop cameras, people should know that Panasonic 12-35/2.8 is not a 24-70/2.8 equivalent, not even close, it's actually 24-70/5.6, and I mean it produces the image of such lens on a full frame body, it's not transforming with sound effects and smoke :)

[@raksh9]

This would be misleading to less informed users. Aperture is a value relative to focal length, not relative to 35mm. A 12-35mm f2.8 on m43 is only equivalent to 24-70mm f5.6 in relation to focal length and depth of field. But it still has f2.8 aperture in relation to its native focal length. Panasonic does not claim that 12-35mm f2.8 is equivalent to 24-70mm f2.8 in terms of depth of field relative to 35mm.

[@jay-by1se]

Great comment. But I think people who buy cropped sensors know they aren't full frame. But I shoot m4/3 Olympus, Nikon DSLR and Canon r6ii. The only photos you can tell apart is the Nikon. The Canon and Olympus are indistinguishable. You do get more bokah with the full frame but I found I don't always want everything a blur. Sometimes just a light blur is amazing.. People know.

[@SMGJohn]

The lens never changes, the aperture number is a scientific number how much light it lets in, that does not mean the output circle is the same, but F4 is F4 is F4. Why you are yapping about noise is beyond me, techno gibberish trying to sound intelligent. The ISO is just sensor overdrive, basically how sensitive it is. What dictates noise is not sensor size, its PIXEL SIZE or pitch as its correctly called, a M43 sensor with 6mp will have same noise as a 12mp full frame sensor IF they are made of the same technology and specs. But in reality you are comparing a 25mp M43 sensor to 24mp full frame, of course it is less noisy, the pixels on its sensor are much larger, I am not gonna go into WHY bigger pixel is less noisy, we be here all day but I can give scientific document in this that explains it in details.

[@fy7589]

Despite being tiny, speedboosters are insanely good for them providing an extra stop of light. Making a really cheap 50mm f/1.8 lens perform almost as good as a very expensive 35mm f/1.2 autofocus lens if that even exists. And since the F/stops are not as accurate as the T/stops used in cinema lenses, I believe most manufacturers who make an f/1.2 lens are likely gluing the speedbooster into the cheap lenses they make and calling it an f/1.2 lens. It's a budget alternative to picking up canon's L series lenses for example. Of course it's probably not gonna be as sharp, but the metabones speedbooster adapter I got for my canon R7 and cheap EF 50mm lens, it remains almost as sharp wide open compared to without the speedbooster and even sharper if I stop down to F/1.6 which is brigther than F/1.8 and definitely a lot brighter than stopping the F/2.4 to achieve a similar level of sharpness without the speedbooster. Very nice and informative video you've got there. Thanks.

[@Imhotep397]

The brightness test is like the nectar of the gods! Tony Northup has been pushing that the speedbooster doesn’t increase the light transmission…even knowing that that the extra glass is effectively a magnifying glass for the lens.

[@alex-stone]

He perpetuates a lot of misconceptions. Don't get me started on what he says about focal lengths and crop factors...

[@gerardferry3958]

a speed booster uses all of the lens while a teleconverter uses the middle only as far as i know

[@davidak_de]

that helps a lot, thanks! i was fascinated by the idea that i could create a lens with the depth of field of a f/0.7 lens by combining a canon EF-mount full frame lens with f/1.0 and a speed booster to APS-C. There could be a speedbooster for canon EF to pentax Q, which would create f/0.5 from a f/1.0, but there is only a nikon f to pentax q. and as we learned here, the depth of field will stay the same. maybe it would be too shallow anyway

[@bluecollar8525]

Yes, but it will perform very similarly to that same lens on its native full frame body. The incoming light will be about the same, and so will the field of view and depth of field. This is why people often say it's like turning your apsc into a full frame. I know he said it's a myth, but he's being pedantic with words. the effect is essentially doing exactly that. You widen the field of view by reducing the focal length, but then the crop factor negates/undoes that effect. You will still probably have slightly more noise and less dynamic range because you can't gain extra dynamic range by going lower than base iso on the smaller sensor, but otherwise it does essentially turn it into a full frame camera in the resulting image

[@davidak_de]

@@bluecollar8525 thank you!

[@PetesShredder]

Nice cage. We are working on c70 cages now, and just done a canon expander mount support for the Komodo. This is a useful explanation as we have had a few arguments in the BT office about this 🤣

[@alex-stone]

Andy has my number, you can ring me next time you have an office barney!

[@PetesShredder]

@@alex-stone maybe we should have done... you should nip in and visit if you are ever down this way.

[@alex-stone]

@@PetesShredder Covid permitting, I shall

[@ernie5229]

If you use enough speed boosters in series can you go back in time? This would be so helpful for those missed shots.

[@wesleymarshall2321]

Thank you for this excellent factual video, just one big question. If I have a 50mm on a super 35 sensor with a 0.7 speed-booster will the perspective or backround "fall off" be that of a 50mm or 35mm? Or is the fundamentals of my question flawed?

[@alex-stone]

It would depend on the aperture you've set. If the speedboosted lens is wide open, then you'll have the same DoF ('fall-off') as the 50mm when wide open, even though the aperture is one stop faster. If you stopped down to the same f-number to match the exposure, then the DoF would be deeper.

[@mykelbankz]

This is a very great video Alex, thanks. My question is, is there a possibility of using speed boosters on full frame camera. If yes, would it still affect the f-stop and the focal length? Thanks

[@noelwangler]

It would work if it is a speedbooster from full frame to medium format

[@alex-stone]

Yes, assuming you can mount it (for example E mount), you could use a speedboster on full frame. It would still affect the f-stop and focal length, but also the image circle. If you speedboosted a full frame lens, the resulting image circle would be APS-C and would therefore porthole (hard vignette) on a full frame sensor. You could however speedboost medium format lenses and these would cover full frame.

[@mykelbankz]

@@alex-stone thank you

[@mykelbankz]

@@noelwangler thank you

[@Metallicaman75]

Best Video ever. Please make new ones. Cheers.

[@zak_ray]

I'm not sure I totally agree that DOF is strictly a math thing and has no subjectivity. After all it's based on CoC which is based on the final projected image size relative to the viewer. We intuitively know this to be true, as an out of focus image looks sharp if scaled down far enough. I can see that the bokeh circle is the same size as BB-8's eye, however because the field of view is smaller, doesn't that make each circle closer to the CoC and therefore more likely to be in focus?

[@alex-stone]

Whilst the physical depth that is in focus is measurably the same distance, you are right that subjectively one can appear more in focus depending on one's threshold for what constitutes in or out of focus.

[@Kazimov100]

Hi Alex. Great video explaining Focal Reducers and their effects. I just have 1 question. Upon using a focal reducer we reduce the focal length of the lens which in turn automatically changes the aperture as well. Now my question is will the distortion of the image follow automatically as well. As an example: While a 50mm f2 on Full Frame and a focal reduced 50mm f2 on S35(hence 35.5 f1. 4) will have the same depth of field and field of view will their distortions vary? I.e will the focal reduced 50mm f2 (35.5mm f1.4) now have the same distortion as a 35mm lens, say when it comes to faces and the like. Thus, in effect the images aren't actually completely identical even though the DOF and FOV look exactly the same. Hope I made sense.

[@alex-stone]

There are two types of lens distortion - barrel distortion and perspective distortion. The first is a property of each individual lens - some lenses will fisheye slightly and others will pincushion and is largely independent of focal length. For example, a 24mm from one brand might render horizons straighter than another 24mm from different brand. The second is to do with field of view - how much faces are stretched or squished (wide angle distortion or telecompression). The wider the lens, the more you'll encounter this 'distortion'. It's not distortion in the sense of it being an imperfection - it's optical physics, it's unavoidable. I've seen reviews of ultra wide lenses (11mm on full frame) complain how they didn't like the distortion because of how much it stretches the edge of the image - they were confusing perspective distortion for barrel distortion. That's not a flaw of the lens, that's just how all lenses that wide look. So to answer your question, a speedboosted 50mm on S35 (35mm) will have the same perspective distortion as the original 50mm on full frame as their field of views are the same. The 50mm on S35 without a speedbooster will have less perspective distortion because it has a narrower field of view. Do note however that when using a speedbooster, the extra optics may introduce some barrel distortion of its own.

[@Kazimov100]

@@alex-stone Perfect explanation. Thanks again!

[@flyingfox2005]

@@Kazimov100 A better way think of it. A 50mm using a 0.71 focal reducer is now a 35mm lens. A 35mm lens on S35 has the same angle of view as a 50mm lens on FF / 135. This new 35mm lens has the same perspective distortion as a native 35mm lens on the S35 camera. Because the 35mm lens on S35 lets you place the camera in the same position as the 50mm lens on FF / 135 - and since perspective distortion is dictated by the distance from the entrance pupil to the subject... the perspective distortion is the same. The 50mm on the S35 camera without a focal reducer will have the same perspective distortion - if - the camera is left in the same position (because distance dictates perspective distortion). However given the narrower angle of view, it is more likely you will move the camera back. If you move the camera - YOU - by changing camera position have changed the perspective distortion, as you have changed the distance from subject to lens entrance pupil.

[@matkovskyp]

This is the best explainer of the subject I've seen. 🙀

[@PaulJakob]

Wow great video man! Glad I found this.

[@jobane2]

Thanks for the video. It’s hard for the “art world” to accept science… here is a question for you: why don’t/can’t they make a speed booster / focal reducer for Sony E-mount lenses from full-frame to APS-C? (Not for A mount to E mount) To clarify, I want to use my Sony E mount full frame glass on Sony E mount APS-C cameras and have them “reduced” and use the full image created.

[@alex-stone]

You can't have an E mount to E mount adapter without compromising the ability to reach infinity focus. Speedboosters make up the difference in flange depth between the two mounts they're adapting - ie EF to E mount is 44mm to 18mm, so the adapter is exactly 26mm thick. An E mount to E mount adapter would act like extension tubes - great for macro work but you wouldn't be able to focus beyond a couple of metres at best.

[@utube321piotr]

I am quite selective and I've spent 2 min debating whether to watch your video, glad I did. Keep it going I am subbed.

[@W1ll88]

Would love to see this video with M43 comparisons as well. Though that would start getting really complicated. A lot if misinformation out there on focal lengths, speed boosters and sensors though

[@alex-stone]

M43 is half the size (quarter the area) of full frame, so any comparisons are simple - for the same field of view, half the focal length is required. To match depth of field, halve the aperture value too. Example - to match the 'look' of a 50mm f/2.8 on full frame, a 25mm f/1.4 is needed on M43. The FoV and DoF will be the same but the exposure on M43 will still be 2 stops brighter.

[@DopTikhonov]

So useful video for many who believes in marketing! Thanks)

[@TheSagnac]

Hello, to calculate the hyperfocal distance, how would it be? for example: with a 50mm 1.4 lens + speedbooster 0.71 + m4/3 sensor (Coc 0.015)

[@alex-stone]

The same as calculating a 35 f/1.0 lens + m4/3 sensor

[@icandi9178]

Good stuff!

[@nowyrozdzial]

Just a thought after watching this video. If we have a super 35 camera so 1.5 crop and use apsc lens but in slowmotion that has 2.0 crop add the speedboster we could possibly have the same image characteristics negating the vignetting of apsc lens with speedboster - right?

[@alex-stone]

That would be an interesting use case. I haven't tested that but in theory I suppose it would work. Do remember that the slomo crop is a digital crop so resolution and image quality (sharpness and in some cases dynamic range, bit depth etc) may take a hit. But yes in terms of just field of view, a speedbooster would help negate the digital crop, just remember to take it off when going back to normal speed ;)

[@nowyrozdzial]

I think in most cases assuming modern 4k super 35 cam it is downscaling from 6k or higher resolution. With digital crop it would still be 4k. Speedbooster could also help to keep the brightness similar as slow motion will reduce amount of light due to shutter speed change.

[@nowyrozdzial]

And obviously you will lose autofocus so it is not ideal in slow motion for wildlife and fast moving objects.

[@jackbarron8880]

What about distortion, Alex? I'm scared to use a speedbooster because I don't want to change he character of my zeiss contax lenses

[@alex-stone]

Any added distortion may depend on the brand of focal length reducer though I didn't notice any when pairing them with a Metabones. Don't be scared to use a speedbooster if that's what it takes to use your lenses - better to use them than not. If you have the means, test their look on a full frame camera and then on an APS-C/S35 camera with speedbooster to compare.

[@jackbarron8880]

@@alex-stone Thank you, Alex. I mean it.

[@jackbarron8880]

@@alex-stone I think for me it's just making sure that the image still holds up on a big screen even with the speedbooster, you know?

[@TFaminu]

My boy 🔥. Thanks for the video! Jesus loves you

[@RaphaelIgrisianu]

Can a APSC lens be used with a speedbooster or does it have to be a full frame lens?

[@noelwangler]

It works sometimes. For example you can modify the Canon EF-S 18-55mm and use it as a 24-55mm lens. 18-23mm has vignetting

[@alex-stone]

Yes, APS-C lenses can be used with a speedbooster however the resulting image circle would become even smaller. It would therefore porthole (hard vignette) on an APS-C sensor.

[@quentagonthornton49]

0.71x speed boosters are made for APS-C lenses and 0.64x ones are for full frame. EF-S lenses generally won't fit in speed boosters without modifying them.

[@bryanbarajasBB]

Is it worth using a Sigma 18-35 with metabones speedbooster on a APS-C camera? Yes😂

[@DIONISIOSGIAKOUMIS]

PREACH

[@cordvision]

Has my comment been deleted about how the sensor size is related to the acceptable CoC, therefore has an influence on DOF? If so, may I ask why it has been deleted?

[@alex-stone]

No comments have been deleted...

[@cordvision]

@@alex-stone Oh ok, something must have gone wrong when I tried posting it. I was commenting on the statement that I don't think the statement at 9.00 is true. DOF is actually affected by sensor size when you watch the recorded image on a screen with a fixed size: To display the image of a MFT sensor on your screen, you need to enlarge it twice as much as an image recorded on a full frame sensor (MFT sensor is half the size of the FF sensor). For that reason, the acceptable CoC on the MFT sensor is about half the size as the acceptable CoC on the FF sensor. For that reason, the DOF on the MFT sensor is about half as deep as the DOF on the FF sensor (I know this sounds counter intuitive, but you can easily test that yourself with a DOF calculator. Choose one like DOFsimulator or PCAM and only change format size without touching the other parameters). Any comments?

[@alex-stone]

@@cordvision The perceptual depth of field changes depending on a number of external and variable factors such as display size and even resolution: the closer you look at an image, the more critical the threshold for what is in or out of focus and therefore the thinner the depth of field is perceived to be: the CoC is what is deemed to be acceptably in focus given these additional parameters. The optical depth of field however - the light projected through the lens - remains unchanged as demonstrated in the video. The above two distinctions are important and are how the Achtel 9x7 camera claims to have thinner depth of field than IMAX film, despite the fact it uses a S35 sensor. Side by side with another camera with the same physical size sensor and, when displayed at the same size, the depth of field characteristics will be identical. Zoom in to each image at 1:1 pixels however and the Achtel will appear shallower simply because the blur has been enlarged more.

[@cordvision]

@@alex-stone Not sure what you mean by "optical depth of field". There is no such thing as "optical depth of field" as far as I know. Without defining an permissible CoC you don't have any depth of field at all (literally only a plane with basically zero depth is "in focus"). Therefore, in order to even have depth of field, you need to define the permissible CoC first. This will then have a direct influence on the observed DOF of the image on your viewing screen (in absolute measurable terms... by measuring the CoC on your viewing screen). If viewing conditions for both formats are the same as well as recording resolution, the DOF of a smaller sensor will be shallower than that of a larger sensor. This is even true in your wall example. For the image projected on a smaller portion of the wall (smaller sensor), you would have to move closer to the wall when looking at it (else the image would be much smaller in your field of view). Shortening the viewing distance affects DOF the same as enlarging the projected image. At the end of the day, DOF is all about perception... it is deciding how large a single point is allowed to be in your image to still be perceived as a single point. You therefore can't ignore viewing distance or viewing screen size. This is why a DOF calculator tells you that your DOF gets cut in half by simply changing from a FF sensor to a MFT sensor (without touching any other parameters).

[@alex-stone]

@@cordvision What I mean by optical depth of field is the size of the blur relative to other objects in the frame - how pronouced the bokeh is. Viewing distance, format size and resolution have no influence on this - it's the image formed by the lens.

[@JeffBourke]

I absolutely hate speed boosters.

demasiado drama para algo tan simple y técnico...

[@JeffBourke]

Canon are actually stupid. They released a high end S35 camera without any lenses so they basically released a camera that’s expected to use a booster. Dub dub dub.

[@bbones3318]

cheer up mate. not watching that.