Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Sandisk 16gb 60MB/s for $115

Here's an article back in Sep showing many of these cards seeming to sustain over 70MBs (remember we only need around 40MBs)
http://www.robgalbraith.com/bins/content_page.asp?cid=7-10043-10255

3330_cf_reader_table.jpg
 
Tim Whitcomb said:
I thought there was mention of some "magic" or "Graeme's pixie dust" on the RED cards?
Click to expand...

Negative. Just that the real-world card speeds back in the day weren't good enough so they had to special order their own (for the 16gb that is). But that was back in the 300x days, when they were really slower than that. But now we're in 400x, 500x, 600x days, even if they are much slower than advertised we should be way in the clear.

Maybe we could get Red unlocked so as to use non-proprietary media? Maybe a special bypass mode acknowledging that I'm using non approved data? Since Red tells me if it dropped a frame, i'd be very content to do my own tests.
 
Those cards have already been tested. They don't work.

A good test of these cards is the AJA benchmark or similar sustained write benchmark that shows transaction performance on a graph. If the graph points go up and down rapidly like a racing heartbeat, then you're probably going to see dropped frames with the cards. There is a difference between sustaining 60MB/s and truly sustaining 60MB/s. If the card is really spitting out high speed bursts like 1MB in 1/100th of a second and then nothing the next 3/100s of a second, you just dropped a frame, even though the average speed may seem to be fast enough.

We're all still waiting on any word about the new Lexar UDMA 600X cards though.
 
Jeff Kilgroe said:
Those cards have already been tested. They don't work.

A good test of these cards is the AJA benchmark or similar sustained write benchmark that shows transaction performance on a graph. If the graph points go up and down rapidly like a racing heartbeat, then you're probably going to see dropped frames with the cards. There is a difference between sustaining 60MB/s and truly sustaining 60MB/s. If the card is really spitting out high speed bursts like 1MB in 1/100th of a second and then nothing the next 3/100s of a second, you just dropped a frame, even though the average speed may seem to be fast enough.

We're all still waiting on any word about the new Lexar UDMA 600X cards though.
Click to expand...


Whoa, that hand waive was so fast i'm still blinking.

Who tested these cards where? I'd like to read such a test. You're saying the Sandisk 600x 90MB/s are too slow??
 
I second Shawn's desire to see the tests.

Better yet I want to do the tests for myself. I'm a stickler for doing my own tests.

So, if you really want to "help" me, then stick a menu option in there that says "Enable unapproved media" and let me go.

Better yet, just remove any code that detects the media's manufacturer.

When it doesn't work... well I won't try that again now will I?

Now, for everyone's sanity I did a test of one of the 60MB/s Sandisk UDMA Extreme cards, using a firewire 800 reader and the AJA System Test utility.

My results? 33.7MB/S Read, 19.4MB/s write.

They won't work, not even for REDCODE 28. Not even close.

By way of comparison a 3.5" 5400RPM 16MB cache SATA drive I have hanging off the FW800 writes at 41.8MB/s and reads at 60.8MB/s. So that drive will meet some performance criteria for RED... but its not quite up to snuff for REDCODE 42. A new 1.5TB drive tests just a touch faster just barely qualifying for REDCODE 42, again over FW800.

That isn't close to saturating the FW800 bus. Its faster over SATA, but interestingly sustained read speeds drop.

Disk bandwidth remains a HUGE problem for RED and similar video/film applications. It requires cutting edge computer engineering.

Based on their past actions, I expect that if generic media meets the performance criteria for RED recording, RED will enable it, to the limit of the technology.

So... the new 90MB/s cards, if they truly offer 50% performance improvement over the 60MB/s cards should allow REDCODE 28 recording.

I can say that for a lot of projects a generic 32GB CF card for REDCODE 28 would be a blessing. About 19 minutes by my back of the envelope calcs. That's near Betacam and the like.

The next majore revision of CF cards should feature SATA instead of UDMA, and they should readily handle REDCODE 28 and 36 speeds, as well as being much larger capacity.

Interesting times.
 
Jarred said they tested them and the cards dropped frames. I've personally bought one of the 64GB 90MB/s to test for a couple other applications. First off it's incompatible with most anything I tried to plug it into. Second of all, it was "fast", at times, but the performance was very erratic. Due to compatibility issues and performance issues the card didn't work for my non-RED purposes either so I sent it back. The card also couldn't truly sustain anywhere close to 90MB/s for reads, let alone for writes, and it really struggled to keep any sustained rate on multiple over-writes of the entire card.

I would like it very much if Jarred, Deanan or someone else at RED could elaborate on their internal testing and criteria.

I have a handful of the 32GB SanDisk 60MB/s cards. I can concur with Alexander's test results. About 36MB/s reads here and 22MB/s writes using an ExpressCard CF reader. I don't know where or how SanDisk came up with their 60MB/s spec, but they're full of crap. These cards also perform erratically, but not as much as the 90MB/s ones. I was unable to perform a complete barrage of tests on the 90MB/s card due to it just not being compatible with anything but the Lexar and SanDisk readers. I have no idea who SanDisk is actually marketing this card to.
 
You are not going to want CF Cards in the near future unless they get WAY faster.

SSD is the future for handling more advanced Wavelet Compression
 
Shawn- You also seem to forget that the REASON the RED CF CARDS cost more is that they are the HIGHEST QUALITY CF CARDS on the planet. This has the MOST to do with the sustained DATA speeds and COST.

COST IS DIRECTLY RELATED TO QUALITY... and when you start working on budgets that cost THOUSANDS or TENS OF THOUSANDS OF DOLLARS A DAY... a couple hundred bucks means nothing when you consider the peace of mind of not using a lower grade, discounted card.

PORSCHE, ROLLS ROYCE, MERCEDES, i.e. YOU PAY FOR THE QUALITY....on the HIGH END MODELS...

so I for one dont care too much about COST, I just want more capacity and faster speeds... and THAT IS ALSO WHY I THINK SSD is the best bet in the future as REDCODE 100, 250 etc start coming into play.
 
Tim Whitcomb said:
Shawn- You also seem to forget that the REASON the RED CF CARDS cost more is that they are the HIGHEST QUALITY CF CARDS on the planet. This has the MOST to do with the sustained DATA speeds and COST.
Click to expand...

To be completely fair, 8GB Lexar cards work just as well as Red 8GB cards. But they are less than half the price.
 
Milan Spasic said:
To be completely fair, 8GB Lexar cards work just as well as Red 8GB cards. But they are less than half the price.
Click to expand...

and neither 8GB card will record Redcode 42 :P

These cards are rated for quality, just like all silicon based tech. The higher quality the card, the faster the read/write
and the MORE expensive they are.

Its a FACT. There should be some sort of agency governing these reported READ/WRITE times, because they are BOGUS>

None of those cards "sustain" those advertised speeds.
 
Martin Weiss said:
The only non-RED certified media is the Lexar 300x 8gb.
Click to expand...
nice to know.. now me already knows which flash memory to buy for my actual & 2010* cameras
 
Last edited:
You must log in or register to reply here.
Back
Top