Sorry...sometimes hard to decipher what people are intending to say (and I'm only on my second cup of coffee!). Resolution can be an interesting subject as it has so many meanings and there is so much confusion on the issue in many instances.
If you would like to post, you'll need to register. Note that if you have a BCG store account, you'll need a new, separate account here (we keep the two sites separate for security purposes).
Sorry...sometimes hard to decipher what people are intending to say (and I'm only on my second cup of coffee!). Resolution can be an interesting subject as it has so many meanings and there is so much confusion on the issue in many instances.
Image quality is in the eye of the beholder to a degree, so not quantifiable.
Since the only benefit of cropping in camera is smaller file sizes while increasing the risk of clipping the subject and missing a better composition I see no point whatsoever of changing from FX to any other form factor. I leave this choice to when I am back in my home studio. Buy big fast cards and plenty of them and you will have no issues. NOW if by going to DX crop one could see a boost from 20fps RAW to 30fps RAW then I would revise my view.
A number of people mentioned image quality loss when using DX mode. I was of the belief that the sensor is what determined image quality, not the full frame or crop mode nor file size, as some people seem to believe. I sent off a question to Nikon, I figured they made the camera and they would be the best place to find out if using DX mode did affect the image quality. I got back an answer and I'm posting it here. As I believed, it is the sensor that determines image quality and their is no loss of image quality in DX mode and no loss of quality with file size.
Response from Stacey G. | |
|
I think it depends on one's definition of resolution. For some it seems to mean pixel density and in that sense, sure the sensor has the same pixels per unit area no matter how much is cropped. For others resolution is number of pixels horizontal times the number of pixels vertical. So in that sense the resolution is different with cropping. If you fill the frame in FX mode you have more pixels on the subject than if you fill the frame in DX mode, so the FX has better image quality.
The resolution is not different when cropping either in-camera or in processing. Where resolution does come into play is in printing. If you have a small file size and you want to push the size up beyond the 1:1 ratio image and print resolution can suffer. But if you don’t believe the company that made the camera, in this case Nikon, you certainly won’t believe me.
I was totally agreeing with you and your Nikon email answerer, if your definition of resolution is pixel density (pixels per inch or whatever). I agree and have always agreed and never disputed that there is no difference between cropping in camera vs. cropping in post.
The difference in quality I was talking about was between filling the frame in crop mode (or by cropping in post) vs. filling the frame with the same size image in full frame mode. Imagine a bird head filling the frame in FX mode on a Z9. You cover the head with 8256x5504 pixels. In crop mode you only cover the head with 5408x3600 pixels. In this sense I meant the resolution was different. When you post them here at 1200x800 pixels the one downsampled ftom the full frame will have better image quality than the other because it had greater resolution to start with.
Here is a quote from Steve from an article on cropping. He uses the more common definition of resolution:
"Additionally, keep in mind that when you crop-in camera (or at home), you are also sacrificing resolution. So, if you switch your 46MP D850 into DX crop mode, you’re only capturing 19.4MP of information, not 46MP in a smaller area."
Last edited:
As I understand it's not only about how many focus points are on the target. It has to do with how many focus points the AF processor has to analyze in order to determine what the target is. In DX mode less than half of the sensor is active so far less data for the processor to analyze. So it improves AF speed more than accuracy. Particularly with subject/eye detection. Presumably if using a small AF area then DX mode has little/no benefit. On the other hand if using auto area it should be a big difference.
The above discussion is a good one, presenting choices and justifications that are all helpful and allow each of us can do what works best for us. The temptations of the crop shooting mode are significant and real. However, I have enough trouble keeping many subjects in the full frame viewfinder, much less a crop version. Knowinng the plus and minus sides, I choose to shoot full frame and crop as appropriate in post. That is the joy of the modern, top of the line gear - we can customize to our own needs.
Agree with previous poster - to each their own. Personally, I shoot my Z9 in DX when I need the extra focal length, which is most of the time when shooting warblers and other small or distant birds, and FX mode when I don't. I choose to shoot DX mode rather than crop in post for a couple reasons. First, when shooting bursts of lots of images, it takes time to zoom in one every single one to decide which to actually keep and process, and then crop the keepers. Even a few seconds per image adds up with the high frame rates this camera shoots. Second, and more importantly, since cropping in LR is non-destructive you still have a 46 MP file even though you're cropping out a lot of it. I just assume work with smaller files for faster downloads, less storage space, and less demand on my computer's resources when editing. Again, this is when I know I am going to need to crop an image before I even shoot. The 20 MP file is not much smaller than my D500 files, and frankly more than I would ever need. I just don't see myself making a 24x36 print of even the best warbler portrait ever taken. It would be weird to have a picture of a warbler on my wall ten times larger than life size 
.
.
You don't gain focal length shooting in crop mode. Nothing changes with the optics. Simply limits the field of view.Agree with previous poster - to each their own. Personally, I shoot my Z9 in DX when I need the extra focal length...I just don't see myself making a 24x36 print of even the best warbler portrait ever taken. It would be weird to have a picture of a warbler on my wall ten times larger than life size
And yes I totally agree that printing tiny birds huge is just.... weird
You gain apparent focal length. Same way that using a 600 mm lens on a crop sensor body effectively gives you a 900 mm lens. Granted, it achieves that by just throwing out the outer 50% of the pixels as opposed to cropping in post, but that's where you get the benefit of the smaller file size.
Shooting in DX mode does not increase the reach of a lens.
Guys you are just going round in circles. Steve Perry was interviewed yesterday and this question came up. His answer was along the lines of: “ back in the DSLR days we used to see the outer area (ie not the DX Crop) greyed out. Now with mirrorless bodies the Dx cropped area 7s magnified so it fills the EVF - however all the camera is doing is turning off the photsites outside the crop reducing the amount of data by more than half (45.u—>~20mp). There is no difference between selecting a DX crop in camera and applying a DX crop to an image later in Lightroom, one is just choosing to not use some of the data. For some cameras using a DX crop can improve eye detection, speed up fps and certainly results in smaller file sizes”.
No change in actual focal length but a reduced area for the camera to look in when using some AF modes.
No change in actual focal length but a reduced area for the camera to look in when using some AF modes.
simonsi
Active member
3 (better AF), may not be so in reality, while the image appears bigger in the frame, it covers exactly the same sensor pixels as before in fx mode so I cant see the AF system seeing anything different save potentially less data to process.
Zof9
Member
With less image for the camera to scan, more chance of latching onto to your intended target?
I usually stay in FX on the Z7II…haven’t used the 9 extensively yet. Since the same AF sites are used but just less of them…seems like the only AF adv the is maybe slightly faster since it is ignoring half of the sites. I find it easier to keep the BIF in the viewfinder in FX on the 7II so just crop in post.
peterlin321
New member
Excellent! However I'm still not clear about your point 5: Effective lens length is increased by 1.5x. In Z9 "still" mode, if I switch from FX to DX image area I can see lens length increased (image appears larger). In video if I set to 4K slow motion (120p) then switch from FX to DX in "video" mode, the image appears to be even larger than the above DX still. It's like it's been cropped twice. So what's the effective lens length increase ratio for video when shooting 4K slow motion DX mode - larger than 1.5x? I may not want image to be double cropped for many practical reasons.
Chris K
Well-known member
I'm not a video expert by any means, so maybe someone who knows video can chime in. I DO know some specific video modes apply a crop, so the camera doesn't have to do complex interpolation.
In other words, if your sensor is 8 pixels wide, you can have an 8-pixel video, but with some really simple interpolation (called "binning") you can also have a 4-pixel or 2-pixel video. The problem comes when you want....say, a 5-pixel video. Now you have to apply a resampling algorithm. Those can look awful with high-frequency details, and can take a significant amount of processing power.
The solution is generally to crop the sensor down, so you're only using the "middle 5 pixels". Now there's no interpolation, but you just "discarded" part of the sensor, and thus have an additional crop.
Some cameras do this more than others. For the Z9, I know video people were pretty happy that it doesn't do much of this cropping, so I suggest checking your video settings (ie: 12-bit H264, 14-bit F.log, and all that other crazy video stuff) to see if you just happen to be in a mode that requires a crop.
In other words, if your sensor is 8 pixels wide, you can have an 8-pixel video, but with some really simple interpolation (called "binning") you can also have a 4-pixel or 2-pixel video. The problem comes when you want....say, a 5-pixel video. Now you have to apply a resampling algorithm. Those can look awful with high-frequency details, and can take a significant amount of processing power.
The solution is generally to crop the sensor down, so you're only using the "middle 5 pixels". Now there's no interpolation, but you just "discarded" part of the sensor, and thus have an additional crop.
Some cameras do this more than others. For the Z9, I know video people were pretty happy that it doesn't do much of this cropping, so I suggest checking your video settings (ie: 12-bit H264, 14-bit F.log, and all that other crazy video stuff) to see if you just happen to be in a mode that requires a crop.
Thanks Peter. DX mode for 4k 120 in video is a bit different. I have just checked this in the menus on my Z9 and I think these are the options we have. You can either do FX mode 4k 120 with no crop at all or you can do 3.8k 120 with a 2.3 crop. I guess in the 2.3 crop mode the camera is doing a pixel for pixel readout and that is why Nikon have gone with this specific crop size.
I think you are spot on here Chris. I think for the FX mode 4k 120 there is no crop and the Z9 is binning pixels. Then for 3.8K 120 there is a crop of 2.3 and the Z9 is using all the pixels in the cropped area and is therefore not having to do any processor intensive interpolation.
ninog
Active member
Good post... anyway I'm not sure if the trim option only reduces the mp... or is something else lost?? According to this graph, dynamic range is lost? I ask... I don't understand the graphs very well and sometimes my experience differs a little from them....
Thanks. Very interesting link which definitely suggests that you get just under 1 stop less dynamic range in relation to the ISO value in DX mode. I wonder if it is as simple as that or if the reduced sensor area has an impact on the dynamic range measurement. If this dynamic range reduction is for real it would be good to know why it happens.
peterlin321
New member
Interesting points. Thank you all.