New Cintel Scanner from BMD

mike burton Apr 7, 2014

  1. Sorry about the fire. Hopefully your system is up and running now.

    Is there any chance that you can provide a sample Cintel RAW (DNG) frame?

    Also if you have any sample footage to show or share?

    Have you also purchased magnetic audio head separately and which one in that case?
  2. Word from an owner is that the optics are cheap, the image quality is generally poor and there are NO Sound options.

  3. I would look at the Scan Station Personal ($50k) or the Xena (starts around $30K) both have sound options.
  4. Sorry for the late reply! I've been busy. Thanks for your support.

    The last firmware updates have done a lot for calibration in terms of getting the levels right for scanning. Last results have been a lot better than before! The downside is, I don't own the stock I've been scanning. So I'm not at liberty to share at the moment, hopefully I can get clearance on some of the material. Would be fun to put on my site as well. The firmware also fixed some issues with the mechanism spooling up slowly and losing it's lock with slow winding so it's much better to cue now.
    The stabilisation is supposedly done on board in the scanner, using the Cintel Imagemill technology. I'm pretty sure 30fps really isn't that much of a problem. I/O is another story but that depends entirely on the capturing machine. The 4K sensors aren't really that bad, but like you point out later, you need to use the sweet spot.

    You're right, it does indeed work by setting the light intensities, changing the settings does nothing to the sensor. The problem with the early version/firmware was that the scopes gave a different reading during scanning than it did after. So it was nearly impossible to capture in the 'sweet spot'. Now that they've fixed that it's a whole different animal.

    That seems a bit harsh. I'm sure you're aware a lot of scanners have very old sensors and rely on multiple exposures to get to an acceptable image. This camera runs at 30fps and does a full area scan with fair sharpness and dynamic range.

    This is just getting bashy. Sprocket drive or not is a valid choice, not a positive or negative. This one doesn't use a tight sprocket like the Arri does, so this one will take a lot of shrinkage. I've just had some nitrate film in it and didn't have to worry about sprockets hitting at all and a missing perforation didn't throw the scanner off. The sprockets are there just as a pace sensor, I think. I wouldn't be too surprised if you can modify it to work with rubber rollers.

    Sure you can, they're not the same beast, and they don't deliver the same file but you're serving roughly the same purpose. I've worked with Scanity files that were relatively hard to work with in comparison to the Cintel But i've also had great results with damaged material [edit: on the Scanity].

    As I pointed out before, it's not tightly sprocketed, they're just there to 'feel' the film. That said, I'd love to try.

    That again depends on several factors. One customer that's currently likely to buy has a massive amount of 16mm archive of readings and sessions that are all stored fairly well and I've tested footage and compared it with an Arri scan. It did surprisingly well. I honestly didn't expect it to but it really did. Major considerations are budget, maintenance costs, file manageability and output possibilities. After quality, and it stood that test well. All of those are pretty good actually, if you're not bound to SDI.

    Thanks, just about one more month and I can sleep in my own bed! Or I can buy a new bed, the last one went out the window. That was more fun than it probably should have been.

    I'll try to get new scans soon. At this time all I've got is the underexposed stuff and some nitrate film I don't have rights to. The properly scanned material is still at the facility but I'll arrange for it to be sent back to me. By the way, the files are CRI, not DNG. Confusing, I know. Kind of a bummer actually, this means compatibility and durability could be an issue.
    No I haven't. The material I've got to work with so far has been either optical audio or mute.

    I think YMMV means something else than what you're going for. The optics aren't bad, the focal depth is better than the Arri but not as good as the Scanity I think. Focussing is a bit of a hassle but I feel that might be our demo unit. It's all the way at the end of the focus range. Sharpness and chromatic abberations seem to be fine and we're running a test soon with just an ND filter on the light source to see if it's really as evenly lit as they claim. We're getting a lot of center illumination but that might actually just be our source material. Overall I wouldn't say the optics are cheap.

    Also, there are sound options and mount points, they've always been advertised and they demo them at all the shows.. So that's a bit of an off claim. It has built in image-to-audio decoding (as a render in Resolve) and there are magnetic and optic heads available, just not directly from BMD but third parties. The connectors are all there and they're supported from within Resolve.

    I'm still not claiming it's the best scanner ever, but since the firmware and software updates came the scanner took a major jump in quality and workability. At 30K this thing is getting more and more amazing. I'm noticing I'm more annoyed now by all my BMD devices not being able to turn off (and they all have horribly small and loud fans) more than any other bug or misplaced feature. So that means I'm getting quite content working with it. I haven't noticed the pattern noise since I can capture in the sweet spot more easily.

    Thanks everyone for replying so seriously and passionately! More questions are most certainly welcome.
    Jason Myres likes this.

  5. I would actually mostly be interested in seeing the extremes: Underexposed reversal, Overexposed neg. Getting a good scan of well-exposed film is the easy part!

  6. You're not wrong there. I've got some pretty difficult material laying there, so I'll do a batch of tests as soon as I can.

  7. Also, if it's possible to post some DPX files, that would be great. We're still on Resolve 11, and I don't think it can open the scanner's native file format.


  8. What Perry said.

    i.e. the sensor being a "Inexpensive" CMOS unit it will like all CMOS sensors introduce some of it's own noise artifacting to the scan. A Northlight, for example, with an "Older" CCD array will introduce little or no noise of it's own to the scan.

    Also does this machine do any kind of light field correction or calibration? Most new machines do an open gate field correction to eliminate any light field variances, dead pixels, dirt on the sensor or optics, etc.

    As for sprockets just being there for film positioning i.e. non drive, that is not how this machine works. The machine has a set of two sprockets to drive the film and the dancer arms feed servo follower motors for the platters, like all Cintel machines.

    I don't hate sprocket drive, we have a 5K (True RGB) Xena with sprocket drive and Pin Registration and a 6K TriLinear CCD Imagica. There are definitely films that can't go through those machines.

    It seems like an odd duck of a machine and the resolution for 16mm is unacceptable in 2016 really.
  9. hello,
    we just got a cintel scanner for testing in here at kafard films, you are welcome to pass and see if you're around paris.
    Don't hesitate if you have any question or test you'd like to perform.
    I'll try to upload some raw footage as soon as possible.
  10. I'm really curious to hear how it does with dense negative. At NAB they happened to have some neg and threaded it up for me on the show floor. I was kind of astonished they did, because it looked *really* bad. It wasn't even that dense, and there was fixed pattern noise in all the shots where there was sky.
  11. God help you if you have to run a dense IP or a theatrical print through an Imagica. My experience is they do not do well with positive material.
    Robert Houllahan likes this.
  12. The Imagica has a odd Toshiba Tri-Linear CCD I ended up scanning a recent (newly shot) Vista-Vision project on the 5K True RGB Pin Registered scanner and it looked allot better than what I could get from the Imagica.

    The Xena 5K uses this CCD:

    Run in 1-tap Monochrome at 14bit.

    No fixed pattern noise, in fact no noise whatever at the exposure times used for film scanning.
  13. hi perry,
    We've started testing with resolve 12 and indeed weird noise and magenta pattern in bright sky but an upgrade to 12.5 have improved that a lot, however with dense negative we've achieved better results by overexposing the neg and then re-balancing the scan in davinci.
    here is some white scan from a fuji 400 negative and an extract from reala(all with an aricam LT), courtesy of kafard films :)
    Those are in prores unscaled for obvious sharing purpose but tell me if you wanna see some CRI.
    I'll post two scan(over and under exposed) of the reala ungraded later(slow internet connection...)

  14. I would like to get a series of test charts shot on film and then scanned. You can tell a lot about what the scanner is doing just by checking D-Min / D-Max under a variety of exposure conditions.

    I think given that the Cintel scanner is only $30K, nobody is expecting it to make pictures as good as a $500,000 ArriScanner (or whatever those cast). But there are some odd omissions and issues with the scanner that have me concerned. I can't say for sure, since I haven't put my hands on one or had to grade anything done on one (yet).

  15. I'll see if that's possible with my DP, not even sure we have any test chart...
    check the reala scan, it has some nice dynamic i think, i'd like your opininion on that.
  16. I've been away for a bit, moved back into my own house and everything. Things are almost back to normal!

    I haven't noticed the color cast and what I'm seeing in your original is something that (I think) should be fixable with the light controls. The light controls function very similarly to offset and I was able to get to a pretty comfortable color using just the log settings.

    Marc, I only have access to materials that clients give me and so far none of them have come along with proper test charts, if you can send me some I'll be happy to get them scanned in with the Cintel.
  17. hi roel,
    that is actually the best result I had, no pattern or noise in that one. blue was easily overcame and neurtralized any magenta in the sky.
    result seems good to me, only 16mm has been a struggle for now, still testing...
  18. I've had some problems with 16 as well, but mostly device control. The image is as good as I'd expect it to be seeing as you're just cropping part of the sensor. What are your problems?

    I've also compared 16mm to the scanner from MWA (I think it was the Choice? Not sure) and the MWA is a lot sharper.
  19. Transport for 16mm needs some serious upgrade indeed, keep loosening...
    Issue I've encountered is in the grain, it looks like noise and have a lot of magenta compared to a telecine made on a spirit.
    Sharpen is ok, did you make the mwa scan yourself? I suspect some tech to add sharp, some admited to denoise in post...

Share This Page