compact v dslr, depth of field and ISO

[richarddd]

Would I have been able to take this picture with the same DOF and better image quality with a DSLR?

I shot this at 28mm, 1/4 sec, f/2 and ISO 200 with a compact. To get the same DOF with a DSLR, I'd likely have to stop down a lot. Let's say f/8 with a DSLR. That's 4 stops slower, so if we leave shutter speed constant, I'd need an ISO of 3200, which would likely be much noisier unless I used a very high end DSLR. Alternatively, I'd need to slow the shutter to 4 secs, which would require much steadier support than I had (or some combination of higher ISO and slower shutter speed).

Am I overstating the issues with a DSLR?





--
http://fruminousbandersnatch.blogspot.com/

 

[hahncholo]

If you want deep depth of field, then yes, a compact does that easier. However, many people see the ability to have a shallow DOF as a larger advantage. It is more difficult to get a large DOF with a DSLR, but it's impossible to get a shallow one with a compact, since the maximum relative aperture is far smaller.

 

[godfrog]

Would I have been able to take this picture with the same DOF and better image quality with a DSLR?

I shot this at 28mm, 1/4 sec, f/2 and ISO 200 with a compact. To get the same DOF with a DSLR, I'd likely have to stop down a lot. Let's say f/8 with a DSLR. That's 4 stops slower, so if we leave shutter speed constant, I'd need an ISO of 3200, which would likely be much noisier unless I used a very high end DSLR. Alternatively, I'd need to slow the shutter to 4 secs, which would require much steadier support than I had (or some combination of higher ISO and slower shutter speed).

Am I overstating the issues with a DSLR?





--
http://fruminousbandersnatch.blogspot.com/

You would not be able to take a considerably better image with a DSLR without longer exposure time. However, with the same exposure time, you could get an image just as good.

The crop factor between 1/1.7" sensors and APS-C is 3, so you'd get the same field of view with a 18mm lens, and you'd get the same depth of field with the lens at F6. This is about 3 stops darker, so you'd need to use ISO1600. The image quality of a K-5 or D7000 at ISO1600 easily equal the S95 at ISO200.

Compact cameras never have an advantage in terms of ability to achieve deep DOF. At best they are "not worse", but as soon soon as there is more light or a tripod, the DSLR will get the advantage by getting the same DOF but much higher dynamic range and better general image quality.

--
My Flickr:
http://www.flickr.com/photos/36164047@N06/

 

[richarddd]

The crop factor between 1/1.7" sensors and APS-C is 3, so you'd get the same field of view with a 18mm lens, and you'd get the same depth of field with the lens at F6. This is about 3 stops darker, so you'd need to use ISO1600.

Aha! You're right - apply the crop factor and you get to f/6 for the same DOF.

The image quality of a K-5 or D7000 at ISO1600 easily equal the S95 at ISO200.

Compact cameras never have an advantage in terms of ability to achieve deep DOF. At best they are "not worse", but as soon soon as there is more light or a tripod, the DSLR will get the advantage by getting the same DOF but much higher dynamic range and better general image quality.

Never might be overstating the case, but so long as DSLR image quality is equal or at the calculated ISO, the DSLR would be no worse. Not every DSLR has the low light image quality of a K-5 or D7000.
--
http://fruminousbandersnatch.blogspot.com/

 

[MichaelAM]

you seem to miss out the point that the DSLR has larger pixels - they suck in more light at the same aperture compared to a compact.

besides, we're not really talking about DOF here - at anything above f/3 on a DSLR most people would have a hard time recognizing the diffirence.

 

[godfrog]

The crop factor between 1/1.7" sensors and APS-C is 3, so you'd get the same field of view with a 18mm lens, and you'd get the same depth of field with the lens at F6. This is about 3 stops darker, so you'd need to use ISO1600.

Aha! You're right - apply the crop factor and you get to f/6 for the same DOF.

The image quality of a K-5 or D7000 at ISO1600 easily equal the S95 at ISO200.

Compact cameras never have an advantage in terms of ability to achieve deep DOF. At best they are "not worse", but as soon soon as there is more light or a tripod, the DSLR will get the advantage by getting the same DOF but much higher dynamic range and better general image quality.

Never might be overstating the case, but so long as DSLR image quality is equal or at the calculated ISO, the DSLR would be no worse. Not every DSLR has the low light image quality of a K-5 or D7000.
--
http://fruminousbandersnatch.blogspot.com/

Yes you are right, an older DSLR compared to the S95 would have trouble keeping up for the same field of view, DOF and shutter speed.

Then again, not every compact has the performance of a S95

--
My Flickr:
http://www.flickr.com/photos/36164047@N06/

 

[PeterLeyssens]

Hi Richard,

Would I have been able to take this picture with the same DOF and better image quality with a DSLR?

Sure.

I shot this at 28mm, 1/4 sec, f/2 and ISO 200 with a compact.

Your compact is a very good one with quite a large sensor for a compact. The S95 has a sensor that has a diagonal that is 4.5x smaller than a 35mm film negative. This translates to about 4.5 stops difference with a full frame camera, 3.5 stops with an APS sensor camera and 2.5 stops with a (Micro) FourThirds camera.

To get the same DOF with a DSLR, I'd likely have to stop down a lot.

Yes, 4.5/3.5/2.5 stops relatively. So that means you would get around f9 (full frame), f6.5 (APS) or f5 (FourThirds).

I'd need an ISO of 3200, which would likely be much noisier

To be precise, around 4000ISO (full frame), 2000 ISO (APS) or 1000 ISO (FourThirds). Note that each of the pixels of these SLRs is larger and gathers more light. Some claim that this corresponds exactly to the increase in ISO, i.e. the ISOs I mention for the respective cameras and your ISO200 would all give equivalent noise. There's been endless discussions on the validity of this statement, but I usually don't care because I get usable results up to at least 800ISO on my Micro FourThirds camera (in colour: in B&W, it's good up to at least 3200), which is more than enough for my needs.

Alternatively, I'd need to slow the shutter to 4 secs, which would require much steadier support than I had (or some combination of higher ISO and slower shutter speed).

I've made similar church photos at 1 or 2 seconds by putting my camera on my wallet, with a few credit cards shifted around until the horizon was fairly horizontal. This was with a film SLR using colour slides and the result was stunning. Don't go looking around for high tech solutions when you can get crazy long exposures using what you have with you.

Am I overstating the issues with a DSLR?

Sure

There are issues with DSLRs, to be sure. Don't try to carry one in a pocket like you would with the S95. But the worries you have are fictions of the mind. As a final thought: is there a reason why you ask this ? Don't you like the S95 or are you worried about the existence of better cameras ?

Peter.

--
gallery at http://picasaweb.google.com/peterleyssens
NAP (Nearly a PAD (Photo a Day)) at http://nap.techwriter.be

 

[sherwoodpete]

For situations like this, a pocket tripod or clamp is extremely useful.

Regardless of the camera, be it film, DSLR or digital compact, I would always consider the first priority is placing the camera on a solid support.

Even if you forget to bring any accessories, there are usually shelves, furniture, books etc. which can be used. Always treating the environment and the people with respect, I've done this many times.

As for the comparison of a compact with a DSLR at the same shutter speed, depth of field etc, that is a pretty clear example of the much discussed "equivalence".

Regards,
Peter

 

[richarddd]

To be precise, around 4000ISO (full frame), 2000 ISO (APS) or 1000 ISO (FourThirds).

For even more precision, see the dialog with godfrog, above.

I've made similar church photos at 1 or 2 seconds by putting my camera on my wallet, with a few credit cards shifted around until the horizon was fairly horizontal.

I used a similar technique with this shot. I've run into issues in other churches.

A tripod would be ideal, but most major churches, cathedrals, museums, castles, etc. don't allow tripods. I've been considering finding a way to disguise a tripod, such as putting it in a backpack, so that it would look like I'm just resting the camera on a backpack. I sometimes use a minipod, but I've been hassled about that.

A 2 second self timer is usually helpful in these situations.

As a final thought: is there a reason why you ask this ? Don't you like the S95 or are you worried about the existence of better cameras ?

I'm very fond of the S95. It produces excellent images at low ISOs and has great controls (better than most low or even middle level DSLRs I've seen).

I'd like to go much wider than the 28mm (35mm equivalent) limit on the S95, which would seem to require a camera with interchangeable lenses. I can do panoramas, but that's not really the same.

--
http://fruminousbandersnatch.blogspot.com/

 

[richarddd]

Even if you forget to bring any accessories, there are usually shelves, furniture, books etc. which can be used. Always treating the environment and the people with respect, I've done this many times.

A few years ago I saw a German tourist move a multi-language "do not touch" sign off an antique table to rest a camera on the table. It offends me every time I think of it.
--
http://fruminousbandersnatch.blogspot.com/

 

[sherwoodpete]

Even if you forget to bring any accessories, there are usually shelves, furniture, books etc. which can be used. Always treating the environment and the people with respect, I've done this many times.

A few years ago I saw a German tourist move a multi-language "do not touch" sign off an antique table to rest a camera on the table. It offends me every time I think of it.
--
http://fruminousbandersnatch.blogspot.com/

Understood.

I did say "with respect", which your example very clearly wasn't.

Regards,
Peter

 

[richarddd]

I wasn't disagreeing with you. I'm a big fan of bracing when it can be done with respect, as you say.

Sorry for the confusion.
--
http://fruminousbandersnatch.blogspot.com/

 

[PeterLeyssens]

Hi Richard,

I've made similar church photos at 1 or 2 seconds by putting my camera on my wallet, with a few credit cards shifted around until the horizon was fairly horizontal.

I used a similar technique with this shot. I've run into issues in other churches.

A tripod would be ideal, but most major churches, cathedrals, museums, castles, etc. don't allow tripods.

I've only once used a tripod inside a building and I thought it was extremely cumbersome. Mostly, it's possible to make photos by stabilising cameras on the ground, against pillars, etc. Sometimes, exposures of seconds may not be possible. That's all part of the fun, though: I don't do buildings for money, so I don't require studio precision.

I'm very fond of the S95. It produces excellent images at low ISOs and has great controls (better than most low or even middle level DSLRs I've seen).

I'd like to go much wider than the 28mm (35mm equivalent) limit on the S95, which would seem to require a camera with interchangeable lenses.

I have the LX3 which goes wider but has less intuitive controls. Then again, I hardly change settings anyway because DOF is so deep I can use f2.0 all the time. The current batch of serious compacts is very enjoyable, indeed.

For my next photo project, I'll be using my E-P1, though. Looking forward to see how it will compare to the film cameras I used before.

Peter.

--
gallery at http://picasaweb.google.com/peterleyssens
NAP (Nearly a PAD (Photo a Day)) at http://nap.techwriter.be

 

[GeraldW]

Peter,

You're absolutely right. I just went over to the Imaging Resource Comparometer and compared the S90 at ISO 200 to the Canon XSi (450D) at ISO 1600. I picked the XSi as the comparison as it has nearly the same number of vertical rows of pixels as the S90, and is relatively close in time, so the image processing technology is similar, being both are Canons. Captured detail is very similar, as is noise appearance at 100%.

Even the 3.5 stop theoretical crop factor ratio is nearly met as the S90 is (according to DPReview) about 1/3 stop overrated. So an ISO 200 to 1600 comparison is closer to 3 1/3 stops rather than 3.
--
Jerry

 

[PaulRivers]

Hmm, I wonder if anyone has a more authoritative link to what the equivalent f-stops for depth of field would be between the s95 and a Canon crop dslr?

This thread seems to have taken on a certain "dslr is always better tone", and while I do not personally know if that is accurate or not, I've been under the distinct impression that f6.0 on a dslr would have trouble getting infinite depth of field, while the s95 has infinite depth of field even at f2.0 at wide angle when focused on anything more than 5 or 10 feet away...

I could definitely be wrong, just wondering if anyone might have a link to a more official source on the topic?

 

[sherwoodpete]

Hmm, I wonder if anyone has a more authoritative link to what the equivalent f-stops for depth of field would be between the s95 and a Canon crop dslr?

This thread seems to have taken on a certain "dslr is always better tone", and while I do not personally know if that is accurate or not, I've been under the distinct impression that f6.0 on a dslr would have trouble getting infinite depth of field, while the s95 has infinite depth of field even at f2.0 at wide angle when focused on anything more than 5 or 10 feet away...

I could definitely be wrong, just wondering if anyone might have a link to a more official source on the topic?

Try here. Remember to choose both camera type and focal length, when comparing different formats/cameras.
http://www.dofmaster.com/dofjs.html

Regards,
Peter

 

[godfrog]

Hmm, I wonder if anyone has a more authoritative link to what the equivalent f-stops for depth of field would be between the s95 and a Canon crop dslr?

This thread seems to have taken on a certain "dslr is always better tone", and while I do not personally know if that is accurate or not, I've been under the distinct impression that f6.0 on a dslr would have trouble getting infinite depth of field, while the s95 has infinite depth of field even at f2.0 at wide angle when focused on anything more than 5 or 10 feet away...

I could definitely be wrong, just wondering if anyone might have a link to a more official source on the topic?

This site explains the concept in detail:
http://www.josephjamesphotography.com/equivalence/

The quick way to think about it is to establish the "crop factor" between the two formats and multiply the aperture by this value. A 1/1.7" has a 9.5mm diagonal, and APS-C around 28.8mm (at 24x16mm sensor, actual sizes are slightly below this, Canon smaller than Sony/Nikon/Pentax), so the crop factor is 3.03, close enough to 3. Equivalent field of view and depth of field is achieved on the APS-C camera by a lens with 3 times the focal length, set at F-stop 3 times what you set on the 1/1.7" camera.

As a side note, the aperture (in millimeters) will be the same with a 6mm lens set at F2 and a 18mm lens set at F6, around 3mm.

--
My Flickr:
http://www.flickr.com/photos/36164047@N06/

 

[godfrog]

Hmm, I wonder if anyone has a more authoritative link to what the equivalent f-stops for depth of field would be between the s95 and a Canon crop dslr?

This thread seems to have taken on a certain "dslr is always better tone", and while I do not personally know if that is accurate or not, I've been under the distinct impression that f6.0 on a dslr would have trouble getting infinite depth of field, while the s95 has infinite depth of field even at f2.0 at wide angle when focused on anything more than 5 or 10 feet away...

I could definitely be wrong, just wondering if anyone might have a link to a more official source on the topic?

Try here. Remember to choose both camera type and focal length, when comparing different formats/cameras.
http://www.dofmaster.com/dofjs.html

That calculator is fairly good for comparing how to set different cameras for the same final output, since it uses the same sharpness criterion for all formats, a CoC of about 1/1500th the sensor diagonal.

If you focus using the hyperfocal distance found with such a CoC though, you will notice that infinity is far from "sharp" at the pixel level of a modern DSLR, with detail smeared over several pixels. Using 1/1500th of the sensor diagonal CoC as a sharpness criterion for a medium or large format digital/film camera would be nuts, setting the level of "sharpness" lower than what a cellphone camera is able to achieve.

--
My Flickr:
http://www.flickr.com/photos/36164047@N06/

 

[alanr0]

Hmm, I wonder if anyone has a more authoritative link to what the equivalent f-stops for depth of field would be between the s95 and a Canon crop dslr?

This thread seems to have taken on a certain "dslr is always better tone", and while I do not personally know if that is accurate or not, I've been under the distinct impression that f6.0 on a dslr would have trouble getting infinite depth of field, while the s95 has infinite depth of field even at f2.0 at wide angle when focused on anything more than 5 or 10 feet away...

I could definitely be wrong, just wondering if anyone might have a link to a more official source on the topic?

This site explains the concept in detail:
http://www.josephjamesphotography.com/equivalence/

The quick way to think about it is to establish the "crop factor" between the two formats and multiply the aperture by this value. A 1/1.7" has a 9.5mm diagonal, and APS-C around 28.8mm (at 24x16mm sensor, actual sizes are slightly below this, Canon smaller than Sony/Nikon/Pentax), so the crop factor is 3.03, close enough to 3. Equivalent field of view and depth of field is achieved on the APS-C camera by a lens with 3 times the focal length, set at F-stop 3 times what you set on the 1/1.7" camera.

As a side note, the aperture (in millimeters) will be the same with a 6mm lens set at F2 and a 18mm lens set at F6, around 3mm.

Godfrog's side note is fundamental to understanding the physics here.

For the same field of view and same depth of field, the physical aperture (in mm) is the same. So if you keep the exposure time constant, the same number of photons reach the sensor.

If both DSLR and compact have the same number of pixels, and microlenses which give close to 100% fill factor, then on average, the same number of photons reaching each corresponding pixel, and photon noise is identical. With an ideal sensor and associated electronics which introduce no additional noise, the images are exactly equivalent with the same signal to noise ratio.

In the real world the larger pixels of the DSLR can be a disadvantage under these circumstances. The larger capacitance of the DSLR leads to higher read-out noise, and potentially poorer performance when the photon count (signal) is identical to that from a smaller sensor. For longer exposure times, sensor leakage current might be a concern too.

In practice, the read noise of recent DSLR sensors at high ISO is pretty low (2-3 electrons RMS), so there may not be much difference - and it could be eliminated if the DSLR has higher quantum efficiency. Read noise for a 2004 vintage compact (Canon S70) is quoted as 2.3 electrons by Roger Clarke, though I would expect the performance of current compacts to be better. http://www.clarkvision.com/articles/digital.sensor.performance.summary/index.html#References

If you can afford to keep the shutter open for longer, then the DSLR will collect a lot more light before saturation kicks in, and image quality will be substantially better.

Cheers
--
Alan Robinson

 

[richarddd]

Hmm, I wonder if anyone has a more authoritative link to what the equivalent f-stops for depth of field would be between the s95 and a Canon crop dslr?

This thread seems to have taken on a certain "dslr is always better tone", and while I do not personally know if that is accurate or not, I've been under the distinct impression that f6.0 on a dslr would have trouble getting infinite depth of field, while the s95 has infinite depth of field even at f2.0 at wide angle when focused on anything more than 5 or 10 feet away...

I could definitely be wrong, just wondering if anyone might have a link to a more official source on the topic?

This site explains the concept in detail:
http://www.josephjamesphotography.com/equivalence/

The quick way to think about it is to establish the "crop factor" between the two formats and multiply the aperture by this value. A 1/1.7" has a 9.5mm diagonal, and APS-C around 28.8mm (at 24x16mm sensor, actual sizes are slightly below this, Canon smaller than Sony/Nikon/Pentax), so the crop factor is 3.03, close enough to 3. Equivalent field of view and depth of field is achieved on the APS-C camera by a lens with 3 times the focal length, set at F-stop 3 times what you set on the 1/1.7" camera.

As a side note, the aperture (in millimeters) will be the same with a 6mm lens set at F2 and a 18mm lens set at F6, around 3mm.

Great article on equivalence. Well worth reading.
--
http://fruminousbandersnatch.blogspot.com/