Experimental 8K super35 GH5 vs SpeedBooster vs. native lens - footage & tutorial - Epic Episode #6
Вставка
- Опубліковано 7 вер 2024
- We are building a crazy rig to show how an 8K super35 sensor would look. That gives us the opportunity to match it up agains a GH5 with a SpeedBooster. How well does the boosted version hold up?
Does the booster cause significant optical degrading? We are going to find out… and of course we are matching these images up against GH5 with a native lens to see, if it was worth it in the first place.
For your viewing pleasure, the native 8K large format composition is uploaded in 8K. The rest of this episode was acquired in 5K HEVC or 5.7K ProRes RAW and upscaled to match.
Please join the ranks of the Media Division:
www.youtube.co...
Instagram/IGTV: mini tutorials, whats cooking, and updates:
/ media.division
Facebook Group: Community, discussion, help... ask everything
/ mediadivision2
You can download the original GH5 8K composition in ProRes 444 here (beware - this has 5,5 GB):
we.tl/t-FwBQln...
If you just want to play with one frame:
we.tl/t-IcoY1X...
The Media Division is dedicated to the art of filmmaking and sharing knowledge about it. Tests, Reviews, Gear, Tutorials, Tips, Tricks, Techniques, Experiments, Behind the Scenes, Experience & Knowledge… for work, for fun, for the love of the craft. For beginners and professionals.
MDEpicEpisodeS1E06
This is reminiscent of the Panasonic GH2 hack! I'm excited to see what can evolve from this technique!
Thanks man... well, this has almost no practical application, so don't expect toooo much ;-)
How to you joining the 8k footage on After effect? Give some tutorial 🍻🍻 i want experiment in 65 large format
@@heyndre It's right in here… but very short as it is so simple… ua-cam.com/video/1T3U0Kkqy1g/v-deo.html
And (as I mention in the end of this video), I did the 65mm experiment already……………… here: ua-cam.com/video/HN29l3ouR64/v-deo.html
I realize it is pretty off topic but do anyone know a good website to watch newly released tv shows online?
You deserve an award for not only the content of this video but the cinematography and presentation. The work you put in to such a specialist video is remarkable, well done!
Thanks a lot Hugh.... very kind of you and very motivating. Trying to get better always always.
4:54 - that finger print on the sensor though...
was waiting for that... as you can see on the very next scene it has been cleaned.
@@MediaDivision Lol 'cleaned'
04:15 That stop motion effort.. Neat.
Thanks man!!!
VERY interesting. I’ll go with 5K boosted, mostly because I already have that equipment. I think most of us (myself included) worry way too much about ideal when what we already have is more than good enough.
Thanks.... and agreed. But hey.... being a nerd is halve the fun ;-)
Media Division Indeed it is!
DOOOODAH!! You know your scat!"Although I will be 63 in 2weeks I want to grow up and be just like you. I only have to watch to your presentation 20 more times so I can remember the word speed throttler. Just remember I'm with you coach! Cheers!
Thanks… rather watch out newer episodes… we got soooooo much better since
Nothing that hasn't been said before on the comments. This is frankly brilliantly executed and educational. Thanks a lot for creating it.
Thanks a lot for that Oni... and thanks a lot for hanging around!
wow, a future-proof video in 2019 ;) I don't even watch YT in 1080p yet...
it might be time for a new TV, Laptop etc...... no, seriously.... diminishing returns and all.... 4K on UA-cam is worth it for the bandwidth
Can you ask Panasonic for an early model of the S1 for a testing video? You would be the perfect person to test the camera for 2 months and then release several in-depth videos on the capabilities of the camera.
thanks man... I happen to be admin of the FB group, too. So yes.... I guess ;-) tell them... but only if I can keep it :-)))))
Dude. This is *amazing* -- the execution, depth of information, polish, etc. are all just breathtaking. I don't really sub to filmmaker videos because it's all basically just HOW TO MAKE YOUR VIDEOS LOOK LIKE OTHER POPULAR VIDEOS but you've really outdone yourself and I can't not give you props for it.
Your color grading/cinematography are really beautiful as well. Would love to see more info/tutorials from you on those. Again, thanks for the really great, high-effort work. Subbed!
That so sweet of you Luke.... very much appreciated. I know what you mean... UA-cam can feel a bit redundant in that way. Glad you think I have developed an own voice. I have quite a bit of tuts and bts that touch your request and more will follow in the future. If you want to know anything specific, just let me know
Thanks for hanging and with that name you need to watch the Goldvader tutorial ;-)
I just wanted to say thank you for producing such high quality, engaging content. Not only does it look beautiful, but the structure and overall pacing of your content makes me excited for what I'm about to experience. I look forward to every one of your videos. Thanks again for producing such striking, quality content.
Thanks Brandon.... that is very motivating. I will try to get better and better.... thanks for hanging around!!
It's shameful how horribly underappreciated this channel is. This is a truly remarkable channel
Thanks a lot Christian..... working on it ;-) give me 2000 years
Dude. Thank you for having a good concept review video. More people need to see this and follow suit. This is AWESOME!
Thank you so much Robert. Naaa.... leave that to me please ;-)
Dude your channel is literally the only one on UA-cam I feel the need to use 4K. Like I have to use 4K or better if I ever get better. Please make more content, and your music sample is perfect since it screws with my childhood melodies (nostalgia).
I try man, I try… but these videos make no money, so this is just a hobby and I got children to feed. There will be episodes comeing… don't worry
@@MediaDivision Eventually you will. I love what Peter does as well at 4m subs which didn't happen overnight. Long road but fast later on, when he got to 1m; 2m, 3m & 4m came really quick.
Speaking engagements and collabs helped push his channel.
Capitalize 👇
I was seriously looking at medium format lenses after watching your video. A sponsorship or affiliation with ProFoto, eBay kickbacks, Amazon, etc.
Right now I own a D300. I'm looking at the Z 50 & Z 6, D3s, D5, D500, D850 Fuji X-T3, GH5, Leica M2, 5DMk 4. I have studio space. I'll be getting my old strobes back but would like to switch over to LED. I'll convert from paper back drops to cloth.
Some don't prefer to do affiliations but personally I don't care. I rarely ever use them but if it's of high interest (photography) like I'm rebuilding my photography kit and studio. Once I've paid a few bills. Nov. at the earliest and Feb. at the latest I'll have the funds to pay for the new kit.
@@MediaDivision I'd love to use medium format digital but it just don't make sense for me right now.
Sure... there is always one who made it big with something... but I have to feed a family now. I know the theory and the progression curve... but it is what it is... theory.
I do use affiliate... you find affiliate links related to my episodes in the descriptions but that is really something close to nothing.
it doesn't really make a meaningful difference at all. Back in the analog days it was a clear boost in resolution, but today, it became unnecessary. FF is really a sweet spot
Great channel. The reason that the depth of field is shallower on the spin-twin (ST) even at the same f-stop is that on the others you zoomed to a smaller focal length. F-stop if the ratio of focal length over aperture, thus at the smaller focal length the same f-stop was a smaller aperture. At the same perspective -- same scene same camera position -- aperture in millimeters determines depth of field.
For the ST you stated focal length of 28mm at f/2.8, so the aperture was 10mm. You didn't state the focal length for the others, just that you zoomed in till the image on the smaller sensor matched.
There may be a secondary effect in that on the higher resolution ST the "circle of confusion" should be smaller. We can resolve smaller variations in focus.
This channel is making the content we really want to see! Great stuff!!
Thanks Erik!!! So nice of you and thanks for hanging!
who on earth gonna make such a epic content. subscribed
Hahaha... thanks man..... and thanks for hanging
I haven't seen anything quite like this before. This should be UA-cam's hit video. Many thanks for your effort. Wish to see more.
Thanks man… I wouldn't mind for sure. Spread the word ;-) There is a lot already and more to come… a little patience… This is not something I can make a living from
@@MediaDivision Sharing your videos with my friends and geek lovers will be a joy and I will.
@@JajabarBangladesh Thats is terribly nice of you! Thanks a lot
I stumbled across this video on reddit, then came here and lost an entire evening on your channel. Excellent videos. Excellent presenting style, and excellent knowledge. Don't stop making videos. Your style is so refreshing when most of youtube is all about fast cuts, stock music and slo mo b roll. You get to the point, explain it well and you do it so well because you understand it. I have to stop typing now because if i continue your wife may get jealous ha ha.
Hey Wagon Wheele.... thanks for the flowers. That is kind to say and very motivating for me. I hope I will improve and do many even better things in the future. I know the wife problem :-)))))))
Love the experimentation. The 8k looks very sharp and I like how the light rays are more visible.
Camilo Pineda thanks Camilo.... in all fairness.... the light rays might change from scene to scene as the haze in the room changes slightly.
Everything about this video is just amazing! This deserves so many thumbs up!! Thank you for taking the time and effort to produce this, real quality video.
Cheers Babebey… trying my best and it can only get better ;-)
Great video.
I think the dof is actually the same despite the different degree of background blur. And there is no flange-distance problem as other have suggested.
Note that dof is defined by the distance that is in focus. And the focus is a matter of reaching a certain threshold of resolution. As long as different resolution curves cross at this point the dof is the same. However different shapes of resolution curves are possible. The larger format here has a steeper one.
I think that's because the speed-booster increases the asymmetry of the optics. Creating a curve that deviates from a flater symmetrical bell shape resulting in a smaller circle of confusion.
Thanks for the flowers and the input. You are talking about optical resolution?? The sensor resolution will certainly not matter. If you are right, and the booster influences strength of DOF blur, that would be interesting to know. First time I ever hear of that.
@@MediaDivision
Optical.
ua-cam.com/video/d1FD4-oBPvY/v-deo.html
~7:30
amazing as always. thank you very much for the video. it actually answers some questions that i always want to know but cannot bring myself to find out.
Thanks Sean... glad you found something useful
I'm loving your experiments, with or without practical application. That's what experimenting is all about!
I'd like point out, and thank specially for putting so much care in the cinematography of your videos and tuts themselves, not only in the examples, but throughout all the video.
There's so many content out there that's good enough in it's content, and when the examples come, but then the "guy talking" is badly lit, overexposed, and overall ugly. Yours are a pleasure to watch all through. Thank you.
Thanks Pablo... always trying to make dry subjects entertaining and seamless... glad you feel we succeeded
eyo a new video! glad to see the channel grow
Me too Ruddy… cheers man
Very cool experiment - the speed bosted is definitely worth the investment, - Your depth of field query... the 8k composite "spin twin", is the only version of the test with the larger flange distance. Flange distance does effect DOF
Thanks man… yeah, flange is a good candidate even though I tried hard to get it right. If that is the solution I am amazed how drastic the effect is and I wonder, why nobody ever^developed a mini tube mount where you can alter the flange in a very small amount (that is not the same as tilt shift). That would totally erase DOF arguments with same sensors?! Of course you can't reach infinity that way, but in cases where you want DO you usually not at infinity.
@@MediaDivision Perhaps use a bellows (commonly used for macro)
@@StewartMarsden I own bellows, but the y don't move sideways and the flange is waaay to short to fit a bellows in between. Even if there where bellows moving sideways and thin enough… they still will not magically give the right flange… quite the opposite as bellows are made to alter the flange
Wow! Revealing. I say that the M4/3 sensor is 70% bigger than the old 16mm film stock of yesteryear. Adding the focal reducer gets you the 35mm movie film stock field of view. The only thing I don't like about my particular the focal reducer is the lens flares. It project a second set of flares in huge bluish-purple hexagonal pattern over the original lens flare. Outside of that I'm surprised at how well the m4/3 with a focal reducer performs.
That sound strange… my booster doesn't do that. Of course, the lens will add an element to the flare, but zooms have so many already... one more doesn't matter to me. Yes, I was suprised how close they are. But between 8k s35 and native, there is a least a relevant difference.
Ok. I’m not even sure what I just watched but Sir it was awesome! Definitely hitting that 🔔
Thanks man… glad to have you… stay with us and we get there… I don't know where there is exactly ;-)
Nicky you're a genius, conducting interesting experiments that I only partially understand but completely entertaining nonetheless.
Thanks a lot Billy!! Genius.... hardly .... but it is a nice thought :-) thanks for hanging
Woah! Great experiment, thanks for that. I like the speedbooster combo not only for the imagequality but for the versatility my other canon lenses (10-18mm eg) give me
Thanks man… sure… same for me… but most importantly for the gained stop.
Well I can tell the 8K is still superior even when displayed on my 1440
screen if not by much. However, I wouldn't consider the comparison
absolute at this point since, the demonstration are under controlled
conditions.
Still, one has to wonder if the results would be as
close out in the real world and what anomalies might pop up when light
starts bouncing around from all angles?
In the end, size does
matter. Although it may be outside the necessary parameters for all
intents and purposes. In any event, this demonstration was superb and
very revealing. Bravo!
Thanks for your input Krane! It gets very apparent sitting close to a large 4K screen BUT of course… that is only so important. Comparing the GH5 at the time, I can say that shows remarkable results, considering they tiny sensel size ;-)
Your videos are great. I think this will be your year of channel growth!
Thanks man.... that would be really sweet
Well done, nice results!
Thanks man
Interesting video! I always feel the higher resolution the camera, the more I have to not make it look like high resolution, for the actors and for the story. Using an 8k RED, for example, looks hopstital clean; using lenses from the 60s or 70s, in addition to filters makes it more pleasing to the eye; my 2cents. If the budget is there, having the flexibility of shooting in 8k or 6k and modifying the image with lenses/filters, is preferred.
thanks man.... agreed. I use Mamiya 645 lenses (many shots in this) and the talkies have a 1/4 black promist on them. Still, I like high res for detail and fine noise. Extremely high res actually makes an image less sharp... I know that is counterintuitive - it is simply because there are more gradients between areas of high contrast.
Check out this test where I use the method to produce a 12K Alexa 65 sized sensor with EVA1
ua-cam.com/video/HN29l3ouR64/v-deo.html
That was GREAT!!!! Loved it. Subscribed and looking forward to more. I love my GH lineup (4,5 & 5s) plus all my speedboosters. Thank you dude!
Thanks man.... glad to have you around. I do few Gh5 specific videos, so I hope that will not be a problem. I think more about doing generally cool stuff. Experiments, cinematography, recreating classic effects... all the is fun.
Media Division looking forward to learning more than about my camera. I think you’ve got cool ideas. There’s a ton of GH5 stuff out there, not a lot of clever stuff like this!
@@sbrown6434 then I'm even more glad to have you around ;-)
Génial, superbe travail merci de partager ces expérimentations dans ce format ! Nice
Cheers Laurent !!!
Neat experiment! My only note is on the first shot the flashing light was driving me crazy. I later realized that it was the projector, but as a new viewer it was distracting. Keep up the great work!
Thanks man… I thought it looked interesting. Doesn't distract me even on a HDR display ;-)
@@MediaDivision You know it might be a good idea to put a strobe warning on this video. As the flickering could trigger a seizure for someone with photosensitive epilepsy.
@@MYTDX4 i don't think it is that bad as it never covers a significant screen portion and it is at a low non stable frequency...... and nobody ever reads the description BEFORE watching the video
Very interesting video, 👍🏼 of course here in México go for a larger sensor means a big economical sacrifice that some times clients don’t even appreciate neither want to paid for it
thanks.... thats not a Mexican but a global problem, but hey, halve the fun is being a nerd
Mate your videos are amazing! Educational and force you to think on the subject after the class
Well, thank you my dear sir……… never thought I would kind of give homework ;-)
Media Division 🌟🌟🌟🌟🌟
Thank you! very informative and very well presented video, as always. Ive been having doubts if buying a speedbooster or not.
cheers Ignacio.... well the 18-35 + booster might not be so relevant in this test, but in low light situation it is gold. I used it on events over and over and over... love it. If you have a GH5s, it is less relevant for sure.
Great Experiment as always!!
Miss your videos man and nice look ;)
Cheers Darko... trying to do more, but even a short one like this takes 3 days in total and I have a small boy at home... trying to get better and faster
I love your gh5 vids. Top notch work. Softer might be related to how much light is reaching the sensor compared to F. F stop transmission maybe uniform, however variance of T stop might occur. Just a guess.
Thanks Faust.... hmmmm, the light transmission is surely inhibited by the booster (very little), but that doesn't change the IRIS of the lens and therefor not the DOF. The image would just be slightly darker, or noisier if compensated in the grade. If my cam had auto iris activated, it would open the iris to compensate (my guess would be still to little to make a difference lake this)... but that would result in a boosted image with shallower DOF... nor vice versa. So, thats not it ;-)
The best explanation so far seems to be a flange distance that is off.
Thanks for your input!!
Camera scientist, nice!
Thanks Robin… and thanks for watching!
I'm super excited for your next project!!
Me tooooo… I already had to change pampers twice… Ultra Panavision 70.......... but I can't promise it will be my next. One of my next. Have to wait for a couple of things!!! Patience!
@@MediaDivision I'm sure it's going to be awesome whenever you finish it!!
@@sheggy5442 I'll make sure…… thanks for hanging
Holy cow.. UA-cam now support 8K! And here I'm using 1366x768 super crap green shifted monitor seasoned with dead pixels here and there.
It has for years.... time for a new screen. Modern TVs work well and a dirt cheap
@@MediaDivision Yeah. But I would like to make a direct jump for a pro monitor... (that's what I like to think.. 😅 I'll settle for a cheap fairly color precise 2K/4K monitor in a few years [I don't earn yet]).
What monitor do you use for color grading?
And looking at the video description I'd like to praise the inclusion of footage again. I'll play with it when I get better PSU (current one is crap.. shuts system down when GPU and CPU are stressed together).
Actually a calibrated Samsung Premium UHD TV. The image is just as good as my 2000$ hardware calibrated 32inch and thats where my contend lives anyways. It is important to circumvent the OS to avoid shifts. I use a blackmagic decklink.
thanks man... glad you have fun with it.
Excellent experiment! I'd really love to be a fly on the wall making direct observations when you fellows produce your work. Perhaps, on the unlikely chance I conjure up shift shaping or find my magic wand, you should lay off on swatting flys or at least flys with a big grin on their fly face.
Thank you... maybe there will come a day, where we will produce an episode as part of a workshop?
Home made Panavision! 😍
Yes..... ultra Panavision 70.... I'm peeing myself already. Camera is on the way
Lets go - That was an interesting video, as usual!
Thanks man!
Awesome idea with that rig and the whole experiment!! :)
Will definitely check out your channel further.
About the DOF, I have a few thoughts:
1. maybe just a user error (your screen showed f3.2 when you put on the camera);
2. f2.8 on the lens would translate to f2.0 in the screen, when the speedbooster is added so f2.8 setting on the screen is actually an f4.
3. There may be a possibility (at least sometimes that happens on canon bodies), that when you remove the lens, or switch off the camera to remove the lens, it opens the aperture wide open automatically, therefore, losing the f2.8 setting.
Those are just guesses, as I also have no other thoughts, why would the dof be different than the SB sversion of the shot.
Thank you...
sure, have a stroll ;-)
1. sure... maybe. Of course, the scenes of the recording are staged after the fact and I didn't care to actually use the same settings on all components... it is more about showing the process.
2. It helps to understand that the lens "creates" the image and the sensor "just" scans it. If the lens indeed had the same iris setting (and I believe it did), the boosted image and the s35 should have the same DOF as it is the same image. A booster just gives a shallower DOF than a native lens, because it allows same framing at longer focal length. The focal length and the image crop are exactly the same for booster and TwinSpin
3. possible... but having done this a couple of times, I believe the Sigma doesn't change focal length if the body is just switched of (unpowering the lens) or disconnected (unpowering the lens as well). It doesn't contain a battery that could allow such an action (as far as I know)
There has been some suggestions here and on Reddit. It might be related to a flange distance that is slightly off.
Thanks a lot for your input Reinis
If you would make a Skillshare course, I would probably sign Up for it.
Your videos are always content wise and quality wise on such a high level
Thanks Carsten.... if there where enough of people that think like you, I would like to do that. But I kind of doubt that ;-)
Love your vides! With regard to the quesiton at 8:59, to my knowledge the speed booster is only supposed to provide the fstop luminance of the S35 equivalent, not the DOF
Thanks man… the question was not about that… the speedbooster makes an equivalent lens according to the factor,... with all the properties including DOF. Thats not what happened here… probably a flange problem.
Its so different compare another UA-cam videos. Film mode all time. Waiting more and longer "Films" from you :) :)
Thank you so much Josh.... encouraging in deed! The only thing I can offer you is my Goldvader Episode... it is the longest and most complicated YT contend that I made so far..... now, real films?!?! Let's keep dreaming :-)))))
I've been deeply researching this lately. I'm a firm believer of owning the gear I really like working with and renting when the production calls for it. I've been trying to figure out the next stage for my GH5S. I've rented the Metabones Speedbooster for it and I like what it does for the field of view and lens options but I feel like I got more distortions in my build using the Rokinon Cine DS lenses that I didn't get on a full frame camera. I'm kind of at a crossroads of purchasing the speedbooster vs a larger format camera such as the EVA1 or one of the other options in the 6-10k USD range. I really appreciate the content you make here, it has definitely influenced my research. Thank you.
Thanks a lot Red... I don't worry to much about distortions, your viewers usually don't care and it is often even thought after... like in wide anamorphic lenses. "La La Land" is a distortion party. All setups have advantages and disadvantages. I will soon get a MAVO LF and I will go deeper into Full Format and boosted medium formats. She is the missing component in my Ultra Panavision 70 setup. EVA1 is definitely a solid all purpose camera for people that work on all front lines and I can recommend her. Maybe you look into the Lumix S when it comes out. I like to have MFT, s35, and FF options in my arsenal... and maybe one day a 65mm ;-)
@@MediaDivision I gotta say I'm extremely excited to hear about that Panavision setup. I feel like subconsciously we relate the distortions from anamorphic lenses to watching a movie yet the distortions from the Speedbooster feel unnatural in a way. I've had both success and minor flubs with it. I hadn't even heard of the Lumix S until now, I will definitely start looking in to it. Seems like a great option for the cost.
@@RedRamzor I can't see any distortions in this test caused by the booster... that was more or less my purpose for the test. I do agree that anamorphic play our subconsciousness to perceive a higher production value. They are mild in the ultra Panavision 70 as it is only a 1.25 stretch, but thats another unique characteristic that separates it from the normal 2x scope
@@MediaDivision Your footage looked quite similar. Perhaps I'm pushing my glass too far which is why I'm seeing some milky distortion in my setup.
@@RedRamzor Thats sounds like the is dirt, fungus or anything else on or in the booster?! Or you use a cheap/bad booster. The Metabones doesn't do anything like that even shot wide open.
6:12 Your editing screen has the size of a planet
Just an ordinary LCD TV… cheaper than most monitors wit 4K HDR and high nits… also, the on a desk the curvature finally makes sense
@@MediaDivision this is a good idea...
Very interesting experiment...
Cheers John
Absolutely incredible man, this channel deserves so much subs
Thanks Tune! Give me a hundred years and I'll make this worth while ;-)
Love watching these from iPhone screen :D
Come on... please watch on a large screen TV with a bear and some popcorn. Especially the newer ones are feature film productions
@@MediaDivision I usually start on the iPhone then move on to PC monitor since on the tiny screen you can’t see what is what. By the way, videos on superfast lenses and anamorphic are FIRE! ❤️
That’s the spirit… yeah… a ton of work went into those. Glad you enjoy… this is possible because of the generosity of people like you. Thank you!!!!
Regarding the depth of field between the speedboosted shot and the 8k, IMO there's a very small difference between them, but could it be that the speedboosted shot was back focused slightly, making the background slightly more in focus? I honestly don't see how the DoF could otherwise be different as mathematically it should be identical to the 8k shot when viewed at the same size.
Edit: thinking about it a little more, I think what most likely happened was that the 8k version ended up being slightly front focused when you were estimating the flange distance
Thanks for the input. Thats what I thought (should be the same)… focus was not touched. Someone on Reddit suggested, that I had a flange distance for the TwinSpin that was slightly off and that could possibly cause the effect.
@@MediaDivision I suppose that the way to be sure would be to do the test on a scene with a very shallow depth of field so it's obvious where the camera is focused and how deep the DoF goes.
Interesting video as always!
@@thatcherfreeman yeah… not going to invest that time ;-) Thanks man
I had the same thought, that there is a back focus issue. Best is to put a rule at the side of the projector in a 45° angle, like that you can see precisely where the focus is.
I have a feeling that if you were to try focusing on something further away with the TwinSplin, you would not be able to get the image sharp. I think you have to adjust back focus for infinity first and then use the Sigma 18-35mm focus to get your subject sharp.
The 8K footage is sharpee and higher dynamic range..
You see the air lighted up when the projector in on ...in the other firmats you don't see it ...😎..
Well done ..great learnings!
Thanks man, but I don’t think that the resolution influences the dynamic range in any other way then giving you a bit more “usable” details in the shadow due to the finer noise and what one would perceived as acceptable. The illuminated air and the difference between the shots is simply a slightly uneven haze in the room.... let’s say... a cloud came by. Sharpness (or Detail like I prefer to call it) is higher in the 8k for sure!
@@MediaDivision haha i was already thinking ..".maybe he was smoking during the 8K shot ..🤣😉
Right... my office is a de facto bong :-)
absolutely outofworld great studied!!!
Thanks Mangesh
Your Dept of field is not shallower course you're not using one big sensor but 2 small sensors... Beautiful video im totally subscribing
thank you and glad to have you around.
The hole purpose of being behind the lens and "scanning" different portions of that projection is to effectively simulate a large sensor. You can solve that with armchair logic for yourself. If you had a Full Format sensor, and you would cut the sensor to pieces, and then you would assemble the images generated by the pieces... would that image assembled in post have another depths of field as the original sensor... no.
It would be differnet if you moved the cam and lens together.
This channel just 🤯🤯🤯
Thanks man
Very impresive test. You have an interesting channel. Finally a serious test of speedboosters versus native lens. One observation though. The shadows were darker and had much less noise with the speedbooster than the 35 and in my opinoon was little better with a native lens as well. I wonder why you didn't use an adapter and corrected the frame like you did with the native lens, or did I miss the explanation in the intro?
Thanks David... I am not sure what you mean. The point of the test is to compare the look of the same object/framing/perspective using a native lens MFT vs boosted s35 vs "real" s35... to evaluate if it is worth it going s35 (or boosted s35). The larger lens collects a bit more light that we compensate with higher ISO on native MFT... so, a bit more noise is to be expected. The real s35 has 2x the resolution, so less light per pixel... the raised noise level is less obvious as the noise is finer... does that help?
Gud Tutorial bro. U open my eye to something Amazing. thanks a million.
Thanks a lot Shae… I'm glad I could do that for you!!!!
Interesting Perspectives, surprised to see such little difference. As you mention... it depends on what application it's for! Only thing that comes to mind with the depth of field... could to be the flange distance?
thanks Richard... yeah, flange is the No 1 candidate at the time.
Excellent, love your content mi amigo, and you remind me of Humphry Bogart. (not important, I know)
Thanks.... recognising faces is other faces is very common, and it doesn't mean I look like Bogart, but that your brain uses the same neuronal network for both of us... a bit like with Pareidolia...I wish I knew what this is called in science but I can't find it.
I think your depth of field was only greater because the lens wasn't physically attached for the "super 35" sensor. Had it been attached, your focus would have been in the same range as it was for the speed booster.
Thanks for the input… yes, many here suggest that a flange distance "error" might be the reason. I am surprised how this very samll change does effect the DOF in such dramatic way, and why nobody releases a mount with flange shift (not tilt shift) to do this on purpose?!
Media Division That’s a cool idea! I’d love to see something like that come to fruition.
I would stack that on a speed booster for bokeh galore lol
Def gonna try to recreate this super 35 with my gh4 and speedbooster so please put out more content like this. I’m a super low budget filmmaker and I love stuff like this that I can try to recreate for little to no cost.
The DOF difference beetween 8k and booster footage is becouse metabones adds around 1 EV - and it shows it as i.e. f1.2 on f1.8 lens (you used sigma 18-35 f1.8). In reality its f1.2 only in terms of transmission, not DOF. So you probably set the same apurture on both tests, which in fact, should be f1.2 on booster. IF for the test you set sigma to f1.8 on booster, it was T1.8 (more or less, probably T2.0) for transmission, but f2.8 DOF :)
Because of a lot of image process details are not given, with best information I can gather from your video about your mechanical set up and image stitching and cropping(assuming you are not maxing out the display limitations of the resolutions), a rough theoretical calculation puts the depth of field of the boosted image at 1.52 times of that of your spintwin, that difference is mostly coming from you stepping down the aperture after the boost. (Maintaining the same f-stop at 2.8 means you have stopped it down since adding the speed booster would have automatically set it at f2 if nothing’s changed, if you have shot the boosted image at f2 then the dof difference would have been only 1.08 so practically identical with the boosted image teeny bit deeper)
Great video, thanks
Thanks Sai.... my pleasure
Excellent and very valuable! Thank you!
Thanks Piotr!!!!!
Subscribed...notifications turned on. Good stuff 👍🏻
Thanks Viral! Nice to have you around
I love this, this is great. If I were going to (finally) invest in a Metabones speedbooster for my GH5, which "mount" would give me the most utility/variety of lens choices? Canon EF mount? I just watched your video on the Mamiya lenses so my wheels are spinning.
Thanks!
Thanks Robert, I thinks that EF is the most versatile and allows most options unless you already have a large set of F mount lenses
Hi, the SpeedBoosters usually would make sharper image, the depth of field is now enlarged with a speed booster when compared with the "original" setups.
Wy would a speed booster deliver a sharper image? It doesn't. It gives you a brighter image. Image might be sharper when you starve the sensor from light. The additional brightness reduces grain and therefor perceived sharpness.
Media Division sorry there might be some wording issues with my bad English :( however I found the official answers about the DoF and SpeedBoosters: The short answer is Speed Booster® on an APS-C sensor gives essentially the same depth-of-field effect as if a full-frame camera body were used.
The long answer is complicated. If we are referring to depth-of-field in the mathematical sense, that depends on the aperture, magnification and circle of confusion (CoC). Magnification in turn depends on distance and focal length. The 50mm lens now becomes a 35mm lens which behaves very differently in terms of perspective. The question is, do we still keep the distance the same? Should the CoC be kept the same? There are many missing variables we need to choose and fill-in before we could get a meaningful answer. When people claim Speed Booster® does not change the depth-of-field, they usually neglect to state the implicit assumption that the distance is kept the same (thereby changing the object size) and the CoC is kept the same. The same logic would lead to the conclusion that an APS-C camera has the same depth-of-field as a full-frame camera, too, which under the same implicit assumptions is mathematically true (the depth-of-field formula is format-size-agnostic, after all), but with which many people would disagree from practical experience.
However, when most people ask about depth-of-field, they are not interested in mathematics, but rather, they are after a certain kind of shallow depth-of-field "look". If this is the case, the short answer above applies.
OK… that is all obvious and right and nothing new to us.
A 35mm and a 50mm and a 100mm lens have all the same "perspective" but another field of view. Perspective is only "generated" by the subjects distance. A lens with a booster has identical perspective compared to the lens without one. DoF is obviously diffrent.
This calculator answers every possible question in that regard.
www.pointsinfocus.com/tools/depth-of-field-and-equivalent-lens-calculator/
Thanks a ton. Love your content
Thanks Aman… it is my pleasure… and thanks for hanging
It would seem that as the booster compresses the light it also compresses the circles of confusion. Tele compressors are known to improve optical transfer function performance, or how the optics project points of light from the object or scene onto the sensor.
en.wikipedia.org/wiki/Telecompressor
en.wikipedia.org/wiki/Optical_transfer_function#Definition_and_related_concepts
en.wikipedia.org/wiki/Point_spread_function
en.wikipedia.org/wiki/Circle_of_confusion
:)
Thanks Scott… of course it does… compared to a native lens. But it shouldn't compared to a larger sensor with the same FOV. It seems to be an issue with the flange distance
shut up and take my like! =)
hahhahahahahhahaha… thanks man!
I know you mentioned that you record with the GH5 a lot for these episodes - what lenses do you tend to use for these? And amazing lighting btw!
Thanks... of this one are shot on EVA1... using a Sigma 18-35 EVA1 looks quite similar to GH5... of course EVA1 looks better and has less noise. Low noise level is important for UA-cam and noise reduction is ineffective in long form formats
I triple dare you to apply this method to your Mavo LF. I want to see down-sampled 12k just for pure masochism.
Been there...
ua-cam.com/video/HN29l3ouR64/v-deo.html
Not with the LF though... we are working on something much bigger with her.... much!
@@MediaDivision You just became my new favorite UA-cam Channel. A+++
@@jamesjenkins33 thats the spirit ;-) thanks man
I think the sensor stabilization is playing a role here somehow. In the other video when you tried to emulate the large sensor with the EVA1 there was shift in perspective. I do not see it here except the bending on the sides but not apparent change of perspective - which doesn't seem right. I felt like a larger sensor apparently gives the image more volume perspective wise similar to a on optical space compression, objects that are further away look smaller by comparison to a M43 sensor. I can see this shift in perspective in GH5 vs GH5s comparison, where not only the bigger sensor gives a wider image, but the perspective is drastically changed, like in this image: i.ytimg.com/vi/ql8szWeElXQ/maxresdefault.jpg And we can not see this in the center on any of the images. Except for the resolution when you pixel peep, I do not see a difference this time, except for the image distortion on the side extremities..
Thanks for your input... the IBIS is turned of and the camera is not moving during recording of the segments. There is no shift in perspective as the camera is not moving. In the EVA65 test I used the same lens to eliminate differences caused by the lens. To get the same framing one has to move the lens... resulting in a different perspective. Here, I achieve the same framing by using a much wider lens on MFT. As I can proof by the overlay at 8:25, there is no magical magnification etc. using a larger sensor. The wider lens produces more barrel distortion... that is all. In your image he would just have to zoom slightly in on the GH5s to get the exact same magnification/perspective. If he moved in, that would change the perspective.
Great video, I think that the video looks so similar in sharpness because of the lens, it is being bottle necked
Thanks... i think the vario 12-35 holds up great and there is little difference in detail unless you zoom in.
...and the Sigma 18-35 is as sharp as modern glass gets. It can resolve twice the resolution with no problem
Media Division oh, i havent ever used that lens but ill take your word for it! Great job on the video, very interesting topic. I also liked your other videos
@@43connor3 Thanks man!
Clarification on dof
8:45 The 8k footage has a shallower dof, I have a theory, that I still not have confirmed, but neither have discarded:
Starting from the concept of sharpness, and we all can confirm, that is achieved when the light "rays" from the lens realign at the sensor/film focus plane. My theory is based on the pixel density, so if we have more pixel to record the disaligned rays, even if we downsample later the recorded out of focus has a much better quality (same like watching this video for me on a 1080p screen at 4k/8k resolution), the sampled information of the out of focus have much more data so the debayer process can retain it as much as possible.
Or maybe when the lens is actually not attached to something the aperture put itself at f1.8 alltime.
That’s one wild theory man ;-) .... but hey... the DOF difference has certainly nothing to do with compression or sixes density. Whater, however... definitely optical. A flange distanced that is off a bit seems to be the most likely explanation. If I tried to focus to infinity , it possibly would not work. Thanks for your input
A shot at answering your depth of field question:
1. The 18-35 sigma art lens is a APS C lens and not a full frame or super 35 lens so you don't get as much of a boost light wise from the speedbooster because the lens area of light is smaller, therefore decreasing the shallower depth of field you would achieve with a lens meant to cover full frame or super 35 sensor.
2. super 35 8k is still a wider sensor and naturally so, a wider lens/image projection than 5k boosted.
3. i'm no expert
oh and 4. flange distance as a commenter previously mentioned.
beautiful videos and experiments you do. Thank you.
Hmmmmm, it does'n work that way.... with a .71 speedbooster it doesn't matter if you through on a asp-c or FF or even medium format lens. The light gathered by the sensor is in all three cases is the same.
The 18-35 covers super35.... super35 and ASP-C are used synonymously today
EVA1 sensor is Super 35 sized (24.60mm x 12.97mm) so even smaller then (some) ASP-C
both formats range in size.
"Sensor sizes range from 20.7×13.8 mm to 28.7×19.1 mm, but are typically about 22.5×15 mm for Canon and 24×16 mm for other manufacturers"
That is the most suggested reason...yeah. Thanks a lot Ejike :-) it is my pleasure
You don't need an 8K display to enjoy the image quality of this video. I only have a FullHD 32" monitor, and I can see a lot more detail with the 8K stream. That's the reason why I keep the UA-cam player constantly on the 4K stream, even with a 1080p display.
Absolutely… of course you don't see anything of the resoluten - the perceived improved image quality is because of the much higher bandwidth that UA-cam grants 4K over HD contend. The jump from 4K to 8K is much smaller and barely worth it in terms of bandwidth.
@@MediaDivision So, the 8K stream we see is not actually 4K or 8K, just better Full HD signal? Now that I think about it, UA-cam's 480p stream is a lot worse than what we used to see on SD televisions in the 90'. UA-cam is basically labeling a stream as 1080p or 4K, but it's always lower than that in terms of image quality and detail...
Thanks for responding...it's so rare nowadays for a UA-camr to interact with his viewers. Wish you the best! Keep up the great content :)
no... the 8K steam is 8k resolution, but if your display has 1080p that is that. You still get a "better" 1080p image as the data you receive has higher bandwidth and is scaled (oversampled) locally. 480p UA-cam is much harder compressed and uses a fraction of SD TV . Just like 1080p UA-cam uses the same codec but much higher compression than BluRay... giving an image that is barely worth calling HD. Image quality and resolution correlate but are not the same. A 4K video with half the bandwidth of a HD video is inferior in image quality no matter the pixel count unless you use a codec that is much more efficient using the given bandwidth (h264 vs h265).
Always man... thanks a lot!
who NEEDS 8K?
Nobody NEEDS film... if we can get over that we can start imagining the possibility and new ways to use the medium in ways we couldn’t before. Like distance agnostic giant video touch screens. There you would need way higher resolution then 8k or even 16k... watch this for inspiration ua-cam.com/video/EUjUxKvXgpw/v-deo.html
Amazing work!
Did you do all this by yourself?
Thanks Saleh… yes of course… I think I got way better since, watch some of our new stuff. Got a little help with the latest two.
Thumbs up and subbed! I have a minor question in regards to simulating film on digital m43 cameras: What would the focal length of 40mm in super35 translate to to in MFT format?
Thanks... and nice to have you. That would be 28mm. You can always use this very convenient online tool: www.pointsinfocus.com/tools/depth-of-field-and-equivalent-lens-calculator/#{%22c%22:[{%22f%22:11,%22av%22:%228%22,%22fl%22:40,%22d%22:3048,%22cm%22:%220%22}],%22m%22:0}
@@MediaDivision You´re a hero! Awesome fun tool! I´ve seen calculations floating around the web saying that if you multiply your m43 focal length by two you get the full frame sensor equivalent(roughly). And Super35´s sensor is even larger right? So to get a focal length AND Fov as close to 35mm or 40mm (on super35 to simulate wes and other filmmakers) would this still be a 28mm m43 lens? Apologies if my question is confusing, i´m pretty bogged up thinking about this! The reason im so passionate about it is because I would like to pick up a budget cine lens for m43 like the meike to practice with and get as close to the actual picture you would have on super35 at 35mm or 40mm. I think this topic is super interesting as well haha!
No problem... s35 much smaller than 35mm (vertical running film instead of horizontal running film)... here is a size comparison I did ua-cam.com/video/rUOYXjH6y7w/v-deo.html.... in the link I send before (paste copy... don't click), you can scroll down and you see a list in sequence of size with formats... all with the calculated values. S35 is 1.39 crop factor. That is spot-on half way between FF and MFT. Now 28mm on MFT equivalent to 40mm on s35 should make sense
Awesome experiment. Could you possibly do something I've been searching for and have yet to find: A real-world projection of GH5 footage onto the big screen - as in cinema screen. It would be great to see how it compares to professional cameras like Red and Alexa. So far, every "filmmakers" review of the GH5 gives no real indication of how it would look if projected in a large movie theatre. They all seem to be thinking "broadcast quality" which is not cinema.
Thanks a lot... well, as I don't have access to a cinema I can obviously not do what you are asking for, and it wouldn't make much sense either. The specific "look" of any cam has nothing to do wether the image is projected or consumed on a TV. If you have the same field of view of any screen and comparable ambient light to peak brightness, appearance of the image should be the same.
Interesting. 👍
indeed... thanks
Found this one year later. I need to say that thank you very much doing this, very very interesting. I have a question which perhaps you would know the answer to. The GH5S uses a wider part of the sensor and so I was wondering if a super 35 lens would vignette when using the 0.71 x speed booster. I have the 0.64 times and for sure it vignettes! Thanks in advance!
Pleasure man....there is no simple answer to that... it very much depends on the lens. longer lenses cover better. Example: a Sigma 18-35 covers Full Frame from about 22 upward... so, it would cover MFT from 22 even with a 0.64. As a 0.64 is not all the way FF to MFT it will certainly still cover GH5. to be sure with lenses that barely cover you will have to try.
Media Division thanks so much for getting back to me, really appreciate it. It’s the gh5s we use which I think shoots a little wider on the sensor than the gh5 if I’m not mistaken?
I had in mind perhaps the sigma 50-100 T2 lens actually. Or perhaps the new dzo film 20-70 or 50-125 Pictor lenses.
I wouldnt mind at all to buy the 0,71x booster instead if that were to help things.
Thing is we operate out of Bali, so access to try these things is limited. In a covid free world we would go to Singapore to do hands on testing. Sorry to pick your brains, but your brains are valuable! Thanks again in advance!!!
@@fritzproduction7113 You are asking me about gear I don't posses. Should the 50-100 work using a 0.71... sure. Why don't you go into the facebook groups with the users that do possess the gear? That way you can be sure facebook.com/groups/pansonicgh5
Media Division thanks so much for the lead! Will check it out. Appreciated!!!
Wow! Very interested in anamorphic lens! )
Hey Cyril.... give me a while. I want to nail it before I release something... but the fitting camera will arrive very soon.
After watching your videos. I am now going to purchase a second-hand GH5 with lots of extras. Is there anything I should look out for? With my drone ( mavic pro air) and that camera, I want to make little crazy videos with the music I have made over the years.
Look out for power lines 😝no, sounds like a solid combo for the enthusiast. Some day you might want to add a Gimbal or step up to the Mavic Pro 2 as the image is so much better
Hello my friend... I have been looking at all the new Panasonic announcements and saw a flood of GH5 cameras is real. I mainly shoot on Blackmagic cameras but to be honest when traveling more and more, and learning more and more about lighting - I realise MFT is actually not that much different from the super35... Especially if you consider good lighting and composition. Really tempted to get a GH5 for at least my UA-cam stuff. What's your thinking on MFT overall? Is it here to stay? Again if you compare the MFT to Super35 it's not as big of a jump as Super35 to full frame. Would be interesting to hear your thoughts.
MFT is just fine... but technology moved a lot since GH5 days... the GH6 is way ahead. Here is our test. ua-cam.com/video/jUd_I8Z3Iqw/v-deo.html
With the brand new S5 mk2 Panasonic finally introduced a good auto focus and even better stabilisation for almoust the same money, so if you are not invested in MFT lense... the jump to full frame is n option.
Richtig richtig cool.
Danke dir!
m4/3 is for on the go. for studio productions you can use a bigger camera. the m4/3 system is realy light and small the lenses are a lot smaller as the 35/FF sensors for handheld work I love my lumix.
stating the obvious?
Looking this episode make me think to capture projector projection as yours in slow motion, maybe you can do such test in 120 / 240 fps? I'm curious how frames repeats in fluid motion as realtime (repositioning) in time (that we can't notice). I'm still beating with motion stutter (judder). Like make raw DSLR records in 60 fps but not for slowmotion, just want to achieve same motion blur and feel as it for 24fps@180degree but utilize display rate properly (maybe I can write some DirectX special rendering filter plugin). It's like prevent computer display that works with 60hz refresh rate in most cases, of simple duplicate (repeat some frames) to feel gaps. I'm think in more depth, like to simulate real cinema look with flashing, so to grab digital file and try to run at 60hz, there is going to 24 frames mapped stretched to 60hz refresh rate. I think if calculate it right like telecine projection (in analog world) by small delay of presented interframe vertical motion blur on repeated frame and 1/2 of pre and post frame. Also tape seems have black gaps between frame that also infuences on time, also it may present some blinking (flashing) inertia effect on projector scope. So idea to make creators life in digital world more wet natural. than that difital dry look. At beginning it can be simulated in after effects.
That would be a test with an audience of one... you ;-). The rest of your comment is a bit incoherent (I guess due to language problems) but let me tell you this: Old projectors like this one have a shutter, so in high fps you would see the shutter opening and closing. As you can see in my video, the light from the projector is flickering. That is on purpose. I set it to 12fps while filming in 24fps.... the flicker is the shutter at work. Higher frame rate will just give a flicker with lower frequency. As the movements are an illusion, the exposure time of your camera will not change the given motion blur of the projection (I don't know if that is what you wanted to know?)
@@MediaDivision I though about it later, yep shutter, almost forget about it :) anyway better to see in motion to imagine how 24fps may be properly shown on 60hz display. Idea is that guys shot 180 degree shutter both for 24fps and 60fps (not slowmotion) so 60 present less motion blur due to more shutter samples. So to preserve 24fps look by motion shutter degree should be 270 or 360 ? :) And second problem is if footage already 24fps showing on 60hz display, why just do 3:2 pulldown etc and show each frame 2.5 times, if we can simulate something like in projector (sem interchange by fading in/out of shutter) :)
@@protasov-by I really having a hard time to understand what you want or need. at 180° shutter, a 60hz diplay looks fine shooting 24fps...and of course 60fps. A REAL ANALOG projection should look just fine, too, as the image is stable. With beamers you have to take the tech in account. Most beamers are DLP and that have to be shot at 360°.... If that is the question (?!) we recreated the Goldfinger intro with a segment about that problem herfe: ua-cam.com/video/iKuf7VSGEGI/v-deo.html
@@MediaDivision Ok I try again, I want to shoot 60fps footage and show at this speed, but preserve filmic look (so mean 180 degree shutter for 24 not same for 60 fps, shutter degree should be more for more blur), here is problem displaying 24fps on 60hz monitor (judder) ua-cam.com/video/RltHHoDd8gU/v-deo.html at the end of video there is footage examples: i.e 1/50 shutter (180degree) show that juddery effect. The second question is how to conform footage to 60fps timeline when editing from 24fps source (if we can't reshoot it at 60) but without frame bleding or optical flow in After Effect, that cause presenting each frame 2 times 60/24 = 2.5, so this thing ruins fluid motion, ofcourse we can set out displays to 24herz or 48herz shoot and produce in 24fps, thus will be better, but users don't do it and will watch on 60fps or even 120fps displays... So, I think travel to world of shoot and edit in 60fps to 1:1 frame sampling, and use frame (fading) trick, that I try to develop for Premiere, thing going to be based on real projector flickering work :) so deal to understang full process in slomo. Sorry if bother much but seems nobody feel problem of juddering?
This would be an interesting way to design a camera.
Thats what Arri did with the Alexa LF and the Alexa 65
Since we are talking about micro-four-thirds what I your thoughts on black magic cinema pocket camera?
Well, I never used it… but it surely looks good on paper and the image looks very good, too. As always, it comes down to what you want to do. The BMPCC 4K is useless for events, run and gun etc for the lack of a handy codec. You just can't produce several Terrabytes of data with every job, if you have one job after the other. If you only do narrative, it looks like the best option for the best price at the moment. If you have to do both, the GH5s looks like the best MFT at this time, just because you can cover both, and the IQ of ProRes HQ does not trump Panasonics 422 10bit h264 or even 420 HEVC. But that is just my opinion.
Nice work, and informative. Regarding the depth of field differences between the 8k and 5k speedbooster images, are you sure you were using the same aperture for both? E.G. F2.8? When you add a speedbooster, it automatically adjusts the reported aperture -- so the same physical opening that was reported as F2.8 without the speedbooster will be reported to the camera as F2 once you add the speedbooster. If you then take the additional step of closing the aperture to F2.8, then the physical opening in the lens is the same size as the opening would be at F4 without the speedbooster, accounting for the deeper depth of field. Long story short -- to compare like for like, you should've shot the 5k speedbooster version at F2, since the speedbooster transforms the Sigma into an F1.2 lens for the purpose of what information is reported to the camera. If you had done that, you would've gotten an equivalent depth of field.
Thanks man.... yes, pretty sure it was 2.8 all the time. I know that i could have gotten the same DOF on native... I only have a wide enough native that is f2.8. Opening a lens to f1.2 will make it considerably softer and that would have voided the purpose of this test. The s35 DOF eems to have been introduced by a flange distance that is off.
@@MediaDivision I didn't mean opening it up to F1.2 (which would've actually given you a shallower depth of field with the speedbooster than the "8k" sensor. What I meant is that with the speedbooster you should've opened it up to F2, which is the same physical aperture 'hole' size in the Sigma 18-35 as F2.8 without the speed booster. Actually, I'm not really sure how you set the aperture on the Sigma in for the 8k sensor recording without connecting the lens to the camera given that aperture control is electronic, not manual -- did you pre-set it somehow with the lens connected to the camera? In any event, if you had just taken the lens after you finished the 8k sensor test, screwed on the speed-booster, and connected it to the camera, it would've registered an F2.0 aperture reading initially. By subsequently stepping down to F2.8 at this point, you're narrowing the physical aperture of the lens relative to the size it was during the "8K" sensor testing, so you're going to get a different depth of field.
Andrei Voinigescu there seem to be some misconceptions. The lens generates the image not the camera. Speedbooster MFT and s35 should look the same at the same aperture. Sigma stays in the last settings it was when you disconnect.
@@MediaDivision, well yes and no, you introduce an additional set of optics with speed booster and say that lens makes the same picture which isn´t the case. you need to match FOW with dof.
1) it´s like stitching a shallow dof panorama - you can shoot 8 photos with 50mm f1.4 lens and have huge fow and shallow dof, and than put a speed booster and wonder why dof on 8 shot looks shallower despite being shot with same lens. the difference is you stocked just 2 images, and difference between stocked images and speed booster is minuscule, that´s also the reason why pictures look so close.
2) another point - resolution. high resolution affect what you consider being out of focus(circle of confusion, etc). i mostly use background blur/bokeh balls as a reference, and they are damn close in both shots.
3) Andrei has a point. if you want to get same FOW and DOF on different formats you need to adjust both focal length and fstop.
MFT 12mm f2.8
APS-C 18mm f4
FF 24mm f5.6
on all three formats, you gonna get absolutely the same image, give or take.
by applying speed booster to your lens, you´re changing the lens. 18mm f2.8 is no longer like that, it behaves like 12mm f2
p.s cropping changes your fow, so you can´t talk about precise measurements.
if you don´t believe me, make some tests.
a) shoot FF 24mm f2.8 and MFT with same settings - the crop of the MFT will fit perfectly into FF picture (thats the case what you mention the lens makes the picture dof is same, blurr is same, but FOW are different, hence 2 different pictures)
b) shoot FF 58mm f5.6 and MFT 24mm F5.6 (the FOW matches, but the dof will be different because of different focal length)
c) FF 58mm F5.6 mft 24mm f2.8 (everything matches, because of math!)
@@vitaliikharkovchuk6384 duuuude... thak you for your long answer but this is way to long. Reading over it just scratching the surface ... I disagree with many point and some are just not about what I'm doing here (for example I'm NOT trying to match DOF but maintain f-stop for boosted and s35 ... and I don't own a faster native lens wide enough to match).
To go through this point by point with reasonable answers would take me about half an hour.... I rather share the limited time I have for this channel with a broader audience. Thank you for understanding.
Cool concept!
Thanks man