liork Posted February 21, 2017 Share Posted February 21, 2017 Why is sensor size relevant to the amount of crop in "IS"? Quote Link to comment Share on other sites More sharing options...
Jimmy Posted February 21, 2017 Author Share Posted February 21, 2017 Is that a serious question? If a digital IS system requires, say, a 5% crop to work it's magic... That is very inconsequential on an s35 sensor... your fast lenses will still be fast, your wide lenses will still be wide. The smaller the sensor, the bigger impact it will have on your glass. Quote Link to comment Share on other sites More sharing options...
liork Posted February 21, 2017 Share Posted February 21, 2017 5% crop will be 5%... if you use 12-35mm on G85 which is 24mm on the wide side (equivalent to full frame) vs using 15-85mm on Canon D77 (which is also 24mm equivalent to full frame on the side side), 5% crop will be the same on both cameras. Quote Link to comment Share on other sites More sharing options...
Jimmy Posted February 21, 2017 Author Share Posted February 21, 2017 No shit.... But you were talking about the XC10, a fixed lens. Even if you do insist on bringing the G85 into a conversation about magic lantern cameras. There are not infinite combinations of lenses that mean you can ALWAYS match the focal length and F stop of one crop to another crop. Considering the vast array of lenses that could go onto the D77, a 5% crop would make very little difference. This becomes especially relevant when talking about lenses that could utilise the AF system of both cameras. Quote Link to comment Share on other sites More sharing options...
Nikkor Posted February 21, 2017 Share Posted February 21, 2017 software stabilization raw Video hack? No. Quote Link to comment Share on other sites More sharing options...
Jimmy Posted February 21, 2017 Author Share Posted February 21, 2017 "No" what? The XC10 does a great job of combining software stablisation and DPAF... Add raw and a large sensor and it could be beautiful. Especially for under $1000. Hopefully the ML guys can think outside the box more than the people on here. Quote Link to comment Share on other sites More sharing options...
Arikhan Posted February 21, 2017 Share Posted February 21, 2017 @Jimmy Quote The XC10 does a great job of combining software stablisation and DPAF Wrong...The XC10 has NO DPAF... Quote Link to comment Share on other sites More sharing options...
Jimmy Posted February 21, 2017 Author Share Posted February 21, 2017 well, it has good AF... if it is not DPAF, then the 77D with raw would be even more enticing. Quote Link to comment Share on other sites More sharing options...
mercer Posted February 21, 2017 Share Posted February 21, 2017 Yeah if the ML team has the inclination, and it's possible, the 77D could be a great option for low cost, handheld ML Raw. The software based IS in the XC10 is the best IS I have ever used. It's better than the IBIS in the a6500 and the GX85... no question. Other than a strap I was pulling taut, there was no other physical stabilization. During this shot(s) I walked forward backward and shuffled sideways... try and do that with a Panny... jpfilmz and Jimmy 2 Quote Link to comment Share on other sites More sharing options...
Nikkor Posted February 21, 2017 Share Posted February 21, 2017 46 minutes ago, Jimmy said: "No" what? The XC10 does a great job of combining software stablisation and DPAF... Add raw and a large sensor and it could be beautiful. Especially for under $1000. Hopefully the ML guys can think outside the box more than the people on here. If the stabilization is Software I don't think it will be Applied to undebayered Data because normally it implies rotation,etc. Maybe they could write the Data from the giroscope to a file and use that for a Software stabilizer on the Computer. But those are many Maybes. Jimmy 1 Quote Link to comment Share on other sites More sharing options...
Jimmy Posted February 21, 2017 Author Share Posted February 21, 2017 You might be right... maybe the IS calculations would be earlier in the chain so as to show in live view... but I'm really not very technical. Quote Link to comment Share on other sites More sharing options...
Marco Tecno Posted February 21, 2017 Share Posted February 21, 2017 Nx1 has dis which works by slightly cropping the image to reduce shakiness. This is also done in 4k. Quote Link to comment Share on other sites More sharing options...
tweak Posted February 22, 2017 Share Posted February 22, 2017 What is the card write speed of the 77D? As all this talk is very meaningless if that isn't high enough (which I highly doubt it will be given the cameras specs and price). Quote Link to comment Share on other sites More sharing options...
liork Posted February 22, 2017 Share Posted February 22, 2017 This is the card write speed of the D80: http://www.cameramemoryspeed.com/canon-80d/sd-card-comparison/ Quote Link to comment Share on other sites More sharing options...
IronFilm Posted February 22, 2017 Share Posted February 22, 2017 On Sunday, February 19, 2017 at 1:24 AM, Jimmy said: I guess the issue is whether the buffer/SD is fast enough to write a usable resolution in raw. Almost certainly will not be good enough. Quote Link to comment Share on other sites More sharing options...
Jimmy Posted February 22, 2017 Author Share Posted February 22, 2017 It seems the 10/12 bit raw compression is coming on well, which might allow close to 1080p.... Obviously lots of ifs, buts and maybes in this thread though. Quote Link to comment Share on other sites More sharing options...
liork Posted February 22, 2017 Share Posted February 22, 2017 Did you really think Canon will give a video powerhouse for a bargain price??? Quote Link to comment Share on other sites More sharing options...
Jimmy Posted February 22, 2017 Author Share Posted February 22, 2017 1 hour ago, liork said: Did you really think Canon will give a video powerhouse for a bargain price??? Of course not... Their DSLRs are the least progressive on the market... that's the whole point of Magic Lantern Quote Link to comment Share on other sites More sharing options...
andrgl Posted February 22, 2017 Share Posted February 22, 2017 17 hours ago, mercer said: Yeah if the ML team has the inclination, and it's possible, the 77D could be a great option for low cost, handheld ML Raw. The software based IS in the XC10 is the best IS I have ever used. It's better than the IBIS in the a6500 and the GX85... no question. Other than a strap I was pulling taut, there was no other physical stabilization. During this shot(s) I walked forward backward and shuffled sideways... try and do that with a Panny... Egads. This is a great video to show why EIS is terrible for movement. You've got motion blur galore (from excessive camera shake) and yet the software has aggressively locked off the frame. It looks completely unnatural and the human brain can pick it out without any knowledge of why it looks "wrong". You should read up on some technical papers on how OIS works. The TLDR implementation of most manufacturers is that it offers 1-2° of correction, which means you have to keep the lenses pointed in the same angle while capturing video. Even excessive handshake will cause the OIS to hunt and create that ping-pong effect you see from most people who complain about OIS not working. This shit ain't magic. At the end of the day it's using data from a gyro and some code to predict what should be happening. Plop a clueless human into the equation and you end up with... mercer 1 Quote Link to comment Share on other sites More sharing options...
mercer Posted February 22, 2017 Share Posted February 22, 2017 2 hours ago, andrgl said: Egads. This is a great video to show why EIS is terrible for movement. You've got motion blur galore (from excessive camera shake) and yet the software has aggressively locked off the frame. It looks completely unnatural and the human brain can pick it out without any knowledge of why it looks "wrong". You should read up on some technical papers on how OIS works. The TLDR implementation of most manufacturers is that it offers 1-2° of correction, which means you have to keep the lenses pointed in the same angle while capturing video. Even excessive handshake will cause the OIS to hunt and create that ping-pong effect you see from most people who complain about OIS not working. This shit ain't magic. At the end of the day it's using data from a gyro and some code to predict what should be happening. Plop a clueless human into the equation and you end up with... Yeah obviously no in camera stabilization, software or hardware based is made for drastic movements... or any movement really. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.