Price/Performance Comparison Between E-Mount and Micro Four Thirds (with Chart)

[pwjazz]

I've owned and happily used an NEX-5N for several years now. I use only manual focus lenses like the Voigtlander Nokton Classic 35mm, and I've been happy with the shooting experience. I particularly appreciate the low-light performance of the camera with this lens, which allows me to get handheld shots like this one:





I do miss more shots than I would like when using manual focus, so due to having a child and getting lazier, I'm inclined to pick up some autofocus lenses. For reasons that I elaborate at the end of this post, I'm not satisfied with the NEX-5N when using autofocus lenses, so I will need to pick up a new body. Because my existing manual focus lenses can be adapted to any mirrorless system, this means that I have the luxury of considering other systems than Sony.

When I bought my NEX, I knew that the selection of E-Mount lenses was somewhat limited, but I didn't care since I was just adapting third-party lenses. A few years have passed, and I assumed that the situation had improved. I was right to a certain extent, but after researching and comparing the E-Mount ecosystem to Micro 4/3 in particular, I'm not satisfied.

The current E-Mount selection does include some well-performing lenses, but they seem generally expensive in comparison to Micro 4/3. To make sure that I'm not misunderstanding the situation, I collected some data on the top performing lenses from both systems per DXOMark and correlated that with price data from B&H Photo. The Micro 4/3 lenses were tested on an OM-D E-M10, and the E-Mount lenses on a Sony NEX-7.

Taking into account max aperture, we see the following:



You can find the data and chart in this Google Spreadsheet.

Looking at this chart, I notice a few things:

1. At the very cheap end, E-mount leads with the Sigma 60mm and Sony 50mm 1.8. Unfortunately these are very similar focal lengths, and both fairly specialized to portraiture. The Sigma 60 is also available on Micro 4/3, where it does pretty well too.

2. The Sigma lenses for both systems are a bright spot, and very affordable. They're not competitive on aperture, but being primes they're still well ahead of many of the zooms on light gathering.

3. Micro 4/3 leads at most other price points. In particular, it is strong in the midrange, with 4 highly rated options right around $500. This is basically the m.zuiko primes, which as a set cover a useful range of focal lengths. Sony's 35mm 1.8 does deserve an honorable mention for hanging in there too.

4. Micro 4/3 has the 2 best performing lenses, price no object.

5. The very expensive FE zoom lenses put on a very poor showing on the APS-C NEX-7 (they're the two red dots at the bottom right). They do get higher scores when tested on the A7R, but in comparison to the primes' performance on the A7R, they're still very bad. Yes, zoom lenses are worse performers than primes, but the Olympus 12-40 2.8 manages to significantly outperform the more expensive FE's, so that excuse only seems to go so far.

6. Conversely, the Zeiss FE prime lenses (2 red dots towards top right) are a bright spot performance-wise, though not price-wise.

Sony's lens roadmap has them focusing exclusively on FE lenses for full frame over the next year or two, which so far seem to be universally expensive and only occasionally good on APS-C. Given that I can put together put together a set of high quality Micro 4/3 primes at midrange prices from existing models, I'll very likely jump to Micro 4/3. It doesn't hurt that the Olympus lenses are currently on sale when purchased with a body, and it also doesn't hurt that at any given aperture, they have a deeper depth of field which makes me less likely to miss focus.

Some points for further discussion:

1. When looking at the dxomark results, well-performing lenses seem to get higher scores on higher resolution sensors (for example, compare the Sonar 55mm on NEX-7 versus A7R). This would imply that higher resolution can improve lens score, presumably because the sharpness score is not resolution-independent. Considering that the data shown here is comparing a 16MB OM-D E-M10 with a 24MP NEX-7, and the lenses on the Olympus are often winning, does that mean that the gap would widen even further if Micro 4/3 ever offers higher resolution sensors?

2. Conversely, there is the truism that higher resolution sensors are harder on lenses. So perhaps for the less stellar lenses, this actually puts the The NEX-7 at a disadvantage in this comparison? That said, it seems that Sony is trending towards higher res sensors. Even the lowly A5000 has a 20MP sensor. So, I think it's fair to judge lens' performance relative to the sensors in Sony's modern bodies, yes?

3. Can FE zoom lenses be made to perform well on APS-C Sonys?

4. Can FE lenses be made affordable? I'm no optical engineer, but it seems to me that Micro 4/3 has the luxury of targeting smaller sensors, which allows for smaller lenses, which I assume would be easier to design and produce with more consistent quality at a lower price than the larger lenses for APS-C and full frame. Can anyone who actually knows what she's talking about explain how sensor/lens size factors into quality and cost?

--- Why I need a new body to go with my autofocus lenses ---

I tried out the Sony 50mm 1.8 OSS, and found that the lens was really nice (sharp, well built, pleasant in hand), but I found that using auto-focus changed my experience of the camera considerably. When shooting at the pace of manual focus, things like having to hit a button before adjusting exposure compensation didn't bother me. However at auto-focus pace, I suddenly expect everything else to keep up. Also, I found that the autofocus wasn't so great in low light, and the poorly-positioned focus-assist light was often useless since the left side of the frame is shaded by the lens (not to mention that my index finger naturally rests right on top of it). Lastly, touch to focus is nice, but having to hit the shutter release afterwards is annoying.

These are all problems that are resolved in more modern cameras like the OM-D E-M10, so I think I just need to upgrade.

 

[audiobomber]

You have some severe flaws in your analysis.

The current E-Mount selection does include some well-performing lenses, but they seem generally expensive in comparison to Micro 4/3. To make sure that I'm not misunderstanding the situation, I collected some data on the top performing lenses from both systems per DXOMark and correlated that with price data from B&H Photo. The Micro 4/3 lenses were tested on an OM-D E-M10, and the E-Mount lenses on a Sony NEX-7.

Taking into account max aperture, we see the following:

But you're not taking equivalent aperture into account. F/2.8 on m4/3 is not equivalent to f/2.8 on APS-C. In DOF, f2.8 on m4/3 is equivalent to f3.7 on APS-C. Noise also needs equivalence to be considered. Sure, you will get a faster exposure with the m4/3 f2.8 lens than you will with f3.7, but that is negated by the fact that you can simply increase ISO to match. Have you missed DPR's important article on equivalence?

http://www.dpreview.com/articles/2666934640/what-is-equivalence-and-why-should-i-care

You can find the data and chart in this Google Spreadsheet.

Looking at this chart, I notice a few things:

1. At the very cheap end, E-mount leads with the Sigma 60mm and Sony 50mm 1.8. Unfortunately these are very similar focal lengths, and both fairly specialized to portraiture. The Sigma 60 is also available on Micro 4/3, where it does pretty well too.

Your comparison is based on Overall Score, which is a fudged number. They don't explain how they get this number, but it seems to be at least partly subjective, and based on format. Just because two lenses score 20 Overall, does not in any way make their real world results equal. Look instead at the actual measurements of the lens, mounted on a camera.

Panasonic Leica DG Nocticron 42.5mm F1.2 ASPH on Panasonic Lumix DMC-GX7 vs Sony E 50mm F1.8 OSS on Sony A6000 vs Olympus M.ZUIKO DIGITAL ED 60mm F2.8 Macro on Olympus OM-D E-M10 | DXOMARK

www.dxomark.com

In the above we see the SEL50 has more than double the resolution of the GX7/42.5mm, and 44% higher than EM10/60 macro. Transmission of the Sony lens is in between the other two, distortion is negligible, vignetting favours m4/3 and CA is wildly better with the Sony.

A good lens, mounted on FF a body with a larger sensor and higher resolution, will outdo anything you can get in APS-C, and the same applies to the best APS-C lenses compared with m4/3. Here's the ideal proof, the same lens mounted on the NEX7 vs. EM10. I predict right now that the a6000 when it's tested by DXOMark will outdo the NEX7 due to a weaker blur filter.

Sigma 60mm F2.8 DN A Sony E on Sony A6000 vs Sigma 60mm F2.8 DN A Micro 4/3 on Panasonic Lumix DMC-GH3 vs Sigma 60mm F2.8 DN A Micro 4/3 on Olympus OM-D E-M10 | DXOMARK

www.dxomark.com

Same lens, almost twice the resolution on 24mp APS-C vs. m4/3 (16mp vs. 9mp) and half as much CA.

4. Micro 4/3 has the 2 best performing lenses, price no object.

I hope I have disabused you of this notion. The SEL50 linked above outperforms any lens on a m4/3 body, and there are several lenses for Sony that outperform the SEL50.

5. The very expensive FE zoom lenses put on a very poor showing on the APS-C NEX-7 (they're the two red dots at the bottom right). They do get higher scores when tested on the A7R, but in comparison to the primes' performance on the A7R, they're still very bad. Yes, zoom lenses are worse performers than primes, but the Olympus 12-40 2.8 manages to significantly outperform the more expensive FE's, so that excuse only seems to go so far.

Look again. The Sony FE zooms reach 8mp vs. 7 for the 12-40. Transmission is surprisingly close. Distortion is wildly in favour of the FE lenses, vignetting is better, CA much better on the Zeiss.

Sony FE 28-70mm F3.5-5.6 OSS on Sony NEX-7 vs Olympus M.ZUIKO DIGITAL ED 12-40mm F2.8 PRO on Olympus OM-D E-M10 vs Sony FE Carl Zeiss Vario-Tessar T* 24-70mm F4 ZA OSS on Sony NEX-7 | DXOMARK

www.dxomark.com

Sony's lens roadmap has them focusing exclusively on FE lenses for full frame over the next year or two, which so far seem to be universally expensive and only occasionally good on APS-C. Given that I can put together put together a set of high quality Micro 4/3 primes at midrange prices from existing models, I'll very likely jump to Micro 4/3. It doesn't hurt that the Olympus lenses are currently on sale when purchased with a body, and it also doesn't hurt that at any given aperture, they have a deeper depth of field which makes me less likely to miss focus.

Lower cost for lower performance, IMO.

Some points for further discussion:

1. When looking at the dxomark results, well-performing lenses seem to get higher scores on higher resolution sensors (for example, compare the Sonar 55mm on NEX-7 versus A7R). This would imply that higher resolution can improve lens score, presumably because the sharpness score is not resolution-independent. Considering that the data shown here is comparing a 16MB OM-D E-M10 with a 24MP NEX-7, and the lenses on the Olympus are often winning, does that mean that the gap would widen even further if Micro 4/3 ever offers higher resolution sensors?

M4/3 sensors cannot match APS-C, and never will. Any advance introduced in one will migrate to the other. So far all of the innovation has been with APS-C first, then migrated to FF and M/4/3. That's because APS-C is the largest ILC market.

2. Conversely, there is the truism that higher resolution sensors are harder on lenses. So perhaps for the less stellar lenses, this actually puts the The NEX-7 at a disadvantage in this comparison? That said, it seems that Sony is trending towards higher res sensors. Even the lowly A5000 has a 20MP sensor. So, I think it's fair to judge lens' performance relative to the sensors in Sony's modern bodies, yes?

The NEX7 has a high resolution sensor, but the new 20 and 24mp sensors have improved corner resolution due to micro-lenses, and higher resolution due to the current trend to weaker blur filters.

3. Can FE zoom lenses be made to perform well on APS-C Sonys?

Pixel density will continue to increase, and larger sensors will continue to outperform smaller sensors.

4. Can FE lenses be made affordable? I'm no optical engineer, but it seems to me that Micro 4/3 has the luxury of targeting smaller sensors, which allows for smaller lenses, which I assume would be easier to design and produce with more consistent quality at a lower price than the larger lenses for APS-C and full frame. Can anyone who actually knows what she's talking about explain how sensor/lens size factors into quality and cost?

The first thing one learns is a marketing course is that price is not driven by cost. Price is set by the market.

--- Why I need a new body to go with my autofocus lenses ---

I tried out the Sony 50mm 1.8 OSS, and found that the lens was really nice (sharp, well built, pleasant in hand), but I found that using auto-focus changed my experience of the camera considerably. When shooting at the pace of manual focus, things like having to hit a button before adjusting exposure compensation didn't bother me. However at auto-focus pace, I suddenly expect everything else to keep up. Also, I found that the autofocus wasn't so great in low light, and the poorly-positioned focus-assist light was often useless since the left side of the frame is shaded by the lens (not to mention that my index finger naturally rests right on top of it). Lastly, touch to focus is nice, but having to hit the shutter release afterwards is annoying.

The a6000 has better controls than any NEX, and better auto-focus than any m4/3 body. That's the one you should be considering.

 

[Mel Snyder]

You have some severe flaws in your analysis.

The current E-Mount selection does include some well-performing lenses, but they seem generally expensive in comparison to Micro 4/3. To make sure that I'm not misunderstanding the situation, I collected some data on the top performing lenses from both systems per DXOMark and correlated that with price data from B&H Photo. The Micro 4/3 lenses were tested on an OM-D E-M10, and the E-Mount lenses on a Sony NEX-7.

Taking into account max aperture, we see the following:

But you're not taking equivalent aperture into account. F/2.8 on m4/3 is not equivalent to f/2.8 on APS-C. In DOF, f2.8 on m4/3 is equivalent to f3.7 on APS-C. Noise also needs equivalence to be considered. Sure, you will get a faster exposure with the m4/3 f2.8 lens than you will with f3.7, but that is negated by the fact that you can simply increase ISO to match. Have you missed DPR's important article on equivalence?

http://www.dpreview.com/articles/2666934640/what-is-equivalence-and-why-should-i-care

You can find the data and chart in this Google Spreadsheet.

Looking at this chart, I notice a few things:

1. At the very cheap end, E-mount leads with the Sigma 60mm and Sony 50mm 1.8. Unfortunately these are very similar focal lengths, and both fairly specialized to portraiture. The Sigma 60 is also available on Micro 4/3, where it does pretty well too.

Your comparison is based on Overall Score, which is a fudged number. They don't explain how they get this number, but it seems to be at least partly subjective, and based on format. Just because two lenses score 20 Overall, does not in any way make their real world results equal. Look instead at the actual measurements of the lens, mounted on a camera.

http://www.dxomark.com/Lenses/Compa...n-Olympus-OM-D-E-M10___1297_901_745_0_897_937

In the above we see the SEL50 has more than double the resolution of the GX7/42.5mm, and 44% higher than EM10/60 macro. Transmission of the Sony lens is in between the other two, distortion is negligible, vignetting favours m4/3 and CA is wildly better with the Sony.

A good lens, mounted on FF a body with a larger sensor and higher resolution, will outdo anything you can get in APS-C, and the same applies to the best APS-C lenses compared with m4/3. Here's the ideal proof, the same lens mounted on the NEX7 vs. EM10. I predict right now that the a6000 when it's tested by DXOMark will outdo the NEX7 due to a weaker blur filter.

http://www.dxomark.com/Lenses/Compa...Olympus-OM-D-E-M10___1104_0_1103_842_1103_937

Same lens, almost twice the resolution on 24mp APS-C vs. m4/3 (16mp vs. 9mp) and half as much CA.

4. Micro 4/3 has the 2 best performing lenses, price no object.

I hope I have disabused you of this notion. The SEL50 linked above outperforms any lens on a m4/3 body, and there are several lenses for Sony that outperform the SEL50.

5. The very expensive FE zoom lenses put on a very poor showing on the APS-C NEX-7 (they're the two red dots at the bottom right). They do get higher scores when tested on the A7R, but in comparison to the primes' performance on the A7R, they're still very bad. Yes, zoom lenses are worse performers than primes, but the Olympus 12-40 2.8 manages to significantly outperform the more expensive FE's, so that excuse only seems to go so far.

Look again. The Sony FE zooms reach 8mp vs. 7 for the 12-40. Transmission is surprisingly close. Distortion is wildly in favour of the FE lenses, vignetting is better, CA much better on the Zeiss.

http://www.dxomark.com/Lenses/Compa...SS-on-Sony-NEX-7___1244_736_1233_937_1253_736

Sony's lens roadmap has them focusing exclusively on FE lenses for full frame over the next year or two, which so far seem to be universally expensive and only occasionally good on APS-C. Given that I can put together put together a set of high quality Micro 4/3 primes at midrange prices from existing models, I'll very likely jump to Micro 4/3. It doesn't hurt that the Olympus lenses are currently on sale when purchased with a body, and it also doesn't hurt that at any given aperture, they have a deeper depth of field which makes me less likely to miss focus.

Lower cost for lower performance, IMO.

Some points for further discussion:

1. When looking at the dxomark results, well-performing lenses seem to get higher scores on higher resolution sensors (for example, compare the Sonar 55mm on NEX-7 versus A7R). This would imply that higher resolution can improve lens score, presumably because the sharpness score is not resolution-independent. Considering that the data shown here is comparing a 16MB OM-D E-M10 with a 24MP NEX-7, and the lenses on the Olympus are often winning, does that mean that the gap would widen even further if Micro 4/3 ever offers higher resolution sensors?

M4/3 sensors cannot match APS-C, and never will. Any advance introduced in one will migrate to the other. So far all of the innovation has been with APS-C first, then migrated to FF and M/4/3. That's because APS-C is the largest ILC market.

2. Conversely, there is the truism that higher resolution sensors are harder on lenses. So perhaps for the less stellar lenses, this actually puts the The NEX-7 at a disadvantage in this comparison? That said, it seems that Sony is trending towards higher res sensors. Even the lowly A5000 has a 20MP sensor. So, I think it's fair to judge lens' performance relative to the sensors in Sony's modern bodies, yes?

The NEX7 has a high resolution sensor, but the new 20 and 24mp sensors have improved corner resolution due to micro-lenses, and higher resolution due to the current trend to weaker blur filters.

3. Can FE zoom lenses be made to perform well on APS-C Sonys?

Pixel density will continue to increase, and larger sensors will continue to outperform smaller sensors.

4. Can FE lenses be made affordable? I'm no optical engineer, but it seems to me that Micro 4/3 has the luxury of targeting smaller sensors, which allows for smaller lenses, which I assume would be easier to design and produce with more consistent quality at a lower price than the larger lenses for APS-C and full frame. Can anyone who actually knows what she's talking about explain how sensor/lens size factors into quality and cost?

The first thing one learns is a marketing course is that price is not driven by cost. Price is set by the market.

--- Why I need a new body to go with my autofocus lenses ---

I tried out the Sony 50mm 1.8 OSS, and found that the lens was really nice (sharp, well built, pleasant in hand), but I found that using auto-focus changed my experience of the camera considerably. When shooting at the pace of manual focus, things like having to hit a button before adjusting exposure compensation didn't bother me. However at auto-focus pace, I suddenly expect everything else to keep up. Also, I found that the autofocus wasn't so great in low light, and the poorly-positioned focus-assist light was often useless since the left side of the frame is shaded by the lens (not to mention that my index finger naturally rests right on top of it). Lastly, touch to focus is nice, but having to hit the shutter release afterwards is annoying.

The a6000 has better controls than any NEX, and better auto-focus than any m4/3 body. That's the one you should be considering.

 

[quezra]

Three big flaws in your analysis: two already mentioned by audiobomber - lack of equivalence adjustment and blind usage of overall scores

A third one is that the NEX-7 is a problematic body because of Sony's micro-lens calibration being off. Pretty much all of the E-mount lenses tested on the lower resolution a5000 perform better than on the NEX-7. More Sony users need to write to DxOMark to push them to test lenses on the newer 24MP bodies of A6000 and A5100 because they will retain the obsolete and flawed marks on the NEX-7 forever otherwise. I can't believe DxOMark wasted time testing all the lenses on A5000 and a3000 (identical sensors) yet gone more than 6 months without testing the A6000 which is now the actual flagship camera for APS-C E-mount.

 

[audiobomber]

Three big flaws in your analysis: two already mentioned by audiobomber - lack of equivalence adjustment and blind usage of overall scores

A third one is that the NEX-7 is a problematic body because of Sony's micro-lens calibration being off.

I did mention a couple of times that I believe an a6000 will outresolve a NEX7.

Micro-lens improvements are highly significant, I agree, but I don't think they impact center resolution, just edges and corners. I believe the superior DXO resolution scores of the a5000 vs. NEX7 are the result of a weaker AA filter in the new bodies. Blur filters have a large effect on resolution.

I can't believe DxOMark wasted time testing all the lenses on A5000 and a3000 (identical sensors) yet gone more than 6 months without testing the A6000 which is now the actual flagship camera for APS-C E-mount.

Yes, hopefully they will test the a6000 or a5100, and also the 16-70mm f4. Those are serious and annoying gaps.

 

[AHM]

It's an interesting analysis. With respect to DxO mark score, you do need a single number, so that seems OK. But you definitely need to use equivalent aperture. Otherwise my iPhone lens outperforms most of these lenses at a much lower price.

 

[SmoothGlass]

You have some severe flaws in your analysis.

The current E-Mount selection does include some well-performing lenses, but they seem generally expensive in comparison to Micro 4/3. To make sure that I'm not misunderstanding the situation, I collected some data on the top performing lenses from both systems per DXOMark and correlated that with price data from B&H Photo. The Micro 4/3 lenses were tested on an OM-D E-M10, and the E-Mount lenses on a Sony NEX-7.

Taking into account max aperture, we see the following:

But you're not taking equivalent aperture into account. F/2.8 on m4/3 is not equivalent to f/2.8 on APS-C. In DOF, f2.8 on m4/3 is equivalent to f3.7 on APS-C. Noise also needs equivalence to be considered. Sure, you will get a faster exposure with the m4/3 f2.8 lens than you will with f3.7, but that is negated by the fact that you can simply increase ISO to match. Have you missed DPR's important article on equivalence?

http://www.dpreview.com/articles/2666934640/what-is-equivalence-and-why-should-i-care

You can find the data and chart in this Google Spreadsheet.

Looking at this chart, I notice a few things:

1. At the very cheap end, E-mount leads with the Sigma 60mm and Sony 50mm 1.8. Unfortunately these are very similar focal lengths, and both fairly specialized to portraiture. The Sigma 60 is also available on Micro 4/3, where it does pretty well too.

Your comparison is based on Overall Score, which is a fudged number. They don't explain how they get this number, but it seems to be at least partly subjective, and based on format. Just because two lenses score 20 Overall, does not in any way make their real world results equal. Look instead at the actual measurements of the lens, mounted on a camera.

http://www.dxomark.com/Lenses/Compa...n-Olympus-OM-D-E-M10___1297_901_745_0_897_937

In the above we see the SEL50 has more than double the resolution of the GX7/42.5mm, and 44% higher than EM10/60 macro. Transmission of the Sony lens is in between the other two, distortion is negligible, vignetting favours m4/3 and CA is wildly better with the Sony.

A good lens, mounted on FF a body with a larger sensor and higher resolution, will outdo anything you can get in APS-C, and the same applies to the best APS-C lenses compared with m4/3. Here's the ideal proof, the same lens mounted on the NEX7 vs. EM10. I predict right now that the a6000 when it's tested by DXOMark will outdo the NEX7 due to a weaker blur filter.

http://www.dxomark.com/Lenses/Compa...Olympus-OM-D-E-M10___1104_0_1103_842_1103_937

Same lens, almost twice the resolution on 24mp APS-C vs. m4/3 (16mp vs. 9mp) and half as much CA.

4. Micro 4/3 has the 2 best performing lenses, price no object.

I hope I have disabused you of this notion. The SEL50 linked above outperforms any lens on a m4/3 body, and there are several lenses for Sony that outperform the SEL50.

5. The very expensive FE zoom lenses put on a very poor showing on the APS-C NEX-7 (they're the two red dots at the bottom right). They do get higher scores when tested on the A7R, but in comparison to the primes' performance on the A7R, they're still very bad. Yes, zoom lenses are worse performers than primes, but the Olympus 12-40 2.8 manages to significantly outperform the more expensive FE's, so that excuse only seems to go so far.

Look again. The Sony FE zooms reach 8mp vs. 7 for the 12-40. Transmission is surprisingly close. Distortion is wildly in favour of the FE lenses, vignetting is better, CA much better on the Zeiss.

http://www.dxomark.com/Lenses/Compa...SS-on-Sony-NEX-7___1244_736_1233_937_1253_736

Sony's lens roadmap has them focusing exclusively on FE lenses for full frame over the next year or two, which so far seem to be universally expensive and only occasionally good on APS-C. Given that I can put together put together a set of high quality Micro 4/3 primes at midrange prices from existing models, I'll very likely jump to Micro 4/3. It doesn't hurt that the Olympus lenses are currently on sale when purchased with a body, and it also doesn't hurt that at any given aperture, they have a deeper depth of field which makes me less likely to miss focus.

Lower cost for lower performance, IMO.

Some points for further discussion:

1. When looking at the dxomark results, well-performing lenses seem to get higher scores on higher resolution sensors (for example, compare the Sonar 55mm on NEX-7 versus A7R). This would imply that higher resolution can improve lens score, presumably because the sharpness score is not resolution-independent. Considering that the data shown here is comparing a 16MB OM-D E-M10 with a 24MP NEX-7, and the lenses on the Olympus are often winning, does that mean that the gap would widen even further if Micro 4/3 ever offers higher resolution sensors?

M4/3 sensors cannot match APS-C, and never will. Any advance introduced in one will migrate to the other. So far all of the innovation has been with APS-C first, then migrated to FF and M/4/3. That's because APS-C is the largest ILC market.

2. Conversely, there is the truism that higher resolution sensors are harder on lenses. So perhaps for the less stellar lenses, this actually puts the The NEX-7 at a disadvantage in this comparison? That said, it seems that Sony is trending towards higher res sensors. Even the lowly A5000 has a 20MP sensor. So, I think it's fair to judge lens' performance relative to the sensors in Sony's modern bodies, yes?

The NEX7 has a high resolution sensor, but the new 20 and 24mp sensors have improved corner resolution due to micro-lenses, and higher resolution due to the current trend to weaker blur filters.

3. Can FE zoom lenses be made to perform well on APS-C Sonys?

Pixel density will continue to increase, and larger sensors will continue to outperform smaller sensors.

4. Can FE lenses be made affordable? I'm no optical engineer, but it seems to me that Micro 4/3 has the luxury of targeting smaller sensors, which allows for smaller lenses, which I assume would be easier to design and produce with more consistent quality at a lower price than the larger lenses for APS-C and full frame. Can anyone who actually knows what she's talking about explain how sensor/lens size factors into quality and cost?

The first thing one learns is a marketing course is that price is not driven by cost. Price is set by the market.

--- Why I need a new body to go with my autofocus lenses ---

I tried out the Sony 50mm 1.8 OSS, and found that the lens was really nice (sharp, well built, pleasant in hand), but I found that using auto-focus changed my experience of the camera considerably. When shooting at the pace of manual focus, things like having to hit a button before adjusting exposure compensation didn't bother me. However at auto-focus pace, I suddenly expect everything else to keep up. Also, I found that the autofocus wasn't so great in low light, and the poorly-positioned focus-assist light was often useless since the left side of the frame is shaded by the lens (not to mention that my index finger naturally rests right on top of it). Lastly, touch to focus is nice, but having to hit the shutter release afterwards is annoying.

The a6000 has better controls than any NEX, and better auto-focus than any m4/3 body. That's the one you should be considering.

--
Dan

I agree OP's post ridiculous, and I agree with audiobomber above.

DXOmark "overall" numbers are mostly useless, though their subcomponents such as field maps of sharpness, etc. are useful. But even there, where does bokeh, micro contrast, etc. factor in? Where does OSS ability factor it? Where does focus throw and smoothness factor in? They don't factor those in, which is a problem.

A 3:2 aspect ratio vs 4:3 is important to some. I prefer 3:2 and a 3:2 sensor also aligns better with HD video (16:9).

Furthermore you need to consider cross-mount lenses like Samyang 12/2, or Sigma 19/30/60.

I used to shoot MFT and left because the bang for buck simply was not there. I actually had to go back to DSLR for a while until the newer sensors came out, at which point I left DSLRs for good and went Canon and Sony mirrorless.

IMHO, for mirrorless:

Canon = best value in lenses & compatible flashes, and if you don't need speed can adapt EF lenses, also legacy lenses... biggest limitation right now is relatively slow AF speed and no bodies w/ EVF

Sony/Samsung = best value in bodies (sensor tech, focus peaking, zebras, decent EVFs), if you don't need speed can adapt EF lenses, also legacy lenses

MFT = mediocre value unless you are talking about discontinued stuff; their bodies get smaller 4:3 sensors, not larger 3:2 sensors, and their lenses are overpriced after accounting for focal/aperture equivalence... you are paying a premium for compactness/portability that is oftentimes not much better than APS-C mirrorless, since in both cases you probably need a camera bag anyway. MFT doesn't adapt wideangle EF or legacy lenses as well since the crop factor is huge.

Fuji X-Trans = Excellent gear but you definitely pay for it. Like MFT, it's also mediocre value in mirrorless after you factor in they cheat at ISO by 1/2-1 stop and X-Trans has noise reduction even in RAW. Their Bayer stuff is better value, though.

Pentax Q and Nikon 1, Samsung NX = worst value, just get a RX1/10/100, FZ1000, G7X, LX100, etc. instead. Only extreme combos like V3 + EVF + 70-300CX make any sense, and they cost a fortune.

Leica = I guess technically worst bang for buck, but the folks who can afford it don't care, so they are in a category of their own...

 

[audiobomber]

Dan: An excellent response. However,

Thanks Mel, however...

1. This poster started with an hypothesis and used the DxO tests - and his interpretation of them - to justify

But his interpretation was flawed. Should we allow him to decide on his next camera based on misinformation?

2. The purchase of an new M4/3 camera

I think anyone who owns an M4/3 camera and is pleased with the results should stick with it.

Everyone has different needs. Like most things, there is no "one true way" when choosing a camera. Note however that the OP has a NEX currently, not m4/3. He is considering a move to m4/3.

The fundamental point of the OP is that by going to M4/3, he gets access to more lenses.

I don't dispute that m4/3 has more lenses available, but I disagree that this was his fundamental point. He also stated that m4/3 lenses are higher performing than E-mount. I strongly disagree, because you must mount a lens on a camera, and a smaller sensor limits performance. He also said that comparable lenses are less expensive in m4/3. I don't know if they are or not, but it is important to consider lenses that truly are equivalent, and one can't do that without knowing about equivalence.

For some people, M4/3 meets their needs and wants. Be happy for them.

I am happy for them. What makes me unhappy is when they make uninformed comparisons. The worst one I've seen lately was comparing an Olympus 300mm f4 with a FF 600mm f4. This cannot be allowed to stand without challenge.

I did my best to help the pwjazz. He obviously put a lot of work into his graph, but I don't think it was helping him to make an informed choice. Changing systems is usually an expensive proposition. It would be a shame to change and possibly make a mistake due to incomplete or flawed information.

 

[Dennis]

It's a mildly interesting chart, but assumes that you can interpret DXOMark's overall scores as "performance" that's meaningful to you.

Further, as has already been pointed out, your bubble size is based on f-stop. And for all practical puposes, f/1.8 on m43 does not give you the same "performance" or benefit as f/1.8 on APS-C.

Also, the chart does not take into account features like IS that you pay for.

As for performance, the NEX-7 was unkind to lenses, at least WA lenses. The 50/1.8 gets a higher score on the 20MP A3000/5000 than it does on the NEX-7, and presumably would do better still on the A6000. I imagine that WAs look worse still on the NEX-7, which was notorious for soft corners on WAs due to the lack of microlenses on the sensor, IIRC.

You base the analysis on DXOMark's overall score, even though, for instance, a lens that scores 14 M-Pix on the 20MP A5000 scores 24 overall while the Panasonic that scores 28 overall registers only 13 M-Pix. So you really have to place a lot of importance on whatever it is DXOMark places importance on to spend your money based on their scores.

I find this kind of data fun to look at, but ultimately completely useless for any practical purpose.

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

 

[AHM]

You have some severe flaws in your analysis.

The current E-Mount selection does include some well-performing lenses, but they seem generally expensive in comparison to Micro 4/3. To make sure that I'm not misunderstanding the situation, I collected some data on the top performing lenses from both systems per DXOMark and correlated that with price data from B&H Photo. The Micro 4/3 lenses were tested on an OM-D E-M10, and the E-Mount lenses on a Sony NEX-7.

Taking into account max aperture, we see the following:

But you're not taking equivalent aperture into account. F/2.8 on m4/3 is not equivalent to f/2.8 on APS-C. In DOF, f2.8 on m4/3 is equivalent to f3.7 on APS-C. Noise also needs equivalence to be considered. Sure, you will get a faster exposure with the m4/3 f2.8 lens than you will with f3.7, but that is negated by the fact that you can simply increase ISO to match. Have you missed DPR's important article on equivalence?

http://www.dpreview.com/articles/2666934640/what-is-equivalence-and-why-should-i-care

You can find the data and chart in this Google Spreadsheet.

Looking at this chart, I notice a few things:

1. At the very cheap end, E-mount leads with the Sigma 60mm and Sony 50mm 1.8. Unfortunately these are very similar focal lengths, and both fairly specialized to portraiture. The Sigma 60 is also available on Micro 4/3, where it does pretty well too.

Your comparison is based on Overall Score, which is a fudged number. They don't explain how they get this number, but it seems to be at least partly subjective, and based on format. Just because two lenses score 20 Overall, does not in any way make their real world results equal. Look instead at the actual measurements of the lens, mounted on a camera.

http://www.dxomark.com/Lenses/Compa...n-Olympus-OM-D-E-M10___1297_901_745_0_897_937

In the above we see the SEL50 has more than double the resolution of the GX7/42.5mm, and 44% higher than EM10/60 macro. Transmission of the Sony lens is in between the other two, distortion is negligible, vignetting favours m4/3 and CA is wildly better with the Sony.

A good lens, mounted on FF a body with a larger sensor and higher resolution, will outdo anything you can get in APS-C, and the same applies to the best APS-C lenses compared with m4/3. Here's the ideal proof, the same lens mounted on the NEX7 vs. EM10. I predict right now that the a6000 when it's tested by DXOMark will outdo the NEX7 due to a weaker blur filter.

http://www.dxomark.com/Lenses/Compa...Olympus-OM-D-E-M10___1104_0_1103_842_1103_937

Same lens, almost twice the resolution on 24mp APS-C vs. m4/3 (16mp vs. 9mp) and half as much CA.

4. Micro 4/3 has the 2 best performing lenses, price no object.

I hope I have disabused you of this notion. The SEL50 linked above outperforms any lens on a m4/3 body, and there are several lenses for Sony that outperform the SEL50.

5. The very expensive FE zoom lenses put on a very poor showing on the APS-C NEX-7 (they're the two red dots at the bottom right). They do get higher scores when tested on the A7R, but in comparison to the primes' performance on the A7R, they're still very bad. Yes, zoom lenses are worse performers than primes, but the Olympus 12-40 2.8 manages to significantly outperform the more expensive FE's, so that excuse only seems to go so far.

Look again. The Sony FE zooms reach 8mp vs. 7 for the 12-40. Transmission is surprisingly close. Distortion is wildly in favour of the FE lenses, vignetting is better, CA much better on the Zeiss.

http://www.dxomark.com/Lenses/Compa...SS-on-Sony-NEX-7___1244_736_1233_937_1253_736

Sony's lens roadmap has them focusing exclusively on FE lenses for full frame over the next year or two, which so far seem to be universally expensive and only occasionally good on APS-C. Given that I can put together put together a set of high quality Micro 4/3 primes at midrange prices from existing models, I'll very likely jump to Micro 4/3. It doesn't hurt that the Olympus lenses are currently on sale when purchased with a body, and it also doesn't hurt that at any given aperture, they have a deeper depth of field which makes me less likely to miss focus.

Lower cost for lower performance, IMO.

Some points for further discussion:

1. When looking at the dxomark results, well-performing lenses seem to get higher scores on higher resolution sensors (for example, compare the Sonar 55mm on NEX-7 versus A7R). This would imply that higher resolution can improve lens score, presumably because the sharpness score is not resolution-independent. Considering that the data shown here is comparing a 16MB OM-D E-M10 with a 24MP NEX-7, and the lenses on the Olympus are often winning, does that mean that the gap would widen even further if Micro 4/3 ever offers higher resolution sensors?

M4/3 sensors cannot match APS-C, and never will. Any advance introduced in one will migrate to the other. So far all of the innovation has been with APS-C first, then migrated to FF and M/4/3. That's because APS-C is the largest ILC market.

2. Conversely, there is the truism that higher resolution sensors are harder on lenses. So perhaps for the less stellar lenses, this actually puts the The NEX-7 at a disadvantage in this comparison? That said, it seems that Sony is trending towards higher res sensors. Even the lowly A5000 has a 20MP sensor. So, I think it's fair to judge lens' performance relative to the sensors in Sony's modern bodies, yes?

The NEX7 has a high resolution sensor, but the new 20 and 24mp sensors have improved corner resolution due to micro-lenses, and higher resolution due to the current trend to weaker blur filters.

3. Can FE zoom lenses be made to perform well on APS-C Sonys?

Pixel density will continue to increase, and larger sensors will continue to outperform smaller sensors.

4. Can FE lenses be made affordable? I'm no optical engineer, but it seems to me that Micro 4/3 has the luxury of targeting smaller sensors, which allows for smaller lenses, which I assume would be easier to design and produce with more consistent quality at a lower price than the larger lenses for APS-C and full frame. Can anyone who actually knows what she's talking about explain how sensor/lens size factors into quality and cost?

The first thing one learns is a marketing course is that price is not driven by cost. Price is set by the market.

--- Why I need a new body to go with my autofocus lenses ---

I tried out the Sony 50mm 1.8 OSS, and found that the lens was really nice (sharp, well built, pleasant in hand), but I found that using auto-focus changed my experience of the camera considerably. When shooting at the pace of manual focus, things like having to hit a button before adjusting exposure compensation didn't bother me. However at auto-focus pace, I suddenly expect everything else to keep up. Also, I found that the autofocus wasn't so great in low light, and the poorly-positioned focus-assist light was often useless since the left side of the frame is shaded by the lens (not to mention that my index finger naturally rests right on top of it). Lastly, touch to focus is nice, but having to hit the shutter release afterwards is annoying.

The a6000 has better controls than any NEX, and better auto-focus than any m4/3 body. That's the one you should be considering.

 

[blue_skies]

You are correlating a single DxOmark interpretation as absolute gospel.

E.g. compare Leica 25/1.4 with Sony 35/1.8, the Leica lens ($500) on the EM-1 and the Sony lens ($450) on the A5000. Scores are almost identical: 11P for the Leica and 12P for the Sony. The overall score favors the Leica: 24 versus 22.

Now go and compare images obtained with both lens/camera setups. What you will find is that there is a specific range in which the m43 sensor performs well (low ISO), but if you broaden your application, you will invariably find that larger sensors provide a larger workable area (high ISO), which ultimately gives you a more usable setup.

Also, if you were to crop the APS-C to m43 format, you'd end up with similar pixel (resolution) count, but you would mitigate the edge/corner softness effect. In other words - you may expect higher scores from the center area in APS-C (which is another way to view m43).

By its own metric, smaller sensors produce larger DOF at same apertures, which can also skew the interpretation of the results - you may in fact prefer this, but it limits your application range. If you want nicer bokeh, you are pretty much looking at larger size sensors.

Within equivalence parameters, technology 'nodes' seem to be indicating which sensor has more noise tolerance, and smaller sensors have come to market faster than larger sensors, skewing some of findings to smaller sensors. This however ignores the fact that on a larger sensor you can always 'win' by trading DOF for less noise with a faster lens.

Or, to put it this way - if your data holds as presented, APS-C would be better than FF, m43 would be better than APS-C, and so on: in the end, we would all end up with a cell phone camera, right?

It does not quite work that way - each smaller format limits your flexibility and takes away some of the 'wow' factor (3D separation, bokeh, DR, lighting/shadow noise, low noise, flare tolerance, highlight/blow-outs, etc.). Pick the format that pleases you the most, and don't look back!

 

[val1]

Three big flaws in your analysis: two already mentioned by audiobomber - lack of equivalence adjustment and blind usage of overall scores

A third one is that the NEX-7 is a problematic body because of Sony's micro-lens calibration being off. Pretty much all of the E-mount lenses tested on the lower resolution a5000 perform better than on the NEX-7. More Sony users need to write to DxOMark to push them to test lenses on the newer 24MP bodies of A6000 and A5100 because they will retain the obsolete and flawed marks on the NEX-7 forever otherwise. I can't believe DxOMark wasted time testing all the lenses on A5000 and a3000 (identical sensors) yet gone more than 6 months without testing the A6000 which is now the actual flagship camera for APS-C E-mount.

AMEN ! Also, single point to point shooting olympus AF and sony are in the same zipcode. AFC the sony a6000/5100 is just head and shoulders beyond. The panasonic GH4 AFC is comparable to the sony but the body is almost 1000$ MORE than the sony. I used the NEX6 and looked at changing to micro 4/3s but then the A6000 came out.

My thinking was:

A6000 afc is much better than mirriorless OTHER than the GH4.

Olympus 3 axis is pretty comprable to sony OSS but the 5 axis is better. (This won't help you chase your kids.)

Video quality: Sony and Panasonic are both good. I think sony is better. Olympus video is bleh.

A6000 1 stop sensor advantage over the olympus. So when looking at oly you need 1 stop advatage to hit the same depth etc. Oly needs 1 stop advatage to hit same exposure. There is always hope for a faster lens but your sensor is replaced with the camera body. (And you can get back decent money selling your lenses but you get crap on sale of a body.)

A6000 24mpx compared to 16mpx. So for a comparable lens the sony will strongly out resolve the oly.

When you check out primes the two systems are on par imo. The olympus primes are better but the better sensor of the sony more than makes up for this difference. So your basically treading water. When it comes to zooms, oly has an advatage depending on your needs. If you want alot of reach oly has more better options. For sony native mount, you can only hit 300ish mm equivalent. 4/3 goes longer but dimmer. Oly has a killer f/2.8 zoom but in price and end of the day performance it is pretty much on par with the zeiss 16-70. Besides, you use a zoom lens for many reasons but not because it is the finest performer.

If your not sure, grab your 50mm 1.8 and take it to worst buy. Pop on your lens and have you kid run at you.

 

[cptobvious]

Having owned both systems (APS-C E-mount and Olympus Micro 4/3), I'd say price/performance in the end is somewhat of a wash. Sony has the bigger sensor and the more bang for the buck bodies, but Olympus has the awesome 5-axis stabilization (at least several stops better than the lens-based Optical SteadyShot) and generally better lenses, including more sub-$500 1.8 primes. In real-life usage, I didn't see a whole lot of difference in low-light performance between the two. APS-C Sony has cleaner noise at equivalent exposures, but I could shoot at 1/2 second handheld with my E-P5 and 12mm f/2 prime.

The biggest reasons I sold my MFT system was 1. I wanted more DOF control and 2. I didn't like the native 4:3 format. The biggest reasons I sold my APS-C Sony was 1. lacked the lenses I wanted and 2. wanted faster AF (this was with the NEX-6). MFT generally had the better control layout (front/rear dials, touchscreen control of AF), while Sony generally had the better ergonomics with the deeper grip. It would be nice if the two companies merged, took the best parts of each system and created the perfect mirrorless camera, but that's just wishful thinking.

 

[MarsObserver]

A very nice, in-depth, analytical analysis of the M4/3 and e-mount lens systems.

I'd like to offer a counter-point if I may.

My dissertation is as follows:

Zeiss

Thank you.



--
http://www.flickr.com/photos/mars_observer/

 

[audiobomber]

Your comparison is based on Overall Score, which is a fudged number. They don't explain how they get this number, but it seems to be at least partly subjective, and based on format. Just because two lenses score 20 Overall, does not in any way make their real world results equal. Look instead at the actual measurements of the lens, mounted on a camera.

 

[Dennis]

Taking DxO mark overall score is probably the most reasonable.

Would you buy a car based on a 1-100 rating ?

Otherwise, you have to give up ranking, which is a valid point of view.

Yup. Stick with the analysis and forget trying to come up with a universal set of priorities in order to rate them.

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

 

[Erik1111]

Having owned both systems (APS-C E-mount and Olympus Micro 4/3), I'd say price/performance in the end is somewhat of a wash. Sony has the bigger sensor and the more bang for the buck bodies, but Olympus has the awesome 5-axis stabilization (at least several stops better than the lens-based Optical SteadyShot) and generally better lenses, including more sub-$500 1.8 primes. In real-life usage, I didn't see a whole lot of difference in low-light performance between the two. APS-C Sony has cleaner noise at equivalent exposures, but I could shoot at 1/2 second handheld with my E-P5 and 12mm f/2 prime.

The biggest reasons I sold my MFT system was 1. I wanted more DOF control and 2. I didn't like the native 4:3 format. The biggest reasons I sold my APS-C Sony was 1. lacked the lenses I wanted and 2. wanted faster AF (this was with the NEX-6). MFT generally had the better control layout (front/rear dials, touchscreen control of AF), while Sony generally had the better ergonomics with the deeper grip. It would be nice if the two companies merged, took the best parts of each system and created the perfect mirrorless camera, but that's just wishful thinking.

Do you really think in body stabilization is several stops better than lens based? Just curious. I know in the old days the claim was lens based stabilization was better, but that may have changed. I do wish my a6000 had in body stabilization for my legacy long primes.

 

[LouMeluso]

I've seen so many examples of the "DxO Mark Fetish" on this forum that it's quite hilarious reading the pundits hurling scores like so many magic spells at each other. But this really elevates it to new level.

 

[tjuster1]

Dan: An excellent response. However,

1. This poster started with an hypothesis and used the DxO tests - and his interpretation of them - to justify

2. The purchase of an new M4/3 camera

I think anyone who owns an M4/3 camera and is pleased with the results should stick with it. They need not be influenced at all by the fact that the rest of the photographic world isn't going in their direction - that the world is going to better and better APS-C and full-frame - and larger - sensor cameras, priced at or below M4/3 bodies - even the A7. I'm a ham radio operator, I love an obsolete 1978 electronic keyer, and while the rest of ham radio has moved on, I bought 3 used ones like I enjoy on ebay to assure my long term supply. Some people love 1959 Chevy Impalas, despite the technical superiority of modern cars.

What makes little sense is when simple arthmetic clearly dominates an argument: Larger and larger sensors being developed in a highly competitive market of multiple vendors, versus the essentially static technology of a much smaller sensor marketed only by two financially strapped companies. Decades ago, photography went through similar large vs. small film debates: half-frame 35mm vs. full frame, 16mm vs. Minox, etc. Bigger "sensors" - whether film or electronic - ALWAYS win, except when technology enables the smaller sensor to challenge the bigger. That happened in film when improved emulsions let 35mm challenge 6x6cm. But no one - certainly not Olympus or Panasonic - is driving a similar challenge of APS-C or full-frame from the M4/3 side. If one believes the rumors, Olympus is designing full-frame camera.

The fundamental point of the OP is that by going to M4/3, he gets access to more lenses. If he goes Nikon or Canon, he gets VASTLY more choices. Sony, Sigma, Samyang and Zeiss - not to mention tens of thousands of legacy lenses - give all those wishing to use either Sony APS-C or FF cameras more lenses than any of us would want to own, much less carry.

For some people, M4/3 meets their needs and wants. Be happy for them.

Oh how condescending, but to be expected.

The OP owns a NEX camera; he's hardly an m43 fanboy.

He's discovered that the NEX body makes a great tool for adapting old lenses. Sony's FF cameras I'm sure are even better, but they're still too pricey for some people (e.g., me).

He's discovered that native E-mount lenses are generally larger, more expensive, and less good than m43 equivalents. As someone who's dabbled in both systems (and still owns one camera from each), I can attest to this. That you disagree is fine, but at least he tried to support his position with evidence. You've offered stupid arguments, like:

• The rest of the world is going in the APS-C direction and abandoning m43. Not if you look at the number of lenses introduced, that's for sure. Are you seriously arguing that Sony supports E-mount with the same vigor that Panasonic and Olympus support m43?
• Bigger sensors ALWAYS win. If you're talking IQ, of course. But larger sensors require a larger image circle and thus require larger lenses--and since the OP is obviously looking at small MILC cameras, this is not really a relevant argument. But you know that.
• He could get more lens choices with Nikon or Canon. Right, but obviously he's looking for a MILC system (small body), so that's a non-starter, as you know as well;
• Sony users have access to "tens of thousands of legacy lenses". Which of course any MILC camera does as well.

It's all about finding the sweet spot between size and IQ. I think there are three of them: 1" sensor cameras (RX100), m43, and FF. I think APS-C is too similar to m43, and m43 lenses are significantly smaller with very little sacrifice in sensor IQ. (The fact that m43 lenses tend to be better overall has nothing to do with size but probably a result of having two manufacturers competing for business.) I reached this conclusion after spending a lot of time with my NEX-5R and Sony/Sigma lenses (the ones I could afford) and comparing the results with output from my m43 camera, so it's not armchair analysis but real-world experience.

Have you ever used an m43 camera for any period of time? Have you ever even held one in your hands?

I'll agree with this sentiment you wrote: for some people NEX meets their needs and wants (like you), and I'm happy for them. But it's not a slam dunk, and--and I know you can't possibly swallow this without gagging--other camera systems DO have advantages over your fetish. Everyone should weigh the plusses and minuses and reach their own conclusion.

 

[cptobvious]

Having owned both systems (APS-C E-mount and Olympus Micro 4/3), I'd say price/performance in the end is somewhat of a wash. Sony has the bigger sensor and the more bang for the buck bodies, but Olympus has the awesome 5-axis stabilization (at least several stops better than the lens-based Optical SteadyShot) and generally better lenses, including more sub-$500 1.8 primes. In real-life usage, I didn't see a whole lot of difference in low-light performance between the two. APS-C Sony has cleaner noise at equivalent exposures, but I could shoot at 1/2 second handheld with my E-P5 and 12mm f/2 prime.

The biggest reasons I sold my MFT system was 1. I wanted more DOF control and 2. I didn't like the native 4:3 format. The biggest reasons I sold my APS-C Sony was 1. lacked the lenses I wanted and 2. wanted faster AF (this was with the NEX-6). MFT generally had the better control layout (front/rear dials, touchscreen control of AF), while Sony generally had the better ergonomics with the deeper grip. It would be nice if the two companies merged, took the best parts of each system and created the perfect mirrorless camera, but that's just wishful thinking.

Do you really think in body stabilization is several stops better than lens based? Just curious. I know in the old days the claim was lens based stabilization was better, but that may have changed. I do wish my a6000 had in body stabilization for my legacy long primes.

Generally I'd agree that lens-based is better, but the Olympus 5-axis IBIS system is pretty eye-opening. I'd recommend borrowing or renting an OM-D to try out.