The one thing that dissapoints me with the 1DX is the buffer size and the number of RAW frames possible to store during a high burst.
The 35 frames takes only 3seconds, 3 !!!!! to fill the buffer..
Anyway, there seem to be something strange going on, since shooting m-RAW mode actually decreases the number of possible burts
shots to 26, while shooting s-RAW mode increases the number to 39. I understand it takes some computing time to resize the RAW, but this will also happen while downsizine to s-RAW. Hence it is strange that the number of frames possible during m-RAW is "only" 26,
which is a ridicilously small number which you will reach within 1.5sec !! Why could not Canon install a 3 x bigger buffer when the rest
of the camera is built for speed?? (shutter/cpu). The D4 can shoot up to 3 times more frames than the 1DX.
I have tested the 1DX with the fastest memory card currently available on the planet (Lexar 128G 1000x Pro UDMA7 150/150MBps),
and the buffer was still full after 35frames continious shooting.
Too bad Canon!!
That seems to conflict with this report:
When I read about the 12fps performance of this camera I was excited – especially given the larger image size – but I was also worried as my 10fps performance of the 1D Mark III & IV has always been significantly limited by its puny buffer. In my early testing with what I was told was a slow CF card, I was able to get 52 full-size RAW frames in burst mode before the buffer started to stutter. That’s up from 30 in the 1D Mark IV, so that’s promising given the significantly larger file sizes. However, I would have really loved to have seen that number closer to 100 for RAW. RAM is pretty cheap, so I’m always frustrated there isn’t a way to add RAM or do something to overcome this limitation. With that gripe aside, the 52 RAW frames is going to be usable enough in most practical scenarios so this is a welcome relief. This also means that sRAW or JPEG only shooters will find themselves with an endless supply of buffer for sports shooting.
Indeed, and check Ole Salomonsen comments back then in that thread!!! (which he (Ron) called BS!! )
I re-read those comments to get a refresh on that, but dug further to some reliable sources. When you tested the 1Dx, where and when did you get to test it?
Here's what I'm finding odd on your comments:
1) the 1Dx manual states in RAW the burst is 35 images with a regular card and 38 with a UDMA 7 card. If the data xfer rate of 167MB/sec (max) for UMDA7 is supported, as stated in the specifications for the 1Dx, then I would expect that Ron's numbers of approx 50-55 RAW images to be about right. Your statements would have you believe that zero data made it out of the Lexar UDMA7 card, something not possible.
2) I know from experience with my 5D3, I can shoot approx 19-20 RAW images at 6 fps on a SanDisk Extreme Pro 32GB UDMA6 card before the buffer stutters. The 5D3 manual states it's RAW burst is 13 images, 18 when using UDMA7. I have not used a UDMA7 card in the 5D3 yet but would expect a little better performance than I have observed.
3) The D4 numbers you claim are highly questionable given reliable testing. For example, they totally conflict with Rob Galbraith's documented testing. That testing shows in a 30 second period the D4 using a Lexar 1000x 32GB UDMA7 was able to write 86 RAW+Jpeg images, with a max RAW data rate of 61.4MB/s. In the same testing with the same card on the 5D3, it could write 75 RAW+Jpeg images with a max RAW data rate of 80.2MB/s .
So, given documented testing by a known method and source of information that is also well know for his like of Nikon, the 5D3 is out performing the RAW xfer rate of the D4 by about 19MB/sec on the same card. I seriously doubt the 1Dx would be slower than the 5D3. In combined RAW+Jpeg, the D4 manages just 11 images more than the 5D3 in 30 seconds of shooting despite the 5D3's RAW+L-Jpeg rated burst capacity of 7 images. And let's not forget the 5D3 is a 22Mpix image....the D4 is 16Mpix, about 30% smaller.
This does not agree with your claims of the 1Dx or the D4's image rates and burst capacity. The D4 is barely outperforming RAW+Jpeg compared to the 5D3. So, can you tell us when you tested the 1Dx, where and when did you get to test it with the Lexar 1000x card?
Further, if anyone is wondering why M-RAWand S-RAW numbers seem odd, it because M-RAW is a 10M image which requires more processing (size reduction to a non-multiple of the RAW and processing to handle that). S-RAW you will note is exactly 1/4 of the original RAW (4.5M vs. 18M), so the math is far simpler, thus the greater burst ability.
Additional hands on that do not agree with the D4 statements by the poster:
1) Ron's numbers are wrong, sadly, check my proof video I posted in this thread. I have used the fastest memory card available, Lexar 128GB CF UDMA7 Pro, which I use daily with my 5D3. The 5D3 Canon also claim to be able to fully utilize any UDMA7 card with up to 167MB/s, which I have found to be wrong.
The 5D3 is capable of almost 100MB/s, about the same as the D4. (tested with speeds which allowed for continous shooting without the buffer stopping)
2) I am capable of shooting 33-34 RAW images at 6fps on my 5D3 with the Lexar UDMA7 card before buffer stops.
3) Seem like I am not entirely correct on the D4, (although I have read several places it could go higher), this guy is achieving 6-7 seconds continous shooting until buffer is full (at 05:45), this should mean 60-70 RAW files:XQD + Nikon D4 full speed test
I have also done extensive speed testing on my own with the 5D3, 5D2, the D4, D800 and D800E, all with the Lexar 128GB UDMA7 CF Pro 1000x, since I work a lot with stop motion video in the dark, and need the cameras to be able to shoot as fast as possible without the buffer stopping. The 5D3 has about the same write speed as the D4. IMO the D4 is actually slightly disappointing in its computing power, taking into account its fast shutter mechanics. Still its buffer seem to be larger than the 1DX, since RAW files of D4 is actually 33MB compared to 28 on the 1DX.
With the price tag, and the shutter speed and the computing power availbale on the 1DX, I have to say I it is disappointing Canon did not implement a bigger buffer, given also how cheap flash ram is these days, even if it might be enough for many people.
For my type of work, I am guessing the 1DX will still outperform the D4 (shooting as fast as possible without the buffer interfering), but thats not the only point. Even if my type of work is not what most people do, I also do regular work, such as wildlife and bird photography (BIF), and what If I am shooting an eagle diving down for a fish? This is a moment which could take much more than 3 seconds, sometimes 7-8seconds. But then the buffer on the 1DX is full..
And if I have to turn down the shooting speed of the 1DX to not fill up the buffer, then what is the point of me spending 7000 dollars on the 1DX rather than just using my 5D3 ?