Pages: 1 [2] 3   Go Down

Author Topic: NEX7 Rolling Review -- conclusion  (Read 20367 times)

Pelao

  • Full Member
  • ***
  • Offline Offline
  • Posts: 198
Re: NEX7 Rolling Review -- conclusion
« Reply #20 on: December 18, 2011, 12:07:57 pm »

As usual Michael's review was, for me, incredibly useful. As befits a landmark camera it was thorough, and especially so in the areas that matter to me: handling & ergonomics, and output to print.

I have been using a M4/3 setup for 2 years now, along with my usual Canon pro gear. I have come to really enjoy having a responsive camera and 2 or 3 primes with me all the time, and with pretty decent output. In fact, in some circumstances the output has been excellent.

Unlike some, I don't find the Nex setup overly large with the native lenses. Certainly larger than the M4/3, and indeed I would like the lenses to be smaller. But when I step into the realm of actual handling, a Nex with the 24 or 50 is, overall, much lighter and less bulky than my 'pro' gear: it will easily be comfortable in the same bag I use for my M4/3 gear.

The Nex 7 is looking good as a candidate to replace both sets of current equipment. I hope they add a native 35 (50 equivalent), but that can be resolved using a MF lens.

Anyway, thanks Michael for concentrating on the stuff I needed to know.

« Last Edit: December 18, 2011, 12:39:51 pm by Pelao »
Logged

AlanG

  • Full Member
  • ***
  • Offline Offline
  • Posts: 195
    • http://www.goldsteinphoto.com
Re: NEX7 Rolling Review -- conclusion
« Reply #21 on: December 18, 2011, 12:22:17 pm »

Hi,

Do you actually know if the Sony NEX-7 has a AA-filter? Has that been stated clearly anywhere?

The IR filter may be a source of problems to, it's normally quite thick and it would affect some Leica lenses. That was the reason that the M8 had very thin IR-filter. For the M9 the engineers came up with a thicker filter, but probably thinner then what is mostly used with other cameras.

Sony has the luxury to be able to construct it's lenses for the stuff sitting in front of the sensor. (CGA, microlenses, AA-filter and IR filter).

Best regards
Erik


It seems to me that if the Nex 7 had no AA filter, that would be stated by Sony somewhere and  be reported in the reviews.

I think future Leica w/a lenses are likely to be more "retro focus" in order to have fewer problems with color shift and vignetting in the corners unless sensors, microlenses or software correction improve.

Since the Nex 7 uses a smaller sensor, the issue is mitigated because the image is cropped before one gets to the parts that would be that big a problem. Also, the M9 has built in color shift and vignetting correction that is optimized for various lenses.

This is the explanation that Leica's CEO, Stefan Daniels, gives about IR filters.

http://www.megapixel.co.il/english/archive/23834

Q: If you look at the DXOMARK results of the M9 you can see that the ISO marks are considerably lower than other full frame cameras. Do you see this as one of the major things you would like to improve in the next M version?

A: The reason why this is lower lies in the special circumstances of the M-System. It is quite complicated, but I will try to make it short: M-lenses hit the sensor in a flat angle, therefore they are the most compact on the market and older lenses are compatible with M8 and M9. This flat angle will not allow us to use interference IR cut coating on the sensor cover glass, as this would cause uncontrollable color fringes. Instead of, we use an absorption filter, which is not sensible to different light angles, but does filter much more of the visible light as interference filters, and therefore the signal from the sensor needs to be amplified much more which results in a lower light sensitivity. Of course, having a higher sensitivity of the sensor is an issue and we put a significant amount of effort in improving this.
Logged
Alan Goldstein
[url=http://www.Goldstein

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: NEX7 Rolling Review -- conclusion
« Reply #22 on: December 19, 2011, 12:38:10 am »

Hi,

The reason I asked about the AA-filter is twofold. Firstly, my Sony Alpha 55 with 16 MP seems to have a very weak AA filter as it has a tendency to moiré. Secondly it makes a lot of sense to drop the AA-filter with shrinking pixel size. It is not exactly cheap, as far as I understand.

Thank you very much for the link to the interview with Stefan Daniels. Very interesting.

Best regards
Erik



It seems to me that if the Nex 7 had no AA filter, that would be stated by Sony somewhere and  be reported in the reviews.

This is the explanation that Leica's CEO, Stefan Daniels, gives about IR filters.

http://www.megapixel.co.il/english/archive/23834


Logged
Erik Kaffehr
 

douglasf13

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 547
Re: NEX7 Rolling Review -- conclusion
« Reply #23 on: December 19, 2011, 12:53:09 am »

Hi Douglas, that's quite re-assuring as my aim is to use the NEX 7 for 3 months or so, mainly with the kit lens before making any decision as to either further lenses (especially manual focus types) or selling my Nikon gear.  Your earlier post showed a lovely leather case for your NEX 5 - would you mind posting a link to the manufacturers so that I can see if anything is available for the 7.

Thanks

John

Don't get me wrong, it's still a cheap kit zoom, but it has very good central sharpness.  My NEX case is made by Ciesta, but I don't know if they have NEX-7 cases.
Logged

douglasf13

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 547
Re: NEX7 Rolling Review -- conclusion
« Reply #24 on: December 19, 2011, 12:55:19 am »

Hi,

Do you actually know if the Sony NEX-7 has a AA-filter? Has that been stated clearly anywhere?

The IR filter may be a source of problems to, it's normally quite thick and it would affect some Leica lenses. That was the reason that the M8 had very thin IR-filter. For the M9 the engineers came up with a thicker filter, but probably thinner then what is mostly used with other cameras.

Sony has the luxury to be able to construct it's lenses for the stuff sitting in front of the sensor. (CGA, microlenses, AA-filter and IR filter).

Best regards
Erik



Hi, Erik.  Joakim "theSuede," who works in the industry, has some pretty specific things to say about the M9's filters. The M9's filters and micro lenses improvements appear to be mostly marketing.  Check it out:
http://www.fredmiranda.com/forum/topic/893332/8&year=2010#8504636
« Last Edit: December 19, 2011, 01:01:51 am by douglasf13 »
Logged

bjanes

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3387
Re: NEX7 Rolling Review -- conclusion
« Reply #25 on: December 19, 2011, 08:59:50 am »

Do you actually know if the Sony NEX-7 has a AA-filter? Has that been stated clearly anywhere?

The IR filter may be a source of problems to, it's normally quite thick and it would affect some Leica lenses. That was the reason that the M8 had very thin IR-filter. For the M9 the engineers came up with a thicker filter, but probably thinner then what is mostly used with other cameras.

If the Nex-7 has a low pass filter, it must be relatively weak as judged by the aliasing pattern noted on a Siemens star on an image published in DPReview. The pattern is very similar to that of the Leica M9, which definitely lacks a low pass filter. The shown images are at 100% with no re-sizing.

One can also calculate the resolution using the Siemens star, according to the method outlined by Bart van der Wolf. The star in the DPReview image has 60 bars per circumference, whereas Bart's chart has 144 bars per circumference, so one has to modify the formula to (60 / pi) / diameter from (144 / pi) / diameter. This method is independent of the image magnification. Bart's chart is sinusoidal, which is better than the sharp bars of the star used by DPReview, but I don't think that this would materially affect the results.

Using this method, I get 58 cy/mm or 0.39 cy/pixel for the Leica and 106 cy/mm or 0.42 cy/mm for the Sony. The picture height of the Sony is 15.6 mm, so the Sony would resolve 1654 cy/ph. Using the same crop for the Leica, one gets 905 cy/ph. If one uses the full frame of the Leica, the resolution is 1392 cy/ph. It does not make sense to crop the Leica if one wants to maintain the image quality for which one is paying.

Regards,

Bill
« Last Edit: December 19, 2011, 10:38:42 am by bjanes »
Logged

AlanG

  • Full Member
  • ***
  • Offline Offline
  • Posts: 195
    • http://www.goldsteinphoto.com
Re: NEX7 Rolling Review -- conclusion
« Reply #26 on: December 19, 2011, 01:03:01 pm »

It does not make sense to crop the Leica if one wants to maintain the image quality for which one is paying.

Regards,

Bill

Yes. But it also shows that the APS format, with a good lens, can deliver the quality that many people require and are used to getting from a Leica M9. It would be interesting to see a 100% comparison using a Summilux 75 on the M9 and a Summilux 50 on a Nex 7 of the same composition. Maybe if those images are close one could conclude that using 50 and longer Leica lenses on a Nex 7 for more reach makes a lot of sense.

I use Canon full frame cameras because my work requires high performance wide angles (17 & 24 TSE lenses especially) and this seems to be the best solution short of using an MF technical camera. But it looks like these higher res APS cameras will suit the needs of many photographers.  When you look at it, there really only are a few models of full frame cameras, so I guess the market spoke long ago. But the gap seems to be narrowing further, especially if you figure that most people won't really need a 50MP full frame model even if the lenses can make use of that resolution.  But perhaps such a camera will pull some users from considering MF.
Logged
Alan Goldstein
[url=http://www.Goldstein

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: NEX7 Rolling Review -- conclusion
« Reply #27 on: December 19, 2011, 04:28:49 pm »

Hi!


18*(1654/1392)^2=25 ;-)

Best regards
Erik



If the Nex-7 has a low pass filter, it must be relatively weak as judged by the aliasing pattern noted on a Siemens star on an image published in DPReview. The pattern is very similar to that of the Leica M9, which definitely lacks a low pass filter. The shown images are at 100% with no re-sizing.

One can also calculate the resolution using the Siemens star, according to the method outlined by Bart van der Wolf. The star in the DPReview image has 60 bars per circumference, whereas Bart's chart has 144 bars per circumference, so one has to modify the formula to (60 / pi) / diameter from (144 / pi) / diameter. This method is independent of the image magnification. Bart's chart is sinusoidal, which is better than the sharp bars of the star used by DPReview, but I don't think that this would materially affect the results.

Using this method, I get 58 cy/mm or 0.39 cy/pixel for the Leica and 106 cy/mm or 0.42 cy/mm for the Sony. The picture height of the Sony is 15.6 mm, so the Sony would resolve 1654 cy/ph. Using the same crop for the Leica, one gets 905 cy/ph. If one uses the full frame of the Leica, the resolution is 1392 cy/ph. It does not make sense to crop the Leica if one wants to maintain the image quality for which one is paying.

Regards,

Bill
Logged
Erik Kaffehr
 

kers

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4388
    • Pieter Kers
Re: NEX7 Rolling Review -- conclusion
« Reply #28 on: December 19, 2011, 04:40:29 pm »

Maybe I have missed it,
but is the Nex7 (Nex 5n) camera very silent?
I am using DSLR's and just cannot take photographs at violinconcerts wich i would like to do..

I know that the new Nikon J1/V1 is soundless, but their quality is below my standards...

anyone?
Logged
Pieter Kers
www.beeld.nu/la

bjanes

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3387
Re: NEX7 Rolling Review -- conclusion
« Reply #29 on: December 19, 2011, 05:22:10 pm »

18*(1654/1392)^2=25 ;-)

Yes, the two cameras resolve pretty much as their pixel count would suggest, not much more nor much less. The lens used for the Nex-7 was the US $125 DT 50 mm f/1.8 and the lens used on the Leica was the US $4995 Summarit-M 75 mm f/2.5. Of course, the measurement was at low MTF, around 10%. The Leica lens might well do better at higher MTF because Leica lenses are usually regarded as favoring contrast over resolution (at least, according to commonly held belief).

Regards,

Bill
Logged

madmanchan

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2115
    • Web
Re: NEX7 Rolling Review -- conclusion
« Reply #30 on: December 19, 2011, 05:40:04 pm »

All the current NEX cameras have (relatively weak) optical low-pass filters, including the 7.
Logged
Eric Chan

michael

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 5084
Re: NEX7 Rolling Review -- conclusion
« Reply #31 on: December 19, 2011, 06:19:40 pm »

My most recent comparison (just published) shows that it must be very low indeed, because it's propensity for colour moire is almost exactly the same as the AAless M9.

Michael
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Re: NEX7 Rolling Review -- conclusion
« Reply #32 on: December 19, 2011, 06:41:11 pm »

But the gap seems to be narrowing further, especially if you figure that most people won't really need a 50MP full frame model even if the lenses can make use of that resolution. 


If the lenses can make use of a 50mp full frame, there will be no shortage of buyers if the camera is priced right. Cropped format cameras have always provided an advantage at the long telephoto end, and full frame cameras an advantage at the wide-angle end.

A full frame 50 or 60mp camera gives one the best of both worlds. The 24mp of the NEX-7 translates to 60mp full frame. Fantastic!

Quote
But perhaps such a camera will pull some users from considering MF.

No doubt, but the fact remains that the larger sensor will always provide that noticeably better SNR at 18% grey, which translates to creamier skin texture for fashion models.  ;D
Logged

douglasf13

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 547
Re: NEX7 Rolling Review -- conclusion
« Reply #33 on: December 19, 2011, 09:00:40 pm »


No doubt, but the fact remains that the larger sensor will always provide that noticeably better SNR at 18% grey, which translates to creamier skin texture for fashion models.  ;D

I would say that's only necessarily the case if the sensor technologies are the same.  A larger sensor can only leverage it's size so much if the technology of the smaller sensor is better.  That's partly why we're seeing the performance of these brand new aps-c sensors bump up against the FF cameras that have been on the market for a few years.  Of course, the new FF cameras coming soon will start the cycle all over again.  Medium format is an interesting beast, because the sensors are so much larger, despite being relatively primitive in design compared to the latest aps-c CMOS sensors.  Man, if Sony would scale up their newest EXMOR tech into 6x6 or 645, then we'd really be seeing something.
Logged

dreed

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1715
Re: NEX7 Rolling Review -- conclusion
« Reply #34 on: December 19, 2011, 11:36:22 pm »

On another web forum, dedicated to Canon cameras, this comment was made in the last couple of days in regards to the change that has come about since the release of the small form factor cameras with interchangeable lenses:

I work at a popular photography store and can tell you that canon is missing out. I have people coming in all day saying "olympus pen?" "Sony Nex?". They dont even know what they are but there is such a buzz surrounding those cameras. I inform people on those cameras and they are a pretty easy sell. To be honest we have been sold out of the olympus pens but could have sold tons if we had them (due to xmas aswell). I also show them the nikons and they say "where is the canon version of these?" The sony nex sells extremely well too.

... I think that speaks quite clearly about what the impact is of Canon's refusal thus far to eat its own lunch.
Logged

dreed

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1715
Re: NEX7 Rolling Review -- conclusion
« Reply #35 on: December 19, 2011, 11:42:45 pm »

You'll need to get an adaptor that has aperture control.

The best is the one from Novoflex...

http://www.ebay.com/itm/Novoflex-Adapter-Nikon-Lenses-Sony-NEX-Cameras-NEX-NIK-/390373197192?pt=LH_DefaultDomain_0&hash=item5ae40e4588#ht_3644wt_922

Michael

There doesn't appear to be an equivalent Novoflex for Canon EOS mount lenses.

Does anyone here have any experience (1st or 2nd hand) on how well the various offerings work and/or which is best?
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Re: NEX7 Rolling Review -- conclusion
« Reply #36 on: December 19, 2011, 11:47:43 pm »

I would say that's only necessarily the case if the sensor technologies are the same.  A larger sensor can only leverage it's size so much if the technology of the smaller sensor is better.  That's partly why we're seeing the performance of these brand new aps-c sensors bump up against the FF cameras that have been on the market for a few years.  Of course, the new FF cameras coming soon will start the cycle all over again.  Medium format is an interesting beast, because the sensors are so much larger, despite being relatively primitive in design compared to the latest aps-c CMOS sensors.  Man, if Sony would scale up their newest EXMOR tech into 6x6 or 645, then we'd really be seeing something.

Of course! It's understood that advanced technology applied to the smaller pixel may result in that smaller pixel having all the attributes of an 'obsolete' larger pixel. Why bring that into the equation. It applies across all fields.

The essential point I'm making can be clearly seen when comparing the Canon 20D with the Canon  5D2 on the DXOMark website. The pixels are about the same size and quality, yet when you toggle between screen and print, the larger 5D2 sensor always has the advantage at print size. At pixel size, in screen mode, both cameras are about equal, although the 20D does seem to have slightly better color sensitivity. On the other hand, the 5D2 has very slightly better DR in screen mode.

The point I'm making is, if Sony were to come out tomorrow with a FF sensor consisting of 60m NEX-7 pixels, the resulting image, downsampled to the 24mp NEX-7 size, would be better in all respects, assuming equality of lenses used.
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: NEX7 Rolling Review -- conclusion
« Reply #37 on: December 20, 2011, 12:58:02 am »

+1

Except downsampling adds artifacts. The truth is that we never use images without resampling, except when pixel peeping at actual pixels.

- When we print, we resample and apply dither.

- When we post we resample

- The human vision system obviously also resamples

Best regards
Erik

Of course! It's understood that advanced technology applied to the smaller pixel may result in that smaller pixel having all the attributes of an 'obsolete' larger pixel. Why bring that into the equation. It applies across all fields.

The essential point I'm making can be clearly seen when comparing the Canon 20D with the Canon  5D2 on the DXOMark website. The pixels are about the same size and quality, yet when you toggle between screen and print, the larger 5D2 sensor always has the advantage at print size. At pixel size, in screen mode, both cameras are about equal, although the 20D does seem to have slightly better color sensitivity. On the other hand, the 5D2 has very slightly better DR in screen mode.

The point I'm making is, if Sony were to come out tomorrow with a FF sensor consisting of 60m NEX-7 pixels, the resulting image, downsampled to the 24mp NEX-7 size, would be better in all respects, assuming equality of lenses used.
Logged
Erik Kaffehr
 

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: NEX7 Rolling Review -- conclusion
« Reply #38 on: December 20, 2011, 01:18:27 am »

Hi,

Two different factors. Shot noise is little dependent on sensor technology and is normally dominant. Bigger sensors captures more photons, less shot noise.

Readout noise affects the darkest parts. Here sensor technology plays a major role, so DR increases with improved technology.

Best regards
Erik


I would say that's only necessarily the case if the sensor technologies are the same.  A larger sensor can only leverage it's size so much if the technology of the smaller sensor is better.  That's partly why we're seeing the performance of these brand new aps-c sensors bump up against the FF cameras that have been on the market for a few years.  Of course, the new FF cameras coming soon will start the cycle all over again.  Medium format is an interesting beast, because the sensors are so much larger, despite being relatively primitive in design compared to the latest aps-c CMOS sensors.  Man, if Sony would scale up their newest EXMOR tech into 6x6 or 645, then we'd really be seeing something.
Logged
Erik Kaffehr
 

douglasf13

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 547
Re: NEX7 Rolling Review -- conclusion
« Reply #39 on: December 20, 2011, 01:22:36 am »

Ray, I was responding to your comment about medium format sensors, which will be quite a bit behind the next round of 135 sensors in tech, although I'm not saying one will be better than the other. As Erik mentions above, shot noise will be better in medium format, just maybe not DR.
« Last Edit: December 20, 2011, 01:25:06 am by douglasf13 »
Logged
Pages: 1 [2] 3   Go Up