Go Back   Steve's Digicams Forums > Digital Cameras (Point and Shoot) > Panasonic / Leica

Reply
 
Thread Tools Search this Thread
Old Jul 28, 2004, 2:46 AM   #1
Senior Member
 
playlong's Avatar
 
Join Date: Jul 2004
Posts: 237
Default


Jerry SD tests



I just conducted a very good test of the FZ10 with two SD cards. The file size of each image was 1.59 for the Panasonic 16mb card and 1.61 for the SanDisk I (not the Ultra). The pictures were taken on a tripod using five picture burst mode, with identical lighting. Because I had to shut off the camera to remove, the card different zoom settings might account for the difference in file sizes. The Panasonic took 7.5 seconds to record 7.95mb or 1.06mbs. The SanDisk I took 6.5 seconds to record 8.05mb or 1.24mbs. Since the files were smaller then 2MB the differences should be more demonstrable with a larger file size. The five write acknowledgement overhead has more statistical impact with smaller files.



In early setup of the experiment, I used another 256 card with the Dane Electric label. This card is rated at 10mb and took 13 seconds to write 7.25mb, or .56mbs.



What is apparent now is The FZ10 is not capable of writing at speeds requiring the emotional purchase of 10mb cards. However, it also demonstrates that various manufactures are flooding the market with unsubstantiated claims of performance, no quality control and a wide performance differences within their own product lines.



What I am alluding to is corporate greed. The temptation of selling a 30-cent item, regardless of capacity for up to $190 is too great. We the public is at the mercy of these manufactures. The only thing we can do to protect ourselves is to buy two identical cards from the same vendor with a return policy for non-performance.



When I get more time I will try to retime the shot with images greater then 3mb. I believe Panasonic's Fz20 may utilize a memory buffer and an increased write speed to overcome the card problem. We shall see.

playlong is offline   Reply With Quote
Sponsored Links
Old Jul 28, 2004, 3:08 AM   #2
Senior Member
 
Join Date: Jul 2004
Posts: 124
Default

It would be good to see some more testing of the difference between standard and so called "hi speed" or "ultra" SD cards when used in an FZ10.

No point wasting money on the hi speed cards if the FZ10 cannot write any faster.

And don't say "but what about the long term investment", because you know hi speed cards will be much cheaper in future.
lawrencew is offline   Reply With Quote
Old Jul 28, 2004, 3:51 AM   #3
Senior Member
 
Klaas's Avatar
 
Join Date: Jun 2004
Posts: 164
Default

It takes my FZ10 2 seconds to process and write a 5 shot burst to a Sandisk Ultra II SD card. I think the high speed cards are well worth the extra money (payd 60 euro for my 256MB Sandisk card).
Klaas is offline   Reply With Quote
Old Jul 28, 2004, 3:52 AM   #4
Senior Member
 
playlong's Avatar
 
Join Date: Jul 2004
Posts: 237
Default


It is really an uncontrollable mess out there. The Sandisk I tested was supposed to be an under 2mb card. The Ultra II is supposed to be 10mb. Maybe it is just a label change. The Panasonic card I used for my test was the one supplied with the camera. Would a purchased Panasonic perform better? I have heard people returning the"better" Ultra II, Simpletech and Lexar because of poor performance.

t is really an uncontrollable mess out there. The Sandisk I tested was supposed to be an under 2mb card. The Ultra II is supposed to be 10mb. Maybe it is just a label change. The Panasonic card I used for my test was the one supplied with the camera. Would a purchased Panasonic perform better? I have heard people returning the Ultra II, Simpletech and Lexar because of poor performance.
playlong is offline   Reply With Quote
Old Jul 28, 2004, 8:38 AM   #5
Senior Member
 
Panasonic's Avatar
 
Join Date: Sep 2003
Posts: 321
Default

If you search the archives I believe this topic was discussed. There are several factors that may skew your numbers.



First, digital still cameras are designed around a 2 MB transfer rate. Purchasing the faster cards is generally not going to offer any advantage. An exception to this rule can occur if you use inexpensive SD cards that are inherently slower than the 2 MB standard. There are also 3 specs; read, write, and copy.



Faster cards are a huge advantage with products that can record real time MPEG2 video, such as the SV-AV100 in our D-snap line. It's designed around a 10MB transfer rate. Larger cards will increase to 20 MB



In your description it didn't mention if the two cards you used were the same size. Generally, as the card's capacity increases the speed decreases slightly. Panasonic broke that trend when we introduced our 256MB card. It was the first to offer 10 MB despite the larger size.



Early <smaller> SD cards use a serial interface Larger and faster cards now also offer a parallel interface. So, to make use of the faster rate the device must use a parallel interface.



The speed can also vary by the amount of free space on a card. As it fills, the time may decrease slightly. This varies from manufacturer to manufacturer.


And finally, any such tests must be done using identical files. An SD card is similar to a hard drive where files are written using memory arranged in clusters. I don't have the exact numbers handy but let me try to illustrate it using a simplified explanation. Let's say each cluster is 4K. If you store a file that's 16K, it occupies 4 clusters. If you store a file that is 17.5K it now consumes 5 clusters. The "left over" is wasted and the time to read, write or otherwise access the file is increased by about 20%. Burst mode will give you a general indication as to how fast you can write data to a card but depending on the variables mentioned above the actual numbers will vary and will not exceed 2MB. Hope you find this useful.
Panasonic is offline   Reply With Quote
Old Jul 28, 2004, 11:24 AM   #6
Member
 
Join Date: Jun 2004
Posts: 49
Default

I do not know if available in the US, but the best performing SD card I have is made by Verbatim. It is only rated at 2MB/s but is consistanly out performs my SanDisk Ultra II and Panasonic cards(all 256MB)in my FZ10.I have owned several Verbatim memory cards in both CF and SD format, they are thesame price as the NON-high speed cards but offer a lifetime warranty and always perform as well as thehigh-speed cards. I havenever tested in a pro DSLR where the high speed cards will probably shine. As Panasonic Bob stated in an earlier thread some cards do not have the same speed for read and write and they only advertise the one which is fastest. I seems Verbatim's cards have the same speed rating no matter what.
DunlopKR is offline   Reply With Quote
Old Jul 28, 2004, 11:33 AM   #7
Member
 
Join Date: Jun 2004
Posts: 49
Default

Just to give you some numbers regarding my pervious post

InH burst mode onFZ10 for 5 frames with 256MD SD cards:

Verbatim ~2s, SanDisk Ultra II ~2s, Panasonic ~3s

In H burst mode on a friends Nikon 5400 for 7 frames with 256MB CF cards:

Verbatim ~5s, SanDisk Ultra II ~4s, SanDisk regular ~10s, Kingston ~9s
DunlopKR is offline   Reply With Quote
Old Jul 28, 2004, 12:08 PM   #8
Senior Member
 
playlong's Avatar
 
Join Date: Jul 2004
Posts: 237
Default

I never expected to hear from Panasonic since the FZ10 is not the issue. Although your thread clouds the postulate of my test, I will answer your comments.



You said:
First, digital still cameras are designed around a 2 MB transfer rate. Purchasing the faster cards is generally not going to offer any advantage. An exception to this rule can occur if you use inexpensive SD cards that are inherently slower than the 2 MB standard. There are also 3 specs; read, write, and copy.

I said:


Since the files were smaller then 2MB the differences should be more demonstrable with a larger file size. The five write acknowledgement overhead has more statistical impact with smaller files.
What is apparent now is The FZ10 is not capable of writing at speeds requiring the emotional purchase of 10mb cards.



You said:
Faster cards are a huge advantage with products that can record real time MPEG2 video, such as the SV-AV100 in our D-snap line. It's designed around a 10MB transfer rate. Larger cards will increase to 20 MB

I am saying:


There must be some mistake we discussing the FZ10 not the SV-AV100.

You said:

In your description it didn't mention if the two cards you used were the same size. Generally, as the card's capacity increases the speed decreases slightly. Panasonic broke that trend when we introduced our 256MB card. It was the first to offer 10 MB despite the larger size.



I said: Panasonic 16mb card. I failed to mention that the SanDisk was 256mb.



I am saying:

To follow your thought process the Panasonic card should have been the clear winner

You said:



Early <smaller> SD cards use a serial interface Larger and faster cards now offer a parallel interface. So, to make use of the faster rate the device must use a parallel interface.



The interface from the FZ10 to the card uses a ninepin serial interface. What you should have said was Matsushita Electric (Panasonic), SanDisk and Toshiba, from the outset, were in control of the card speeds by how many data lines on the "serial device" were activated. One data line 2.5, two data lines 5.0, four data lines 10.0 and this was a method of mid cycle performance bumps. The specification called for a handshake between the female port and the card to determine the amount of data lines in use.

You said:


The speed can also vary by the amount of free space on a card. As it fills, the time may decrease slightly. This varies from manufacturer to manufacturer.

I am saying:


How can a random access device without moving parts be substantially degraded by data location? There is no average access, time seek, or rotational delay. The only delay in writing to the card could be caused by data fragmentation due to deleted files; access to the FAT for alternate locations is not one of motion and therefore is inconsequential. Keeping the FAT in FZ10 memory would improve on slow read speeds but I suspect would produce negligible gains since there are no moving parts. The tests were conducted on formatted cards so there was no fragmentation.

You said:

And finally, any such tests must be done using identical files. An SD card is similar to a hard drive where files are written using memory arranged in clusters. I don't have the exact numbers handy but let me try to illustrate it using a simplified explanation. Let's say each cluster is 4K. If you store a file that's 16K, it occupies 4 clusters. If you store a file that is 17.5K it now consumes 5 clusters. The "left over" is wasted and the time to read, write or otherwise access the file is increased by about 20%. Burst mode will give you a general indication as to how fast you can write data to a card but depending on the variables mentioned above the actual numbers will vary and will not exceed 2MB. Hope you find this useful.



I said:

The pictures were taken on a tripod using five picture burst mode, with identical lighting. Because I had to shut off the camera to remove, the card different zoom settings might account for the difference in file sizes.



I am saying:

Yes, I have a complete understanding of cluster sizes and it really does not matter what the manufacture implemented. Cluster size is a trade off between buffer sizes; storage waste, data read/ write acknowledgements and rewrite speeds. If a manufacturer wanted to slow down his card below the one data line specification of 2.5MBS, they could implement a tiny cluster size. How devious, since the lowest card specification should work fine in the FZ10.

Yes, any such test should be conducted with the same size file but in the grand scheme of things the file sizes and results between the two manufactures inconsequential. Yes, maybe if I were using a three place decimal recording it would show. If you want to claim Panasonic the winner, please do. However, the low-level ScanDisk competed very favorably against the 10mb Panasonic card.



The one thing you did not challenge me on is how I arrived at the timings



The main point of my test was to show a) The FZ10 writes at 2mbs and faster then 2MB media performance is a waste of money. b) Various manufactures are flooding the market with unsubstantiated claims of performance, no quality control and a wide performance differences within their own product lines.



The only thing we can do to protect ourselves is to buy two identical cards from the same vendor with a return policy for non-performance. People please do not go out and buy SanDisk unless the vendor has a return policy. The forums indicate wide discrepancies in product performance with the regular and Ultra II cards.



While I have your attention, I have some questions about the FZ20. Did Panasonic solve this card / performance problem with a large temporary memory cache, increased the write speed and write size using the faster processor?



Could you see if you or someone else can answer my thread called "Viewfinder delay question for Panasonic Bob."



Thank you



Jerry
playlong is offline   Reply With Quote
Old Jul 28, 2004, 3:11 PM   #9
Senior Member
 
Klaas's Avatar
 
Join Date: Jun 2004
Posts: 164
Default

Panasonic wrote:
Quote:
First, digital still cameras are designed around a 2 MB transfer rate. Purchasing the faster cards is generally not going to offer any advantage. An exception to this rule can occur if you use inexpensive SD cards that are inherently slower than the 2 MB standard. There are also 3 specs; read, write, and copy.

Bob, the FZ10 can write at a much higher speed than 2MB/sec, it takes 2 seconds to write a 5 shot burst to the fast SD cards, 10 seconds to the slower types.

So I think the fast cards are well worth the extra money.
Klaas is offline   Reply With Quote
Old Jul 28, 2004, 4:25 PM   #10
Senior Member
 
playlong's Avatar
 
Join Date: Jul 2004
Posts: 237
Default

Panasonic must have produced 4mb write FZ20s and slipped them into the market place in FZ10 skins. You people are lucky to have these special cameras. It takes 2 seconds to capture five burst mode images into memory therefore leaving you with zero seconds to write to the SD card.
playlong is offline   Reply With Quote
 
Reply


Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off



All times are GMT -5. The time now is 10:51 PM.