WorkHorse Boson
highest possible careful tweaked settings for very high image quality with enough stability for short takes up to 5 - 8 minutes.
Improved for best low light noise and textures in all modes, perfect for high iso, and excelent image texture in good light.
Arri Alexa image quality inside the GH2, (just missing the Alexa dynamic range of course)
default version:
*SH and H slots are better for good light in low iso and for more detailed scenes in good light
*SH slot is higher quality and more stable with manual focus legacy lenses
*H slot more stable with native lenses with electronic contacts
*H slot also better for use with image stabilization
*hbr is better in good light than in low light
*default version is more stable in SH and H 720p and in HBR than the boson version
*24p 24H and 24L are the same of boson
*great quality in all modes
boson version:
*SH slot improved for low light textures and noise
*H slot better for good light in low iso, better for more complex details scenes in good light
*H slot is more stable for native lenses
*HBR does the low light / good light balance automaticaly for better noise in low light or better detail in good light
*24p has same noise quality in both good light or low light for constant image texture this keeps same random noise in all light conditions, good for intercut with SH
*a little less stable in 720p and HBR than the default version
*best quality in all modes
seta = workhorse default
setb = workhorse boson
setc = workhorse default lower datarate in SH and H for more stability in 720p and hbr
setd = workhorse boson lower datarate in SH and H for more stability in 720p and hbr
sete = same as seta, but the 24H and 24L are optimized to deliver sharpest image for steady complex detail scenes, it does the automatic adaptation in 24p for low light noise or good light steady complex detail scenes
setf = same as setc, but the 24H and 24L are optimized to deliver sharpest image for steady complex detail scenes, it does the automatic adaptation in 24p for low light noise or good light steady complex detail scenes
all tests performed in NTSC with the Sandisk 32GB SDHC Class 10 U1 45Mbps card.
transcode the original mts files to Apple ProRes 422 full range using the 5DtoRGB free software before editing to avoid the diagonal pattern rain problem. Try the ProRes HQ, normal and LT versions to perceive the best balance between quality/file size.
as of september 05 2014 all versions are modified flowmotion 2.02 with extensive careful tweaking
please donate to www.personal-view.com
Would you show few tests of sete and setf in 24H/24L??
I think the versions e and f do not make sense for cinema production. e and f use the panasonic stock gop table for 24p and this makes the I frames to change the size: when there is low detail with noise the I frames are small and B frames increase in size, in the death charts which are full of detail the I frames are very big and the B frames very small, so I perceived that I can see more details in the death chart. But I can see this with the eyes very close to the screen as a pixel peeper seeking for perfection. before answer your question I was thinking about if e and f would be useful someway, but I perceived that I does not deserves changing firmware. If the gop table was tweaked to keep the I frames very big all the time so yes it would be more useful for a very sharp image, but I did not tweaked the gop table this way and the behavior of I frames size is automatic in the e and f, I will not use them.
for cinema production it is better to have the same random noise no matter the light or details of the scene, so the other versions deserves the use, they are careful tweaked for the purposes described in the first post.
at this moment I do not have enough time to tweak the e and f to keep the I frames very high all the time. this would make them useful for good light use without noise, with the camera quiet, to get a very sharp and detailed image for pixel peepers.
if you or someone else would like to tweak the e and f versions this is the way to do:
the versions e and f are an attempt to keep the I frames very high and the B frames very low in 1080p24, this makes the image more sharp and detailed for steady shoots. this is done in the gop tables. the gop tables are the stock in these patches, I did not find a way to avoid the automatic ratio variation yet.
In my math calculations, eight I frames at around 1250000 size each and sixteen B frames at around 300000 size each will deliver a great sharp and detailed image. the compression in the I frames will be only 2.5 and in B frames will be 10 which is in a excelent low value for a B frame when the I frame is big.
the datarate will be around 128000000, so to tweak these versions there is the need to change the 24H datarate to 128000000 and the 1080p24 top limit to 128000.
the maximum limit value for an I frame is in the sweet spot with a 2.5 compression ratio which allows this size, so now it is a matter of learn how to manipulate the gop tables, but I do not know how to do this yet, needs tests.
the correct numbers in gop tables will allow the I frames to keep in 1250000 size all the time and the B frames to keep at 300000 size all the time. this will deliver the sharpest and most detailed image for steady shoots.
after you find the best settings for each hack patch, all the secrets are in the GOP TABLES, they manipulate the size of the I, B, P frames and how much they will vary in size depending on image details, and consequently the ration between them. knowing how to do this makes possible to extract the best form the datarate to get what you want from the image. mainly you can have priority to sharpness and detail or to fine random noise, or an automaticaly adjust between both. for cinema a constant fine random noise is better, for advertising a sharp detailed image is better.
I think that the scalling matrices and deblocking matrices are important, but after you have a good matrix then the gop tables is what does the job.
these last patches I uploaded are very careful tweaked in all settings, they are a perfect base to play with gop tables and understand what they do. a good thing to do would be to map the gop tables, starting with the panasonic stock values and increasing or decreasing each number one step each time. if we map the gop tables, we will can do whatever we want with the image.
at this moment I cannot do this because I am full of other tasks, very busy...
these are the sete and setf with higher datarate for 24H 128000000, the 24L is now 96000000
the gop table for 24p is the panasonic stock: "1080p24 GOP Table=1, 3, 0, 0, 8, 0" this gop table makes the I frames very high when there is lots of details in the image, but when there is no details the I frames decrease in size due to the variable bit rate.
to implement the idea of keep the I frames always very high (around 1.200.000) and the B frames always very low (around 300.000) I think it will be done tweaking the gop table above.
I do not know if it is possible to turn the codec into constant bitrate, but at least to keep the I frames as big as possible to record as much detail as it can, this will make the steady low iso shoots more sharp and detailed due to less compression in the I frames.
I have good news, I am using the weekend to try to map and understand the GOP TABLES.
all the secret for image encoding is here, the magic happens here.
of course the magic in gop tables happens after finding the best for all other settings, which is already done.
I have a better understanding about the gop tables now, far from dominated, but good notions.
advice for PAL shooters: boson is not tweaked yet for work in PAL, I am doing attempts on it.
workhorse boson, versions, 07 september 2014, works in NTSC or PAL
highest possible careful settings for very high image quality with enough stability for short takes up to 5 - 8 minutes.
improved to deliver the best textures and fine random noise in all modes
boson:
*SH slot improved for best textures and fine random noise (this is what makes the boson very special)
*H slot has slightly less quality than SH but it is more stable
*HBR is improved for great textures and fine random noise (very special)
*24p is improved for best textures and fine random noise (just best cinema)
workhorse:
*slightly less quality in 720p SH slot, almost unoticeable, but it is more stable than the boson in 720p
*HBR has same boson settings but can be more stable also due to share some 720p settings
*same image quality of boson in 24p
from seta to sete = boson (5 versions)
the difference is: seta has the highest datarate for 720p and hbr, the 720p and hbr datarate decreases a little in each version, this is for you to find the best stable version for your camera+card+lens combo. try the seta first and change if needed.
from setf to setj = workhorse (5 versions)
the difference is: setf has the highest datarate for 720p and hbr, the 720p and hbr datarate decreases a little in each version, this is for you to find the best stable version for your camera+card+lens combo. try the setf first and change if needed.
tips: manual focus legacy lenses allows more stability. disable image stabilization and i-dynamic when using native electronic lenses for more stability or load a lower datarate version.
Tested with the Sandisk 32GB SDHC Class 10 U1 45Mbps card. Very fast cards recomended. Use TOP memory cards.
as of september 07 2014 all versions are modified flowmotion 2.02 with extensive careful tweaking
please donate to www.personal-view.com
Stunning settings, together with nebula-workh!
I'm doing some testing of Boson (seta), which I guess you'd call your A+ version, just using the Stray color chart with the stock 14-42 lens. All tested using 'Standard' -2,-2,0,-2; idynamic, etc. all off. The SD card used is the SanDisk 64Gb Extreme Pro 280MB/s. I'm in PAL land, and as I note that your most recent settings state both PAL and NTSC hopefully it'll work.
Started with (again, seta) 24p and 24H setting. Formatted the card, in camera, after loading the hack/patch. Shot just over 12 mins before I manually stopped, creating two spanned 4Gb files and another one of about 800mb; longer than your "short takes up to 5 - 8 minutes", but being a death chart test maybe not 'real life' usage.
StreamParser info of the 2nd MTS file - the middle one of the spanned file.
Stream: Mode: 1080/24p Clip Duration: 0:05:33.202 TS Clock: 27MHz TS Duration: 0:06:09.276 Average PCR Timing Interval: 0.093 seconds TS Max Bitrate: 102,400,000 Average Overall Bitrate: 93,136,685 Average Video Bitrate: 102,607,015 Average Audio Bitrate: 239,986 Max GOP Size: 1,742,976 Bytes (13,943,808 bits) Video Frame Count (All-I/P/B): 7,989 - 2,663 / 0 / 5,325 Video Frame Size (Min/Ave/Max): 226,560 / 534,890 / 1,112,064 I Frame Size (Min/Ave/Max): 794,112 / 947,764 / 1,112,064 P Frame Size (Min/Ave/Max): 0 / 0 / 0 B Frame Size (Min/Ave/Max): 226,560 / 328,514 / 406,080 *Video/Audio calculations based on TS Packets
Tried camera playback, didn't work. So, tried the camera off-camera on trick, still no luck. However, I forgot to try the 'record a small clip' trick, before reformatting and testing 24L!
Is the above of any use/help??
OK, now tried 24p 24L setting - all other testing as above (e.g. seta, death/colour chart, SD card, etc) Card reformatted (in cam) before test. I stopped recording at 14 mins before the camera stopped. Two spanned files produced, second one being about 3.8Gb.
StreamParser info of the 1st MTS file
Stream: Mode: 1080/24p Clip Duration: 0:07:22.852 TS Clock: 27MHz TS Duration: 0:08:03.354 Average PCR Timing Interval: 0.093 seconds TS Max Bitrate: 76,942,486 Average Overall Bitrate: 71,103,533 Average Video Bitrate: 77,001,937 Average Audio Bitrate: 240,015 Max GOP Size: 1,937,088 Bytes (15,496,704 bits) Video Frame Count (All-I/P/B): 10,618 - 3,539 / 1 / 7,077 Video Frame Size (Min/Ave/Max): 100,800 / 401,420 / 1,167,744 I Frame Size (Min/Ave/Max): 414,336 / 696,040 / 1,167,744 P Frame Size (Min/Ave/Max): 188,736 / 188,736 / 188,736 B Frame Size (Min/Ave/Max): 100,800 / 254,176 / 426,240
Tried camera playback, didn't work Tried the camera off-camera on trick, still no luck Tried 'record a small clip' trick, got "Cannot Record, File Number limit exceeded"!! After formatting the card, I took the camera outside and tried recording a few short clips (about 30s-1min each) to see if I'd get the same 'file number' error; it didn't reappear and in cam playback worked (without any tricks!).
@frullaccia thanks! hope it will be useful for you
@JayUKB thanks for the tests and measurements! the 24p 24H and 24L are expected to record more stable. The most interesting tests will be in the 720p SH and in the HBR because these modes are the best improvements in the Boson and are the modes more fragile to stability. this is the reason I created 5 versions lowering the datarate a little bit so people can find the best datarate for the memory card in these modes. As you own a super fast card it will be useful information to test the seta "A+ version" in SH and HBR with your card. Many thanks!
some tips: manuall focus lenses allows more stability. native lenses with electronic contacts can stop recording when using the i-Dynamic or the image stabilization in the SH and in HBR.
HBR 25p test - again using the death/colour chart. Card reformatted (in cam) before test. I stopped recording at about 12 mins. Two spanned files produced, second one being about 1.8Gb.
StreamParser info of the 1st MTS file
Stream: Mode: 1080/50i Clip Duration: 0:08:17.920 TS Clock: 27MHz TS Duration: 0:08:18.075 Average PCR Timing Interval: 0.093 seconds TS Max Bitrate: 102,400,000 Average Overall Bitrate: 68,979,804 Average Video Bitrate: 68,389,810 Average Audio Bitrate: 240,015 Max GOP Size: 1,190,016 Bytes (9,520,128 bits) Video Frame Count (All-I/P/B): 12,448 - 4,149 / 1 / 8,297 Video Frame Size (Min/Ave/Max): 45,504 / 341,940 / 853,824 I Frame Size (Min/Ave/Max): 336,576 / 820,142 / 853,824 P Frame Size (Min/Ave/Max): 115,008 / 115,008 / 115,008 B Frame Size (Min/Ave/Max): 45,504 / 102,879 / 344,064
Tried camera playback, worked first time...odd!
@JayUKB great to see you got spanning, this shows how important is a fast card!
the playback in camera is not expected to work for all recorded files, until now I got no problem in computer, all files I recorded can playback in the PC.
I will get some rest now, after intensive work.
Many thanks!
@JayUKB keep up the great tests! It is important information about the performance in fast card. Your reports are very useful. thanks.
The spanning and no stop recording are showing that the camera hardware can handle the heavy data even in HBR. this is amazing. the streamparser information shows is very useful to perceive the performance. this results on a death charts are proving a reliable performance for real world shoots.
I am curious to see about the 720p because it is the most difficult mode for the camera and card to handle. I just own the 45Mbps card so these tests with a fast card are very important information.
After some rest, I did some tests with the SH and H 720p in sun light foliage. The H mode works great for very detailed scenes in low iso while the SH is better for situations in low light and high iso.
The SH shares more data to the P frames so the noise is better, the H shares more data to the I frames so for complex details it is better. Good thing is that the Boson can do both things without change the firmware, just using SH and H for different situations.
I tested the sete also which has the lower datarate for 720p and the quality is very good for noise in SH also, good surprise. it keeps around 48Mbps while the seta keeps around 70Mbps
I did not get stop recording yet in the seta 70Mbps but I just did short takes until now.
I am curious to see if a fast card will hold long time recording with the seta in SH and H to get the highest quality in 720p without stop recording, but I just have the 45Mbps card...
Interesting information in the stramparser is the average size of I frames and B/P frames. when the I frames are high the codec is doing a good job for details, when the B/P frames are high the codec is doing a good job for noise and textures. when both I and P/B frames keeps more constant the image texture will be constant for noise/details, 24p works this way. If the settings I did works OK, the HBR will change the I B frames size for details or noise when needed, this first reports shows it is working for details in the death chart because I frames are high. So it will be interesting to see how it will perform in low light also, how will be the size of I and P/B frames.
Nebula-Boson (the Scalling Tables and Deblocking Tables from Nebula T8 are merged with the Boson)
I did no tests yet so I do not know if it will be stable...
I did careful tests with Nebula T8 matrix in the death charts and in low light.
Two positive feedback, one negative feedback (the negative feedback makes it a "no go"):
a little more sharp in the death charts.
in low light the noise is nore fine in dark areas, beautiful, makes the very dark areas more filmic.
there are some stuck pixels (bright pixels) sometimes pulsing in the image, more noticeable in dark areas, so this makes the Nebula matrices unfeasible. I noticed this in 24p. There is also a report about this in the topic discussing the Nebula.
It seems the flowmotion matrices are in the sweet spot, extract more from the sensor makes the stuck pixels shows up.
Boson / WorkHorse - Flowmotion and Nebula matrices versions - 08 september 2014:
careful settings for very high image quality with enough stability for short takes up to 5 - 8 minutes.
depending on the version and on the recording mode you can get:
*best textures and fine random noise
*best results for complex details
works in NTSC or PAL
boson:
*SH 720p slot: improved for best textures and fine random noise in low light or high iso
*H 720p slot: same datarate quality of SH, but improved for complex details and low iso
*HBR: can do both things, good textures and fine random noise or complex details
*24p: balanced between constant good noise and good resolution, sweet spot for cinema
workhorse:
*a little more stable in 720p than boson
*both SH and H slots have slightly less noise quality than boson, but are great for details
*the SH slot has higher datarate than H slot, H slot is an option for better stability
*HBR has same boson settings but can be more stable also due to share some 720p settings
*same image quality of boson in 24p
from seta to sete = boson (5 versions)
the difference is: seta has the highest datarate for 720p and hbr, the 720p and hbr datarate decreases a little in each version, this is for you to find the best stable version for your camera+card+lens combo. try the seta first and change if needed.
from setf to setj = workhorse (5 versions)
the difference is: setf has the highest datarate for 720p and hbr, the 720p and hbr datarate decreases a little in each version, this is for you to find the best stable version for your camera+card+lens combo. try the setf first and change if needed.
tips: manual focus legacy lenses allows more stability. disable image stabilization and i-dynamic when using native electronic lenses for more stability or load a lower datarate version.
Tested with the Sandisk 32GB SDHC Class 10 U1 45Mbps card. Very fast cards recomended. Use TOP memory card, it can improve stability.
these versions of september 08 2014 have the boson and workhorse with Flowmotion or Nebula Scalling Tables and Deblocking Tables
please donate to www.personal-view.com
I am tweaking the fallback tables in the nebula versions to make it better. this will make the frames more constant size along the frames's flux.
Boson and WorkHorse Patches for GH2 Hack
Boson:
WorkHorse:
From seta to sete of Boson (5 versions):
The difference is: seta has the highest datarate for 720p and hbr, the 720p and HBR datarate decreases a little in each version, this is for you to find the best stable version for your camera+card+lens combo. Try the seta first and change if needed.
From setf to setj in WorkHorse (5 versions):
The difference is: setf has the highest datarate for 720p and HBR, the 720p and hbr datarate decreases a little in each version, this is for you to find the best stable version for your camera+card+lens combo. Try the setf first and change if needed.
Memory Cards: Developed with the Sandisk 32GB SDHC Class 10 U1 45Mbps card. Very fast cards recomended. Use TOP memory card, it can improve stability for more continuous time recording, and for use image stabilization and i-Dynamic.
Tips: Manual focus legacy lenses allows more stability. Disable image stabilization and i-dynamic when using native electronic lenses for more stability or load a lower datarate version or use a fast memory card.
Matrix Tables: These versions of september 09 2014 have the Boson and WorkHorse with Flowmotion or Nebula Scalling Tables and Deblocking Tables. Fallback Tables tweaked to deliver more constant size in the frames along the frames's flux.
Please donate to www.personal-view.com
this is experimental, to be safe use the 09 september versions.
Boson - WorkHorse for Super Fast Cards:
This is an attempt to improve the image quality even more increasing the datarate
This can help the 720p SH in Boson to be better for complex details
This can help the 720p SH in WorkHorse to be better for textures and fine random noise
Try these patches in 80Mbps cards and above
Try to record a 4GB file in each mode, feedbacks welcome!
About the datarate increase:
BF and BN setf = one step above seta
BF and BN setg = two steps above seta
WF and WN setd = two steps above setf
WF and WN sete = one step above setf
BF = Boson Flowmotion
BN = Boson Nebula
WF = Workhorse Flowmotion
WN = Workhorse Nebula
This is a complement for the september 09 versions, there is no new settings, just datarate increase, experimental
720p (PAL/50) SH test...using the seta in "workhorse_boson_07_september_2014_ntsc_pal.zip", from a few days ago. Using the death/colour chart again, same Sandisk 64GB 280 card. The camera stopped after recording a 4GB file, so no spanning.
StreamParser info.
Stream: Mode: 720/50p Clip Duration: 0:06:53.100 TS Clock: 27MHz TS Duration: 0:06:53.242 Average PCR Timing Interval: 0.093 seconds TS Max Bitrate: 102,400,000 Average Overall Bitrate: 83,027,811 Average Video Bitrate: 82,114,013 Average Audio Bitrate: 240,007 Max GOP Size: 862,848 Bytes (6,902,784 bits) Video Frame Count (All-I/P/B): 20,655 - 6,885 / 13,769 / 0 Video Frame Size (Min/Ave/Max): 22,464 / 205,279 / 553,728 I Frame Size (Min/Ave/Max): 248,256 / 331,523 / 553,728 P Frame Size (Min/Ave/Max): 22,464 / 142,167 / 199,872 B Frame Size (Min/Ave/Max): 0 / 0 / 0
What's BF and BN?
thanks for the 720p SH test. it shows that the 720p SH in seta can be reliable with a fast card. streamparser is showing good results.
I edited the post above about what is BF, BN, WF, WN, please take a look. Would be great you do a try on these high datarate versions on your card, as it is one of the fastest card available today. If it works on your card it will be amazing because they improve image due to more datarate.
Many thanks for your useful reports!
It looks like you're new here. If you want to get involved, click one of these buttons!