Jump to content


  • Content Count

  • Joined

  • Last visited

About scrambled

  • Rank

Contact Methods

  • Website URL

Profile Information

  • Location
    Birmingham, UK
  1. Congratulations to romania and bulgaria! Sure there will be short term and long term affects, but I'm sure everyone hopes that in the long term this will be a positive thing. Although there is no euro yet, I believe there are lots of other changes involved with becoming part of europe, things like being subject to european laws, ease of movement of people, jobs, goods etc across the borders, and huge confidence boost for europeans to invest within the new entrants, because they know there is far reduced beaurocracy between european countries. One effect of this has been europeans buying up 'bargain' property in the new entrants, I'm sure the house prices have ballooned since a timescale for entry into the EU was announced. Us europeans need 'lebensraum' as hitler put it. So I guess locals not being able to afford housing so easily could be a cost of entering the EU. But then I guess the idea is that it is only a matter of time before foreign multinationals move into the new entrants taking advantage of the job market, until eventually things start to 'equalize'.
  2. scrambled

    Virus Software

    First line of defence is a regular scan with spybot search & destroy, then adaware. If you get something and these aren't picking it up, I also recommend the free online scan at http://www.trendmicro.com . It takes ages but manages to catch all kinds of things that other scanners miss. 'HijackThis' is also another program that's very useful but only if you know what you are doing. But the single most important thing for avoiding malware is to install firefox. Also, once you have firefox installed, load up IE, and set security settings to maximum, close it, then never touch it again (unless you have to). Removing the shortcut and program list entry for internet explorer is also a good idea in case anyone else uses your machine. A lot of applications load IE on the sly (like msn, or some file sharing apps) so I haven't found a way to totally lock IE out of your system (thanks to such helpful design by microsoft, no wonder they got told off by the european court). However setting all the internet explorer security settings to maximum is the best I've found so far. Also I find a couple of firefox extensions handy - adblock plus and noscript, for determining which websites have permission to run stuff on your machine. As an aside Norton is imo, worse than useless, I've been called in to remove it from several people's machines ... it's more like scumware itself as it makes itself painful to remove, uses up your PCs resources and slows it down, AND doesn't provide any significant protection against malware. As a rule, most microsoft software is a security risk as they tend to include unnecessary 'features' that make it easy for hackers to take over your machine. If you do use things like word etc, be very aware of where your files are coming from, and maybe scan them before use. File types that are normally safe: jpg, gif, avi, txt, bmp, html, mpg, mp3, rtf, wav Some file types that should raise security fears: exe, wmv, asf, doc, dll, mov, qt, pif, pdf, scr
  3. 134 accepted, 3 rejected so around 98%. Had one rejected for overly simple vector, one was a joke pic, and much as I hate to admit it, I even had one rejected for noise or artefacts. So let me say when people complain of outrageous rejections for noise, I FEEL YOUR PAIN.
  4. I'd have thought you'd have needed a neutral third party as a witness ... that's the idea of a witness. I usually use a neighbour. IANAL, and you'd have to consult a lawyer for reliable advice, but this is my understanding: Think of it like this, a model release is a document showing that the model has consented to use of their image. A witness is there to provide a certain amount of protection against photographers forging releases (being a third party), and models making counter claims that a release is forged (when it is in fact genuine). A photographer signing as photographer AND witness is obviously useless for this purpose. A husband or relative isn't ideal either (although may be accepted, depends on the agency) .. the best witness is someone that would be hard to prove would collude with you, and someone of professional standing would be ideal, such as a doctor or solicitor, because the idea is they'd have a lot to lose by faking a release.
  5. On a CRT, imo the most important thing about refresh rates is that low ones will give you headaches (60-75 hz or so). If you can get 100 hz or higher you'll appreciate it. On an LCD, refresh rate is less important as the screen holds it's colour more over the refresh, so you don't get flicker (which is what causes the headaches). But these days, I'd really go for an LCD, it's a no brainer, CRTs are soooo last century (a lot of places don't even sell CRTs anymore). I think you can pick up decent 19 inch ones for around 200 squid, I have a 19 inch iiyama prolite, it's fantastic. Have a nose around on ebuyer.com and pcnextday.co.uk. Did I mention they are also much easier to spot noise and artefacts on lol .
  6. scrambled

    Is this noise?

    I can't see any noise .. but wouldn't say that this particular area of the photo is in focus (I'm assuming this is a 100% crop).
  7. Nope. Correct. Not in principle (providing camera manafacturers refrain from any hard baking of colour balance into their RAW files). Bear in mind however that setting e.g. a white card white balance can be easier if you do it at the shoot. In addition auto-white balance on most cameras may not be available in the same way at post processing time (although personally I only use auto white balance for snapshots now, it's so all over the place). Correct. Indeed, that's the whole idea behind RAW. Bottom line is, everyone shoots RAW, it's just a question of whether they prefer the conversion to happen on the camera, or on the PC afterwards.
  8. It's quite confusing, because some articles seem to quote a very high quantum efficiency for ccds: http://micro.magnet.fsu.edu/primer/digitalimaging/concepts/quantumefficiency.html And above they define quantum efficiency as: However susan's article suggests that several successful 'captures' of photons may be necessary for the registering of a unit of light on current ccds (if the new chip is really '2,000 times more sensitive to light'), not sure which articles to believe really (as you never know what is marketing hype). Maybe there is a stage after the quantum efficiency where further efficiency is lost, and this is where the new sensor improves things. Or perhaps the increased number of electrons reduces the relative effect of noise generated by the amplifier, or the relative effect of dark current. Here's another good read: http://www.pixcellent.com/CCDROLE5.htm
  9. Add my agreement to fuzzbass here. Interesting point from davidcrehner about increasing the base sensitivity of a sensor - however, ultimately the limit is physics, unless your talking star trek technology lol. As far as I remember from my physics lessons(!) and experiments with programming radiosity, photons are particles (with wave like properties), and there is a lot of randomness involved as to whether a photon will be emitted in any particular direction at any moment by a light source. As you move further from a light source the amount of photons hitting a given surface area (part of your lens) decreases. Once you get down to shadows and low light, you are dealing with a finite number of photons behaving in a stochastic fashion. Even with a 100% efficient conversion of each photon received, you are still going to have stochastic 'noise' unless you increase the exposure time, capture more photons and average it out (as I suggested in my post here: http://submit.shutterstock.com/forum/viewtopic.php?t=13683) Come to think of it I'm sure some bright physicist / mathematician could calculate some real figures here : and in that vein I've found a few links which suggests that ccds can be very efficient at detecting photons as fuzzbass suggests: http://www.andor-tech.com/library/digital_cameras/?app=327 I hope too that one day Nikon will click on this new-fangled 'particulate nature of light' idea, and make a full frame sensor. I can see it now in the offices at Nikon HQ 'so that's why people have been using medium format cameras all these years, doh!! suddenly it all makes sense' lol...
  10. muchliang: If you are able to post scaled down versions and 100% crops (a pixel for pixel crop of the original) of your images in here, we'll do our best to help you learn to identify any technical problems with your images, and how to avoid them. It's difficult at first I'll admit, but once your eyes know what to look for, it becomes quite easy to get past the issues of noise and artefacts. It's would be a very good idea to post your next 10 up for critique before you submit them, because usually other members will be able to pick up on any problems the reviewers might spot. Also (shameless plug) have a look at my faq hopefully it may help a little: http://www.geocities.com/scrambled_egg_on_toast/StockPhotographerBasics.html
  11. Agree this kind of thing seems to be the best way. After a bit of experimentation I now do neat image on a separate top layer (everything is affected), create a layer mask for this layer, invert the mask (nothing is affected), then paint into the mask areas where I want neat image to affect (just noise, nothing else). This has the benefit of being easily reversible for any area of the image.
  12. If you want to see how to really get rid of noise ... have a look at how the space telescope type things are made to work. http://en.wikipedia.org/wiki/Astrophotography You have 2 main types of noise .. fixed pattern noise, and random noise. To remove fixed pattern type noise you can do things like take a series of similar 'null exposures' (at a similar shutter speed and conditions), then subtract the average noise from your photo. Afaik this is what noise reduction does on nikon cameras when you take a long exposure and it says 'job noise reduction' (this is incidently very different from the noise reduction done in post processing by e.g. NeatImage and NoiseNinja). For more random noise, you need a 'longer exposure to the subject' so you can average out the noise. If you could somehow tweak the gain of a standard camera down to say 'ISO 1', that should give you an idea. Trouble is most camera sensors aren't designed that way - people expect to get results from a quick exposure. Alternatively you could simply take a series of identical photos, and combine them in photoshop by averaging (much like HDR photography). Problem there is that any slight physical movement will lose you sharpness. And of course with any of these long / multiple exposure techniques, you really need a static subject.
  13. Snap, I use photobox to get my prints too ... the image only seems to show when I'm logged in though, so you probably need to host it somewhere else.
  14. A beautiful scene indeed with the mountain and the giant ducks. I had this one yesterday:
  15. This will be the kind of thing they were referring to, I'd imagine:
  • Create New...