SONY A9 Gallery out. Not impressed!

[Marianne Oelund]

Here is beauty at ISO 1000.....not good at all

Shutter 1/2000 at f/2.8 and ISO 1000 is about LV 10.5.

I'm lucky if I can get LV 7 in a rink, and the D5 does better than this example in that light.

--
Source credit: Prov 2:6
- Marianne

 

[OtiGuy]

Sony A9 only has 12 bit files in continuous shooting mode. Another Sony fail.

Loading…

blog.kasson.com

 

[NickInLA]

Here is beauty at ISO 1000.....not good at all

Shutter 1/2000 at f/2.8 and ISO 1000 is about LV 10.5.

I'm lucky if I can get LV 7 in a rink, and the D5 does better than this example in that light.

--
Source credit: Prov 2:6
- Marianne

I hope you were being sarcastic :-D Actually that is one of the better examples. I personally feel out of the two photographers, Carey's shots are much more bearable than Rishi's shots. Also I do not see why someone would go -2 EV and not bump up the ISO. I can understand that with film but why on digital?

 

[d750guy]

Tony Northrup says the A9 sharpness, dynamic range, and high iso are not as good as the A7RII. If that's true, then the A9 is considerably inferior to my D750 high iso, since the D750 beats the A7RII on dpreviews iso comparison tool. And the D750 costs one third of the A9.

A7RII is 42 MP. Hard to believe Sony will release a 24MP sensor inferior to 42MP in high ISO. Let's wait for RAW samples

--
Thanks
Jemini Joseph
http://wildbirdimages.com/

What difference will raw files make? If you watched Tony Northrup's review, he shot the same scene, same lighting, same time, from both cameras and the A9 was SIGNIFICANTLY worst than the 42mp A7RII.

 

[Rishi Sanyal]

https://www.dpreview.com/samples/0552112277/sony-a9-samples-gallery

I know this is early but those pictures in the gallery viewed at 100% show terrible noise and lack of detail and smudging even at ISO 1000. I believe they are OOC jpgs but I don't think RAWs will be much different.

Also the pictures shot by Carey seem better than the ones shot by Rishi in terms of clarity and it could be in camera settings but overall the early results are not impressive at all. When viewed at the fit screen mode they really look OK but not when 100% pixel peeped.

Two potential reasons for why many of Carey's shots may have more clarity than many of mine:

1. Carey mostly shot at F2.8 with the 70-200 GM; the majority of my shots were at F5.6 on the long-end of the 100-400 (there were only a couple 100-400s available, so we split up our shooting; Carey with the 70-200, me mostly with the 100-400). I also tended toward higher shutter speeds to ensure I froze the action. 2-3 stops makes a big difference when you're already at ISOs in the thousands.

2. I added some contrast back into my shots because I find Sony's standard JPEG rendering to be a bit flat with DRO Auto enabled. That will accentuate noise. I meant to indicate this in each individual photo, but at 6 am last night after 10 days of non-stop work, I hit my wall. Apologies for that; I will annotate each image with the adjustments shortly.

Hope this helps,

Rishi
--------------------------
Rishi Sanyal, Ph.D
Technical Editor | Digital Photography Review
dpreview.com (work) | rishi.photography (personal)

 

[Rishi Sanyal]

Sony A9 only has 12 bit files in continuous shooting mode. Another Sony fail.

http://blog.kasson.com/the-last-word/a9-drops-to-12-bit-precision-in-continuous-mode/

We reported this on the launch date - I'm confused why you're spamming this same message all over our site.

But since you are, here's a little perspective you may wish to consider:

Though we haven't measured it yet, the dynamic range in 12-bit mode is still likely to be higher on the Sony a9 than the Canon 1D X or the Nikon D5. And dynamic range is the only thing affected by 12-bit (no, nothing to do with 'smooth gradations' or whatnot other marketing BS companies would like you to believe).

So how, exactly, is this a failure, when most sports shooters do just fine with a 1D X and D5?

Please, help me understand. That said, I do wish Sony cameras didn't drop to 12-bit in continuous shooting, because their sensors are clean enough (have low enough read noise) to take advantage of the extra bit-depth, unlike previous Canon cameras that could've gotten away with 12-bit ADCs with no image quality loss (there's no point in sampling at 14-bit depth if your read noise is so high that you only ever have 12 EV of dynamic range, which is more than most Canon cameras had at a pixel level before they finally went on-chip ADC, like Sony).

My main point is that the only camera that will do better at these high frame rates is the 1D X II, and perhaps some of the older Nikons (D4s). But since these cameras are typically shot at high ISO, there will be no difference whatsoever. 12-bit doesn't matter at high ISOs where you have far less dynamic range than 12 EV. This is why people need to understand the science, not just spew specs that help them troll one brand against another.

In other words: the drop to 12-bit will have literally no bearing on most photographers, unless extra noise is introduced due to a lack of correlated double sampling or something (we haven't had a chance to test this yet).
--------------------------
Rishi Sanyal, Ph.D
Technical Editor | Digital Photography Review
dpreview.com (work) | rishi.photography (personal)

 

[NickInLA]

https://www.dpreview.com/samples/0552112277/sony-a9-samples-gallery

I know this is early but those pictures in the gallery viewed at 100% show terrible noise and lack of detail and smudging even at ISO 1000. I believe they are OOC jpgs but I don't think RAWs will be much different.

Also the pictures shot by Carey seem better than the ones shot by Rishi in terms of clarity and it could be in camera settings but overall the early results are not impressive at all. When viewed at the fit screen mode they really look OK but not when 100% pixel peeped.

Two potential reasons for why many of Carey's shots may have more clarity than many of mine:

1. Carey mostly shot at F2.8 with the 70-200 GM; the majority of my shots were at F5.6 on the long-end of the 100-400 (there were only a couple 100-400s available, so we split up our shooting; Carey with the 70-200, me mostly with the 100-400). I also tended toward higher shutter speeds to ensure I froze the action. 2-3 stops makes a big difference when you're already at ISOs in the thousands.

2. I added some contrast back into my shots because I find Sony's standard JPEG rendering to be a bit flat with DRO Auto enabled. That will accentuate noise. I meant to indicate this in each individual photo, but at 6 am last night after 10 days of non-stop work, I hit my wall. Apologies for that; I will annotate each image with the adjustments shortly.

Hope this helps,

Rishi
--------------------------
Rishi Sanyal, Ph.D
Technical Editor | Digital Photography Review
dpreview.com (work) | rishi.photography (personal)

Rishi I appreciate your input on this thread. Your clarification between selection of lenses is indeed helpful. When I said Carey's are better I was actually alluding to perceived clarity. At 100 percent they too are little smudgy. Not what we've come to expect from gear of this Calibre. There is noise and then there is missed focus. Yours show both as evidenced.

 

[tqlla]

Tony Northrup says the A9 sharpness, dynamic range, and high iso are not as good as the A7RII. If that's true, then the A9 is considerably inferior to my D750 high iso, since the D750 beats the A7RII on dpreviews iso comparison tool. And the D750 costs one third of the A9.

A7RII is 42 MP. Hard to believe Sony will release a 24MP sensor inferior to 42MP in high ISO. Let's wait for RAW samples

 

[Rishi Sanyal]

Here is beauty at ISO 1000.....not good at all

Shutter 1/2000 at f/2.8 and ISO 1000 is about LV 10.5.

I'm lucky if I can get LV 7 in a rink, and the D5 does better than this example in that light.

--
Source credit: Prov 2:6
- Marianne

This shot was pushed at least 1.33 EV in post, so I'd dial that LV 10.5 down to 9.17. Furthermore, I've added contrast to this image because Sony JPEGs by default appear a bit flat (and green) to me. So consider it an ISO 2500 image with some added contrast.

It's entirely possible the D5 would have slightly less noise and, therefore, better detail retention in its JPEG. The D5 is one of the best low light performers out there in terms of Raw noise, as we've stated in our review, despite being somewhat contradictory to DXO's confusing results where the D5 falls behind the 1D X II (something we disagree with).

But I doubt it'd be much better or worse. Especially considering that Sony's JPEG detail retention is amongst the best - something the entire IQ testing industry thus far agrees on (including quantitative tests using the latest detail retention metrics set forth by the ISO committee on image quality).

My point is that I don't expect other full-frame cameras to do significantly better in terms of image quality in the same situation. But we'll know more once we run the a9 through our standard studio test scene. Its BSI design should give it a leg up over the D5, and I'd assume it's dual gain would help as well. But we won't know for sure until we've done controlled studies.

My apologies if the lack of description of how I processed the files is causing confusion.

Best wishes,
Rishi

--------------------------
Rishi Sanyal, Ph.D
Technical Editor | Digital Photography Review
dpreview.com (work) | rishi.photography (personal)

 

[HBowman]

Can you please, guys, stop speaking about sony products on Nikon forum ?? I'm tired of those comparisons. I'm Nikon shooter, only Nikon shooter and when I come on this forum I want to speak about Nikon FX products, past/present/future... not about other brands trying to hunt on the same ground. I could not care less about this.

--
Kind regards - http://www.hulyssbowman.com -

 

[antoineb]

Yeah, let's all get really worked up about pre-release stuff!

Frankly based on a broader sample of reviewers than just DPR, the IQ is clearly there, for a 24mp sensor.

But instead of getting all worked up, just take a deep breath, and wait until we have officiel reviews based on the actual final product, several months from now.

https://www.dpreview.com/samples/0552112277/sony-a9-samples-gallery

I know this is early but those pictures in the gallery viewed at 100% show terrible noise and lack of detail and smudging even at ISO 1000. I believe they are OOC jpgs but I don't think RAWs will be much different.

Also the pictures shot by Carey seem better than the ones shot by Rishi in terms of clarity and it could be in camera settings but overall the early results are not impressive at all. When viewed at the fit screen mode they really look OK but not when 100% pixel peeped.

 

[[email protected]]

My understanding:

Dynamic range is defined as the full well capacity divided by the noise, it doesn't have anything to do with the bit depth.

Bit depth defines how many steps the dynamic range is divided up into.

Any dynamic range can be represented by any bit depth, lower bit depth leads to more likelihood of posterization.

maljo

 

[scokill]

https://www.dpreview.com/samples/0552112277/sony-a9-samples-gallery

I know this is early but those pictures in the gallery viewed at 100% show terrible noise and lack of detail and smudging even at ISO 1000. I believe they are OOC jpgs but I don't think RAWs will be much different.

Also the pictures shot by Carey seem better than the ones shot by Rishi in terms of clarity and it could be in camera settings but overall the early results are not impressive at all. When viewed at the fit screen mode they really look OK but not when 100% pixel peeped.

Two potential reasons for why many of Carey's shots may have more clarity than many of mine:

1. Carey mostly shot at F2.8 with the 70-200 GM; the majority of my shots were at F5.6 on the long-end of the 100-400 (there were only a couple 100-400s available, so we split up our shooting; Carey with the 70-200, me mostly with the 100-400). I also tended toward higher shutter speeds to ensure I froze the action. 2-3 stops makes a big difference when you're already at ISOs in the thousands.

2. I added some contrast back into my shots because I find Sony's standard JPEG rendering to be a bit flat with DRO Auto enabled. That will accentuate noise. I meant to indicate this in each individual photo, but at 6 am last night after 10 days of non-stop work, I hit my wall. Apologies for that; I will annotate each image with the adjustments shortly.

Hope this helps,

Rishi
--------------------------
Rishi Sanyal, Ph.D
Technical Editor | Digital Photography Review
dpreview.com (work) | rishi.photography (personal)

Sorry, had an incorrect comment.

 

[NickInLA]

Can you please, guys, stop speaking about sony products on Nikon forum ?? I'm tired of those comparisons. I'm Nikon shooter, only Nikon shooter and when I come on this forum I want to speak about Nikon FX products, past/present/future... not about other brands trying to hunt on the same ground. I could not care less about this.

 

[jeffwunder]

It's a fair request and understandable but we're not talking about an older Sony. This camera is a FX camera capable of doing some wonderful things and we're just analyzing pictures taken with this camera. It'll die down soon but as of now it's a Nikon competitor and nothing wrong in discussing pros and cons of any other system.

The A9 has been hyped here. The hype got my attention, so I went into this with an open mind and viewed alot of A9 images from dpreview and image-resource. I have to say, I'm not all that impressed. These pics could have come from any decent semi-pro camera on the market today. Nothing stands out to me. Sure, the A9 has many great features: speed, focusing, silence, no blackouts etc. Probably great video too. But if IQ is not at D5 or 1DXII level, is it really in the same league? Is 20fps really that important vs 12fps? There are plenty of important features a camera has to have, but IQ is still #1 to me. The A9 seems to be more of a statement of what is eventually possible with mirrorless, and I guess that's cool.

 

[NickInLA]

Was expecting way better than this!

Loading…

www.dpreview.com

https://www.dpreview.com/samples/0552112277/sony-a9-samples-gallery

I know this is early but those pictures in the gallery viewed at 100% show terrible noise and lack of detail and smudging even at ISO 1000. I believe they are OOC jpgs but I don't think RAWs will be much different.

Also the pictures shot by Carey seem better than the ones shot by Rishi in terms of clarity and it could be in camera settings but overall the early results are not impressive at all. When viewed at the fit screen mode they really look OK but not when 100% pixel peeped.

Two potential reasons for why many of Carey's shots may have more clarity than many of mine:

1. Carey mostly shot at F2.8 with the 70-200 GM; the majority of my shots were at F5.6 on the long-end of the 100-400 (there were only a couple 100-400s available, so we split up our shooting; Carey with the 70-200, me mostly with the 100-400). I also tended toward higher shutter speeds to ensure I froze the action. 2-3 stops makes a big difference when you're already at ISOs in the thousands.

2. I added some contrast back into my shots because I find Sony's standard JPEG rendering to be a bit flat with DRO Auto enabled. That will accentuate noise. I meant to indicate this in each individual photo, but at 6 am last night after 10 days of non-stop work, I hit my wall. Apologies for that; I will annotate each image with the adjustments shortly.

Hope this helps,

Rishi
--------------------------
Rishi Sanyal, Ph.D
Technical Editor | Digital Photography Review
dpreview.com (work) | rishi.photography (personal)

 

[LiveFromPhilly]

The problem with the DPR tool is that it shows you 100% views for both cameras despite the A7RII having a much higher megapixel count, meaning noise will be more apparent since your seeing a smaller portion of the A7RII's photo. When you print the photos at the same size the A7RII files don't show any more noise but have more fine detail.

Same goes for the D810. Who wants to be downsizing all the time. Much of the time we tend to crop the larger photo. Makes it nice when the you don't have to worry about downsizing to get cleaner images. The thing is today's cameras are way ahead of cameras 10 years ago. I don't care for DXO magic sauce results. Plus there comparison of sensors. Wow sensor is rates at 2900 and the other 3000. So the 3000 must be so much cleaner then the 2900 one. Though it is better then the subjective resolutions for lenses. Anyway as you can tell I'm not much on DXO results. If I can see it in real world photos means more.

The last part may have more to do with D750 has an AA filter and the Sony 7RII doesn't. Also the tool doing the downsizing tend to add some sharpening when downsizing. I know Photoshop advises it. I have seen those results between my D810 and D750. I normally have to add some sharpening to my D750 images because of the AA filter.

Update: Went back and checked the comparative. It does equalize them out between 6400 and 25600. I don't see either as the clear winner. Still means you have to downsize to do it. The sharpness difference still can be attributed to AA filter and no filter.

It depends on how you use the camera, I suppose. When I share photos online they generally have to be normalized for a size much smaller than their maximum output regardless of the camera (I think my old D60 is about the only camera with file sizes small enough by default). In the instances I print, I'm printing to the same size as well. You're correct that the D810 also gets that benefit.

You make a good point about the AA filter. I forgot about that.

 

[StillLearning]

The problem with the DPR tool is that it shows you 100% views for both cameras despite the A7RII having a much higher megapixel count, meaning noise will be more apparent since your seeing a smaller portion of the A7RII's photo. When you print the photos at the same size the A7RII files don't show any more noise but have more fine detail.

Same goes for the D810. Who wants to be downsizing all the time. Much of the time we tend to crop the larger photo. Makes it nice when the you don't have to worry about downsizing to get cleaner images. The thing is today's cameras are way ahead of cameras 10 years ago. I don't care for DXO magic sauce results. Plus there comparison of sensors. Wow sensor is rates at 2900 and the other 3000. So the 3000 must be so much cleaner then the 2900 one. Though it is better then the subjective resolutions for lenses. Anyway as you can tell I'm not much on DXO results. If I can see it in real world photos means more.

The last part may have more to do with D750 has an AA filter and the Sony 7RII doesn't. Also the tool doing the downsizing tend to add some sharpening when downsizing. I know Photoshop advises it. I have seen those results between my D810 and D750. I normally have to add some sharpening to my D750 images because of the AA filter.

Update: Went back and checked the comparative. It does equalize them out between 6400 and 25600. I don't see either as the clear winner. Still means you have to downsize to do it. The sharpness difference still can be attributed to AA filter and no filter.

It depends on how you use the camera, I suppose. When I share photos online they generally have to be normalized for a size much smaller than their maximum output regardless of the camera (I think my old D60 is about the only camera with file sizes small enough by default). In the instances I print, I'm printing to the same size as well. You're correct that the D810 also gets that benefit.

You make a good point about the AA filter. I forgot about that.

I came from the film days and the digital cameras far exceed them. So I'm just happy to have my pick of what I like. I really could careless on who is king of the hill. Only downside is I'm putting way more time in post processing.

I would prefer the Sony A7Rii over the A9 for my type of photography. I just don't like how the camera feels in my hands. Too small and too many sharp edges. They need to hire an Italian designer. Don't like having to use LR for RAW conversion. Don't like the higher prices though the A7Rii is coming down. I appreciate that I can pick the qualities I want in a camera. I prefer the larger cameras. Feels better in my XL sized hands.

 

[Dr Blackjack]

Seems normal for early released camera galleries. Pictures usually improve in more capable hands. I did notice the the NR setting seems aggressive. But then again I keep mine set on low. I'll take extra grain to preserve more detail.

Your comment has inspired a new (maybe not) idiom for me in context of photography. "Take this with a grain of sensor noise"

 

[StillLearning]