Why a small sensor will give you sharper images

[v1fan]

There are two types of sharpness, resolution/detail (sensor and lens) and focus/clarity (depth of field).

Resolution and sharpness of the sensor/lens is useless if the camera is not focused properly or the depth of field is lacking. On the other hand a low quality sensor and lens with good depth of field provides acceptable sharp images with everything in focus and also can be sharpened later.....

*You CAN sharpen for resolution and detail but you CANNOT sharpened for depth of field.





The small sensor gives you the focus and depth of field in even wide apertures and that means you can keep the ISO at minimum and shoot at wide open apertures. The D3s needs to shoot at ISO 650 in order to have the same depth of field as V1 at ISO100 for the same shutter speed.

 

[Jared Huntr]

v1fan wrote:
The D3s needs to shoot at ISO 650 in order to have the same depth of field as V1 at ISO100 for the same shutter speed.

Yes, but the assumption is that you suffer no IQ loss by shooting at ISO 650 on a full frame. In fact, some would argue that ISO 650 is cleaner, has less noise, and therefore captures more detail than ISO 100 on the V1. I don't know if this is necessarily true, but that is the crux of the argument against your original premise.

 

[ehr444]

Tony Sweet says in his lectures that to get sharp images you need to shoot at f/22, well past the defraction limit. Why does he say this? Because he needs depth of field!

He shoots a D3s. Now if he were to shoot with a V1 or a V2, he could accomplish the same depth of field by shooting at f/22 / 2.7 = f/8. This allows shooting at a much lower ISO or shorter shutter speed.

Intrinsic noise of the sensor may be an issue that needs quantification. However it seems inevitable that the smaller sensor eventually is going to outperform.

 

[Martin Datzinger]

even at the same apertures values as low as, let's say f/4, on both cameras.

Why? Because there are 3 Nikon brand tilt-shift lenses for FX, allowing Scheimpflug at suitable FOVs and there are none für CX, not even third party.

Best regards,

Martin

 

[ehr444]

Martin Datzinger wrote:

even at the same apertures values as low as, let's say f/4, on both cameras.

Why? Because there are 3 Nikon brand tilt-shift lenses for FX, allowing Scheimpflug at suitable FOVs and there are none für CX, not even third party.

Tilt-shift lenses help in specialized situations. They won't help get a deep scene in focus from foreground to background, like a whole room or landscape.

And I haven't tried it but they should work on the N1 system via the FT1 adapter.

---


Best,


Edward H Russell

 

[Martin Datzinger]

Of course they do. That's what they're made for. Vertical foreground objects might become a problem, I'll give you that.

And the CX format applies a 2.7x FL multiplicator. So a 24 PC-E behaves like a 65mm, not exactly wideangle and what one has spent 1900€ for in the first place.

 

[Martin Datzinger]

Again: According to dxomark (sorry, I don't have other means to compare), normalised to 8MP (the print setting on their compar-o-meter, the D800E can be stopped down 3 to 4 stops until it worsens to V1 levels, in all their measured IQ parameters. Which incidently completely evens out the maximum DoF advantage, that is claimed here.

 

[ehr444]

Martin Datzinger wrote:

Again: According to dxomark (sorry, I don't have other means to compare), normalised to 8MP (the print setting on their compar-o-meter, the D800E can be stopped down 3 to 4 stops until it worsens to V1 levels, in all their measured IQ parameters. Which incidently completely evens out the maximum DoF advantage, that is claimed here.

 

[Jared Huntr]

ehr444 wrote:

Martin Datzinger wrote:

Again: According to dxomark (sorry, I don't have other means to compare), normalised to 8MP (the print setting on their compar-o-meter, the D800E can be stopped down 3 to 4 stops until it worsens to V1 levels, in all their measured IQ parameters. Which incidently completely evens out the maximum DoF advantage, that is claimed here.

 

[ehr444]

Jared Huntr wrote:

It would be interesting to add a more practical reality to the test...

Keep our budget realistic and use only entry level or non-professional dSLRs . The rationale being that a typical V1 owner has either upgraded from an older entry level dSLR e.g. D80/D90/D5000 or is trying to choose between the V1 and a current entry level dSLR like the D5100 or D3200.

If the V1 can hold its ground against the D3x, I expect it would do even better against the entry level models. But the D5100 and D3200 are 1.5x crop DX cameras. They don't have as wide a spread in DOF against the V1 as the FX cameras do.

Another realistic test is to use a fast prime + FT1 on the V1 (say a 35mm 1.8) comparing to the same prime on the dSLR and frame identically by walking closer/farther to the subject. This would be another test that more closely matches a typical use case in reality.

Changing the shooting position will have a major effect on DOF. That is why I shot with the same (zoom) lens and from the same distance.

"Lab tests" that don't reflect how you are going to use the camera in the field don't count for much. They are just theoretical curiosities that prove nothing.

I think the test I did shows that at the same shooting distance and same frame coverage, the CX format camera can perform well, giving equivalent DOF at 2.7 stops smaller aperture and 2.7x smaller focal length, than an FX format camera.


---

Best,

Edward H Russell

 

[Kaj E]

Martin Datzinger wrote:

Again: According to dxomark (sorry, I don't have other means to compare), normalised to 8MP (the print setting on their compar-o-meter, the D800E can be stopped down 3 to 4 stops until it worsens to V1 levels, in all their measured IQ parameters. Which incidently completely evens out the maximum DoF advantage, that is claimed here.

 

[Jared Huntr]

ehr444 wrote:

Jared Huntr wrote:

It would be interesting to add a more practical reality to the test...

Keep our budget realistic and use only entry level or non-professional dSLRs . The rationale being that a typical V1 owner has either upgraded from an older entry level dSLR e.g. D80/D90/D5000 or is trying to choose between the V1 and a current entry level dSLR like the D5100 or D3200.

If the V1 can hold its ground against the D3x, I expect it would do even better against the entry level models. But the D5100 and D3200 are 1.5x crop DX cameras. They don't have as wide a spread in DOF against the V1 as the FX cameras do.

Another realistic test is to use a fast prime + FT1 on the V1 (say a 35mm 1.8) comparing to the same prime on the dSLR and frame identically by walking closer/farther to the subject. This would be another test that more closely matches a typical use case in reality.

Changing the shooting position will have a major effect on DOF. That is why I shot with the same (zoom) lens and from the same distance.

Yes, I realize that but it is a realistic use case for many owners coming from legacy dSLRs whose high ISO performance was the same or worse than the V1. Being forced to reframe with a cropped sensor adds another dimension to the issue because it changes the perspective and deepens DOF. They can shoot at the same shutter speed and aperture as before with their dSLR but now be able to capture more depth while filling the entire frame.


For example, compare the V1 against the much larger sensor dSLR D200 at ISOs 800 and above:

Loading…

www.2cameraguys.com

Small sensors have come a long way and surpass dSLR standards that we wouldn't even question several years ago. That is why I recently called out a D200 owner that criticized the V1 sensor as typical of a small sensor with low dynamic range that blows high-lights.

 

[pengch]

ehr444 wrote:

Martin Datzinger wrote:

Because there are 3 Nikon brand tilt-shift lenses for FX,

Tilt-shift lenses help in specialized situations. They won't help get a deep scene in focus from foreground to background, like a whole room or landscape.

some day i'd like to see a T/S adapter, i have seen some one made a sony nex t/s adapter.