D3S owners - does it REALLY get 9 FPS??

<p>The reason I ask is because neither my D300 nor my D700 get the advertised frame rates - even with the big battery.<br /> I like to shoot indoor sports, and there is a big difference between 5 and 7 and 9 FPS in things like a volleyball spike. I am very interested in the increased light sensitivity, but would like to hear from someone who has one and verified it does actually do 9 FPS.<br /> (the way I measured these other cameras was simple - just took a few bursts, and looked at the time on the images, and saw how many were captured per second).<br /> Many thanks.<br /> John</p>

<p>I'm not sure about the D300, D700 or D3-series, but the frame rate of my D2H is highly dependent on several factors. Those include shutter speed (so available light is a factor), autofocus options (including but not limited to release priority vs. focus priority), resolution and format (JPEG vs. NEF). While 7-8 fps is the maximum rated speed, actually achieving that is possible only when all settings and available light permit. It will certainly do that in bright daylight with the correct settings - I've shot sequences of wrecks at car races to confirm this.</p>

<p>Aside from meeting technical specs, I usually found that timing was more important than frame rates in capturing the moment of peak action. It'll still take movie camera frame rates of 16-24 fps to rely on sheer frame rates to reliably snag that perfect moment. I don't do much action photography these days but the instantaneous response of a single shot seems to work as well as the maximum frame rates for most of my people photography.</p>

<p>John, a more accurate measurement is to record the burst with your PC microphone using <a href="http://audacity.sourceforge.net/">Audacity</a> (Windows). You can then expand the time scale of the recording to millisecond precision.</p>

<p>I was certainly getting 9 frames/sec with the D3 (not D3S) when I captured this sequence two years ago: <a href="http://www.photo.net/photo/7073740">http://www.photo.net/photo/7073740</a></p>

<p>As Lex says, you need to make sure that the camera does not need to AF between frames. If you use focus priority, it could slow it down. So does a slow shutter speed .... :-)</p>

<p>John, my previous D300 and D3, and my current D700 achieve their advertised frame rates. With both D300 and D700 you have to have the battery grip and use either AAs or the EN-EL4(a) battery(s). Also, has to be in 12-bit mode, NR off (if I recall) and (darn, I forget what it is.) As Lex noted, there are certain settings that affect frame rate.</p>

<p>Thanks, Michael - I will try that.<br>

Thanks, Lex - of course, some of my best shots are single anticipated frames. That is particularly so in sports I know the best. In soccer for example, I most often take single exposures, but I spend more time watching soccer than I do almost anything else (including working, it seems). For volleyball, though, I have found it VERY hard to get a hand contacting a ball in a spike - (not that that is the only shot I want to get).<br>

The more sports I shoot, the more I realize you can't anticipate everything, and I force myself to keep the camera in front of my face AFTER the anticipated action is over - and that is when you get some really amazing shots, when unexpected things happen - which you miss entirely if you have brought the camera down to admire your work in the LCD screen. (Keep it pointed at the action, stupid is something I have said to myself WAY too many times. In these circumstances, rapid sequence shots often give you nice surprises.<br>

Curiously, if you take say 3 or 4 second bursts a few times, you notice it does vary even within a single burst at the same exposure factors by 1 or 2 FPS. And, not always getting slower. Not sure why. Getting slower as the battery peak charge diminishes would make sense, but it sometimes is a bit faster in the 2nd or 3rd second. ??<br>

Thanks - but has anyone actually measured their D3S yet?<br>

J</p>

<p>I have a loaner D3S at home. I'll check it tonight.</p>

<p>How could it be checked? I mean.... I really don't think anyone can count to 9 in one second! :) <br>

But if it is possible I would like to know how.... <br>

Good luck Shun!</p>

<p>I am going to use a stop watch to time my continuous shooting for 3 seconds and see whether I end up with about 27 frames. The D3S has a much deeper memory buffer so that it should be able to handle 30+ consecutive NEF frames.</p>

<p>Rene', see my post above. </p>

<p>Measurement is easy. You can grab a D3 shutter sound from the web: <a href="http://dc.watch.impress.co.jp/static/2007/09/25/d3.mp3"><MP3 Link></a></p>

<p>Open it in Audacity:<br>

<img src="../users/MichaelChang/ShutterSound/Capture_61.jpg" alt="" /></p>

<p>Then expand the timeline for more precision - one click, about 110mS, which is 9fps:<br>

<img src="../users/MichaelChang/ShutterSound/Capture_62.jpg" alt="" width="507" height="306" /></p>

 

<p>Heh! I've used Audacity for some interesting stuff but it never occurred to me to apply it to photography.</p>

<p>if you take a cue from some of the suggestions above -- like turning off NR and 14-bit color, putting focus to AF-S -- a camera like my D700 w/grip can fire off the advertised 8 fps. for the heck of it, i just now tried, and -- no real surprise here -- it easily cranked out the frames. i just flipped through images on the LCD, counting how many were shot in the span of one second. it didn't take 2 or 3 attempts, so i imagine you will get the same results if you DO try this at home. as for the D3s, i have little doubt it is capable of doing all that's claimed -- just not unconditionally.</p>

<p>Lex, there was once a forum link (which I can't find) to a slo-mo shutter release video (or animation) from (I believe) a Nikon. Sync it to Audacity and one can correlate time vs. event based on sounds it makes. </p>

<p>I tested the D3S sample as I described earlier. I used a 35mm/f1.4 AI-S so that no AF is involved. I do get approximately a 27-frame burst when I hold down the shutter release for 3 seconds, or as close to 3 seconds as I can.</p>

<p>I am using ISO 200 and 14-bit RAW capture with lossless compression. It takes about 30 seconds to write 27 RAW files onto a SanDisk Extreme 4 CF card. Therefore, it is a bit more than one second per frame.</p>

<p>The D3, D700, and D3S are not affected by 14-bit capture; they remain at 9 frames/second. However, the D300, D300S, and D3X will drop down to about 2 frames/second if you use 14 instead of 12-bit capture.</p>

<blockquote>

<p>The D3, D700, and D3S are not affected by 14-bit capture; they remain at 9 frames/second. However, the D300, D300S, and D3X will drop down to about 2 frames/second if you use 14 instead of 12-bit capture.</p>

 

</blockquote>

<p>Shun, thank you for clarifying. Seems I have been a source if mis-information lately. I was obviously remembering my D300's limitations and applying it to the D3 and D700.</p>

<p>This might be wrong, but I think the looking at the exif data might be wrong way to calculate fps. The cameras are limited by the buffer to a certain number of frames. The exif data could be reporting when the file was written to the card not actually captured. I guess you could test this by filling the buffer, 27 or so frames, let half the buffer dump to the CF card then snap another frame. Look at the timing between the 2nd to last and the last shot if there is no difference in time than between those two and the two previous, it is writing the time written, not time taken. I would test this but, my camera is only 3fps so the timing would be less obvious.</p>

<p>Each image file should have several time stamps in the EXIF data, one time stamp indicating the time the image is captured and another one for when the file is written. If you copy the image file from one hard drive to another, the latter time stamp can change.</p>

<p>BTW, officially, for 14-bit capture, the D300/D300S drop to 2.5 frames/second while the D3X drops to 1.8 frames/second. That is why I almost never set the D300 to 14-bit capture. There is no such issue with the 12MP FX models.</p>

<p>In my experience with my D300, the biggest hit to maximum capture rate that was left out above is Active D-Lighting. It may be that this does not impact the D3/D3s (!#$%%@$<shakes fist>), but it sure does slow down the D300.</p>

<p>Although having high framing rates is all well and good, when you let rip on a fast moving subject you are turning the chance you will get the best shot over to the camera, not a good thing. The key to getting the best photo is not blazing away at high rates, but a knowledge of the subject and a honed sense of timing. Several well timed successive shots is far better than just holding the shutter button down and being along for the ride. I used to shoot a lot of sports when I was in college. I used to set the framing rate to the highest (non-mirror lockup) speed (5 fps) on my Nikon F2 motors to accomplish one thing; to advance the film as quickly as possible so I would be ready for the next shot when it came. I cannot ever remember just holding the shutter button down and being along for the ride and hoping I got the best shot. Not to mention, with only 36 exposure rolls of film, I would be going through a ton of film and might very well <em>miss the best shot</em> while I was busy loading more film.</p>

<p>Granted, this was far more important for film cameras than digital, since now you have essentially unlimited frames available to you, but the concept of <em>you using the camera</em>, rather than <em>the camera using you</em> is still as applicable today as ever.</p>

<p>Poor man's Audacity: photograph a digital stopwatch running.</p>

<p>You'll need sufficient light of course to keep the shutter high, but you can get very easy and accurate results. If hundredths of a second aren't precise enough, shoot a burst of ten frames and divide the total time span by ten, etc.</p>

<p>Andy: thanks for the most obvious suggestion! Why wasn't that the first thing suggested? I'll bet everyone else feels pretty stupid now. I'm always astounded by the simple solutions that put to shame all of the complicated technology that people are thinking around in circles.</p>

<p>Is it true that the only way you can get rated FPS on a Canon is to shoot JPEGS, not RAW?</p>

<blockquote>

<p>Is it true that the only way you can get rated FPS on a Canon is to shoot JPEGS, not RAW?</p>

</blockquote>

<p>Canon makes excellent DSLRs. You wouldn't think they are that bad, would you?</p>

<blockquote>

<p>Is it true that the only way you can get rated FPS on a Canon is to shoot JPEGS, not RAW?</p>

</blockquote>

<p>Believe whatever you want, but it doesn't make it true. Don't get caught up in believing that all Canons work one way and all Nikons work another. Remember: each model camera is created by an individual engineering initiative. Each camera has different parts, a different sensor, different processor, different shutter, different layout, metering, autofocus, viewfinder, etc. Some cameras will share certain components with one or two others, but in general you will find differences among the different cameras that will make one model as different from another as a Canon from a Nikon.</p>

<p>For example, if you are frustrated by Canon because you have only shot a Digital Rebel, it would be rash to pass judgment on Canon as a whole, as the Rebels do NOT represent Canon's standard of excellence. Only their cheapest technology goes into the Rebels.</p>

<blockquote>

<p>Poor man's Audacity: photograph a digital stopwatch running.</p>

</blockquote>

<p>A stopwatch is way more expensive than Audacity ;)</p>