Jump to content

Alpicat

Members
  • Posts

    104
  • Joined

  • Last visited

Everything posted by Alpicat

  1. With the latest build Danne posted the max resolution that I can set is 1184x2000 (i.e. 3552x2000) in 1x3 mode - I don't know why, maybe beyond that it becomes too unstable and he set that limit, or maybe I'm doing something wrong - but for me it's a nice resolution, it comes out exactly as 16:9 aspect ratio.
  2. It's incredible how much progress has been made on the EOS M in the last few months! I've tried several builds with the 1x3 binning setting but was getting corrupt footage every time, until today - with Danne's latest build I'm finally getting good results in this mode. The max res in 1x3 seems to be 1184x2000 in this build (which stretches out to 3552x2000) - I'm getting continuous shooting in 14-bit lossless. As @Brian Williams mentioned the rolling shutter is really pronounced though, and there's noticeable fixed pattern noise in the shadow areas which I hadn't seen this camera do before. Don't know if these things can be improved in future builds.
  3. 16:9 would be 1736x976 - in the magic lantern menu select 16:9 aspect ratio and that's the resolution that should appear. You'll probably get close to continuous recording without the sd card hack in 12 bit lossless at this resolution. I've just seen that the eos m can now shoot 3k (3032x1436) at 24fps in 5x zoom mode. Look forward to trying this later this month when I can. The crop in this mode is 2.76x, or 2x with the speedbooster. Record times will probably only be a few seconds even in 9 or 10 bit lossless raw.
  4. @henricus If you want to shoot this resolution you have to enable crop mode (I think it's called mv1080p 3x3) - otherwise you're limited to 720p vertical resolution with excessive aliasing. The aspect ratio is what you see in the video, I guess it's close to 1.5:1 since this mode nearly uses the entire aps-c sensor space. I get continuous recording at this resolution in 14-bit lossless raw if the sd card hack is enabled - i.e. you can record for minutes, until the camera starts getting hot and then it's probably a good idea to stop. Without the card overclock at 10 or 12 bit lossless raw you should still be able to get fairly decent record times from what I remember.
  5. Hi yes I tried it and it works fine. although liveview is a bit crazy with the screen heavily stretched vertically. When corrected in post, the image comes out at an aspect ratio of 4.6:1 which isn't very useable. You can crop the sides off down to 2.35:1 ratio but then you loose quite a bit of resolution. I'll test it further though. There's someone who stuck an anti-aliasing filter into their EF to EF-M adapter using sellotape, sounds like an interesting solution! I might try that one day!
  6. At long last I've received the Viltrox speed booster for my EOS M, which means it's now a full frame raw video camera! Seems to work nicely so far, here's my first footage using the whole sensor area. Sorry I keep doing videos of the same thing, hopefully one day I'll do a more interesting video: Youtube description: First test using the Viltrox Speedbooster with Magic Lantern raw video on my Canon EOS M. Resolution is 1736x1042 (14-bit lossless raw) in 3x3 mv1080p mode, using the full sensor area. The Viltrox speed booster provides 0.71x magnification which makes the EOS M camera practically full frame (1.14x crop compared to 35mm full frame). As always, youtube compression makes the footage look mushy compared to what I exported from davinci resolve. The Viltrox seems sharp enough for video, although it doesn't perform so well in the corners for photography, which may bother some people. I used a Tamron adaptall 35-70mm f3.5 for this video. I didn't use the SD US overclock card hack as at this resolution you usually get decent recording times without the hack. Also I forgot to remove the focus pixels on mlvapp. Sorry for the shaky video, this was a very quick test!
  7. Thanks Andrew for this eye opening article, didn't realise the situation was that bad. Interesting to hear about what you experienced in Berlin. It isn't yet but I've had to start preparing for it (for commercial projects) - had to redo contracts last minute before a shoot to take GDPR and article 13 stuff into account, and had to get everyone to sign again. Have to be extra careful on set now with getting release forms signed from absolutely everyone even if they're hardly visible in the background. If buildings are also considered copyrighted material, that's going to make things even harder!
  8. They did have a bunch of cameras at the event for people to try out, which I did, but they didn't allow us to record onto our SD cards. The camera is fairly weighty (tried it with a veydra lens), but that does make it quite nice to hold.
  9. They haven't said anything unfortunately, they just showed the films. One person who ordered from cvp really early on told me he hasn't had any updates from them at all, which sounds a bit worrying. I personally haven't ordered yet, will only get it in the new year if I need it for work.
  10. I'm at the pocket camera launch party in London, they've shown this film on the big screen along with a couple other shorts, looks nice! Anyone else at the event?
  11. I have the ef-s 17-55mm and you can just pull the back part off with your fingers, no need to unscrew anything. You do need to use some force but it seems safe to do - I think it's different to newer ef-s lenses where you do have to use a hacksaw. With that part removed, it should be fine to use on a speedbooster. The focus ring on this lens has soft stops, it's nice to use - with a slightly longer focus throw than the Tamron or Sigma equivalents. The stabilisation is only 3 stops but it's very smooth for video - it doesn't jump around abruptly like some Tamron lenses or Lumix lenses seem to do when panning. The build quality isn't that great unfortunately, it feels a bit fragile considering how much it weighs.
  12. I'd never buy this camera but for certain small jobs where I'm shooting myself and don't have a camera team I may consider renting it for its auto focus capabilities, should be cheap enough to rent hopefully. I'm wondering if the DPAF will be as effective with continuous video autofocus when using adapted EF or EF-S lenses as it will be with native RF lenses - will be interesting to test this out. @hijodeibn it could be a while before we see a speedbooster, judging by how long it's taking Metabones and Viltrox to release one for EF-M mount, which both companies have been planning to make for a really long time. It just takes time to design and manufacture stuff. Hopefully the EOS R will generate more interest and they'll be quicker! Although again, I'm wondering how effective video autofocus will be with an adapter.
  13. @Brian Williams with MLVFS I think there might be an option to remove focus pixels in the same way that MLV App does (with a focus pixel map) but don't remember what the option was called. Chroma smoothing supposedly does degrade image quality slightly - but probably not noticeable most of the time. @mercer I've tried converting mlvs to prores (using the LogC curve) on MLV App, and I agree the results are great! Interestingly with the EOS M I found that when I exported some shots to prores, it got rid of aliaising and moire, which remained an issue when exporting to cdng... but in some other cases where there was a different type of aliaising it didn't make a difference and possibly the cdngs were better! There haven't been updates for the bitrate hack that I know of. We were extremely lucky that the magic lantern forum members did so much work with the EOS M these last few months. It's gone a bit quiet now and will probably be a while before we see any new updates for that camera, but that's ok. Now I'm just waiting for Viltrox to release their EF-M speedbooster so I can turn my EOS M into a 'full-frame' camera using the mv1080p mode! I think the 70D also benefits from SD card overclocking but don't know if that camera can record lossless raw yet.
  14. @Brian Williams I'm using Windows and was also getting corrupt files with MLVFS. I don't think it's compatible with lossless raw files, but need to check if there's an updated version. The patterns you were seeing are the focus pixels? MLV App works fine for me and it gets rid of focus pixels automatically without using chroma smoothing, but would be interesting to see if MLVFS produces different results at all with the exported DNGs (in terms of colour and image quality).
  15. @Brian Williams these tests are great and the colours look good! Would be interesting to know what your post production process was.
  16. @tupp @youshouldtry11 thanks for this, had never heard of the cameflex mount! Shooting in raw is stable even with the SD overclock hack - I rarely had problems with corrupt footage, unless it's an extremely experimental build. The only issue with the SD hack is activating it, which can be a bit risky, but after activation, the camera is fine to use as normal. If you want to avoid the SD card hack and want to use s8 or s16 lenses, just download the latest 4k lossless experimental build from the magic lantern website, select "movie crop mode" in 10 or 12 bit lossless raw up to 1800x1030 resolution (4.66x crop). That mode is nice to use as it gives you a properly framed liveview, and should give you fairly continuous recording times. Of course you can also select lower resolutions whilst in movie crop mode if you want a bigger crop or a different aspect ratio.
  17. @youshouldtry11 c-mount is a screw thread type mount and I don't see any screw here. I'm unsure what makes you think the above is c-mount, to me it looks like it doesn't have a mount - perhaps it's been physically removed. I sold my schneider lens last month as I wasn't using it, however below are some photos of it with c-mount: https://thumbs.worthpoint.com/zoom/images3/1/0915/23/schneider-kreuznach-optivaron-66mm-1_1_ba20e52d51029e2677ec30cad36cabc4.jpg https://www.picclickimg.com/d/w1600/pict/263721207841_/Schneider-Kreuznach-Beaulieu-Optivaron-18-6-66-C-Mount-Zoom.jpg
  18. @youshouldtry11 The lens is pretty small, but heavy as it's metal - feels front heavy on the eos m. It's a c-mount lens -I think this is the adapter that works best to make it work parfocally, though haven't personally tried it: https://fotodioxpro.com/products/c-eosm Don't know what mount the lens in your photo has - it doesn't look like it has any mount at all, maybe it was unscrewed or removed, or it's a custom built one?
  19. @youshouldtry11 if you plan to use super 8 lenses, you're best off using "movie crop mode" and cropping to 4:3 aspect ratio in post to get rid of most of the vignetting. Movie crop mode in h.264 (1920x1080 resolution) is similar or same to the field of view you get with movie crop mode in raw (1080x1030 resolution). This is a 4.66x crop from full frame - which is larger than super 8, but cropping to 4:3 aspect ratio will get rid of most of the vignetting. Below are tests I did with the eos m and schneider 6-66mm f1.8 which I had posted here a few pages back. I shot this in raw at 1800x1024 in movie crop mode. For this lens I think the fotodiox adapter is the best one that will give you parfocal lens performance, from what I heard. It's tough to find an adapter that works as they all seem to be really different. If you want to use super 16mm lenses and shoot raw, there's also the option of using 5x zoom mode with a max stable resolution of 2520x1080 (3.33x crop from full frame), this is close to 16mm (rather than s16). However, it isn't possible to shoot this resolution with regular h.264 video.
  20. Thanks @tupp. I guess this update makes the title of this thread slightly out of date now! Here's the build I used: https://bitbucket.org/Dannephoto/magic-lantern/downloads/magiclantern-Nightly.2018Jul03.EOSM202_mv1080p.zip With this build you're stuck with shooting full sensor only (3x3 pixel binning), you can't switch to 5x zoom mode. They may have partially fixed that in the newest build but unsure if you can then access the full sensor height, I'll need to test. See here: https://www.magiclantern.fm/forum/index.php?topic=19300.msg204564#msg204564 I shot that footage at iso 100 and iso 400, f3.5 and above. I later noticed that if I export to prores on MLVapp I get less noise in the shadows that exporting to cdng (which can look really excessive if you push up the shadows). However I don't know if I tested this fairly and will have to experiment further. The above video is cdng. Right now the main issue is liveview preview is streched vertically 1.6x (preview looks like you have an anamorphic lens attached) and you can only see the top half of the image. It's still usable though and half-pressing the shutter gives you a low fps / pixelated preview of the whole frame. The moire/aliaising can get fairly bad depending on what you're filming, and can become very obvious if cropping in on an image. I guess it's now the same amount as you get with all the other aps-c Canon DSLRs, or it's similar to the bmpcc. Exporting to prores rather than cdng on mlvapp can sometimes reduce the amount of moire - I need to test this further though.
  21. @mercer thanks! I do need to make a short film and should watch Rear Window, can't believe I haven't seen that yet! Yes I'm using Juan Melara's method on resolve, at least for clips where any highlights are blown out otherwise it's not needed. It's the easiest way to get highlight rolloff to look ok - I have it all set up on resolve so it only takes a click and minor exposure / colour balance adjustments after that.
  22. Thanks! It will certainly be slightly similar to the 5D once the speed booster is available! At the moment, the resolutions I've mentioned above require the use of highly experimental builds that aren't even available in the experimental downloads page yet. It's a bit of a hassle getting the different modes to run, however progress is being made pretty quickly.
  23. Here's my first quick test with mv1080 full sensor recording on the eos m (1736x1120), no crop and continuous recording in 14 bit lossless raw. I'm happy with the image quality and moire isn't too bad. I think youtube softens the image slightly, so have uploaded at 1440p which hopefully looks a bit better: @webrunner5 In crop mode at 2520x1080 or 2520x1386 the ef-m 15-45mm lens is 50-150mm full frame equivalent (3.33x crop). With a 0.71x speed booster that will be a 2.36x crop - so using an ef-s 10-18mm (modified to fit speed booster), that would result in 24-43mm full frame equivalent. The ef-m 15-45mm won't work with the speed booster. The river thames video I shot at 2224x1200, which has a crop factor of 3.75x, so the 15-45mm lens was 56-170mm equivalent there. The crop factor changes depending the resolution selected - values can be calculated here: http://rawcalculator.bitballoon.com/calculator_desktop
  24. Have tested it and it's working fully. The preview is stretched vertically 1.6x for now - as if there were an anamorphic lens attached, but other than that it's fine. I never thought they would manage to do this! Also, seperately, here's the most extensive test I've done so far with the new crop rec 2.5k mode (max 2520x1386) on the eos m. I recorded at 2224x1200 to get usable record times of around 10 - 20 seconds per clip. It was a very warm day so the camera got pretty hot!
×
×
  • Create New...