Velocity Reviews - Computer Hardware Reviews

Velocity Reviews > Newsgroups > Computing > Digital Photography > Does faster CF card make significant difference for D70s?

Reply
Thread Tools

Does faster CF card make significant difference for D70s?

 
 
Roger N. Clark (change username to rnclark)
Guest
Posts: n/a
 
      09-23-2006
ASAAR wrote:

> On Sat, 23 Sep 2006 13:22:59 -0600, Roger N. Clark (change username
> to rnclark) wrote:
>
>
>>If you go back and READ, you will see I did not take
>>sides. I simply stated my observations from the data
>>I had and that the answer was likely in between the two bounds.
>>
>>I would think the OP would see the simple differences
>>in published card speeds and calculate for the types of images
>>he makes, what the speed difference would be.
>>
>>My D70s test data show:
>>
>>iso200, uniformly lit wall, all pixels saturated giving one value:
>> nef file = 4,081,448 bytes (the jpeg is only 127 kbytes)
>>
>>iso200, uniformly lit wall unsaturated exposure: 5,496,832 bytes
>> (jpeg = 804 kbytes)
>>
>>iso1600 1/8000 second dark frame: 5,742,592 bytes
>> (jpeg = 779 kbytes) Note this frame looks black when displayed.
>>
>>The maximum uncompressed file slightly more than 9 mbytes.
>>
>>Normal images would likely fall between the the ~5.5 mbytes and 9 mbytes.
>>That is all I said. I made no implications as to which end I thought
>>images come out at.
>>
>>But again, you seem to feel a need to go off the deep end.
>>Perhaps if you approached people differently, you might get
>>better reactions.

>
>
> No, you're quite wrong. I didn't go off the deep end anywhere.
> It's you that resorted to the ranting that you once tried to accuse
> me of doing. You still don't have a clear understanding of what
> Bill and I were discussing. It was only about how long it would
> take for the D70s to write RAW image files to the flash card in the
> camera. Determining how large the images would be in the computer's
> memory when expanded is completely irrelevant and adds nothing to
> the discussion, other than to show that you have some experience
> performing tests and measuring things. Thank you for this time
> adding that these test shots included bundled jpeg images. When
> they're subtracted from the total NEF file size, the *all* are
> noticeably smaller than the 5.4MB size I got from dpreview. (Note:
> dpreview may have actually used 5,400kb which is smaller still, but
> still larger than your results).
>
> If getting a better reaction from you was of any concern to me I
> might ignore your irrelevant and misleading replies. But it's not.
> Your overreaction just makes you look bad and does little for your
> reputation. Why you think it's important to show that a compressed
> RAW file will be expanded in the computer's memory is a mystery that
> only you can explain. It may be an interesting factoid, but it's
> nothing that we didn't already know, and it did nothing to support
> the mistaken claim that the D70s would take 2 seconds to write its
> RAW files. You may say that you didn't intend to support the claim
> that way, but if you didn't, then what was the purpose of that bit
> of irrelevancy? Anyone reading this thread with an open mind can
> come to their own conclusions, and they probably already have.
> Maybe to stir the pot and hope that I'd overreact the way you have?
>


1) The file sizes I cited are those generated by the CAMERA, not
expanded in computer memory. The raw file sizes do have relevance to
write times. Where do you get the idea the values I posted
were expanded computer memory image sizes?
It should be obvious that a 6-megapixel camera would generate
about an 18 megabyte in computer RGB 8-bit/channel image, or 36 MByte
RGB 16-bit/channel image. Therefore statements like nef
file = x megabytes is not in computer memory file sizes,
nor is maximum raw file size ~ 9 megabytes.

2) the jpegs are converted from the raw files post processing, and not
related to the raw file size.

3) you are still going on about a mistaken claim that Bill has
already tried to clear up, yet you still want to beat it to death.
And I didn't support ANY claim; you jumped to that conclusion.
I simply provided information to bound the problem.

Amazing!

Roger
 
Reply With Quote
 
 
 
 
ASAAR
Guest
Posts: n/a
 
      09-23-2006
On Sat, 23 Sep 2006 15:49:24 -0600, Roger N. Clark (change username
to rnclark) wrote:

> 1) The file sizes I cited are those generated by the CAMERA, not
> expanded in computer memory. The raw file sizes do have relevance to
> write times. Where do you get the idea the values I posted
> were expanded computer memory image sizes?


Are you misunderstanding again? The RAW files that you cited
(about 5MB or so) were the sizes assumed to have been written to the
cards. Any expansion that I mentioned was based entirely on your
insistence on pointing to their 9MB file size:

> The maximum uncompressed file slightly more than 9 mbytes.
>
> Normal images would likely fall between the the ~5.5 mbytes and 9 mbytes.


Yes, you're correct that they would expand to a much greater
degree in memory. But you keep ignoring the fact that the D70's RAW
files are generally smaller than your stated lower bound. In fact,
as I pointed out, based on the sizes you cited, they ALL were
smaller than 5.5MB. Two of the three that you cited were smaller
even without subtracting the size of the embedded jpeg file.


> 2) the jpegs are converted from the raw files post processing, and not
> related to the raw file size.


Then it's your fault for not making that clear. Dpreview showed
file sizes for pictures taken only including RAW files and also gave
file sizes for RAW+JPEG. In trying to guess why your even your file
size for presumably the most highly compressible file (evenly lit
wall) would be larger than dpreview's typical file size for the D70,
I tried giving you the benefit of the doubt, that you had included
the size of an embedded jpeg file. You see what happens when you
refuse to answer questions or explain your atypical results? I've
asked several times why your highly compressed RAW file was so
small, and you've refused to speculate. Well, I can speculate, and
perhaps it's because your data isn't as accurate or reliable as
you'd have us believe. Otherwise why wouldn't you want to verify
your results (I don't have a D70 or D70s so I'm unable to do so) or
at least give a plausible reason for what appears to be an unusually
large file size. One possibility that comes to mind (although I
don't pretend that it is the correct answer) is that Nikon may
preallocate a minimum amount of memory for all RAW images. If you
can present a better theory I'd welcome hearing it, and I'm sure
that others would also.


> 3) you are still going on about a mistaken claim that Bill has
> already tried to clear up, yet you still want to beat it to death.
> And I didn't support ANY claim; you jumped to that conclusion.
> I simply provided information to bound the problem.


Uh, that *I* initially tried to clear up, and Bill has since done
so as well. I'd say that it's you that are trying to beat it to
death, by adding irrelevant calculations that have nothing to do
with an already "cleared up" bone of contention. The information
you provided was, as usual, not only irrelevant but incomplete,
saying nothing about a lower bound, and nothing about whether the
upper bound could ever be approached. If your second dark frame
only produced a 5,742,592 byte file (~5.48MB) when do you think
you'll ever find or produce a file that approaches your "maximum
uncompressed file slightly more than 9 mbytes"? Such red herrings
do nothing to further the discussion or help figure out real world
values for how quickly the D70 series cameras write RAW files.


> Normal images would likely fall between the the ~5.5 mbytes
> and 9 mbytes. That is all I said. I made no implications as to
> which end I thought images come out at.


Technically correct, but that's an example of what would make for
a good entry in the old classic "How To Lie With Statistics". It
would have been far more accurate to have said "Normal images would
likely fall between 4.0 mb and 6.0 mb", and it wouldn't have led
any to mistakenly assume that RAW NEF files as large as 7, 8 or 9MB
represent normal images. In case you care to argue this point, note
that by "normal" I mean "typical", not that a 7MB NEF file should
somehow be construed to be abnormal.


> Amazing!


Indeed, sir!

 
Reply With Quote
 
 
 
 
=?iso-8859-1?Q?Rita_=C4_Berkowitz?=
Guest
Posts: n/a
 
      09-24-2006
Roger N. Clark (change username to rnclark) wrote:

> Normal images would likely fall between the the ~5.5 mbytes and 9
> mbytes. That is all I said. I made no implications as to which end I
> thought images come out at.


It doesn't matter to ASAAR anyway; he's waiting for Seagate to release the
new U320 SCSI CF cards next year. These babies will be *FASTER* than fast!






Rita

 
Reply With Quote
 
ASAAR
Guest
Posts: n/a
 
      09-24-2006
On Sat, 23 Sep 2006 20:00:08 -0400, Rita Berkowitz wrote:

>> Normal images would likely fall between the the ~5.5 mbytes and 9
>> mbytes. That is all I said. I made no implications as to which end I
>> thought images come out at.

>
> It doesn't matter to ASAAR anyway; he's waiting for Seagate to release the
> new U320 SCSI CF cards next year. These babies will be *FASTER* than fast!


That would be of no interest to me. Unless that is, Nikon manages
to release next year either a D80sc or D200sc upgrade. Then things
get interesting. Only the much larger Nikon pro models will have the
ability to insert up to eight of those CF cards, which would allow
them to create a *faster* than *FASTER* than fast SCSI CF RAID array
inside the camera. But because of the camera's excessive size and
weight I wouldn't be tempted to get one of those beasts anyway! But
I got a breaking story and it didn't come from Drudge.

Manufacturers have slowly been phasing out CF cards. If you cut
open some of the newest 8GB CF cards, you'll see that they actually
contain a couple of SDHC cards. Olympus's new large CF cards have a
slightly lower maximum capacity for now, but that's because
internally they're using the same number of xD cards. But that will
soon change because a CF card can hold more xD than SD cards.

 
Reply With Quote
 
Roger N. Clark (change username to rnclark)
Guest
Posts: n/a
 
      09-24-2006
ASAAR wrote:
> On Sat, 23 Sep 2006 15:49:24 -0600, Roger N. Clark (change username
> to rnclark) wrote:
>
>
>>1) The file sizes I cited are those generated by the CAMERA, not
>>expanded in computer memory. The raw file sizes do have relevance to
>>write times. Where do you get the idea the values I posted
>>were expanded computer memory image sizes?

>
> Are you misunderstanding again? The RAW files that you cited
> (about 5MB or so) were the sizes assumed to have been written to the
> cards. Any expansion that I mentioned was based entirely on your
> insistence on pointing to their 9MB file size:
>
>>The maximum uncompressed file slightly more than 9 mbytes.
>>
>>Normal images would likely fall between the the ~5.5 mbytes and 9 mbytes.


That's it, take it out of context again. Go back and READ.
It is obvious I was citing the maximum size of an uncompressed
NEF file.

> Yes, you're correct that they would expand to a much greater
> degree in memory. But you keep ignoring the fact that the D70's RAW
> files are generally smaller than your stated lower bound. In fact,
> as I pointed out, based on the sizes you cited, they ALL were
> smaller than 5.5MB. Two of the three that you cited were smaller
> even without subtracting the size of the embedded jpeg file.


Try a google search for actual raw file sizes. For example,
Ken Rockwell cites sizes up to 6.2 mbytes in his D70 NEF files.
Other cameras I have more experience with have compressed raw files
that vary about a factor of 2 in size depending on ISO and
image content.

>>2) the jpegs are converted from the raw files post processing, and not
>>related to the raw file size.

>
> Then it's your fault for not making that clear. Dpreview showed
> file sizes for pictures taken only including RAW files and also gave
> file sizes for RAW+JPEG.


The reviews I've read all indicate the jpeg is a separate
file so there is no embedded jpeg.

> In trying to guess why your even your file
> size for presumably the most highly compressible file (evenly lit
> wall) would be larger than dpreview's typical file size for the D70,
> I tried giving you the benefit of the doubt, that you had included
> the size of an embedded jpeg file. You see what happens when you
> refuse to answer questions or explain your atypical results? I've
> asked several times why your highly compressed RAW file was so
> small, and you've refused to speculate. Well, I can speculate, and
> perhaps it's because your data isn't as accurate or reliable as
> you'd have us believe. Otherwise why wouldn't you want to verify
> your results (I don't have a D70 or D70s so I'm unable to do so) or
> at least give a plausible reason for what appears to be an unusually
> large file size. One possibility that comes to mind (although I
> don't pretend that it is the correct answer) is that Nikon may
> preallocate a minimum amount of memory for all RAW images. If you
> can present a better theory I'd welcome hearing it, and I'm sure
> that others would also.


Irrelevant. See above.

>>3) you are still going on about a mistaken claim that Bill has
>>already tried to clear up, yet you still want to beat it to death.
>>And I didn't support ANY claim; you jumped to that conclusion.
>>I simply provided information to bound the problem.

>
> Uh, that *I* initially tried to clear up, and Bill has since done
> so as well. I'd say that it's you that are trying to beat it to
> death, by adding irrelevant calculations that have nothing to do
> with an already "cleared up" bone of contention. The information
> you provided was, as usual, not only irrelevant but incomplete,
> saying nothing about a lower bound, and nothing about whether the
> upper bound could ever be approached. If your second dark frame
> only produced a 5,742,592 byte file (~5.48MB) when do you think
> you'll ever find or produce a file that approaches your "maximum
> uncompressed file slightly more than 9 mbytes"? Such red herrings
> do nothing to further the discussion or help figure out real world
> values for how quickly the D70 series cameras write RAW files.
>
>>Normal images would likely fall between the the ~5.5 mbytes
>>and 9 mbytes. That is all I said. I made no implications as to
>>which end I thought images come out at.

>
> Technically correct, but that's an example of what would make for
> a good entry in the old classic "How To Lie With Statistics". It
> would have been far more accurate to have said "Normal images would
> likely fall between 4.0 mb and 6.0 mb", and it wouldn't have led
> any to mistakenly assume that RAW NEF files as large as 7, 8 or 9MB
> represent normal images. In case you care to argue this point, note
> that by "normal" I mean "typical", not that a 7MB NEF file should
> somehow be construed to be abnormal.


Perhaps some others will post file NEF sizes for scenes
taken at ISO 200 to 1600, e.g. some with lots of detail,
like trees, leaves, or sandy beaches (where detail is
resolved down to the pixel). That would indicate
likely upper limits.

I would ask my friend with the D70s, but he is now in
Afghanistan.

Roger
 
Reply With Quote
 
ASAAR
Guest
Posts: n/a
 
      09-24-2006
On Sat, 23 Sep 2006 20:09:38 -0700, Roger N. Clark (change username
to rnclark) wrote:

>>>1) The file sizes I cited are those generated by the CAMERA, not
>>>expanded in computer memory. The raw file sizes do have relevance to
>>>write times. Where do you get the idea the values I posted
>>>were expanded computer memory image sizes?

>>
>> Are you misunderstanding again? The RAW files that you cited
>> (about 5MB or so) were the sizes assumed to have been written to the
>> cards. Any expansion that I mentioned was based entirely on your
>> insistence on pointing to their 9MB file size:
>>
>>>The maximum uncompressed file slightly more than 9 mbytes.
>>>
>>>Normal images would likely fall between the the ~5.5 mbytes and 9 mbytes.

>
> That's it, take it out of context again. Go back and READ.
> It is obvious I was citing the maximum size of an uncompressed
> NEF file.


Very little that you say is obvious. Take the above, for
instance. I could be wrong, but it's my impression that there is no
switch on the D70 that lets the user choose between writing
compressed and uncompressed NEF files. So the only NEF files that
the D70 or D70s writes would be compressed, and assuming a worst
case image, the file would not be able to be compressed very much
and might approach 9MB in size. If you were speaking of an
uncompressed NEF file, the question would be *why*?


> Try a google search for actual raw file sizes. For example,
> Ken Rockwell cites sizes up to 6.2 mbytes in his D70 NEF files.
> Other cameras I have more experience with have compressed raw files
> that vary about a factor of 2 in size depending on ISO and
> image content.


Sure, but an extreme case is far from typical. You're going on,
and on, and on and on for practically no good reason. You still
don't get it. You're trying to finesse your way from the heart of
the discussion/disagreement, which was all about the rate that the
D70s could be expected to write RAW files. You're all about only
pointing out the "upper bound", which would be an unlikely
occurrence for most people. Perhaps you should take up your case
with Dpreview and other review web sites. After all, they presented
a typical RAW file size of about 5.4MB, and that is consistent with
the sizes you cited. Now you're doing everything possible to show
that in some cases the sizes could be much higher. So why not try
to have figures published that say something like

> In our tests comparing a wide range of scene types, we found that
> over 98% of the files were 5.4MB +/- 0.45MB.


If review web sites followed your lead, they might say this:

> In our tests comparing a wide range of scene types, we found that
> file sizes ranged from a low of 4.2MB to a high of 8.6MB


The latter is much less informative, so harping on an upper bound
isn't very helpful at all. But anyone reading this thread probably
has a good idea by now what's behind your fixation.


> The reviews I've read all indicate the jpeg is a separate
> file so there is no embedded jpeg.


That seems to be correct, and any ambiguity is removed for the D80
by dpreview, which states:

> *1 File size reported here is the size of the RAW and JPEG files
> added together.


But the RAW file size for the 10mp D80 is given as 8,600 KB. No
range, lower bound or upper bound. Probably because there's no
pressing need.


> Perhaps some others will post file NEF sizes for scenes
> taken at ISO 200 to 1600, e.g. some with lots of detail,
> like trees, leaves, or sandy beaches (where detail is
> resolved down to the pixel). That would indicate
> likely upper limits.


I'll leave that to you and any other measurebators that feel a
need to join in. It could provide some unexpected or interesting
results, but I have better things to measure if I were so inclined.
If any of your Canon cameras save uncompressed RAW files you could
test them. It would be better though if the same person or group
tested all of the camera makes so that they could be compared
fairly, with each brand taking pictures of the same subject.


> I would ask my friend with the D70s, but he is now in
> Afghanistan.


I wish him well and hope that he keeps his head low. Conditions
have been steadily deteriorating there over the last year or two.

 
Reply With Quote
 
=?iso-8859-1?Q?Rita_=C4_Berkowitz?=
Guest
Posts: n/a
 
      09-24-2006
ASAAR wrote:

>> It doesn't matter to ASAAR anyway; he's waiting for Seagate to
>> release the new U320 SCSI CF cards next year. These babies will be
>> *FASTER* than fast!

>
> That would be of no interest to me. Unless that is, Nikon manages
> to release next year either a D80sc or D200sc upgrade. Then things
> get interesting. Only the much larger Nikon pro models will have the
> ability to insert up to eight of those CF cards, which would allow
> them to create a *faster* than *FASTER* than fast SCSI CF RAID array
> inside the camera. But because of the camera's excessive size and
> weight I wouldn't be tempted to get one of those beasts anyway! But
> I got a breaking story and it didn't come from Drudge.


LOL!

> Manufacturers have slowly been phasing out CF cards. If you cut
> open some of the newest 8GB CF cards, you'll see that they actually
> contain a couple of SDHC cards. Olympus's new large CF cards have a
> slightly lower maximum capacity for now, but that's because
> internally they're using the same number of xD cards. But that will
> soon change because a CF card can hold more xD than SD cards.


Interesting! I never knew this. This is going to suck when the new D3x is
introduced since I have a load of CFs that I'm not will to toss just yet.
Hopefully Nikon will put both a CF and SD slot in their future models.







Rita


 
Reply With Quote
 
John Turco
Guest
Posts: n/a
 
      09-25-2006
"Roger N. Clark (change username to rnclark)" wrote:
>
> ASAAR wrote:


<edited, for brevity>

> > I don't know why you seem to be trying to resurrect Bill's bogus
> > file write rate, unless the D70s actually saves uncompressed RAW
> > files that are nearly twice as large as those saved by the D70. If
> > that's the case, then you and Bill would be correct. But I haven't
> > heard that the D70s made that change, and it would seem that if the
> > improved version of the D70 took twice as long to write RAW files,
> > more than one person would have complained long and loudly.
> >

>
> You accuse people of having reading comprehension problems; how ironic.
> Go back and look what I wrote. I computed the UPPER LIMIT to the
> file size. Then I gave real data for flat test scenes of
> 5.4 mbytes/raw file, so a lower limit for normally exposed scenes
> (dark frames are lower yet). Reality is likely in between flat test
> scenes and the upper limit. The D70s is what the OP has.
>
> Get a life, and stop foaming at the mouth!
> Over some simple math you turn this thread into a major attack on people.
> The mark of a troll.
>
> Roger



Hello, Roger:

I'm a bit shocked by your unchararcteristic outburst, above. If ASAAR
really is trolling, you're playing right into his hands.

As a distinguished scientist and mature adult, you must rise above the
crowd!


Cordially,
John Turco <(E-Mail Removed)>
 
Reply With Quote
 
 
 
Reply

Thread Tools

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 On
Pingbacks are On
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
most significant and less significant address a01lida VHDL 2 11-16-2008 12:28 PM
Significant performance difference of function in and outside of WebServices - Can't explain Greg Sohl ASP .Net Web Services 0 03-14-2007 09:44 PM
Significant Digits (Significant Figures) SMH Javascript 0 01-07-2007 09:52 AM
Q: how exactly is a 40x flash card "faster" than a 4x flash card? Paul Bennett Digital Photography 3 06-26-2004 09:32 PM
Which is faster - USB-enabled Lexar card with Jumpshot, or conventional USB2 card reader? Alan F Cross Digital Photography 4 07-15-2003 08:52 PM



Advertisments