Measuring Camera Speed Performance With A Microphone
Memory cards have their speed rating systems. For example, class 6 is the recommended base class for 1080p HD video coming from DSLRs. Those classes however, don't tell you what is the burst rate on individual shots. Mostly because each image has a different MB size to it depending on many factors.
Jaroslav over at Crazy Lab found an interesting way of measuring the burst rate and comparing different factors that affect the camera to card writing speed. TO make the test constant he covered the lens of his Canon T3i and took pictures of darkness. By recording the shutter sound (or music as some call it) and displaying the waveform in Audacity Jaroslav was able to compare burst-rates of different ISOs, capturing modes and cards.
There are three tests performed Each test has an initial part where the burst rate is almost identical until the camera's internal buffer gets filled. Then an image is taken only when the buffer has enough space after dumping data to the memory card.
Here are some of Jaroslav's interesting findings:
Higher ISOs create bigger. This means that the buffer is filled faster (less images) and it takes longer for each image to clear the buffer.
Raw files are bigger, but for some reason, the camera can take faster RAWs, than JPGs (I am not sure if this has to do with the memory card, though).
And the last comparison of SanDisk Extreme HD Video, Hama 150x, Samsung Essential MicroSDHC and SanDisk ultra II. Head over to Crazy lab to see that one.
[Measuring the performance of DSLR cameras and why the ISO setting could impact the burst length at crazylab.de]
[stopwatch by julianlimjl]
Get the DIYP greatness via RSS, newsletter and Twitter
Connect with the community: Facebook Page, Discussions
Share Ideas, Setups, Images and Projects on DIYP's Flickr, visit Readers Photos
Comments
Re: Measuring Camera Speed Performance With A Microphone
I suspect higher ISOs put out larger JPG files (and thus slow down burst mode) because of the noise and hot pixels that is recorded as part of high ISO photography. At low ISO's, you generally have little noise, so it is more likely that colors adjacent to a pixel are close or the same. So the image optimization within JPG will treat these as being the same, and it can encode this with a shorter sequence. While at high ISO with noise, every so often adjancent pixels will be completely diffent, and this breaks up the optimization.
The observation that RAW files are faster at burst mode but produce larger files is likely due to less processing needed on the file, since all that is done for most RAW files is encode the file, and possibly do a lossless compression. JPG files need to do all of the normal processing, white balance, sharpening, etc. that goes on, and this takes some amount of processing. Typically embedded CPUs are fairly wimpy, due to either cost constraints, or power constraints (the higher clocked a CPU is, the more power it generates, which drains batteries faster, and also produces more heat). With slower memory cards you might see JPG files write faster because the time it takes to write the SD/CF card becomes a larger part of the total time to process and write an image, and the RAW image's size will take longer.
"or music as some call it"
That is hilarious!!
I can't stop laughing :))
Post new comment