GH5 SD Cards That Work at 400 Mbps...

T

TheDingo

Veteran
While we wait for better availability of ADATA, Kingston, Panasonic, SanDisk, Sony V90 cards, I thought we should start a list of SD cards that support 400 Mbps continuous shooting.

I know that lots of cards will work for 5-20 seconds, but which non-Vxx cards are good for 15 minutes or more of 400 Mbps shooting ?

On the V90 front, I am thinking of trying one of these with a SD adapter...

ADATA_V90_MicroSD.jpg
 
Lexar 128GB Professional 1000x UHS-II SDXC Memory Card .jpg

There have been reports of Lexar 1000x 150 MB/s SDXC U3 UHS-II cards working with the new GH5 400 Mbps profiles.

I own six of these 128 GB cards and used the free ATTO Storage benchmark to test my cards with a Delkin UHS-II card reader.

One of my cards did pretty bad on this test...

Lexar Worst.jpg

Here's the average result for four of my cards...

Lexar Average.jpg

And one of my cards ( the "miracle" card ) had a write speed that was almost twice as fast as everything else...

Lexar Best.jpg

Next I tested all six cards in my GH5 recording at 400 Mbps, and only the "miracle" card worked with no problems.

So if you get lucky and can find the right batch of Lexar 1000x cards they will work with the new GH5 400 Mbps profiles, but keep in mind that you might get stuck with cards that only work at 150 Mbps. :nads:
 
Hi

TheDingo said:
View attachment 126899

There have been reports of Lexar 1000x 150 MB/s SDXC U3 UHS-II cards working with the new GH5 400 Mbps profiles.

I own six of these 128 GB cards and used the free ATTO Storage benchmark to test my cards with a Delkin UHS-II card reader.

One of my cards did pretty bad on this test...

View attachment 126900

Here's the average result for four of my cards...

View attachment 126901

And one of my cards ( the "miracle" card ) had a write speed that was almost twice as fast as everything else...

View attachment 126902

Next I tested all six cards in my GH5 recording at 400 Mbps, and only the "miracle" card worked with no problems.

So if you get lucky and can find the right batch of Lexar 1000x cards they will work with the new GH5 400 Mbps profiles, but keep in mind that you might get stuck with cards that only work at 150 Mbps. :nads:
Click to expand...

Memory cards are a commodity item, with companies like Lexar just buying the cheapest NAND memory on the day, so there is no consistency, except for the speed stickers claiming all sorts of course. NAND memory gets binned for various speed and quality as they come off the production line, and it will be luck of the draw what we may get. If demand is low and yields are high we get lucky with good performing flash cards, sometimes not so. Usually it's all been good enough or speed variations hidden as slower cards would still be fast enough, but now we have cameras wanting constant write speeds of 400Mbits/sec, the lack of consistency and any concern at meeting a minimum quality is showing up.

Hence why we have V ratings for guaranteed minimum speeds now, and why V60/90 cards are few and far between and more expensive, because they need the best quality NAND memory and yields will be lower. Because the best NAND will now be taken for V60/90 rated cards, cards not rated V60/90 will probably continue for some time to soak up the lesser quality stuff. This means as time goes on cards that don't show the V60/90 rating, are going to be less likely to work even though we might have similar cards that do now.

It should also be noted that benchmark software may be misleading when testing V60/90 cards, as they will not be testing the V60/90 part of the SD specification, as this requires hardware that supports the new protocol and benchmark software to enable it. The likelihood is bench mark tests may be slower than we would expect a V90 card to be, or slower than an equivalent non V90 card that is much cheaper, yet the cheaper card is then found not to work. This is because a V60/90 indicates the card supports a new recording protocol, which is different to the protocol that a computer/memory card reader or cameras before use. V60/90 isn't just about the raw speed, it indicates the card supports other features required in order to ensure real-time recording happens without pauses needed by the card.

Regards

Phil
 
Phil_D said:
It should also be noted that benchmark software may be misleading when testing V60/90 cards, as they will not be testing the V60/90 part
Click to expand...

Yes, I've long known that the performance of memory cards ( and devices like USB keys ) is a bit of crap-shoot because it all depends on the performance of the actual NAND parts coming off the factory line. I posted my experiences with the LEXAR 1000x cards so that people don't assume that these cards are guaranteed to work at 400 Mbps, because it's a bit of a gamble if they will or not. ( in my case only 1 out of 6 cards worked at 400 Mbps )

I am hoping that some friendly programmer will write a minimum write speed test for memory cards, that will test every bit of memory space and then report back the slowest writing speed. It will be a slow and boring test to run, but the results could be very useful for video production.
 
Using the Lexar Professional 1800x MicroSDXC w/ UHS-II SD Adapter here with no problems so far.

image.png


http://www.lexar.com/products/memory-cards/microsd/Lexar-Professional-1800x-microSDHC-microSDXC-UHS-II-cards.html
 
View attachment 126882

The $289 US ADATA 256GB MicroSD UHS-II U3 V90 memory card I ordered from NewEgg arrived this morning, so I put it through it's paces.

Here's the ATTO Benchmark score...

ADATA 256 GB MicroSD V90.jpg

This is the best benchmark I've seen from any card I own. 164 MB/Sec Write speed, and 241 MB/Sec Read speed.

In my GH5 camera it can record for 83 minutes at 400 Mbps ALL-I.

It's a bit weird using a MicroSD card instead of a full size SDXC, but it's the only V90 card I could order in Canada that offered 256GB for $299 Canadian. ( from NewEgg.ca )

The full size SD card version of this card is supposed to have a Write speed of 200+ MB/sec, but this little MicroSD card will more than do the trick for me. :love4:
 
I just tested two Lexar 2000X 64GB cards with GH5 set to 4096x2160,23.98,422 10 bit,all-I, 400Mbs. Each card took 20 minutes of video which filled the card. It switched from slot 1 to slot 2 for 40 minutes of recording. These cards were purchased about two months ago but I don't know which revision they were.
 
I expected trouble, but my Lexar 1000x's are working. Crftsho on youtube, a group doing a lot of GH5 testing also confirmed this, saying the beta version did not work and the release does (on 1000x Lexar cardS).
 
Kellar42 said:
Crftsho on youtube, a group doing a lot of GH5 testing also confirmed this, saying the beta version did not work and the release does (on 1000x Lexar cardS).
Click to expand...

Based on my own testing LEXAR 1000x performance is completely dependent on the batch your card came from. Only one out of my six LEXAR 1000x cards works at 400 Mbps, and this also shows when tested with the free ATTO drive benchmark app. LEXAR 1000x cards are a gamble when it comes to 400 Mbps recording.
 
I'm still trying really hard to determine whether 400Mbps has any advantages over 150 IPB in ways other than being less processor intensive to edit and render. Anyone have any insight?
 
jtflesher said:
I'm still trying really hard to determine whether 400Mbps has any advantages over 150 IPB in ways other than being less processor intensive to edit and render. Anyone have any insight?
Click to expand...

I think this is going to take some time to fully explore. My guess is that 400 Mbps will improve panning movement, which can be a bit "juddery" when shooting with IPB formats. ( also want to check if the h.265 CODEC is better when panning )
 
You must log in or register to reply here.
Back
Top