when he touched the lense, it kinda tilted, "first panic attack" then he moved it closer to the edge "second panic attack" your are living on the edge Viz xD
25:40 The two linearly polarised filters are arranged at ~90 degrees out of phase to kill all the light coming from the reflection / specular highlights. If they were aligned they’d be in phase and all light would get trough.
Iconasys has photography turntables that can automatically capture up to 8,000 images per rotation. Iconasys photography software can also remove the background automatically from all the images. Has Agisoft Metashape integration too. Makes the whole process much easier and quicker.
Hey, it doesn't seem that you answer a lot of these questions, but I'll give it a try. How would you get pictures of the under side of the object? just placing it sideways on the platform won't work because the tracking markers won't change rotation. Taking pictures of the underside without the markers might work, but you would still need to give the program reference points. I know you can unlock the other tutorials on Patreon but i just have this one question. Anyways, thanks for the awesome tutorial!
Hi, sorry I'm late to this conversation, I'm a brand new student in photogrammetry. What about creating a seperate model with the object lying on its side, and then using the reference markers to get the 2 objects on the same scale, merge them together with any modelling software? Could that work, maybe? 🤔
How high ISO can be used for photogrammetry in order to make an 3d-model for wood carving? I have a very light sensitive Sony A7 III camera, but my ring light is only led and not a flash with strong light so with a rotary table I need to use high ISO in order to get higher shutter speed.
You mention aperture in your description, but consider this- if you’re scanning smaller objects (like what you’re doing with the turntable), your depth of field doesn’t need to extend more than a couple feet. You’ll have the sharpest image and no loss of DoF in your turntable space. F8-16.
Awesome tutorial, thx a lot! But as a photographer I want to give an advice - read about optical diffraction. You don't wanna have an apperture too small, for the majority of lenses it must be around f/10-14. The thing is at some point sharpness of the image
Hey. Thank you. Im aware that smaller aperture sizes reduces sharpness and I mistakenly left that out of the video. But in this case, a long depth of field is more important than sharpness in my opinion. Out of focus parts may end up in the textures and I’d rather have the whole texture a little unsharp rather than having islands of very blurry texture. Its a trade off.
@@TextureSupply of course, good dof is a great deal! 🙂 Btw I wonder is there a way to monetise this 3D models? I’m not so good at 3D modelling, and right now I don’t wanna dive deep into this topic (I mean learning Blender or 3D Max or Maya etc). So is there a practical use for raw photogrammetry models?
One thing to consider is that after certain aperture, difraction starts blurring a little bit your images. So maybe not going as high might be better (first testing that you've got enough depth of field using lower apertures, not just cranking it up all the way to f22). Amazing content as always.
Eduardo López that is true and sadly not possible to overcome, usually the sharpest aperture are around 8-11. However if we're using that low aperture for photogrammetry we will have shorter depth of field and that will cause a lot more problem than the sharpness we lose with a smaller aperture. Having shallow depth of field will make your textures more or less unusable år you will manually have to mask out images for texture creation as, at least agisoft, will otherwise pick parts of the texture from the blurred part of the images. So it's either sacrifice a little sharpness or add a lot of manual work. 😔 (i understand that you probably know that already but responding so that others who might not will know why).
Then perhaps ideally do not use a full frame camera, as depth of field (DOF) decreases with the size of your image circle (and focal length). Meaning an aperture value of 5,6 on a 50mm lens on a MFT sensor, has deeper focus (less depth of field) than that of the same aperture value and lens on a full frame sensor, unless it all comes down to field of view anyway, meaning that the DOF on a smaller sensor is “perceived” as deeper because of the narrower field of view? However I do suspect that a smaller sensor camera would be preferable for sharpness purposes, all else being equal, as the quality of the lens obviously play a large role as well. And so if focal length does not matter for the end result, then definitely use a wide angle lens, as the wider your angle of view the deeper your depth of field will be, meaning that an image shot on the same camera (in your case Sony a7R) with the 35mm lens at f. 1:8 will have significantly deeper focus than your 105mm at same aperture.
Hey there, great tutorial! I'd like to see the others.. just a question before buying the patreon: did you do all the photogrammetry course? I didn't see it on your patreon page. Thank you!
Awesome video. I noticed you did not mention 3df Zephyr. I find it has a much more user friendly interface and workflow. Many thanks for this outstanding tutorial.
nice job man... might mention that you plan on going through a lot of common mistakes and showing solutions/alternatives. Almost closed video thinking this guy doesn't know what he's talking about haha I find that as much I think I might know.. There's always a chance somebody knows something better.. something new. You mentioned link to model for filter holder though? I don't see that. I've seen a few and wanted to print but they are of many pieces... I love the idea yours only has one!! :) Again.. great job! Any plans for updates? You could get away with doing all the same steps but later methods/techniques.
8:00 *Little depth of field. The higher your f.stop a.k.a. aperture number, the smaller your lens aperture opening (as measured in fractions or divisions of the total lens diameter) the more “depth” you have of your “field” of view, in other words, a higher aperture value decreases your lens opening and increases your subjects focus area. As a general rule, the higher your aperture number, the sharper the image. Also, equally important in this case, the wider your field of view (shorter focal length), the deeper your focus, or depth of field (a.k.a DOF for short). And to get even deeper into the subject, the size of your image circle also matters, meaning that the physical size of the cameras image sensor also matters since this effects the field of view. Meaning that smaller sensor cameras has a narrower field of view, hence a deeper focus across the image area. Therefore, not taking optical quality, noise, bit depth, resolution etc into consideration, I would assume a smaller sensor (MFT or APS-C) mirrorless camera with a wide angle lens would hit the sweet spot for photogrammetry?
Hi! I see in your intro that you have also created a model of a plant with many leaves. I have been trying so hard to photoscan a plant and have a nice 3D model of it, but unfortunately never succeeded. However, I am not an expert... Is there any chance you could help me or show in any way how photoscanning of plants work? Thank you!
Best 3D Scanning TH-cam video so far! the good part is you use Sony a7 :))) but got a few question here! 1: how can I get film for the flash? I have heard you can remove it from broken lcd , right? 2.double ISO wont affact IQ for RC ?I also have heard you should not go above iso 100 for photogrammetry processing! 3. what is your focus setting ?Is it wide?camera dose not miss focus in wide mode?
Not sure exactly what you mean. Black out in terms of painting all walls black and empty it, absolutely. That'd probably help a lot, if I could 😀 Black out in terms of killing all ambient light however, makes no difference as the fast shutter speed and small aperture ensures no ambient light is captured.
Title says "part 2-5 in description". Description says "All videos are accessible on the Viz.Guru Patreon page, www.patreon.com/vizguru.". No fake at all. 😉
Bro, thank you very much! Exactly the advice I've been searching for the past 5 days, even though I don't need even 10% of the quality you're achieving. All I want to do is to capture parts in 3d from which I can design other parts to fit on so I can 3d print them, and all I have is an android phone with free photogrammetry software. (The free software doesn't allow me to take the pictures seperately with a proper camera.) Gonna buy a cheap turntable and a cheap light box with a dark background, print a sort of cow pattern for the turntable, 3d print a stand for the phone, and that should do the trick!
Wouldn't you have been better off using two cheap softbox lights instead of that ring flash? You could have those for $100 on Amazon and I think they'd give a much softer more even lighting from two sides. I'd put the object being photographed on a reflective mat white card. to get some light from below. Other than that great tutorial.
to shoot outdoors surfaces do you need to shoot in hss? or just connected to the camera is it enough? I understand that connected to the camera the hss function is not compatible
First of all, Thank you!! it is wonderful. Plus, could you please share your resources if there is, any resource related to photogrammetry in technical side? Deep learning, computer graphics course, any lecture records, any book? Any help would be appreciated!!
Hello sir I'm a big fan of your fstorm work specially in the facebook group. However, as a student I cannot afford monthly subscription so I decided to try the alpha ver. since its free and wanted to ask if it still worth getting the old version? Thanks and Godbless
Thank you for video! Tell me please, what focal lenght lenses do you usually use to shot (scan) such small objects? And what focal lenght lens did you used in this particular case? Also would like to ask, is it ok if use to shot (scan) object like in the video with the lens 50mm (for example Sony FE 50mm F2.5 G ) ?
I've tried every combination. IMO Small aperture is a better trade-off than ISO. And I think the ISO capabilities of this camera is highly overrated. It may be better than many other cameras, but the images get noisy nonetheless.
Iconasys has photography turntables that can automatically capture up to 8,000 images per rotation. Iconasys photography software can also remove the background automatically from all the images. Has Agisoft Metashape integration too. Makes the whole process much easier and quicker.
Because in my opinion short focus is worse than a little soft image. With optimal aperture (9-11 ish) there is higher risk that out-of-focus areas will end up in the textures, causing random parts of the 3d model to have blurry texture. Then it's better to have half-decent sharpness all over and sharpen in post. A soft image caused by small aperture can be sharpened, but blurry islands caused by DoF can not be easily fixed.
Hi, I'm totally new to this, but my thinking for the bottom is to lie the object on its side, start capturing a totally new model, and then merge the two objects together on the computer using reference points. Not sure if it will work, gonna try it.
@@TextureSupply I just signed up (base level) there to get access to them and it's now asking me to go from the base level up to the Top level. Problem is it says could take up to 7 days to process my original payment (PayPal) before I can upgrade. Seems like a convoluted process Patreon has created making it difficult for a new member to simply upgrade to see desired content. ?????
Here's the message I get when trying to upgrade: You can't make changes to your pledge because you have a pending transaction from February 3, 2022. Pending payments typically take about 7 business days to settle. You'll be able to change your pledge once this payment is settled.
@@Greynerd on patreon as stated in both the video and description. :-) However I've been planning on making them public outside of patreon too, just haven't gotten around to it yet.
I'm new to photography. I want to do a 3d model of my dog. It has short black hair. I dont need color texture but detail. Would doing black and white photos be good?
Anyone else had a panic attack when he moved the lens close to the edge of the table at 5:32?
Haha it makes me a bit uncomfortable too :D
when he
touched the lense, it kinda tilted, "first panic attack" then he moved it closer to the edge "second panic attack" your are living on the edge Viz xD
Just wanted to comment on the same thing... very stressful moment. xD
25:40 The two linearly polarised filters are arranged at ~90 degrees out of phase to kill all the light coming from the reflection / specular highlights. If they were aligned they’d be in phase and all light would get trough.
Wow it's nice to see an extremely professional tutorial on TH-cam. Thank you so much for this it helped me out a lot.
I miss the free STL file for the flash filter.
Do you still have this course available on Patreon?
Iconasys has photography turntables that can automatically capture up to 8,000 images per rotation. Iconasys photography software can also remove the background automatically from all the images. Has Agisoft Metashape integration too. Makes the whole process much easier and quicker.
Hey, it doesn't seem that you answer a lot of these questions, but I'll give it a try. How would you get pictures of the under side of the object? just placing it sideways on the platform won't work because the tracking markers won't change rotation. Taking pictures of the underside without the markers might work, but you would still need to give the program reference points.
I know you can unlock the other tutorials on Patreon but i just have this one question. Anyways, thanks for the awesome tutorial!
Hi, sorry I'm late to this conversation, I'm a brand new student in photogrammetry.
What about creating a seperate model with the object lying on its side, and then using the reference markers to get the 2 objects on the same scale, merge them together with any modelling software? Could that work, maybe? 🤔
How high ISO can be used for photogrammetry in order to make an 3d-model for wood carving? I have a very light sensitive Sony A7 III camera, but my ring light is only led and not a flash with strong light so with a rotary table I need to use high ISO in order to get higher shutter speed.
You mention aperture in your description, but consider this- if you’re scanning smaller objects (like what you’re doing with the turntable), your depth of field doesn’t need to extend more than a couple feet. You’ll have the sharpest image and no loss of DoF in your turntable space. F8-16.
Great video, where can I find the 3D printer file for the device you made?
Awesome tutorial, thx a lot! But as a photographer I want to give an advice - read about optical diffraction. You don't wanna have an apperture too small, for the majority of lenses it must be around f/10-14. The thing is at some point sharpness of the image
Hey. Thank you. Im aware that smaller aperture sizes reduces sharpness and I mistakenly left that out of the video. But in this case, a long depth of field is more important than sharpness in my opinion. Out of focus parts may end up in the textures and I’d rather have the whole texture a little unsharp rather than having islands of very blurry texture. Its a trade off.
@@TextureSupply of course, good dof is a great deal! 🙂
Btw I wonder is there a way to monetise this 3D models? I’m not so good at 3D modelling, and right now I don’t wanna dive deep into this topic (I mean learning Blender or 3D Max or Maya etc). So is there a practical use for raw photogrammetry models?
You mentioned that you could share the printer file of the flash filter support. Do you have the link for I can download? Thanks!
One thing to consider is that after certain aperture, difraction starts blurring a little bit your images. So maybe not going as high might be better (first testing that you've got enough depth of field using lower apertures, not just cranking it up all the way to f22). Amazing content as always.
Eduardo López that is true and sadly not possible to overcome, usually the sharpest aperture are around 8-11. However if we're using that low aperture for photogrammetry we will have shorter depth of field and that will cause a lot more problem than the sharpness we lose with a smaller aperture. Having shallow depth of field will make your textures more or less unusable år you will manually have to mask out images for texture creation as, at least agisoft, will otherwise pick parts of the texture from the blurred part of the images.
So it's either sacrifice a little sharpness or add a lot of manual work. 😔
(i understand that you probably know that already but responding so that others who might not will know why).
Then perhaps ideally do not use a full frame camera, as depth of field (DOF) decreases with the size of your image circle (and focal length). Meaning an aperture value of 5,6 on a 50mm lens on a MFT sensor, has deeper focus (less depth of field) than that of the same aperture value and lens on a full frame sensor, unless it all comes down to field of view anyway, meaning that the DOF on a smaller sensor is “perceived” as deeper because of the narrower field of view? However I do suspect that a smaller sensor camera would be preferable for sharpness purposes, all else being equal, as the quality of the lens obviously play a large role as well. And so if focal length does not matter for the end result, then definitely use a wide angle lens, as the wider your angle of view the deeper your depth of field will be, meaning that an image shot on the same camera (in your case Sony a7R) with the 35mm lens at f. 1:8 will have significantly deeper focus than your 105mm at same aperture.
Is photogrammetry worth it if you want to make some money on the side?
Best photogrammetry tutorial I have seen on TH-cam so far, saving a lot of time avoiding mistakes and optimizing the setup, thanks for sharing!
You did not see much then
The first video of the course was really great! Any idea when the rest will be available on Gumroad?
instaBlaster.
@@isaacchristian2652? 🤔
Hi, where is the file for 3D printing?
Some great information here, thanks for the tips. What tier of the Patreon is this series? It's uclear from the tier information. Cheers
Hi, great video! What polorize filter are you using?
Hey there, great tutorial! I'd like to see the others.. just a question before buying the patreon: did you do all the photogrammetry course? I didn't see it on your patreon page. Thank you!
Are you using auto focus or manual focus as I think I missed that in the video ?
What about the paid version of Zephyr??
I have no idea. :-)
Awesome video. I noticed you did not mention 3df Zephyr. I find it has a much more user friendly interface and workflow. Many thanks for this outstanding tutorial.
Hello,RealityCapture Meshroom Metashape which is better?(I use RTX 3060 12G)
nice job man... might mention that you plan on going through a lot of common mistakes and showing solutions/alternatives. Almost closed video thinking this guy doesn't know what he's talking about haha I find that as much I think I might know.. There's always a chance somebody knows something better.. something new.
You mentioned link to model for filter holder though? I don't see that. I've seen a few and wanted to print but they are of many pieces... I love the idea yours only has one!! :) Again.. great job! Any plans for updates? You could get away with doing all the same steps but later methods/techniques.
8:00 *Little depth of field.
The higher your f.stop a.k.a. aperture number, the smaller your lens aperture opening (as measured in fractions or divisions of the total lens diameter) the more “depth” you have of your “field” of view, in other words, a higher aperture value decreases your lens opening and increases your subjects focus area.
As a general rule, the higher your aperture number, the sharper the image.
Also, equally important in this case, the wider your field of view (shorter focal length), the deeper your focus, or depth of field (a.k.a DOF for short).
And to get even deeper into the subject, the size of your image circle also matters, meaning that the physical size of the cameras image sensor also matters since this effects the field of view.
Meaning that smaller sensor cameras has a narrower field of view, hence a deeper focus across the image area.
Therefore, not taking optical quality, noise, bit depth, resolution etc into consideration, I would assume a smaller sensor (MFT or APS-C) mirrorless camera with a wide angle lens would hit the sweet spot for photogrammetry?
I see you use the Tamron 28-75 but what focal length? I didn’t hear that part. Thanks!
Hi! I see in your intro that you have also created a model of a plant with many leaves. I have been trying so hard to photoscan a plant and have a nice 3D model of it, but unfortunately never succeeded. However, I am not an expert... Is there any chance you could help me or show in any way how photoscanning of plants work? Thank you!
Soooo,where do we find that polarizing film u used in front of the flash?
the same question!!! :(
any ring flash you recommend to buy now since that specific one is now $500+ or would a ring flash made for your camera be a better option?
How do you make the flames in your Fstorm render?
What is phogrametry job or income opportunity now in 2022 or in future?
Didnt see them on your patreon.. can you post direct link
Best 3D Scanning TH-cam video so far! the good part is you use Sony a7 :)))
but got a few question here! 1: how can I get film for the flash? I have heard you can remove it from broken lcd , right?
2.double ISO wont affact IQ for RC ?I also have heard you should not go above iso 100 for photogrammetry processing!
3. what is your focus setting ?Is it wide?camera dose not miss focus in wide mode?
when doing polarized lighting you should black out the room.
Not sure exactly what you mean. Black out in terms of painting all walls black and empty it, absolutely. That'd probably help a lot, if I could 😀
Black out in terms of killing all ambient light however, makes no difference as the fast shutter speed and small aperture ensures no ambient light is captured.
Could you do this with a Windows Surface 7? It has the infered scanner
Y los otros videos? Title fake
Title says "part 2-5 in description". Description says "All videos are accessible on the Viz.Guru Patreon page, www.patreon.com/vizguru.".
No fake at all. 😉
Bro, thank you very much!
Exactly the advice I've been searching for the past 5 days, even though I don't need even 10% of the quality you're achieving.
All I want to do is to capture parts in 3d from which I can design other parts to fit on so I can 3d print them, and all I have is an android phone with free photogrammetry software. (The free software doesn't allow me to take the pictures seperately with a proper camera.)
Gonna buy a cheap turntable and a cheap light box with a dark background, print a sort of cow pattern for the turntable, 3d print a stand for the phone, and that should do the trick!
Polarizing filter or not? What do you recommend?
Great tutorial, but paywalling the rest of it? Come on dude
Wouldn't you have been better off using two cheap softbox lights instead of that ring flash? You could have those for $100 on Amazon and I think they'd give a much softer more even lighting from two sides. I'd put the object being photographed on a reflective mat white card. to get some light from below. Other than that great tutorial.
Thank you! Very thorough and informative!
Im trying to buy the course in Gumroad but is not avaliable :(
Could you please make available in GUMROAD?
Has been on my to-do for weeks. Thanks for reminding :-)
hello where can i find the file for printing
to shoot outdoors surfaces do you need to shoot in hss? or just connected to the camera is it enough? I understand that connected to the camera the hss function is not compatible
First of all, Thank you!! it is wonderful. Plus, could you please share your resources if there is, any resource related to photogrammetry in technical side? Deep learning, computer graphics course, any lecture records, any book? Any help would be appreciated!!
Thank you so much for creating this series of tutorials, I really appreciate the effort and work you have put into it. Thank you.
Hello sir I'm a big fan of your fstorm work specially in the facebook group. However, as a student I cannot afford monthly subscription so I decided to try the alpha ver. since its free and wanted to ask if it still worth getting the old version? Thanks and Godbless
what microphone do you use to record the video?
Hi! Where can i find the file for printing? :)
Thank you for video! Tell me please, what focal lenght lenses do you usually use to shot (scan) such small objects? And what focal lenght lens did you used in this particular case?
Also would like to ask, is it ok if use to shot (scan) object like in the video with the lens 50mm (for example Sony FE 50mm F2.5 G ) ?
Hello,RealityCapture Meshroom Metashape which is better?(I use RTX 3060 12G)
very interesting ! other question: Where did you get your jacket from ? :-)
Will he tell us the make and model of all the equipments he uses here?
All I can say is awesome explanation thank you very much sir
lol where’s part two
🙄
How you got the bottom of the model ? cheers
Beautiful!
How is the focus set? Manual on first photo, or automatic for all photos?
Great video! What tripod do you use ?
With that camera, you can crank the ISO up quite a bit before noice even becomes remotely a problem
I've tried every combination. IMO Small aperture is a better trade-off than ISO. And I think the ISO capabilities of this camera is highly overrated. It may be better than many other cameras, but the images get noisy nonetheless.
@@TextureSupply You may be right, its the RIII, right? Higher megapixel count means more noice at higher ISO's
th-this isnt the hub
After 2 years where is 4 other parts?
Hi Datec. Reading the descriptions sometimes helps. Part 2-5 has been available since day 1 on Patreon.
Really good tutorial, been doing the same facing all the same issues. I like your lazy Suzie much better then my
Iconasys has photography turntables that can automatically capture up to 8,000 images per rotation. Iconasys photography software can also remove the background automatically from all the images. Has Agisoft Metashape integration too. Makes the whole process much easier and quicker.
Why are you using f22? it causes diffraction!
Because in my opinion short focus is worse than a little soft image. With optimal aperture (9-11 ish) there is higher risk that out-of-focus areas will end up in the textures, causing random parts of the 3d model to have blurry texture. Then it's better to have half-decent sharpness all over and sharpen in post. A soft image caused by small aperture can be sharpened, but blurry islands caused by DoF can not be easily fixed.
@@TextureSupply ok, tks for the answer
How do you get the bottom and top of objects?
Hi, I'm totally new to this, but my thinking for the bottom is to lie the object on its side, start capturing a totally new model, and then merge the two objects together on the computer using reference points.
Not sure if it will work, gonna try it.
Pretty good man! Tnx
Hi fstrom is free
wow
Are all 5 Parts of this series still available on your Patreon page still?
Yes indeed. I've been wanting to make them available outside of patreon too, just haven't gotten around to make it happen yet.
@@TextureSupply I just signed up (base level) there to get access to them and it's now asking me to go from the base level up to the Top level. Problem is it says could take up to 7 days to process my original payment (PayPal) before I can upgrade. Seems like a convoluted process Patreon has created making it difficult for a new member to simply upgrade to see desired content. ?????
Here's the message I get when trying to upgrade: You can't make changes to your pledge because you have a pending transaction from February 3, 2022. Pending payments typically take about 7 business days to settle. You'll be able to change your pledge once this payment is settled.
HOW about the other parts of t his course
Yes what about them :)
@@TextureSupply where are they?
@@Greynerd on patreon as stated in both the video and description. :-) However I've been planning on making them public outside of patreon too, just haven't gotten around to it yet.
I'm new to photography. I want to do a 3d model of my dog. It has short black hair.
I dont need color texture but detail.
Would doing black and white photos be good?