Reading mode: Light Dark

Why is the DR of the D3s no better than the D300?

Go to last post
J

jfriend00

Veteran Member
Messages
13,711
Solutions
4
Reaction score
4,980
Location
San Francisco Bay Area, US
I'm contemplating getting a used D3s for lower light sports shooting. When looking at the DXO measurements for it, I was surprised by a couple things:

1. The D3s has basically the same DR as the D300 (12.2 vs. 12). How can that be? With such a lower noise floor (almost 2 stops better noise performance) and bigger pixels and same resolution, how can the D3s not also have better DR? I was thinking that an FX camera would have to have better DR than a DX camera of similar generation.

2. Is the D610 really 2.4evs higher in DR than the D3s (14.4 vs. 12)?

Obviously the D3s is great at high ISO in low light. But, I was surprised it was so low in DR. Why would that be?
 
Flashlight said:
jpeg or RAW? DPR says 'ACR best' is D300 10.1 vs D3s 12.2.
Click to expand...
I care about RAW and it's my understanding that RAW is what DXO measures though I don't know what their methodology is. I have no interest in JPEG measurements since I never shoot JPEG.

10.1 vs 12.2 might be more what I would expect, but why does DXO measure something so different for the D300?
 
jfriend00 said:
I'm contemplating getting a used D3s for lower light sports shooting. When looking at the DXO measurements for it, I was surprised by a couple things:
  1. The D3s has basically the same DR as the D300 (12.2 vs. 12). How can that be? With such a lower noise floor (almost 2 stops better noise performance) and bigger pixels and same resolution, how can the D3s not also have better DR? I was thinking that an FX camera would have to have better DR than a DX camera of similar generation.
  2. Is the D610 really 2.4evs higher in DR than the D3s (14.4 vs. 12)?
Obviously the D3s is great at high ISO in low light. But, I was surprised it was so low in DR. Why would that be?
Click to expand...
My guess is that to achieve the insane high-ISO DR of the D3S (three stops more the D300 at 6400) they had to trade in som base-ISO DR.
Maybe the level of readnoise limited the base-ISO DR of that "old" sensor in D3S?
Regarding D610 vs D3S, the D3S has better DR from ISO800.

If you look at D7000 vs D4 you see the same thing. D7000 ahead a base ISO but then it loses roughly a stop of DR for every stop of ISO.
The D4 can take almost four stops of increasing ISO before it loses one stop of DR.

Adam

PS
English is not my first language but I try... = )
 
Last edited:
If you look at D4 x D7100 or D7000, both smaller sensors, Toshiba and Sony, beat the D4 re DR at base ISO, but not as ISO moves up. Just very different sensor/ADC designs.
 
I like Bill's plot of PDR to show what's going on between sensors. the D3S moves away from the D300 starting at ISO 200, and is 2 stops better above ISO 800

a438df89893142db9443dbf469fd77fe.jpg
 
Last edited:
DR is not a constant measure - it falls as the ISO goes up. The summary chart only shows the DR rating at base ISO.

IF you want to use DXO as your reference source. You need to look at the full graph across the ISO range.

DXO shows the D300 DR dropping like a rock as you go up the iso range.

The D3s can still hold onto approx 11 stops at ISO 1600. The D300 drops to about 8.5 stops at the same ISO rating.

So if all you shoot is at base ISO - go for the D300, but if you like to shoot at high ISO - forget about the D300.

jfriend00 said:
I'm contemplating getting a used D3s for lower light sports shooting. When looking at the DXO measurements for it, I was surprised by a couple things:

1. The D3s has basically the same DR as the D300 (12.2 vs. 12). How can that be? With such a lower noise floor (almost 2 stops better noise performance) and bigger pixels and same resolution, how can the D3s not also have better DR? I was thinking that an FX camera would have to have better DR than a DX camera of similar generation.

2. Is the D610 really 2.4evs higher in DR than the D3s (14.4 vs. 12)?

Obviously the D3s is great at high ISO in low light. But, I was surprised it was so low in DR. Why would that be?

--
John
Gallery: http://jfriend.smugmug.com
Click to expand...
 
jfriend00 said:
I'm contemplating getting a used D3s for lower light sports shooting. When looking at the DXO measurements for it, I was surprised by a couple things:

1. The D3s has basically the same DR as the D300 (12.2 vs. 12). How can that be? With such a lower noise floor (almost 2 stops better noise performance) and bigger pixels and same resolution, how can the D3s not also have better DR? I was thinking that an FX camera would have to have better DR than a DX camera of similar generation.

2. Is the D610 really 2.4evs higher in DR than the D3s (14.4 vs. 12)?

Obviously the D3s is great at high ISO in low light. But, I was surprised it was so low in DR. Why would that be?

--
John
Gallery: http://jfriend.smugmug.com
Click to expand...
DxO's headline DR mark is only at base ISO, makes sense for a camera like the D800 but for a camera like the D3s the higher ISO DR is surely what people are buying it for?

At ISO 200 its already better than the D300 by around half a stop, at ISO 400 by over a stop then around 2 stops by ISO 800.
 
Last edited:
Print vs. Sensor dynamic range.

In print dynamic range, each sensor output is resized to an 8"x12" 300 dpi print

Sensor dynamic range reflects the dynamic range for each sensor pixel.

According to DxO measurements, the only time the D300 matches the D3S dynamic range is at base ISO. For sports shooting, the D3S will excel, and even surpass the D800 above ISO 1000





[ATTACH alt="Print Dynamic Range 8MP 8"x12" 300 dpi print "]351864[/ATTACH]
Print Dynamic Range 8MP 8"x12" 300 dpi print





Sensor Dynamic Range 100% zoom to pixel level
Sensor Dynamic Range 100% zoom to pixel level
 

Attachments

  • f826ddcd24fe4ed7b5441b305c170a83.jpg
    f826ddcd24fe4ed7b5441b305c170a83.jpg
    46.8 KB · Views: 0
jfriend00 said:
I'm contemplating getting a used D3s for lower light sports shooting. When looking at the DXO measurements for it, I was surprised by a couple things:

1. The D3s has basically the same DR as the D300 (12.2 vs. 12). How can that be? With such a lower noise floor (almost 2 stops better noise performance) and bigger pixels and same resolution, how can the D3s not also have better DR? I was thinking that an FX camera would have to have better DR than a DX camera of similar generation.

2. Is the D610 really 2.4evs higher in DR than the D3s (14.4 vs. 12)?

Obviously the D3s is great at high ISO in low light. But, I was surprised it was so low in DR. Why would that be?
Click to expand...
The D3s uses an outboard amp/A-D package, which leaves a higher noise floor at base ISO in favor of a lower noise-floor at high gain. But the D300 only matches DR at base ISO. From ISO400 up, the D3s is far ahead. It exceeds the theoretical maximum for APS-C (zero noise, 100% QE) in that range.

The D610 uses an on-chip amp/A-D configuration which is very low noise at base ISO, and mostly linear all the way up. The only penalty is additional thermal noise, seen especially at high gain settings.

The D4 has an enhanced full-well capacity at ISO100 of approx 120k-e, almost double what the D3s sensor collects. It has a DR at base ISO about a stop or so above its predecessor, in spite of still retaining the outboard amp/A-D configuration. The D4 is much better at base ISO than the D3s.
 
Luke Kaven said:
The D3s uses an outboard amp/A-D package, which leaves a higher noise floor at base ISO in favor of a lower noise-floor at high gain. But the D300 only matches DR at base ISO. From ISO400 up, the D3s is far ahead. It exceeds the theoretical maximum for APS-C (zero noise, 100% QE) in that range.

The D610 uses an on-chip amp/A-D configuration which is very low noise at base ISO, and mostly linear all the way up. The only penalty is additional thermal noise, seen especially at high gain settings.

The D4 has an enhanced full-well capacity at ISO100 of approx 120k-e, almost double what the D3s sensor collects. It has a DR at base ISO about a stop or so above its predecessor, in spite of still retaining the outboard amp/A-D configuration. The D4 is much better at base ISO than the D3s.
Click to expand...
Very interesting. Where do you get this type of info from?
 
Thanks folks for the graphs as you raise the ISO. I know the D3s wins handily for both noise and DR as you raise the ISO. My question was meant to be about base ISO and why there was little difference in DR at base ISO with such a bigger sensor.

Luke's answer points to a tradeoff in the design that sacrifices some noise at base ISO in order to have much lower noise at higher ISO values.
 
Hi John,

Please, anyone reading my comment here don't take it the wrong way.

I'm trying to be as honest as I can, without trolling or something.

That's the whole problem when people buy gear and trust DXO (or any graphs) more than they trust their own eyes, while actually shooting.

That's a trap.

Please, be a photographer, shoot more, and forget graphs. Or take graphs with the largest rock of salt you can find.

I have a D700 and had both D300 and D300s.

All of them are rated as 12 stops of DR at base ISO.

I'm almost willing to bet my car that the D700 is far ahead at base ISO DR. So far ahead it's not even funny.

ISO 200 D700's shadows are almost noiseless, while D300's are a noise fest.*

*what I mean by noise fest, is that without any shadow lifting, you can already see lots of grain both in blue skies, and deep shadows, in D300(s) files.

So I ask you, is DXO right?

Both have 12 stops? Really??

And what about the D3s, with a better sensor than D700's? 12 stops too?

IMO, trust your eyes, not graphs.

Edit that:

I believe the D700 and D3s might be close enough, for their difference in DR shows only after a certain gain level.

But D3s and D300? No way on Earth they're the same.
 
marcio_napoli said:
Hi John,

Please, anyone reading my comment here don't take it the wrong way.

I'm trying to be as honest as I can, without trolling or something.

That's the whole problem when people buy gear and trust DXO (or any graphs) more than they trust their own eyes, while actually shooting.

That's a trap.

Please, be a photographer, shoot more, and forget graphs. Or take graphs with the largest rock of salt you can find.
Click to expand...
Look, I'm thinking about buying a used D3s or D700 to complement my D300 and I'm doing some research about what I can expect in terms of performance (quite frankly do decide how much I think it's worth to spend and how long I'll be happy with the performance vs. newer generation cameras). I don't own these prospective cameras. I can't shoot with them. I could rent each of them and do my own shooting tests (at the cost of several hundred dollars and a lot of time), but I'd rather see what existing research has already been done on them (preferably quantitative, not just some shooter's opinion) in order to better evaluate my choices first.
marcio_napoli said:
IMO, trust your eyes, not graphs.
Click to expand...
I don't own these cameras. I can't shoot with them. FYI, I know the D3s kills the D300 in noise at all ISO values. My question was about DR at base ISO.
 
jfriend00 said:
I'm contemplating getting a used D3s for lower light sports shooting. When looking at the DXO measurements for it, I was surprised by a couple things:

1. The D3s has basically the same DR as the D300 (12.2 vs. 12). How can that be? With such a lower noise floor (almost 2 stops better noise performance) and bigger pixels and same resolution, how can the D3s not also have better DR? I was thinking that an FX camera would have to have better DR than a DX camera of similar generation.

2. Is the D610 really 2.4evs higher in DR than the D3s (14.4 vs. 12)?

Obviously the D3s is great at high ISO in low light. But, I was surprised it was so low in DR. Why would that be?
Click to expand...
 
Kabe Luna said:
jfriend00 said:
I'm contemplating getting a used D3s for lower light sports shooting. When looking at the DXO measurements for it, I was surprised by a couple things:

1. The D3s has basically the same DR as the D300 (12.2 vs. 12). How can that be? With such a lower noise floor (almost 2 stops better noise performance) and bigger pixels and same resolution, how can the D3s not also have better DR? I was thinking that an FX camera would have to have better DR than a DX camera of similar generation.

2. Is the D610 really 2.4evs higher in DR than the D3s (14.4 vs. 12)?

Obviously the D3s is great at high ISO in low light. But, I was surprised it was so low in DR. Why would that be?
Click to expand...
Click to expand...
 
nothing wrong with that. D700 is so slightly better. But... what a difference in practical results.
 
Just to try to clarify my original question (it is so annoying that DPR doesn't allow me to clarify the actual text of the original question), my question is about the DR of the D300 vs. D3s at base ISO only.

I know that the D3s is loads better for noise and DR as the ISO starts going up.

I found it unusual that the DR at base ISO of the much bigger D3s sensor would be the same as the D300 so I wanted to know if this is really the case and, if so, why?
 
Hi John,

Sorry if my reply sounded I was accusing you of not shooting. That's not what I meant, honestly.

I'm sorry if I sounded like that, I really am.

I was just saying in general, do not always trust measured numbers.

Graphs do not tell the real story always.

I can see shadow noise in D300's files before any file torture, while on D700's I see noise only when shadows are lifted a couple stops.

Sorry again if I sounded too harsh, do not intended that.
 
You must log in or register to reply here.

Keyboard shortcuts

f
Forum
m
My threads
Mobile site
About
Editorial content
Cameras & Lenses
Community
All content, design, and layout are Copyright © 1998–2025 Digital Photography Review All Rights Reserved.
Reproduction in whole or part in any form or medium without specific written permission is prohibited.
When you use DPReview links to buy products, the site may earn a commission.
©GPS Media - Guides, Products, Services.
Back
Top