Nikon D3 - weird ISO counts

[5711]

<p>hi there</p>

<p>i have noticed, when using the nikon d3 and the nikon d3x on location, taking shots with those two cameras of basically the same thing (different focal lentgh though) that the nikon d3 tends to overexpose a bit which basically resulted in me having the d3 on EV -0.5 in bright daylight, always.<br /> i took shots of the same thing with the same lens and saw that my d3 overexposed a bit.<br /> i do not care rly to send it in, it has a shuttercount of 370k and will be replaced by a d4s.</p>

<p>anyway. my question concernes this:</p>

<p>today i too this shot and looked at the file info</p>

<p>it says ISO 2540<br /> the camer's screen says iso 2500</p>

<p>just before i did this<br /> i did take longexposures using two big stoppers and the camera was set to:</p>

<p>D lightning : normal<br /> long exp. noise reduction: on<br /> 14 bit raw<br /> uncompressed</p>

<p>any thoughts on this?</p>

<p>this is my profile on nat.geo. your shot, one of those was published in nat geo in oct. 2012.<br /> http://yourshot.nationalgeographic.com/profile/160621/</p>

<p>screenshot of the file info concerning my question:<br /> http://www.photo.net/photo/17689537&size=lg</p>

<p><img src="/photo/17689537&size=lg" alt="" /></p>

<p>thanks in advance</p>

[CvhKaar]

<p>mm Try the same with D-lighting of, D-lighting adjusts some of your settings , but i'm not sure which ones exactly...</p>

[5711]

<p>had this idea when smoking a cigarette right now and did test fot that.<br>

it is not the d lightning</p>

<p>basically everything switched off.</p>

<p>matrix meter, no correction<br>

nothing. only autofocus changes.</p>

 

[5711]

<p>i have this camera for a long time.<br>

serial nr starts with 2020xxx</p>

<p>i do have the buffer upgrade.<br>

and had it in service to adjust the viewfinder, which was off since it once slammed into the ground pretty hard.</p>

<p>but this rly dsnt have anything to do with what i am up against here.</p>

<p>anyone has a d3 and can take a shot at iso 2500 and then look at the file info?</p>

<p>thanks</p>

[ShunCheung]

<p>By any chance you were using auto ISO?</p>

[5711]

<p>nope, i do not use that, ever. however, i have set a range, but i do not use it. <br /> i will disable the iso range used by auto iso and test again tomorrow.<br>

will do a complete swipe of the sutom settings too and try that.</p>

<p>anyone got a d3 and can test if that error i am getting can be reproduced?</p>

[kari_oinonen]

<p>What lens did you use and at what aperture values?<br>

For some lenses Nikons DSLRs do a trick to compensate for the actually "missing" full aperture.<br>

This was seen with 1.4/50G - only two clicks from f2 to f1.4 with 1/3 steps when using a film camera (F100). No ISO compensation available in that case.<br>

So I have heard, that now the situation with a DSLR is: three clicks from f2.0 to f1.4 BUT the last step is compensated by the ISO setting. Using that lens. In the lens practically nothing changes between f1.4 and f1.6<br>

Maybe the case or similar?<br>

If you can see aperture blades to move at aperture setting fully open minus 1/3 stops, then you may have another "problem".</p>

 

[elliot1]

<p>I just tested my D3 - no issue. Mine is older also, serial number starts with 2028xxx.<br>

<br>

I suggest you reset your camera and try again.</p>

[5711]

<p>that is my iso-wonderland:<br>

Lo 1- iso 100<br>

lo 07 - iso 126<br>

lo 03 - iso 159<br>

200 - iso 200<br>

250 - iso 252<br>

320 – iso 317<br>

400 – iso 400<br>

500 – iso 504<br>

640 – iso 635<br>

800 - iso 800<br>

1000 – iso 1008<br>

1250 – iso 1270<br>

1600 – iso 1600<br>

2000 - iso 2016<br>

2500 – iso 2540<br>

3200 – iso 3200<br>

4000 – iso 4032<br>

5000 – iso 5080<br>

6400 – iso 6400</p>

<p>testet with<br>

14-24 @ 2.8 -4<br>

24 @ 1.4 -4<br>

24-70 @2.8 -4<br>

50 @1.8 -4<br>

70-200 @2.8- 4</p>

<p>camera reset completely, updated, nothing changed to it.<br>

nothing changed, no problems like you mentioned @ kari</p>

<p>i will send it in and have the shutter renewed.<br>

will ask them what this is providing my iso list..</p>

<p>i get a feeling, we will never know ...</p>

<p> </p>

[5711]

<p>sent nikon a mail proving my test series with changes to the camera and the list that is absolutely the same after the reset and asked them what the hell this is.<br>

over time i've worked around this and i learned to deal with it like i started to rly get the weird autofocus of the canon 1d mk 3.<br>

but having a couple of days off left me with noting to do..so ..well..we will see what nikon tech support tells me.</p>

<p>send it in, we will repair it for you.<br>

this will cost you this and that<br>

i bet..</p>

<p>with new products they're rly good, but with old used ones..i will find out :)</p>

<p>thank you guys for your time and responses, i will let you know what nikon tells me.</p>

<p>cheers</p>

[5711]

<p>sent nikon a mail proving my test series with changes to the camera and the list that is absolutely the same after the reset and asked them what the hell this is.<br>

over time i've worked around this and i learned to deal with it like i started to rly get the weird autofocus of the canon 1d mk 3.<br>

but having a couple of days off left me with noting to do..so ..well..we will see what nikon tech support tells me.</p>

<p>send it in, we will repair it for you.<br>

this will cost you this and that<br>

i bet..</p>

<p>with new products they're rly good, but with old used ones..i will find out :)</p>

<p>thank you guys for your time and responses, i will let you know what nikon tells me.</p>

<p>cheers</p>

[5711]

<p>yeah, as i thought</p>

<p>send it in they said.</p>

 

[mike_halliwell]

<p>Well there's definitely a pattern in those numbers!<br /> 200 - iso 200<br />250 - iso <strong>252</strong><br />320 – iso <em>317</em><br />400 – iso 400<br />500 – iso <strong>504</strong><br />640 – iso <em>635</em><br />800 - iso 800<br /> ....<br /> 1600 – iso 1600<br />2000 - iso <strong>2016</strong><br />2500 – iso <em>2540</em><br />3200 – iso 3200<br />4000 – iso <strong>4032</strong><br />5000 – iso <em>5080</em><br />6400 – iso 6400<br /> <br /> The same doubling going on....but not quite in the usual 1/3 EV way. <br /> <br /> Take the last set iso 3200 to 6400. The true 1/3 stops are 4266 and 5322...but have always been 4000 and 5000 for simplicity. 4032 and 5080 are just wrong.<br /> <br /> However, the wrong numbers do have their associated doubles....252>504....2016>4032 etc<br /> <br /> Equally, the conventional series of say....200...250...320...400 has always been wrong, it should be closer to 200...266...333...400, but no-one really cares as long as the pic comes out!</p>

<p>I don't think the reported ISO and the overexposure can be connected, at lest not in the normal way as the whole stops report correctly so it's not a consistent off-set. It would imply if you took shots with the ISO set on 100, 200, 400, 800, 1600 etc, they'd be just peachy. I don't think that's true!</p>

<p>LATE EDIT. You say you dialled in -1/2EV on the D3, can you do that when the camera's working in 1/3EV steps? Never thought about it! Maybe it's the numerical difference?</p>

[5711]

<p>i dnt use 1/3 EV steps, i use 0.5</p>

<p>however, when i set the camera to iso 1/3 steps the problem is consistant.<br>

if it is not set to iso 1/3, but full steps, the problem only is in the high iso values.</p>

<p>so as far as i figured it out by now, this has something to do with the iso settings.</p>

<p>however it is changing a bit and values appear and reappear depending on what i do or not.<br>

it is weird as it seems to be a bit random too.</p>

<p>the metering difference between the d3 and the d3x has to be the fact that those two cameras use different metering and i very much like the d3x better actually as the d3 tends to overexpose a bit.</p>

<p>i managed to get the pretty same exposure if the d3 is set on 1/3 EV steps and iso on full steps. and a constant ´-.3 EV dialed in.</p>

<p>the nikon guy stated that it is totally in order that the iso changes a bit when set to 1/3 steps.<br>

however my model seems to be a bit off, so it is a specific problem to my camera which they gladly will take a look at.</p>

<p>i also think it is cool that you actually noticed the numbers chnaging in this pattern, mike.</p>

<p>this actually spawned the idea of trying different iso settings and EV depending on each other and actually solved this</p>

<p>so thank you very much<br>

if we ever meet, i'll buy you some drinks!<br>

and i mean it :)</p>

<p>cheers!</p>

 

[rodeo_joe1]

<blockquote>

<p>"Equally, the conventional series of say....200...250...320...400 has always been wrong, it should be closer to 200...266...333...400, but no-one really cares as long as the pic comes out!"</p>

</blockquote>

<p>Sorry Mike, but that's not right. Stops are measured as powers of two, so a one third stop increase (2^0.33333) is a linear increase of 1.26 times, and two-thirds of a stop (2^0.6667) is a linear multiplier of 1.587. That makes the series from base 200 = 200....252....317.5, and if you then multiply 317.5 by 1.26 (1/3rd stop) you get to 400. Well, allowing for rounding errors you do.</p>

<p>Therefore the OP's reported ISO figures are absolutely correct to the nearest whole number. It's the approximated conventional series that's a bit out.</p>

<p>In any case, a couple of points on the ISO scale in 250 would make two-tenths of naff all difference to the exposure. Let me consult Windows Calculator ........ 0.8% or 1/87th of a stop to be precise! I doubt that a densitometer or lux-meter could reliably detect that after being digitised, let alone the human eye.</p>

[mike_halliwell]

<p>RJ, Right of course ..Doh..! I'd forgotten about that bit. The halving and doubling sequence is the easy (but not accurate) bit!</p>

<p>So how come the OP's read-out is TRUE, but no-one elses reads out like that? Different EXIF reader?</p>

[5711]

<p>i ve tried different tools to do that</p>

<p>windows<br /> photoshop<br /> irfanview</p>

<p>all showed same results.</p>

<p>the maths you've profided seems to make sense.<br>

thx for pointing that out, haven't thought of actually doing that myself, cheers</p>

 

[rodeo_joe1]

<p>No idea why you're getting a more accurate report of ISO speed than the rest of us Norbert. I usually set my ISO to 1/2 stop intervals, and set the aperture and shutter controls to 1/3rd stop. That way I figure I can get to within 1/6th of a stop of the "correct" exposure by varying the ISO as well as aperture and/or shutter. However, I just set my D800 to 1/3rd stop ISO intervals and 250 ISO. The EXIF data still reported the ISO as 250, and not 252; so I don't really know what's going on with your D3. Maybe you still have an early firmware version loaded? Just a guess.</p>

[5711]

<p>nope, lates version of the firmware.<br>

also newest acr for photoshop/lightroom.</p>

<p>nikon service guy told me that those iso settings are normal for 1/3 iso stops.<br>

well at least on a d3/d3x it is supposed to be like this. </p>