a7S high ISO claim - - -over hyped or not ?

Started Jul 3, 2014 | Discussions
LSHorwitz1 Senior Member • Posts: 2,492
a7S high ISO claim - - -over hyped or not ?

As a long time satisfied NEX-7 owner contemplating the possible purchase of an a7S, I am trying to clarify just how much of an improvement in low light performance I might expect.

Using DxOMark measurements as a starting point, I see that my NEX-7 is rated at 1016 and that the a7S is measured at 3702.

One might conclude that this measured less than a 4-fold increase in ISO should result in less than a 2 stop benefit, since each stop is a doubling / halving.

The comparison of the published specifications of two camera bodies as described by Sony however suggests a much larger gain, since one camera, the NEX-7, was specified as having a max sensitivity of 16,000 ISO, while the new a7S is described as having a max sensitivity of 102,400, over a six-fold improvement.

Is this claim by Sony over-stated, and thus the promised 6+ fold ISO gain increase claimed only really measureses to less than a 4 fold increase?

I would be disappointed to go from my current NEX-7 and all my associated glass to the a7S at considerable expsnse, only to find that the DxO measured improvement of less than 2 stops was all that I gained.

To avoid the confusion which may arise from Sony's offering extended ISO beyond 102,400 to 409,000, I want to ask anyone who replies to please leave this extended claim out of the discussion. I do realize that further claims by Sony to extend the a7S to yet higher extended ISO of 409,000 makes the a7S appear to have yet additional sensitivity, but I would judge this to be no more than manually "push processing" the sensor further, no different from manually setting my NEX-7 to expose a couple more stops underexposed. I am trying to compare both bodies based entirely on their measured sensitivities for their normal claimed range of ISO.

Thanks for any insights here,

Larry

Sony Alpha 7S Sony Alpha NEX-7
If you believe there are incorrect tags, please send us this post using our feedback form.
JonTafferOfPhotography Regular Member • Posts: 230
Re: a7S high ISO claim - - -over hyped or not ?
1

LSHorwitz1 wrote:

As a long time satisfied NEX-7 owner contemplating the possible purchase of an a7S, I am trying to clarify just how much of an improvement in low light performance I might expect.

Using DxOMark measurements as a starting point, I see that my NEX-7 is rated at 1016 and that the a7S is measured at 3702.

One might conclude that this measured less than a 4-fold increase in ISO should result in less than a 2 stop benefit, since each stop is a doubling / halving.

The comparison of the published specifications of two camera bodies as described by Sony however suggests a much larger gain, since one camera, the NEX-7, was specified as having a max sensitivity of 16,000 ISO, while the new a7S is described as having a max sensitivity of 102,400, over a six-fold improvement.

Is this claim by Sony over-stated, and thus the promised 6+ fold ISO gain increase claimed only really measureses to less than a 4 fold increase?

I would be disappointed to go from my current NEX-7 and all my associated glass to the a7S at considerable expsnse, only to find that the DxO measured improvement of less than 2 stops was all that I gained.

To avoid the confusion which may arise from Sony's offering extended ISO beyond 102,400 to 409,000, I want to ask anyone who replies to please leave this extended claim out of the discussion. I do realize that further claims by Sony to extend the a7S to yet higher extended ISO of 409,000 makes the a7S appear to have yet additional sensitivity, but I would judge this to be no more than manually "push processing" the sensor further, no different from manually setting my NEX-7 to expose a couple more stops underexposed. I am trying to compare both bodies based entirely on their measured sensitivities for their normal claimed range of ISO.

Thanks for any insights here,

Larry

based on the simple dimensions of a full frame sensor, you will gain more than a stop in SNR performance holding everything constant.

then if you factor in just pixel pitch, there is roughly another stop there, if you are comparing at 100%.

finally, if you consider 2 years of sensor and more importantly image processing advancements, I'd estimate another stop if not 1.5 stops based on image samples I've seen compared to my d4

id say you can expect something like a 3 stop improvement over what you have in the nex-7 in terms of SNR and probably more gradual DR and CD fall off as well.

this method above is better for thinking about practical ISO performance gains, rather than comparing max ISO which is, by in large, misleading and manipulated.

also remember dxo ISO ratings are normalized, so a high MP body will benefit due to downsampling. so in reality the ISO performance of the nex-7 is actually far worse at 100% than you would think based just on dxo's ISO rating. the d800 folks can't seem to wrap their heads around that, but it is true. it becomes really obvious once you start cropping.

 JonTafferOfPhotography's gear list:JonTafferOfPhotography's gear list
Nikon D4 Sony a6000 Nikon AF-S Nikkor 24mm f/1.4G ED Nikon AF-S Nikkor 85mm f/1.4G Nikon AF-S Nikkor 35mm f/1.4G +7 more
Pixel Pooper Senior Member • Posts: 1,805
Re: a7S high ISO claim - - -over hyped or not ?

LSHorwitz1 wrote:

As a long time satisfied NEX-7 owner contemplating the possible purchase of an a7S, I am trying to clarify just how much of an improvement in low light performance I might expect.

Using DxOMark measurements as a starting point, I see that my NEX-7 is rated at 1016 and that the a7S is measured at 3702.

One might conclude that this measured less than a 4-fold increase in ISO should result in less than a 2 stop benefit, since each stop is a doubling / halving.

The 2 stop benefit is only at those ISO settings. At higher ISO the A7S will be more than 2 stops better.

The comparison of the published specifications of two camera bodies as described by Sony however suggests a much larger gain, since one camera, the NEX-7, was specified as having a max sensitivity of 16,000 ISO, while the new a7S is described as having a max sensitivity of 102,400, over a six-fold improvement.

The max ISO setting is 6 stops higher, but that doesn't mean the performance is 6 stops better. I don't think Sony ever claimed that both cameras would perform the same at their max ISO settings.

Is this claim by Sony over-stated, and thus the promised 6+ fold ISO gain increase claimed only really measureses to less than a 4 fold increase?

I would be disappointed to go from my current NEX-7 and all my associated glass to the a7S at considerable expsnse, only to find that the DxO measured improvement of less than 2 stops was all that I gained.

To avoid the confusion which may arise from Sony's offering extended ISO beyond 102,400 to 409,000, I want to ask anyone who replies to please leave this extended claim out of the discussion. I do realize that further claims by Sony to extend the a7S to yet higher extended ISO of 409,000 makes the a7S appear to have yet additional sensitivity, but I would judge this to be no more than manually "push processing" the sensor further, no different from manually setting my NEX-7 to expose a couple more stops underexposed. I am trying to compare both bodies based entirely on their measured sensitivities for their normal claimed range of ISO.

With Sony sensors there is no difference between raising ISO and push processing apart from the lower DR at higher ISO. There is only one sensitivity, and every setting above base ISO is push processed.

Thanks for any insights here,

Larry

 Pixel Pooper's gear list:Pixel Pooper's gear list
Sony RX100 Sony Alpha NEX-6 Sony a6000 Sony E 18-55mm F3.5-5.6 OSS Tamron SP AF 60mm F/2 Di II LD IF Macro +3 more
(unknown member) Veteran Member • Posts: 9,509
Re: a7S high ISO claim - - -over hyped or not ?

LSHorwitz1 wrote:

As a long time satisfied NEX-7 owner contemplating the possible purchase of an a7S, I am trying to clarify just how much of an improvement in low light performance I might expect.

Using DxOMark measurements as a starting point, I see that my NEX-7 is rated at 1016 and that the a7S is measured at 3702.

One might conclude that this measured less than a 4-fold increase in ISO should result in less than a 2 stop benefit, since each stop is a doubling / halving.

DXO is measuring low light performance, not high ISO.....some cameras get a score under 100 ISO.

The difference between the NEX-7 and A7S at higher ISOs will actually be greater.

They calculate the low light ISO score...

"The SNR indicates how much noise is present in an image compared to the actual information (signal). The higher the SNR value, the better the image looks, because details aren't drowned by noise. SNR strength is given in dB, which is a logarithmic scale: an increase of 6 dB corresponds to doubling the SNR, which equates to half the noise for the same signal.

An SNR value of 30dB means excellent image quality. Thus low-light ISO is the highest ISO setting for a camera that allows it to achieve an SNR of 30dB while keeping a good dynamic range of 9 EVs and a color depth of 18bits.

A difference in low-light ISO of 25% represents 1/3 EV and is only slightly noticeable.

As cameras improve, low-light ISO will continuously increase, making this scale open."

http://www.dxomark.com/Cameras/Compare/Side-by-side/Sony-A7S-versus-Sony-NEX-7___949_736

Dynamic range drops away on the NEX-7 more steeply.     Signal to noise seems to drop away similarly and colour depth drops away similarly after short slighter drop for the A7S.

Every review I have seen for the A7S says it is the low light king and most think it is pretty special.

Surprising to me since it seems to be more aimed as a video camera and it might get panned for that for rolling shutter and NOT having 15 stops of DR (even though its DR is better than anything like it).

I know I want one and may well get one at years end.

I would be disappointed to go from my current NEX-7 and all my associated glass to the a7S at considerable expsnse, only to find that the DxO measured improvement of less than 2 stops was all that I gained.

I am not sure how much better it will be for you but I don't think there is anything better currently....maybe Canon will have something in a few months.

Larry

tqlla Veteran Member • Posts: 3,754
Re: a7S high ISO claim - - -over hyped or not ?

LSHorwitz1 wrote:

As a long time satisfied NEX-7 owner contemplating the possible purchase of an a7S, I am trying to clarify just how much of an improvement in low light performance I might expect.

Using DxOMark measurements as a starting point, I see that my NEX-7 is rated at 1016 and that the a7S is measured at 3702.

One might conclude that this measured less than a 4-fold increase in ISO should result in less than a 2 stop benefit, since each stop is a doubling / halving.

The comparison of the published specifications of two camera bodies as described by Sony however suggests a much larger gain, since one camera, the NEX-7, was specified as having a max sensitivity of 16,000 ISO, while the new a7S is described as having a max sensitivity of 102,400, over a six-fold improvement.

Is this claim by Sony over-stated, and thus the promised 6+ fold ISO gain increase claimed only really measureses to less than a 4 fold increase?

I would be disappointed to go from my current NEX-7 and all my associated glass to the a7S at considerable expsnse, only to find that the DxO measured improvement of less than 2 stops was all that I gained.

To avoid the confusion which may arise from Sony's offering extended ISO beyond 102,400 to 409,000, I want to ask anyone who replies to please leave this extended claim out of the discussion. I do realize that further claims by Sony to extend the a7S to yet higher extended ISO of 409,000 makes the a7S appear to have yet additional sensitivity, but I would judge this to be no more than manually "push processing" the sensor further, no different from manually setting my NEX-7 to expose a couple more stops underexposed. I am trying to compare both bodies based entirely on their measured sensitivities for their normal claimed range of ISO.

Thanks for any insights here,

Larry

I dont think the DXO numbers correlate in that manner.

 tqlla's gear list:tqlla's gear list
Sony Alpha DSLR-A850 Sony Alpha a99 Sony a99 II Sony 16-35mm F2.8 ZA SSM Carl Zeiss Vario-Sonnar T* Sony 24-70mm F2.8 ZA SSM Carl Zeiss Vario-Sonnar T* +9 more
Pixel Pooper Senior Member • Posts: 1,805
Re: a7S high ISO claim - - -over hyped or not ?

tqlla wrote:

LSHorwitz1 wrote:

As a long time satisfied NEX-7 owner contemplating the possible purchase of an a7S, I am trying to clarify just how much of an improvement in low light performance I might expect.

Using DxOMark measurements as a starting point, I see that my NEX-7 is rated at 1016 and that the a7S is measured at 3702.

One might conclude that this measured less than a 4-fold increase in ISO should result in less than a 2 stop benefit, since each stop is a doubling / halving.

The comparison of the published specifications of two camera bodies as described by Sony however suggests a much larger gain, since one camera, the NEX-7, was specified as having a max sensitivity of 16,000 ISO, while the new a7S is described as having a max sensitivity of 102,400, over a six-fold improvement.

Is this claim by Sony over-stated, and thus the promised 6+ fold ISO gain increase claimed only really measureses to less than a 4 fold increase?

I would be disappointed to go from my current NEX-7 and all my associated glass to the a7S at considerable expsnse, only to find that the DxO measured improvement of less than 2 stops was all that I gained.

To avoid the confusion which may arise from Sony's offering extended ISO beyond 102,400 to 409,000, I want to ask anyone who replies to please leave this extended claim out of the discussion. I do realize that further claims by Sony to extend the a7S to yet higher extended ISO of 409,000 makes the a7S appear to have yet additional sensitivity, but I would judge this to be no more than manually "push processing" the sensor further, no different from manually setting my NEX-7 to expose a couple more stops underexposed. I am trying to compare both bodies based entirely on their measured sensitivities for their normal claimed range of ISO.

Thanks for any insights here,

Larry

I dont think the DXO numbers correlate in that manner.

The DXO numbers are ISO settings and can be compared as such.

 Pixel Pooper's gear list:Pixel Pooper's gear list
Sony RX100 Sony Alpha NEX-6 Sony a6000 Sony E 18-55mm F3.5-5.6 OSS Tamron SP AF 60mm F/2 Di II LD IF Macro +3 more
Rishi Sanyal
Rishi Sanyal dpreview Admin • Posts: 713
Re: a7S high ISO claim - - -over hyped or not ?
5

Pixel Pooper wrote:

With Sony sensors there is no difference between raising ISO and push processing apart from the lower DR at higher ISO. There is only one sensitivity, and every setting above base ISO is push processed.

While generally true for Sony, this is actually not the case with the A7S. There appears to be some downstream read noise that limits shadow performance at lower ISOs, and this is what caps the low ISO pixel-level dynamic range below that of the A7R.

Also, to the OP: this article we recently published may give you some idea of real-world ISO performance. If you make it through to the 2nd page, you'll note that ISO performance can't simply be talked of in terms of 'this many stops better', because different tones in the image are affected differently by the various sources of noise. For example, brighter tones (higher signals) will really only improve if the sensor gained efficiency (sometimes referred to as 'effective quantum efficiency'), while shadow performance will increase both by this and by lower read noise. We've only seen a tiny increase in actual efficiency, so we believe most of the benefit comes from optimizations in representing very low signals (lower aggregate read noise due to less pixels, lower upstream read noise, lower quantization error, or any combination of these & other factors). That means most of the benefits of the A7S will only be seen in deep shadows of high ISO shots, or in overall performance above extremely high ISOs (circa ISO 51,200 & above). But have a look yourself - different folks have different cutoffs for what they consider 'significant improvement'

That said - and someone already mentioned this earlier - the increase in sensor size alone will give you at least a stop better ISO performance than the NEX-7. Simply b/c a full-frame sensor literally can capture ~2.25x as much light as a similar APS-C sensor. Factor in other optimizations, and you do really have a 'low light king'. But, again, as we say in our conclusions of that article, the benefit of the A7S over, say, the A7R can only really be appreciated at very high ISOs. And more appreciated in the shadows of these high ISOs.

There's also a dynamic range benefit at higher ISOs. At base ISO, however, there's a dynamic range cost compared to the A7R.

Hope this helps,

Rishi

-- hide signature --

“It is a capital mistake to theorize before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts.” -- Sherlock Holmes

 Rishi Sanyal's gear list:Rishi Sanyal's gear list
Sony RX100 IV Nikon D810 Canon MP-E 65mm f/2.5 1-5x Macro Nikon AF-S Nikkor 85mm f/1.8G Sigma 35mm F1.4 DG HSM Art +3 more
andrewD2 Veteran Member • Posts: 8,941
Re: a7S high ISO claim - - -over hyped or not ?

Rishi Sanyal wrote:

Pixel Pooper wrote:

With Sony sensors there is no difference between raising ISO and push processing apart from the lower DR at higher ISO. There is only one sensitivity, and every setting above base ISO is push processed.

While generally true for Sony, this is actually not the case with the A7S. There appears to be some downstream read noise that limits shadow performance at lower ISOs, and this is what caps the low ISO pixel-level dynamic range below that of the A7R.

Also, to the OP: this articlewe recently published may give you some idea of real-world ISO performance. If you make it through to the 2nd page, you'll note that ISO performance can't simply be talked of in terms of 'this many stops better', because different tones in the image are affected differently by the various sources of noise. For example, brighter tones (higher signals) will really only improve if the sensor gained efficiency (sometimes referred to as 'effective quantum efficiency'), while shadow performance will increase both by this and by lower read noise. We've only seen a tiny increase in actual efficiency, so we believe most of the benefit comes from optimizations in representing very low signals (lower aggregate read noise due to less pixels, lower upstream read noise, lower quantization error, or any combination of these & other factors). That means most of the benefits of the A7S will only be seen in deep shadows of high ISO shots, or in overall performance above extremely high ISOs (circa ISO 51,200 & above). But have a look yourself - different folks have different cutoffs for what they consider 'significant improvement'

That said - and someone already mentioned this earlier - the increase in sensor size alone will give you at least a stop better ISO performance than the NEX-7. Simply b/c a full-frame sensor literally can capture ~2.25x as much light as a similar APS-C sensor. Factor in other optimizations, and you do really have a 'low light king'. But, again, as we say in our conclusions of that article, the benefit of the A7S over, say, the A7R can only really be appreciated at very high ISOs. And more appreciated in the shadows of these high ISOs.

There's also a dynamic range benefit at higher ISOs. At base ISO, however, there's a dynamic range cost compared to the A7R.

Hope this helps,

Rishi

-- hide signature --

“It is a capital mistake to theorize before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts.” -- Sherlock Holmes

The A7R is reading better in THEIR DR tests because of the resolution.
What limits DR? It's the detail you can see in the shadows given that the highlights can not be exposed any higher.
So what is better 36Million super clean pixels or 12Million super super clean pixels. In THEIR test the former.
I have a great way of directly comparing DR between cameras, doesn't give you a DR value but it does give you a winner. DXOmark is WAY off on the difference between 5DIII and 6D. So I don't trust their values.
My guess is that for 13 stop scenes the A7R will win out but for 20 stop scenes I'd have to find out.

Andrew

ProfHankD
ProfHankD Veteran Member • Posts: 4,965
Re: a7S high ISO claim - - -over hyped or not ?
3

andrewD2 wrote:

The A7R is reading better in THEIR DR tests because of the resolution.

Read the full set of tests at DxO.  Rather than viewing "scores," click on "measurements" and then "dynamic range."  Two plots are available there. The "print" one shows the dynamic range at each ISO when all images are first scaled to 8MP (the default graph). The "screen" one shows the dynamic range at the native pixel resolution -- not giving the A7R (or A7) an advantage, but essentially penalizing them for their higher resolution.

At marked ISO 100 (which is really a bit lower), the A7S has less dynamic range than the A7 or A7R even in the "screen" graph -- there, the A7 beats the A7R and both beat the A7S. However, the A7S dynamic range drops much slower as the ISO is increased, and at higher ISOs the A7S has a couple of Ev more DR.

I have no affiliation with DxO, but they have earned my respect and I have yet to find a case where they were simply wrong based on my own precision measurements. There are some things I like to measure that DxO doesn't (yet?), but unsupported claims that DxO is wrong really annoy me. I think DxO is providing a huge service to the community and should be applauded for making so much high-quality data freely available.

 ProfHankD's gear list:ProfHankD's gear list
Canon PowerShot A640 Canon PowerShot A720 IS Canon PowerShot S70 Canon PowerShot G1 Canon PowerShot G5 +27 more
andrewD2 Veteran Member • Posts: 8,941
Re: a7S high ISO claim - - -over hyped or not ?

ProfHankD wrote:

andrewD2 wrote:

The A7R is reading better in THEIR DR tests because of the resolution.

Read the full set of tests at DxO. Rather than viewing "scores," click on "measurements" and then "dynamic range." Two plots are available there. The "print" one shows the dynamic range at each ISO when all images are first scaled to 8MP (the default graph). The "screen" one shows the dynamic range at the native pixel resolution -- not giving the A7R (or A7) an advantage, but essentially penalizing them for their higher resolution.

At no time can the A7s resolve 36MP. That means if the detail is resolved in a moderate to high DR scene by the A7r DESPITE a little more noise per pixel at ISO100 it will still be an apparent winner. On an extreme DR scene who knows. 10 stop push back? Anyone tried that?

Andrew

Jefenator
Jefenator Senior Member • Posts: 2,363
Re: a7S high ISO claim - - -over hyped or not ?

I don't usually follow high ISO technical debates very closely. All I know is, going from the NEX-7 to the now-lolwly A7, I felt a very distinct improvement. The sensor seems at least one stop cleaner and the improved EVF makes manual focusing much easier under difficult conditions. That's a pretty big deal right there and I already feel spoiled with what I got.

A7 NEX7 EOS-M ISO Comparison

I would consider the maximum usable ISO on the A7 to be 12800 (and that can be iffy - would rather stay lower if at all possible). Testimonials I've read thus far seem to suggest that the A7s should get you an additional two stops over that: ISO 51200 usable (YMMV). That, along with enhanced AF reliability in low light, could be a huge deal for those who spend most of their time using available light in crazy low quantities.

I could use all this on occasion, but most of the time, I like having the greater resolution, more fidelity at low ISOs (especially the dynamic range) and not least of all an extra $1000 with the plain old A7. I would love to try an A7s as a second body eventually (maybe swoop in on a closeout special).

FWIW Steve Huff seems to be really loving his A7s in all kinds of light. (He is notoriously bubbly about most new acquisitions but this one does seem like a real standout for him.)

http://www.stevehuffphoto.com/2014/07/01/the-sony-a7s-digital-camera-review-wow-period/

 Jefenator's gear list:Jefenator's gear list
Sony Alpha 7 Canon EF-M 11-22mm f/4-5.6 IS STM Sony FE 55mm F1.8 Sony FE 90mm F2.8 macro Canon EOS M +3 more
OP LSHorwitz1 Senior Member • Posts: 2,492
Re: a7S high ISO claim - - -over hyped or not ?

Jefenator wrote:

I don't usually follow high ISO technical debates very closely. All I know is, going from the NEX-7 to the now-lolwly A7, I felt a very distinct improvement. The sensor seems at least one stop cleaner and the improved EVF makes manual focusing much easier under difficult conditions. That's a pretty big deal right there and I already feel spoiled with what I got.

A7 NEX7 EOS-M ISO Comparison

I would consider the maximum usable ISO on the A7 to be 12800 (and that can be iffy - would rather stay lower if at all possible). Testimonials I've read thus far seem to suggest that the A7s should get you an additional two stops over that: ISO 51200 usable (YMMV). That, along with enhanced AF reliability in low light, could be a huge deal for those who spend most of their time using available light in crazy low quantities.

I could use all this on occasion, but most of the time, I like having the greater resolution, more fidelity at low ISOs (especially the dynamic range) and not least of all an extra $1000 with the plain old A7. I would love to try an A7s as a second body eventually (maybe swoop in on a closeout special).

FWIW Steve Huff seems to be really loving his A7s in all kinds of light. (He is notoriously bubbly about most new acquisitions but this one does seem like a real standout for him.)

http://www.stevehuffphoto.com/2014/07/01/the-sony-a7s-digital-camera-review-wow-period/

The improvement from a NEX-7 is substantial, and especially if the majority of use is indoors with a relatively slow zoom lens, my typical situation. Seems like an a7S is an appropriate next step for me, but hate to "start over again" with new full frame lenses. My NEX-7 and 5 lenses have depreciated very quickly over a couple years of ownership.

Jefenator
Jefenator Senior Member • Posts: 2,363
Re: a7S high ISO claim - - -over hyped or not ?

LSHorwitz1 wrote:

The improvement from a NEX-7 is substantial, and especially if the majority of use is indoors with a relatively slow zoom lens, my typical situation. Seems like an a7S is an appropriate next step for me, but hate to "start over again" with new full frame lenses. My NEX-7 and 5 lenses have depreciated very quickly over a couple years of ownership.

That's a tough call. I know something of what you speak - I took a real bruising with the SEL24 and I bought mine used! 

Going to a larger format can be a good way to experience the full brunt of the law of diminishing returns. (How many thousands do we want to spend per extra stop of usable ISO?)

Using legacy glass was the big trick for me. If these A7 bodies could use only FE lenses, I would probably have to wait another couple years. (As it is, tempting as they seem, part of me is hesitant to invest in the official Sony lenses. I have definite concerns about QC, depreciation and future viability. All the while my legacy glass is doing great and I can still jump to Fuji or even Canon with a couple of cheap adapters.)

 Jefenator's gear list:Jefenator's gear list
Sony Alpha 7 Canon EF-M 11-22mm f/4-5.6 IS STM Sony FE 55mm F1.8 Sony FE 90mm F2.8 macro Canon EOS M +3 more
OP LSHorwitz1 Senior Member • Posts: 2,492
Re: a7S high ISO claim - - -over hyped or not ?

Jefenator wrote:

LSHorwitz1 wrote:

The improvement from a NEX-7 is substantial, and especially if the majority of use is indoors with a relatively slow zoom lens, my typical situation. Seems like an a7S is an appropriate next step for me, but hate to "start over again" with new full frame lenses. My NEX-7 and 5 lenses have depreciated very quickly over a couple years of ownership.

That's a tough call. I know something of what you speak - I took a real bruising with the SEL24 and I bought mine used!

Going to a larger format can be a good way to experience the full brunt of the law of diminishing returns. (How many thousands do we want to spend per extra stop of usable ISO?)

Using legacy glass was the big trick for me. If these A7 bodies could use only FE lenses, I would probably have to wait another couple years. (As it is, tempting as they seem, part of me is hesitant to invest in the official Sony lenses. I have definite concerns about QC, depreciation and future viability. All the while my legacy glass is doing great and I can still jump to Fuji or even Canon with a couple of cheap adapters.)

The diminishing return of comparatively large costs to achieve perhaps marginal gains is especially exacerbated in the case of migrating from APSC with my NEX-7 to an a7S, since the a7S trades resolution for low light SNR, and I am not thrilled with losing cropping and editing flexibility of a 24 Mpixel body for only 12 Mpixels.

Moreover, Sony has the recurring habit of nearly obsoleting each new model so rapidly with yet another improved version that resale values are very weak. I have a NEX-7, 2 Zeiss APSC lenses (including the 24mm f/1.8), 3 other lenses also APSC, and they have dropped in total nearly 50% in 2 years. The replacement lenses I would consider are Zeiss pricing yet only Sony G or worse quality, the two FE deluxe zooms 24-70 and 70-200, both very good but neither really outstanding AFAIK.

Spending another $5K and taking a 50% loss on my prior $5K investment after 2 years is quite a steep hit. And why oh why did Sony switch their hot shoe on both my NEX-7 and a77 to yet another design, forcing me to replace my superb Metz flash guns? My dad's Leicas and Rolleiflexes appreciated and had a decade or more of enduring value and prestige, while these Sony marvels are ephemeral with no market cache' after 18 months....

Ah......decisions, decisions........

Rishi Sanyal
Rishi Sanyal dpreview Admin • Posts: 713
Re: a7S high ISO claim - - -over hyped or not ?

andrewD2 wrote:

ProfHankD wrote:

andrewD2 wrote:

The A7R is reading better in THEIR DR tests because of the resolution.

Read the full set of tests at DxO. Rather than viewing "scores," click on "measurements" and then "dynamic range." Two plots are available there. The "print" one shows the dynamic range at each ISO when all images are first scaled to 8MP (the default graph). The "screen" one shows the dynamic range at the native pixel resolution -- not giving the A7R (or A7) an advantage, but essentially penalizing them for their higher resolution.

At no time can the A7s resolve 36MP. That means if the detail is resolved in a moderate to high DR scene by the A7r DESPITE a little more noise per pixel at ISO100 it will still be an apparent winner. On an extreme DR scene who knows. 10 stop push back? Anyone tried that?

Andrew

I'm not sure what you're trying to say, but if one camera has more base ISO DR than the other for one scene, it'll also have higher base ISO DR for any scene, no matter what the DR of that scene.

Apologies in advance if I'm misunderstanding what you're trying to say.

-Rishi

 Rishi Sanyal's gear list:Rishi Sanyal's gear list
Sony RX100 IV Nikon D810 Canon MP-E 65mm f/2.5 1-5x Macro Nikon AF-S Nikkor 85mm f/1.8G Sigma 35mm F1.4 DG HSM Art +3 more
andrewD2 Veteran Member • Posts: 8,941
Re: a7S high ISO claim - - -over hyped or not ?

Rishi Sanyal wrote:

andrewD2 wrote:

ProfHankD wrote:

andrewD2 wrote:

The A7R is reading better in THEIR DR tests because of the resolution.

Read the full set of tests at DxO. Rather than viewing "scores," click on "measurements" and then "dynamic range." Two plots are available there. The "print" one shows the dynamic range at each ISO when all images are first scaled to 8MP (the default graph). The "screen" one shows the dynamic range at the native pixel resolution -- not giving the A7R (or A7) an advantage, but essentially penalizing them for their higher resolution.

At no time can the A7s resolve 36MP. That means if the detail is resolved in a moderate to high DR scene by the A7r DESPITE a little more noise per pixel at ISO100 it will still be an apparent winner. On an extreme DR scene who knows. 10 stop push back? Anyone tried that?

Andrew

I'm not sure what you're trying to say, but if one camera has more base ISO DR than the other for one scene, it'll also have higher base ISO DR for any scene, no matter what the DR of that scene.

Apologies in advance if I'm misunderstanding what you're trying to say.

-Rishi

A dynamic range measurement has to have a cut off at which you say 'this isn't good enough quality shadow detail to be counted as useful DR'. In other words you can say, this camera has 13stop DR of a high quality, or it could be measured as 15 stops DR if the quality threshold is lower.
DXO marks DR results are hopeless, a 5DIII still registers some info at a similar DR to a 6D but the banding of the 5DIII makes the last few stops of DR of very poor quality. Some of the Sony sensors show good detail but the limitation is colour accuracy.

On a 15 stop scene it maybe they (A7R and A7S) both pull up acceptably noisy detail from the deepest shadows and the higher megapixels resolves more detail. However if you pulled up detail from a more extreme scene (or even the same scene shot underexposed) it maybe (I'd have to test it) the lower megapixel camera is still showing some detail whereas the higher megapixel camera gives you nothing useful.

Andrew

viking79
viking79 Forum Pro • Posts: 14,137
If you shoot RAW, I would look at DXO Low Light Sports score

LSHorwitz1 wrote:

As a long time satisfied NEX-7 owner contemplating the possible purchase of an a7S, I am trying to clarify just how much of an improvement in low light performance I might expect.

Using DxOMark measurements as a starting point, I see that my NEX-7 is rated at 1016 and that the a7S is measured at 3702.

What this is saying is based on RAW measurements, if you like the NEX 7 at ISO 1000, you should like the A7S at ISO 3700.

Understand that DXO resizes to a common output size. These are similar to the results you would get resizing your NEX 7 images to 12 MP before doing a pixel level comparison.

One might conclude that this measured less than a 4-fold increase in ISO should result in less than a 2 stop benefit, since each stop is a doubling / halving.

Yes, this is true. There is an approximate 2 stop flexibility advantage to the A7S. This is better than the 1 1/3 stop inherit in full frame vs APS-C, so this is saying the sensor in the A7S is in fact slightly better than the NEX 7 (but in line with A6000 or slightly better).

The comparison of the published specifications of two camera bodies as described by Sony however suggests a much larger gain, since one camera, the NEX-7, was specified as having a max sensitivity of 16,000 ISO, while the new a7S is described as having a max sensitivity of 102,400, over a six-fold improvement.

These numbers are for marketing purposes.

Is this claim by Sony over-stated, and thus the promised 6+ fold ISO gain increase claimed only really measureses to less than a 4 fold increase?

What claim by Sony? Those are not claims, just available ISO settings. The slightly less than 4 fold increase is what you would expect. This is a huge difference. It is like buying an f/1.4 lens instead of an f/2.8 lens (or maybe like f/1.4 and f/2.5).

I would be disappointed to go from my current NEX-7 and all my associated glass to the a7S at considerable expsnse, only to find that the DxO measured improvement of less than 2 stops was all that I gained.

That is a huge improvement. As an example, if you are shooting at your max tolerable ISO two stops means shooting at 1/250 instead of 1/60 (this means motion blur vs no motion blur for a walking person). In terms of lenses 1 stop is usually 2 times the price, so 2 stops would usually be a minimum of 4 times the price (but this is very approximate as lens price is influenced largely by market demands).

At the same time, if you don't need two stops flexibility, it is an extreme waste of money. I.e. if you are using a tripod for landscape work.

The A7S has a very specific market in mind, those that shoot events, etc, and need the high ISO capability, along with videographers who don't need more than 8 MP resolution (4K).

Nothing I say has anything to do with JPEG quality settings. I have no idea how they perform in JPEG. DXO mark should only be considered for RAW shooters (or those looking at sensor performance rather than JPEG engine performance).

Eric

 viking79's gear list:viking79's gear list
Sony Alpha 7R Samsung NX1 Samsung NX 30mm F2 Pancake Samsung NX 85mm F1.4 ED SSA Samsung NX 60mm F2.8 Macro ED OIS SSA +5 more
Rishi Sanyal
Rishi Sanyal dpreview Admin • Posts: 713
Re: a7S high ISO claim - - -over hyped or not ?

andrewD2 wrote:

I'm not sure what you're trying to say, but if one camera has more base ISO DR than the other for one scene, it'll also have higher base ISO DR for any scene, no matter what the DR of that scene.

Apologies in advance if I'm misunderstanding what you're trying to say.

-Rishi

A dynamic range measurement has to have a cut off at which you say 'this isn't good enough quality shadow detail to be counted as useful DR'. In other words you can say, this camera has 13stop DR of a high quality, or it could be measured as 15 stops DR if the quality threshold is lower.

I understand that. DxO uses a cutoff of SNR=1. You may or may not agree with that cutoff, but it does account for read noise difference between cameras. Higher SNR cutoffs for DR measurements lead to DR measurements being more & more affected by shot noise - providing less and less discrimination between similar-size sensors with different read noise performances. Ultimately telling you more about sensor size, not sensor performance.

In other words, there's some good logic behind using low SNR cutoffs. Incidentally, our SNR cutoff is ~3 for our JPEG tone curve tests. We could probably debate cutoffs till the cows come home, which ultimately may be part of the reason DxO settled on 1.

DXO marks DR results are hopeless, a 5DIII still registers some info at a similar DR to a 6D but the banding of the 5DIII makes the last few stops of DR of very poor quality. Some of the Sony sensors show good detail but the limitation is colour accuracy.

I wouldn't call them hopeless, as they correlate quite well to usability of shadow detail in RAW files. For example, according to their data, base ISO DR goes as follows: a7R > a7S >> 5DIII. Which correlates exactly with my real-world experience. You do bring up a good point about banding, which make the 5DIII shadows even less useful. That's a tough problem to tackle, though.

On a 15 stop scene it maybe they (A7R and A7S) both pull up acceptably noisy detail from the deepest shadows and the higher megapixels resolves more detail. However if you pulled up detail from a more extreme scene (or even the same scene shot underexposed) it maybe (I'd have to test it) the lower megapixel camera is still showing some detail whereas the higher megapixel camera gives you nothing useful.

No, because while it's theoretically possible that the lower megapixel camera might have more DR b/c of higher FWC (if pixel-level read noise remains the same), the whole point of the RAW DR test is to see if this is actually the case. In this case, read noise swamps the lower signals on the A7S, so it cannot match the performance of the A7R (at base ISO).

The particular read noise swamping low signals on the A7S at base ISO is typically referred to as 'downstream' read noise - b/c it's introduced downstream of the ISO amplification step. The reason the A7S DR is higher at higher ISOs is b/c amplification of the signal off the sensor decreases the impact of this downstream read noise. That is, upstream read noise becomes the dominant factor limiting SNR of lower signals (ignoring shot noise, b/c this is similar across similar-sized sensors). This is noise introduced upstream of the ISO amplification step - typically the read noise of the imaging sensor itself. This upstream, sensor-level, noise is rather low on the A7S, but it's pretty low - at the pixel-level - on the A7R and 5DIII as well. The aggregate upstream read noise, however, is likely lower on the A7S due to it having fewer pixels. Furthermore, due to some other optimizations, it seems the upstream read noise may be even lower on the A7S than the A7R and/or 5DIII at the highest ISOs as well. Both of these factors (fewer pixel count, lower pixel-level read noise) would contribute to the A7S pulling ahead at higher ISOs.

But at base ISO - or, put another way, for very low signals *not* further amplified before digitization - the A7R wins. And that means it has more base ISO DR than the A7R for any scene.

-Rishi

Andrew

-- hide signature --

“It is a capital mistake to theorize before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts.” -- Sherlock Holmes

 Rishi Sanyal's gear list:Rishi Sanyal's gear list
Sony RX100 IV Nikon D810 Canon MP-E 65mm f/2.5 1-5x Macro Nikon AF-S Nikkor 85mm f/1.8G Sigma 35mm F1.4 DG HSM Art +3 more
andrewD2 Veteran Member • Posts: 8,941
Re: a7S high ISO claim - - -over hyped or not ?

DXO marks DR results are hopeless, a 5DIII still registers some info at a similar DR to a 6D but the banding of the 5DIII makes the last few stops of DR of very poor quality. Some of the Sony sensors show good detail but the limitation is colour accuracy.

I wouldn't call them hopeless, as they correlate quite well to usability of shadow detail in RAW files. For example, according to their data, base ISO DR goes as follows: a7R > a7S >> 5DIII. Which correlates exactly with my real-world experience. You do bring up a good point about banding, which make the 5DIII shadows even less useful. That's a tough problem to tackle, though.

It's THE problem. Not tackling it makes the numbers utterly worthless in the real world. They put 5DIII and 6D 0.4 stop apart. Really? Not in my tests, no where near that close. It would suggest that all you had to do to make the 5DIII's shadows as useable as the 6D's was by using an expose 1/3rd of a stop higher. Anyone with both cameras knows that is not true. So if they are miles off there why would I trust the A7r/A7s findings? I certainly have no more reason to trust it over Sony's DR value.

No point in talking read noise this / SNR that if you can't actually use the DR in a real image.
Why not do your own COMPARATIVE test rather than relying on DXO?

Andrew

Rishi Sanyal
Rishi Sanyal dpreview Admin • Posts: 713
Re: a7S high ISO claim - - -over hyped or not ?

andrewD2 wrote:

DXO marks DR results are hopeless, a 5DIII still registers some info at a similar DR to a 6D but the banding of the 5DIII makes the last few stops of DR of very poor quality. Some of the Sony sensors show good detail but the limitation is colour accuracy.

I wouldn't call them hopeless, as they correlate quite well to usability of shadow detail in RAW files. For example, according to their data, base ISO DR goes as follows: a7R > a7S >> 5DIII. Which correlates exactly with my real-world experience. You do bring up a good point about banding, which make the 5DIII shadows even less useful. That's a tough problem to tackle, though.

It's THE problem.

No. It's a problem. And, trust me, not one that is lost on me, given that I use a Canon 5DIII for my people/wedding photography. Sony for everything else.

Not tackling it makes the numbers utterly worthless in the real world. They put 5DIII and 6D 0.4 stop apart. Really? Not in my tests, no where near that close. It would suggest that all you had to do to make the 5DIII's shadows as useable as the 6D's was by using an expose 1/3rd of a stop higher. Anyone with both cameras knows that is not true. So if they are miles off there why would I trust the A7r/A7s findings? I certainly have no more reason to trust it over Sony's DR value.

If you've had so much experience with these cameras, can you please enlighten us as to an exact EV cost the FPN on the 5DIII has on DR? So that we can pass that along to DxO to add simply as an offset such that their existing infrastructure of tests/metrics - which many find immensely useful - can still remain intact.

Utterly worthless? DxO data shows the Canon 5DIII a full 3 stops behind the a7R in terms of base ISO DR. Which pans out in real-world shooting, though is a bit of an underestimate b/c of the FPN you mention. Does that make it worthless? No.

If assigning some sort of EV penalty for FPN in DR measurements were that simple & widely accepted, perhaps DxO would've had a way to account for this already. If you yourself know of any such solution, please do bring it to our/their attention though.

Also - you do realize that different copies of bodies have varying degrees of FPN, right? How does one address that?

No point in talking read noise this / SNR that if you can't actually use the DR in a real image.
Why not do your own COMPARATIVE test rather than relying on DXO?

Andrew

Actually, there is very much a point. B/c the general trends that emerge are entirely valid. Sure, there's an additional FPN cost for certain cameras that hurts DR a bit more. Does that make the information provided that shows a ~3 stop base ISO DR advantage for the a7R over the 5DIII useless, or the information that shows a ~1 stop base ISO DR advantage of the a7R over the a7S useless? Or are the results showing the higher ISO DR advantage of the a7S over the a7R invalid/useless? Absolutely not.

To argue that not accounting for FPN suddenly makes all the studies useless/untrustworthy is simply misleading.

By the way, do you have any idea of what the overhead of doing comparative tests between all possible combinations is? One of the major points of setting up controlled studies in the first place is so that you can meaningfully compare cameras in the future with cameras in the past. Even the high ISO one-off test I did here is not something we can continue to do in any meaningful sort of manner, which is why we're working on new in-house tests.

And, by the way, I'm going to let you in on a little secret: we don't hang on to every camera we test forever.

I'd ask that in the future, before fervently responding with all caps, to please consider these sorts of things. That said, I do agree with you that accounting for FPN in some manner would be useful. Although, more and more we're seeing manufacturers ensure that FPN is subdued to minimal levels. There will, of course, always be exceptions.

-Rishi

 Rishi Sanyal's gear list:Rishi Sanyal's gear list
Sony RX100 IV Nikon D810 Canon MP-E 65mm f/2.5 1-5x Macro Nikon AF-S Nikkor 85mm f/1.8G Sigma 35mm F1.4 DG HSM Art +3 more
Keyboard shortcuts:
FForum MMy threads