Please help with the Processing Speed

Comments

12 comments

  • Ulf Liljegren
    Capture One is not a memory hungry program if you compair to PhotoShop as an example.
    Fuji S2 and S3 images does takes longer to develop then for example Canon and Nikon due to the diagonal readout of the chip.

    Weather or not this is the correct time is hard to say when it is depends on CPU speed, Memory Speed, Harddrive speed, Configuration of computer, other programs running, Services running.

    Any users on the forum that has some processing speed examples?
    0
    Comment actions Permalink
  • erikhoffman
    Its straight forward math calculation
    Why dont you take a look how long it takes for one photo and multiply by 380 so it looks like it takes about one and a half minutes to process one pic which is far too long and as you did not indicate what processor you have i Guess may be 2Gh
    For this sort of work a dual Xeon is a must and with 3Gh i process my P25 at 25 seconds per pic
    so I cannot see that Fuji will take as long as you indicate

    Erik Hoffman
    0
    Comment actions Permalink
  • mikeh
    Sorry I thought I was little more clear I have a pentium 4 processor that is a single processor at 3.4GHz. I have 2GB worth of ram. It takes about a minute and a half per image. I could only wish it was closer to the 25 seconds the last individual stated. Is this all the better I will get??
    0
    Comment actions Permalink
  • mikeh
    I am running a 10,000 rpm Hard drive and have anti virus, and all other programs turned off when I processs the batch.
    0
    Comment actions Permalink
  • MattFahrner
    I use a Fuji S2 and find C1 Pro takes an exceptional amount of time compared to Adobe Camera Raw or Fuji EX software. It's probably at least 3 times longer if not more.

    This is on a 1.7 ghz Pentium with 512mb ram running XP.

    Ok, I did some timings. For Phase One's Capture One:

    2:50 for full conversion with TIFF write to disk
    (45 seconds to show\"100%\", 45 more seconds for progress bar to
    stop, 80 seconds more to actually complete)


    With Adobe Camera Raw:

    21 seconds
    (15 seconds for conversion, 6 seconds for manual TIFF write)


    With Fuji EX:

    52 seconds for full conversion with TIFF write to disk

    All had about the same amount of color correction, exposure adjustment, and sharpening.

    The weird thing is is that the majority of time is spent after it reads \"100%\" and after the progress bar reaches the end (which oddly is after reading 100%).

    As you can see, C1 is an order of magniture slower than ACR. Looking at the images though, there is little difference, at least with this particular image, in any of the renderings (they are all excellent). Generally I find C1 a little softer than ACR, but I think this has to do with DSLR noise reduction not having an \"off\" setting.

    Unfortunately, clearly the argument of the honeycomb pixel layout slowing things down does not hold - Adobe faces the same issue and rips through the conversion.

    I really want to buy C1, but there are these little quirks like this that are making me waver. I really suspect that there is an issue in the actually writing of data or something because of how much time is spent at the end of the conversion when it claims to be \"complete\".
    0
    Comment actions Permalink
  • mikeh
    But the truth is you are right. ACR is quicker. However I have found the opposite in the clarity or sharpness of the images. C1 tends to be a better file when done, for me anyways. Maybe it is time to make that switch to a Mac or a Canon system 😕
    0
    Comment actions Permalink
  • MattFahrner
    Ha! You wish!

    I ain't gonna get locked into one of those overpriced proprietary, can't buy from anyone else but Apple, Macs (though they are pretty dang nice these days).

    Nor am I gonna trade in my (also dang nice and usually well detailed and sharp) Fuji S2 just to make a $99 piece of software run better! And a Canon? I'm a Nikon man - those are fightin' words! 😊

    Actually I'd be perfectly happy to have a Canon if I didn't have beaucoup dollars in expensive Nikon glass. But I would also be equally happy to have a Nikon, which is why I'm staying. Naw, my next stop is either an S3 or a fancier Nikon, but that won't be a while because I already pushed my luck with my better half buying this one (though if I bought it for her that might um...).

    Anyway, I wouldn't hold your breath over me switching to a Mac or a Canon. In the mean time, I think the conversions should be faster and sharper, and I suspect some of it may be related to Fuji S2 alone (whereas a Canon might render faster and be sharper). Clearly there are a lot more Canons out there than Fujis or even Nikons for that matter and I'm sure they've got more attention.

    Thanks for the response (and getting my blood up!). 😊

    I think it would be an interesting poll to see if Nikon users tend to prefer PCs, while Canon users prefer Macs. Wonder if there's a mindset there.
    0
    Comment actions Permalink
  • MattFahrner
    By the way, using a different image (of a landscape, not a face) I couldn't really see any discernable difference in sharpness. It may be that it depends on the subject matter (or I'm just doing something stupid - equally as possible).

    I do think the speed, as noted by my timings, is unacceptably higher.
    0
    Comment actions Permalink
  • Nick 1
    Mikeh,

    I use a dual 2.6 HP workstation with 2.5Gb RAM and ultra scsi drives - I would suspect our systems should be fairly close in processsing speed?

    My timings for an 11Mpix Canon 1Ds are 20sec to 100% 16bit TIFF. On a 1.6 Centrino laptop with 768 RAM, the time is 38 sec.

    I don't have a Fuji, but would strongly suspect an issue on your system / install of C1? 3.5.2 is fairly stable and also fairly quick too. It may not beat CS on the clock, but in terms of quality and flexibilty, even using Russell's script, it is a better package.

    HTH?

    Nick WB

    BTW my Mac Dual 2Ghz machine also processes the same file in 20 sec 😉
    0
    Comment actions Permalink
  • MattFahrner
    I honestly don't think it's an install problem. I have exactly the same issue with my Fuji S2 images with no improvement moving to C1 LE 3.6. I think it's Fuji S2 conversion specific.

    I'm also convinced there is nothing wrong with my \"workstation\" as it were. I have no issues with any other software, conversion or otherwise. Given that we both have the same issue, I think it would be worth Phase One looking at it. It would be pretty easy to test.

    I'm also willing to supply a Fuji RAF for anyone to try.
    0
    Comment actions Permalink
  • MattFahrner
    I did a test - with a completely fresh install on a Dell 2.6 ghz machine (512mb memory, 7200 RPM disk, XP Pro) where C1 had never been installed before. It took C1 roughly 2 minutes 40 seconds to convert a Fuji S2 RAF. It was actually longer than my 1.7ghz laptop tests - but then again this latest was with 3.6, not 3.5.2.

    Again, ACR is less than 30 seconds and Fuji EX is like 50 seconds. C1 is pretty extreme.

    I also thing the LE output size should either be 12mp or have a choice for 6mp and 12mp (that would be my favorite, but I know LE doesn't allow you to choose sizes). The S2 default size is 12mp.

    There is also that funky focus offset bug...
    0
    Comment actions Permalink
  • MattFahrner
    After restarting the 2.6ghz workstation and making sure nothing else was running (I had forgotten about an audio app), I got it down to 1 minute 50 seconds. That while quite a bit better still seems too long.

    All I'm aiming for is getting someone at Phase One to try a supplied RAF (any RAF really) on a Windows XP box with 3.6 LE and see if they too think it's too long. If so I'll be happy as long they:
    • Either find that I'm wrong.

    • Or queue it for a future fix.
    Thanks.
    0
    Comment actions Permalink

Please sign in to leave a comment.