p.2 #1 · XT5 using Digital Teleconverter and further crop to get Canon 1D - like files
AleDiSciascio wrote:
I should point out that obviously the subject of the image will be a big determination of how or if the image can be used.
That said I took the image you shared that I guess was shot with the 2 DTC, I then cropped it square and then cropped the square image to 3:2, basically doing what I mentioned.
I then applied two quick and dirty unsharp mask (one for microcontrast, one for actual sharpness) and to my eyes if the photo were something worth printing (no offense to your masterpiece but I like people more than trucks) I think it would make a completely decent 8x10 print without a doubt.
And in my experience once you take viewing distance into account that means you can print it basically at any size provided it won't be "pixel peeped"
Here is the resulting image:
Ok I don't know how to link an image from imgur clicking the link gives "bad link".... but if you copy and paste this in a browser it will work: https://imgur.com/a/HeuMH6b
And to see the actual image you'd have to download it. It really is printable ...Show more →
Printing 8x10 isn't really anything special and certainly not the metric by which I judge "printability". I'd rather have the full size RAW file and use other tools to crop in/upsize than relying on the DTC, but that's me.
You clearly have your mind made here, and will find out for yourself soon enough if it's good enough to meet your requirements.
p.2 #2 · XT5 using Digital Teleconverter and further crop to get Canon 1D - like files
ruthenium wrote:
Re "Meanwhile if I shoot the viltrox at F1.4 i should be getting an image that is equivalent to ..... 75 * 1.5 (APS C) * 2.0 (Digital TC) * 1.5 (final crop) .... which works out to 337.5mm at F2 equivalent on Full frame.
Where am I wrong?"
You FF FL equivalent looks correct. You correctly converted 75mm to FF FL, then applied the "manual" crop factor which is 3.
For the equivalent FF DoF, the F1.4 must be multiplied by the same crop factor = 1.4 x1.5 x 3 = F6.3
The equivalent FF ISO is obtained by multiplying the camera ISO by the crop factor squared = 20.25. For example, if the camera ISO is 200, then the cropped image should be equivalent to one obtained with a FF camera set to the same SS, ISO 4000, FL 337.5mm F6.3.
For photographically equivalent images obtained using FF and cropped sensor cameras, the PDR is typically similar, and can be practically the same (check Photonstophotos for examples).
I suggest running the cropped images through Gigapixel or Topaz Photo AI to upscale these to, say 20MP. You may well see a significant improvement in IQ....Show more →
Maybe I'm using the depth of field calculator wrong but if I'm using it right it's even (MUCH?) worse than you said
on one screen I put a Fuji X camera with a 75mm lens at f 1.4 I set my subject distance to 25 meters. the calculator reported a DOF of 6.37 meters
on the other screen I put a canon 5D series cameara first with a 112.5mm lens (only accounting for the crop sensor) and addjusted the F stop to bring the dof as close as possible to 6.37 meters. That would be somewhere between F 2.0 and F 2.2 which isn't bad.
I then accounted for the 2x Teleconverter, so I changed the canon lens to a 225mm lens and I started increasing the F stop to bring the dof as close as possible to 6.37 meters and... YIKES... it's somewhere between F8 and F9
Accounting for the final crop... lens changed to 337.5 .... yields an F stop of F/19 ....
suddenly the 70-300's 5.6 doesn't look so bad
I kind of hope I did this wrong... because F/19 .... jeez.
I suppose on the positive side if I stop it down to F4 20 meters worth of stuff are in focus... even a Fuji can handle AF in those circumstances ;P
p.2 #3 · XT5 using Digital Teleconverter and further crop to get Canon 1D - like files
Jack Flesher wrote:
FTR, ISO is ISO and doesn’t “change” with crop. The ISO noise factor may change, and theoretically it happens at the square of the format factor, but in reality at normal ISO settings it seems to be negligible, or far less than the theoretical value. IME I don’t see it in Fuji X files until ISO 3200 or higher, and then the relative impact seems about linear with crop factor. FWIW. My only point here is I wouldn’t worry about it very much unless you’re at very high ISO.
While the ISO doesn't change on the camera body with crop, the PDR does. When asking the specific question on "how an equivalent FF image should look like?", the answer is it should look like one obtained on a FF camera set to the ISO which is the crop factor squared-times higher. The SNR of the cropped image is going to be like that of the high-ISO FF image ( when viewed at the same size from the same distance). Thus, the consideration of the effect of cropping on the ISO helps to get an idea of how the cropped should look like, if it is going to be usable in terms of the SNR.
p.2 #4 · XT5 using Digital Teleconverter and further crop to get Canon 1D - like files
AleDiSciascio wrote:
Where am I wrong?
I’m late to the party and have not read the entire thread, so this may have already been addressed.
I didn’t check your math, but I suspect that you are correct when it comes to pixel dimensions. However, you’ll be using a much smaller area of the sensor for your actual image, which means (among other things) that is it equivalent to using a smaller-than-APS-C sensor camera with the downsides that come with smaller sensors.
Generally it is better to use lenses that project the image size you need onto the full sensor.
Basically, your approach could be OK for an occasional emergency situation where you don’t have a long enough lens and you have no other choice, but as a regular approach it isn’t a great idea. (OK, I’ll grant you that you could make very small images for things like social media use and probably no one will notice.)
Bottom line — you are getting the camera, right? Just try it and see what it looks like. If the results are good enough for you, so be it.
(From the sounds of your replies, which I have now read, it almost sounds like you aren’t here so much for feedback on your idea — which your original post suggested was your goal — as you are to tell us why your idea is great and perfect. In that case, no need to engage here, just do your thing and be happy or not with it.)
p.2 #5 · XT5 using Digital Teleconverter and further crop to get Canon 1D - like files
gdanmitchell wrote:
I’m late to the party and have not read the entire thread, so this may have already been addressed.
I didn’t check your math, but I suspect that you are correct when it comes to pixel dimensions. However, you’ll be using a much smaller are of the sensor for your actual image, which means (among other things) that is it equivalent to using a smaller-than-APS-C sensor camera with the downsides that come with smaller sensors.
Generally it is better to use lenses that project the image size you need onto the full sensor.
Basically, your approach could be OK for an occasional emergency situation where you don’t have a long enough lens and you have no other choice, but as a regular approach it isn’t a great idea. (OK, I’ll grant you that you could make very small images for things like social media use and no one will notice.)...Show more →
Oh for sure. I wouldn't routinely go shoot soccer games with the 75 hoping to get 300mm results, but for instance we have a trip coming up at the end of july and I figure I'll bring maybe 2 lenses ... and I might decide to bring the 75 so I can get some nice portraits of my friends back home and their kids... and still have a "long" lens available to me in a pinch if I "need it"
p.2 #6 · XT5 using Digital Teleconverter and further crop to get Canon 1D - like files
AleDiSciascio wrote:
…I've printed images for clients where the file was cropped to the point that the source image was at a resolution of 150ppi ROUTINELY and I remember printing one as low as 120ppi which to me and the clients looked great when printed.
I print a lot, and I would occasionally go as low as 180ppi with pretty good results, though I regarded that as a worst-case floor, not as a regular target.
I suspect that there may have been a few (very few) cases where someone really needed a print that required 150ppi (though I can’t actually think of one offhand), and at 150ppi the pixelization of the image really does start to become visible even to those who aren’t quite nose-to-the-print-inspection fanatics. 120ppi? No thanks. That’s fine on the screen, but it a serious print of any sort, nope.
But again, if you have made up your mind that cropping a small APS-C image down to the level of a much smaller sensor and printing it at 120ppi is just dandy, that’s your choice. No reason to ask for advice here.
p.2 #7 · XT5 using Digital Teleconverter and further crop to get Canon 1D - like files
gdanmitchell wrote:
(From the sounds of your replies, which I have now read, it almost sounds like you aren’t here so much for feedback on your idea — which your original post suggested was your goal — as you are to tell us why your idea is great and perfect. In that case, no need to engage here, just do your thing and be happy or not with it.)
Sorry to come across that way, wasn't my intention at all.
I don't have the camera and my X-E2 doesn't have the digital TC. I guess I'm half excited about getting camera etc... and half concerned about what kit to bring on the trip since we'll have luggage size restrictions.
And yes admittedly for the past 20 years or so... basically ever since I've been ordering prints on photo paper I've found myself "arguing" with peeps who don't seem to believe you can print a photo that will look great on a wall or in an album unless your source image is AT LEAST 300ppi ... and my walls and the walls of clients and their albums, clearly tell a different story
p.2 #8 · XT5 using Digital Teleconverter and further crop to get Canon 1D - like files
If you are comfortable with cropped sensor cameras, and need a relatively compact camera system, consider micro-four-thirds, like OM-3 that is IP53 weather sealed and has a good IBIS, a fast stacked sensor. The 40-150mm F4 (FF equivalent 80-300mm) weighs 382 g and is 99.4mm long.
p.2 #9 · XT5 using Digital Teleconverter and further crop to get Canon 1D - like files
Had the same impression of someone not really looking for feedback..
For what it’s worth:
Cropping a 75mm lens to 200mm (300mm FF equivalent) on the X-T5 leaves you with a file that’s 2.905x1.937. Any upscaling is hit and miss so I wouldn’t count on that.
If set to f/1.4 the DOF would be equivalent to f3.7 (f5.6 FF equivalent)
Imo a much better option would be getting a MFT camera with the 40-150mm f2.8, or even the 35-100mm f2.8 and cropping in.
But if you only want to crop in in a pinch and usually stay closer to the original 75mm, then I guess yes you can do smaller prints with those 5.6mp files
p.2 #10 · XT5 using Digital Teleconverter and further crop to get Canon 1D - like files
ruthenium wrote:
While the ISO doesn't change on the camera body with crop, the PDR does. When asking the specific question on "how an equivalent FF image should look like?", the answer is it should look like one obtained on a FF camera set to the ISO which is the crop factor squared-times higher. The SNR of the cropped image is going to be like that of the high-ISO FF image ( when viewed at the same size from the same distance). Thus, the consideration of the effect of cropping on the ISO helps to get an idea of how the cropped should look like, if it is going to be usable in terms of the SNR. ...Show more →
Show it to us with image samples taken originally on FF, properly exposed at say ISO 800. No fair adding noise in post to prove your point.
p.2 #11 · XT5 using Digital Teleconverter and further crop to get Canon 1D - like files
fjablo wrote:
Had the same impression of someone not really looking for feedback..
For what it’s worth:
Cropping a 75mm lens to 200mm (300mm FF equivalent) on the X-T5 leaves you with a file that’s 2.905x1.937. Any upscaling is hit and miss so I wouldn’t count on that.
If set to f/1.4 the DOF would be equivalent to f3.7 (f5.6 FF equivalent)
Imo a much better option would be getting a MFT camera with the 40-150mm f2.8, or even the 35-100mm f2.8 and cropping in.
But if you only want to crop in in a pinch and usually stay closer to the original 75mm, then I guess yes you can do smaller prints with those 5.6mp files ...Show more →
I guess the reality is that:
A. I DID want feedback on things like "your math is wrong" or "I've tried something like this and cropped in the Viltrox 75 1.2 is SHIT" or for instance "keep in mind that while the viltrox is awesome shot normally even at 1.2 ... if you were to use that lens with the Digital Converter and then applied an additional crop to the resulting file, the DOF performance of that gorgeous 1.2 lens at say a distance of 20 meters from the subject would be complete ASS..." <--- that is feedback I got and when I went looking with a DOF calculator unless I'm doing the math wrong it's even FAR WORSE than I was told... if I'm calculating correctly... shot at F1.4 the 75mm viltrox then cropped in to be a 337 equivalent would have the same DOF as a full frame lens shot from the same position at something insane like F19. I also welcomed feedback about using upsampling software and being careful about high iso
B. I DID NOT want feedback on things like "not enough pixels for a good print" since I've run into pixel-peepers who obsess over 300+ppi for prints for almost two decades and while I respect their right to their opinion I disagree with their conclusions IN MOST normal circumstances.
Either way... the feedback I've received has in fact changed my mind from "this could be a useful trick to use whenever I want" to "this might end up being just an emergency solution and definitely not something I want to use with any frequency" ... so regardless of how I came across, the feedback was useful and it impacted me.
p.2 #12 · XT5 using Digital Teleconverter and further crop to get Canon 1D - like files
AleDiSciascio wrote:
I guess the reality is that:
A. I DID want feedback on things like "your math is wrong" or "I've tried something like this and cropped in the Viltrox 75 1.2 is SHIT" or for instance "keep in mind that while the viltrox is awesome shot normally even at 1.2 ... if you were to use that lens with the Digital Converter and then applied an additional crop to the resulting file, the DOF performance of that gorgeous 1.2 lens at say a distance of 20 meters from the subject would be complete ASS..." <--- that is feedback I got and when I went looking with a DOF calculator unless I'm doing the math wrong it's even FAR WORSE than I was told... if I'm calculating correctly... shot at F1.4 the 75mm viltrox then cropped in to be a 337 equivalent would have the same DOF as a full frame lens shot from the same position at something insane like F19. I also welcomed feedback about using upsampling software and being careful about high iso
B. I DID NOT want feedback on things like "not enough pixels for a good print" since I've run into pixel-peepers who obsess over 300+ppi for prints for almost two decades and while I respect their right to their opinion I disagree with their conclusions IN MOST normal circumstances.
Either way... the feedback I've received has in fact changed my mind from "this could be a useful trick to use whenever I want" to "this might end up being just an emergency solution and definitely not something I want to use with any frequency" ... so regardless of how I came across, the feedback was useful and it impacted me....Show more →
Pretty sure that DOF calculation is incorrect. A 75mm lens at f/1.4 has an aperture opening of 53.6mm, exactly the same as a 300mm lens at f/5.6. For your 337mm example it's more like f/6.3. Shot from the same position, they should give the same DOF.
I suspect the DOF calculators fall short here because they're always based on viewing the full image, not at pixel-level. So they put out a higher DOF number for the 75mm image on the Fuji because that's what might look "acceptably sharp" at a normal viewing distance. Cropping in essentially magnifies part of the image and reveals that a smaller area is "acceptably sharp".
Ultimately, the appearance of background blur and subject isolation (to avoid the term DOF as it's a fuzzy concept with lots of assumptions) only depend on the aperture size (so focal lengths / aperture value) and subject distance. Sensor size or cropping are only relevant for the FOV.
p.2 #13 · XT5 using Digital Teleconverter and further crop to get Canon 1D - like files
One reason I bought Xt5 is for crop possibilities, can do that anyway without resorting to fiddling with menus, one reason why I never used thesports mode on the Xt3.
A real digital tc would interpolate to restore higher res, THAT would be useful!
p.2 #14 · XT5 using Digital Teleconverter and further crop to get Canon 1D - like files
fjablo wrote:
Pretty sure that DOF calculation is incorrect. A 75mm lens at f/1.4 has an aperture opening of 53.6mm, exactly the same as a 300mm lens at f/5.6. For your 337mm example it's more like f/6.3. Shot from the same position, they should give the same DOF.
I suspect the DOF calculators fall short here because they're always based on viewing the full image, not at pixel-level. So they put out a higher DOF number for the 75mm image on the Fuji because that's what might look "acceptably sharp" at a normal viewing distance. Cropping in essentially magnifies part of the image and reveals that a smaller area is "acceptably sharp".
Ultimately, the appearance of background blur and subject isolation (to avoid the term DOF as it's a fuzzy concept with lots of assumptions) only depend on the aperture size (so focal lengths / aperture value) and subject distance. Sensor size or cropping are only relevant for the FOV....Show more →
Nice! So ... assuming I can get what I want in focus, it's not AS BAD as I feared
I'm once again excited about the possibilities (in a pinch! in a pinch!)
p.2 #16 · XT5 using Digital Teleconverter and further crop to get Canon 1D - like files
fjablo wrote:
Pretty sure that DOF calculation is incorrect. A 75mm lens at f/1.4 has an aperture opening of 53.6mm, exactly the same as a 300mm lens at f/5.6. For your 337mm example it's more like f/6.3. Shot from the same position, they should give the same DOF.
I suspect the DOF calculators fall short here because they're always based on viewing the full image, not at pixel-level. So they put out a higher DOF number for the 75mm image on the Fuji because that's what might look "acceptably sharp" at a normal viewing distance. Cropping in essentially magnifies part of the image and reveals that a smaller area is "acceptably sharp".
Ultimately, the appearance of background blur and subject isolation (to avoid the term DOF as it's a fuzzy concept with lots of assumptions) only depend on the aperture size (so focal lengths / aperture value) and subject distance. Sensor size or cropping are only relevant for the FOV....Show more →
DOF calculators are not really all that accurate or useful. If anything they might give us an idea of the relationship between different apertures on the same lens, but as far as accurately predicting the useful, acceptable DOF they fall way short.
The fact is that at the edges of the DOF zone we aren’t talking about things being sharp — it is actually a matter of whether the out-of-focus quality is or is not too big to be acceptable in some particular image reproduced in some particular way and some particular size. There’s a ton of subjective stuff there! And it isn’t captured by the calculators.
p.2 #17 · XT5 using Digital Teleconverter and further crop to get Canon 1D - like files
gyoung143 wrote:
One reason I bought Xt5 is for crop possibilities, can do that anyway without resorting to fiddling with menus, one reason why I never used thesports mode on the Xt3.
A real digital tc would interpolate to restore higher res, THAT would be useful!
Gerry
I could be wrong about this but I think "Some" older Fuji X cameras may have done this. I saw one guy say it on a youtube video. I think he was talking about the Fuji X100V.
I went to look at the X100V manual and compared it to the X100VI manual.
The X100V talks about the digital teleconverter and DOES NOT make any mention of the image resolution being reduced.
The X100VI manual, like the XT5 manual states that using the 1.4x DTC shrinks L images to M and the 2.0x DTC shrinks all images to S
So either Fuji didn't think it was worth warning X100V users that their images would be smaller, or Fuji changed how they do this.
What would be cool is if the camera gave you the OPTION to do one of those things where it takes 2-3 shots in really fast burst and then blends them to increase the effective resolution... and THEN applied the crop. That would be (in the right circumstances with static subjects anyway) even better than interpolating.
EDITED TO SAY - I see now that RoamingScott confirmed that the previous generation camears with the lower MP sensor did in fact interpolate.
p.2 #19 · XT5 using Digital Teleconverter and further crop to get Canon 1D - like files
gdanmitchell wrote:
Interpolate in camera or interpolate in post — either way you don’t “get back the lost detail.” It is gone.
True. But the current crop of interpolation algorithms do a very good job of manufacturing plausible data to fill the gaps. So while you don’t get any real data back, you often get a decent interpretation of what might have been there. This is especially true with a well captured and processed raw file; and not so much with a JPEG.
