VIOFO A119S Modified Firmware MODs Archive

[Caerus]

same color background?
After the MOD 7G testing pick a MOD and this will be the Logo.

Yeah leaving it as is, is fine.

 

[dacbuc]

A few snapshots:
1. FWA119S_MOD-5K_v1.01+25Mbps+IQ-K
2. FWA119S_MOD-7K_v1.01+27Mbps+IQ-K
3. FWA119S_Req02_Custom-7G_v1.01+27.5Mbps+Sharp36+Edge178+Sat10



I'd say the 3rd one looks best, with good sharpness, almost no 'halo', just a bit too much saturation:
https://drive.google.com/open?id=1pWfHZC42oqSHpfnv4OPmKg2rOWU_7oS3

Would you make another custom mod: v1.01+25Mbps+Sharp36+Edge178+Sat5?
Thank you!

 

[gse]

I'd say the 3rd one looks best, with good sharpness, almost no 'halo', just a bit too much saturation:

Increase all the pictures, look at the big tree, where the black BMW and grass next to the white minibus, red brickwork at a distance, etc. No. 5 in the "porridge"

 

[BCHobbyist]

A few snapshots:
1. FWA119S_MOD-5K_v1.01+25Mbps+IQ-K
2. FWA119S_MOD-7K_v1.01+27Mbps+IQ-K
3. FWA119S_Req02_Custom-7G_v1.01+27.5Mbps+Sharp36+Edge178+Sat10

View attachment 41140View attachment 41141View attachment 41142

I'd say the 3rd one looks best, with good sharpness, almost no 'halo', just a bit too much saturation:
https://drive.google.com/open?id=1pWfHZC42oqSHpfnv4OPmKg2rOWU_7oS3

Would you make another custom mod: v1.01+25Mbps+Sharp36+Edge178+Sat5?
Thank you!

The Custom 7G @27.5Mbps has improved v1.01, is very good, crisp, but not too vivid. Need more footage before making adjustements, Saturation seems fine for old v1.01 which required huge boost from default.
On my way out on a job, try new Stable MOD-7E.

 

[ljw2k]

FWA119S_MOD-7K_v1.01+27Mbps+IQ 1080p@60fps With CPL

https://www.dropbox.com/s/ekf8gi8d3ah4as8/2018_0828_135954_064.MP4?dl=0

 

[BCHobbyist]

FWA119S_MOD-7E_v2.61+27Mbps+IQ-E.bin = phase-0 Sharpness, not as good as 7K, cloudy smoke haze 1080P@30fps CPL on, EV 0, WDR off
probably needs boost to saturation otherwise not bad. The K7-2.61 with phase-4 sharpness control should always have better IQ compared to 7D&E phase-0. Difference using phase-0, it has higher object outline definition, but also right at 'halo' mark. Once we all agree K7 is the way to proceed suggest increasing to phase-5 (1 point increase across all 624 values) to boost the crispness.

https://drive.google.com/open?id=1kal18uNsdGIlIR1uE1qtlm_UQeDhpnU2

 

[dacbuc]

Too much 'edge halo' and ‘false’ sharpness increase the ‘uneveness’ of color patches (e.g. cars) rising the need for more saturation. Correctly adjusted sharpness (like in 7k) makes colors appear more ‘solid’.

 

[c4rc4m]

Some footage - Viofo 1.2 firmware vs Mod 7:

https://dashcamtalk.com/forum/threa...l-firmware-vs-bc-mod-7-stable-firmware.35908/

There's are halos at night, but I'm not sure they're any worse than in the official Firmware. I think a lot of the halos are simply a product of a low light sensor and bright light source.

 

[BCHobbyist]

A119S Major Sharpness Control Discovery
As many know @gse and I plus a few others use the old v1.01 official firmware for its non-over-sharpened and night neutral balanced Image Quality. Unknown to all I've been digging deeper into v1.01 to improve it for my own use and possibly made a major discovery. In the next few hours will conclude my Alpha testing to confirm these new Sharpness controls. To demonstrate the IQ improvement v1.01 has over new v2.61 the next v1.01 MOD-7H will leave Sharpness phase-0 to 5 at default removing all adjustments. My discovery requires one road test and if successful confirms there is no need to adjust the 624 data values in either v1.01 or v2.61 or future releases. The key to this discovery is the old v1.01 with default IQ before the outrageous higher adjustments began with v1.2 to v2.61 current. After comparing night footage recorded yesterday it became obvious why the MODs i've made for v2.61 do not work the same on v1.01, each adjustment instead dulled and softened which was the reverse of what was expected. It was a shocking discovery, something I should have noticed 20 days ago but was focused on improving v2.61 where all adjustments had intended effects. This is further proof why phase1-5 Sharpness became so hard to balance even when taking them down 20 points, to be specific, reducing all 624 Sharpness values down 1-10 points had zero effect in removing 'halo'. My Alpha testing confirmed it.
Conclusion: Sharpness table is balanced because they are almost identical between v1.01 & v2.61, therefore another Sharpness control must be responsible and detectable by comparing 'other' tables between v1.01 & v2.61 and all other DVR models, all have similar initial 624 values. I've been adjusting the wrong Sharpness tables for the effects desired. (said the man with few words, evidence is needed be back shortly)

Deduction Confirmed
Wanted to save time in Alpha testing so the 'halo' effect is at maximum limit before getting ugly, showing only on nearby thin objects like wires not distant. Also Saturation is at maximum limit before affecting detail. These two limits are demonstrated in the footage below. Want feedback from your own footage to decide on how much to lower each image control.

MOD-7H-Alpha_v1.01+27.5Mbps+CrispIQ-H_014.MP4 >> sample footage
https://drive.google.com/open?id=1GZxBdHVLPuFUKkJBwLr_NyqvYaZjYpCQ

MOD-5H-Alpha_v1.01+25Mbps+CrispIQ-H_025.MP4 >> sample footage
https://drive.google.com/open?id=1I2zOP0rZO8SuJf9aHgFDCVZZz8URnhrq

Currently no ChangeLog for any Alpha is written and considered a work in progress. Try the Bit Rate of preference and record day & night footage for comparison.
FWA119S_MOD-7H-Alpha_v1.01+27.5Mbps+CrispIQ-H.bin
FWA119S_MOD-5H-Alpha_v1.01+25Mbps+CrispIQ-H.bin
DOWNLOAD https://drive.google.com/open?id=1kg3Ak0e4i7dY2CyT4Lk-CMfxhqubD4ti

@lukasz.pl @Caerus @c4rc4m @dacbuc @borik @Dadi_T @gse @ljw2k @Koganiz

 

[BCHobbyist]

Some footage - Viofo 1.2 firmware vs Mod 7:

https://dashcamtalk.com/forum/threa...l-firmware-vs-bc-mod-7-stable-firmware.35908/

There's are halos at night, but I'm not sure they're any worse than in the official Firmware. I think a lot of the halos are simply a product of a low light sensor and bright light source.

It's always more difficult to assess IQ values from night footage due to increased motion blurring but when the settings are correct it will be crisp and POP, which I just discovered with the help of old firmware. The only 'halo' effect to confirm over-sharpening is on thin wires and object outlines, not light sources. Ignore all bright lights and look at detail around edges to see the 'halo'. when its extreme the outline of the object seems to double in thickness, that is called over-sharpening and as @dacbuc & @gse and others pointed out must be avoided for maximum image quality. We are all learning about image control and we all have preferred levels of Sharpness, my goal is to find the limit before 'halo'.

 

[Caerus]

Beware massive file sizes! Also, tons of construction on campus...

These are daytime full runs at about 10 minutes each around the UM campus. They were done on different days, but close to the same time of day, and weather was comparable, save for a few more clouds here and there. Night ones will follow once I do one more run with the other FW.

Custom 7G 2.61: 1080p@30fps + CPL + EV 0 + WDR Off + Frequency 60Hz
https://drive.google.com/open?id=1nX2nkAkZERCDitDV1RyvVDiLALq-vXwP

Custom 7G 1.01: Same settings
https://drive.google.com/open?id=1ID-tF0CgN6ZEF-j8V3zHcKFz-jQ4imRZ

These should give a good representation of the daytime video.

From my view I think the colours were a more "accurate" representation (loosely referenced) on 2.61, 1.01 was warmer/slightly orange? Both pushed the limit. In terms of picture quality though, despite the picture being much warmer, I thought 1.01 was clearly the better look overall. Clarity is better, edges are smoother, and 2.61 showed hints of graininess, while none of that was present on 1.01.

 

[jokiin]

Beware massive file sizes! .

would be better to use the 1 minute loop setting while collecting sample videos, easier to share

 

[Caerus]

would be better to use the 1 minute loop setting while collecting sample videos, easier to share

That's usually what I do, or I clip them to be around 1 minute.

A UM campus detour was requested. So I figured a full run through the route would be appropriate in this case.

Other videos will go back to the normal format afterward.

 

[BlackCloud]

Ok!!!
Now I have a couple of interesting bits of information that may decide on who may want what particular mod.

My A119S.v.2 was out of action the past few days due to a complication in diagnosing a left edge focus issue. In that time I had a new ribbon cable sent and a new A119S.v.2 was ordered and received. The ribbon cable arrived and installed successfully. I also know that the ribbon cable does not and has not solved the focus issue it did prevent the camera from working at all till the new ribbon arrived. I now have 2 A119S.v.2 and my original v1 A119. The v1 A119 is still at the rear of my vehicle.

However I now have 2 working A119S.v.2 These cameras are not the same. The Newest one does NOT have the same lens as the now replaced ribbon one. Rumor has it, that sometime in 2018 the lens in the A119S.v.2 was changed to the same lens in the A129. I believe there was no change in the image sensor just the physical lens. It has a slightly wider field of view in the new one and there is some vignetting (as reported in another thread) with the original CPL in use.

New camera A119S.v.2 Clip.

The Top right corner is the dot matrix on the windshield. This was not visible with my other A119S.v.2 mounted in the same gps mount. There seems to more of my "A-pillar" in view than there was before. You might even be able to see my vehicle registration sticker in the bottom left of the screen that was not there before either.

I think I have another clip already posted from the same highway "flyover" at same time of day. with the previous camera.

 

[BCHobbyist]

Beware massive file sizes! Also, tons of construction on campus...

These are daytime full runs at about 10 minutes each around the UM campus. They were done on different days, but close to the same time of day, and weather was comparable, save for a few more clouds here and there. Night ones will follow once I do one more run with the other FW.

Custom 7G 2.61: 1080p@30fps + CPL + EV 0 + WDR Off + Frequency 60Hz
https://drive.google.com/open?id=1nX2nkAkZERCDitDV1RyvVDiLALq-vXwP

Custom 7G 1.01: Same settings
https://drive.google.com/open?id=1ID-tF0CgN6ZEF-j8V3zHcKFz-jQ4imRZ

These should give a good representation of the daytime video.

From my view I think the colours were a more "accurate" representation (loosely referenced) on 2.61, 1.01 was warmer/slightly orange? Both pushed the limit. In terms of picture quality though, despite the picture being much warmer, I thought 1.01 was clearly the better look overall. Clarity is better, edges are smoother, and 2.61 showed hints of graininess, while none of that was present on 1.01.

@Caerus, I owe you big time for your time and effort to record exactly what I requested, Excellent, A+ !!
This video comparison reveals many differences between v1.01 & V2.61 no one here has ever seen.
- proves A119S can record 27.5Mbps and save as single 10min clip without video corruption (was doubtful).
- proves without a doubt v1.01 has different contrast and HUE color than v2.61
- proves the Engineering Building on left is still the ugliest building on campus.
- proves Engineering Students still walk around with beer bottles and don't give a damn.
- proves strange things still happen in front of Art Building as guy walks his pet rat on a string under small trees.
- proves people in that City still don't know how to drive.

prefer v1.01
I appreciate the effort in recording that footage, a trip down memory lane decades ago & 3,000 miles away, thank you @Caerus

 

[BlackCloud]

This picture shows the two A119S.v.2 that I now have side by side the newest one is the one on the right!

 

[BCHobbyist]

This picture shows the two A119S.v.2 that I now have side by side the newest one is the one on the right!

View attachment 41152

Now you can mount twin DVRs and record true side-by-side Firmware Comparisons or keep one as backup. The new A119S with F1.6 A129 Lens and metal base is a keeper.

 

[BCHobbyist]

A suggestion to change the colored DICE Logo to represent everyone involved, and/or the coding, and/or hardware, and/or text "Modified Firmware".
ideas, comments ?

 

[BlackCloud]

A suggestion to change the colored DICE Logo to represent everyone involved, and/or the coding, and/or hardware, and/or text "Modified Firmware".
ideas, comments ?

I would support the idea of "modified firmware" text somehow in a splash screen.

 

[Koganiz]

Probably too large, something like DT would fit.