Rob Galbraiths Rubbery figures.

Started Aug 24, 2012 | Discussions thread
Tom2572
Tom2572 Senior Member • Posts: 1,129
Re: Rob Galbraiths Rubbery figures.

thubleau7 wrote:

I'm not going to shoot you but the point just seems to be getting lost in translation.

My point is this, how come the A57 produces a high result comparable to the Canon 5DMK11 Mk111 Canon 1D over the 30 second burst and yet the low mark on the 20mb card is still the same as the 95mb/card.

But when you look at the Canon scores the high mark is no better than the A57 using the Pro cards 1000 and yet the low mark for say the 20mb/card is nearly half the result of the 20mb/card in the A57.

It just does not add up.

if the tests were accurately done and not just a calculation based on mb/sec then why is the A57 producing real world results for the 20/mb card and the 95mb/card which are the same allowing for 5 percent measurment error and yet the results in the Poweful canon group shows the 20mb/card is way down on speed compared to the top card?

To me it sticks out and says there is something not right with Robs results.....its almost like a scale test in Mythbusters.where they prove that you just can't take a 1/10th scale object and compare the results to a full size object...the calculations will not add up because results are flawed once you go under a certain size ratio.the same with these cards applying what a 1000 card can do and then doing some calculations or relying on a manufacturers web site to provide detail on what a 20mb card can do in camera without actually doing a test in camera..

One thing it looks like you may not have taken into account as a possible explanation is sample variation of the SD cards themselves. In the SD card, this industry makes something that wouldn't have been considered possible 20 years ago and shrunk it to the size it is for under $50 retail and lower in some cases, which further translates to a probable $2-$4 total production cost per unit made by the lowest bidder in China.

With those parameters, expecting products like this to have zero variation over millions of units produced is a lesson in futility. For experiments like this to be done properly would take testing multiple units from verified separate production runs and verified separate production facilities if they are produced as such. Since this is the Internet though and no one questions testing methods anymore, I always read "reviews" of any sort with a grain of salt even when they are done by well known and respected sources.

As a side note I think it's great what you're doing and don't let others discourage you with their playschool comments about getting a life. Doing something you enjoy and aren't getting paid for is the very definition of hobby no matter what anyone else thinks of it.

Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow