Teleconverter versus DX mode for mirrorless camera

[macwalter]

Many claim that DX allows the AF to work better on things the camera is programmed to recognize like birds.

I tried this today as the light was low so I used the bare 400 f4.5 without the 1.4tc. Yesterday using the lens with the 1.4tc I was able to easily track the various warblers with single point and 3d tracking both with bird id on. Today without the TC I found the camera struggling to focus especially with 3d tracking until I switched to DX mode where it improved significantly. Later as the light improved it worked better in fx but not as good as with the 1.4tc. I was surprised by this. I do think that a larger subject really helps especially in less than ideal light.

 

[JoelKlein]

Subject recognition needs a decent sized object to work out what it is, and where the body/head/eye is.

 

[Anjin San]

I tried this today as the light was low so I used the bare 400 f4.5 without the 1.4tc. Yesterday using the lens with the 1.4tc I was able to easily track the various warblers with single point and 3d tracking both with bird id on. Today without the TC I found the camera struggling to focus especially with 3d tracking until I switched to DX mode where it improved significantly. Later as the light improved it worked better in fx but not as good as with the 1.4tc. I was surprised by this. I do think that a larger subject really helps especially in less than ideal light.

It does…and despite thinking about it a fair bit and googling I’ve yet to find a decent explanation beyond ‘better because it’s bigger in the frame’…but the pixel size is the same, you’re just using fewer of them so I’ve just accepted that DX helps if AF is struggling and ain’t worrying about it any more. The recovering engineer in me wants a better answer…but the reality based photographer see it jus’ don’t matter.

 

[Nimi]

It does…and despite thinking about it a fair bit and googling I’ve yet to find a decent explanation beyond ‘better because it’s bigger in the frame’…but the pixel size is the same, you’re just using fewer of them so I’ve just accepted that DX helps if AF is struggling and ain’t worrying about it any more. The recovering engineer in me wants a better answer…but the reality based photographer see it jus’ don’t matter.

It relates to AI. It's easier for the processor to identify an "eye" or "face" in the projected image against its learned library of eyes and faces.

 

[Anjin San]

It relates to AI. It's easier for the processor to identify an "eye" or "face" in the projected image against its learned library of eyes and faces.

I’ve read that before…but of there are say 25 pixels on the face in FX…there are still 25 in DX…so other than being apparently larger in the EVF I don’t understand how DX should make a difference…Nikon obviously knows why but they’ve not chosen to explain why it is so. As you know…engineers and even recovering engineers hate unanswered questions…and just because is rarely a good answer. Maybe the AF algorithm iterates faster in DX because there re fewer pixels to analyze…more ‘looks’ per second assuming it’s set up that way would decrease the time required to achieve lock maybe.

 

[macwalter]

I think in terms of image quality not much if any difference unless somehow the camera doesn’t let in as much light. I can say that a larger subject is easier for the camera to see in low light especially. But if I know final size of the subject will be the same I prefer FX so if the bird flies my chances of getting it in flight improve.

 

[Nimi]

I’ve read that before…but of there are say 25 pixels on the face in FX…there are still 25 in DX…so other than being apparently larger in the EVF I don’t understand how DX should make a difference…Nikon obviously knows why but they’ve not chosen to explain why it is so. As you know…engineers and even recovering engineers hate unanswered questions…and just because is rarely a good answer. Maybe the AF algorithm iterates faster in DX because there re fewer pixels to analyze…more ‘looks’ per second assuming it’s set up that way would decrease the time required to achieve lock maybe.

It's not at the pixel level. It's the collection of pixels forming an image as a percent of the frame compared to the library. The key is the %, and therefore the DX-effect.

 

[macwalter]

But I wonder that the z8 is not a true crop sensor camera and when you are only using part of the frame in DX mode what advantage that really offers except in terms of the camera having an eas time seeing smaller subjects. As was said it’s the same number of pixels and you can’t crop any deeper either way.

 

[apakett]

I’ve read that before…but of there are say 25 pixels on the face in FX…there are still 25 in DX…so other than being apparently larger in the EVF I don’t understand how DX should make a difference…Nikon obviously knows why but they’ve not chosen to explain why it is so. As you know…engineers and even recovering engineers hate unanswered questions…and just because is rarely a good answer. Maybe the AF algorithm iterates faster in DX because there re fewer pixels to analyze…more ‘looks’ per second assuming it’s set up that way would decrease the time required to achieve lock maybe.

I have also noticed that in DX mode distant birds focus easier. You can probably achieve the same results by selecting a smaller focusing area.

 

[macwalter]

You can but even on single point the camera tends to struggle more on smaller subjects in lo light. There is less problem in better light. That being said I still prefer FX most times unless the situation is worse than usual. Today shooting warblers in lousy light I shot in FX using single point and a large wide area mode that actually did pretty well, better tha 3d or auto af. I was able to use subject recognition most of the time. I try to practice daily getting used to the new firmware.

 

[apakett]

You can but even on single point the camera tends to struggle more on smaller subjects in lo light. There is less problem in better light. That being said I still prefer FX most times unless the situation is worse than usual. Today shooting warblers in lousy light I shot in FX using single point and a large wide area mode that actually did pretty well, better tha 3d or auto af. I was able to use subject recognition most of the time. I try to practice daily getting used to the new firmware.

The AF system works on the displayed image. Cropping will give AF more subject pixels to work with.

 

[markymark]

@Steve covered this nicely with blog posts and videos a while back. From the image quality standpoint this info still holds up in the mirrorless world with perhaps the IQ resulting from TC use being a bit better these days only because the newest mirrorless TCs are a bit better than some of the older TCs. This doesn't get into focus acquisition differences but IQ, DoF impacts, effective ISO impacts are still impacted the way those things were in DSLRs.

Use A Teleconverter Or Crop The Image? - Backcountry Gallery

Think fast! You’re out searching for wildlife – your camera in hand – when you spot some critter you’ve always wanted to bag off in the distance. Sadly, you know that your lens is just not going to be quite long enough to get the shot you want on its own. So, here’s the question – […]

backcountrygallery.com

Just a question, that article was done on although very good equipment, but it wasn’t done on the mirrorless system & I was just wondering if the results would be the same say using a Z9 coupled with a 600mm TC Z lens?

 

[bleirer]

The AF system works on the displayed image. Cropping will give AF more subject pixels to work with.

That's the part I'm not understanding. There aren't any more sensor photosites, cropped or uncropped the bird or whatever is covered by the same number and size of photosites so the improvement has to come through the dual stream evf somehow. But Canon doesn't have dual stream but still shows the same af improvement in crop.

 

[apakett]

That's the part I'm not understanding. There aren't any more sensor photosites, cropped or uncropped the bird or whatever is covered by the same number and size of photosites so the improvement has to come through the dual stream evf somehow. But Canon doesn't have dual stream but still shows the same af improvement in crop.

This is my understanding. The display is 2.1 MP. The photosites are down-sampled to that resolution regardless of whether the camera is in FX or DX mode. Dual stream makes what is happening more obvious. At some point in the processing Canon has to split off the display images. That is what the AF system uses. I would love to see a paper from Nikon explaining exactly what is going on. I am also a retired engineer.

 

[JoelKlein]

This is my understanding. The display is 2.1 MP. The photosites are down-sampled to that resolution regardless of whether the camera is in FX or DX mode. Dual stream makes what is happening more obvious. At some point in the processing Canon has to split off the display images. That is what the AF system uses. I would love to see a paper from Nikon explaining exactly what is going on. I am also a retired engineer.

Interesting that an engineers brain is having a hard time understanding the fact that the AF is reading the EVF as is.

 

[macwalter]

you guys are all way beyond me! All I know is if the bird is bigger in the frame the AF can have an easier time getting on it but the actual size is no different in the endgame. I don’t know if you lose some dynamic range using DX mode as I don’t know how the camera actually goes from fx to DX. I prefer fx as it gives me more room to capture flying subjects and crop deeper if necessary.

 

[Chappy]

My best bet is to always move closer to the subject to fill the frame. If I'm far enough away that I can't get a good pick in FX mode or Dx mode at 19 mp then it is more likely that i won't print. i might put it on Social Media though.

 

[macwalter]

I would agree but sometimes you can’t get as close as you like. I think people need to be realistic about what their gear can deliver and also the situation. For instance On a birding trip one can’t take the time needed to get the ideal photo and so it comes down to intent. These shots would not meet printable criteria but knowing that one might still want to take the shot. also light often plays into these scenarios. I can’t afford the big glass nor do I want to lug all that gear around when I’m out hiking. So handholdable lenses are what I use. Also at least for me pixel perfect is not always the only criteria I go by. I love photography and being out in nature is Sometimes enough reward for me.

 

[Chappy]

I bought the 600pf because it fits my style of photography. I rarely shoot on a tripod and always try to move closer to the subject. i know it can't always be done but i try. I learned my lesson about cropping too much and then trying to print years ago. While the 1.4 teleconverters have gotten much better, I don't like putting them on or taking them off in the field.

 

[EricBowles]

I’ve read that before…but of there are say 25 pixels on the face in FX…there are still 25 in DX…so other than being apparently larger in the EVF I don’t understand how DX should make a difference…Nikon obviously knows why but they’ve not chosen to explain why it is so. As you know…engineers and even recovering engineers hate unanswered questions…and just because is rarely a good answer. Maybe the AF algorithm iterates faster in DX because there re fewer pixels to analyze…more ‘looks’ per second assuming it’s set up that way would decrease the time required to achieve lock maybe.

I think the image you are using is the display - not the pixels. So you may have an 8000 pixel width image that is displayed at 1200 pixels whether it is FX or cropped to DX - the image is downsized to fit the EVF display. The 25 pixels of the subject face in FX would be displayed as 37 pixels in a DX viewfinder and for the AF system that uses the same feed. The area the camera has to look for a subject is also smaller and has eliminated the edges of the frame. More importantly, the camera starts with subject recognition so once it finds a subject, it looks for the head and then the eye. Probably a better scenario is not that the head is 25 pixels but that the subject is 250 pixels in the FX display, and 375 pixels in the DX display. As far as AF is concerned, it's like getting closer.

 

[bleirer]

I think the image you are using is the display - not the pixels. So you may have an 8000 pixel width image that is displayed at 1200 pixels whether it is FX or cropped to DX - the image is downsized to fit the EVF display. The 25 pixels of the subject face in FX would be displayed as 37 pixels in a DX viewfinder and for the AF system that uses the same feed. The area the camera has to look for a subject is also smaller and has eliminated the edges of the frame. More importantly, the camera starts with subject recognition so once it finds a subject, it looks for the head and then the eye. Probably a better scenario is not that the head is 25 pixels but that the subject is 250 pixels in the FX display, and 375 pixels in the DX display. As far as AF is concerned, it's like getting closer.

I get how that would work with the Nikon dual stream. But why does it work with Canon which says af is based on each sensor pixel having 2 halves that are compared. There must be more to how the magic works.

 

[Anjin San]

I think the image you are using is the display - not the pixels. So you may have an 8000 pixel width image that is displayed at 1200 pixels whether it is FX or cropped to DX - the image is downsized to fit the EVF display. The 25 pixels of the subject face in FX would be displayed as 37 pixels in a DX viewfinder and for the AF system that uses the same feed. The area the camera has to look for a subject is also smaller and has eliminated the edges of the frame. More importantly, the camera starts with subject recognition so once it finds a subject, it looks for the head and then the eye. Probably a better scenario is not that the head is 25 pixels but that the subject is 250 pixels in the FX display, and 375 pixels in the DX display. As far as AF is concerned, it's like getting closer.

Could be…I need to look at the EVF pixels and compare to the FX and DX dimensions…the larger in the viewfinder is clearly why AF works better in DX but engineers always want to know why. Alternatively…mostly I just added it to the lengthy list of things I just accept without fully understanding them.

 

[macwalter]

I bought the 600pf because it fits my style of photography. I rarely shoot on a tripod and always try to move closer to the subject. i know it can't always be done but i try. I learned my lesson about cropping too much and then trying to print years ago. While the 1.4 teleconverters have gotten much better, I don't like putting them on or taking them off in the field.

I have no doubt the 600pf will add to my ability to reach out further. I agree that in most cases 600mm is a sweet spot for birds. I do like having the better low light performance of the 400 f4.5 though as today it was dark yet good action and I could still get some decent shots. I also like that I can get 560mm at f6.3 and 800mm at f9. Surprisingly I found the 2x did a decent job in good light with many tack sharp images of BIF. It doesn’t rate well in the reviews and I do sometimes see chromatic Aberration in hi contrast images. The 600pf is no doubt better in this regard but you’re still at f9. The 800pf is too big and too long. I have the 500pf and though it is sharper than the 400 F4.5 with the 1.4tc I find in looking at real world images the difference is really small and it doesn’t handle as nice. I handhold 98% of the time. I can travel with the 100-400zoom for versatility and the 400 f4.5 for birds with both TCs. I have problems swapping on TCs in the field with the curtain on.

 

[Chappy]

B

I have no doubt the 600pf will add to my ability to reach out further. I agree that in most cases 600mm is a sweet spot for birds. I do like having the better low light performance of the 400 f4.5 though as today it was dark yet good action and I could still get some decent shots. I also like that I can get 560mm at f6.3 and 800mm at f9. Surprisingly I found the 2x did a decent job in good light with many tack sharp images of BIF. It doesn’t rate well in the reviews and I do sometimes see chromatic Aberration in hi contrast images. The 600pf is no doubt better in this regard but you’re still at f9. The 800pf is too big and too long. I have the 500pf and though it is sharper than the 400 F4.5 with the 1.4tc I find in looking at real world images the difference is really small and it doesn’t handle as nice. I handhold 98% of the time. I can travel with the 100-400zoom for versatility and the 400 f4.5 for birds with both TCs. I have problems swapping on TCs in the field with the curtain on.

I am from a dry climate where there is lots of dust and the wind gets up now and then that gets the dust to flying. When I am shooting out in the open I don’t bother bringing the teleconverter as I have learned not to take the lens off of the camera in the outdoors. Even in my truck I rarely take the lens off because I drive around the turnrows with my windows down. I find in most cases that I can get the shot in DX mode.

I have had to clean the sensors and lenses in my cameras way too many times over the year because of the dust. When the 300pf and the 500pf came out it changed my style where I could find ways to move closer to the wildlife or birds. It became a new challenge for me.

I can understand using teleconverters in different situations. I thought about using one when going to the Texas coast. After a few minutes I got the film on my eyeglasses from the wind and salt water in the air. I’ve gotten some good pics with teleconverters and think about using them. Below is one I took back in 2016 with a D500 with the Sigma 150-600 sport with a Sigma 1.4 Teleconverter.

You can only see EXIF info for this image if you are logged in.

 

[macwalter]

Dry conditions are an issue for sure. I’ve so far had little problems with the z8 but significant problems with my dslrs. The ability to close the sensor on the z8 makes a difference. Also the new TCs are clearly superior. I rare used them on my dslr. Also I live in a very wet region.