Full-frame/APS-C Lens equivalence?

[brettgoulder]

Hi,

I know full-frame vs APS-C topics are done-to-death on this forum, but I was hoping someone could clarify my thinking. I am currently comparing a Fujifilm X-H2S and a Sony A7IV and had questions about how to compare lenses.

One thing that gets brought up quite a bit on this forum is that full-frame lenses are more costly and bigger than APS-C. However, my understanding is that the DOF difference between an APS-C lens and a full-frame lens is one stop. Also, the noise performance for full-frame is a 1-1.5 stops better than APS-C. Is it not more appropriate to compare an APS-C f/2.8 lens to a full-frame f/4 lens?

So in this case, these two would be roughly equivalent:
- Fujinon 16-55mm f/2.8 ($1099), 655g, 83.3 x 106 mm
- Sony 20-70mm f/4 ($1099), 488g, 78.7 x 99 mm

Rather than comparing the Fuji lens to $2200 Sony's 24-70 GM II. In this case, the Sony lens actually is smaller, same price, and beats the Fuji on the wide-end.

Since the noise performance is better by 1-1.5 stops on the full frame sensor, then I can just bump the ISO by a stop to get the same image quality.

Is my thinking flawed somehow? If someone could clarify this for me, I would appreciate it.

 

[sybersitizen]

... my understanding is that the DOF difference between an APS-C lens and a full-frame lens is one stop.

Close, though it's actually a bit more.

Also, the noise performance for full-frame is a 1-1.5 stops better than APS-C. Is it not more appropriate to compare an APS-C f/2.8 lens to a full-frame f/4 lens?

It's appropriate.

So in this case, these two would be roughly equivalent:
- Fujinon 16-55mm f/2.8 ($1099), 655g, 83.3 x 106 mm
- Sony 20-70mm f/4 ($1099), 488g, 78.7 x 99 mm

Roughly, yes.

Since the noise performance is better by 1-1.5 stops on the full frame sensor, then I can just bump the ISO by a stop to get the same image quality.

Is my thinking flawed somehow?

It's not flawed in theoretical terms, though it could be flawed in practice. That would depend on the actual performance characteristics of the specific cameras and lenses.

 

[chrisfisheye]

Hi,

I know full-frame vs APS-C topics are done-to-death on this forum, but I was hoping someone could clarify my thinking. I am currently comparing a Fujifilm X-H2S and a Sony A7IV and had questions about how to compare lenses.

One thing that gets brought up quite a bit on this forum is that full-frame lenses are more costly and bigger than APS-C. However, my understanding is that the DOF difference between an APS-C lens and a full-frame lens is one stop. Also, the noise performance for full-frame is a 1-1.5 stops better than APS-C. Is it not more appropriate to compare an APS-C f/2.8 lens to a full-frame f/4 lens?

So in this case, these two would be roughly equivalent:
- Fujinon 16-55mm f/2.8 ($1099), 655g, 83.3 x 106 mm
- Sony 20-70mm f/4 ($1099), 488g, 78.7 x 99 mm

Rather than comparing the Fuji lens to $2200 Sony's 24-70 GM II. In this case, the Sony lens actually is smaller, same price, and beats the Fuji on the wide-end.

Since the noise performance is better by 1-1.5 stops on the full frame sensor, then I can just bump the ISO by a stop to get the same image quality.
Is my thinking flawed somehow? If someone could clarify this for me, I would appreciate it.

I made the switch from the Sony 16-55mm aps-c to Sony 20-70mm FF and I am very happy with it.

So I use now a FF camera. (A7CR)

It is important to use an high res sensor if you want to have more versatility with your lenses. I can also keep my aps-c lenses (like the 70-350mm)

 

[MJ_Photo38]

Hi,

I know full-frame vs APS-C topics are done-to-death on this forum, but I was hoping someone could clarify my thinking. I am currently comparing a Fujifilm X-H2S and a Sony A7IV and had questions about how to compare lenses.

One thing that gets brought up quite a bit on this forum is that full-frame lenses are more costly and bigger than APS-C. However, my understanding is that the DOF difference between an APS-C lens and a full-frame lens is one stop.

Actually, for DOF equivalence (and only DOF!) you can apply the crop factor to the aperture to get an idea of the lens needed on full frame to achieve the exact same image.

For example, a 23mm f/2 on APS-C would require a 35mm lens on full frame to achieve the FOV, and, it would require you to set the aperture on that lens to f/3 to have the exact same DOF (roughly. That's not 100% correct). Closest value to this are f/2.8 and f/3.2, so it's between 1 and 1.33 stop difference.

Also, the noise performance for full-frame is a 1-1.5 stops better than APS-C. Is it not more appropriate to compare an APS-C f/2.8 lens to a full-frame f/4 lens?

It's not always the case, it really depends on the sensors you're comparing. If you compare a Sony 61MP sensor with a 26MP APS-C, the pixel pitch is identical and the noise on a pixel level will be the exact same.

However, if you're comparing two 24MP sensors, one being FF, the other being APS-C, well this is where you see that 1/1.3 stop difference (it's rarely 1.5 stop. On the few testings I did, I was always closer to 1.0 stop)

So in this case, these two would be roughly equivalent:
- Fujinon 16-55mm f/2.8 ($1099), 655g, 83.3 x 106 mm
- Sony 20-70mm f/4 ($1099), 488g, 78.7 x 99 mm

Rather than comparing the Fuji lens to $2200 Sony's 24-70 GM II. In this case, the Sony lens actually is smaller, same price, and beats the Fuji on the wide-end.

I guess this is all a matter of needs. If we're comparing the lenses only, then sure. If we put the camera bodies into the mix, that's another debate.

Both the X-H2 and X-H2S beat the A7IV on specs alone, by quite a large margin when it comes to burst rates and video specs. APS-C isn't competitive with the lenses alone.

Since the noise performance is better by 1-1.5 stops on the full frame sensor, then I can just bump the ISO by a stop to get the same image quality.

Technically, yes.

 

[PhotoTeach2]

Hi,

I know full-frame vs APS-C topics are done-to-death on this forum, but I was hoping someone could clarify my thinking. I am currently comparing a Fujifilm X-H2S and a Sony A7IV and had questions about how to compare lenses.

One thing that gets brought up quite a bit on this forum is that full-frame lenses are more costly and bigger than APS-C. However, my understanding is that the DOF difference between an APS-C lens and a full-frame lens is one stop. Also, the noise performance for full-frame is a 1-1.5 stops better than APS-C. Is it not more appropriate to compare an APS-C f/2.8 lens to a full-frame f/4 lens?

So in this case, these two would be roughly equivalent:
- Fujinon 16-55mm f/2.8 ($1099), 655g, 83.3 x 106 mm
- Sony 20-70mm f/4 ($1099), 488g, 78.7 x 99 mm

You are correct that the Field/Angle-of-View (FOV/AOV) is roughly equivalent (FF 24-84mm would be closer).

And you are correct that the DOF and NOISE would be equivalent.

But ... the "exposures" are different, as the f/2.8 is 1-stop "faster" vs the (slower) f/4.

So either the SS or ISO would need to be +/- 1-stop between them.

Equivalency only applies FOV/AOV & DOF & NOISE.

Rather than comparing the Fuji lens to $2200 Sony's 24-70 GM II. In this case, the Sony lens actually is smaller, same price, and beats the Fuji on the wide-end.

Yes, but the Fuji beats it on the TELE end.

Since the noise performance is better by 1-1.5 stops on the full frame sensor, then I can just bump the ISO by a stop to get the same image quality.

Yes, and you will indeed need to bump it by 1-stop to get the same exposure (light/unit-area), or use 1-stop longer Shutter-Speed.

The Fuji has the advantage if you want a faster SS (at same ISO).

Is my thinking flawed somehow? If someone could clarify this for me, I would appreciate it.

You are correct as far as you went, but sometimes a faster SS is critical/needed, especially in low-light when you may not want to increase ISO, (and risk additional noise).

 

[John Crowe]

Your explanation coincides with all the techies, and for those who only use one or the other format.

I have used APS-C alongside FF for 8 of my 17 digital years and never think in terms of the points that you have made. I use APS-C for telephoto photography and I use FF for wideangle photography. When I use the APS-C I am taking advantage of the crop factor to get more MP on the subject than I can with FF. With APS-C I am generally shooting at the lowest possible ISO, wide open, with the fastest possible shutter speeds to stop motion in high speed sports. I only consider the maximum possible aperture in terms of being able to select the fastest possible shutter speed, and not anything to do with relative DOF.

For me a 400/2.8 lens effectively becomes a 600/2.8 or 640/2.8 on each APS-C. I don't care how it may change the DOF and since I shoot below ISO 800, I don't care about the slight difference in noise either.

This is a different perspective for those using both systems to their strengths.

 

[ThrillaMozilla]

Hi,

I know full-frame vs APS-C topics are done-to-death on this forum, but I was hoping someone could clarify my thinking. I am currently comparing a Fujifilm X-H2S and a Sony A7IV and had questions about how to compare lenses.

One thing that gets brought up quite a bit on this forum is that full-frame lenses are more costly and bigger than APS-C. However, my understanding is that the DOF difference between an APS-C lens and a full-frame lens is one stop. Also, the noise performance for full-frame is a 1-1.5 stops better than APS-C. Is it not more appropriate to compare an APS-C f/2.8 lens to a full-frame f/4 lens?

So in this case, these two would be roughly equivalent:
- Fujinon 16-55mm f/2.8 ($1099), 655g, 83.3 x 106 mm
- Sony 20-70mm f/4 ($1099), 488g, 78.7 x 99 mm

Rather than comparing the Fuji lens to $2200 Sony's 24-70 GM II. In this case, the Sony lens actually is smaller, same price, and beats the Fuji on the wide-end.

Since the noise performance is better by 1-1.5 stops on the full frame sensor, then I can just bump the ISO by a stop to get the same image quality.
Is my thinking flawed somehow? If someone could clarify this for me, I would appreciate it.

Your question will probably prompt dissertations, but yes, you got it right. It's a pleasure to see that.

The difference is about 1 1/6 stop. Canon APS-C sensors are slightly smaller, so it's about 1 1/3 stop for Canon. 1 stop difference is close enough for practical purposes.

Full frame lenses do tend to be bigger and more expensive, but that is not always true. The difference in size is usually greater for the longer lenses, but may not apply to the shorter lenses.

 

[sybersitizen]

You are correct that the Field/Angle-of-View (FOV/AOV) is roughly equivalent (FF 24-84mm would be closer).

And you are correct that the DOF and NOISE would be equivalent.

But ... the "exposures" are different, as the f/2.8 is 1-stop "faster" vs the (slower) f/4.

The exposures would be the same when the full frame ISO is raised accordingly, as the OP said ... and so would the noise. But you know that.

So either the SS or ISO would need to be +/- 1-stop between them.

Equivalency only applies FOV/AOV & DOF & NOISE.

And EXPOSURE through the use of ISO. ISO is the thing that allows the other things to remain equivalent.

 

[Dennis]

Is it not more appropriate to compare an APS-C f/2.8 lens to a full-frame f/4 lens?

That's a fair comparison *IF* you want to compare equivalent lenses. For some people, the point of buying a FF camera is to exploit the larger sensor in low light.

Since the noise performance is better by 1-1.5 stops on the full frame sensor, then I can just bump the ISO by a stop to get the same image quality.
Is my thinking flawed somehow? If someone could clarify this for me, I would appreciate it.

In terms of equivalence, that's an entirely reasonable comparison. Outside of equivalence, you have all the other pros & cons of any two pieces of equipment .

When I chose my mirrorless camera to replace my DSLR, I'd honed in on wanting a good 24-200ish lens to cover a lot of my shooting needs without needing to change lenses often. I dismissed the Sony and Canon 24-240s as being mediocre lenses (and wasn't interested in compromising on the wide end with a Tamron 28-200). That left the OM 12-100/4 and the Nikon 24-200. The Nikon kit was smaller, lighter, cheaper and the FF sensor makes up for the slower max aperture. There's not much low light advantage, but there doesn't need to be. I had enough other reasons for choosing it. I've always relied on faster primes for low light anyway; preferring a zoom with a wider range.

--
Dennis
Gallery at http://kingofthebeasts.smugmug.com

 

[kolyy]

Hi,

I know full-frame vs APS-C topics are done-to-death on this forum, but I was hoping someone could clarify my thinking. I am currently comparing a Fujifilm X-H2S and a Sony A7IV and had questions about how to compare lenses.

One thing that gets brought up quite a bit on this forum is that full-frame lenses are more costly and bigger than APS-C. However, my understanding is that the DOF difference between an APS-C lens and a full-frame lens is one stop. Also, the noise performance for full-frame is a 1-1.5 stops better than APS-C. Is it not more appropriate to compare an APS-C f/2.8 lens to a full-frame f/4 lens?

So in this case, these two would be roughly equivalent:
- Fujinon 16-55mm f/2.8 ($1099), 655g, 83.3 x 106 mm
- Sony 20-70mm f/4 ($1099), 488g, 78.7 x 99 mm

Rather than comparing the Fuji lens to $2200 Sony's 24-70 GM II. In this case, the Sony lens actually is smaller, same price, and beats the Fuji on the wide-end.

Since the noise performance is better by 1-1.5 stops on the full frame sensor, then I can just bump the ISO by a stop to get the same image quality.
Is my thinking flawed somehow? If someone could clarify this for me, I would appreciate it.

Yes, that's quite correct and not really rocket science. Beware that some people do not like this (usually users of cameras with smaller sensors) and try to create a smokescreen by talking about "exposure" etc, which has zero substance to it.

 

[Bob]

Look at it this way: The sensor size is irrelevant. Just compare the two products as-is and forget the specs.

• Which feels better in your hands?
• Which produces better images for your taste?
• Which has a good selection of lenses that feel good in your hands.
• Which produces the kind of DOF and bokeh you like?

Specs exist to distract and confuse. This in turn stimulates consumption. People get fixated on numbers. If "3" is good then "3.1" must be better.

 

[PhotoTeach2]

Your explanation coincides with all the techies, and for those who only use one or the other format.

I have used APS-C alongside FF for 8 of my 17 digital years and never think in terms of the points that you have made. I use APS-C for telephoto photography and I use FF for wideangle photography. When I use the APS-C I am taking advantage of the crop factor to get more MP on the subject than I can with FF. With APS-C I am generally shooting at the lowest possible ISO, wide open, with the fastest possible shutter speeds to stop motion in high speed sports. I only consider the maximum possible aperture in terms of being able to select the fastest possible shutter speed, and not anything to do with relative DOF.

For me a 400/2.8 lens effectively becomes a 600/2.8 or 640/2.8 on each APS-C. I don't care how it may change the DOF and since I shoot below ISO 800, I don't care about the slight difference in noise either.

This is a different perspective for those using both systems to their strengths.

BINGO !!!

Equivalency is a fool's errand because both larger and smaller sensors have both advantages and disadvantages.

If you want lowest noise and highest IQ, use FF (or even MF).

If you wang longest TELE, use 1/2.3" (aka Nikon P950 & 1000 with 2000/3000mm-EFL).

Both 1"-type and m4/3, (and DX/APS) are COMPROMISES between these.

DX/APS was an "economical" compromise because FF sensors were originally prohibitively expensive ... and were the smallest sensor that still allowed usable OVF.

 

[PhotoTeach2]

You are correct that the Field/Angle-of-View (FOV/AOV) is roughly equivalent (FF 24-84mm would be closer).

And you are correct that the DOF and NOISE would be equivalent.

But ... the "exposures" are different, as the f/2.8 is 1-stop "faster" vs the (slower) f/4.

The exposures would be the same when the full frame ISO is raised accordingly, as the OP said ... and so would the noise.

OP said: "bump the ISO by a stop to get the same image quality"

It looked to me like he was (only) referring to IQ (noise) ... and not "exposure"

The fact is that he HAS to bump the FF- ISO by one stop to retain the same lightness (lost due to the slower f/4).

However, increasing the FF-ISO results in MORE "noise" than possible with a lower-ISO. So you have then lost the inherent higher-IQ (noise) advantage from the larger-sensor due to the "slower"-lens.

But you know that.

There are 3 different contexts that the word "exposure" is associated with.

1. "Exposure" as light/unit-area (and the f/2.8 provides 1-stop greater light/unit-area -- irregardless of sensor size)

2. Exposure-"settings" (which can be the exact same between different sensor sizes).

3. Exposure = "lightness" (of a final image)

ONLY #1 is the most accurate ... #3 is the least-accurate since it is really only correcting "lightness" -- after a change (in lightness) due to "exposure" change.

So either the SS or ISO would need to be +/- 1-stop between them.

Equivalency only applies FOV/AOV & DOF & NOISE.

And EXPOSURE through the use of ISO. ISO is the thing that allows the other things to remain equivalent.

ISO does not (directly) affect "exposure", (only "lightness").

It may indirectly affect exposure if shooting in "auto" because the camera will change exposure to correct for lightness change.

NOTE that Exposure-Meters and the Sunny-16 rule apply equally to all sensor sizes, (aka 1/2.3" to FF and MF and even 8"x10" etc.).

So changing between f/2.8 & f/4 will require a change in either SS or ISO to maintain same exposure(SS) or lightness(ISO), irregardless of sensor size.

 

[Dennis]

The Fuji has the advantage if you want a faster SS (at same ISO).

But why would anyone insist on shooting the same ISO regardless of format?

Is my thinking flawed somehow? If someone could clarify this for me, I would appreciate it.

You are correct as far as you went, but sometimes a faster SS is critical/needed, especially in low-light when you may not want to increase ISO, (and risk additional noise).

You're not risking additional noise by shooting FF 1 stop higher than APS-C.

 

[Dennis]

So you have then lost the inherent higher-IQ (noise) advantage from the larger-sensor due to the "slower"-lens.

Yes, that's the whole idea behind equivalence. No advantage; no disadvantage. Equvalent.

Some people want FF for the supposed low light advantage (which requires compromises). But the OP sounds like he's just comparing two kits and looking to make sure he's not missing out on anything by going with the FF option, which offers a lens range he seems to like better.

 

[sybersitizen]

You are correct that the Field/Angle-of-View (FOV/AOV) is roughly equivalent (FF 24-84mm would be closer).

And you are correct that the DOF and NOISE would be equivalent.

But ... the "exposures" are different, as the f/2.8 is 1-stop "faster" vs the (slower) f/4.

The exposures would be the same when the full frame ISO is raised accordingly, as the OP said ... and so would the noise.

OP said: "bump the ISO by a stop to get the same image quality"

It looked to me like he was (only) referring to IQ (noise) ... and not "exposure"

If the exposures are not kept constant in the comparison as described above, image quality can go all over the place in any direction. The OP was clearly talking about equivalence, where exposures are kept constant.

And yes, discussions about equivalence are concerned with exposure as it relates to total light captured, not exposure as it relates to the Sunny 16 rule.

The fact is that he HAS to bump the FF- ISO by one stop to retain the same lightness (lost due to the slower f/4).

You can say it that way or you can say he has to raise ISO to maintain the same exposure. Makes no difference.

However, increasing the FF-ISO results in MORE "noise" than possible with a lower-ISO. So you have then lost the inherent higher-IQ (noise) advantage from the larger-sensor due to the "slower"-lens.

We're talking about equivalence. Same noise. We are not talking about advantages.

So either the SS or ISO would need to be +/- 1-stop between them.

Equivalency only applies FOV/AOV & DOF & NOISE.

And EXPOSURE through the use of ISO. ISO is the thing that allows the other things to remain equivalent.

ISO does not (directly) affect "exposure", (only "lightness").

A change in ISO permits a change in exposure so the two comparisons can actually be equivalent: same lightness, same shutter speed, same DOF, same angle of view, same noise.

 

[PhotoTeach2]

You are correct that the Field/Angle-of-View (FOV/AOV) is roughly equivalent (FF 24-84mm would be closer).

And you are correct that the DOF and NOISE would be equivalent.

But ... the "exposures" are different, as the f/2.8 is 1-stop "faster" vs the (slower) f/4.

The exposures would be the same when the full frame ISO is raised accordingly, as the OP said ... and so would the noise.

OP said: "bump the ISO by a stop to get the same image quality"

It looked to me like he was (only) referring to IQ (noise) ... and not "exposure"

If the exposures are not kept constant in the comparison as described above, image quality can go all over the place in any direction. The OP was clearly talking about equivalence, where exposures are kept constant.

But which definition of exposure are you using, (#1, #2, or #3) ??? (only #1 is correct)

And yes, discussions about equivalence are concerned with exposure as it relates to total light captured, not exposure as it relates to the Sunny 16 rule.

Exposure is light/unit-area. Total-light-captured is more related to the the size (area) of the sensor -- but still same light/unit-area.

The fact is that he HAS to bump the FF- ISO by one stop to retain the same lightness (lost due to the slower f/4).

You can say it that way or you can say he has to raise ISO to maintain the same exposure. Makes no difference.

Then you are using definition #3, which is really "lightness". And it can potentially increase noise (if the level becomes visibly-noticeable).

However, increasing the FF-ISO results in MORE "noise" than possible with a lower-ISO. So you have then lost the inherent higher-IQ (noise) advantage from the larger-sensor due to the "slower"-lens.

We're talking about equivalence. Same noise. We are not talking about advantages.

If you did not use longer-SS (to compensate for the smaller f/stop), then the potential of noticably-objectionable noise is higher.

So either the SS or ISO would need to be +/- 1-stop between them.

Equivalency only applies FOV/AOV & DOF & NOISE.

And EXPOSURE through the use of ISO. ISO is the thing that allows the other things to remain equivalent.

ISO does not (directly) affect "exposure", (only "lightness").

A change in ISO permits a change in exposure so the two comparisons can actually be equivalent: same shutter speed, same DOF, same angle of view, same noise.

I agree you can do that with only the minor differences between DX/APS to FF ... but can you apply the same rationale to 1/2.3" vs FF ???

Both have advantages and disadvantages.

You muddy the waters of a very simple concept every single time this comes up.

 

[PhotoTeach2]

The Fuji has the advantage if you want a faster SS (at same ISO).

But why would anyone insist on shooting the same ISO regardless of format?

You always want to shoot at "lowest" ISO whenever possible.

If you have a slower lens (aka f/4 vs f/2.8), you have to use a higher ISO to compensate for the lower light/unit-area, (or longer SS). Both negative compared to the f/2.8.

Is my thinking flawed somehow? If someone could clarify this for me, I would appreciate it.

You are correct as far as you went, but sometimes a faster SS is critical/needed, especially in low-light when you may not want to increase ISO, (and risk additional noise).

You're not risking additional noise by shooting FF 1 stop higher than APS-C.

Of course you are ... but I agree that 1-stop may not be enough to notice (especially since I have espoused that there is NO-noise until it becomes noticeably-objectionable).

But without changing SS, there is "lower" exposure from the f/4, (requiring an increased ISO to compensate for lower "lightness").

And sensor-size is irrelevant in the context of f/2.8 vs f/4 light/unit-area (exposure).

 

[sybersitizen]

You are correct that the Field/Angle-of-View (FOV/AOV) is roughly equivalent (FF 24-84mm would be closer).

And you are correct that the DOF and NOISE would be equivalent.

But ... the "exposures" are different, as the f/2.8 is 1-stop "faster" vs the (slower) f/4.

The exposures would be the same when the full frame ISO is raised accordingly, as the OP said ... and so would the noise.

OP said: "bump the ISO by a stop to get the same image quality"

It looked to me like he was (only) referring to IQ (noise) ... and not "exposure"

If the exposures are not kept constant in the comparison as described above, image quality can go all over the place in any direction. The OP was clearly talking about equivalence, where exposures are kept constant.

But which definition of exposure are you using, (#1, #2, or #3) ??? (only #1 is correct)

And yes, discussions about equivalence are concerned with exposure as it relates to total light captured, not exposure as it relates to the Sunny 16 rule.

I withdraw the use of the term exposure. Now, exactly what point do you wish to make?

 

[Great Bustard]

For me a 400/2.8 lens effectively becomes a 600/2.8 or 640/2.8 on each APS-C. I don't care how it may change the DOF and since I shoot below ISO 800, I don't care about the slight difference in noise either.

...if you don't care about the [1.3 stop] difference in DOF and noise, then the 400 / 2.8 lens on APS-C "effectively becomes" (i.e. "is equivalent to") a 640 / 4.5 on FF, not a 640 / 2.8. That is, 400mm f/2.8 1/800 ISO 400 on APS-C has the same angle of view, same DOF, same noise (for similar sensor tech), and same motion blur as 640mm f/4.5 1/800 ISO 1000 on FF.