outerbeat

190 mbps looks absolutely amazing

Update on this nx500 hack package - last version is here

Thank to Kino Seed on dpr, here is great mod pack for nx500 with MENU for all options, so no more need for editing files to change settings! There are some key-combination changes: EV+... AEL: immediate hibernation UP: load UP profile DOWN: load DOWN profile Half-Press-Shutter: focus pull mobile: telnet EV: Main Menu ############# Main Menu - video birtates - custom functions - profiles - settings - hibernate - sleep Video Bitrates lists all set bitrates, and allows for changing them persistently (that means settings are kept even after camera restart) Custom Functions - focus stacking - focus buttons - batch recording (continuous recording) Profiles save/load UP or DOWN profile save/load FullBackup (all camera settings) Settings - Silent Shutter (nx500 only) - VGA to 2.5K (nx500 only) - Uninstall Mod NOTES: - for the moment this mod utilizes only the BT-patch (loading from Sdcard) - "checkbuttons" used for nx500 are not "holding" the preset values at the moment - options / buttons starting with a "-" are not yet implemented How to install: 1) start with a clean memory card (no files) 2) unzip the contents of the zip into the root of the sdcard 3) put card in camera, and restart x) to uninstall mod - from custom menu> settings> Uninstall Mod Download link _____________________ I've tried this one on NX1, it installs OK but does not apply any settings yet, need some bug-fixing. I think in nx500 it must work great. though. Please try it and share experience

Well, I think card can deal with heat, cause in card reader it heating actually more than in camera. I think it's not about short circuit, but about heating itself inside the camera. But from another point of view - sensor is far from sd-card, and it has heat-proof management, camera just shut down itself that's all. But let's this never happens

Me too, in most cases I don't need to shoot long, but rarely I do, and know that card is heating. So far it's no problem, just reminder to keep an eye, "if anything" Yes, that is correct. If you set up bluetooth hack before run bitrate hack - it's works this way, yes, you get 140 mbps without editing the nx-ba.gdb file

Thanks for info here. How about long shoots like 30 min (or more with record limit off) - is sd-card heating more? I'd say that all my cards little bit hot after long run, need to keep an eye on this maybe, or maybe not? Yes I agree, it tends to be improving as bitrate increses, but not so much. I did the little test of DIS without OIS, maybe I nees to redid this with OIS on? I have a 16-50 OIS S lense, do you think DIS and OIS works better together?

Thanks man! As for you question I'd say your Lexar can do 150-160 mbps video in NX1 tops, which is pretty great, it's like two times better that original NX1 "pro" option which is 80 mbps. But I can be totally wrong in my prediction, don't let me stop you, just try and test it by yourself. Please, don't hesitate to update any info on this matter, because for now an sd-cards for decent 4K shooting are some sort of bottleneck in case of NX1 bitrate hack If you going to try bitrate hack - here is best option, just read the readme file and do as it said, it's pretty much straightforward

@Soul-Brother thanks! Can you provide 4kb tests also? For science @sandro here is the test footages, 15 sec each, just a roundtrip with camera and shaking 3 times with incresing intense. One is 50 mbps HQ 4K, other is 190 mbps

Oh, good question. It's need to be tested, I don't remember when I use DIS last time. Maybe there is progress here... so, how exactly we can provide reliable test fo DIS?

Test count 5, I dont think this test needs more Test Size 100, since bitrate hack record usually drops at 180-200 Mb in my experience

As for me I think that NX1 tends to forcing the operator to use more or less exposure in some scenes, which actually doesn't need to - I think in most cases MBL manipulations do the trick. I'm even considering to find a way to script MBL into one of the conrollers on camera, in order to set up MBL on the fly if this even possible, but it's only hopes for me yet )) And I did not see any criminal oversharpeness in foliage in your footage, so, who knows, only you That is very true for me also

@kidzrevil fine footage, there is no dropped frames I think, just hand-record without DIS and this is cool. I liked your grading not maybe in terms of aestetic but in technical way that shows pretty much wide possibilities of camera+bitrate. Without hack it can't go so well after hand-shooting it must be far more jerky and "strobed" as it seems now I think

Please do, it might be very helpful for others

At first I'm actually was going to disagree with you, but then I'm remembered that I did exactly the same at one moment - disable record limit, and only after this I did try new bitrate parameters, increasing from 160 to 190 on my Sandisk. So... right now I run DEV mode and enable record limit back. Testing video... Records two files without errorrs, three files more than 1 minute lenght, - it is all ok with 190 mbps bitrate hack. So, it might worth trying to disable record limit in DEV menu Hi Soul-Brother. Well, first of all, I am talking only about usage some of sd-cards while running bitrate hack for NX1 cameras. It increases bitrate up to enormous 320 mbps, which can't be operate at any of sd-cards, no matter UHS-II or not. And according to that table in Blackmagics "sd card speed" document, which I've posted earlier, the brand new Lexar one is not good for record with such a high bitrates. Still, it can be useful for usual "PRO" bitrate in NX1, but I have zero experience on this, sorry.

Please report this results here, since it might be a common problem with cards/hacks and cameras.

Maybe it's not about exact 160, maybe it's, say, 158 - OK, and 159 - is not OK, so it's not codec issue, but more speed limits Thanks, man, I will try this, but brand new Kingston that failed me is already moneyback'ed It is very true, I've tested enough to say all of my cards been formatted in NX1 and in PC, and in PC they all been formatted in all of the block sizes. It's actually a huge test yep. I can say that you can easily format sd-card in PC, just use "standart block size" option, I think NX1 do the same, but slill, better do format in NX1

Specs says otherwise - UHS-II, but still, you can be right

Well, IDK yet dependences between real record algorithms in NX1 and any of tested parameters, maybe I'm totally wrong, but for me it seems that random 4kb write tests shows far more consistent results to actual allowed bitrate, than vendors "write speed ABC Mb\s". It is just theory proof test, I might be totally wrong on that. But again, vendors write speeds is total BS for NX1 4K record imho. Maybe it is a NX1 thing, yes, but Blackmagic's document actually provides some proof of that not only NX1 have this troubles.

Well, new Beta 12.5 doesn't have this problem with HEVC at all. It just doesn't support h.265 now

Let's just test something, if we can? There is huge BS going around sd-cards performance, and I just return one to distributor - Kingston claims that one of their products can do 80 Mb/s write speed, so I bought it. Naturally, I can't record not only any DCI4K or UHD, but all 1080p in pro bitrate, it can do 1080p30 in HQ mode at best. But any test in Windows or MacOS tells me that card works perfectly with great 80 Mb/s which can provide smooth record for video bitrate up to 640 (!!!) mbps. It's just a total crap actually, this record speed in sd-card specs doesn't say much for us. It seems that vendors of sd-cards can't test their products in real applications such a cameras or so, or just doesn't want to share this info with us. It seems only for this matter exists such documents, like Blackmagic chart of compatible sd-cards (page 20 of this document) where we can find table of tested sd-cards, which can be used in their 4K recorder. It records in ProRes, so we can easily establish actual speeds of this cards, cause ProRes bitrates are well known. So, for any use of bitrate hack I'd highly recomended use one of this cards from Blackmagic document: Transcend SDXC UHS II 180MB/s Lexar SDXC UHS II 2000x 300MB/s Delkin Devices SD 633X SDHC UHS I 80MB/s Delkin Devices SD 633X SDXC UHS I 80MB/s Sandisk Extreme Pro UHS I 95MB/s Lexar SDXC UHS II 1000x 150MB/s Lexar SDXC UHS II 1000x 150MB/s Also this ones might be not so good option: Sandisk Extreme Pro UHS II 280MB/s Lexar SD 633x SDHC UHS I 95MB/s Strange, you see, that in real life test brand new Sandisk Extreme Pro UHS II with astounishing 280MB/s is totally loses it's predecessor Sandisk Extreme Pro UHS I with only 95MB/s So, I've tested main parameter that I could imagine to connect with actual situation. You can do this also, it's very easy. Download any app that can measure IOPS by 4kb blocks random write, I use this one: CrystalDiskMark. Here is my findings: Kingston SDXC Class 10 UHS-I U3 90/80 MB/s Sequential Read (T= 1) : 93.726 MB/s Sequential Write (T= 1) : 22.861 MB/s Random Write 4KiB (Q= 1,T= 1) : 0.111 MB/s [ 27.1 IOPS] Random Write 4KiB (Q= 32,T= 1) : 0.119 MB/s [ 29.1 IOPS] Transcend SDXC Class 10 UHS-I U3, 95/60 MB/s Sequential Read (T= 1) : 94.381 MB/s Sequential Write (T= 1) : 67.108 MB/s Random Write 4KiB (Q= 1,T= 1) : 1.054 MB/s [ 257.3 IOPS] Random Write 4KiB (Q= 32,T= 1) : 1.292 MB/s [ 315.4 IOPS] SANDISK Extreme Pro SDXC Class 10 UHS-I U1 95/90 MB/s Sequential Read (T= 1) : 94.381 MB/s Sequential Write (T= 1) : 93.421 MB/s Random Write 4KiB (Q= 1,T= 1) : 1.224 MB/s [ 280.9 IOPS] Random Write 4KiB (Q= 32,T= 1) : 1.398 MB/s [ 331.2 IOPS] V3 and V2 are the same in case of NX1 In V2 which I use all the time there is option for changing bitrate for 5 slots, but I can't use any HQ mods. for some reason it works on my NX1 only with PRO bitrates, which is sad but I'm gonna wait for it

Yes, I need to add that delay appears not all the time, only a few times, but noticable, like 3-4 seconds between I push the button and camera stops. Also that applies to turn-off process, it's longer now with the bitrate hack

At first my Sandisk Extreme Pro can do only 160 mbps, and I shoot all my stuff in 150 mbps for little headroom, it works vwry well without crashing or lags. After some time situation changes, now I run with 180 mbps, but tested 190 and it goes well on the same sd-card. Maybe you can try this too, but maybe it just some deviation, not much to say How is it appears? Little delayed reaction between you press the record button and it actually stops the video? If so - count me in, same experience here And this one is the "motion cadence" also

It's just about sd-cards mostly. This bitrate is perfectly utilize nowadays sd-cards speed and works without errors or lags. I've tested now like ten 64 GB sd-cards wide variety of prices and vendors, they pretty much the same when it comes to speed of NX1 sd-card driver (? idk how it calls) and top bitrate which I can use is no more 190 for now, but I'm going to test further Sandisk Extreme Pro UHS I 95MB/s 64GB is fastest one that I've tested. I can use it with 190 mbps easily, it works fine, but to be sure I just fly with 180 mbps. Strange thing, when I did first testing of bitrate hack, all of my cards goes with "card is too slow" message when I tried to record video, and first bitrate, that any card allowed me is 140-160 mbps. But IDK how, when after some shooting with 140-160 mbps bitrate I've put the same sd-card but with 180-190 mbps option, run record, it goes with same error after 3-4 secons video... and after that allows you to go with 180-190. After that it goes smooth and pretty fine with 190. Can't say the reason, I don't know it, but I can recommend this way to any testers - just go first with 200 mbps, wait for "card is too slow" message, go second time, go third, if message still appears, turn off the camera, set 190 mbps, run 1-2 record with error message, and maybe now you can do 190. If not - go to settings and decrease bitrate one step more, repeat previous step "with 1-2 errors".. Second sweet spot of bitrates - is size of sd-card. Actually, this one works with "record limit" hack. This hack allows you to record more than 30 min video. For UHD@30 it's 76 minutes with sound and not limited without audio (one user on dpreview said that but I didn't test this yet). So, now you have 76 minutes limit, ok. And now we can set 120 mbps bitrate for UHD@30 and this is a truly sweet spot for this two hacks - now you can record like 74-75 minutes 120 mbps UHD@30 video without any fear to get the error message or crash without message at all (NX1 did this with 4K video that exceed time limit 76 min). It is very cool! One page before I've posted the test results of my bitrate hack test. There is 4K video file that I create from two 4K footages one 80 and other 150 mbps. They just heavily destructed (only luma curve + brightness + contrast) for more visual reference, left side - 150 mbps source, right side - 80. Both recorded in UHD@30 with identical setting and environment, only difference is the bitrate. Scale to 300% for better proof and to eliminate influence of export bitrate. Can you see the difference between them? For me it's obvious, and this test - is how I actually open my eyes to see all the power of High Efficiency Video Codec implemented in H265. Highly recommend to learn about it, worth knowing. If by "motion cadence" cinematorgaphers means some sort of undescribed mixture of shutter speed, framerate and limits of modern encoding process, but from "artistic point of view", then I can tell, bitrate hack definitely improve this, yep. I'd say HEVC theoretically wins in this field when it comes to any lossy codec, and it improves greatily from bitrate hack. Overall quality of picture is improved, each frame is far more stable in terms of color, light - all this parameters, that goes to codec and being encoded, and this goes more smoothly, this also applies to the serial set of frames.

+1 here, it might be useful for testing

Yes, I agree, removing NR could help to reduce MB, but it's also very depends. To be clear HEVC actually doesn't use MB at all, this codec have new feature - CTU, and this one goes far more then usual MB. Maybe, just maybe, NR is better for HEVC. because of basic structure and size of single block, which can be just 4x4 px and grouped in "trees" - variable larger block structures up to 64x64 px, so It is actually might increase visual noise when it "smoothing" single noise pixel by 4x4 single prediction or transform block when NR is removed. Actually this Kino Seed's nx500 2,5K footage without NR seems to be slightly affected by this, but I can't tell that I'm 100% shure of this, it's just might be To be clear I don't know this for real and I know nothing actually about videocoding in terms of programming but I'm trying to use my very little knowledge to explain this things that my eyes seeng at least for myself )) Don't take my word on it, do test yourself or use mine - actually there is clear difference between block structure between 80 and 150 mbps footages