D 300 14-bit mode shutter delay...

[greg_morgan]

After shooting with my D300 in available light and studio strobe situations, as well as with SB800's.

I've noticed quite a bit of shutter "delay", or drag in RAW/14 bit mode. As 14-bit was a selling point

of the camera, this is semi-usable feature. Put your D300 in Raw file mode, files at 12-bit, and

shoot in Manual mode, at 1/1000 and faster. Now, switch 'Nef Raw Recording' to 14-bit, and shoot

at the same fast shutter speed... the shutter/mirror sounds like 1/20 of a second! I called Nikon

about this and was told it's the buffer... wouldn't that be post-exposure, not during? If you try

to shoot in 14-bit with a speedlight, the pre-flash to flash exposure is very delayed. BTW, this

occurs with any compression setting in RAW. This delay has prevented sports shooters from using

14-bit, as it throws off exposure timing, as well as limiting you to 2.5 FPS.

 

Mind you, the D300's images in 12-bit are beautiful, so I've left it in 12-bit. Has anyone else

been disappointed with the 14-bit performance, and does anyone have any info as to what's

actually taking place with the shutter delay?

 

Thanks, guys...

 

Greg

[tonybeach]

http://forums.dpreview.com/forums/read.asp?forum=1039&message=26245613

 

"As 14-bit was a selling point of the camera, this is semi-usable feature."

 

> You should be pretty aware of the limitation of 14 bits for shooting sports on a D300 based on your own observation that using it limits you to 2.5 fps.

 

I am myself investigating its usefulness for non-sporting applications and may have more to say about this if I can detect any advantages. Based on this post: http://forums.dpreview.com/forums/read.asp?forum=1021&message=26325741 there may be an advantage I have not yet been able to extract from the files. Based on Julia Borg's experience with the D3, my thinking is that the primary advantage of 14 bits is experienced at base ISO and as noise increases by using higher ISOs, any advantages diminish as greater accuracy merely makes a more precise measurement of noise.

[mjt]

(Anthony) ... i'm not too sure they understand the difference between what 12-bit and 14-bit gives you - their argument is that you get more DR going from 12 to 14-bit. i tried, fruitlessly, to get the point across. for example: http://forums.dpreview.com/forums/read.asp?forum=1021&message=26324831

 

but this thread is not about DR :) for whatever reason, Nikon has chosen to "throttle" the D300 when it's in 14-bit mode. folks aren't seeing any real degradation with the D3, between 12-bit and 14-bit. i suspect it's something in the firmware [of the D300].

[tonybeach]

MJ, Julia understands the concept of 14 bits versus 12 bits and the role it plays in DR perfectly well. It is a matter of accuracy in measurement. I don't recall the post where she said it, but it is no different than what you said and the analogy was a very good one when (to paraphrase her here) she said: If you measure the ceiling to the floor in one foot increments you will have a less accurate measure of the distance than you will if you measure it in inches. So if you only measure down to the last full foot you will have less distance than if you measure down to the last full inch.

 

Regarding Nikon "throttling" the D300 and your suggestion that this is firmware related -- I am inclined to agree with what Thom Hogan said about that. Specifically, the D3 and the D300 have different sensors and different ADC circuitry. Sony uses essentially the same sensor in the A700 but only handles the data in 12 bits. Nikon is moving the data off in 14 bits but has to give the sensor a chance to settle down prior to and after the exposure or it will be no different than just using 14 bits.

 

The question in this thread is (as I read it) "Has anyone else been disappointed with the 14-bit performance, and does anyone have any info as to what's actually taking place with the shutter delay?" I cannot say I have been disappointed with 14 bits, but my expectations were not high to begin with. In using it for its intended purpose, I am not bothered by the noticeably increased shutter lag. I also provided in my first link to Thom Hogan's post at DPR his current explanation for why this occurs; I have yet to read a better explanation.

[greg_morgan]

Thanks for the link, Anthony... looks like I'm not alone on this subject! Mike made a point I

was talking about regarding pre-flash/flash delay... it's very noticeable... your subject

might think the shot's taken and blink/move. Studio strobe, hard-wired, it just feels like

you're dragging the shutter. I haven't tested with PocketWizards yet...TBD. Just a theory,

but maybe the D300 is using an electronic shutter scheme in 14-bit, IE normal shutter

opens, electronic shutter activates in 14-bit, closes, then normal shutter closes... judges?

 

BTW, I'm not a sports shooter... I work at base iso's for location/environmental portraits,

but some sports guys have commented on the delay, say, for a volleyball spike, where

even the slightest delay will throw them off the peak of the action.

 

Thom's assertion that Nikon has been "up front" about this... where? Not in the brochure.

Not on Nikon's websites. I like Thom's real-world reviews, so I hope to see his in-depth

review, including the 14-bit vs 12 bit issue.

 

Thanks and looking forward to more comments/opinions.

 

Greg

[Brad_]

Dynamic range is set primarily by the sensor - not ADC resolution. You're probably not seeing much

difference because it's difficult for the eye to casually discern that level (14 bits) of tonal resolution.

Now, the number of ADC bits does drive best-case signal to noise ratio, SNR, (which if low can

restrict/mask sensor DR) - and that is further tempered with the notion of ENOB, or effective number of

bits. As an example, you may have an ADC with 14 data lines that wiggle, but its effective number of

bits based on real SNR performance will (always) be less - could be 12 1/2 equivalent bits. Also, there

are several ways of expressing dynamic range that can be influenced/limited by ADC selection - the top

end is clear, but one can choose different bottom specs.

[tonybeach]

"I was talking about regarding pre-flash/flash delay... it's very noticeable..."

 

> It is noticeable, as noticeable as what I experienced with my D70 cameras. In fact, I just tested the D300 in 14 bit side by side with the D70 and they are indistinguishable. I would guess that we are talking about a shutter lag on the D300 of around 100 milliseconds using 14 bits versus 45 milliseconds using 12 bits.

 

Just a theory, but maybe the D300 is using an electronic shutter scheme in 14-bit"

 

> There are two reasons this theory is unfounded. First, there would be blooming issues. Second, it would be published and reported by the websites like Image Resource and DPR. Check page two of the specifications on the cameras at DPR and you will see for instance that the D70 has a "Combined mechanical and CCD electronic shutter" while the D300 has an "Electronically-controlled vertical-travel focal plane shutter".

 

"...some sports guys have commented on the delay, say, for a volleyball spike, where even the slightest delay will throw them off the peak of the action."

 

> Naturally for those applications the camera should be used in 12 bits. Honestly, is anyone seeing a difference between 12 bits and 14 bits? I just spotted a difference in a pair of files taken today, when viewed at 500% you could actually see it, at 200% it was much harder for me to meaningfully distinguish the files-- I'm still working on identifying any advantages in DR, but I can see where detail might be separated from noise at lower EVs if the file is properly handled -- for me this is still a theoretical advantage though.

 

"Thom's assertion that Nikon has been 'up front' about this... where?"

 

> Thom said "Nikon has been pretty up-front that 14-bit comes with some limitations." You are taking that out of context when you substitute "this" for what he actually said. Nikon publishes that 14 bit only works at 2.5 fps. It is mentioned in the D300 specifications page DPR published nearly two months before the camera was released.

 

No, Nikon does not mention that shutter lag increases using 14 bits; but lets put this in perspective. At 6.5 fps the shutter actuates every 154 milliseconds, you have 45 milliseconds of shutter lag followed by 100 milliseconds of mirror black out, which leaves 10 milliseconds between shots; at 2.5 fps the shutter actuates every 400 milliseconds, and it appears that there is about 100 milliseconds of shutter lag followed by 100 milliseconds of mirror black out, which leaves about 200 milliseconds between shots.

[greg_morgan]

Thanks for the responses...

 

Anthony - That's right about the D 70 comparison... in 12-bit the D300 is very fast for

flash photography.

 

12 vs 14- bit quality - I can't say I see a definite difference in an average file, but I've seen

some large prints (fashion) side by side, done at 12 and 14 bit, D300, and the 14-bit

might have had a little better shadow detail, skin tone a little smoother...

 

Sports shooters I've talked to have expressed their interest in higher resolution shooting,

that's why I brought it up... you've got guys shooting skateboarding with film Hasselblads

with a 30mm fisheye! Not to go off-topic...

 

My original post was regarding shutter delay, not FPS. I was aware of the 2.5 FPS limitation

but my concern is with the delay in single shot mode. The D300 is a great DSLR, so having

this limitation in 14-bit, as others in the DPR post have mentioned, is strange, indeed.

 

Cheers,

 

Greg

[t_n1]

Check out this thread over at nikonians....<br/>

<br/>

<a href='http://www.nikonians.org/cgi-bin/dcforum/dcboard.cgi?az=show_thread&om=18129&forum=DCForumID71&omm=0&viewmode=threaded'>Thread at Nikonians.org</a>

[._._z]

"My original post was regarding shutter delay, not FPS."

 

Also several discussions on this topic at dpreview.com

[tonybeach]

Hi Morgan,

 

"My original post was regarding shutter delay, not FPS."

 

> I understand that, but you can't imply that Thom Hogan is saying that Nikon has been open specifically about the shutter lag since he was referring to the general limitations of 14 bit mode.

 

"...having this limitation in 14-bit...is strange, indeed."

 

> It is unexpected, but explainable -- I think Thom Hogan has presented a possible explanation. Perhaps we should be upset about the Live View Mode's implementation if that turns out to be the culprit. Would you be willing to give up that feature in exchange for no shutter lag in 14 bit mode? For me it wouldn't matter, I would not miss Live View (I haven't used it yet) and this shutter lag issue does not bother me.

[greg_morgan]

Hi all...

 

" Would you be willing to give up that feature (Live View) in exchange for no shutter lag in

14 bit

mode?"

 

In a heartbeat... I wouldn't miss Live View either, although I'm not sure that's the culprit.

 

The chart on the Nikonians thread is interesting, as it shows at ISO 100 that 12 and 14 bit

DR about the same, but I'm sure there will be 'real world' tests posted soon enough.

 

Re: Thom Hogan... yes, he was probably referring to the general limitations of 14-bit, but

this was a statement made in the DPR thread about the shutter delay.. nothing implied or

out of context. Anyways... Thom's knowledge of all things Nikon is without question...

if anybody can discover what causes the shutter delay, he will. Great no B.S. reviews,

recommendations, and opinions.

 

Thanks, guys...

 

Greg

[t_n1]

"The chart on the Nikonians thread is interesting, as it shows at ISO 100 that 12 and 14 bit DR about the same, but I'm sure there will be 'real world' tests posted soon enough."

 

Were you looking at D3 or D300? For the D300, that's the only iso that showed any significant difference between 12 and 14 bit.

[greg_morgan]

TN - My bad... I was looking at D3 vs D300... so the gap between both cameras is fairly

close in 14-bit at ISO 100... and there is more DR at 14-bit..(8.9 vs 8.6? ) So if you can get

close to a D3 at ISO 100 in 14-bit, all the more reason to solve the shutter delay issue... if

it's possible.

 

FYI, I tested the D300 in the studio today with Pocket Wizards in 14-bit mode, with studio

strobes, and they trigger at the end of the shutter delay, not the beginning, as was the

case with hard-wired strobes via sync socket. Not so good for fleeting expressions in

people/portrait/fashion... in 12-bit mode, it's instantaneous.

 

Thanks,

 

Greg

[tonybeach]

Hi Greg,

 

This "problem" with the D300 is not going to be solved. It is not a firmware issue and is an inherent property of the D300's hardware. If you want a Nikon camera that shoots at the same speed with the same shutter lag at 14 bits as it does at 12 bits, you have to bite the bullet and buy a D3.

 

Frankly, you are spending a lot of time niggling over 1/3 of stop of DR against 50 milliseconds of shutter lag. Nikon has delivered a capable camera for a reasonable price; and I have to wonder how many complaints there would be if it only delivered 12 bit RAW files.