madwand
-
Posts
405 -
Joined
-
Last visited
Content Type
Profiles
Forums
Blogs
Events
Downloads
Gallery
Store
Posts posted by madwand
-
-
<p>I've owned and tried a number of FD 50 f/1.8, 1.4 and SSC versions, and consider the f/2 version to be as good and sometimes better than the others -- based on informal subjective judgment. The f/2 version is the smallest, lightest, least sensitive to flare, and also the cheapest. IMO it's worth keeping and actually using, although I wouldn't pay much money for one.</p>
-
<p>Look into the Benro Travel Angle series. E.g. the TRA-169. There are also carbon-fiber versions. It's very compact, relatively light, and expands a decent height for the compactness. The included head is cheap, but the alternatives tend to be bigger and heavier. You can use other heads of course. The legs are functional and decent for the price/size/weight.</p>
-
<p>I suggest considering the EF-S 15-85 instead as an all-in-one solution with greater utility and quality than the existing kit lens. It gives a FF-equivalent field of view of 24-135 which is very nice, better than the 24-105 (on a FF) I last took on travel, which served the majority of my shots while the 70-200 f/4 rarely made it out of the hotel. In this view, which is of course personal and not entirely universal, a long lens is often more of a liability than a necessity in travel, and one's better served by a better and wider standard lens, with the long lens as an after-thought if that, optimized for portability (e.g. the EF-S 55-250 would be preferred as it's more portable and cheaper).</p>
-
<p>The 17, with its bulging exposed element and no support for hoods, has the obvious consequence of being very susceptible to flare -- both general contrast reduction and floating colored balls. You would notice this in the field, especially in summer. A tripod and a manual shade can deal with much of it, but the 24 is much better in this regard.</p>
<p>The 17's acclaimed optics are also best when the shift / tilt is not used, negating part of the appeal of going ultra-wide and then "correcting" for it. Besides, you can't correct every dimension.</p>
<p>I'd suggest the 17 only if you really benefit from the greater field of view -- the 24 will be better than a cropped 17, is a nice focal length on a full frame, and has other advantages.</p>
-
Apologies if my URLs are broken...<br><br>
<br><br>
<a href="http://www.baltimoresun.com/business/bal-bz.petition07oct07,0,1275184.story?coll=bal-business-headlines">L.A. Times staff fights cuts</a><br><br>
<i>"Tribune, struggling to lift its shares after four straight quarters of declining profit, has cut 200 Times news jobs in the past five years and ousted Johnson, who had publicly opposed further staff reductions"</i><br><br>
-
I had the same itch some time ago, and bought a 28 f/2.8 to complement my 17-40, and did some detailed comparisons. I found the 28 to be slightly brighter, noisier, less reliable in focusing, and at best equal to the 17-40 -- disappointing overall, and so I returned the 28.
My tests were done mostly if not entirely at close range and considered sharpness and contrast. I didn't notice distortion to be significant for my needs. I used a 20D, semi-live monitoring, an angle-finder for manual and assisted focusing, tripod, checked for focusing accuracy, etc..
Since then, I've also noticed that my 17-40's auto-focus at wide angles isn't always satisfactory (although it's difficult to notice this using just the little viewfinder).
I'm afraid the 28 f/2.8 is no Distagon 21... but it might be worth your trouble to do a test for yourself with option to return; that way you'll get results you can personally rely upon, and perhaps your findings will be more positive than mine. But if I did this again, I'd probably try the 24 first.
-
I believe that the case for 64-bit performance in Photoshop CS2 has been overstated in this thread, but 64-bit is a 100% correct way to exceed the 2-3 GiB memory limit of Win32, which is of utility to some photographers / Photoshop users. That support from Adobe is a positive indication here.
Driver performance is generally irrelevant here, because it's not going to make scanning faster.
64-bit may or may not make money. I agree that the consumer need isn't really there, but I don't know what marketing might achieve. It certainly helps the limited number of applications with such memory needs that it's widely supported, and of course should be supported by those devices which serve such applications.
I believe, on customer service grounds, that it's more correct for Nikon to provide this support than not to, however they can do what they want. I believe it's incorrect for photographers to take the negative in this matter, though of course some are doing so in order to try to educate.
-
<i><q>Yaron Kidron Photo.net Patron Prolific Poster, may 31, 2006; 09:28 p.m.
Intel is not entirely clear on what its 64 bit hardware is destined to be. They are still making a lot of money from their highly optimized 32 bit hardware: the fact is in sales: 64 bit doesn't make a dent in their sales spreadsheets. AMD has a good solution, but still, it clearly doesn't bring in enough interest for Nikon to port (and optimize) their drivers.</q></i><br><br>
Pardon me, but this seems like misinformation which is perhaps based on a confusion with Itanium. All current Xeon, Pentium D, Pentium 6xx, many Pentium 5xx, and many Celerons CPU's have and all Conroe, Mermon, and Woodcrest CPU's reportedly will have support for EM64T, which is essentially the same technology as AMD64. The fact that Intel has provided such broad support for a technology pioneered by a nemesis and in conflict with its own 64-bit Itanium speaks volumes about that technology. A notable but insignificant exception is in mobile processors -- none of the current ones with Intel branding support EM64T. But this doesn't really matter and moreover it's said that this will change with the release of Mermon.<br><br>
If all these processors don't make a dent Intel's sales spreadsheets, I'd ignore those spreadsheets for this topic. If all these CPU's don't convince you that Intel is behind this technology, then I can't imagine anything that will.<br><br>
Here are some Intel docs on the subject: <br><br>
<a href="http://www.intel.com/business/bss/products/server/64-bit_tipping_point.pdf?iid=ibe_64bit+rl_64bitadv">The 64-bit Tipping Point</a><br>
And a wiki:
<a href="http://en.wikipedia.org/wiki/EM64T">http://en.wikipedia.org/wiki/EM64T</a><br>
-
<i><q>Yaron Kidron Photo.net Patron Prolific Poster, may 31, 2006; 09:28 p.m.
There's no doubt the future is 64bit, but it's not that time yet. I'd rather wait for someone (be it Apple if not MSFT) to come out with a solid 64 bit solution, that would be widely supported by 3rd party developers; or, you can wait for Microsoft to make film scanners...</q></i><br><br>
It's not going to get much more "solid" than it is now, for numerous reasons. It is a solid solution at present, with excellent backwards capability, except of course for the kernel mode driver issue, which MS has put the onus on HD vendors to solve, which they can, with a little effort and expense.<br><br>
The arguments against the current 64-bit solutions IMO pretty much place those adherents into the back of the bus in terms of adoption timeframe. I'd say we're at the tail end of early adoption, approaching mainstream, and still a long ways away, if we ever get there, from the almost mandatory and climb on board or find yourself limited phase. For some people, including some above posters, the latter's pretty much what's needed for their adoption. I'm not trying to force the issue on them; I've got a number of 32-bit only processors myself, and would promote greater flexibility.<br><br>
However, to throw darts at the earlier adopters trying to make a go of it for stated reasons, IMO shows their perspective and attitude towards adoption much more than it indicates the state of the art. 64-bits is here, now, and has been for some time already. Talking about mainstream adoption as if that's what it takes for you to benefit from it merely indicates a follower mentality IMO, whether given by a consumer or a producer.
-
64-bit OS's aren't for the lemmings just now, and they won't be for lemmings for some time. They're of value to photographers specifically who use large amounts of memory, as is the OP. (BTW, 64-bit XP and 2003 also get beyond the 2 TB drive storage limit.)
64-bit XP, sharing history with 2003, is actually pretty stable. The early adopter days of it ain't ready so we won't make drivers for it are long gone. It's never going to be a dominant desktop OS, as evidenced by the naysayers here already, and for the simple facts that it doesn't have much to offer non-high-end users, and suffers driver availability problems.
Linux support would be awesome simply because it would get more source code in the field, which could facilitate the availability of third-party Windows drivers. So, I'd have no objections there. (And those drivers should compile for 64 bit CPU's too -- as many of us are already running 64-bit Linux.)
Nikon is possibly hurting too much in the film arena to bother with new expenditure on scanners. A second machine which scans to TIFF files is probably a good workaround in the meanwhile; possibly for the lifetime of the scanner.
I would however find it funny if next year people buy/install the 64-bit version of Vista in droves; call up their HW vendors asking for compatible drivers, telling them that they paid for 64 bits and want to use every one of them.
-
6 GiB of RAM is said by some to be a magic number for Photoshop CS2 running on a 64-bit platform (64-bit CPU, 64-bit OS). However, that's not exactly cheap -- you're into the server/workstation multi-physical CPU configuration, and paying through the nose for that rarified market.
Under consumer hardware, 3-4 GiB will be the upper limit. /3GB in the boot.ini will allow CS2 under supporting OS's to use as much as 3 GiB. Otherwise, the max that a single 32-bit application can use under a 32-bit OS is 2 GiB, while the rest of the system, OS and file cache in particular can use some of the remaining RAM.
There are issues with some motherboards not supporting > 2 GiB RAM well, reduced timing needed to exceed 2 GiB, and not seeing the entire 4 GiB when you have 4 GiB installed. If you're brave, and know that you have such memory needs, of course you can pursue this further, and also consider a 64-bit OS for doing this better. Otherwise, 2 GiB is probably a sensible limit.
-
<i><q>Steven Clark Prolific Poster, may 30, 2006; 02:56 p.m.
As far as I know you don't even need to use the 64-bit mode to write a windows 64 driver. You just need to NOT use the 16-bit mode. So drivers written for XP 64 and Vista 64 could be tested on XP 32.</q></i><br><br>
No, you have to have 64-bit drivers.<br><br>
<a href="http://www.devx.com/amd/Article/20342">http://www.devx.com/amd/Article/20342</a><br><br>
-
It's probably around the same amount of effort to make XP-64 drivers as Vista-64 drivers, and manufactuers would be smarter to beta test their 64-bit drivers now rather than a year from now with Vista GA and face that onslaught. It may even be less effort to develop under XP-64, because Vista ain't ready.
I think this is a sign of general scanner death as much as anything else here, and also a potential sign that Vista 64 will have the same issues in the future, perhaps marginally reduced, but likely not.
Why would issues with Vista-64 be reduced? It shares history and code with XP-64 and 2003. It's not as if 64-bit will somehow magically become more powerful and useful when it comes with a Vista sticker. The main differences will be marketing and Microsoft push, and a greater potential for new sales. I can't see how the market will be so inclined to sell native 64-bit support, and provide that at the driver level in addition to 32-bit support, but marketing doesn't bear a strong relationship with logic.
Epson provides 64-bit drivers for their scanners.
This might also show that this is a symptom of life and death in the field -- Minolta's dead, Nikon scanning is possibly dying, Epson's still kicking.
Another alternative here would be to see if any third-party scanner software vendors take this opportunity for new sales. Silverfast hasn't so far.
-
It's generally been hard to determine the panel type, but you can sometimes make some headway (until they switch the panel underneath you, as Dell sometimes does...). Viewsonic Europe for example gives such details, and you can find 8-bit panels at 20" widescreen and among the "pro" series. This may be a factor -- that 19" panels have become more mainstream, so for higher-end image quality, you might have to go to 20" or larger panels.<br><br>
Here's another article on the subject:<br><br>
Note however that the troubled Dell 20" widescreens are said to have 8-bit S-IPS panels, and they're said to have recently fixed the widely-reported banding issue, which was apparently due to Faroudja video processing.<br><br>
-
<i><b>Allen Herbert wrote:</b></i><br><br>
<i><q>In their controlled tests they came to the conclusion that the image quality of the 10D was superior to the 20dD and 30D. They felt that the noise control software was detrimental to the image and preferred it without it.</q></i><br><br>
This is plausible, as the original concern with the release of the 20D was that it would suffer in noise performance due to the smaller wells. However, it was shown that the noise performance was good -- so perhaps the next thought was that it was due to a sacrifice in image quality. Yes, this flies in the face of general opinion, and for that reason we should test it more broadly and not accept a single journalist's report, regardless of reputation.<br><br>
In their service of the public good / their readers did the UK magazine compare the 10D performance to a 300D? Did they compare the 300D performance to a 20D/30D?<br><br>
Did they give any indication of how they determined that this was due to the noise reduction? Did they indicate how we could do that ourselves?<br><br>
(BTW, I've heard a nasty rumour that some UK journalists sensationalize in order to sell copy. Perish the thought!)
-
Certainly one of the great bargains -- doesn't command the retail price premium of the metal bodied cameras.
I've even liked the plastic kit 50 f/2.0 lens better than f/1.8 and even SSC f/1.4 samples I've come across. Can't say that about the FD f/1.4 though -- that seems better.
-
To clarify a couple of points in my above post:
I used RAW + JPEG large fine in the recent tests. The original tests were done with just RAW and black images (lens cap on) for consistency and reproducibility. To more closely match other tests above, I just used random non-black images for today's measurements.
I also just timed the erasing of 230 or so mostly RAW + a handful of RAW + JPEG large fine images in camera, and that took around 75s.
-
<i><q>Well, I've got the same 4GB Seagate Microdrive (ST1) in Seattle (from Frys) and I noticed the same issue with my 20D:<br><br>
- start-up time 5...15 seconds (depending on the number of files on the card) - writing one RAW + JPEG takes about 40...50 seconds - writing a burst of 6 RAW + JPEG photos takes 4...5 minutes - formatting takes ~15 minutes - reviewing photos is very slow and a painful process<br><br>
I formatted the CF on a PC using 32KB cluster size: no visible effect in my Canon 20D. </q></i><br><br>
Something's definitely wrong. In my Oct. 24/2004 post of Seagate MD performance in a 20D, linked above, I measured 40s to write 20 RAW frames.<br><br>
My old Seagate MD was probably pulled off some audio device and bought off eBay (I've forgotten the details). It's a 5GB ST1.<br><br>
Doing tests matching the above, I get:<br><br>
Startup (with >100 images on card): near instantaneous.<br><br>
Single image RAW + JPEG fine write time: 6s.<br><br>
Burst of 6 RAW + JPEG fine write time: 30s.<br><br>
Formatting -- not tested, formatted externally using FAT32, 32KiB sectors.<br><br>
Image browsing (RAW) -- about 1 per second, a little sluggish, but very usable; faster when flipping through multiple images. <br><br>
In your place, I'd send the card back to Fry's and move on -- my almost 2-year old 5GB Seagate that was not intended to be used in a camera, and was not expected by Canon, performs much better, and my even older Hitachi performs even better. BTW, I have one of the first production 20D's and am using the 2.02 firmware at present.
-
<i><q>I might get spanked for this but I'm going to go non-P.C. and voice what others are thinking......so here goes. Who cares about what lens goes on a 20D? Someone who REALLY cares about image quality would find a way to buy an EOS 1dsmk II, and mate it with some FINE glass.</q></i><br><br>
Of course, even the plebs can afford a 1Ds-II, the ones who really care already have 39 MP MF camears or better. You can see this in the fine print on Canon boxes -- <i><q>If you really care about resolution, please buy a Phase One."</q></i><br><br>
This line of reasoning is not only non-PC, it's irrational.<br><br>
-
Looks good; warrants comparison to the 17-40 f/4 L.
One potential issue comes to mind -- the Tamrom XR series tend to have a (yellow-ish) colour cast in comparison to the Canons -- it'd be helpful to know whether or not this lens also has this, which can be an issue when switching lenses. Presumably in two directions -- if you're switching between a lens with a colour cast or if you're matching another lens that you have with a similar colour cast. With digital, this wouldn't necessarily be a major issue, but it has been apparent with some other Tamron XR's, at least some time ago.
-
Current Dells (xx07) models have widely-reported banding issues, so should be avoided until this issue is fixed.<br><br>
The Viewsonic VX2025WM is a tolerable 20" widescreen with a 8-bit P-MVA panel which can typically be found for a bit more than $300 USD (sometimes less).<br><br>
The following reviewer prefers a Philips (uncalibrated):<br><br>
The "X" might stand for the backlight pattern which is visible in the dark:<br><br>
Good hunting!
-
If you connect the card to a card reader, then you should be able to see it as a regular drive in the OS. Then you can format it using the command line for example in XP. I'll assume drive "G" in my example -- change it to whatever drive your card maps to.<br><br>
You can check the format using chkdsk G: <br><Br>
Here's an example output; it indicates FAT32 32K sectors:<br><br>
The type of the file system is FAT32.<br>
Volume Serial Number is 6003-13E0<br>
Windows is verifying files and folders...<br>
0 percent completed. <br>
50 percent completed.. <br>
100 percent completed... <br>
File and folder verification is complete.<br>
Windows has checked the file system and found no problem.<br>
4,874,464 KB total disk space.<br>
32 KB in 1 folders.<br>
4,874,400 KB are available.<br>
<br>
32,768 bytes in each allocation unit.<br>
152,327 total allocation units on disk.<br>
152,325 allocation units available on disk.<br>
<br><br>
format /? gives the command-line help.<br>
<br>
To (quick) format FAT32 32K sectors, you can use e.g.:<br>
<br>
format G: FS:FAT32 /A:32K /Q<br>
<br>
Your drive may be defective. See here for a test method I previously used together with performance results. If your performance is much below that, you might have a bad card. You can also use such a test in stores to check/compare performance.<br><br>
<a href="http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=009t7K">Seagate vs. Hitachi microdrive performance in 20D</a>
-
The 4 GB Hitachi microdrives used to be the absolute best performance/price, and have far outperformed the standard Sandisks for example. The Seagates haven't been as good in my experience, but it's good to have an alternative, and they still have lots of capacity for the dollar.<br><br>
I've used a 4 GB Hitachi microdrive for years now; I've never had an issue with its speed, and its performance/price advantage was extraordinary originally. However the advantage has lessened over time with CF cards coming down in price greatly.<br><br>
Galbraith's site has a strange absence of the Seagate microdrives, which makes me suspicious that something's amiss with them. I've had problems with a 5GB one that disconnects during card-reader to computer transfers. It's also not as fast as the Hitachi; it typically sits in the bag as a backup/overflow card should the Hitachi be exceeded in a shoot. Here's a partial answer from Galbraith:<br><br>
<i><q>The Seagate ST1 5GB sustained a fall of about 2 inches onto a desk here, midway through testing. The drop proved fatal - the card's miniature hard drive will no longer spin up. Also harmed irreparably was our desire to track down a replacement ST1. While it could have been an unfortunate fluke that the card didn't survive its encounter with our desktop, it could also mean that Seagate's new CompactFlash offering isn't designed to withstand minor bumps and jolts unless it's nestled inside a portable music player (the CF card has been shipping as an embedded part of MP3 devices for several months, and it was a card pulled from a music player that we tested). Either way, the ST1 is off our To Do list for the foreseeable future.<br><br>
Update, January 19, 2005: After this article was first published, a Seagate representative got in touch to discuss the paragraph above. John Paulsen, Sr. Manager in Corporate Communications at the company, indicates that the version of the Seagate CompactFlash card we tested is intended for use in embedded applications only, and that the design of the upcoming retail version, to be known as the Seagate CompactFlash Photo Hard Drive, is different enough from the ST1 that we should rethink our decision to exclude it from further testing. As we have no particular axe to grind with Seagate, we're likely to have a look at the Seagate CompactFlash Photo Hard Drive when one comes available, in the hopes that it will in fact withstand slight rough handling better than our tester ST1 did.</q></i><br><br>
Formatting the Hitachi microdrives to 32K sector sizes can improve performance noticeably -- as shown in the Galbraith chart.<br><br>
The OP seems to indicate a high performance need, perhaps higher than what even the Hitachi's have been capable of -- the performance database could be used to identify top-end cards for such application; they'll be CF.<br><br>
-
For comparison, let's consider 35mm format using a 4000 dpi scanner. Scanning 24mm x 36mm film at 4000 dpi, I get around 21.4 MP. This is particularly interesting, because it's very close to the resolution of the Canon 20D if you consider just the center crop (a full frame 20D would be 21.0 MP).
You can try to argue that the film + 4000 dpi scanner somehow has greater resolution than a 4000 dpi sensor's center crop, but as a matter of experience to many people, this is not true. It's not true for me. The 100% detail of a 20D is more usable than the 100% detail of a 4000 dpi scan of 100 ISO film, due to the relative lack of noise / grain. It's also a matter of experience to many people that they get better results from their DSLR than 35mm film scanned at 4000 dpi.
However, full frame 35mm digital at the resolution of the 20D doesn't exist anywhere near affordably, so you could argue that scanned full frame 35mm film still has technical superiority to affordable DSLR's. I don't happen to believe it, but since my simple argument doesn't have the necessary detail or ability to extrapolate, I'll simply argue that this is an indication of things to come -- where full frame is someday as affordable and as high resolution as scanners+film these days.
I do not consider scanners to be the limiting factor at this point because they scan into the limitations of film itself. The lack of further advances in film itself in contrast to the ongoing well-funded development of digital sensors give the clear answer as to where the technical superiority lies if not affordably now, in the probable near future.
What about scanned MF film? Some would argue that it's technically superior to cropped DSLR's at least. I wouldn't disagree. However MF film is clearly in decline and is not relevant to the vast majority of users, and has issues of its own in terms of handling, etc.
I personally like working with film cameras, and would have continued doing so exclusively had I not seen superior results in digital small format. I think that the basic resolution argument holds, and that film has an advantage only when you enlarge the sensor well beyond the affordable digital sensor sizes. While this is a valid technique, and something to take advantage of at present, it's not a sustainable one, assuming that digital sensor resolution continues to improve while film doesn't.
I further argue that the vast majority of people can get acceptable performance in whatever technology they choose. The availability of digital doesn't somehow invalidate the film techniques that have been developed and used for so long.
I still say however, that film and its scanning days are numbered, because the resolution has been matched and exceeded at the small format, and the larger formats are irrelevant to the vast majority. I think 35mm is the most important format, and it's been lost. The digicam format is mostly irrelevant from a technical / resolution argument standpoint, because most of those users don't act in manners consistent with caring about resolution, or even if they do, then the large scale migration to digital in that format is a good enough answer.
I would love to see film make significant advances. I would love to see scanners do the same alongside digital sensors. I don't believe the latter will happen, or matter without the former, and I don't believe that the former will happen.
Starting out in FD, which lense?
in Canon FD Mount
Posted