Jump to content

jan_brittenson

Members
  • Posts

    569
  • Joined

  • Last visited

Posts posted by jan_brittenson

  1. By the way, the light leak doesn't have to be near the film gate, it can be anywhere along the path. The film wasn't necessarily fogged while positioned at the gate. So even though it's on the left side of a print it could still be caused by something around the left side of the path, like the foam around the window.
  2. Sounds like a pinhole light leak along the bottom. Maybe a dent along the edge? The ones showing it are likely the frames where the film sat still in the camera for some time (hours, days?); the ones not showing it are likely the ones where the film got wound past somewhat expeditiously. Could be the foam around the window, as well.
  3. I do like Steve, lightly add a little tension with my left thumb while winding up to the start mark. I've always done the same with all my 120/220 cameras I've ever owned. I also treat 220 just like 120 and have carried rolls for weeks in a clear plastic ziploc bag. Never had a problem with either kind of film. I suspect if you get loosely wound 220 in your Mamiya 7 there's something out of whack. I'd get it fixed. (My 7II winds both 120/220 nice and tight.)
  4. Took the R9+DMR to the Chinese New Year's parade in San Francisco

    tonight. Fun!

    <center>

    <p>

    <img border=0

    src="http://www.rockgarden.net/download/cny2006/L1090853-00000.jpg"><br>

    <i>Carrying the dragon's head, Summilux 80</i><br>

    <br>

    <p>

    <img border=0

    src="http://www.rockgarden.net/download/cny2006/L1090842-00000.jpg"><br>

    <i>Year of the dog, Summilux 35</i><br>

    <br>

    <p>

    <img border=0

    src="http://www.rockgarden.net/download/cny2006/L1090827-00000.jpg"><br>

    <i>Dancing dragon, Summilux 35</i><br>

    <br>

    <p>

    <img border=0

    src="http://www.rockgarden.net/download/cny2006/L1090835-00000.jpg"><br>

    <i>Teasing the dragon! Summilux 35</i><br>

    <br><p>

    <img border=0

    src="http://www.rockgarden.net/download/cny2006/L1090869-00000.jpg"><br>

    <i>Neatly arranged, Super-Elmarit 15</i><br>

    <br><p>

    <img border=0

    src="http://www.rockgarden.net/download/cny2006/L1090882-00000.jpg"><br>

    <i>Falun Gong staging a protest at being excluded, Super-Elmarit

    15</i><br>

    <br><p>

    <img border=0

    src="http://www.rockgarden.net/download/cny2006/L1090850-00000.jpg"><br>

     

    <i>Hurrah! Summilux 80</i><br>

    </center>

  5. Noise is not a floor, unless it's white noise. In reality, noise is a form of distortion; the original signal is still there, except it's overlaid with pink noise. There may be noise, but to our eyes the image is still there. Most dynamic range testing assumes some arbitrary noise point as the blackpoint for the measurement, but this problematic because it's not a blackpoint. In reality what matters is both the total range which can be distinguished, as well as the characteristics (not just amount) of the noise distortion. Most are familiar with the notion for instance that luminance noise is much preferable over chroma noise. So fundamentally the numbers we can obtain my analysis are pitted against aesthetics. And guess what... in the end the proof is always in the pudding.
  6. I too have done enough business with Poon to feel perfectly comfortable recommending him.

    <p>

    The P67II is a very nice camera. I don't use mine much, and have sold most of the lenses, but the body and finder never sold -- so I decided to hold onto it. I especially like using it for all kinds of experimental work, where once I figure it out visually and technically I can switch to 4x5 instead. It's a very easy camera to hold and operate, and you can cycle through cheap lenses until you figure out what you really end up using in the long term.

    <p>

    <a href="http://www.photo.net/photo/2493888">Stuff like this light painting. (P67II, 90-180, Efke 25, Microdol-X.)</a>

  7. Cool, thanks for the data sheet! I like the idea of the output crossbar on the device, that should allow using a separate ADC for each channel. Which means each channel can get its own gain and offset, allowing you to white balance in hardware -- just like scanners. If I were you I'd seriously consider a dial for WB and sensitivity. Not sure how to do that cleanly though without suffering feeping creaturism.
  8. I prefer to do my own testing and evaluation. Generally, if the numbers, checkboxes, and reviews don't come up with a showstopper issue but otherwise might indicate a suitable fit for some need I'll find one to try out. Either in a store, an acquaintance, or I buy used. Sometimes I'll buy new with the understanding that I can return it if dissatisfied. The reason I do my own evaluation is nobody else seems to have my particular needs, care about the same things, and most reviews are concerned with purchasing value or predicting what "most people" will think -- which is pretty much useless to me. (My name isn't Most People, I don't share Mr People's needs or interests, and what's a good value is perhaps the most irrelevant point that can possibly be made.)
  9. Stop-down metering only works with some Canon bodies, depending on which type of metering it uses. It didn't work at all with my Canon 1D, but worked fine with the 1Ds and 1Ds2 for instance. I used these with both Contax and Leica R lenses, with no difference in results. I suspect the 20D's meter needs to know what aperture it's metering at, or it might be close to the edge of the image so needs calibration (linearization) data from the lens.
  10. The reason it looks dark is that your screen isn't a linear device. To double the perceived brightness you need to quadruple the signal. This is known as gamma 2.0. It's perfectly normal for linear files to look dark since nobody other than you know that they're linear and not gamma 2.0. Profiles and color spaces declare among other things what the gamma of a file is, so if you assign a linear profile (one that declares the data is linear, aka gamma 1.0) to it it will look fine since it's now known what its actual gamma is and no assumption needs to be made. If you've turned off all color management, then the implicit assumption is that the gamma and other attributes of the file are the same as your screen. If your file is linear it will look very dark because the screen approximates gamma 2.0 (NTSC) while the file encodes data as gamma 1.0.

     

    Going from gamma 1.0 to 2.0 is a pretty significant change, one that needs to be done in 16 bit depth. If you assign a gamma 1.0 input profile to the file you will likely see significant posterization, this is because the preview generation (aka simulation) is designed to be fast, not accurate, so is usually done in 8 bit with little or no artifact suppression. What you get from an actual conversion (recalculation of image data to go from one space/profile to another) will likely differ. So linear files aren't generally viewable or workable, other than for automated processing (or PS color adjustments and other work "by the numbers"). Also, it has been my experience that curves is a much better way to make large gamma adjustments than using color management engines, the latter simply don't seem to be designed to do this well and tend to produce artifacts. (A single curve point of (128,192) will multiply the gamma by 2, BTW.)

  11. It's a general-purpose tool that simulates one device using another. Like a printer using a screen, a printer using a different printer, or a screen using a different screen. Technically, more complex configurations are possible as well (such as simulating what a simulation of your printer would look like on your TV, assuming you have a profile for both). It also allows you to check what something will look like on a device for which you have a profile but which may not be color managed per se, such as say a particular DVD player the TV set. You might be able to work out a profile describing it, but the DVD player is never going to pay attention to color management.

     

    It also allows you to first edit and make all changes (in a larger working space), with gamut warnings, and then when you like what you see do the final conversion to the device space. This is especially useful when the device has a relatively small or quirky gamut.

     

    I strongly recommend Real World Color Management by Fraser as an intro to CM. It really makes a lot of this more clear (although it's still not all that easy to understand always).

  12. I've been trying to set up a Mac Mini to run as an ImagePrint 6

    spooler, but I just can't seem to get it to work.

    <p>

    The Mini is configured as shipped; it came with 10.4.2, I upgraded it

    to 10.4.3.<br>

    I've installed the ImagePrint 6 Tiger Update.<br>

    The dongle is attached and shows in IP Setup.<br>

    I've installed the Epson 7600 drivers, which work fine. The printer

    shows 480Mbps (in System Profiler) so is USB2 protocol enabled. I

    wanted these drivers just to make sure all is good, but IIRC they're

    not required for IP6.<br>

    It accepts the encryption key when I hit enable, but then issues a

    string of "Socket create error in ip_config" or something cryptic like

    that. If I mistype the encryption key it gives in a message about

    that, so it seems the key is good.<br>

    After displaying these messages in IP Setup the printer remains

    disabled.<br>

    <p>

    Anyone else had the same problem?

  13. <i>Imported (Grey Market) items from B&H are warranted and serviced through B&H. Nikon cameras and such imported through side channels (not official Nikon USA) will not be serviced by Nikon USA service centers in or out of warranty. There is an explanation on the B&H website:</i><p>

    B&H don't explain how to get IMP Nikon equipment repaired once it's out of warranty. Is B&H committing to always service them (for pay) in perpetuity, since Nikon USA won't touch them? Does this committment apply when the item is sold, or is IMP Nikon essentially unservicable once sold?

  14. What problem are you solving by going to Leica? The 24-70 is a fairly mediocre lens as far as quality optics go. The Canon primes are significantly better. The difference between Canon zooms and Canon primes, "L" or not, is much greater than the difference between Canon primes and Leica primes. If you're not looking to address something specific and understand the tradeoffs, you're likely to be disappointed.
×
×
  • Create New...